Psi Chi Journal Volume 31.3 | Special Issue

Health Psychology in Applied Contexts

PSI CHI JOURNAL OF PSYCHOLOGICAL RESEARCH SPECIAL ISSUE 2026 | VOLUME 31, NUMBER 3

EDITOR

ROBERT R. WRIGHT, PHD

Brigham Young University-Idaho

Email: wrightro@byui.edu

ASSOCIATE EDITORS

TIFANI FLETCHER, PHD West Liberty University

STELLA LOPEZ, PhD University of Texas at San Antonio

TAMMY LOWERY ZACCHILLI, PhD Saint Leo University

ALBEE MENDOZA, PhD

Delaware State University

JULEE POOLE, PhD Purdue University Global

KIMBERLI R. H. TREADWELL, PhD University of Connecticut

EDITOR EMERITUS

DEBI BRANNAN, PhD Western Oregon University

MANAGING EDITOR BRADLEY CANNON

DESIGNER

JANET REISS

EDITORIAL ASSISTANT EMMA SULLIVAN

ADVISORY EDITORIAL BOARD

GLENA ANDREWS, PhD RAF Lakenheath USAF Medical Center

AZENETT A. GARZA CABALLERO, PhD Weber State University

MARTIN DOWNING, PhD Lehman College

HEATHER HAAS, PhD University of Montana Western

ALLEN H. KENISTON, PhD University of Wisconsin–Eau Claire

MARIANNE E. LLOYD, PhD Seton Hall University

DONELLE C. POSEY, PhD Washington State University

LISA ROSEN, PhD Texas Women's University

CHRISTINA SINISI, PhD Charleston Southern University

ABOUT PSI CHI

Psi Chi is the International Honor Society in Psychology, founded in 1929. Its mission: "recognizing and promoting excellence in the science and application of psychology." Membership is open to undergraduates, graduate students, faculty, and alumni making the study of psychology one of their major interests and who meet Psi Chi’s minimum qualifications. Psi Chi is a member of the Association of College Honor Societies (ACHS), and is an affiliate of the American Psychological Association (APA) and the Association for Psychological Science (APS). Psi Chi’s sister honor society is Psi Beta, the national honor society in psychology for community and junior colleges.

Psi Chi functions as a federation of chapters located at senior colleges and universities around the world. The Psi Chi Headquarters is located in Chattanooga, Tennessee. A Board of Directors, composed of psychology faculty who are Psi Chi members and who are elected by the chapters, guides the affairs of the Organization and sets policy with the approval of the chapters.

Psi Chi membership provides two major opportunities. The first of these is academic recognition to all inductees by the mere fact of membership. The second is the opportunity of each of the Society’s local chapters to nourish and stimulate the professional growth of all members through fellowship and activities designed to augment and enhance the regular curriculum. In addition, the Organization provides programs to help achieve these goals including conventions, research awards and grants competitions, and publication opportunities.

JOURNAL PURPOSE STATEMENT

The twofold purpose of the Psi Chi Journal of Psychological Research is to foster and reward the scholarly efforts of Psi Chi members, whether students or faculty, as well as to provide them with a valuable learning experience. The articles published in the Journal represent the work of undergraduates, graduate students, and faculty; the Journal is dedicated to increasing its scope and relevance by accepting and involving diverse people of varied racial, ethnic, gender identity, sexual orientation, religious, and social class backgrounds, among many others. To further support authors and enhance Journal visibility, articles are now available in PsycInfo, EBSCO, Clarivate Emerging Sources Citation Index, Crossref, and Google Scholar databases. In 2016, the Journal also became open access (i.e., free online to all readers and authors) to broaden the dissemination of research across the psychological science community.

JOURNAL INFORMATION

The Psi Chi Journal of Psychological Research (ISSN 2325­7342) is published quarterly in one volume per year by Psi Chi, Inc., The International Honor Society in Psychology. For more information, contact Psi Chi Headquarters, Publication and Subscriptions, 651 East 4th Street, Suite 600, Chattanooga, TN 37403, (423) 756­2044. https://www.psichi.org; psichijournal@psichi.org

Statements of fact or opinion are the responsibility of the authors alone and do not imply an opinion on the part of the officers or members of Psi Chi.

Advertisements that appear in Psi Chi Journal do not represent endorsement by Psi Chi of the advertiser or the product. Psi Chi neither endorses nor is responsible for the content of third­party promotions. Learn about advertising with Psi Chi at https://www.psichi.org/Advertise

COPYRIGHT

Permission must be obtained from Psi Chi to reprint or adapt a table or figure; to reprint quotations exceeding the limits of fair use from one source, and/or to reprint any portion of poetry, prose, or song lyrics. All persons wishing to utilize any of the above materials must write to the publisher to request nonexclusive world rights in all languages to use copyrighted material in the present article and in future print and nonprint editions. All persons wishing to utilize any of the above materials are responsible for obtaining proper permission from copyright owners and are liable for any and all licensing fees required. All persons wishing to utilize any of the above materials must include copies of all permissions and credit lines with the article submission.

199 Introduction to the Special Issue: Health Psychology in Applied Contexts

Robert R. Wright

Department of Psychology, Brigham Young University–Idaho

200 Marijuana Use by College Students: Relationships With Personality Traits and Protective Factors of Resilience

Laney S. Sims and Shelia M. Kennison*

Department of Psychology, Oklahoma State University

208 Understanding Medication Nonadherence as Risk-Taking: Insights From College Students With Chronic Conditions

Annah L. Boone and Shelia M. Kennison* Department of Psychology, Oklahoma State University

215 Biogenetic Endorsements of Depression: The Enduring Influence of the Chemical Imbalance Theory

Saoirse E. Ward and James W. Diller* Department of Psychological Science, Eastern Connecticut State University

223 The Recollections of Childhood Food Parenting and Adult Mindful Eating

Abigail Brighton, Adelyn Sherrard*, and Cin Cin Tan** Department of Psychology, The University of Toledo

230 Hikikomori and Internet Addiction in U.S. College Students

Mai P. N. Tran1 and Cameron S. Kay*1,2

1Psychology Department, Union College 2Environmental Social Sciences Department, Stanford Universityn

237 Exploring Factors That Support Sense of Belonging in Undergraduates at a Large University in the United States

Kylie S. Sambirsky, Kit E. Ganzle, Sienna M. Russell, Ian J. McGillicuddy, Mandira T. Gowda, Yohana A. Markos, Joseph Levy, Leah Teeters*, and Winnie Zhuang*

Renée Crown Wellness Institute, University of Colorado Boulder

246 Emotion Regulation Strategies and Sleep Quality as Predictors of Depression in College Students

Luke Stanton and Colleen Georges* Department of Psychology, Rutgers University

252 “Working Out” the Relationship Between Mental Health and Exercise: A Cross-Sectional Study of United States Adults

Robert A. Bourne1, Craig A. Warlick1,2, Parker Gardner2*, and Ashley C. T. Jones1*

1School of Psychology, University of Southern Mississippi

2Department of Psychological Sciences, Texas Tech University

263 The Impact of Tardive Dyskinesia on Quality of Life in Individuals With Severe Mental Illness: A Meta-Analysis

Madelynn D. Loring*1, Trinity Lumbard*1, Brian Drwecki**1, and Kaylyn McAnally Star**2 1Department of Psychology and Neuroscience, Regis University

2School of Rehabilitative and Health Science, Regis University

272 Exposing the Falseness of Instagram: The Impact of Instagram vs. Reality on Body Dissatisfaction and Body Appreciation

Isabella Cerase1, Mylene Feiler*2, Sean Dougherty*1, and Madeline Dougherty*1, 1Department of Psychology, Florida State University

2Department of Language Arts, International Studies Preparatory Academy

281 The Relationships of Perceived Burdensomeness, Loneliness, and Religious Strain With Suicidal Ideation in LGBTQ+ and Non-LGBTQ+ College Students

Skyler Woolley, Abby Allen, Grace Collier, Jada Nagel, Jacob Schultz, Sarah Loertscher, Daniel Hatch*, Kirsten L. Graham*, and Bryan L. Koenig* Department of Psychology, Southern Utah University

Introduction to the Special Issue: Health Psychology in Applied Contexts

Robert R. Wright

Department of Psychology, Brigham Young University–Idaho

Iam pleased to introduce a special issue of Psi Chi Journal of Psychological Research with the title: Health Psychology in Applied Contexts. In efforts to promote the journal and support our author contributors, we have combined 11 studies in this unique special issue to showcase some recent efforts in applying psychological concepts in varied settings to promote health and wellness. As the journal has become more widely recognized, we have recently received an abundance of high­quality research at all levels (i.e., undergraduate, graduate, faculty), demonstrating an increase in submission quantity and quality. Moreover, consistent across many of these articles was an application theme in various domains of biopsychosocial health, a core tenet of health psychology.

The reader will find herein a collection of studies representing a wide range of topics in the field of health psychology. For instance, marijuana use among college students is presented alongside concerns for medication nonadherence as risk­taking behavior. Moreover, prominent health concerns such as depression, addiction, suicide, body appreciation, loneliness, and belonging, as well as mental illness are investigated. And, in line with traditional health psychology, the major health behaviors of diet, sleep, and exercise along with sociocultural and familial influences are also represented in this collection. As such, we are pleased to offer this special issue to highlight these very applicable and important topics to promote and protect health and well­being.

Author Note

Robert R. Wright https://orcid.org/0000000241017840

Correspondence concerning this article or the special issue should be addressed to Robert R. Wright, Department of Psychology, Brigham Young University–Idaho, 210 West 4th South, Rexburg, ID 834602140. Telephone: 2084964085. Email: wrightro@byui.edu

Marijuana Use by College Students: Relationships With Personality Traits and Protective Factors of Resilience

Laney S. Sims and Shelia M. Kennison* Department of Psychology, Oklahoma State University

ABSTRACT. We examined marijuana use by college students and how use was related to personality traits and protective factors of resilience. The sample included 116 undergraduates who reported using marijuana multiple times in the past year. Participants completed questions about marijuana use, personality traits, and resilience in an online survey. Frequency of use was high with over half using multiple times per week. Personality traits and protective factors of resilience were related to motives and consequences of use, but not frequency of use. In a multiple regression in which negative consequences of use was predicted, the only significant predictor was cognitive protective factors of resilience. Higher resilience was related to fewer consequences. Marijuana use was high and related to negative consequences. Colleges should develop strategies for reducing marijuana use by students.

Keywords: marijuana, college students, personality traits, resilience, marijuana legalization

In recent years, marijuana use in the United States has increased, as changes in laws in some states now permit the use of marijuana for recreational or medical purposes (National Institute on Drug Abuse, 2023; National Institutes of Health, 2021). Research has documented that marijuana use has risen most sharply for those between 19 and 30 years of age, which is an age group that includes most college students. In 2018, Oklahoma legalized marijuana for medical purposes (Oklahoma Secretary of State, 2022). Prior to the change in legal status, marijuana use by youth in Oklahoma was ranked 43rd nationally, and use by adults in Oklahoma was ranked 39th (Texoma HIDTA, 2025). After 2018, Oklahoma ranked 3rd in the nation in marijuana use by youth and 20th by adults. To our knowledge, there is little to no research on marijuana use by college students in Oklahoma since its legalization. Our aim was to assess college students’ frequency of use, motives for use, and negative consequences of use and how these aspects of use related to personality traits and protective factors of resilience.

Prior research has documented that frequent marijuana use (also referred to as utilization) is associated with multiple negative consequences (Evins et al., 2012; Gunn et al., 2020; Jeffers et al., 2024). Frequent marijuana use has been linked to both physical and mental health problems. A recent study found that

frequent marijuana use was associated with increased risk of heart attack and stroke even after controlling for other cardiovascular risk factors (e.g., tobacco use and demographic variables; Jeffers et al., 2024). In adolescents, frequent marijuana use has been associated with an increased risk for developing schizophrenia (Evins et al., 2012). Other studies have found that more frequent marijuana use is associated with greater anxiety and depression symptoms (Denson & Earleywine, 2006; Hayatbakhsh et al., 2006; Scholes­Balog et al., 2013) At least one study has failed to observe the relationship (Green & Ritter, 2000). In a study by Duckworth et al. (2023), young adults responded to survey questions twice a day (i.e., morning and evening) for five extended periods, each over two weeks. The results showed that marijuana use was associated with having less engagement in school or work the day following use. Although some negative effects of marijuana may subside after abstinence, problems with risk­taking and decision­making may continue to be observed (Crean et al., 2011). One study reported that those who use marijuana while they are alone are the most at risk of experiencing negative consequences (Okey et al., 2022). The most serious negative consequences of frequent marijuana use occurs when one develops a dependency on the drug (Gunn et al., 2020) or develops an addiction, such as Cannabis Use Disorder (CUD; Hasin et al., 2013;

Schulenberg et al., 2018; Shi, 2014). In a study conducted in the United States, the results showed that approximately 3 in 10 people who use marijuana regularly meet the criteria for CUD (Hasin et al., 2015). In the most recent revision of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2013), substance use disorders are diagnosed using a checklist of 11 symptoms, including signs of tolerance, cravings, negative social and personal consequences, and withdrawal, which have been present for 12 months or more (Patel & Marwaha, 2022).

Prior research has documented numerous motives for marijuana use (Boys et al., 2001; Drazdowski et al., 2021; Shi, 2014). The most common motives for using marijuana include changing mood, improving physical effects such as help with sleep, improving social experiences, making an activity less boring, and managing the effects from other substances (Boys et al., 2001). Depressed individuals who use marijuana to modulate their mood are more likely to become dependent (Shi, 2014). In another study, researchers found that almost half of their sample reported using marijuana to help them sleep (Drazdowski et al., 2021). They also found that using marijuana as a sleep aid led to higher use rates and more problematic use. Although using marijuana for enjoyment is less likely to lead to problematic use (Naegele et al., 2022), using it to cope with anxiety and stress is associated with drug dependence and other negative outcomes (Lee et al., 2009; Naegele et al., 2022; Phillips et al., 2017; Schultz et al., 2019).

Understanding which students may be most at risk of frequent marijuana use and developing a dependency may be useful for colleges, as they may want to target those individuals with messaging that raises awareness of the dangers of frequent marijuana use, or about the availability of physical and mental health services on campus. Numerous brief interventions to reduce marijuana use exist and have been shown to be effective (Hone et al., 2024). Costs to institutions could be reduced if resources and programming were directed only to those students identified at the highest risk level.

Personality traits and other personal characteristics may be useful in identifying individuals at risk of marijuana abuse. Individuals higher in sensation seeking personality traits (e.g., thrill­and­adventure seeking, boredom susceptibility, experience seeking, and disinhibition) have been shown to use drugs more often than others (LaSpada et al., 2020; Zuckerman, 2007). Individuals high in sensationseeking have shown a higher preference for stimulants and hallucinogens (Galizio et al., 1983) and a greater likelihood of experiencing adverse consequences associated with marijuana use (Parnes et al., 2024). Research also has shown that marijuana use is related to Big Five personality traits (i.e., extraversion, agreeableness, conscientiousness,

neuroticism, and openness to experience; John et al., 2008). Those with lower levels of conscientiousness and agreeableness or higher levels of neuroticism are more likely to suffer from CUD (Dash et al., 2019). A recent study found that more frequent marijuana use was related to lower agreeableness, lower conscientiousness, lower extraversion, and higher levels of neuroticism/emotional instability (Jones et al., 2022).

College students most at risk for using substances and experiencing negative consequences of substance use (e.g., addiction, problems with daily functioning) may be those with the lowest levels of resilience. Resilience refers to the ability of an individual to function at a higher level than expected in the context of experiencing one or more obstacles (Garmezy et al., 1984; Masten et al., 2004). Prior research has shown that individuals higher in resilience are less likely to use substances (Myntti & Armstrong, 2022; Wingo et al., 2014) and less likely to experience the negative effects of stress (Myntti & Armstrong, 2022).

The Present Study

The aim of the present research was to investigate marijuana use by college students at a large public university in Oklahoma since the legalization of medical marijuana. In 2018, the use of marijuana for medical purposes became legal in Oklahoma for those 18 years of age and older through State Question 788 (Oklahoma Secretary of State, 2022). Under the law, the medical conditions qualifying for a medical marijuana card were not specified (Aston, 2024). By 2023, approximately 9 percent of the population in Oklahoma (over 368,000 people) had obtained medical marijuana cards (Marijuana Policy Project, 2023). Oklahoma ranks 10th nationally for marijuana use among individuals aged 18 or older, with about 26% reporting past­year use.

We aimed to determine the level of marijuana use by college students in Oklahoma, the common motives, and the negative consequences of use. Second, we aimed to determine how marijuana use, motives for use and consequences of use were related to individual differences in personality traits (i.e., sensation seeking and Big Five traits). We expected to observe that higher frequency of use and more negative consequences of use would be related to higher levels of sensation­seeking personality traits, lower levels of conscientiousness, lower levels of agreeableness, and higher levels of neuroticism (Jones et al., 2022). Third, we aimed to determine how protective factors of resilience related to use, motives for use and negative consequences of use. We expected to observe that those using marijuana the most and experiencing the most negative consequences would be those with the lowest levels of resilience (Myntti & Armstrong, 2022; Wingo et al., 2014).

Method

Participants

The participants were 141 undergraduates (114 women, 23 men, 1 other, 3 preferred not to respond) who were enrolled in psychology and speech courses at Oklahoma State University. They received course credit as compensation for their participation. The average age of our participants was 21.66 years (SD = 5.28). The sample was mostly White (62%). Other groups in the sample include Latinx (4%), Asian/Asian American (3%), Black/African American (3%), Native American (4%) and more than one group (24%). In the sample, 44.6% reported having a marijuana card from Oklahoma; 3% reported having a marijuana card from another state. The mean score on the cut down, annoyed, guilty, eye-opener – adapted to include drugs (CAGE­AID, Brown & Rounds, 1995) questionnaire was 1.93 out of maximum score of 4 (SD = 1.25).

Materials

We assessed frequency of usage with one item that we constructed for the study: “How many times have you used marijuana in the past 12 months? Select the response that most closely approximates your frequency of usage.” The options were (a) only 1 or 2 times in the past 12 months, (b) more than twice, but less than once a month, (c) about once a month, (d) multiple times a month, but less than once a week, (e) about once a week, (f) multiple times a week, but less than once a day, (g) about once a day, (h) multiple times a day, and (i) prefer not to respond. Participants were also asked to indicate whether they had a medical marijuana card from Oklahoma or another state.

We used established measures to assess marijuana use, motives, consequences, sensation­seeking personality traits, and Big Five personality traits. We assessed the level of marijuana use using the 4­item CAGE­AID questionnaire adapted to include drugs (CAGE­AID, Brown & Rounds, 1995). The questions were: (a) “Have you ever felt you ought to cut down on your marijuana use?” (b) “Have people ever annoyed you by criticizing your marijuana use?” (c) “Have you ever felt bad or guilty about your marijuana use?” and (d) “Have you ever used marijuana first thing in the morning to steady your nerves or get rid of a hangover?” Participants responded yes or no. The number of yes responses were summed with higher scores reflecting greater likelihood of CUD. The CAGE­AID measure improved upon the CAGE measure in terms of providing reliable measurement for participants are different types of participants (e.g., sex, income, and education; Brown & Rounds, 1995).

Motives for using marijuana were assessed using 18­items from an established measure (Boys et al., 2001). The questions focused on six motives: (a) changing mood (e.g., “When using marijuana, I use it mostly to

make myself feel less down or depressed”), (b) physical effects (e.g., “When using marijuana, I use it mostly to help me stay awake”), (c) social purposes (e.g., “When using marijuana, I use it mostly to enjoy the company of my friends”), (d) facilitate activity (e.g., “When using marijuana, I use it mostly to enhance an activity, such as listening to music or playing a game or sport”), and (e) manage effects of other substances (e.g., “When using marijuana, I use it mostly to improve the effects of other substances”). Participants responded on a 5­point scale from 1 (never) to 5 (always). Prior research has found that the questions have good internal consistency (Cronbach’s alpha α = .78; Boys et al., 2001). In the present study, we observed good internal consistency (Cronbach’s alpha α = .83).

We assessed the negative consequences of marijuana use using the 50 ­ item Marijuana Consequences Questionnaire (MACQ, Simons et al., 2012). There were eight categories of negative consequences (i.e., social, impaired control, self­perception, self­care, risk behaviors, academic, physical dependence, and blackout consequences). Scores were determined by the total “yes” responses in each subgroup. This questionnaire provided detailed insights into the potential negative outcomes linked to marijuana consumption. Prior research has found that the scale has acceptable internal consistency (Cronbach’s alpha α = .98; Simons et al., 2012). In the present study, we observed good internal consistency (Cronbach’s alpha α = .93).

We assessed sensation­seeking personality traits using the Brief Sensation Seeking Scale (BSSS, Hoyle et al., 2002). Participants scored how well they identified with each statement on a 5­point scale (1 = disagree strongly to 5 = agree strongly). Statements described different sensation­seeking behaviors (e.g., “I like to do frightening things”, and “I like wild parties”). Prior research has found that the scale has acceptable internal consistency (Cronbach’s alpha α = .74; Hoyle et al., 2002). In the present study, we observed acceptable internal consistency (Cronbach’s alpha α = .70).

We assessed the Big Five Personality traits using the 40 ­ item mini ­ markers questionnaire (Saucier, 1994). Participants are asked to indicate how accurately 40 adjectives describe them using a 9 ­ point scale (1 = extremely inaccurate to 9 = extremely accurate). Each trait was assessed using eight adjectives. For each trait, scores were averaged with higher averages reflecting more of the trait. Prior research has found that the scale has good internal consistency (Cronbach’s alphas ranging from α = .76 to α = .86; Saucier, 1994). In the present study, we observed good internal consistency (Cronbach’s alpha ranging from α = .70 to α = .87).

We assessed protective factors of resilience using

Marijuana Use by College Students | Sims and Kennison

the 24­item scale of protective factors questionnaire (Ponce­Garcia et al., 2015), which includes four factors with two related to social protective factors of resilience (i.e., social support and social skills) and two related to cognitive protective factors of resilience (i.e., goal efficacy and planning and prioritizing behavior). Each factor corresponded to six items, with each item rated on a 7­point scale (1 = not at all like me, 7 = exactly like me). Sample items included “I am confident in my ability to achieve goals”, and “I am good at working with others as a team.” Responses to items were summed for each factor. Those with the highest resilience level had the highest summed scores. Prior research has reported that the measure has high internal consistency (Cronbach’s alphas ranging from α = .90 to α = .93; Hamilton et al., 2021; Kennison & Spooner, 2023). We also observed high internal consistency for the measure in the present study (Cronbach’s alphas ranging from α = .83 to α = .91). We computed overall resilience by summing scores for each of the four protective factors. One question was included to check how closely participants were paying attention during the survey. The question was phrased as follows: “For this question, we would like to measure how much participants are paying attention to the survey. Please select the response “usually not busy at all” for this question ” The options were as follows: (a) usually extremely busy, (b) usually very busy, (c) usually somewhat busy; (d) usually not busy at all, (e) never busy, and (f) prefer not to respond.

Procedure

After we obtained IRB approval for the study from the Institutional Review Board at Oklahoma State University, we recruited participants from the Sona system in the department of psychology, which allows researchers to post studies and allows student participants to make appointments to complete studies. To be eligible for the study, participants must have used marijuana multiple times in the last 12 months. Approximately 30 percent of the students enrolled in Sona were eligible to participate in the study. We used a convenience sampling design. The study was described on Sona as follows: “The study explores motives for using marijuana and experiences with the drug as well as personality, attitudes, life experiences, and demographic variables, such as gender, ethnicity, age, as well as others.” The study was constructed using a professional license of Qualtrics. All participants provided their informed consent before beginning the survey. Participants completed the study online. Participants received the measures in the same order (Big Five personality traits, sensation­seeking personality traits, motives, CAGEAID, consequences, and demographics, including

Consequences_impaired_control

Note. SS = Sensation-Seeking.

Summary of Correlation Results

Consequences_impaired_control

Consequences_blackout

Note. Freq = frequency of marijuana use; CAGE = CAGE-AID; E = Extroversion; A = Agreeableness; C = Conscientiousness; N = Neuroticism; O = Openness; SS = Sensation-Seeking; SPF = Scale of Protective Factors; values for Freq reflect Spearman’s rank coefficients. * p < .05. ** p < .01. *** p < .001.

whether they have a card for medical marijuana). The order of items within each scale was randomized for each participant. We mitigated demand characteristics in the study by ordering personality questions before questions about sensitive questions about marijuana use and consequences and keeping the reference to marijuana use minimal. Data were collected anonymously with a small set of demographic variables, which makes reidentification of participants less likely to occur. Data analyses were performed using IBM SPSS 29.

Results

Initially, data were screened for inattentive responding. Twenty­five participants were excluded from the data set due to incorrectly answering the attention check question or not completing at least 80 percent of the questions in the study. The remaining dataset contained data from 116 participants (93 women, 19 men, 1 other, and 3 prefer not to respond). Responses were used to compute the descriptive statistics for the variables. Table 1 displays these results. Most variables were approximately normally distributed (excluding motives related to managing the effects of other drugs). The results indicated a high level of marijuana use. These results are displayed in Figure 1. Over 50% of the sample reported using marijuana multiple times a week. Approximately 35% reported using multiple times a day.

We examined the relationships among frequency of usage, motives for usage, and negative consequences of usage using correlational analyses. Table 2 displays these results. More frequent marijuana use was positively correlated to four of the five common motives for using marijuana: change mood ( r = .44, p < .001), social purposes (r = .22, p = .02), facilitate activity (r = .52, p < .001), and physical effects (r = .43, p < .001). CAGEAID scores were positively correlated to all five of the common motives for using marijuana: change mood (r = .29, p = .002), social purposes (r = .34, p < .001), facilitate activity ( r = .33, p < .001), physical effects (r = .33, p < .001), and manage effects of other drugs (r = .25, p = .007). More frequent use was related to six of the eight common negative consequences of marijuana use: academic (r = .30, p = .001), social (r = .21, p = .03), self­care (r = .45, p < .001), risk­behavior (r = .32, p = .001), impaired control ( r = .42, p < .007), and physical dependency (r = .61, p = .007). Those reporting the most frequent use also had significantly higher CAGE­AID scores (r = .55, p < .001).

Personality traits (i.e., Big Five and sensationseeking) were not related to frequency of usage (rs < .17, ps > .07). Those reporting higher levels of cognitive protective factors of resilience reported less frequent usage (r = ­.22, p = .020). Social protective

factors of resilience were unrelated to frequency of usage. Personality traits were related to motives and consequences of use. These results are displayed in Table 2. The results showed that lower levels of conscientiousness were significantly related to higher levels of the change mood motive for using marijuana (r = ­.30, p = .001) and higher levels of six of the eight common negative consequences: social (r = ­.25, p = .008), self­care (r = ­.31, p = .001), self­perception (r = ­.25, p = .008), impaired control (r = ­.27, p = .003), physical dependency (r = ­.26, p = .006), and blackout (r = ­.26, p = .006). Lower levels of agreeableness were related to significantly higher levels of two of the five common motives: change mood (r = ­.21, p = .020) and manage effects of other drugs (r = ­.28, p = .002). Higher levels of neuroticism were significantly related only to higher levels of social negative consequences (r = ­.25, p = .007).

The results showed that higher levels of sensationseeking personality traits were significantly related to one of the five common motives for marijuana use: facilitate activity (r = .28, p = .002). Higher levels of sensation ­ seeking personality traits were also significantly related to higher levels of one of the eight common negative consequences of marijuana use: risk behavior (r = ­.26, p = .005). Sensation seeking personality traits were not related to frequency of marijuana use or CAGE ­ AID scores. Higher levels of extraversion were related to significantly lower levels of two of the common motives: social purposes (r = ­.19, p = .04) and change mood (r = ­.26, p = .006) and to lower levels of three of the common negative consequences: academic (r = ­ .21, p = .020), physical dependency ( r = ­ .26, p = .006), and blackout (r = ­.23, p = .020). Higher levels of

Frequency of Marijuana Use Among Participants

Note. This chart illustrates the distribution of frequency responses from our sample. 23% of participants reported using multiple times a day, 12% reported once a day, etc. Individuals scored their responses with 1 = only 1 or 2 times a year to 8 = multiple times a day

openness were significantly related to lower levels of the three of the eight common negative consequences: social (r = ­.19, p = .045), impaired control(r = ­.25, p = .006), and blackout (r = ­.36, p = .001).

The results showed that there were relationships between the protective factors of resilience and motives and consequences of marijuana use. Higher levels of social protective factors were significantly related to higher levels of three motives when using marijuana: social ( r = ­ .21, p = .020), change mood ( r = ­ .31, p = .001), and manage the effects of other drugs ( r = ­ .20, p = .030) and lower levels of one of the negative consequences: blackout ( r = ­ .28, p = .002). Those higher in cognitive protective factors of resilience reported significantly lower levels of four of the five motives for marijuana use: change mood (r = ­.34, p = .001), facilitate activity (r = ­.21, p = .020), physical effects (r = ­.20, p = .030), and manage the effects of other drugs (r = ­.24, p = .007). Higher levels of cognitive protective factors of resilience also reported significantly being less likely to experience all eight of the negative consequences marijuana use: academic (r = ­.21, p = .020), social (r = ­.27, p = .002), self­perception (r = ­.30, p = .001), self­care (r = ­.42, p < .001), risk behavior (r = ­.20, p = .030), impaired control ( r = ­ .32, p = .001), physical dependency (r = ­.24, p = .007), and blackout (r = ­.32, p = .001).

To explore further how individual differences in personality traits and resilience predicted negative consequences of marijuana use, we conducted a multiple regression analysis in which the dependent variable was the sum of the eight categories of negative consequences. The independent variables were the Big Five personality traits (i.e., extraversion, agreeableness, conscientiousness, mood instability, and openness), sensation­seeking, and protective factors of resilience (i.e., social and cognitive factors). The variables were entered simultaneously. The assumptions for multiple regression were met (Field, 2024). The model was significant, F(8, 115) = 3.33, p = .002, with 14% of variance observed for negative consequences of use was explained (i.e., adjusted R2 = .14). Cognitive protective factors of resilience were the only significant predictor (β = ­.32, p = .018). Table 3 displays a summary of these results.

Discussion

The study examined marijuana use by college students in Oklahoma. The use of marijuana for medical purposes became legal in Oklahoma in 2018. The results showed a high level of marijuana use with 23.2% reporting using marijuana daily and over 50% using multiple times per week. Over 62% of participants had a CAGE­AID score above 2, indicating a high and concerning level

of use. Approximately 48% of the sample had a medical marijuana card either from Oklahoma or another state. Lower levels of extraversion were related to higher levels of social motives and motives related to mood for using marijuana. Lower extraversion was also related to higher levels of three negative consequences (i.e., academic, physical dependency and blackout). Lower levels of openness were related to higher levels of three negative consequences (i.e., social, impaired control, and blackout). Higher levels of sensation­seeking personality traits were related to one motive for use (i.e., facilitating activity) and one negative consequence (i.e., risk ­ behavior). Those higher in cognitive protective factors of resilience reported less frequent use. Personality traits were not related to the overall frequency of usage or CAGE­AID scores. In a multiple regression analysis predicting total negative consequences of marijuana use, only cognitive protective factors of resilience emerged as a significant predictor. Those lower in cognitive protective factors of resilience (i.e., goal efficacy and planning behavior) experienced more negative consequences of use. These results are the first to document the level of marijuana use by college students in Oklahoma since the legalization of medical marijuana in the state in 2018. We hope that these results can serve as an impetus for future research with the population of college students in the region. We speculate that colleges and universities are likely to experience higher attrition, stemming from marijuana use interfering with students’ academic performance. Furthermore, with increased marijuana use by college students, there may be increased demand for mental health resources available at colleges and universities.

The present study also provides new knowledge about how protective factors of resilience are related to

TABLE 3

marijuana use by college students and the experienced consequences. Among the most noteworthy findings is the fact that protective factors of resilience are not equal in predicting which students will have the most negative consequences of marijuana use. We assessed social and cognitive protective factors of resilience using the scale of protective factors (Ponce­Garcia et al., 2015). Social protective factors of resilience include social support and social skills. In contrast, cognitive protective factors of resilience include goal efficacy (e.g., belief one can achieve goals) and planning and prioritizing behavior (e.g., engaging in planning to achieve goals). Only cognitive protective factors of resilience emerged as a significant predictor of total negative consequences reported by students. Based on these findings, we believe that colleges and universities may be able to implement interventions that provide students with opportunities to strengthen their planning behavior and, if possible, goal efficacy. A recent study discussed the benefit of strengthening skills related to resilience (Hamilton et al., 2021).

In the present study, we did not find that personality traits were related to frequency of marijuana use or use meeting criteria for disordered use (i.e., using CAGE­AID scores). Prior studies have shown that Big Five personality traits (Dash et al., 2019; Jones et al., 2022) and sensationseeking traits were related to drug use (Galizio et al., 1983; LaSpada et al., 2020; Parnes et al., 2024; Zuckerman, 2007). We speculate that because our sample of participants were using marijuana at relatively high rates, individual differences in personality traits mattered less. In contexts in which many people are using substances and obtaining substances is easy and relatively inexpensive, personality traits may explain less variances in use.

There were multiple limitations in the study. Foremost, the study relied on self­report. Self­report responses may be inaccurate due to lapses in memory or a tendency to underreport behaviors that may be perceived as socially undesirable. Participants’ report about marijuana use and consequences may be inaccurate. In addition, the sample may be affected by self­selection bias, as some who were eligible for the study may not have participated due to concerns about possible social or legal ramifications. The campus at which this study was conducted lists marijuana as prohibited, even with a medical marijuana card, in its student code of conduct. Consequently, the use of marijuana among college students may be even higher than estimated in this study. Also, our sample was majority White and primarily included individuals from the midwestern region of the United States. The results may not generalize to other types of populations. Lastly, we asked students about their marijuana use with a question that we constructed rather than using an established measure from prior

research studies. In future research studies, researchers should use a validated measure of marijuana use. Future research is needed to understand whether marijuana use by college students increases during the transition from high school to college. Greater autonomy may lead some college students to obtain a medical marijuana card without their parents’ knowledge. Students’ perceptions about the benefits and risks of marijuana may be influenced by peers and a misperception that the drug has no risk and everyone is using it. Reducing marijuana use by college students will likely involve campuses implementing empirically supported interventions (Hone et al., 2024). Interventions may differ in their effectiveness. Furthermore, the effectiveness of interventions may vary across populations. For example, it is unclear whether interventions would be comparably effective in states with and without laws legalizing medical or recreational marijuana use.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425596

Aston, A. (2024, February 10). How do I get a medical marijuana card in Oklahoma? What to know about eligibility, fees. The Oklahoman. https://www.oklahoman.com/story/news/healthcare/2024/02/10/ medical-marijuana-card-oklahoma-what-to-know-about-licenseseligibility/72520981007/

Boys, A., Marsden, J., & Strang, J. (2001). Understanding reasons for drug use amongst young people: A functional perspective. Health Education Research, 16(4), 457–469. https://doi.org/10.1093/her/16.4.457

Brown, R. L., & Rounds, L. A. (1995). Conjoint screening questionnaires for alcohol and other drug abuse: Criterion validity in a primary care practice. Wisconsin Medical Journal, 94(3), 135–140. https://doi.org/10.1080/10401339509539722

Crean, R. D., Crane, N. A., & Mason, B. J. (2011). An evidence-based review of acute and long-term effects of cannabis use on executive cognitive functions. Journal of Addiction Medicine, 5(1), 1–8. httsp://doi.org/10.1097/ADM.0B013E31820C23FA

Dash, G. F., Slutske, W. S., Martin, N. G., Statham, D. J., Agrawal, A., & Lynskey, M. T. (2019). Big Five personality traits and alcohol, nicotine, cannabis, and gambling disorder comorbidity. Psychology of Addictive Behaviors, 33(4), 420–429. https://doi.org/10.1037/adb0000468

Denson, T. F., & Earleywine, M. (2006). Decreased depression in marijuana users. Addictive Behaviors, 31(4), 738–742.https://doi.org/10.1016/j.addbeh.2005.05.052

Drazdowski, T. K., Kliewer, W. L., & Marzell, M. (2021). College students’ using marijuana to sleep relates to frequency, problematic use, and sleep problems. Journal of American College Health, 69(1), 103–112. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7061072/ Duckworth, J. C., Graupensperger, S., Schultz, N. R., Gilson, M. S., Fairlie, A. M., Patrick, M. E., & Lee, C. M. (2023). Alcohol and marijuana use predicting next-day absenteeism and engagement at school and work: A daily study of young adults. Addictive Behaviors, 142, 107670. https://doi.org/10.1016/j.addbeh.2023.107670

Evins, A. E., Green, A. I., Kane, J. M., & Murray, R. M. (2012). The effect of marijuana use on the risk for schizophrenia. Journal of Clinical Psychiatry, 73(11), 1463–1468. https://doi.org/10.4088/JCP.12012co1c

Field, A. (2024). Discovering statistics using SPSS (6th Ed.). Sage. Galizio, M., Rosenthal, D., & Stein, F. A. (1983). Sensation seeking, reinforcement, and student drug use. Addictive Behaviors, 8(3), 243–252. https://doi.org/10.1016/0306-4603(83)90019-9https://doi.org/10.1016/03064603(83)90019-9

Garmezy, N., Masten, A. S., & Tellegen, A. (1984). The study of stress and competence in children: A building block for developmental psychopathology. Child Development, 55(1), 97–111. https://doi.org/10.2307/1129837

Green, B. E., & Ritter, C. (2000). Marijuana use and depression. Journal of Health

and Social Behavior, 40–49. https://doi.org/10.2307/2676359

Gunn, R. L., Aston, E. R., Sokolovsky, A. W., White, H. R., & Jackson, K. M. (2020). Complex cannabis use patterns: Associations with cannabis consequences and cannabis use disorder symptomatology. Addictive Behaviors, 105, 106329. https://doi.org/10.1016/j.addbeh.2020.106329

Hamilton, S. L., Baraldi, A., & Kennison, S. M. (2021). ADHD symptoms, stress, and resilience in college students. Journal of College Counseling, 24, 49–62 https://doi.org/10.1002/jocc.12176

Hasin, D. S., Saha, T. D., Kerridge, B. T., Goldstein, R. B., Chou, S. P., Zhang, H., Jung, S., Pickering, R. P., Ruan, W. J., Smith, S. M., Huang, B., & Grant, B. F. (2015). Prevalence of marijuana use disorders in the United States between 2001–2002 and 2012–2013. JAMA Psychiatry, 72(12), 1235–1242. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5037576/ Hayatbakhsh, M. R., Najman, J. M., Jamrozik, K., Mamun, A. A., Alati, R., & Bor, W. (2007). Cannabis and anxiety and depression in young adults: A large prospective study. Journal of the American Academy of Child & Adolescent Psychiatry, 46(3), 408–417. https://doi.org/10.1097/chi.0b013e31802dc54d

Hone, L. S., Boness, C. L., Helle, A. C., Kilmer, J. R., Morgan-Daniel, J., Chance, S. M., Rohleen, T. K., & Cronce, J. M. (2024). A scoping review of interventions for US college students’ harmful cannabis use and a call for a college cannabis intervention matrix. International Journal of Mental Health and Addiction, 1–29. https://doi.org/10.1007/s11469-024-01361-6

Hoyle, R. H., Stephenson, M. T., Palmgreen, P., Lorch, E. P., & Donohew, R. L. (2002). Reliability and validity of a brief measure of sensation seeking. Personality and Individual Differences, 32(3), 401–414. https://doi.org/10.1016/S0191-8869(01)00032-0

Jeffers, A. M., Glantz, S., Byers, A. L., & Keyhani, S. (2024). Association of cannabis use with cardiovascular outcomes among US adults. Journal of the American Heart Association, e030178. https://doi.org/10.1161/JAHA.123.030178

John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative big five trait taxonomy. Handbook of Personality: Theory and Research, 3(2), 114–158. Jones, S. A., Van Fossen, R. P., Thompson, W. K., Baker, F. C., Clark, D. B., & Nagel, B. J. (2022). Developmental trajectories of Big Five personality traits among adolescents and young adults: Differences by sex, alcohol use, and marijuana use. Journal of Personality, 90(5), 748–761. https://doi.org/10.1111/jopy.12694

Kennison, S. M., & Spooner, V. H. (2023). Childhood relationships with parents and attachment as predictors of resilience in young adults. Journal of Family Studies, 29(1), 15–27. https://doi.org/10.1080/13229400.2020.1861968

LaSpada, N., Delker, E., East, P., Blanco, E., Delva, J., Burrows, R., Lozoff, B., & Gahagan, S. (2020). Risk taking, sensation seeking, and personality as related to changes in substance use from adolescence to young adulthood. Journal of Adolescence, 82, 23–31. https://doi.org/10.1016/j.adolescence.2020.04.011

Lee C. M., Neighbors C., Hendershot C. S., Grossbard J. R. (2009). Development and preliminary validation of a comprehensive marijuana motives questionnaire. Journal of Studies on Alcohol and Drugs, 70(2), 279–287. https://doi.org/10.15288/jsad.2009.70.279

Marijuana Policy Project. (2023). Medical cannabis patient numbers. https://www.mpp.org/issues/medical-marijuana/state-by-state-medicalmarijuana-laws/medical-marijuana-patient-numbers/ Masten, A. S., Burt, K. B., Roisman, G. I., Obradovic, J., Long, J. D., & Tellegen, A. (2004). Resources and resilience in the transition to adulthood. Continuity and Change. Development and Psychopathology, 16(4), 1071–1094. https://doi.org/10.1017/S0954579404040143

Myntti, W. W., & Armstrong, S. B. (2022). Bouncing back from COVID-19: The role of resilience in depression, substance use, and loneliness in college students amidst the pandemic. Journal of American College Health, 1–10. https://doi.org/10.1080/07448481.2022.2135377

Naegele, H., Betzler, F., Viohl, L., Koslowski, M., Ernst, F., & Petzold, M. B. (2022). Cannabis use, use motives and cannabis use disorder among Berlin college students. Journal of Drug Issues, 52(4), 568–584. https://doi.org/10.1177/00220426221086877

National Institute on Drug Abuse. (2023, August 17). Marijuana and hallucinogen use, binge drinking reached historic highs among adults 35 to 50 https://nida.nih.gov/news-events/news-releases/2023/08/marijuana-andhallucinogen-use-binge-drinking-reached-historic-highs-among-adults-35-to-50

National Institutes of Health. (2021). Marijuana use at historic high among college-aged adults in 2020. News Releases. Retrieved July 12, 2024 from https://www.nih.gov/news-events/news-releases/marijuana-use-historichigh-among-college-aged-adults-2020

Okey, S. A., Waddell, J. T., & Corbin, W. R. (2022). I smoke alone: Indirect effects of

solitary cannabis use on negative consequences through coping motives. Journal of Studies on Alcohol and Drugs, 83(5), 721–730. https://doi.org/10.15288/jsad.21-00200

Oklahoma Secretary of State. (2022, January 4). State Question No. 820, Initiative Petition No. 434. https://www.sos.ok.gov/documents/questions/820.pdf

Parnes, J. E., Tyskiewicz, A. J., Prince, M. A., & Conner, B. T. (2024). Daily cannabis use: Do impulsivity and sensation seeking predict negative cannabis related consequences? Journal of Drug Issues, 00220426231226221. https://doi.org/10.1177/00220426231226221

Patel, J., & Marwaha, R. (2022). Cannabis use disorder. National Library of Medicine https://www.ncbi.nlm.nih.gov/books/NBK538131/

Phillips, K. T., Lalonde, T. L., Phillips, M. M., & Schneider M. M. (2017). Marijuana use and associated motives in Colorado university students. The American Journal on Addictions, 26(8), 830–837. https://doi.org/10.1111/ajad.12640

Ponce-Garcia, E., Madewell, A. N., & Kennison, S. M. (2015). The development of the scale of protective factors: Resilience in a violent trauma sample. Violence and Victims, 30(5), 735–755. https://doi.org/10.1891/0886-6708.VV-D-14-00163

Saucier, G. (1994). Separating description and evaluation in the structure of personality attributes. Journal of Personality and Social Psychology, 66(1), 141. https://doi.org/10.1037/0022-3514.66.1.141

Scholes-Balog, K. E., Hemphill, S. A., Patton, G. C., & Toumbourou, J. W. (2013). Cannabis use and related harms in the transition to young adulthood: A longitudinal study of Australian secondary school students. Journal of Adolescence, 36(3), 519–527. https://doi.org/10.1016/j.adolescence.2013.03.001

Schulenberg, J. E., Johnston, L. D., O’Malley, P. M., Bachman, J. G., Miech, R. A., & Patrick, M. E. (2018). Monitoring the future national survey results on drug use, 1975–2017. Volume II, College Students & Adults Ages 19–55. Institute for Social Research. University of Michigan.

Schultz, N. R., Bassett, D. T., Messina, B. G., & Correia C. J. (2019). Differential role of cannabis use motives in predicting impairment across three measures. Journal of Studies on Alcohol and Drugs, 80(1), 26–31. https://doi.org/10.15288/jsad.2019.80.26

Shi, Y. (2014). At high risk and want to quit: Marijuana use among adults with depression or serious psychological distress. Addictive Behaviors, 39(4), 761–767. https://doi.org/10.1016/j.addbeh.2013.12.013

Simons, J. S., Dvorak, R. D., Merrill, J. E., & Read, J. P. (2012). Dimensions and severity of marijuana consequences: Development and validation of the Marijuana Consequences Questionnaire (MACQ). Addictive Behaviors, 37(5), 613–621. https://doi.org/10.1016/j.addbeh.2012.01.008

Texoma HIDTA. (2025). Marijuana in Oklahoma. https://www.texomahidta.org/files/DDF/Texoma%20HIDTA%20 Oklahoma%20Marijuana%20Report%202025.pdf

Wingo, A. P., Ressler, K. J., & Bradley, B. (2014). Resilience characteristics mitigate tendency for harmful alcohol and illicit drug use in adults with a history of childhood abuse: A cross-sectional study of 2024 inner-city men and women. Journal of Psychiatric Research, 51, 93–99. https://doi.org/10.1016/j.jpsychires.2014.01.007

Zuckerman, M. (2007). Sensation seeking and risk. American Psychological Association. https://doi.org/10.1037/11555-000

Author Note

Shelia M. Kennison https://orcid.org/0000­0001­9298­3152

Laney S. Sims is now a student at Oklahoma State University College of Osteopathic Medicine.

The datasets generated and analyzed during the current study are not publicly available in order to protect participants’ anonymity as survey questions may identify them. The data are available from the corresponding author on reasonable request by researchers.

The authors have no conflict of interest to report in relation to the research in this report.

Laney S. Sims played a leading role in developing the conceptualization, data collection, original writing, and editing, and a supporting role in data analysis. Shelia Kennison played a leading role in conceptualization, data collection, data analysis, original writing, and editorial assistance.

Correspondence concerning this article should be addressed to Shelia M. Kennison, Ph.D., 116 Psychology Building, Department of Psychology, Oklahoma State University, Stillwater, OK 74078. Email: shelia.kennison@okstate.edu

Understanding Medication Nonadherence as Risk-Taking: Insights From College Students With Chronic Conditions

Annah L. Boone and Shelia M. Kennison* Department of Psychology, Oklahoma State University

ABSTRACT. We examined medication nonadherence among college students with chronic illnesses. We hypothesized that medication nonadherence is a form of risk taking and would be related to variables that have been shown to be associated with risk taking (e.g., personality traits and relationships with parents in childhood). The sample included 151 undergraduates with chronic illnesses for which they take medication. Participants were enrolled at Oklahoma State University. Participants completed an online survey, providing information about their medication adherence, beliefs about medicines, personality traits, general risk­taking in daily life, early childhood relationships with parents, and demographics. The results showed that 30% of the variance in medication nonadherence was accounted for in a multiple regression model (adjusted R2 = .30, p < .001). Significant predictors were sensation seeking personality traits, childhood relationship with mother, and medication concerns. Medication nonadherence can be viewed as a form of risk­taking. Those most likely to engage in medication nonadherence may be identified through screening of personality and relationships with parents during childhood.

Keywords: medication nonadherence, risk­taking, personality traits, parenting

Medication nonadherence occurs when one does not take prescription medication as prescribed (Barber et al., 2004; DiMatteo & DiNicola, 1982; DiMatteo et al., 2012; Marcum et al., 2013; Osterberg & Blaschke, 2005). Examples include not taking the medication at all, taking more or less than prescribed, and stopping medication early. Medication nonadherence is observed in approximately 30% to 50% of people taking long ­ term prescribed medication (Martin et al., 2019) and is observed across age groups (Mårdby et al., 2007). Research from the last decade showed that 45.8% of people in the United States had taken a prescription medication in the last month (Martin et al., 2019). Estimates of the costs of medication nonadherence in the United States exceed $100 billion annually (OECD, 2018). The World Health Organization recognized medication nonadherence as a global concern, as it can lead to serious health consequences (World Health Organization, 2003).

Historically, medication adherence has typically been viewed as strongly influenced by the provider–patient relationship and the communication that

happens at the time that a medication is prescribed (DiMatteo & DiNicola, 1982; DiMatteo et al., 2012; Martin et al., 2010). Approaches to reduce medication nonadherence typically focus on improving or increasing provider–patient communication. The information­motivation­strategy model characterizes provider–patient communication about medications as including three types of content: (a) information, which refers to how well providers and patients build a trusting relationship, and how comfortable patients feel sharing concerns and asking questions; (b) motivation, which refers to the providers’ ability to motivate the patient to view the medication as beneficial and necessary; and (c) strategy, which refers to concrete steps taken by the provider to reduce any barriers to compliance, such as reminders, cost­lowering measures, written instructions, as well as others (DiMatteo et al., 2012).

Studies of medication noncompliance have focused on a specific medical condition (e.g., diabetes, cancer) or a specific subpopulation of interest (e.g., different genders, age groups, ethnicities; Gast & Mathes, 2019; Shahin et al., 2019). Across subpopulations, patient

beliefs about health and medication are related to how well patients take medication as prescribed. Multiple factors are related to medication adherence (e.g., beliefs and knowledge about medication, health literacy, education level, social support; World Health Organization, 2003).

Despite the breadth of research on medication nonadherence, few studies have examined the problem in college students (Curtis & Kaugars, 2022; Hammonds et al., 2015; Hill, 2023; Labig et al., 2005). These studies have examined the role of beliefs (Labig et al., 2005); social stigma, anxiety, and peer support (Curtis & Kaugars, 2022); and other patient ­ related factors (e.g., health locus of control; Hammonds et al., 2015; Hill, 2023). Hammonds et al. (2015) conducted a randomized control trial to test medication adherence in college students who were prescribed an antidepressant. Participants were randomly assigned to an experimental group, which used a smartphone app to receive reminders to take their medication, or a control group, which did not use a smartphone app or any other addition to their daily routines. Medication adherence was higher for the experimental group. Other factors that were related to medication adherence were use of illegal drugs, health beliefs, and type of health care services being received (e.g., mental health care).

Prior studies have not approached medication nonadherence as a form of health­related risk­taking. Risk­taking is defined as any act in which the outcome is uncertain and may be adverse (Hansson, 2011). When one takes too much or too little of a medication, or takes the medication differently than it has been prescribed, one is taking a health risk. Although there are no prior studies exploring medication nonadherence as a type of risk taking, Peters et al. (2012) showed that medication nonadherence occurs more in college students who use substances (i.e., alcohol and marijuana), a form of risk­taking behavior. Moreover, studies that have demonstrated that some forms of risk­taking can be reduced through interventions have shown promise (Hennessy et al., 2013; Zimmerman et al., 2007), indicating that such approaches could be adapted to reduce medication nonadherence.

Prior research on risk­taking by college students has demonstrated that students’ retrospective reports of their past relationships with parents were related to current risk­taking (Kennison et al., 2016; Schwartz et al., 2009; Wood et al., 2021). For example, Schwartz et al. (2009) examined 12 health ­ related behaviors (e.g., drug use, binge drinking, risky sexual behavior, misuse of prescription medication) of college students and aspects of their prior relationships with their parents, including connection, disrespect, nurturance, and psychological control. The results showed higher levels of

positive relationships with parents were related to lower levels of risk­taking. The results are consistent with the notion that positive relationships with parents during childhood promote health development of children (Bowlby, 1988) and may help children learn to assess and avoid risks (Del Giudice, 2009). Other research has suggested that parents influence many aspects of the child’s psychosocial context in which the child learns new behaviors, some involving risk (Chisholm et al., 2005; Quinlan, 2007).

The present study examined medication nonadherence as a form of risk ­ taking in a sample of college students with chronic illnesses. A common definition of chronic illness is any condition lasting more than a year (Perrin et al., 2017). Approximately 27.6% of college students reported having a chronic illness (American College Health Association, 2016). These students face unique challenges during the transition from high school to college, including managing their health appointments and medications (Arnett, 2000), which may lead to increased medication nonadherence. Chronic illness in this population has been associated with lower levels of well­being (Wodka & Barakat, 2007) and greater risk of poor academic performance or dropping out (Maslow et al., 2011). Simultaneously, the transition from high school to college can involve increased risk­taking due to reduced parental supervision and expanded opportunities for novel experiences (Fromme et al., 2008; Romm et al., 2022).

We hypothesized that medication nonadherence is a type of risk­taking behavior and would be related to personality traits and early childhood factors previously identified to be related to risk­taking. Among personality traits, sensation seeking, which refers to a tendency to seek out new and intense experiences, has been consistently shown to relate to risk­taking of many types (Zuckerman, 2014). Research with college students has shown that those reporting higher levels of sensation seeking also reported higher levels of health­related risk­taking (Kennison & Messer, 2017; Kennison et al., 2016; Popham et al., 2011). Accordingly, we expected to observe that higher levels of sensation seeking would be related to higher levels of medication nonadherence.

We also hypothesized that medication nonadherence may be related to childhood relationships with parents. Prior research has shown that negative relationships with mothers during childhood have been associated with higher levels of risk taking (Kennison et al., 2016; Wood et al., 2021). Based on this literature, we hypothesized that negative maternal relationships during childhood would be associated with higher levels of medication nonadherence in college students.

Method

Participants

The sample included 151 undergraduates (127 women, 17 men, 2 nonbinary) at Oklahoma State University who were enrolled in psychology or speech communication courses and who self­identified as having a chronic illness. Participants received course credit for participating. Participants were 22.76 years old on average (SD = 7.21). In terms of ethnicity, most of the sample was White (67%). Other groups included in the sample are as follows: Native American (4%), Black/African American (4%), Latinx (2%), Asian/Asian American (2%), or belonging to more than one group (21%). Ninety­six percent of participants reported being raised by their biological mothers, and 89%, by their biological father. Numerous chronic illnesses were reported by participants. Most students reported more than one. The top 10 most reported illnesses are summarized in Table 1.

Materials

We used established measures in the study. We assessed medication nonadherence using the 4­item Morisky medication­taking adherence scale (MMAS, Morisky et al., 1986). Sample items include “Do you ever forget to take your medicine?” and “When you feel better do you sometimes stop taking your medicine?” Participants responded either yes or no. Participants received a point for each yes response for each item. Participants’ scores were summed with higher values indicating higher medication nonadherence. In research investigating medication adherence for participants being treated for high blood pressure, the measure was found to have adequate internal consistency (α = .74; Morisky et al., 1986). Morisky et al. (1986) also demonstrated concurrent validity for the measure in showing that medication adherence was related to participants’ blood pressure at baseline. In the present study, we observed acceptable internal consistency (α = .70).

We measured medication beliefs using the beliefs about medicines questionnaire (BMQ; Horne et al., 1999). The measure has two factors (i.e., necessity beliefs and medication concerns) , each involving 5 items. Participants rated each item using a 5­point scale from 1 (strongly disagree) to 5 (strongly agree). We summed scores for each factor. Higher sums reflected more beliefs about the medicine being necessary and higher levels of concern. Research has shown the measure’s internal consistency to be adequate with Cronbach’s alphas ranging from α = .65 to α = .89 (Çınar et al., 2018). Horn et al. (1999) demonstrated concurrent validity by showing that the measure was related to medication nonadherence in samples of patients undergoing treatment for heart issues, kidney dialysis, and cancer. In the present

study, we observed internal consistency ranging from α = .79 to α = .85.

We assessed sensation seeking personality traits using the 8­item brief sensation seeking scale (BSSS, Hoyle et al., 2002). Participants rated statements using a 5­point scale from 1 (disagree strongly) to 5 (agree strongly). Example items include “I like wild parties” and “I get restless when I spend too much time at home.” The measure has been shown to have acceptable internal consistency (α = .74; Hoyle et al., 2002). Hoyle et al. (2002) demonstrated concurrent validity for the measure in showing that responses were related to drug­related attitudes and self­reported drug­related behaviors. We also observed acceptable internal consistency in the present study (α = .80).

We measured general risk­taking in daily life using the 30­item domain­specific risk­taking attitude scale (DOSPERT, Blais & Weber, 2006). Each item described a type of risk ­ taking for one of the five domains of risk ­ taking (i.e., health/safety, recreational, social, ethical, and financial; e.g., passing off someone else’s work as your own). Participants were instructed to rate how likely they were to engage in that behavior using a 5­point scale from 1 (very unlikely) to 5 (very likely). We computed sums for the six items for each domain and then summed each domain to create a total risk­taking score. Higher sums reflected more risk­taking. Prior research showed that the measure had good reliability. Cronbach’s alphas ranged from α = .68 to α = .80 (Shou & Olney, 2020). Shou and Olney (2020) also demonstrated the measure’s concurrent validity in a study in which participants’ responses were related to self­reported sensation seeking personality traits and performance in a behavioral inhibition task in which risk­takers typically

Summary of Top 10 Chronic Illnesses

perform differently than non­risk­takers. In the present study, we observed good internal consistency for the DOSPERT total score (α = .85).

We quantified relationships with parents during childhood using the 24­item parents as social context questionnaire (PASCQ, Skinner et al., 2005). The measure has six factors: three positive (i.e., warmth, autonomy support, and structure) and three negative (i.e., coercion, rejection, and chaos). We asked participants to rate their relationship with the mother and father (or equivalent caregiver). For each question, participants used a 4­point rating scale (1 = not at all true, 2 = not very true, 3 = sort of true, and 4 = very true). We computed the sum for each category of questions (i.e., six categories for relationship with mother and six categories for relationship with father). We then averaged the three positive and three negative factors for mother ratings and father ratings, resulting in four composite variables: (a) positive relationship with mother, (b) positive relationship with father, (c) negative relationship with mother, and (d) negative relationship with father (Kennison & Byrd­Craven, 2020; Kennison & Spooner, 2023). Research has documented that the measure has good internal consistency (Cronbach’s alphas between α = .72 and α = .90; Kennison & Spooner, 2023). In the present study, good internal consistency was also observed (Cronbach’s alphas between α =. 79 and α = .93). Skinner et al. (2005) demonstrated the measure’s concurrent validity in showing that participants’ responses were related to a variety of outcomes (e.g., social competence, academic competence, and behavioral problems). We also asked participants to indicate who their primary male and female caregivers were during childhood using the following options: (a) biological parent; (b) adopted parent; (c) stepparent; (d) grandparent; (e) family friend; or (f) other.

Procedure

After obtaining IRB approval, we recruited participants from the Sona system in a department of psychology. Courses participating in the Sona system included psychology and speech communication courses. Most classes satisfied general education requirements; thus, they attract students of all undergraduate majors. Students are able to access Sona, review availability studies, and decide whether to participate. The present study was carried out as an online survey, which was constructed using a professional license of Qualtrics. Informed content involved participants viewing the first page of the survey, which provided information about the study, known risks, benefits, compensation, and how their data would be protected. Those indicating that they wished to volunteer for the study were instructed to

advance the survey. Those who decided not to volunteer were instructed to close their browser. The study used a cross­sectional design. The ordering of measures was the same for all participants (i.e., childhood relationships with parents, sensation­seeking personality, general risktaking, medication beliefs, medication adherence, and demographics). Demographic questions included: age in years, gender (i.e., man, woman, other), and racial/ ethnic group. Participants could skip any question by selecting the option “prefer not to answer.” The mean time that participants spent completing the survey was 46.3 minutes (median = 25 minutes). We used IBM SPSS Statistic 29 to analyze the data.

Results

We computed means for the variables (i.e., medication nonadherence, beliefs in medicine, necessity and concerns, sensation seeking, five domains of general risk­taking, attachment, childhood relationships with parents, and three subtypes of ACEs) and found that they were approximately normally distributed. Table 2 displays a summary of these results as well as a summary of the correlations between variables.

We tested the hypothesis that medication nonadherence would be related to sensation ­ seeking personality traits and general risk­taking in daily life using correlations. The results partially supported the hypothesis as that medication nonadherence was positively correlated with sensation ­ seeking traits (r = .29, p < .001). Medication nonadherence was not related to general risk­taking.

Next, we tested the hypothesis that medication TABLE 2

Summary of Descriptive Statistics

nonadherence would be related to childhood relationships with parents. The hypothesis was partially supported as medication nonadherence was positively related to higher levels of negative mother relationships (r = .31, p < .001) and negative father relationships (r = .23, p = .004) and lower levels of positive mother relationships (r = ­.14, p = .040).

We also found that beliefs about medication were related to medication nonadherence and childhood relationships with parents. Medication nonadherence was positively correlated with concerns about medication (r = .42, p < .001). Concerns about medication were also positively correlated with negative relationships with mothers (r = .48, p < .001) and fathers (r = .37, p < .001). Concerns about medication were negatively related to positive relationships with mothers (r = .40, p < .001) and fathers (r = .29, p = .001). Beliefs and concerns about medication were not related to sensation seeking nor general risk­taking.

We explored how the personality, family, and belief variables were related to medication nonadherence by carrying out a multiple regression analysis using medication nonadherence as the dependent variable. We entered the predictor variables simultaneously. These were gender (1 = male, 0 = female)1 sensationseeking personality traits, general risk­taking, childhood relationships with parents (i.e., positive mother, positive father, negative mother, and negative father), and beliefs in medicine (BMQ) (i.e., BMQ necessity and BMQ concerns). We confirmed that the assumptions for multiple regression were met (Field, 2024). Tolerance and VIF values were within acceptable ranges. Table 3 displays a summary of the results. The model was significant, F(9, 126) = 6.89, p < .001, ηp2 = .21, accounting for 29.6% of the variance (adjusted R2 = .296). There were four significant predictors: sensation seeking ( β = .26, p = .012), negative mother relationship (β = .38, p = .005), positive mother relationship (β = ­.288, p = .027), and BMQ concerns (β = .31, p < .001).

Discussion

In the study, we investigated medication nonadherence among college students with chronic illnesses. The results confirmed the two hypotheses: a) medication nonadherence would be higher for those with higher levels of sensation­seeking personality traits, which have been linked to risk taking (Zuckerman, 2014) and b) medication nonadherence would be higher for those with more negative relationships with parents during childhood, which have been linked to some forms of risk taking (Kennison et al., 2016; Wood et al., 2021). The results showed that negative mother relationships during childhood were related to taking risks with medication.

The results also showed that those reporting higher levels of concerns about medication reported higher levels of nonadherence.

These findings are the first to support the idea that medication nonadherence is a type of health­related risk taking, similar to other types of health­related risktaking (e.g., using/abusing substances, engaging in risky sexual behavior, not using safety devices when driving or riding vehicles). Conceptualizing medication nonadherence as a form of risk­taking may lead to new approaches for designing interventions to improve adherence among college students as well as other populations. For example, existing frameworks for reducing risk­taking in substance use or sexual health could be adapted to address medication adherence. Colleges and universities could also incorporate medication nonadherence into risk­reduction programs on campus. Such programs may reduce medication nonadherence and contribute to positive student outcomes, such as improved academic performance, retention, and degree completion.

An unexpected finding in the present study was that general risk­taking in daily life was not related to medication nonadherence. We speculate that the measure that we used to assess general risk ­ taking (i.e., the DOSPERT, Blais & Weber, 2006) was unable to capture typical risk­taking activities common to young adults with chronic illness. For example, the health/safety category of risk­taking in the DOSPERT represents six aspects of potential health­related risk­taking each

1Data from nine nonbinary participants were excluded from the analysis due to there being an inadequate number of participants to constitute a third gender category.

TABLE 3

with a separate item (i.e., drinking heavily, engaging in unprotected sex, driving without a seatbelt, riding a motorcycle without a helmet, sunbathing without sunscreen, and walking alone at night in an unsafe area). Chronic illness can reduce one’s physical activity in general, which may limit opportunities for engaging in these particular forms of risk­taking. Future research is needed to examine whether individuals with and without chronic illnesses engage in different types of risk taking and how sensation­seeking personality traits are related to the different types of risk­taking behaviors.

The results are consistent with Hammonds and colleagues’ (2015) research showing medication nonadherence for antidepressants was related to use of illicit drugs, a form of health­related risk­taking. Future research is needed to investigate whether medication nonadherence and its relationships with sensationseeking personality traits, beliefs about medication, and childhood­relationships with parents depend on the specific type of medication being used and the type of chronic illness that the person has. In the present study, in our sample of participants, there was a variety of chronic illnesses represented, for which the prescribed medications also varied. Future research is needed to develop and test interventions to improve medication adherence in college students.

The present results also add to the literature demonstrating relationships between risk­taking and childhood relationships with parents (Kennison et al., 2016; Schwartz et al., 2009; Wood et al., 2021). Such results are compatible with theories of development emphasizing the benefits of positive parental relationships with children (Bowlby, 1988). Future research is needed to understand how relationships with parents relate to their children’s long­term patterns of beliefs or behavior related to taking medication. We speculate that the relationships may be complex and involve factors not addressed in the present research. Understanding these relationships could lead to the development of screeners that healthcare providers could use to identify those who are most at risk of medication nonadherence as well as interventions designed to increase medication adherence.

There are multiple limitations in this study. Foremost, we assessed medication nonadherence using a self­report questionnaire. Responses on self­report measures may be inaccurate, due to participants’ misremembering their medication behaviors as well as due to participants’ possibly responding in a socially desirable manner. The latter aspect of inaccurate responding would likely lead to an estimate of medication nonadherence that is lower than what is the case. Second, our study recruited college students with a variety of different chronic illnesses, which were treated with different types

of medications with different regimens (e.g., required daily, multiple times a time, weekly). It is possible that taking risks with medications are more common with some types of medication and some types of regimens than others. Third, most of our participants reported having both female and male caregivers during childhood. It is possible that the results may not generalize to samples of individuals who have different family histories. Fourth, our sample was majority White and female; thus, the results may not generalize to other populations and samples. Lastly, other factors which were not examined in the present study (e.g., financial constraints, religious beliefs, family or peer influence) are likely to be related to medication nonadherence in some people.

There are several possible directions for future research. Examining the relationships from this study with different populations is needed to determine whether the outcomes are generally observed or specific to college students. It is likely that rates of risk­taking with medications would differ depending on aspects of the type of medication as well as the nature of the illness for which the medication is prescribed. Lastly, future research is needed to determine whether interventions characterizing medication nonadherence as risk­taking would be effective in decreasing it. Such studies ideally would utilize direct methods of assessing adherence (e.g., pill counting or other direct measurements of medication use) in addition to participant’s self­reports.

References

American College Health Association. (2016). National college health assessment II: Reference group executive summary Fall 2015. Silver Spring, Maryland: American College Health Association. https://www.acha.org/wp-content/uploads/2024/07/NCHA-II_Fall_2015_ reference_group_executive_summary.pdf

Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55(5), 469–480. https://doi.org/10.1037/0003-066X.55.5.469

Barber, N., Parsons, J., Clifford, S., Darracott, R., & Horne, R. (2004). Patients’ problems with new medication for chronic conditions. BMJ Quality & Safety, 13(3), 172–175. https://doi.org/10.1136/qshc.2003.005926

Blais, A., & Weber, E. U. (2006). A domain-specific risk-taking (DOSPERT) scale for adult populations. Judgment and Decision Making, 1(1), 33–47. https://doi.org/10.1017/S1930297500000334

Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. Routledge.

Chisholm, J. S., Quinlivan, J. A., Petersen, R. W., & Coall, D. A. (2005). Early stress predicts age at menarche and first birth, adult attachment, and expected lifespan. Human Nature, 16, 233–265. https://doi.org/10.1007/s12110-005-1009-0

Çınar, M., Tekgöz, E., Çınar, F. İ., & Yilmaz, S. (2018). Reliability and validity of the Turkish translation of the Beliefs about Medicines Questionnaire (BMQ-T) in patients with Behçet’s Disease. Turkish Thoracic Journal, 19(4), 231–232. https://doi.org/10.5152/TurkThoracJ.2018.180108

Curtis, C., & Kaugars, A. S. (2022). Correlating barriers to medication adherence with trait anxiety, social tigma, and peer support in college students with chronic illness. Psi Chi Journal of Psychological Research, 27(3), 178–184. https://doi.org/10.24839/2325-7342.JN27.3.178

Del Giudice, M. (2009). Sex, attachment, and the development of reproductive strategies. Behavioral and Brain Sciences, 32(1), 1–67. https://doi.org/10.1017/S0140525X09000016

DiMatteo, M. R., & DiNicola, D. D. (1982). Achieving patient compliance

Pergamon. https://hdl.handle.net/10822/792169

DiMatteo, M. R., Haskard-Zolnierek, K. B., & Martin, L. R. (2012). Improving patient adherence: A three-factor model to guide practice. Health Psychology Review, 6(1), 74–91. https://doi.org/10.1080/17437199.2010.537592

Field A. (2024). Discovering statistics using SPSS (6th Ed.). Sage.

Gast, A., & Mathes, T. (2019). Medication adherence influencing factors—An (updated) overview of systematic reviews. Systematic Reviews 8 112. https://doi.org/10.1186/s13643-019-1014-8

Fromme, K., Corbin, W. R., & Kruse, M. I. (2008). Behavioral risks during the transition from high school to college. Developmental Psychology, 44(5), 1497–1504. https://doi.org/10.1037/a0012614

Hammonds, T., Rickert, K., Goldstein, C., Gathright, E., Gilmore, S., Derflinger, B., Bennett , B., Sterns, A., Drew, B. L., & Hughes, J. W. (2015). Adherence to antidepressant medications: A randomized controlled trial of medication reminding in college students. Journal of American College Student Health, 63(3), 204–208. https://doi.org/10.1080/07448481.2014.975716

Hansson, S. O. (2011). Risk: Objective or subjective, facts or values. Risk Analysis, 13(2), 231–238. https://doi.org/10.1080/13669870903126226

Hennessy, M., Romer, D., Valois, R. F., Vanable, P., Carey, M. P., Stanton, B., Brown, L., DiClemente, R., & Salazar, L. F. (2013). Safer sex media messages and adolescent sexual behavior: 3-year follow-up results from project iMPPACS. American Journal of Public Health, 103(1), 134–140. https://doi.org/10.2105/AJPH.2012.300856

Hill, E. M. (2023). An examination of health locus of control in relation to antibiotic resistance perceptions and antibiotic use behaviors among college students. Journal of American College Student Health, 1–9. https://doi.org/10.1080/07448481.2023.2208227

Horne, R., Weinman, J., & Hankins, M. (1999). The beliefs about medicines questionnaire: The development and evaluation of a new method for assessing the cognitive representation of medication. Psychology & Health, 14(1), 1–24. https://doi.org/10.1080/08870449908407311

Hoyle, R. H., Stephenson, M. T., Palmgreen, P., Lorch, E. P., & Donohew, R. L. (2022). Reliability and validity of a brief measure of sensation seeking. Personality and Individual Difference, 32, 401–414 https://doi.org/10.1016/S0191-8869(01)00032-0

Kennison, S. M., & Byrd-Craven, J. (2020). Childhood relationship with mother as a precursor to ageism in young adults. Current Psychology, 39(4), 1331–1339. https://doi.org/10.1007/s12144-018-9838-2

Kennison, S. M., & Messer, R. H. (2017). Cursing as a form of risk-taking. Current Psychology, 36, 119–126. https://doi.org/10.1007/s12144-015-9391-1

Kennison, S. M., & Spooner, V. H. (2023). Childhood relationships with parents and attachment as predictors of resilience in young adults. Journal of Family Studies, 29(1), 15–27. https://doi.org/10.1080/13229400.2020.1861968

Kennison, S. M., Wood, E. E., Byrd-Craven, J., & Downing, M. L. (2016). Financial and ethical risk-taking by young adults: A role for family dynamics during childhood. Cogent Economics & Finance, 4(1), 1232225. https://doi.org/10.1080/23322039.2016.1232225

Labig, C. E., Zantow, K., & Peterson, T. O. (2005). Factors affecting students’ medicine-taking habits. Journal of American College Student Health, 54(3), 177–183. https://doi.org/10.3200/JACH.54.3.177-184

Marcum, Z. A., Sevick, M. A., & Handler, S. M. (2013). Medication nonadherence: A diagnosable and treatable medical condition. JAMA, 309(20), 2105–2106. https://doi.org/10.1001/jama.2013.4638

Mårdby, A. C., Åkerlind, I., & Jörgensen, T. (2007). Beliefs about medicines and self-reported adherence among pharmacy clients. Patient Education and Counseling, 69(1–3), 158–164. https://doi.org/10.1016/j.pec.2007.08.011

Martin, C. B., Hales, C. M., Gu, Q., & Ogden, C. L. (2019). Prescription drug use in the United States, 2015–2016. National Center for Health Statistics. https://stacks.cdc.gov/view/cdc/78184

Martin, L. R., Haskard-Zolnierek, K. B., & DiMatteo, M. R. (2010). Health behavior change and treatment adherence: Evidence-based guidelines for improving healthcare. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195380408.001.0001

Maslow, G. R., Haydon, A., McRee, A. L., Ford, C. A., & Halpern, C. T. (2011). Growing up with a chronic illness: Social success, educational/vocational distress. Journal of Adolescent Health,49(2), 206–212. https://doi.org/10.1016/j.jadohealth.2010.12.001

Morisky, D. E., Green, L. W., & Levine, D. M. (1986). Concurrent and predictive validity of a self-reported measure of medication adherence. Medical Care, 24(1), 67–74. https://doi.org/10.1097/00005650-198601000-00007

OECD. (2018). Investing in medication adherence improves health outcomes

and health system efficiency: Adherence to medicines for diabetes, hypertension, and hyperlipidaemia. OECD Health Working Papers, 105. https://doi.org/10.1787/8178962c-en

Osterberg, L., & Blaschke, T. (2005). Adherence to medication. New England Journal of Medicine, 353(5), 487-497. https://doi.org/10.1056/NEJMra050100

Quinlan, R. J. (2007). Human parental effort and environmental risk. Proceedings of the Royal Society of London B, 274, 121–125. https://doi:10.1098/rspb.2006.3690

Perrin, J. M., Bloom, S. R., & Gortmaker, S. L. (2017). The increase of childhood chronic conditions in the United States. JAMA, 297(24), 2755–2759. https://doi.org/10.1001/jama.297.24.2755

Peters, E. N., Leeman, R. F., Fucito, L. M., Toll, B. A., Corbin, W. R., & O’Malley, S. S. (2012). Co-occurring marijuana use is associated with medication nonadherence and nonplanning impulsivity in young adult heavy drinkers. Addictive Behavior, 37(4), 420–426. https://doi.org/10.1016/j.addbeh.2011.11.036

Popham, L., Kennison, S. M., & Bradley, K. I. (2011). Ageism, sensation-seeking, and risk-taking in young adults. Current Psychology, 30, 184–193 https://doi.org/10.1007/s12144-011-9107-0

Romm, K. F., Metzger, A., Gentzler, A. L., & Turiano, N. A. (2022). Transitions in risk-behavior profiles among first-year college students. Journal of American College Student Health, 70(7), 2210–2219. https://doi.org/10.1080/07448481.2020.1846048

Schwartz, S. J., Zamboanga, B. L., Ravert, R. D., Kim, S. Y., Weisskirch, R. S., Williams, M. K., Mersamin, M., & Finley, G. E. (2009). Perceived parental relationships and health‐risk behaviors in college‐attending emerging adults. Journal of Marriage and Family, 71(3), 727–740. https://doi.org/10.1111/j.1741-3737.2009.00629.x

Shahin, W., Kennedy, G. A., & Stupans, I. (2019). The impact of personal and cultural beliefs on medication adherence of patients with chronic illnesses: A systematic review. Patient Preference Adherence, 13, 1019–1035. https://doi.org/10.2147/PPA.S212046

Shou, Y., & Olney, J. (2020). Assessing a domain-specific risk-taking construct: A meta-analysis of reliability of the DOSPERT scale. Judgment and Decision Making, 15(1), 112–134. https://doi.org/10.1017/S193029750000694X

Skinner, E., Johnson, S., & Snyder, T. (2005). Six dimensions of parenting: A motivational model. Parenting: Science and Practice, 5(2), 175–235. https://doi.org/10.1207/s15327922par0502_3

Wodka, E. L., & Barakat, L. P. (2007). An exploratory study of the relationship of family support and coping with adjustment: Implications for college students with a chronic illness. Journal of Adolescence, 30(3), 365–376. https://doi.org/10.1016/j.adolescence.2006.04.002

Wood, E. E., Garza, R., Kennison, S. M., & Byrd-Craven, J. (2021). Parenting, cortisol, and risky behaviors in emerging adulthood: Diverging patterns for males and females. Adaptive Human Behavior and Physiology, 7, 114–132. https://doi.org/10.1007/s40750-021-00164-6

World Health Organization. (2003). Adherence to long-term therapies: Evidence for action. World Health Organization. https://iris.who.int/bitstream/handle/10665/42682/9?sequence=1

Zimmerman, R., Palmgreen, P., Noar, S., Lustria, M. L. A., Lu, H., & Horosewski, M. L. (2007). Effects of a televised two-city safer sex mass media campaign targeting high-sensation-seeking and impulsive-decision-making young adults. Health Education &Behavior, 34(5), 810–826. https://doi.org/10.1177/109019810729970

Zuckerman, M. (2014). Sensation seeking: Beyond the optimal level of arousal Psychology Press. https://doi.org/10.4324/9781315755496

Author Note.

Shelia M. Kennison https://orcid.org/0000­0001­9298­3152

The authors do not have any conflicts of interest to report for the research described in this report. The dataset will be made available to others upon a request directed to the corresponding author.

Annah L. Boone played a leading role in developing the conceptualization, data collection, original writing, and editing and a supporting role in data analysis. Shelia Kennison played a leading role in conceptualization, data collection, data analysis, original writing, and editorial assistance.

Correspondence concerning this article should be addressed to Shelia M. Kennison, 116 Psychology Building, Department of Psychology, Oklahoma State University, Stillwater, OK 74078. Email: shelia.kennison@okstate.edu

Biogenetic Endorsements of Depression: The Enduring Influence of the Chemical Imbalance Theory

Saoirse E. Ward and James W. Diller*

Department of Psychological Science, Eastern Connecticut State University

ABSTRACT. The chemical imbalance theory has been scrutinized for its simplistic hypothesis yet remains popular among the public. This study examined whether familiarity with the chemical imbalance theory correlates with biogenetic endorsements of depression and how these beliefs relate to treatment preferences, stigma, and prognostic pessimism. A total of 69 undergraduate students completed an online survey assessing beliefs about depression, familiarity with the theory, and measures of treatment preferences, stigma, and prognostic pessimism. Students familiar with the chemical imbalance theory endorsed biogenetic beliefs significantly more strongly than those unfamiliar with it,  t (63) = 2.50,  p  = .01,  d  = 1.01. A significant effect of treatment preference on biogenetic endorsements was found,  F(3, 61) = 4.19,  p = .01, η² = .17; students who preferred medication or combined treatments endorsed biogenetic causes more strongly than those preferring therapy or no treatment. Biogenetic beliefs were moderately correlated with greater prognostic pessimism, ρ = –.41,  p < .001, ρ² = .17, but not with stigma. These findings highlight the ongoing influence of the chemical imbalance theory on treatment preferences and attitudes towards depression.

Keywords: chemical imbalance, depression, etiological beliefs, treatment preferences, prognostic pessimism

Questions about the origins and the development of mental disorders have captivated human curiosity for centuries. Throughout history, explanations for the origin of mental health disorders have changed from ideas such as an imbalance of bodily fluids to unconscious conflicts and traumas in early childhood development. In recent decades, the causes of mental disorders have shifted to be attributed as biomedical diseases in relation to biogenetic factors (Lebowitz & Appelbaum, 2019). One widely accepted explanation is the chemical imbalance theory (Lebowitz & Appelbaum, 2019; White 2013), which serves as a common explanation for the biological cause of depression and is influential in shaping the public’s perceptions (Pescosolido et al., 2010; Schomerus et al., 2012). Alongside the dispute on the origins of depression, the efficacy of antidepressants as a treatment is also controversial, as researchers found that when compared to the placebo, the differences are minimal and may not be clinically significant. Rather, the effects can be explained by factors unrelated to the expected

therapeutic benefits of the medication (Jakobsen et al., 2019; Munkholm et al., 2019), further emphasizing that the mediating role of neurotransmitters in depression is not fully understood.

The dissemination of the chemical imbalance theory first began in the 1960s, during which Schildkraut (1965) posited that the neurotransmitter norepinephrine played a role in depression. Later, Coppen (1967) theorized a different neurotransmitter, serotonin, was more relevant. The exploration of the biochemistry of affective disorders, particularly depression, sought to measure the levels of serotonin, norepinephrine, hormones, and metabolic indicators within the blood and cerebrospinal fluid of individuals diagnosed with these disorders. Results from biochemical testing analyzed the correlation between biochemical markers and the symptoms of the disorders and suggested “imbalances” in the neurotransmitter serotonin to be a contributing factor in the pathophysiology of depression (Coppen, 1967). Through this research, the serotonin hypothesis influenced the cultivation of the chemical imbalance

theory, which aimed to answer the questions on the causes of depression and to revolutionize the treatment of individuals with mental disorders.

Numerous researchers have attempted to alter neurochemical levels within patients diagnosed with depression to identify its causes. Heninger and colleagues (1996) addressed the role of monoamines such as serotonin, norepinephrine, and dopamine in depression by depleting monoamine levels in patients with a history of depression. The mood of patients was assessed both before and after depletion, revealing that monoamine levels played no direct causal role in the causation of depression, rather a regulatory role. So, although they can influence mood, neurotransmitters cannot determine states of depression. This highlights that the cause of depression cannot be reduced to mediating factors such as neurochemical imbalances, and instead points to the complexity of the disorder and the need to adjust our understanding towards a multicausal explanation.

The chemical imbalance theory’s popularity was further advanced through the development of Selective Serotonin Reuptake Inhibitors (SSRIs; Ang et al., 2022), which were marketed by pharmaceutical companies as agonists which inhibit the reuptake of serotonin, thereby increasing the availability of serotonin in the synaptic cleft. The marketization of the antidepressant further promoted the story of depression being caused by a serotonin deficiency, with SSRIs serving as the key to correcting the deficiency and alleviating depression symptoms. However, in the United States, the prevalence of mental illnesses continued to rise, even as more individuals were prescribed antidepressants (Whitaker, 2010). This trend has led to growing concerns about the effectiveness of the biomedical approach and its treatments.

One concern Whitaker (2010) described is that hypotheses attributing depression primarily to genetic or biomedical factors have limited empirical support. The sensationalized research on the biochemistry of affective disorders was largely driven by pharmaceutical companies, which used the chemical imbalance narrative to market SSRIs such as Prozac and Zoloft (Lacasse & Leo, 2005; Leo & Lacasse, 2008). Lacasse and Leo (2005) noted, “the primary literature is mixed and plagued with methodological difficulties such as very small sample sizes and uncontrolled confounding variables” (p. 1221). An and colleagues (2009) found that direct­to­consumer marketing of prescription drugs affects individuals’ views of depression and preferred treatment. The media also engaged in framing depression as a neurochemical disorder, as journalists frequently referenced the chemical imbalance theory but failed to provide credible sources, according to Leo and Lacasse (2008). Additionally, the lack of a distinction between causation and correlation is

problematic, as decreases in neurotransmitter levels may not directly cause a behavioral disorder. An emphasis on biochemistry may have led to the overlooking of causal environmental and psychosocial factors.

Proponents of the brain disease models of psychopathology argue that biogenetic explanations reduce self­blame among individuals (Illes et al., 2008); nonetheless, this comes at the expense of increased stigma toward people with mental health disorders (Kvaale et al., 2013), and an adoption of negative beliefs about recovery (Kemp et al., 2014; Lebowitz et al., 2021; Schroder et al., 2020). Previous studies comparing the biopsychosocial model to the biomedical model assessed credibility, selfstigma, and treatment expectancies among students and found that although the biomedical model was associated with less self­blame, it was related to a worse expected prognosis (Deacon & Baird, 2009; Kemp et al., 2014; Kvaale et al., 2013; Lebowitz et al., 2013; Phelan, 2005; Phelan et al., 2006). The justification of the biomedical model is harmful, as research has shown that biogenetic explanations increase preferences for treatments such as medication (Deacon, 2013; Iselin & Addis, 2003; Phelan et al., 2006; Schomerus et al., 2012). This is a concern, because medication and psychotherapy offer similar effects in improving symptoms of depression (Cuijpers et al., 2013; Imel et al., 2008) and the effects of therapy have been shown to outweigh those of medication once treatment is discontinued (DeRubeis et al., 2008). A recent meta­analysis by Cuijpers and colleagues (2023) evaluated the efficacy of cognitive behavioral therapy (CBT) relative to other models of therapy, and found that CBT was as effective as pharmacotherapy in the short term, but more effective in the long term. Additionally, combined treatments were more effective than pharmacotherapy alone in both the short and long term. However, they did not outperform CBT alone. Moreover, Davis (2021) found that patients may find comfort in biogenetic explanations, as these can justify their suffering. However, the relationship between explanations on the causality of depression and treatment preferences are mixed. Some research suggests that beliefs related to neurochemical factors predict more positive attitudes and preferences towards medication (Kemp et al., 2014; Watson & Beshai, 2020). In contrast, other studies suggest that it does not. They observed that endorsements of the biological model were associated with preferences for psychotherapy (Goldstein & Rosselli, 2003; Salem et al., 2019).

Mental disorders arise from a complex interaction of many factors, making a single explanation or treatment approach insufficient. For treatment to be effective, researchers emphasize the importance of using a multicausal and multifactorial framework (France et al., 2007). Dar­Nimrod and Heine (2011) and Haslam

(2011) describe the problem that genetic essentialism may pose on individuals’ perceptions of their mental health, suggesting it could be a potential explanation for why biological explanations of depression evoke pessimistic beliefs regarding prognosis. They explain further how genetic essentialism can undermine agency, increase stigma, and reduce motivation for self­improvement or treatment. More recent findings by Berent and Platt (2021) demonstrated that people spontaneously essentialize psychiatric conditions linked to the brain, even when the biological explanation includes no mention of genetic factors, illustrating the strength of the essentialist bias. By conceptualizing depression through this narrow perspective, individuals are prompted to think in a deterministic way about biological processes.

Patient preference could be one mechanism that accounts for treatment selection. Consumer advertisements of antidepressants in the United States do not require preapproval from the Food and Drug Administration (FDA). Instead, the FDA often first views these advertisements around the same time as the public. If they find the advertisement violates the law, the drug company receives a letter asking for the advertisements to be stopped. However, by that point, the impact which these advertisements have on the public has already been made. Kravitz et al. (2005) assessed the influence of patient requests for the antidepressant Paxil, which was advertised directly to consumers, and revealed that participants (who were paid actors pretending to have depression) were frequently prescribed the drug by their physician if they inquired specifically about Paxil.

Understanding where individuals are being exposed to beliefs about depression can help give insight to professionals on ways to approach conversations about treatment options. Sources of information such as the classroom (Schroder et al., 2022) and media (France et al., 2007) were found to be prevalent among college students’ exposure to the chemical imbalance theory. In addition to information sources, addressing professionals’ role is also vital in understanding biases towards treatment. Acker and Warner (2020) considered the importance of social workers as they help navigate their clients’ decisions for treatment; they found that most social workers reported using a chemical imbalance explanation for depression. Furthermore, Lebowitz and Ahn (2014) found that the biological explanation evoked significantly less empathy from clinicians than psychosocial explanations of patients’ symptoms. Despite these disadvantages, the biomedical model of mental disorders remains, even though there is insufficient evidence to suggest that mental disorders are caused by chemical imbalances or that medication works by correcting

imbalances in the brain (Deacon, 2013).

The present study serves to replicate and extend the existing literature by examining the relationship between biogenetic endorsements of depression and familiarity with the chemical imbalance theory. Furthermore, by considering how biological attributes for depression are associated with self­stigma and how these attitudes guide individuals’ preference for treatment as empirical research in this area remains limited, indicating the need to extend and clarify these findings.

It was hypothesized that endorsements of biogenetic causes of depression will be higher among those who prefer medication treatment than among those who prefer therapy. Furthermore, it was expected that greater endorsements of biogenetic causes will correlate with more prognostic pessimism and self­stigma.

Method

Participants

A total of 69 undergraduate students enrolled at a small

TABLE 1

Demographic Characteristics of Participants

New England public university completed the online survey for partial course credit in Psychology courses. Of these, 54 participants (82%) identified as female and 46 (78%) identified as White. The average age of the sample was 19 years. See Table 1 for additional demographic data.

Measures

Reasons for Depression

The Reasons for Depression Questionnaire (RFD; Addis et al., 1995) consists of 48 items to assess beliefs about the causes of depression. Participants were asked to rate the following statements as to the likelihood they caused depression. Each item was rated on a 4­point scale from 1 (definitely not a reason) to 4 (definitely a reason). The full RFD scale consists of nine subscales (i.e., Characterological, Achievement, Interpersonal Conflict, Intimacy, Existential, Childhood, Physical, Relationship, and Biological). A sample item from the characterological subscale is, “I’ve always been this way.” Thwaites et al. (2004) reported internal consistency estimates of α = .73 and α = .94. In the present study, reliability was α = .89 for the biological subscale.

Chemical Imbalance Theory

The chemical imbalance theory questionnaire consisted of two items to measure familiarity with the chemical imbalance theory of depression. Participants were given the following prompt: “Have you heard of the chemical imbalance theory of depression?” Participants chose between “Yes” and “No.” Those who responded yes were provided with a list of five sources of information and asked to: “Please indicate the extent to which you have learned about the chemical imbalance theory of depression from each source” on a 5­point scale from 0 (to no extent) to 4 (to a very great extent). The five sources to choose from were: school/classroom (e.g., abnormal psychology class), internet/media (e.g., Google, social media, books), healthcare providers (e.g., psychiatrist, counselor), personal experience (e.g., in my personal life), and not sure. This measure was developed for this study, and it was based on previous research by Schroder et al. (2022).

Treatment Preference

A single item was used to measure treatment preference. Participants were given the following prompt: “If you were to struggle with mental health problems (e.g., anxiety or depression) and had a choice between no treatment, medication, individual therapy, or combined medication and therapy, to help you with your mental health problems, which one would you choose?” This measure was developed for this study, and it was based on previous research by Schroder et al. (2022).

Prognostic Pessimism and Blame

The Prognostic Pessimism and Blame Scale was used to measure prognostic pessimism and stigma (PPBS; Deacon & Baird, 2009). The measure consists of 13 items rated on a 5 ­ point scale from 1 ( not at all agree) to 5 (extremely agree). The full PPBS consists of four subscales (Credibility, Stigma, Prognosis, and Treatment). For the current study, only the prognosis and stigma subscales were used, and each consisted of four items. A sample item from the stigma subscale is, “To what extent would you feel personally responsible for having developed depression?” Deacon and Baird (2009) reported internal consistency estimates for the stigma subscale as α = .83 for the chemical imbalance explanation and α = .79 for the biopsychosocial explanation. For the prognosis subscale, internal consistency was α = .73 for the chemical imbalance explanation and α = .68 biopsychosocial explanation. In the present study, reliability for the stigma subscale was α = .75 and α = .34 for the prognosis subscale.

Procedure

Prior to data collection, all procedures were approved by our institutional review board for human subjects research. When participants entered the online survey, they were presented with an informed consent screen and clicked “next” on the consent page before moving on to questionnaires. The questionnaires were administered in this order: RFD questionnaire, chemical imbalance theory questionnaire, treatment preference questionnaire, and lastly the PPBS. Upon completing the final measure, participants filled out demographic data and were presented with a debriefing screen explaining the study’s purpose.

Results

For all statistical analyses an alpha level of .05 was adopted. To test Hypothesis 1, that endorsements of biogenetic causes of depression would be higher among students who prefer medication over therapy, a one­way ANOVA was conducted comparing the means of four groups: no treatment (n = 3), medication (n = 4), therapy (n = 31), and combined medication and therapy (n = 27). A significant difference was found among treatment preferences, F(3, 61) = 4.19, p = .010, η2 = .17. Tukey’s LSD was used to determine the nature of the differences between the treatment preferences (d = ­1.85, p = .016). This analysis showed that students who opted for no treatment had lower biogenetic endorsements of depression (M = 1.08, SD = 0.14) compared to those who chose medication (M = 2.94, SD = 1.30) and combined treatments (medication and therapy; M = 2.66, SD = 0.92; see Figure 1). Additionally, students who opted for

therapy had lower biogenetic endorsements of depression (M = 2.02, SD = 1.02) compared to those who chose combined treatments. There was no significant difference between students who chose medication compared to those who chose therapy.

To test Hypothesis 2, that higher endorsements of biogenetic causes of depression will correlate with prognostic pessimism and stigmatizing attitudes, a Spearman’s rho correlation was conducted. A moderate negative correlation was found , ρ = ­ .41, n = 65, p < .001, ρ2 = .17, indicating a significant correlation between biogenetic endorsements and prognostic pessimism. However, stigma was not correlated with biogenetic endorsements, ρ = .02, n = 65, p = .86. Participants in our study scored an average of 2.31 on a 4­point scale on biogenetic endorsements, and 3.63 on a 5­point scale for prognostic pessimism, suggesting a moderately high endorsement of these beliefs. For stigma, participants scored an average of 2.22 on a 5­point scale, suggesting a low endorsement of these items.

To determine whether endorsements of biogenetic causes differed between those who were familiar with the chemical imbalance theory ( n = 30) and those who were not (n = 35), an independent samples t test was conducted. Endorsement of biogenetic causes of depression was significantly higher among students familiar with the chemical imbalance theory (M = 2.63, SD = 1.08), than students who were not ( M = 2.01, SD = 0.95), t(63) = 2.50, p = .01, with a large effect size (Cohen’s d = 1.01).

Discussion

The aim of the study was to examine whether biogenetic endorsements of depression were associated with stigma, prognostic pessimism, and treatment preferences among college students. Additionally, it examined the difference in endorsement of biogenetic causes of depression between students who were familiar with chemical imbalance theory and those who were not.

Hypothesis 1 was not supported, as there was no significant difference in biogenetic endorsement of depression among those who preferred medication and those who preferred therapy. However, there was a significant difference among other treatment preferences. Students who opted for no treatment had lower biogenetic endorsements of depression compared to those who chose medication and combined treatments (medication and therapy). Additionally, students who opted for therapy have lower biogenetic endorsements of depression compared to those who chose combined treatments. These findings are consistent with previous research (Kemp et al., 2014; Watson & Beshai, 2020), which found biogenetic beliefs increase preferences

for pharmacological treatments such as medication. For beliefs about the causes of depression, participants provided mean ratings indicating moderate to high endorsement (with a range 2.10 to 2.90 on a 4­point scale) of factors assessed by the instrument we used.

Hypothesis 2 was partially supported; biogenetic endorsements were negatively correlated with prognostic pessimism whereas it was not correlated with stigma. In line with previous findings, biological explanations can reinforce the perception that mental illnesses are unfixable, promoting a sense of permanence and lack of control over the disorder leading to prognostic pessimism (Haslam, 2011; Kvaale et al., 2013). The failure to find a correlation between stigma and biogenetic endorsements could suggest that stigma may be influenced by many other factors beyond just causal beliefs. Stigma itself is a complicated construct which may have multiple components that are important as they may influence choice of interventions. For instance, there may be self­stigma (e.g., Deacon & Baird, 2009) experienced by the individual seeking treatment. Alternatively, external stigma may arise (e.g., Goldstein & Rosselli, 2003), wherein individuals without mental illness develop negative perceptions of those experiencing symptoms or opting for specific interventions. Future research should clarify the relations between these various aspects of stigma and the selection of interventions.

Consistent with prior research (e.g., Schroder et al., 2022) students familiar with the chemical imbalance theory showed significantly higher levels of biogenetic endorsements of depression than those who were not.

Furthermore, the most common source of information through which students reported learning about the chemical imbalance theory was personal experience ( M = 2.87). This was followed closely by school/ classroom (M = 2.80), internet/media (M = 2.72), and healthcare providers (M = 2.45). These findings contribute to the ongoing discourse on the relationship between biogenetic explanations of depression and reflect how such explanations, despite their scientific limitations, have become embedded in the public’s understanding of mental health. The prominence of personal experience likely reflects students’ direct or indirect exposure with mental health challenges, such as their own experiences with depression or those experienced by people close to them. These encounters may have shaped their beliefs through casual discussions or personal interpretation of their symptoms and treatment. Because these experiences are often poignant, they may have a lasting impact on individuals’ understanding of depression, even in the absence of professional input.

Understanding where students most often encounter these beliefs can provide valuable insight for mental health professionals when engaging in conversations about treatment. Prior literature has highlighted the classroom (e.g., Schroder et al., 2022) and media (e.g., France et al., 2007) as common source of exposure, and these were indeed the second and third most reported sources in the current study. The relatively lower mean for healthcare providers suggests that clinical encounters may not be the primary source of students’ beliefs about depression, which has implications for how such beliefs are formed and reinforced outside of these contexts. The variation in how students encounter information about depression may also help explain why some individuals are more likely to endorse the chemical imbalance theory than others. For instance, a student who has a friend with depression might hear firsthand accounts of their struggles, leading to a deeper emotional connection. Conversely, a student who learns about mental health in a psychology class might encounter various theories and studies that challenge simplistic views of mental health. These patterns highlight the importance of considering both the type and source of mental health information when seeking to understand public beliefs and their implications for treatment choices.

Some strengths of the study are its inclusion of a sample that is not explicitly limited to individuals with a formal diagnosis of depression or history of depression. This broader sampling approach enhances the generalizability of the findings by allowing for a more inclusive examination of how biogenetic beliefs are endorsed and operate within the general population. Such a design helps to capture a wider range of beliefs,

providing insight into how these conceptualizations of mental health are formed outside of clinical settings. Additionally, the use of the RFD as a preexisting measure helps strengthen the methodology of study. Established measure like the RFD contribute to the reliability and validity of the results. They ensure that the constructs being measured are operationalized consistently and accurately. This minimizes measurement error and enhancing replicability. Moreover, given the limited amount of research exploring the relationship between biogenetic beliefs, treatment preferences, stigma, prognostic pessimism, and sources of information, this study fills an important gap in the literature. This helps establish stronger evidence base and informs the development of educational and clinical interventions aimed at improving mental health literacy and tailoring treatment approaches. However, limitations should be noted. Self­report measures may be subject to social desirability, particularly regarding attitudes towards stigma and treatment preferences. However, the use of anonymous surveys likely mitigated some of these effects. Lastly, Cronbach’s alpha for the prognosis scale was low, indicating potential reliability concerns for this measure. Low reliability is problematic because it suggests that the items on the scale may not be consistently measuring the same construct, which can compromise the validity of the any conclusions drawn from the scale. All items on the subscale had low alpha values, and removing individual items would have further reduced reliability. A possible explanation for the low alpha could be due to possible participant confusion about the meaning of items and the limited number of items. Additionally, this could reflect that prognostic pessimism is more complex and multifaceted construct than the measure represents. Future research should consider refining the prognosis measure or using an alternative measure to assess prognostic pessimism more accurately and should evaluate samples with different levels to assess how treatment decisions are influenced.

These findings have several practical implications such as the need for a more refined framework in which education about mental health presents depression as a multicausal disorder, addressing biological, psychological, and social factors. Such an approach could help mitigate the deterministic thinking often associated with biogenetics narratives and encourage individuals to consider a wider range of treatment options. Given the correlation between biogenetic beliefs and pessimistic views on recovery, mental health providers should consider integrating biopsychological perspectives when discussing treatment options with patients. Future research should explore interventions to address deterministic thinking. Because biogenetic beliefs

were correlated with prognostic pessimism, future studies should examine how educational intervention that emphasize biopsychosocial models might reduce pessimistic attitudes about recovery.

References

Acker, J. K., & Warner, L. A. (2020). Let’s talk about depression: Social workers’ use of chemical imbalance explanations. Social Work in Mental Health, 18(3), 237–256. https://doi.org/10.1080/15332985.2019.1700872

Addis, M. E., Truax, P., & Jacobson, N. S. (1995). Why do people think they are depressed?: The Reasons For Depression Questionnaire. Psychotherapy: Theory, Research, Practice, Training, 32(3), 476–483. https://doi.org/10.1037/0033-3204.32.3.476

An, S., Jin, H. S., & Brown, J. D. (2009). Direct-to-consumer antidepressant ads and young adults’ beliefs about depression. Health Marketing Quarterly, 26(4), 259–278. https://doi.org/10.1080/07359680903303981

Ang, B., Horowitz, M., & Moncrieff, J. (2022). Is the chemical imbalance an ‘urban legend’? An exploration of the status of the serotonin theory of depression in the scientific literature. SSM - Mental Health, 2, 100098. https://doi.org/10.1016/j.ssmmh.2022.100098

Berent, I., & Platt, M. (2021). Essentialist biases toward psychiatric disorders: brain disorders are presumed innate. Cognitive science, 45(4), e12970. https://doi.org/10.1111/cogs.12970

Coppen, A. (1967). The biochemistry of affective disorders. The British Journal of Psychiatry, 113(504), 1237–1264. https://doi.org/10.1192/bjp.113.504.1237

Cuijpers, P., Miguel, C., Harrer, M., Plessen, C. Y., Ciharova, M., Ebert, D., & Karyotaki, E. (2023). Cognitive behavior therapy vs. control conditions, other psychotherapies, pharmacotherapies and combined treatment for depression: a comprehensive meta‐analysis including 409 trials with 52,702 patients. World Psychiatry, 22(1), 105–115. https://doi.org/10.1002/wps.21069

Cuijpers, P., Sijbrandij, M., Koole, S. L., Andersson, G., Beekman, A. T., & Reynolds, C. F. (2013). The efficacy of psychotherapy and pharmacotherapy in treating depressive and anxiety disorders: A meta‐analysis of direct comparisons. World Psychiatry, 12(2), 137–148. https://doi.org/10.1002/wps.20038

Dar-Nimrod, I., & Heine, S. J. (2011). Genetic essentialism: On the deceptive determinism of DNA. Psychological Bulletin, 137(5), 800–818. https://doi.org/10.1037/a0021860

Davis, J. E. (2021). ‘The explanation you have been looking for’: Neurobiology as promise and hermeneutic closure. Culture, Medicine and Psychiatry, 46(1), 76–100. https://doi.org/10.1007/s11013-021-09737-2

Deacon, B. J., & Baird, G. L. (2009). The chemical imbalance explanation of depression: Reducing blame at what cost? Journal of Social and Clinical Psychology, 28(4), 415–435. https://doi.org/10.1521/jscp.2009.28.4.415

Deacon, B. J. (2013). The biomedical model of mental disorder: A critical analysis of its validity, utility, and effects on psychotherapy research. Clinical Psychology Review, 33(7), 846–861. https://doi.org/10.1016/j.cpr.2012.09.007

DeRubeis, R. J., Siegle, G. J., & Hollon, S. D. (2008). Cognitive therapy versus medication for depression: Treatment outcomes and neural mechanisms. Nature Reviews. Neuroscience, 9(10), 788–796. https://doi.org/10.1038/nrn2345

France, C. M., Lysaker, P. H., & Robinson, R. P. (2007). The “chemical imbalance” explanation for depression: Origins, lay endorsement, and clinical implications. Professional Psychology: Research and Practice, 38(4), 411–420. https://doi.org/10.1037/0735-7028.38.4.411

Goldstein, B., & Rosselli, F. (2003). Etiological paradigms of depression: The relationship between perceived causes, empowerment, treatment preferences, and stigma. Journal of Mental Health, 12(6), 551–563. https://doi.org/10.1080/09638230310001627919

Haslam, N. (2011). Genetic essentialism, neuroessentialism, and stigma: Commentary on Dar-Nimrod and Heine (2011). Psychological Bulletin, 137(5), 819–824. https://doi.org/10.1037/a0022386

Heninger, G. R., Delgado, P. L., & Charney, D. S. (1996). The revised monoamine theory of depression: A modulatory role for monoamines, based on new findings from monoamine depletion experiments in humans. Pharmacopsychiatry, 29(01), 2–11. https://doi.org/10.1055/s-2007-979535

Illes, J., Lombera, S., Rosenberg, J., & Arnow, B. (2008). In the mind’s eye: Provider and patient attitudes on functional brain imaging. Journal of Psychiatric Research, 43(2), 107–114. https://doi.org/10.1016/j.jpsychires.2008.02.008

Imel, Z. E., Malterer, M. B., McKay, K., & Wampold, B. E. (2008). A meta-analysis of psychotherapy and medication in unipolar depression and dysthymia. Journal of Affective Disorders, 110(3), 197–206. https://doi.org/10.1016/j.jad.2008.03.018

Iselin, M. G., & Addis, M. E. (2003). Effects of etiology on perceived helpfulness of treatments for depression. Cognitive Therapy and Research, 27, 205–222. https://doi.org/10.1023/A:1023513310243

Jakobsen, J. C., Gluud, C., & Kirsch, I. (2019). Should antidepressants be used for major depressive disorder? BMJ Evidence-Based Medicine, 25(4), 130. https://doi.org/10.1136/bmjebm-2019-111238

Kemp, J. J., Lickel, J. J., & Deacon, B. J. (2014). Effects of a chemical imbalance causal explanation on individuals’ perceptions of their depressive symptoms. Behaviour Research and Therapy, 56, 47–52. https://doi.org/10.1016/j.brat.2014.02.009

Kvaale, E. P., Haslam, N., & Gottdiener, W. H. (2013). The ‘side effects’ of medicalization: A meta-analytic review of how biogenetic explanations affect stigma. Clinical Psychology Review, 33(6), 782–794. https://doi.org/10.1016/j.cpr.2013.06.002

Kravitz, R. L., Epstein, R. M., Feldman, M. D., Franz, C. E., Azari, R., Wilkes, M. S., Hinton, L., & Franks, P. (2005). Influence of patients’ requests for direct-toconsumer advertised antidepressants. JAMA, 293(16), 1995. https://doi.org/10.1001/jama.293.16.1995

Lacasse, J. R., & Leo, J. (2005). Serotonin and depression: A disconnect between the advertisements and scientific literature. PLOS Medicine, 2(12), 392. https://doi.org/10.1371/journal.pmed.0020392

Lebowitz, M. S., & Ahn, W. (2014). Effects of biological explanations for mental disorders on clinicians’ empathy. Proceedings of the National Academy of Sciences, 111(50), 17786–17790. https://doi.org/10.1073/pnas.1414058111

Lebowitz, M. S., Ahn, W., & Nolen-Hoeksema, S. (2013). Fixable or fate? Perceptions of the biology of depression. Journal of Consulting and Clinical Psychology, 81(3), 518–527. https://doi.org/10.1037/a0031730

Lebowitz, M. S., & Appelbaum, P. S. (2019). Biomedical explanations of psychopathology and their implications for attitudes and beliefs about mental disorders. Annual Review of Clinical Psychology, 15(1), 555–577. https://doi.org/10.1146/annurev-clinpsy-050718-095416

Lebowitz, M. S., Dolev-Amit, T., & Zilcha-Mano, S. (2021). Relationships of biomedical beliefs about depression to treatment-related expectancies in a treatment-seeking sample. Psychotherapy, 58(3), 366–371. https://doi.org/10.1037/pst0000320

Leo, J., & Lacasse, J. R. (2008). The media and the chemical imbalance theory of depression. Society, 45(1), 35–45. https://doi.org/10.1007/s12115-007-9047-3

Munkholm, K., Paludan-Müller, A. S., & Boesen, K. (2019). Considering the methodological limitations in the evidence base of antidepressants for depression: A reanalysis of a network meta-analysis. BMJ Open, 9(6), e024886. https://doi.org/10.1136/bmjopen-2018-024886

Pescosolido, B. A., Martin, J. K., Long, J. S., Medina, T. R., Phelan, J. C., & Link, B. G. (2010). “A disease like any other”? A decade of change in public reactions to schizophrenia, depression, and alcohol dependence. American Journal of Psychiatry, 167(11), 1321–1330. https://doi.org/10.1176/appi.ajp.2010.09121743

Phelan, J. C. (2005). Geneticization of deviant behavior and Consequences for stigma: The case of mental illness. Journal of Health and Social Behavior, 46(4), 307–322. https://doi.org/10.1177/002214650504600401

Phelan, J. C., Yang, L. H., & Cruz-Rojas, R. (2006). Effects of attributing serious mental illnesses to genetic causes on orientations to treatment. Psychiatric Services, 57(3), 382–387. https://doi.org/10.1176/appi.ps.57.3.382

Salem, T., Winer, E. S., Jordan, D. G., & Dorr, M. M. (2019). Doubting the diagnosis but seeking a talking cure: An experimental investigation of causal explanations for depression and willingness to accept treatment. Cognitive Therapy and Research, 43(6), 971–985. https://doi.org/10.1007/s10608-019-10027-w

Schildkraut, J. J. (1965). The catecholamine hypothesis of affective disorders: A review of supporting evidence. American Journal of Psychiatry, 122(5), 509–522. https://doi.org/10.1176/ajp.122.5.509

Schomerus, G., Schwahn, C., Holzinger, A., Corrigan, P. W., Grabe, H. J., Carta, M. G., & Angermeyer, M. C. (2012). Evolution of public attitudes about mental illness: A systematic review and meta‐analysis. Acta Psychiatrica Scandinavica, 125(6), 440–452. https://doi.org/10.1111/j.1600-0447.2012.01826.x

Schroder, H. S., Block, S. R., & Moser, J. S. (2022). Chemical imbalance and etiological beliefs about depression among college students. Journal of American College Health, 1–8. https://doi.org/10.1080/07448481.2022.2098037

Schroder, H. S., Duda, J. M., Christensen, K., Beard, C., & Björgvinsson, T. (2020). Stressors and chemical imbalances: Beliefs about the causes of depression in an acute psychiatric treatment sample. Journal of Affective Disorders, 276, 537–545 https://doi.org/10.1016/j.jad.2020.07.061

Thwaites, R., Dagnan, D., Huey, D., & Addis, M. E. (2004). The Reasons for Depression Questionnaire (RFD): UK Standardization for clinical and non‐clinical populations. Psychology and Psychotherapy Theory Research and Practice, 77(3), 363–374. https://doi.org/10.1348/1476083041839367

Watson, L. M., & Beshai, S. (2020). Causal explanations of depression on perceptions of and likelihood to choose cognitive behavioural therapy and antidepressant medications as depression treatments. Psychology and Psychotherapy Theory Research and Practice, 94(2), 201–216. https://doi.org/10.1111/papt.12300

Whitaker, R. (2010). Anatomy of an epidemic: Magic bullets, psychiatric drugs, and the astonishing rise of mental illness in America. Crown Publishers/ Random House.

White, R. (2013). The globalisation of mental illness. The Psychologist, 26(3), 182–185. https://www.bps.org.uk/psychologist/globalisation-mental-illness

Author Note.

Saoirse E. Ward https://orcid.org/0009­0005­8254­3125

James W. Diller https://orcid.org/0000­0002­4517­4405

We have no known conflict of interest to disclose. The authors confirm that the research presented in this article met the ethical guidelines, including adherence to the legal requirements, of United States of America and received Eastern Connecticut State University Institutional Review Board approval. This study was supported by the Jean Thoresen ECSU­AAUP Scholarship.

Saoirse E. Ward played a lead role in conceptualization, research design, data collection, data analysis and interpretation, and substantive original writing. James W. Diller played a lead role in supervision and editorial assistance, and a supporting role in data collection, data analysis, and interpretation.

Correspondence concerning this article should be addressed to James W. Diller, Department of Psychological Science, Eastern Connecticut State University, 83 Windham St, Willimantic, CT 06226, United States. Email: dillerj@easternct.edu

The Recollections of Childhood Food Parenting and Adult Mindful Eating

Abigail Brighton, Adelyn Sherrard*, and Cin Cin Tan** Department of Psychology, The University of Toledo

ABSTRACT. Mindful eating is associated with reduced maladaptive eating patterns (Warren et al., 2017). However, there has been little research on individual differences in mindful eating, particularly in relation to early childhood factors. One possible factor is early childhood feeding experiences such as food parenting. Food parenting refers to the strategies a parent uses when discussing and engaging with food while feeding their children. Food parenting is associated with eating outcomes during childhood and adulthood (Liu et al., 2023). Yet, less is known about whether early childhood food parenting is associated with adults’ mindful eating. Thus, this study examined the relationship between retrospective food parenting and adults’ mindful eating. Participants completed the Mindful Eating Questionnaire (Framson et al., 2009) and the Comprehensive Feeding Practices Questionnaire (Musher­Eizenman & Holub, 2007). Pearson’s correlation analyses revealed that recalled coercive control food parenting, r(150) = ­ .21, p = .008, was negatively associated with mindful eating, but recalled autonomy ­ promoting, r (150) = .29, p < .001, and structure ­ based food parenting, r(150) = .20, p = .013, were positively associated with mindful eating. Linear regression analysis, F(4,124) = 3.71, p = .007, R2 = .11, revealed that high autonomy­promoting food parenting was associated with higher adult mindful eating in women, B = .12, t(126) = 3.66, p < .001, but not for men, B = .00, t(126) = 0.07, p = .946. Sex assigned at birth did not moderate the association between recalled coercive control, F(4,124) = 1.63, p = .172, R2 = .05, or structure­based food parenting, F(4,124) = .95, p = .449, R2 = .03, and mindful eating. Findings from the current study show that early childhood feeding experiences may explain adulthood eating patterns, particularly for women.

Keywords: mindful eating, food parenting, adult eating behaviors, retrospective

Emotional eating is defined as when one eats in increased amounts due to negative emotions and includes difficulty to maintain or lose weight (Bruch, 1973; Frayn & Knäuper, 2017). Emotional eating is associated with poorer psychological health, eating disorder symptoms, and difficulty regulating emotions (Braden et al., 2018). Binge eating is defined as eating a significantly larger than an average amount of food in a short period of time, while feeling a lack of control (Katterman et al., 2014; Marcus et al., 2005). Individuals who have binge eating disorders have reported an earlier onset of obesity, a history of more severe dieting, and greater weight fluctuations (Katterman et al., 2014; Marcus et al., 2005). Similar to emotional eating, individuals who are overweight and have binge eating disorder also report

lower self­esteem, feeling less in control of their eating, having more fear of weight gain and preoccupation with food, and being generally less satisfied with their bodies compared to individuals who are overweight without binge eating disorder (Didie & Fitzgibbon, 2005; Marcus et al., 2005; Yanovski et al., 1993).

Mindfulness

It is possible that mindfulness may play a role in eating behaviors. Mindfulness is recognized as a positive intervention for various psychological (e.g., psychosis, anxiety) and physical health problems, including chronic pain (Abba et al., 2008; Mars & Abbey, 2010; Sauer et al., 2012). Mindfulness is the ability to experience life in the present moment without judgment, often

through meditation practices (Sauer et al., 2012). Greater mindfulness is associated with less eating disorder psychopathology, particularly in the areas of binge eating, emotional eating, and body dissatisfaction (Sala et al., 2020). In addition, mindfulness is a mediating factor between early childhood experiences and later disordered eating. For example, research has shown that adverse childhood experiences (ACEs) are related to disordered eating among young adults (Hughes et al., 2017; Royer et al., 2023). However, the associations between ACEs and disordered eating are mediated by mindfulness. Specifically, ACEs are associated with lower mindfulness, which subsequently explains greater disordered eating (Royer et al., 2023). This finding suggests that it is important to examine the role of mindfulness. Furthermore, studies support the effectiveness of mindfulness­based interventions in changing obesity­related eating behaviors (O’Reilly et al., 2014).

Recent research has focused on mindfulness in the context of eating, which is known as mindful eating (Kristeller & Jordan, 2018). Mindful eating is defined as making conscious food choices, developing healthy hunger cues, and responding to those cues with healthy foods; it also involves being in the present moment, awareness of sensations, and being in a meditative­like state (Warren et al., 2017). Research has shown that mindful eating is negatively associated with weight and disordered eating (Bennett & Latner, 2022; Carrière et al., 2017; Forman et al., 2016; Warren et al., 2017). Specifically, developing habits of mindful eating can help to change negative eating patterns, especially emotional eating and binge eating (Warren et al., 2017). Research has shown that disinhibition, which is the ability to stop eating when full, is negatively related to loss of control while eating (Bennett & Latner, 2022). This suggests that individuals who are not mindful eaters tend to engage in more overeating. Furthermore, findings have also shown that individuals can be trained to become mindful eaters. For example, after training individuals to become more aware of the sensory, emotional, and cognitive processes that influence their eating, they experience a significant reduction in eating salty snacks over the following week (Forman et al., 2016). Mindfulness­based interventions are also effective in reducing weight and improving obesity­related eating behaviors in those with overweight or obese statuses (Carrière et al., 2017). Although there is research supporting the idea that adults can alter their negative eating behaviors by practicing mindfulness or mindful eating, there is limited research that examines the individual differences in mindful eating. One possible factor is early childhood feeding experiences (i.e., food parenting).

Food Parenting

Food parenting is defined as the strategies a parent uses when talking about and engaging with food (Vaughn et al., 2016). There are three general constructs of food parenting practices: coercive control, autonomy­promoting, and structure­based (Vaughn et al., 2016). Coercive control is a food parenting practice that attempts to impose the parents’ thoughts and beliefs on their child; it involves pressuring a child to eat, restricting the child from eating for weight or health concerns, and using food as a reward (Vaughn et al., 2016). Responsive food parenting includes autonomy ­ promoting and structure­based food parenting. Autonomy­promoting food parenting is a responsive and supportive food parenting style that involves encouraging the child to participate in meal planning, as well as educating and praising the child. Structure­based is a responsive and consistent parenting style that involves set mealtimes, monitoring what the child eats, and modelling healthy eating behaviors (Małachowska & Jeżewska­Zychowicz, 2021; Vaughn et al., 2016). It has been found that food parenting is associated with eating outcomes during childhood and adulthood (Leuba et al., 2022; Liu et al., 2023; Małachowska & Jeżewska­Zychowicz, 2021; Tan et al., 2016).

Different styles of food parenting (i.e., coercive and responsive) are associated with different child outcomes. Tan and colleagues (2016) found that recalled restriction for weight (i.e., restricting certain foods from children over concerns about their weight) during childhood is associated with more food preoccupation, which is then related to emotional eating in adulthood. For men, increased levels of child control (i.e., letting the child control their eating behaviors) are associated with lower levels of emotional eating. Another study found that higher rates of pressuring children to eat and using food as a reward with child control was associated with higher intake of sweet and salty snacks. Also, using food as a reward may result in emotional eating in adulthood (Małachowska & Jeżewska­Zychowicz, 2021). Taken together, these studies suggest that coercive control food parenting is associated with less diverse diets and maladaptive eating behaviors (e.g., emotional eating).

On the other hand, responsive food parenting, including setting a healthy eating environment, modeling healthy eating behaviors, and allowing the child to participate in food planning and buying, is associated with higher consumption of fresh fruits and vegetables (Małachowska & Jeżewska­Zychowicz, 2021). Previous research has shown that parents’ modelling of healthy eating is correlated with children’s dietary intake and fruit and vegetable preference, and it is recommended that parents model healthy eating behaviors to encourage

healthy eating in their children (Draxten et al., 2014). One study found that adults who model fruit eating as a snack are more likely to have children who meet the recommended fruit and vegetable intake (Draxten et al., 2014). Cross­sectional studies have suggested that positive structure during mealtimes supports adolescent development and maintenance of healthy eating, and another study found that positive food parenting, such as monitoring (i.e., parents allow the consumption but keep track of the intake of unhealthy foods), is associated with lower child BMI (Balantekin et al., 2020). Authoritative parents, who use structure, monitoring, acceptance, warmth, and involvement, are more likely to use mindful feeding, which is associated with positive eating outcomes, such as less food fussiness and less emotional overeating (Goodman et al., 2020). Prior research has suggested that responsive parenting is associated with more adaptive adult eating behaviors. Thus, it is possible that responsive food parenting may promote mindful eating.

Although there has been no study focusing on food parenting and adult mindful eating, there has been research on the association between food parenting and intuitive eating. Intuitive eating is a practice similar to mindful eating, but without meditation. Adults who remember their parents using healthy eating guidance and monitoring practices are predisposed to be intuitive eaters in adulthood (Małachowska & Jeżewska­Zychowicz, 2023). Given that the evidence of intuitive eating’s benefits in altering eating behaviors is limited, but intervening on mindful eating has shown to reduce binge eating, it is important to examine factors associated with mindful eating (Warren et al., 2017). Prior research has revealed sex differences in the associations between recalled food parenting and intuitive eating. Małachowska and JeżewskaZychowicz (2023) found that women who experience high restriction (i.e., parents prevented them from consuming certain foods) rely less on hunger and satiety cues, give themselves less permission to eat, and have increased emotional eating. Also, this research found that women with higher recall of monitoring are more likely to choose healthy foods to eat (Małachowska & Jeżewska­Zychowicz, 2023). Men who experience low levels of restriction are more likely to give themselves more permission to eat and consume unhealthy foods. A similar study found that greater recall of parental concern of child’s overeating is associated with lower intuitive eating in women, and both greater recall of parental concern of child’s overeating and pressure to eat are associated with greater disordered eating in men (Liu et al., 2023). Taken together, these research studies suggest that it is likely that the associations between

food parenting during childhood and mindful eating may differ by sex.

The Current Study

and Hypotheses

This study aimed to examine the association between adults’ recollection of what food parenting strategies their parents used with them (i.e., coercive control, structure ­ based, or autonomy ­ promoting) during childhood and their current mindful eating. First, we examined whether recalled coercive control food parenting was related to mindful eating in adulthood. Based on previous research, we hypothesized that more recalled coercive control food parenting would be related to lower mindful eating (Leuba et al., 2022; Liu et al., 2023; Małachowska & Jeżewska­Zychowicz, 2021; Małachowska & Jeżewska­Zychowicz, 2023; Tan et al., 2016). Second, we examined whether recalled responsive food parenting related to mindful eating in adulthood. Based on previous research, we hypothesized that recollection of more responsive food parenting, specifically autonomy­promoting and structure­based, would be related to more mindful eating (Berge et al. 2020; Lebua et al., 2022; Małachowska & JeżewskaZychowicz, 2021; Małachowska & Jeżewska­Zychowicz, 2023; Musher­Eizenman et al., 2019). Less research has examined whether associations between recalled food parenting and eating behaviors vary by participants’ sex. Therefore, this study explored this research question. As research in this area is limited, no specific hypothesis was proposed.

Method

Participants and Procedure

Participants were recruited in September 2024–October 2024 using Sona, an online platform for university researchers to recruit and administer studies to adult participants enrolled in an introductory psychology course. Adults (N = 152; M age = 20.74 years, SD = 3.15) completed an online survey. In this study, participants reported their sex assigned at birth. Most participants were assigned female at birth (64.5%) and 35.5% were assigned male at birth. Most participants identified as White (68.4%), 8.6% as Black/African American, 13.8% as Asian American, 3.9% as Biracial or multiracial, 3.9% as other, and 1.3% declined to answer. BMI (kg/m2) was collected as a standard descriptive measure and calculated using participants’ self­reported height and weight. The sample had an average body mass index of 25.29 (SD = 5.62). Upon approval of the procedure from the university’s institutional review board, students interested in participating followed the link from Sona to the survey on QuestionPro. Participants were presented with informed consent and indicated consent by clicking

“next” on the survey. Responses were monitored and screened depending on survey completion, overall completion time, and incorrect responses to control questions. Participants that met the inclusion and exclusion criteria received one Sona credit upon completion and submission of the survey.

Measures

Food Parenting

The Comprehensive Feeding Practices Questionnaire (Musher­Eizenman & Holub, 2007) was adapted for adult participants to report their childhood food parenting experiences rather than for parents to report their current food parenting practices. The questions were revised from present tense to past tense and from first person (parent perspective) to third person (child perspective) to reflect that adult participants were reporting on their parents’ past practices. For example, the original item “If my child says, ‘I’m not hungry,’ I try to get him/her to eat anyway” was adapted to “If I said, ‘I’m not hungry,’ my parent tried to get me to eat anyway.” Coercive control food parenting (22 items; α = .84) was averaged across the Pressure to Eat (4 items), Food as a Reward (3 items), Restriction for Weight Control (8 items), Restriction for Health (4 items), and Emotion Regulation (3 item) subscales. Autonomy­promoting food parenting (15 items; α = .78) was averaged across the Child Control (5 items), Involvement (3 items), Modeling (4 items), and Teaching About Nutrition (3 items) subscales. Structure­based food parenting (12 items; α = .82) was averaged across the Encourage Balance and Variety (4 items), Monitoring (4 items), and Environment (4 items) subscales. Participants were asked to reflect on their experiences of food parenting from when they were between 5 and 10 years old and indicate their experiences on a 5­point scale from 1 ( never/disagree ) to 5 ( always/agree ). Higher scores indicate higher levels of the specified food parenting (i.e. coercive control, autonomy­promoting, structurebased), where lower scores indicate lower levels of the food parenting.

Mindful Eating

The Mindful Eating Questionnaire (Framson et al., 2009) is a questionnaire that assesses mindful eating. It consists of the following subscales: Awareness (7 items), Distraction (3 items), Disinhibition (8 items), Emotional Response (4 items), and External Cues (6 items). Total mindful eating (28 items; α = .67) was assessed by reverse scoring the Emotional Response and Distraction subscales and 5 items in the Disinhibition subscale, then averaging the scores across all subscales. Participants were asked to complete the questionnaire and indicate

their eating behaviors on a 4­point scale ranging from 1 (usually/always) to 4 (never/rarely). Higher mean scores indicate higher levels of mindful eating and lower mean scores indicate lower levels of mindful eating.

Statistical Analyses

First, we examined the relationship between recalled coercive food parenting and mindful eating in adulthood by using a correlation analysis. Second, we examined the relationship between recalled responsive food parenting and mindful eating in adulthood by using a correlation analysis. Lastly, we examined whether sex moderates the associations between recalled food parenting and mindful eating in adulthood by using Aiken and West’s (1991) approach to linear regression analyses. Age was included as a covariate in the analyses. The predictor and covariate variables were standardized into z­scores. The interaction term between sex and recalled food parenting was created and entered into regression models to predict mindful eating in adulthood.

Results

Pearson’s correlation analyses are presented in Table 1. Consistent with our hypotheses, Pearson’s correlation analyses revealed a negative relationship between recalled coercive food parenting and adult mindful eating, r(150) = ­ .21, p = .008. Pearson’s correlation analyses also revealed a positive relationship between recalled autonomy­promoting food parenting and adult mindful eating, r(150) = .29, p < .001. Similarly, Pearson’s correlation analyses revealed a positive relationship between recalled structure­based food parenting and adult mindful eating, r(150) = .20, p = .013. Additionally, Pearson’s correlation analyses revealed a negative relationship between age and autonomy­promoting food parenting, r(128) = ­.20, p = .020, although there were no other significant correlations between age and the other key study variables. Age was included as a covariate in subsequent linear regression analyses.

Linear regression analyses examining autonomypromoting food parenting are presented in Table 2. Autonomy­promoting food parenting interacted with sex to predict adult mindful eating, F(4,124) = 3.71, p = .007, R2 = .11. A simple slopes analysis was conducted to examine the interaction effect in Figure 1. For men, the relationship between autonomy­promoting food parenting and mindful eating was not significant, B = .00, t(126) = 0.07, p = .946. However, for women, the slope was statistically significant, B = .12, t(126) = 3.66, p < .001, indicating a positive relationship between autonomy ­ promoting food parenting and mindful eating. Specifically, for women, findings revealed high autonomy­promoting food parenting was associated with higher adult mindful eating compared to those with low recalled autonomy­promoting food parenting. Linear regression analyses examining coercive control and structure­based food parenting are presented in Tables 3–4. Linear regression analysis revealed n o significant interaction between coercive control food parenting and sex to predict mindful eating, F(4,124) = 1.63, p = .172, R2 = .05. Similarly, there was no significant interaction between structure­based food parenting and sex to predict mindful eating, F(4,124) = .95, p = .449, R2 = .03.

Discussion

Taken together, the results of this study support our hypotheses and demonstrate that recalled food parenting is associated with adult mindful eating. Specifically, recalled coercive control food parenting was negatively associated with adult mindful eating. This was consistent with previous research connecting coercive control food parenting and maladaptive eating patterns in adulthood (Leuba et al., 2022; Małachowska & JeżewskaZychowicz, 2021; Tan et al., 2016). Further, research has suggested that general parenting is associated with food parenting practices and child eating outcomes (Sleddens et al., 2011). For example, Lopez et al. (2018) found that authoritative parenting provides more mealtime structure, which subsequently explains better child diet quality. Findings from prior studies, as well as the current study, suggest that parental behaviors during mealtimes are crucial in understanding a child’s eating patterns, both in the present and for the future.

Conversely, recalled responsive food parenting was positively associated with adult mindful eating. This was consistent with past research linking responsive food parenting and adaptive eating behaviors (Balantekin et al., 2020; Draxten et al., 2014; Goodman et al., 2020; Małachowska & Jeżewska­Zychowicz, 2021). Responsive food parenting may foster greater awareness of bodily sensations and promote self­regulation. Previous studies

TABLE 2

Linear Regression Analyses of Recalled AutonomyPromoting Food Parenting and Sex in Predicting Adult Mindful Eating (n = 152)

TABLE 3

Linear Regression Analyses of Recalled Coercive Control Food Parenting and Sex in Predicting Adult Mindful Eating (n = 152)

Note. CI = confidence interval for B.

FIGURE 1

Interaction Between Recalled Autonomy-Promoting Food Parenting and Sex on Adult Mindful Eating

Note ** p < .001.

have shown that autonomy ­ promoting parenting is related to better executive function in preschool children (Distefano et al., 2018). Moreover, executive function and mindfulness are strongly correlated in adults (Short et al., 2016). Taken together, these findings suggest that responsive food parenting may enhance greater awareness of bodily sensations and support self­regulation, ultimately fostering mindful eating behaviors.

Additionally, we found sex differences in the associations between food parenting and mindful eating. Although we did not have a specific hypothesis regarding the direction of these differences, the presence of such differences aligns with prior studies that highlighted sex differences in intuitive eating and disordered eating (Liu et al., 2023; Małachowska & Jeżewska­Zychowicz, 2023). Specifically, women were more likely to be mindful eaters when they experience higher autonomy­promoting food parenting compared to lower autonomy­promoting food parenting. This suggests that autonomy­promoting food parenting meaningfully relates to mindful eating in adult women. These findings may be due to social norms, as women are expected to be more sensitive to others› expectations (Määttä & Uusiautti, 2020). Societal and parental expectations of thinness in women may also impact eating behaviors (Peyer et al., 2015). Women may benefit more from being raised in a manner that promotes autonomy as it provides a supportive buffer against societal expectations. However, there was no significant association between autonomy­promoting food parenting and mindfulness for men. This suggests that autonomy ­ promoting food parenting does not meaningfully relate to mindful eating in men. Men may not benefit as much from this supportive buffer, as societal expectations tend to already value their autonomy (Määttä & Uusiautti, 2020).

Additionally, there were no sex differences in the associations between coercive control or structure­based food parenting and mindful eating. It is unknown why sex did not moderate these associations. One possibility is that, regardless of the child’s sex, parents did not differ in how they used coercive control and structure­based food parenting. Given that the current study did not find sex to be a moderator in these associations, future research could explore whether other child characteristics, such as executive function, might play a role in the association between food parenting and mindful eating.

The current study had several strengths. First, this study had a relatively diverse sample. Specifically, 30.3% of the sample are non­White, and 35.5% are men. Second, this study used well­established, reliable, and valid measures to measure retrospective food parenting and current mindful eating. Third, the results from this study were consistent with prior research in the area.

Yet, this study’s retrospective design was a limitation, as participants were asked to recall their childhood experiences, which could lead to memory bias or inaccuracies. To address this, future research might consider a longitudinal approach, tracking parent–child dyads from early childhood into adulthood to better understand the long­term relationships between food parenting practices and mindful eating. Solely using participant sex assigned at birth as a moderator was another limitation. Future research may consider examining gender identity to provide a more comprehensive understanding of potential moderating effects. Additionally, other individual characteristics such as socioeconomic status, racial identity, and/or adverse childhood experiences may impact the relationship between food parenting and mindful eating. Further, BMI in this study was based on self­reported height and weight, which may be subject to reporting bias and inaccuracy (Gosse, 2014). Future studies could address this limitation by collecting height and weight in a controlled setting. This would also allow for the exploration of related research questions, such as whether objectively measured weight moderates the association between recalled food parenting and mindful eating.

Nevertheless, this study provided valuable insights into how early childhood feeding experiences relate to eating behaviors in adulthood. It highlighted the potential long­term effects of food parenting on mindful eating and emphasized the importance of promoting adaptive food parenting to reduce maladaptive eating patterns, such as emotional eating and binge eating. By identifying factors that contribute to positive eating practices, this research can guide parents, researchers, and clinicians in fostering healthier eating habits in both children and adults. Specifically, intervention studies educating parents on the benefits of responsive food parenting—compared to negative consequences of coercive control—could support healthier eating habits as children grow into adulthood.

References

Abba, N., Chadwick, P., & Stevenson, C. (2008). Responding mindfully to distressing psychosis: A grounded theory analysis. Psychotherapy Research, 18(1), 77–87. https://doi.org/10.1080/10503300701367992

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Sage Publications, Inc.

Balantekin, K. N., Anzman-Frasca, S., Francis, L. A., Ventura, A. K., Fisher, J. O., & Johnson, S. L. (2020). Positive parenting approaches and their association with child eating and weight: A narrative review from infancy to adolescence. Pediatric Obesity, 15(10), e12722. https://doi.org/10.1111/ijpo.12722

Bennett, B. L., & Latner, J. D. (2022). Mindful eating, intuitive eating, and the loss of control over eating. Eating Behaviors, 47, 101680. https://doi.org/10.1016/j.eatbeh.2022.101680

Berge, J. M., Miller, J., Veblen-Mortenson, S., Kunin-Batson, A., Sherwood, N. E., & French, S. A. (2020). A bidirectional analysis of feeding practices and eating behaviors in parent/child dyads from low-income and minority households. The Journal of Pediatrics, 221 93–98. e20.

https://doi.org/10.1016/j.jpeds.2020.02.001

Braden, A., Musher-Eizenman, D., Watford, T., & Emley, E. (2018). Eating when depressed, anxious, bored, or happy: Are emotional eating types associated with unique psychological and physical health correlates? Appetite, 125, 410–417. https://doi.org/10.1016/j.appet.2018.02.022

Bruch, H. (1973). Psychiatric aspects of obesity. Psychiatric Annals, 3(7), 6–9. https://doi.org/10.3928/0048-5713-19730701-03

Carrière, K., Khoury, B., Günak, M. M., & Knäuper, B. (2017). Mindfulness-based interventions for weight loss: A systematic review and meta-analysis. Obesity Reviews, 19(2), 164–177. https://doi.org/10.1111/obr.12623

Didie, E. R., & Fitzgibbon, M. (2005). Binge eating and psychological distress: Is the degree of obesity a factor? Eating Behaviors, 6(1), 35–41. https://doi.org/10.1016/j.eatbeh.2004.08.007

Distefano, R., Galinsky, E., McClelland, M. M., Zelazo, P. D., & Carlson, S. M. (2018). Autonomy-supportive parenting and associations with child and parent executive function. Journal of Applied Developmental Psychology, 58, 77–85. https://doi.org/10.1016/j.appdev.2018.04.007

Draxten, M., Fulkerson, J. A., Friend, S., Flattum, C. F., & Schow, R. (2014). Parental role modeling of fruits and vegetables at meals and snacks is associated with children’s adequate consumption. Appetite, 78, 1. https://doi.org/10.1016/j.appet.2014.02.017

Forman, E. M., Shaw, J. A., Goldstein, S. P., Butryn, M. L., Martin, L. M., Meiran, N., Crosby, R. D., & Manasse, S. M. (2016). Mindful decision making and inhibitory control training as complementary means to decrease snack consumption. Appetite 103, 176–183. https://doi.org/10.1016/j.appet.2016.04.014

Framson, C., Kristal, A. R., Schenk, J. M., Littman, A. J., Zeliadt, S., & Benitez, D. (2009). Development and validation of the mindful eating questionnaire. Journal of the American Dietetic Association, 109(8), 1439–1444. https://doi.org/10.1016/j.jada.2009.05.006

Frayn, M., & Knäuper, B. (2017). Emotional eating and weight in adults: A review. Current Psychology, 37(4), 924–933. https://doi.org/10.1007/s12144-017-9577-9

Goodman, L. C., Roberts, L. T., & Musher-Eizenman, D. R. (2020). Mindful feeding: A pathway between parenting style and child eating behaviors. Eating Behaviors, 36, 101335. https://doi.org/10.1016/j.eatbeh.2019.101335

Gosse, M. A. (2014). How accurate is self-reported BMI? Nutrition Bulletin, 39(1), 105–114. https://doi.org/10.1111/nbu.12075

Hughes, K., Bellis, M. A., Hardcastle, K. A., Sethi, D., Butchart, A., Mikton, C., Jones, L., & Dunne, M. P. (2017). The effect of multiple adverse childhood experiences on health: A systematic review and meta-analysis. The Lancet Public Health, 2(8), 356–366. https://doi.org/10.1016/s2468-2667(17)30118-4

Katterman, S. N., Kleinman, B. M., Hood, M. M., Nackers, L. M., & Corsica, J. A. (2014). Mindfulness meditation as an intervention for binge eating, emotional eating, and weight loss: A systematic review. Eating Behaviors, 15(2), 197–204. https://doi.org/10.1016/j.eatbeh.2014.01.005

Kristeller, J. L., & Jordan, K. D. (2018). Mindful eating: Connecting with the wise self, the spiritual self. Frontiers in Psychology, 9, 1271. https://doi.org/10.3389/fpsyg.2018.01271

Leuba, A. L., Meyer, A. H., Kakebeeke, T. H., Stülb, K., Arhab, A., Zysset, A. E., Leeger-Aschmann, C. S., Schmutz, E. A., Kriemler, S., Jenni, O. G., Puder, J. J., Munsch, S., & Messerli-Bürgy, N. (2022). The relationship of parenting style and eating behavior in preschool children. BMC Psychology, 10(1), 275. https://doi.org/10.1186/s40359-022-00981-8

Liu, Y., Cui, T., Barnhart, W. R., Wang, Q., Yu, Y., & He, J. (2023). Associations among retrospective parenting styles, retrospective food parenting, and current eating behaviors in Chinese adults. Appetite, 106512. https://doi.org/10.1016/j.appet.2023.106512

Lopez, N. V., Schembre, S., Belcher, B. R., O’Connor, S., Maher, J. P., Arbel, R., Margolin, G., & Dunton, G. F. (2018). Parenting styles, food-related parenting practices, and children’s healthy eating: A mediation analysis to examine relationships between parenting and child diet. Appetite, 128, 205–213. https://doi.org/10.1016/j.appet.2018.06.021

Määttä, K., & Uusiautti, S. (2020). Nine contradictory observations about girls’ and boys’ upbringing and education – The strength-based approach as the way to eliminate the gender gap. Frontiers in Education, 5. https://doi.org/10.3389/feduc.2020.00134

Małachowska, A., & Jeżewska-Zychowicz, M. (2021). Does examining the childhood food experiences help to better understand food choices in adulthood? Nutrients, 13(3), 983. https://doi.org/10.3390/nu13030983

Małachowska, A., & Jeżewska-Zychowicz, M. (2023). Retrospective reports of parental feeding practices and current eating styles in polish adults.

Nutrients, 15(19), 4217. https://doi.org/10.3390/nu15194217

Marcus, M. D., Kalarchian, M. A., & Levine, M. D. (2005). Eating disorders: Binge eating. Encyclopedia of Human Nutrition, 80–85. https://doi.org/10.1016/b0-12-226694-3/02095-0

Mars, T. S., & Abbey, H. (2010). Mindfulness meditation practise as a healthcare intervention: A systematic review. International Journal of Osteopathic Medicine, 13(2), 56–66. https://doi.org/10.1016/j.ijosm.2009.07.005

Musher-Eizenman, D. R., Goodman, L., Roberts, L., Marx, J., Taylor, M., & Hoffmann, D. (2019). An examination of food parenting practices: Structure, control and autonomy promotion. Public Health Nutrition, 22(5), 814–826. https://doi.org/10.1017/S1368980018003312

Musher-Eizenman, D. R., & Holub, S. (2007). Comprehensive Feeding Practices Questionnaire: Validation of a new measure of parental feeding practices. Journal of Pediatric Psychology, 32(8), 960–972. https://doi.org/10.1093/jpepsy/jsm037

O’Reilly, G. A., Cook, L., Spruijt-Metz, D., & Black, D. S. (2014). Mindfulness-based interventions for obesity-related eating behaviours: A literature review. Obesity Reviews: An Official Journal of the International Association for the Study of Obesity, 15(6), 453–461. https://doi.org/10.1111/obr.12156

Peyer, K. L., Welk, G., Bailey-Davis, L., Yang, S., & Kim, J.-K. (2015). Factors associated with parent concern for child weight and parenting behaviors. Childhood Obesity (Print), 11(3), 269–274. https://doi.org/10.1089/chi.2014.0111

Royer, M. F., Cosgrove, K., & Wharton, C. (2023). Absent mindfulness: Mediation analyses of the relationship between adverse childhood experiences and disordered eating among young adults. Frontiers in Child and Adolescent Psychiatry, 2, 1148273. https://doi.org/10.3389/frcha.2023.1148273

Sala, M., Shankar Ram, S., Vanzhula, I. A., & Levinson, C. A. (2020). Mindfulness and eating disorder psychopathology: A meta‐analysis. International Journal of Eating Disorders, 53(6). https://doi.org/10.1002/eat.23247

Sauer, S., Walach, H., Schmidt, S., Hinterberger, T., Lynch, S., Büssing, A., & Kohls, N. (2012). Assessment of mindfulness: Review on state of the art. Mindfulness, 4(1), 3–17. https://doi.org/10.1007/s12671-012-0122-5

Short, M. M., Mazmanian, D., Oinonen, K., & Mushquash, C. J. (2016). Executive function and self-regulation mediate dispositional mindfulness and wellbeing. Personality and Individual Differences, 93, 97–103. https://doi.org/10.1016/j.paid.2015.08.007

Sleddens, S. F., Gerards, S. M., Thijs, C., De Vries, N. K., & Kremers, S. P. (2011). General parenting, childhood overweight and obesity-inducing behaviors: A review. International journal of pediatric obesity, 6(sup3), e12–27. https://doi.org/10.3109/17477166.2011.566339

Tan, C. C., Ruhl, H., Chow, C. M., & Ellis, L. (2016). Retrospective reports of parental feeding practices and emotional eating in adulthood: The role of food preoccupation. Appetite, 105, 410–415. https://doi.org/10.1016/j.appet.2016.06.009

Vaughn, A. E., Ward, D. S., Fisher, J. O., Faith, M. S., Hughes, S. O., Kremers, S. P. J., Musher-Eizenman, D. R., O’Connor, T. M., Patrick, H., & Power, T. G. (2016). Fundamental constructs in food parenting practices: A content map to guide future research. Nutrition Reviews, 74(2), 98–117. https://doi.org/10.1093/nutrit/nuv061

Warren, J. M., Smith, N., & Ashwell, M. (2017). A structured literature review on the role of mindfulness, mindful eating and intuitive eating in changing eating behaviours: Effectiveness and associated potential mechanisms. Nutrition Research Reviews, 30(2), 272–283. https://doi.org/10.1017/S0954422417000154

Yanovski, S., Nelson, J. E., Dubbert, B. K., & Spitzer, R. L. (1993). Association of binge eating disorder and psychiatric comorbidity in obese subjects [published erratum appears in Am J Psychiatry 1993 Dec;150(12):1910].

American Journal of Psychiatry, 150(10), 1472–1479. https://doi.org/10.1176/ajp.150.10.1472

Author Note

Adelyn Sherrard https://orcid.org/0000­0001­7039­8403

Cin Cin Tan https://orcid.org/0000­0001­6642­5691

We have no conflicts of interest to disclose. This study was supported by the Office of Undergraduate Research at the University of Toledo.

Abigal Brighton led conceptualization, research design, funding acquisition, data collection, data analysis and interpretation, and drafted the manuscript. Adelyn Sherrard led supervision and editorial support, contributed to analysis and interpretation, and assisted with data collection and writing. Cin Cin Tan led supervision, conceptualization, research design,

Hikikomori and Internet Addiction in U.S. College Students

Mai P. N. Tran1 and Cameron S. Kay*1,2

1Psychology

Department, Union College

2Environmental

Social Sciences Department, Stanford University

ABSTRACT. Past research has found that hikikomori (i.e., an extreme form of social withdrawal) is positively associated with internet addiction (i.e., an inordinate preoccupation with the internet). However, this association has only been established in Japan and a small selection of other countries. The goal of the present study was to extend this research to the United States. We administered the Hikikomori Questionnaire and Young’s Internet Addiction Test to 437 U.S. college students recruited through Prolific. We found a sizeable positive correlation between overall hikikomori scores and internet addiction (r = .30, 95% CI [.21, .38], p < .001), with the correlation not significantly differing from that observed among Japanese college students in prior work (z = 1.63, p = .104). We also found that the isolation factor of hikikomori demonstrated the largest association with internet addiction (r = .32, 95% CI [.23, .40], p < .001), followed by the lack of emotional support factor (r = .27, 95% CI [.17, .35], p < .001) and the disinterest in socializing factor (r = .20, 95% CI [.11, .28], p < .001). Although more research is required to further understand these relations, the present results indicate that extreme social withdrawal may be a risk factor for internet addiction in the United States (and vice versa).

Keywords: hikikomori, social withdrawal, internet addiction, college students

The United States is facing an internet addiction crisis. Internet addiction consists of excessive or poorly controlled preoccupations, urges, and behaviors of internet access and computer use that ultimately lead to impairment and distress in the user (Shaw & Black, 2008; Vondráčková & Gabrhelík, 2016). This impairment and distress can take the form of depression (Lam, 2014), physical aggression when individuals perceive that their internet usage is being restricted (Flisher, 2010), and time disruption that interferes with academic work, professional performance, and daily routines (Chou et al., 2005). There have even been reports of individuals collapsing and passing away as a result of nonstop online gaming activities (BBC, 2005).

Preregistration, Open Data, and Open Materials badges earned for transparent research practices. Preregistration can be viewed at https://osf.io/jpks6/overview Materials and data can be accessed at https://osf.io/cn6tp/overview

A 2018 cross­sectional survey found that the prevalence of internet addiction in U.S. college students was 8.0% (Tang et al., 2018), which translates to approximately 1.77 million college students, based on college enrollment data from the U.S. Census Bureau (2018). In addition, the average prevalence of internet addiction globally grew from 3.5% in 1999 to 32.4% in 2020–2021 (Meng et al., 2022), suggesting that the rates for internet addiction in the United States have likely also increased. One potential (but unexamined) reason for the concerning rate of internet addiction in the United States is hikikomori Hikikomori is a syndrome characterized by extreme social withdrawal, with one of its most notable features being physical isolation in one’s living space (Kato et

al., 2019). To be diagnosed with hikikomori, a person must meet the following criteria: they must leave their home fewer than three days a week, this social isolation must last for at least six months, and it must cause significant functional impairment or distress. Although often assessed as a singular construct, hikikomori has been decomposed into three subconstructs: lack of socialization, isolation, and a lack of emotional support (e.g., Teo et al., 2018). These subconstructs refer respectively to the absence of meaningful socialization with others (e.g., the avoidance of social interactions; discomfort and disinterest in being surrounded by others), the act of isolating oneself (e.g., spending most of one’s time alone), and the absence of perceived emotional support from others (e.g., a lack of people to share one’s personal thoughts with).

The social isolation associated with hikikomori is detrimental to people’s health. People with hikikomori are at risk of developmental disruption (Stavropoulos et al., 2019) and mood disorders (Koyama et al., 2010), as well as schizophrenia, social anxiety disorder, personality disorders, post­traumatic stress disorder, and autism spectrum disorder (Kato et al., 2019). In fact, hikikomori was recognized as a mental health disorder by Japan’s Ministry of Health, Labour and Welfare in 2010 (Saito, 2010), with about 1.2% of people between the ages of 20 and 49 in Japan reporting having experienced hikikomori in their lifetime (Koyama et al., 2010). As of February 2024, hikikomori has also been labelled as a cultural concept of distress under the category “cultural and psychiatric diagnosis” in the DSM­V­TR.

There is good reason to believe that hikikomori and internet addiction are connected. According to Davis’s (2001) cognitive ­ behavioral model of pathological internet use, pre­existing psychosocial problems, which presumably includes hikikomori, are risk factors of maladaptive internet usage. Moreover, Muris et al. (2023) notes that severe addiction is associated with social problems: addiction can lead to self­isolation, and the stress of self­isolation can lead to addiction. As applied to internet addiction and hikikomori, it is possible that maladaptive internet use could lead to hikikomori and hikikomori could, in turn, lead to maladaptive internet use.

Consistent with these theoretical suppositions, prior research has found that hikikomori and internet addiction are, indeed, linked. The first study to provide support for this association was conducted by Tateno and colleagues (2019). The researchers found that hikikomori demonstrated a large positive correlation with internet addiction (r = .39) in a sample of 478 Japanese undergraduate students.1 Since that first study, the positive association has been found in several other

1Because the present project considers individual differences, we use the effect size thresholds for individual differences research: .10 is a small correlation, .20 is a medium correlation, and .30 is a large correlation (Funder & Ozer, 2019; Gignac & Szodorai, 2016).

samples, including a sample of 1,141 Italian undergraduate students (r = .37; Orsolini et al., 2022) and a sample of 2,767 Slovakian primary school students (ρ = .33; Miriam et al., 2024).2 Incidental evidence for the association between hikikomori and internet addiction is also provided by the substantial positive association observed between hikikomori and internet gaming disorder (Dell’Osso et al., 2023; Stavropoulos et al., 2019) and between hikikomori and addiction generally (Muris et al., 2023). Concerning the subconstructs of hikikomori, a lack of emotional support (e.g., Karaer & Akdemir, 2019) and social isolation (e.g., Puri & Sharma, 2019) have both been found to be positively correlated with internet addiction. Although not exactly socialization, social self­efficacy has also been found to be negatively correlated with internet addiction (Baturay & Toker, 2019).

Despite this prior work, no studies have, to our knowledge, examined the association between hikikomori and internet addiction in the United States. Part of the reason for this may be the fact that hikikomori was originally believed to be culture­bound to Japan (Teo & Gaw, 2010), where it was first identified. As a consequence, hikikomori has remained largely unrecognized in the United States (Kato et al., 2019), with its symptoms instead being attributed to mood disorders, substance use disorders, and anxiety disorders (Teo et al., 2015b). However, as additional cases have been identified around the world (Coeli et al., 2023), hikikomori has come to be known as a modern­societybound syndrome, being tied more to the presence of information technologies in a society than to historical cultural differences (Kato et al., 2019). With this change, hikikomori is increasingly being recognized in the United States (Kato et al., 2019; Taku et al., 2023; Teo et al., 2015a), although not yet in relation to internet addiction.

Current Study

The present preregistered study examines the association between hikikomori and internet addiction among college students living in the United States. We focus on college students here because it allows us to provide a better comparison with the existing research on hikikomori and internet addition in other countries, which has focused almost entirely on college students. We also focus on this population because the two constructs of interest are particularly relevant to college students:

2A second recent study from Italy did not find a significant association between the two constructs (Amendola et al., 2021), but the study was underpowered. The study relied on a nonclinical sample of 47 participants (r = .28) and a clinical sample of 19 participants (r = -.16). By our estimates, samples of these sizes would only be able to detect a correlation of .30 or larger 54.5% and 24.4% of the time, respectively.

hikikomori tends to peak in one’s twenties (Imai et al., 2020) and college students are more likely to be addicted to the internet (Kandell, 1998). We hypothesized that there will be a significant positive correlation between hikikomori and internet addiction in a sample of U.S. college students recruited through Prolific (H1). Moreover, we hypothesized that the correlation we observe between hikikomori and internet addiction among the U.S. college students will not significantly differ from that observed among the Japanese college students in Tateno and colleagues’ (2019) original study (H2). By testing these two hypotheses, the present study provided an understanding of whether hikikomori and internet addiction are linked in the United States, and also indicated whether there are country ­ level differences in the relationship between hikikomori and internet addiction. We also examined the association between the factors of hikikomori (i.e., a disinterest in socializing, isolation, and a lack of emotional support) and internet addiction, but these analyses were conducted in an exploratory fashion.

Method

Participants and Procedure

The study was determined to be exempt from review by the IRB committee at the authors’ institution (E24018). Participants were recruited from the on­demand data collection platform Prolific to complete a Qualtrics survey, with informed consent provided at the beginning of the survey. We only recruited participants who indicated that they were (a) a student, (b) enrolled in an undergraduate degree program, and (c) 18 to 28 years of age. To achieve the specific goals of the present study, we also filtered the participants to only recruit those who (d) were located in the United States. An a priori power analysis indicated that 319 participants would be required to detect at least a moderate effect (r = .20) 95% of the time that such an effect existed with an alpha level of .05 (two­tailed). We aimed to collect at least 350 participants to account for potential misspecification of the power analysis.3 Participants were excluded if they failed two or more of four instructed response items embedded in the survey (n = 2; Curran, 2016), completed the survey in less than one­third of the median response time (n = 4; Bedford­Petersen & Saucier, 2021), provided the same response to over half of the survey items in a row (n = 10; Johnson, 2005), exhibited a response standard deviation of less than 0.50 (n = 0; Thalmayer & Saucier, 2014; see also Dunn et al., 2018), or provided an average response of 0 or greater to four infrequency/

3As described in our preregistration, we intended to continue collecting data until we had exhausted the funds we were granted to conduct the project, which resulted in a sample size larger than we had initially intended to collect.

frequency items embedded in the survey (n = 11; Kay & Saucier, 2023). After exclusions, we had a total sample of 437 valid participants (see Table 1).

Measures

The English­version of the Hikikomori Questionnaire (HQ­25; Teo et al., 2018) was used to assess hikikomori. The HQ­25 has 25 items (e.g., “I stay away from other

TABLE 1

Demographic Information for the Sample (N = 437)

Note. Additional demographic information (e.g., current

be found in the

people”) with each item scored on a scale ranging from 0 (strongly disagree) to 4 (strongly agree). The 25 items can be combined to form either an overall hikikomori score or divided into three subscales: (a) Disinterest in Socializing (e.g., “I stay away from people”), (b) Isolation (e.g., “I shut myself in my room”), and (c) Lack of Emotional Support (e.g., “There are few people I can discuss important issues with”). In the present study, the overall scale demonstrated sufficient internal consistency (α = .92), as did the subscales (αSocialization = .91; αIsolation = .82; αEmotionalSupport = .80). The Japanese version of this scale was used by Tateno et al. (2019) to investigate the association between hikikomori and internet addiction in their original study. Young’s Internet Addiction Test (YIAT; Young, 1998) was used to assess pathological internet use (Young, 1998). The YIAT has 20 items (e.g., “How often do you find that you stay online longer than you intended?”) with each item scored on the following scale: 1 (rarely), 2 (occasionally), 3 (frequently), 4 (often), 5 (always).4 In the present study, the scale demonstrated sufficient internal consistency (α = .91). This scale was also used in the original study by Tateno et al. (2019).

Results

To examine the association between hikikomori and internet addiction, we conducted a Pearson’s r correlation (see Figure 1; see Table 2). We found that hikikomori was significantly positively correlated with internet addiction, r(435) = .30, 95% CI [.21, .38], p < .001. This aligned with our hypothesis that there would be a significant positive correlation between hikikomori and internet addiction in the sample of U.S. college students (H1). We also conducted Pearson’s r correlations to examine whether the three individual factors of hikikomori are differentially associated with internet addiction. We found that a disinterest in socializing (r(435) = .20, 95% CI [.11, .28], p < .001), isolation (r(435) = .32, 95% CI [.23, .40], p < .001), and a lack of emotional support (r(435) = .27, 95% CI [.17, .35], p < .001) were all significantly positively correlated with internet addiction. Moreover, we entered the three individual factors of hikikomori as predictors of internet addiction in a multiple linear regression model (see Table 3). The overall regression model was statistically significant, R2 = .11, F(3, 433) = 19.7, p <.001. The results showed that, when controlling for their shared variance, isolation (β = .26, p <.001) and a perceived lack of emotional support (β = .13, p = .005) significantly predicted internet addiction, but a disinterest in socializing did not (β = ­.06, p = .210).

4It is unclear whether “often” is actually more frequent than “frequently.” To maintain consistency with prior studies, we did not change the response scale.

To compare the correlation observed between hikikomori and internet addiction in the United States to that observed in Japan, we conducted Fisher’s r­to­z transformations and compared the resulting z­scores. The z­score for the correlation obtained in the present study (z = .30) was not significantly different from the z­score for the correlation observed in Tateno et al.’s

Scatterplot Showing the Distributions of and Positive Association Between Hikikomori and Internet Addiction

Note. The plots show the HQ-25 score (Teo et al., 2018) and IAT score (Young, 1998) for each of the 437 valid participants. There is a positive correlation between the two scores as indicated by the red line. The correlation coefficient was r = .30 (95% CI [.21, .38], p < .001).

TABLE 2

Descriptive Statistics for and Zero-Order Correlation Among Internet Addiction, Overall Hikikomori, and the Three Hikikomori Subscales

Note * p < .001.

TABLE 3

Model Coefficients for a Multiple Linear Regression Model Predicting Internet Addiction From Socialization, Emotional Support, and Isolation

Predictor

(2019) original study (z = .41), z = 1.63, p = .104. This is consistent with our hypothesis that there would not be a significant difference between the correlations found among the U.S. and Japanese college students (H2). We conducted exploratory independent samples t tests investigating gender differences in our variables. We have also conducted exploratory independent samples t tests evaluating differences in hikikomori scores between participants who reported using the internet for a specific purpose versus those who did not report using the internet for a specific purpose (e.g., instant messaging). The results of these analyses can be found in the Supplementary Material.

Discussion

Past research has indicated that hikikomori may be a potential risk factor for internet addiction. However, this research has been limited to a select few countries (e.g., Japan, Italy). The purpose of the present study was to examine whether the association also exists in the United States. We hypothesized that we would find a significant positive correlation between hikikomori and internet addiction among a sample of U.S. undergraduate students (H1) and that the correlation would not be significantly different from that observed by Tateno and colleagues (2019) in Japan (H2).

There was a large positive correlation between the two constructs, such that participants at a high risk for hikikomori were more likely to be addicted to the internet than participants at a low risk for hikikomori. This aligned with our first hypothesis that a correlation would be found within the U.S. population. It is also consistent with what has been found in past research; the more at­risk for hikikomori individuals are, the more likely they are to be addicted to the internet (e.g., Miriam et al., 2024; Orsolini et al., 2022; Tateno et al., 2019). This correlation indicates that hikikomori may be a risk factor for internet addiction in the United States. Those who have been driven to socially withdraw may use the internet to escape feelings of pain and to cope with the situation. For example, they may not reject relationships with other people entirely but, instead, turn to online platforms to interact with others in a context where the opportunity to be criticized or judged is limited. Likewise, this finding may indicate that internet addiction contributes to hikikomori, such that people become so consumed with their online lives that they end up socially withdrawing from their offline lives. If a person is excessively preoccupied with the internet, it follows that they would likely spend much of their time on the internet and, depending on what they are using the internet for, encounter fewer opportunities to socialize, leading to greater levels of social withdrawal.

In either case, this finding provides additional support for the close link between self­isolation and addiction (Muris et al., 2023).

Exploratory correlation analyses further showed that consistent with past research, each of the three factors of hikikomori were individually associated with internet addiction (e.g., Karaer & Akdemir, 2019; Puri & Sharma, 2019). This may suggest that those with hikikomori are drawn to the internet because it allows them to avoid others and isolate from their real lives, as well as because they lack emotional support. Alternatively, as levels of internet addiction rise, people may find that they are more isolated, less likely to receive emotional support, and less interested in socializing.

Interestingly, exploratory regression analysis showed that isolation and a lack of emotional support, but not a disinterest in socializing, were significantly associated with internet addiction when controlling for their shared variance. This finding may be due to the perils of partialling, whereby a construct no longer represents itself when its shared variance with other constructs is removed (Lynam et al., 2006). However, it is also possible that self­isolation and a lack of emotional support are the primary predictors of internet addiction, with a disinterest in socializing only contributing to internet addiction insofar as it tends to co­occur with self ­ isolation and a lack of emotional support. This could be because the internet can sometimes provide the opportunity to socialize with others. In any case, we recommend additional work investigating the unique contributions of these three constructs to internet addiction.

We further found that the correlation we observed between overall hikikomori scores and internet addiction in the United States was not significantly different from that observed by Tateno and colleagues (2019) in Japan. This finding aligned with our second hypothesis that there would be little to no cultural differences between the United States and Japan in terms of the relationship between hikikomori and internet addiction. Consequently, hikikomori may be just as closely linked to developing an addiction to the internet in the United States as it is in Japan. Researchers and practitioners interested in curbing internet addiction may, therefore, benefit from considering the role of extreme social withdrawal in the condition, regardless of whether they are practicing in the United States or Japan.

Limitations and Future Directions

The present study is not without its limitations. First, a Prolific sample of U.S. undergraduate students is not necessarily generalizable to the entire population of undergraduate students in the United States (nor to

the entire population of the United States in general). We, therefore, encourage future work examining the association between hikikomori and internet addiction in samples drawn from multiple sources.

Second, our sample was similar but not identical to that used by Tateno et al. (2019). For example, although both studies recruited college students, our sample included fewer women than Tateno and colleagues’ study (56.5% versus 72.89%) and the average age was higher (22.14 versus 19.40 years). Moreover, our sample recruited participants from across the United States, whereas Tateno et al. (2019) focused on only one region of Japan. Although the correlation of hikikomori with internet addiction in the present study did not significantly differ from that reported by Tateno et al. (2019), it is worth noting that any difference in the magnitudes of the two correlations could be due to differences in the sample characteristics.

Third, the hikikomori scale that we used—the HQ­20 (Teo et al., 2018)—is only one way of assessing hikikomori. The exact definition of hikikomori is still a topic of much debate, especially given the overlap of hikikomori and existing disorders included in the DSM (e.g., mood disorders; Amendola, 2024), so the measure we used may not perfectly capture the construct. We encourage future research using additional testing methods, such as interviews.

Finally, we cannot infer causality in our study. Rather, this study only provides evidence for the correlation between hikikomori and internet addiction in the United States. It is possible (and, in fact, quite plausible) that there is a bidirectional relationship between the two constructs, such that they are mutually reinforcing. Future studies using longitudinal designs (e.g., crosslagged models) could provide a valuable contribution to understanding how hikikomori and internet addiction influence each other over time.

Beyond addressing the limitations above, future studies could test different interventions for reducing hikikomori and internet addiction. One recently proposed intervention for hikikomori is the use of virtual reality (VR; Aguglia et al., 2024). With more than half of our participants using the internet for online gaming (57.2%), there is a possibility that hikikomori can be combatted by progressively increasing the amount of social interaction required of high ­ risk hikikomori individuals in social­interaction VR games. Furthermore, future work could consider whether the student status of participants affects the association observed between hikikomori and internet addiction. Questions remain as to whether an association between hikikomori and internet addiction would be found for participants who are not currently enrolled in an institution of higher education, as well as for participants who only take classes online.

Conclusions

According to past research, the more at­risk a person is for hikikomori, the more at­risk they are for internet addiction. However, this prior research was restricted to only a few select countries. Our research found further support for the relationship between hikikomori and internet addiction in the United States. As in other countries, the internet may be seen by socially withdrawn U.S. users as a place of refuge, where they can be alone, find emotional support, and socialize. However, this relationship can also potentially turn maladaptive, leading to a preoccupation with the internet that not only causes impairment and distress for the user but that can also lead to even greater social withdrawal. The present study, therefore, highlights the need for greater consideration of both hikikomori and internet addiction in the United States, both in the context of research and in clinical practice, as well as at both individual and institutional levels.

References

Aguglia, A., Masini, F., Franziska De Michiel, C., Amerio, A., Orsolini, L., Volpe, U., & Serafini, G. (2024). Virtual reality as a novel therapeutic approach to Hikikomori. International Journal of Social Psychiatry, 70(8), 1554–1556. https://doi.org/10.1177/00207640241270748

Amendola, S., Cerutti, R., Presaghi, F., Spensieri, V., Lucidi, C., Silvestri, E., Giorgio, V. D., Conti, F., Martorelli, A., Izzi, G., & Teo, A. (2021). Hikikomori, problematic internet use and psychopathology: Correlates in non-clinical and clinical samples of young adults in Italy. Journal of Psychopathology, 27(2), 106–114. https://doi.org/10.36148/2284-0249-412

Amendola, S. (2024). Clarifying the position of hikikomori in mental health: Is hikikomori a variant of already-known mental health disorders? A review of the literature. Journal of Pacific Rim Psychology, 18, 18344909241274808. https://doi.org/10.1177/18344909241274808

Baturay, M. H., & Toker, S. (2019). Internet addiction among college students: Some causes and effects. Education and Information Technologies, 24(5), 2863–2885. https://doi.org/10.1007/s10639-019-09894-3

BBC. (2005, August 10). Technology | S Korean dies after games session. BBC News. https://news.bbc.co.uk/2/hi/technology/4137782.stm

Bedford-Petersen, C., & Saucier, G. (2021). Identifying contrasting themes that orchestrate personality expression across situations. Personality and Individual Differences, 171, 110495. https://doi.org/10.1016/j.paid.2020.110495

Chou, C., Condron, L., & Belland, J. C. (2005). A review of the research on Internet addiction. Educational Psychology Review, 17, 363–388. https://doi.org/10.1007/s10648-005-8138-1

Coeli, G., Planas-Lladó, A., & Soler-Masó, P. (2023). The relevance of educational contexts in the emergence of Social Withdrawal (hikikomori). A review and directions for future research. International Journal of Educational Development, 99, 102756. https://doi.org/10.1016/j.ijedudev.2023.102756

Curran, P. G. (2016). Methods for the detection of carelessly invalid responses in survey data. Journal of Experimental Social Psychology, 66, 4–19. https://doi.org/10.1016/j.jesp.2015.07.006

Davis, R. A. (2001). A cognitive-behavioral model of pathological Internet use. Computers in Human Behavior, 17(2), 187–195. https://doi.org/10.1016/S0747-5632(00)00041-8

Dell’Osso, L., Amatori, G., Muti, D., Giovannoni, F., Parri, F., Violi, M., Cremone, I. M., & Carpita, B. (2023). Autism spectrum, Hikikomori syndrome and internet gaming disorder: Is there a link? Brain Sciences, 13(7), 1116. https://doi.org/10.3390/brainsci13071116

Dunn, A. M., Heggestad, E. D., Shanock, L. R., & Theilgard, N. (2018). Intraindividual response variability as an indicator of insufficient effort responding: Comparison to other indicators and relationships with individual differences. Journal of Business and Psychology, 33(1), 105–121. https://doi.org/10.1007/s10869-016-9479-0

Flisher, C. (2010). Getting plugged in: An overview of internet addiction. Journal of Paediatrics and Child Health, 46(10), 557–559. https://doi.org/10.1111/j.1440-1754.2010.01879.x

Funder, D. C., & Ozer, D. J. (2019). Evaluating effect size in psychological research: Sense and nonsense. Advances in Methods and Practices in Psychological Science, 2(2), 156–168. https://doi.org/10.1177/2515245919847202

Gignac, G. E., & Szodorai, E. T. (2016). Effect size guidelines for individual differences researchers. Personality and Individual Differences, 102, 74–78. https://doi.org/10.1016/j.paid.2016.06.069

Imai, H., Takamatsu, T., Mitsuya, H., Yoshizawa, H., Mitsuya, H., & Furukawa, T.A. (2020). The characteristics and social functioning of pathological social withdrawal, “Hikikomori,” in a secondary care setting: A one-year cohort study. BMC Psychiatry, 20, 1–11. https://doi.org/10.1186/s12888-020-02660-7

Johnson, J. A. (2005). Ascertaining the validity of individual protocols from Webbased personality inventories. Journal of Research in Personality, 39, 103–129. https://doi.org/10.1016/j.jrp.2004.09.009

Kandell, J. J. (1998). Internet addiction on campus: The vulnerability of college students. Cyberpsychology & Behavior, 1(1), 11–17. https://doi.org/10.1089/cpb.1998.1.11

Karaer, Y., & Akdemir, D. (2019). Parenting styles, perceived social support and emotion regulation in adolescents with internet addiction. Comprehensive Psychiatry, 92, 22–27. https://doi.org/10.1016/j.comppsych.2019.03.003

Kato, T. A., Kanba, S., & Teo, A. R. (2019). Hikikomori: Multidimensional understanding, assessment, and future international perspectives. Psychiatry and Clinical Neurosciences, 73(8), 427–440. https://doi.org/10.1111/pcn.12895

Kay, C. S., & Saucier, G. (2023). The Comprehensive Infrequency/Frequency Item Repository (CIFR): An online database of items for detecting careless/ insufficient-effort responders in survey data. Personality and Individual Differences, 205, 112073. https://doi.org/10.1016/j.paid.2022.112073

Koyama, A., Miyake, Y., Kawakami, N., Tsuchiya, M., Tachimori, H., Takeshima, T., & World Mental Health Japan Survey Group. (2010). Lifetime prevalence, psychiatric comorbidity and demographic correlates of “Hikikomori” in a community population in Japan. Psychiatry Research, 176(1), 69–74. https://doi.org/10.1016/j.psychres.2008.10.019

Lam, L. T. (2014). Risk factors of Internet addiction and the health effect of internet addiction on adolescents: A systematic review of longitudinal and prospective studies. Current Psychiatry Reports, 16, 1–9. https://doi.org/10.1007/s11920-014-0508-2

Lynam, D. R., Hoyle, R. H., & Newman, J. P. (2006). The perils of partialling: Cautionary tales from aggression and psychopathy. Assessment, 13(3), 328–341. https://doi.org/10.1177/1073191106290562

Meng, S. Q., Cheng, J. L., Li, Y. Y., Yang, X. Q., Zheng, J. W., Chang, X. W., Shi, Y., Chen, Y., Lu, L., Sun, Y., Bao, Y. P., & Shi, J. (2022). Global prevalence of digital addiction in general population: A systematic review and meta-analysis. Clinical Psychology Review, 92, 102128. https://doi.org/10.1016/j.cpr.2022.102128

Miriam, N., Karina, Z., & Annamária, Š. (2024). Prevalence of the Hikikomori Syndrome in the context of internet addictive behaviour among primary school pupils in the Slovak Republic. TEM Journal, 13(1). https://doi.org/10.18421/TEM131-49

Muris, P., van de Pasch, V., van Kessel, J., & Peet, J. (2023). The relationship between addiction and hikikomori tendencies: A case-control study. Frontiers in Psychiatry, 14, 1273865. https://doi.org/10.3389/fpsyt.2023.1273865

Orsolini, L., Yılmaz-Karaman, I. G., Longo, G., Bellagamba, S., Kato, T. A., & Volpe, U. (2022). Sex-differences in Hikikomori traits as predictors of problematic internet use in Italian university students. Journal of Psychiatric Research, 155, 211–218. https://doi.org/10.1016/j.jpsychires.2022.08.015

Puri, A., & Sharma, R. (2016). Internet usage, depression, social isolation and loneliness amongst adolescents. Indian Journal of Health & Wellbeing, 7(10). https://openurl.ebsco.com/EPDB%3Agcd%3A1%3A2363073/detailv2?sid=e bsco%3Aplink%3Ascholar&id=ebsco%3Agcd%3A119657611

Saito, K. (2010). Hikikomori no hyouka: shien Ni kansuru gaidorain. Retrieved March 20, 2026, from https://kohnodai.jihs.go.jp/subject/100/22ncgm_hikikomori.pdf Shaw, M., & Black, D. W. (2008). Internet addiction: Definition, assessment, epidemiology and clinical management. CNS Drugs, 22, 353–365. https://doi.org/10.2165/00023210-200822050-00001

Stavropoulos, V., Anderson, E. E., Beard, C., Latifi, M. Q., Kuss, D., & Griffiths, M. (2019). A preliminary cross-cultural study of Hikikomori and Internet Gaming Disorder: The moderating effects of game-playing time and living with parents. Addictive Behaviors Reports, 9, 100137. https://doi.org/10.1016/j.abrep.2018.10.001

Taku, K., Rhodes, J., & Zeigler-Hill, V. (2023). Associations that Hikikomori had with thwarted belongingness and perceived burdensomeness in the United States and Japan. Personality and Individual Differences, 214, 112355. https://doi.org/10.1016/j.paid.2023.112355

Tang, C. S. K., Wu, A. M. S., Yan, E. C. W., Ko, J. H. C., Kwon, J. H., Yogo, M., Gan, Y. Q., & Koh, Y. Y. W. (2018). Relative risks of Internet-related addictions and mood disturbances among college students: A 7-country/region comparison. Public Health, 165, 16–25. https://doi.org/10.1016/j.puhe.2018.09.010

Tateno, M., Teo, A. R., Ukai, W., Kanazawa, J., Katsuki, R., Kubo, H., & Kato, T. A. (2019). Internet addiction, smartphone addiction, and Hikikomori trait in Japanese young adult: Social isolation and social network. Frontiers in Psychiatry, 10, 455. https://doi.org/10.3389/fpsyt.2019.00455

Teo, A. R., & Gaw, A. C. (2010). Hikikomori, a Japanese culture-bound syndrome of social withdrawal?: A proposal for DSM-5. The Journal of Nervous and Mental Disease, 198(6), 444–449. https://doi.org/10.1097/NMD.0b013e3181e086b1

Teo, A. R., Fetters, M. D., Stufflebam, K., Tateno, M., Balhara, Y., Choi, T. Y., Kanba, S., Mathews, C. A., & Kato, T.A. (2015a). Identification of the Hikikomori syndrome of social withdrawal: Psychosocial features and treatment preferences in four countries. International Journal of Social Psychiatry, 61(1), 64–72. https://doi.org/10.1177/0020764014535758

Teo, A. R., Stufflebam, K., Saha, S., Fetters, M. D., Tateno, M., Kanba, S., & Kato, T. A. (2015b). Psychopathology associated with social withdrawal: Idiopathic and comorbid presentations. Psychiatry Research, 228(1), 182–183. https://doi.org/10.1016/j.psychres.2015.04.033

Teo, A. R., Chen, J. I., Kubo, H., Katsuki, R., Sato‐Kasai, M., Shimokawa, N., Hayakawa K., Umene-Nakano, W., Aikens, J. E., Kanba, S., & Kato, T. A. (2018). Development and validation of the 25‐item Hikikomori Questionnaire (HQ‐25). Psychiatry and Clinical Neurosciences, 72(10), 780–788. https://doi.org/10.1111/pcn.12691

Thalmayer, A. G., & Saucier, G. (2014). The questionnaire Big Six in 26 nations: Developing cross-culturally applicable Big Six, Big Five and Big Two inventories. European Journal of Personality, 28, 482–496. https://doi.org/10.1002/per.1969

U.S. Census Bureau. School enrollment in the United States: 2018. (2018). https://www.census.gov/content/dam/Census/library/publications/2020/ demo/p20-584.pdf

Vondráčková, P., & Gabrhelík, R. (2016). Prevention of Internet addiction: A systematic review. Journal of Behavioral Addictions, 5(4), 568–579. https://doi.org/10.1556/2006.5.2016.085

Young, K. S. (1998). Internet Addiction Test [Database record]. Retrieved from APA PsycTests. https://dx.doi.org/10.1037/t41898-000

Author Note

Mai P. N. Tran https://orcid.org/0009­0000­4004­8429

Cameron S. Kay https://orcid.org/0000­0002­5210­427X

This study was preregistered at https://osf.io/jpks6/overview

The materials, data, and analytic code are provided at https://osf.io/cn6tp/overview. This study was supported by a Union College Student Research Grant. We have no known conflicts of interest to disclose.

Mai P.N. Tran played a lead role in conceptualization, research design, data collection, data analysis and interpretation, and substantive original writing. Cameron S. Kay played a lead role in research design, data collection, data analysis and interpretation, and substantive original writing, and a supporting role in conceptualization.

Correspondence concerning this article may be addressed to Mai P. N. Tran. Email: maitran20050417@gmail.com

Exploring Factors That Support Sense of Belonging in Undergraduates at a Large University in the United States

Kylie

S.

Sambirsky,

Kit

E. Ganzle,

Sienna M. Russell, Ian J. McGillicuddy, Mandira T. Gowda, Yohana A. Markos, Joseph Levy, Leah Teeters*, and Winnie Zhuang*

Renée Crown Wellness Institute, University of Colorado Boulder

ABSTRACT. Prior literature suggests that sense of belonging is an important predictor of undergraduate success. However, the extent to which individual­level and social factors, which play crucial roles in operationalizing belonging, relate to belonging across college contexts (e.g., global, university, departmental) is unclear. In this study, we examined factors that predicted students’ sense of belonging at a large public university in the United States. We administered questionnaires measuring sense of belonging to college, department, and global contexts, demographic factors, and mental health indicators to undergraduate students (n = 138) at the university. Results suggested that (a) departmental belonging is greater than university and global belonging (p = .002), (b) global sense of belonging and general self­efficacy are strong predictors of departmental belonging ( p < .001; p = .007), and (c) loneliness and perceived social support from friends marginally predicted university and global belonging (p < .001). The results of this study advocate for supporting student belonging within their department by bolstering general self­efficacy and at the university­ and global­levels through social factors, such as loneliness and social support.

Keywords: belonging, college, department, loneliness, peer support, self­efficacy, student, undergraduate, underrepresented minority, university

Belonging, defined as feeling valued, accepted, and that one’s traits align with the surrounding environment (Anistranski & Brown, 2021), is a key determinant of college students’ academic success and psychosocial well­being. A strong sense of belonging is linked to a wide range of positive student outcomes. These include academic persistence and degree completion, mental well­being, professional development, school identification, mastery ­ oriented goals, self ­ esteem, behavioral engagement, and higher performance related to grades and standardized testing (Allen et al., 2021; Anistranski & Brown, 2021; Arslan, 2021; Korpershoek et al., 2019; McBeath, 2016; Moeller et al., 2020).

The impact of belonging is especially important given that significantly higher rates of mental health issues have been documented among U.S. college students in recent years. From 2007 to 2017, the percentage of U.S. students with diagnosed mental health conditions increased from 21.9% to 35.5% (Lipson et al., 2019). According to a 2020 study on the impact of

COVID­19 on college students’ mental health in the U.S., 40.9% reported symptoms of depression, 31% symptoms of anxiety, 17.7% suicidal ideation, 30.5% mental health concerns that impacted academics, and 47% reported loneliness in 2020 (HMN & ACHA, 2020). Many enrolled college students are struggling with mental health issues, despite potentially not having the resources to obtain an official diagnosis. Given these concerning trends, fostering student sense of belonging may be a key pathway to shrink mental health challenges and support well­being. Several factors are associated with students’ sense of belonging. Loneliness has long been established as a predictor of sense of belonging and mental health outcomes (Hughes et al., 2004). For example, in a study of 454 undergraduates, higher loneliness was correlated with greater symptoms of depression, anxiety, stress, and general poor mental health over time (Richardson et al., 2017). Similarly, self­efficacy is associated with better mental health and psychological well­being and

with lower levels of psychological distress (Gull, 2016). Defined as someone’s perception of their ability to succeed in a specific task or goal, self­efficacy is important for academic success, degree completion, and student sense of belonging (Tinto, 2017). However, there has been little prior research conducted into whether loneliness and general self­efficacy are generalizable across student context, which is essential to consider when university provides many outlets (e.g., within academic departments and within the institution overall) for students to manifest either of these factors.

Peer support has also been implicated as a core facet of sense of belonging for college students. In previous research, students indicated that peer support and sense of belonging were vital to their well­being and mental health (Kiefer et al., 2015; McBeath et al., 2018). Moreover, students often turn to other students when they are in distress before reaching out to counseling services for support. Peers are not only important for fostering a sense of belonging on campus, but also for supporting overall mental health (Morse & Schulze, 2013).

Previous research on sense of belonging has illustrated its importance for thriving; however, there is not a single, standardized measure of belonging (Mellinger et al., 2023). Belonging has often been treated as a single, uniform construct, overlooking its complexity and the fact that it arises from multiple sources and contexts (e.g., global sense of belonging being overall disposition to experience a sense of belonging regardless of context). Accordingly, it is essential to consider sense of belonging across contexts in tandem with the students’ own diverse perspectives and social identities, such as race and generational status, which impacts how belonging is felt in a space (Duran et al., 2020). Analysis of the interplay of belonging across contexts with academic and social factors is sparse, yet it can provide valuable insight into student experiences, denoting both their needs and challenges, that can further be used to recommend and implement changes benefiting students across the university and United States.

The current investigation explores patterns of student belonging at a large public university in the United States. We examined undergraduate students’ sense of belonging and how it may differ across contexts, including within their respective departments, their university, and globally. We explored the factors that may predict their sense of belonging, which include loneliness, self­efficacy, and peer relationships. We also collected demographic data, such as year in school, major(s), first ­ generation student status, minority status, and financial stress, which appear to be especially salient in promoting a sense of belonging among college students (Budge et al., 2020; Gillen­O’Neel, 2021; Reid

et al., 2020; Salusky et al., 2024; Zimmerman & Parker III, 2023). Our first hypothesis was that participants’ average sense of belonging to their university and their department (i.e., specific contexts) would be less than their overall disposition to experience a sense of belonging (i.e., global contexts). We predicted this due to the potential that student identity within this specific context may not be fully formed; especially if students are new to campus, they may not have had appropriate time to find identity and belonging within department­ and university­specific contexts. Our second hypothesis was that sense of belonging across contexts (i.e., global, university, and departmental belonging) would be directly correlated with general self­efficacy and perceived social support from friends, but inversely related to loneliness (McBeath et al., 2018; Mellinger et al. 2023). Finally, we explored the extent to which demographic factors predict students’ sense of belonging across contexts. By examining undergraduate students’ sense of belonging across specific contexts, we aimed to generate a more nuanced understanding of how supportive factors are associated with belonging in different contexts. In turn, this contributes to a better understanding of the needs of students to best support them in their personal and academic pursuits, informing interventions that could foster a stronger sense of belonging, community, and inclusion for students.

Methods

Design

Study data were collected and managed online using REDCap electronic data capture tools hosted at the University of Colorado Boulder (Harris et al., 2009; 2019) . We originally planned to recruit a total of n = 225 participants based on an a priori power analysis using the ‘pwrss’ package in R (version 0.3.1; Bulus, 2023). This sample size allowed for 80% power when accounting for 15% attrition based on the effect sizes found by Mellinger et al. (2023). REDCap was utilized to create an online consent process where participants are reminded that they could withdraw from the study at any time. This study was approved by the local Institutional Board of Research and conducted in compliance with the approved protocol.

Participants

The present study recruited a convenience sample of adult undergraduate students at a large public university in the Mountain West. Students recruited through the general psychology subject pool received one class credit for participating in the study. Students recruited from the broader community through campus fliers, local business, and social media posts received a $15 gift card for completing the survey.

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

A total of 156 students, initially recruited for the online questionnaire, gave informed consent, and the final sample consisted of only those who completed the entire survey (N = 138; nGeneral Psychology = 101, nCommunity = 37). The average age of the final sample was 19.9 years old (SD = 5.3 years); the largest reported categories of participant identities were White, women, members of the Department of Psychology and Neuroscience, full­time students, and employed. Underrepresented minority (URM) status was defined based on participants’ self­reported, multiselect race/ethnicity categories. Participants identifying exclusively as White were coded as having majority status (non­URM), in line with the majority demographic representation of both the university population and the broader U.S. context. All other racial/ethnic identities were grouped as URM. This operationalization reflects analytic constraints related to data structure and sample sizes rather than a conceptual definition of underrepresentation. Full details of the demographic information can be found in Table 1.

Measures Demographics

Basic demographic information was collected from participants, including sex/gender, age, race/ethnicity, class year, major(s), credit hours (part/full time), and first­generation status. All participants were asked to enter their school identifier to ensure that they attended the university; however, to maintain anonymity, identifiers were removed prior to data processing and analysis.

UCLA 3-Item Loneliness Scale

The UCLA 3 ­ Item Loneliness Scale (Hughes et al., 2004) measures general feelings of loneliness through 3 items designed to examine relational connectedness, self ­ perceived isolation, and social connectedness. Respondents were asked “Indicate how often you feel the way described in each of the following statements” through the 3­point rating system of 1 (hardly ever), 2 (some of the time), and 3 (often). Total sum scores range from 3 to 9, and higher scores reflect greater loneliness. The scale had good internal consistency (α = .82) in the current study.

General Self-Efficacy 6-Item Scale

The General Self­Efficacy 6­Item Scale (GSE­6) evaluates an individual’s belief in their capacity to effectively handle challenges and execute complex or novel tasks (Romppel et al., 2013). Respondents assess the personal relevance of statements such as “It is easy for me to stick to my aims and accomplish my goals” using a 4­point scale of response options ranging from 1 ( not at all true) to 4 (exactly true). The 6­item version of the GSE

was chosen to reduce participant burden since it has similar psychometric properties to the original 10­item version developed by Schwarzer & Jerusalem (1995). Total sum scores range from 6 to 24, with higher scores reflecting greater self­efficacy. The scale had good internal consistency (α = .82) in the current study.

Perceived Social Support From Friends Scale

The Perceived Social Support from Friends Scale (PSS ­ Fr) was implemented to examine participants perceived social support from friends/peers (Procidano & Heller, 1983). These 20 items specifically ask about respondents’ experiences/feelings in relation to their friendships. To capture more nuanced insights, we chose to follow the approach by Glozah and Pevalin (2017) to modify the original version’s three possible answer choices (Yes, No, Don’t know) to assess items

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

on a 5­point scale of 1 (strongly disagree) to 5 (strongly agree). Six items were reverse scored. Total sum scores range from 20 to 100, with higher scores reflecting greater perceived social support. The scale had excellent internal consistency (α = .91) in the current study.

Sense of Belonging 8-Item Scale

The Sense of Belonging 8 ­ Item Scale (SBS ­ 8) was employed to assess sense of belonging in different contexts (Mellinger et al. 2023). This scale is intended to examine participants’ perceptions of belonging globally and concerning a given context, which in this case were belonging to their department and university. We administered the trait (i.e., not state) forms as this version represents a stable perception, which was better suited to address our aims. We also included the trait global form of the SBS­8 to examine students’ overall disposition to experience a sense of belonging regardless of context. Participants assess 8 items on a 5­point scale of 1 (strongly disagree) to 5 (strongly agree). An example of one of the statements participants responded to was “When you think about your life, to what extent do you feel a general sense of belonging and fitting in?” Four items were reverse scored. Total mean scores range from 1–5, with higher scores reflecting a greater sense of belonging. Both the global and university belonging scales had excellent internal consistency (Global SBS α = .92; University SBS α = .94) in the current study, and departmental belonging had good internal consistency (Department SBS α = .86) in the current study.

Procedures

Participants were free to take this survey at a location and time, and on a device of their choosing, but it was recommended that they complete the survey via computer. All participants completed measures in the following order: demographics, Department SBS­8, UCLA 3 ­ Item Scale, GSE ­ 6, PSS ­ Fr, Global SBS ­ 8, and University SBS­8. The survey took approximately 15 minutes to complete. Additionally, three attention checks were embedded throughout the survey to assess response quality. Nine participants failed the first attention check, ten failed the second, and twelve failed the third, with minimal overlap among these individuals; only two participants failed all three. Notably, sensitivity analyses excluding participants who failed any attention checks did not change the direction or significance of the results, indicating that these participants did not meaningfully influence the observed effects.

Data Analysis

For each analysis, participants with missing data relevant to the analysis were removed. After each analysis, we

identified outliers using the identify_outliers() function from the rstatix package in R, which identifies potential outliers based on interquartile range (IQR). To examine our assumptions of normality and homoscedasticity, we generated normal quantile­quantile plots and spreadlocation plots. Data were transformed appropriately to meet assumptions for parametric tests. For each analysis, we tested whether any demographic variables predict the outcome of interest and included these predictive variables as covariates in subsequent analyses.

Analyses were conducted using the statistical coding language R, version 4.3.3 (R Core Team, 2024). For the first hypothesis, a one­way repeated­measures analysis of variance (RM ­ ANOVA) was used to determine whether there were statistically significant differences between global, CU, and departmental SBS­8 scores. If such a difference was identified, we intended to use a pairwise t­test for a priori hypotheses and Tukey HSD for post hoc comparisons, respectively. For our second hypothesis, a Pearson product­moment correlation was used to examine the relationship between the variables of interest. Finally, we examined multivariate regressions between SBS ­ 8 scores and loneliness, general self­efficacy, and peer relationships to identify significant predictors of belonging.

Results

We examined descriptive statistics and bivariate correlations among the variables of interest, which can be found in Table 2 A Pearson’s r test was used to inform us

Note M and SD are used to represent mean and standard deviation, respectively. Values in square brackets indicate the 95% confidence interval for each correlation. The confidence interval is a plausible range of population correlations that could have caused the sample correlation (Cumming, 2014). PSS-Fr = Perceived Social Support from Friends, GSE-6 = General Self-Efficacy, SBS = Sense of Belonging. * indicates p < .05. ** indicates p < .01.

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

of the strength and direction of the relationships between belonging contexts and variables of interest. We found that loneliness, general self­efficacy, and peer relationships were all significantly correlated in the directions that we expected across all measures of belonging, which provides support for their associations in our sample (Table 2) The data was found to be normally distributed, and no extreme outliers were identified using the Shapiro­Wilk test and visual inspection of a boxplot and Quantile­Quantile plot.

First, we examined the relationship between sense of belonging across contexts (global, college, and department). A Mauchly’s test indicated that the assumption of sphericity was violated, p < .001, so a Greenhouse­Geiser correction was applied. A repeated­measures ANOVA revealed a statistically significant effect of context on students’ sense of belonging, F (1.8, 246.15) = 8.60, p < .001, η² = .06. Follow up pairwise t tests with Bonferroni corrections showed that departmental belonging ( M = 3.94, SD = 0.72) was significantly greater than university belonging (M = 3.72, SD = 0.87; p = .002) and that departmental belonging was significantly greater than global belonging (M = 3.73, SD = 0.82; p = .005) There was no significant difference between university and global belonging, p > .05. These results illustrate that students’ sense of belonging is greater in departmental contexts than in university or global contexts (see Figure 1).

Next, we examined whether URM status was associated with sense of belonging across different contexts. We conducted a 2 (URM vs. Non­URM) x 3 (global, university, and department) mixed ­ model ANOVA with the first factor (URM status) varying between participants and the second factor (context) varying within participants. The predicted two­way interaction between URM status and belonging context was not significant, F(1.8, 244.71) = 0.97, p = .373. There was no significant main effect of URM status, F(1, 136) = 0.18, p = .674. However, there was a significant main effect of SBS context, F(1.8, 244.71) = 4.60, p = .014, η² = .03. Although there was no significant relationship between URM status and SBS context and no significant main effect of URM status, sense of belonging context was still shown to predict belonging as reported previously.

Finally, we conducted a multivariate regression analysis to predict departmental belonging for all students with general self­efficacy, loneliness, and perceived social support from friends while controlling for participants’ disposition to generally experience a sense of belonging (i.e., global trait belonging). The model was fitted on a standardized version of the dataset to obtain standardized parameters, and a Wald t distribution approximation was used to compute p values. The overall model explained

TABLE 3

Regression Results Predicting Sense of Belonging by Department, University, and Global

Predictor

(Intercept) 2.34** [1.13, 3.55]

Loneliness -0.05 [-0.14, 0.03] -.13 [-.33, .07] .01 [-.02, .03] .210

PSS-Fr -0.01 [-0.02, 0.00] -.11 [-.29, .07] .01 [-.02, .03] .244

GSE-6 0.05** [0.01, 0.09] .21 [.06, .36] .04 [-.01, .09] .007**

Global SBS 0.38** [0.19, 0.58] .44 [.21, .67] .07 [.00, .15] < .001**

R2 = .327**

95% CI[.19,.42]

(Intercept) 1.16* [0.07, 2.25]

Loneliness -0.07 [-0.15, 0.00] -.15 [-.30, .01] .01 [-.01, .03] .060

PSS-Fr 0.00 [-0.01, 0.01] .05 [-.09, .18] .00 [-.01, .01] .484

GSE-6 0.01 [-0.02, 0.04] .03 [-.08, .15] .00 [-.01, .01] .553

Global SBS 0.67** [0.49, 0.85] .64 [.47, .81] .15 [.07, .24] < .001**

R2 = .626**

95% CI[.52,.69] (Intercept) 3.20** [2.32, 4.07]

Loneliness -0.28** [-0.33, -0.22] -.58 [-.69, -.46] .27 [.16, .38] < .001**

PSS-Fr 0.02** [0.01, 0.03] .31 [.19, .44] .07 [.02, .13] < .001**

GSE-6 0.02 [-0.01, 0.05] .07 [-.04, .19] .00 [-.01, .02] .195 R2 = .623**

95% CI[.52,.69]

Note. A significant b-weight indicates the beta-weight and semi-partial correlation are also significant. b represents unstandardized regression weights. bindicates the standardized regression weights. sr2 represents the semi-partial correlation squared. p represents statistical significance. LL and UL indicate the lower and upper limits of a confidence interval, respectively. PSS-Fr = Perceived Social Support from Friends, GSE-6 = General Self-Efficacy, SBS = Sense of Belonging. * indicates p < .05. ** indicates p < .01.

FIGURE 1

Mean Sense of Belonging Scores Across Contexts

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

a statistically significant and substantial proportion of variance, R2 = .33, F(4, 133) = 16.17, p < .001, adj. R2 = .31. Among the predictors, global sense of belonging and general self­efficacy both emerged as significant positive predictors of departmental belonging, b = .38, SE = 0.10, t(133) = 3.82, p < .001; b = .05, SE = 0.02, t(133) = 2.72, p = .007, indicating that students who reported higher levels of these qualities were more likely to report a higher sense of belonging to their academic department. In contrast, perceived social support from friends and loneliness did not significantly predict belonging in this model, though both trended negatively. Overall, broader psychological factors may play a more central role in shaping belonging than peer support or perceived feelings of isolation. A full breakdown of the multiple regression results can be found in Table 3.

Exploratory Analyses

In further exploratory analyses, we conducted a multivariate regression to examine predictors of students’ university belonging. Global sense of belonging emerged as a significant positive predictor of university belonging, b = .67, SE = 0.09, t(133) = 7.41, p < .001, indicating that individuals who report a higher sense of global belonging are more likely to report higher university belonging. Further, loneliness trended as a marginally significant predictor as well, b = ­.07, SE = 0.04, t(133) = ­1.90, p = .060. In contrast, general self­efficacy and perceived social support from friends did not significantly predict university belonging in this model (see Table 3). These findings suggest that in the university context, loneliness may have more of an impact on belonging than other factors.

Finally, we conducted a multiple regression to examine predictors of students’ global belonging. Loneliness and perceived social support from friends emerged as significant predictors (negative and positive, respectively) of global belonging, b = ­.28, SE = 0.44, t(134) = ­9.79, p < .001; b = .02, SE < .01, t(134) = 5.04, p < .001. In contrast, general self ­ efficacy did not emerge as a significant predictor in this model, inversely reflecting the multiple regression results predicting departmental sense of belonging. Overall, these findings suggest that as sense of belonging becomes broader, social factors become more important. A full breakdown of the exploratory analyses results can be found in Table 3.

Discussion

This project aimed to determine whether students’ sense of belonging differs across context (i.e., university, department, and global) and whether sense of belonging would be related to various individual and social

factors (i.e., URM status, academic department, general self­efficacy, loneliness, and peer support). Students’ sense of belonging was shown to be greater within their home departments than at their university or globally. Although URM status was not associated with sense of belonging, global sense of belonging and general selfefficacy predicted sense of belonging to departments. These findings reveal important factors that shape university students’ sense of belonging to different levels of their institution.

Contrary to our hypothesis, the results indicate that sense of belonging in participants’ home departments was greater than sense of belonging at the university or globally. Our results suggest that sense of belonging may be more salient for specific contexts and may be inversely related to the size of the context where it is being measured. Previous research supports this finding (Knekta & McCartney, 2021; Kuh et al., 2006) and indicates that this could be due to increased autonomy within departments to enact change and impact student belonging.

The current study reported a lack of group effects for URM students, which counters our original hypothesis. Previous research notes the importance of social identities, such as race, ethnicity, sexuality, and first­generation status, on college education (Duran et al., 2020; Evans et al., 2017; Garcia, 2019; MacNear & Hunter, 2025), which contradicts the current findings. Kuh et al. (2008) also established the effect that engagement in one’s department can have on sense of belonging for URM students. The current study’s null findings may be due to insufficient power to detect statistically significant differences between groups and to the skewed representation of our sample. In our sample, students who reported holding a minority social identity were far fewer in number than students who reported holding the majority social identity for each demographic variable measured. Interestingly, explorations of belonging by URM status showed that students who were not underrepresented minorities experienced higher sense of belonging within department and the university; however, underrepresented minority students had slightly higher global sense of belonging than their counterparts. Future work should explore belonging across different contexts to reveal niches where efforts to increase sense of belonging may be concentrated.

Finally, our multivariate regression analyses implicated global belonging and general self­efficacy as significant predictors of departmental belonging. Although departmental belonging may have been higher than global or university belonging, it appears that having a higher global sense of belonging also has significant implications for how one may feel belonging in more

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

specific contexts. This suggests that a general, broad sense of inclusion (e.g., in the institution or community as a whole) influences and reinforces belonging within more specific contexts (i.e., within one’s home department). Although it makes sense to support department belonging, interventions aimed at increasing students’ global sense of belonging may have downstream effects on university, department, potentially other group­level sense of belonging. The findings suggest that belonging reinforces itself and provides support across a wide range of life situations. Future research may explore the relationship between belonging context to establish the directionality of this relationship. Additionally, the significance of general self­efficacy in predicting departmental belonging suggests that possessing confidence in one’s ability to succeed plays a role in how students feel belonging within departments. It is possible that interventions focusing on bolstering general self­efficacy and global sense of belonging of college students may help to generate stronger sense of belonging on campus.

Peer support did not emerge as a significant factor in our main analysis of sense of belonging, contrary to previous research (McBeath et al., 2018; Kiefer et al., 2015). However, as we zoomed out of department into a broader context through our exploratory analyses, it was revealed that peer support and loneliness have varying levels of associations depending on the belonging context that is being measured. These findings reinforce the importance of context­specific belonging. In a global sense, social relationships (loneliness, perceived social support, etc.) are an integral part of identity (Roberts, 2007).

The findings of this study must be seen considering some limitations. Firstly, we faced restrictions that came with the small sample size; as informed by our a priori power analysis, our sample size (n = 138) may not have been sufficiently powered to detect some of our smaller effects of interest. Further, this study had limited representation of minority demographic groups and diverse gender identities. Future research should strive to attain a larger sample that includes stronger representation of diverse populations to analyze group differences, including differences across sex or gender.

Secondly, the cross­sectional nature of the current study also limits our ability to cultivate a holistic view of sense of belonging as it may shift and evolve across time. To consider belonging as a fixed state among all students facilitates a lack of consideration for each student’s time, place, space, and unique experiences (Gravett & Ajjawi, 2022). There is a need to view belonging as a dynamic experience that evolves throughout a student’s time in higher education, as it may depend more on context and social environment (Allen et al., 2022).

Thirdly, we collected self ­ reported data from

participants who self­selected to engage in the study at one university. Collection of self­report data is liable to various biases, including but not limited to socially desirable responding and recall bias. At the time of conceptualization, alternative measurement methods, such as collecting observational data, did not emerge as better candidates to measure the variables of interest; however, the generalizability of the findings of this study is limited due to these factors.

Finally, to make positive change, belonging must be conceptualized by the students, and this study was unable to accomplish that. Researchers were unable to gather the necessary qualitative data to help understand what belonging means to students. Future studies should include qualitative input from students. Along with this, future studies should strive to focus on departments to allow for a greater understanding of students’ experiences to accommodate their needs.

Conclusion

The current study showed that sense of belonging varies across contexts (i.e., department, university, global), with students reporting a higher sense of belonging to their academic department than the other belonging contexts. We also identified global sense of belonging and general self­efficacy as significant predictors of belonging to one’s department, and exploratory analyses revealed social factors (i.e., loneliness and social support) as significant predictors of global belonging. Ultimately, belonging is not one­size­fits­all; it is a dynamic concept influenced by context, internal factors, and social factors, and these internal and social factors vary in importance depending on the context that is being measured. Universities should consider multi­level interventions to support their students belonging where efforts to strengthen global and departmental belonging may have positive effects on each other; the department level is the most immediate and modifiable context to begin these interventions. Future research should seek to include data from larger, more diverse samples, implement longitudinal, mixed­methods data collection, and collect insight on student’s own definitions of sense of belonging. The emerging need for sense of belonging research must be tailored to support the diverse lived experiences of students across institutional layers.

References

Allen, K. A., Kern, M. L., Rozek, C. S., McInerney, D. M., & Slavich, G. M. (2021). Belonging: A review of conceptual issues, an integrative framework, and directions for future research. Australian Journal of Psychology, 73(1), 87–102. https://doi.org/10.1080/00049530.2021.1883409

Anistranski, J. A., & Brown, B. B. (2021). A little help from their friends? How social factors relate to students’ sense of belonging at a large public university. Journal of College Student Retention: Research, Theory & Practice, 25(2), 305–325. https://doi.org/10.1177/1521025120985107

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

Arslan, G. (2021). School belongingness, well-being, and mental health among adolescents: Exploring the role of loneliness. Australian Journal of Psychology, 73(1), 70–80. https://doi.org/10.1080/00049530.2021.1904499

Budge, S. L., Domínguez Jr, S., & Goldberg, A. E. (2020). Minority stress in nonbinary students in higher education: The role of campus climate and belongingness. Psychology of Sexual Orientation and Gender Diversity, 7(2), 222–229. https://doi.org/10.1037/sgd0000360

Bulus, M. (2023). Pwrss: Statistical power and sample size calculation tools. R package version 0.3, 1.

Duran, A., Dahl, L. S., Stipeck, C., & Mayhew, M. J. (2020). A critical quantitative analysis of students’ sense of belonging: Perspectives on race, generation status, and collegiate environments. Journal of College Student Development, 61(2), 133–153. https://doi.org/10.1353/csd.2020.0014

Evans, R., Nagoshi, J. L., Nagoshi, C., Wheeler, J., & Henderson, J. (2017). Voices from the stories untold: Lesbian, gay, bisexual, trans, and queer college students’ experiences with campus climate. Journal of Gay & Lesbian Social Services, 29(4), 426–444. https://doi.org/10.1080/10538720.2018.1378144

Garcia, C. E. (2019). “They don’t even know that we exist”: Exploring sense of belonging within sorority and fraternity communities for Latina/o members. Journal of College Student Development, 60(3). https://doi.org/10.1353/csd.2019.0029

Gillen-O’Neel, C. (2021). Sense of belonging and student engagement: A daily study of first- and continuing-generation college students. Research in Higher Education, 62, 45–71. https://doi.org/10.1007/s11162-019-09570-y

Glozah, F. N., & Pevalin, D. J. (2017). Psychometric properties of the perceived social support from family and friends scale: Data from an adolescent sample in Ghana. Journal of Child and Family Studies, 26(1), 88–100. https://doi.org/10.1007/s10826-016-0535-1

Gravett, K., & Ajjawi, R. (2022). Belonging as situated practice. Studies in Higher Education, 47(7), 1386–1396. https://doi.org/10.1080/03075079.2021.1894118

Gull, M. (2016). Self-efficacy and mental health among professional students: A correlational study. International Journal of Modern Social Sciences, 5(1), 42–51. Harris, P. A., Taylor, R., Minor, B. L., Elliott, V., Fernandez, M., O’Neal, L., McLeod, L., Delacqua, G., Delacqua, F., Kirby, J., & Duda, S. N. (2019). The REDCap consortium: Building an international community of software platform partners. Journal of Biomedical Informatics, 95, 103208. https://doi.org/10.1016/j.jbi.2019.103208

Harris, P. A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., & Conde, J. G. (2009). Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. https://doi.org/10.1016/j.jbi.2008.08.010

Healthy Minds Network & American College Health Association (HMN & ACHA). (2020). The impact of COVID-19 on college student well-being. https://healthymindsnetwork.org/wp-content/uploads/2020/09/Healthy_ Minds_NCHA_COVID_Survey_Report_FINAL.pdf

Hughes, M. E., Waite, L. J., Hawkley, L. C., & Cacioppo, J. T. (2004). A short scale for measuring loneliness in large surveys: Results from two populationbased studies. Research on Aging, 26(6), 655–672. https://doi.org/10.1177/0164027504268574

Kiefer, S. M., Alley, K. M., & Ellerbrock, C. R. (2015). Teacher and peer support for young adolescents’ motivation, engagement, and school belonging. Rmle Online, 38(8), 1–18. https://doi.org/10.1080/19404476.2015.11641184

Knekta, E., & McCartney M. (2021). What can departments do to increase students’ retention? A case study of students’ sense of belonging and involvement in a biology department. Journal of College Student Retention: Research, Theory & Practice, 22(4), 721–742. https://doi.org/10.1177/1521025118788351

Korpershoek, H., Canrinus, E. T., Fokkens-Bruinsma, M., & de Boer, H. (2019). The relationships between school belonging and students’ motivational, socialemotional, behavioural, and academic outcomes in secondary education: A meta-analytic review. Research Papers in Education, 35(6), 641–680. https://doi.org/10.1080/02671522.2019.1615116

Kuh, G. D., Cruce, T. M., Shoup, R., Kinzie, J., & Gonyea, R. M. (2008). Unmasking the effects of student engagement on first-year college grades and persistence. The Journal of Higher Education, 79(5), 540–563. https://doi.org/10.1080/00221546.2008.11772116

Kuh, G., D., Kinzie, J., Buckley, J., A., Bridges, B., K., & Hayek, J., C. (2006). What matters to student success: A review of the literature. In “Commissioned Report for the National Symposium on Postsecondary Student Success: Spearheading a Dialog on Student Success.” Washington, DC: National Postsecondary Education Cooperative. Retrieved from

https://nces.ed.gov/npec/pdf/ Kuh_Team_Report.pdf

Lipson, S. K., Lattie, E. G., & Eisenberg, D. (2019). Increased rates of mental health service utilization by US college students: 10-year population-level trends (2007–2017). Psychiatric Services, 70(1), 60–63. https://doi.org/10.1176/appi.ps.201800332

MacNear, K. A., & Hunter, C. D. (2025). Racial identity and sense of belonging: Moderators of Black college students’ institutional race-related stress and anxious arousal. Cultural Diversity and Ethnic Minority Psychology, 31(1), 76–87. https://doi.org/10.1037/cdp0000609

McBeath, M., Drysdale, M. T., & Bohn, N. (2018). Work-integrated learning and the importance of peer support and sense of belonging. Education & Training, 60(1), 39–53. https://doi.org/10.1108/ET-05-2017-0070

Mellinger, C., Fritzson, A., Park, B., & Dimidjian, S. (2023). Developing the Sense of Belonging Scale and understanding its relationship to loneliness, need to belong, and general well-being outcomes. Journal of Personality Assessment, 0(0), 1–14. https://doi.org/10.1080/00223891.2023.2279564

Moeller, R. W., Seehuus, M., & Peisch, V. (2020). Emotional intelligence, belongingness, and Mental Health in college students. Frontiers in Psychology, 11. https://doi.org/10.3389/fpsyg.2020.00093

Morse, C. C., & Schulze, R. (2013). Enhancing the network of peer support on college campuses. Journal of College Student Psychotherapy, 27(3), 212–225. https://doi.org/10.1080/87568225.2013.798222

Procidano, M., & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal of Community Psychology, 11, 1–24. https://doi.org/10.1007/BF00898416

R Core Team (2024). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Reid, M., Jessop, D. C., & Miles, E. (2020). Explaining the negative impact of financial concern on undergraduates’ academic outcomes: Evidence for stress and belonging as mediators. Journal of Further and Higher Education, 44(9), 1157–1187. https://doi.org/10.1080/0309877X.2019.1664732

Richardson, T., Elliott, P., & Roberts, R. (2017). Relationship between loneliness and mental health in students. Journal of Public Mental Health, 16(2), 48–54. https://doi.org/10.1108/JPMH-03-2016-0013

Roberts, L. M. (2007). From proving to becoming: How positive relationships create a context for self-discovery and self-actualization. In Exploring positive relationships at work: Building a theoretical and research foundation (1st ed., pp. 29–45). Essay, Psychology Press.

Romppel, M., Herrmann-Lingen, C., Wachter, R., Edelmann, F., Düngen, H.-D., Pieske, B., & Grande, G. (2013). A short form of the General Self-Efficacy Scale (GSE-6): Development, psychometric properties and validity in an intercultural nonclinical sample and a sample of patients at risk for heart failure. GMS PsychoSocial-Medicine, 10 https://doi.org/10.3205/PSM000091

Salusky, I., Monjaras-Gaytan, L., Ulerio, G., Forbes, N., Perron, G., & Raposa, E. (2024). The formation and role of social belonging in on-campus integration of diverse first-generation college students. Journal of College Student Retention: Research, Theory & Practice, 26(2), 473–499. https://doi.org/10.1177/15210251221092709

Schwarzer, R., & Jerusalem, M. (1995). Generalized self-efficacy scale. In J. Weinman, S. Wright, & M. Johnston (Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs, 35–37. NFER-NELSON. https://doi.org/10.1037/t00393-000

Tinto, V. (2017). Through the eyes of students. Journal of College Student Retention: Research, Theory & Practice, 19(3), 254–269. https://doi.org/10.1177/1521025115621917

Zimmerman, T., & Parker III, E.T. (2023). Does financial stress influence students’ campus climate perceptions and sense of belonging? Journal of College Student Retention: Research, Theory & Practice, 0(0), 1–24. https://doi.org/10.1177/15210251231210855

Author Note

Kylie Sambirsky https://orcid.org/0009­0008­2090­6017

Kit Ganzle https://orcid.org/0009­0006­6504­8381

Sienna Russell https://orcid.org/0009­0006­5014­0927

Ian McGillicuddy https://orcid.org/0009­0008­8782­9861

Mandira Gowda https://orcid.org/0009­0009­3322­1188

Joey Levy https://orcid.org/0000­0002­4480­7549

Leah Teeters https://orcid.org/0000­0003­1442­9310

Winnie Zhuang https://orcid.org/0000­0001­7436­1958

We have no known conflicts of interest to disclose.

Sense of Belonging in Undergraduates | Sambirsky, Ganzle, Russell, McGillicuddy, Gowda, Markos, Levy, Teeters, and Zhuang

Funding for this research was provided by the Sapp Family through the Undergraduate Research Fellows Program at the Renée Crown Wellness Institute at the University of Colorado Boulder.

All authors played an equal role in conceptualization, research design, data collection, analysis, and interpretation. Sambirsky, Ganzle, and Russell played an equal role in substantive original writing. Levy, Teeters, and Zhuang played an equal role in editorial assistance and supervision.

Positionality Statement: Sambirsky, Ganzle, Russell, McGillicuddy,

Gowda, and Markos were undergraduate students at the time of conceptualization, data collection, and analysis. This project was designed by students with the intent of impacting the lives of other students. All authors acknowledge that their perspectives are influenced by their positions within the dimensions of this identity.

Correspondence concerning this article should be addressed to Kylie Sambirsky, Renée Crown Wellness Institute, 1135 Broadway, Boulder, CO, 80302, United States.

Email: Kylie.Sambirsky@colorado.edu

Emotion Regulation Strategies and Sleep Quality as Predictors of Depression in College Students

Luke Stanton and Colleen Georges* Department of Psychology, Rutgers University

ABSTRACT. Depression among college students has climbed sharply in recent years, underscoring the need to identify modifiable risks. Previous research has identified sleep quality (SQ) and emotion regulation (ER) strategies as important factors in the development and persistence of depression. This study investigated whether SQ, expressive suppression (ES), and cognitive reappraisal (CR) independently predicted depressive symptoms in a sample of college students, where ES is the inhibition of outward emotion expression and CR is reframing the meaning of a situation to alter its emotional impact. Data were drawn from the publicly available ECSMP dataset and included (n = 88) college students with a mean age of 23.69 years (SD = 2.14). Participants completed self­report measures of SQ using the Pittsburgh Sleep Quality Index, ER using the Emotion Regulation Questionnaire, and depressive symptoms using the Zung Self­Rating Depression Scale. A hierarchical linear regression analysis was conducted to examine the predictive value of SQ and ER strategies, while controlling for age and sex. The final model, F(5, 82) = 7.43, p < .001, explained 31.2% of the variance in depressive symptoms (R² = .31). Poorer SQ (β = 1.81, p < .001) and greater use of ES (β = 0.68, p = .001) were both significant independent predictors. CR was not significantly associated with depressive symptoms (β = ­0.23, p = .18). These findings emphasize the importance of both SQ and ER in understanding depression among college students and the potential efficacy for interventions focused on improving SQ and reducing reliance on ES.

Keywords: depression, college students, sleep quality, emotion regulation

Depression is a prevalent concern among college students with some surveys indicating that there has been up to a 134 percent increase in symptoms of depression in the past decade (Lipson et al., 2022). This upward trend in depressive symptoms is alarming, as it not only affects students’ academic performance, but also their overall quality of life (Liu & Wang, 2024). The transition to college life introduces various stressors, including academic pressures, social challenges, and increased independence, which can contribute to the development or exacerbation of depressive symptoms (Deng et al., 2022; Pedrelli et al., 2015). Importantly, increases are not uniform across all symptoms (Lipson et al., 2022). Pronounced rises have been observed in anhedonia, low mood, sleep disturbance, and concentration difficulties (Lipson et al., 2022). To frame the present study, we first define the emotion regulation (ER) strategies of interest as

well as sleep quality (SQ). We then summarize evidence on ER strategies and SQ, describe how their relation contributes to depressive symptoms, outline proposed mechanisms that follow from poor SQ and ER, and identify the specific gap addressed here.

Emotion regulation (ER) refers to the processes by which individuals influence the emotions they have, when they have them, and how they experience and express these emotions (Gross, 1998). Difficulties in ER have been consistently linked to higher levels of depression among college students (Cutuli et al., 2014). Specifically, maladaptive ER strategies, such as expressive suppression (ES), where individuals inhibit the outward expression of emotions, have been associated with increased depressive symptoms (Gross & John, 2003). Conversely, adaptive strategies like cognitive reappraisal (CR), which involve changing the way one thinks about a potentially emotion­eliciting event, are linked to lower

levels of depression (Gross & John, 2003; Meyer et al., 2012). These findings underscore the importance of ER in understanding and addressing depression in the college student population.

Sleep quality (SQ) is another critical factor influencing mental health among college students. Studies have shown that poor SQ is prevalent in this demographic, with significant associations found between sleep disturbances and increased depressive symptoms (Dinis & Bragança, 2018). Factors contributing to poor SQ include irregular sleep schedules, academic stress, and lifestyle choices common in college settings (Hershner & Chervin, 2014). These factors are often attributed to maladaptive coping behaviors such as irregular bed and wake times, late night screen exposure, caffeine and energy drink use, and variable work or study shifts (Exelmans & Van den Bulck, 2016; Hershner & Chervin, 2014; Lund et al., 2010). Moreover, SQ disturbances can impair cognitive functions and emotional processing, further exacerbating depressive symptoms (Palmer & Alfano, 2017). Addressing SQ is therefore essential in the context of mental health interventions for college students.

Evidence linking ER and SQ to depression shows that SQ relates most strongly to symptom clusters that include anhedonia, concentration and indecisiveness, and low mood with irritability (Dinis & Bragança, 2018; Palmer & Alfano, 2017). On the reward side, curtailed or irregular SQ is associated with blunted ventral striatal responses during reward anticipation and altered valuation, which maps onto reduced positive affect and anhedonia (Krause et al., 2017; Goldstein & Walker, 2014). In the cognitive domain, poorer SQ in student and community samples tracks with worse attention, working memory, and decision making, which aligns with concentration problems and indecisiveness (Killgore, 2010; Lim & Dinges, 2010; Lo et al., 2016).

For mood and irritability, sleep deprivation heightens negative affect, increases amygdala reactivity, and weakens functional connectivity with medial prefrontal regions that normally regulate emotion, which provides a plausible pathway to low mood and irritability in daily life (Goldstein & Walker, 2014; Yoo et al., 2007). Complementary models implicate immune and stress physiology as additional links between disturbed SQ and depressive symptoms, including upregulation of proinflammatory signaling and activation of stress response systems that bias affect toward negativity (Irwin & Opp, 2017; Krause et al., 2017).

Emerging research suggests a bidirectional relationship between SQ and emotional processing (Kahn et al., 2013). Poor SQ can impair the ability to effectively regulate emotions, leading to increased reliance on maladaptive strategies like ES (Palmer & Alfano, 2017).

Conversely, difficulties in ER can contribute to SQ disturbances, creating a cyclical pattern that heightens the risk of depression (Palmer & Alfano, 2017). This interplay underscores the value of examining SQ and ER together rather than in isolation.

Given the intertwined nature of SQ and ER in influencing depression, there is a pressing need for targeted interventions that concurrently address these domains. Although existing studies have explored the individual effects of SQ and ER on depression, few have examined their combined impact, particularly focusing on specific ER strategies like CR and ES. The current study aims to fill this gap by investigating how SQ and distinct ER strategies independently and collectively contribute to depressive symptoms among college students. By elucidating these relationships, this study seeks to inform the development of comprehensive, targeted interventions that can more effectively support the mental health of college students.

We hypothesized that poorer SQ and greater use of ES would each be associated with higher depressive symptoms, and that greater use of CR would be associated with lower depressive symptoms. We further expected that SQ and ES would explain unique variance in depressive symptoms when modeled together while adjusting for age and sex, with SQ showing the strongest association. As a secondary hypothesis, we anticipated that men would report higher mean ES than women, consistent with prior literature finding this relationship.

Methods

Participants and Procedure

Data were drawn from the publicly available emotion, cognition, sleep, and multi ­ model physiological signals (ECSMP) dataset (Gao et al., 2021). ECSMP is a single cohort, one time laboratory dataset collected at Southeast University in Nanjing, China during the academic year. All participants were enrolled students at Southeast University. The release includes self­report questionnaires, laboratory tasks, and multimodal physiology recorded during rest, emotion inductions, and cognitive testing. The sample for the present analyses consisted of (n = 88) healthy college students. Participants had a mean age of 23.69 (SD = 2.14) and were primarily women (n = 57) compared to (n = 31) men (Gao et al., 2021). Race and ethnicity were not recorded in the released ECSMP dataset. As a result, we cannot characterize the racial or ethnic composition of the sample. All participants were healthy volunteers with no known psychiatric or sleep disorders. After providing informed consent, participants completed standardized self­ report questionnaires during an in­lab visit as part of a larger study (Gao et al., 2021). The measures relevant

to this study included an assessment of depressive symptoms, an ER inventory, and a SQ index (Gao et al., 2021). All procedures were approved by the Institutional Ethics Committee for Clinical Research of Zhongda Hospital, affiliated with Southeast University (Approval No. 2019ZDSYLL073­P01) prior to data collection (Gao et al., 2021). All tables and figures are original to this manuscript. Instrument content is summarized with source citations rather than reprinted.

Measures

Depression

Depression levels were measured with the Zung Self­Rating Depression Scale (SDS; Zung, 1965). The SDS is a 20­item self­report scale assessing affective, psychological, and somatic symptoms of depression (e.g. “I feel down­hearted and blue”). Items are rated on a 4­point frequency scale from 1 (a little of the time) to 4 (most of the time). Ten items are reverse scored to control for acquiescence. A total SDS score is computed and typically multiplied by 1.25 to yield an index ranging from 25 to 100, with higher scores indicating greater depressive symptom severity. Scores from 25 to 49 are generally interpreted as falling in the nondepressed range, 50 to 59 as mild depression, 60 to 69 as moderate depression, and 70 or higher as severe depression. The SDS demonstrated strong internal consistency, α = .84, after reverse­scoring applicable items. The SDS has wellestablished validity and reliability in targeted samples (Biggs et al., 1978; Shafer, 2006).

Emotion Regulation Strategies

Habitual use of ES and CR was measured by the Emotion Regulation Questionnaire (ERQ; Gross & John, 2003).

The ERQ is a 10­item instrument with two subscales: ES (4 items; e.g. “I keep my emotions to myself”) and CR (6 items; e.g. “I change the way I’m thinking about a situation to make myself feel better”; Gross & John, 2003).

Participants rate their agreement with each statement on a 7­point Likert scale from 1 (strongly disagree) to 7 (strongly agree). Subscale scores are summed totals for ES and CR, with higher scores indicating greater habitual use of that strategy (Gross & John, 2003). The ERQ has demonstrated acceptable internal consistency in both the CR and ES subscales (Gross & John, 2003; Preece et al., 2020). Additionally, the CR subscale showed good internal consistency, Cronbach’s α = .78, as did the ES subscale, α = .78. In the present sample, the ERQ was used to quantify each participant’s typical use of CR and ES.

Sleep Quality

SQ was assessed with the Pittsburgh Sleep Quality Index (PSQI; Buysse, 1989). The PSQI is a widely

used self­report questionnaire that evaluates SQ and disturbances over a 1­month interval (Dinis & Bragança, 2018). It consists of 19 items covering seven components of sleep (i.e., subjective sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction).

These component scores are summed to yield a global PSQI score, ranging from 0 to 21, where higher scores indicate worse SQ (Buysse, 1989). The PSQI has demonstrated good reliability and validity for identifying sleep dysfunction in clinical and nonclinical samples (Mollayeva et al., 2016). In the present study, five PSQI component scores with sufficient variance (Sleep Latency, Sleep Duration, Sleep Disturbances, Hypnotic Use, and Daytime Dysfunction) yielded acceptable internal consistency (α = .69). In this study, we utilized the PSQI global score as an index of overall SQ for each participant.

TABLE 1

Means, Standard Deviations, and Pearson’s r Correlations Among Study Variables

TABLE 2

Hierarchical Linear Regression Predicting Depressive Symptoms From Emotion Regulation and Sleep Quality

Note. Sex coded as 0 = Male, 1 = Female. ERQ = Emotion

Analytic Plan

All statistical analyses were conducted using R version 4.6. Descriptive statistics and graph visualizations were used to evaluate the distributional properties of outcome variables; data transformations were considered where appropriate. We conducted a hierarchical multiple regression analysis to examine whether ES, CR, and SQ independently predicted depressive symptoms after controlling for demographic covariates in our sample. First, we calculated descriptive statistics and Pearson zero­order correlations among all primary variables, including depression, SQ, ES, CR, age, and sex. The hierarchical regression was conducted in two steps. In Step 1, age and sex were entered as covariates to control for their potential confounding effects.

In Step 2, ES, CR, and SQ (PSQI score) were added to evaluate their unique and combined contributions to depressive symptoms. This approach allowed us to assess whether ER strategies and SQ explained additional variance in depression symptoms beyond basic potential confounding factors. All analyses were conducted using two­tailed tests with α = .05. Table 1 presents descriptive statistics, and the correlation matrix in Table 2 summarizes the regression results, including standardized (β) coefficients, standard errors, t values, p values, R values, and model R² values. Effect sizes were interpreted using standardized beta coefficients (β) from the regression models and zero­order correlations (r). Proportion of variance explained was assessed using R² and ΔR² values for each step of the hierarchical regression

Results

Table 1 displays the sample’s descriptive statistics. Participants reported an average SDS depression score of M = 44.83 (SD = 9.03). The PSQI global score averaged M = 7.55 (SD = 2.12) on a scale of 0–21. Participants had an average ES score of M = 12.60 (SD = 4.62) and a CR score of M = 31.97 (SD = 5.01). In this sample, SDS scores ranged from 25.00 to 77.50 (range = 52.50), ES scores ranged from 4 to 25 (range = 21), and CR scores ranged from 18 to 42 (range = 24). Table 2 shows the zero­order Pearson correlations among all variables. Depressive symptoms were positively correlated with ES, r(86) = .33, p = .002, and poorer SQ, r(86) = .23, p = .031. CR was not significantly associated with depressive symptoms, r(86) = –.10, p = .368. Sex was significantly associated with ES, r(86) = .46, p < .001, such that men reported greater use of ES than women.

A hierarchical linear regression model was computed to examine the impact of ES, CR, and SQ on depressive symptoms, while controlling for covariates. In the first step, age and sex were entered into the regression model. This initial model was not significant, F(2, 85) = 0.45, p

= .640 and accounted for only 1.0% of the variance in depressive symptoms (R² = .01). Within this model, neither age ( β = –0.32, p = .500) nor sex ( β = 1.06, p = .610) significantly predicted depressive symptoms. In the second step, ES, CR, and PSQI total scores were added to the model. The overall model was significant, F (5, 82) = 7.43, p < .001, accounting for a total of 31.2% of the variance in depressive symptoms (R² = .31). The inclusion of the ER and SQ variables significantly improved the model fit (ΔR² = .30, p < .001). In this final model, ES (β = 0.68, p = .001) and poor SQ (β = 1.81, p < .001) were significant predictors of depressive symptoms, but CR was not significantly related to depressive symptoms (β = – 0.23, p = .180). Age and sex remained nonsignificant in the final model. See Table 2 for full model results.

Discussion

The present study examined whether SQ, ES, and CR independently predicted depressive symptoms in college students. As hypothesized, poorer SQ and greater use of ES were significantly associated with higher depression scores. These two variables each explained unique variance in depressive symptoms, highlighting their importance even when considered simultaneously. In contrast, CR was not significantly associated with depression in this sample. Each result is discussed below. The association between SQ and depression was substantial.

Students with higher PSQI scores, indicating worse self­reported sleep, had significantly greater depressive symptoms. This aligns with extensive literature linking sleep disturbance with depressive symptoms (Dinis & Bragança, 2018; Palmer & Alfano, 2017). In our regression model, SQ had the strongest standardized effect, reinforcing the view that sleep problems are a key correlate of depressive symptoms in college students. The bidirectional relationship between sleep and depression has been supported by both theoretical models and longitudinal studies (Kahn et al., 2013). Although our cross­sectional design prevents causal conclusions, the magnitude of the association suggests that sleep health should be a target for mental health interventions. Even in a nonclinical college population, variations in sleep accounted for a significant portion of depressive symptom variance. To combat this, practical strategies such as consistent sleep routines and reduced nighttime screen use may improve both sleep and mood (He et al., 2020; Okano et al., 2019).

Consistent with prior findings, ES was positively associated with depressive symptoms (Cutuli et al., 2014; Gross & John, 2003). Students who frequently inhibited their emotional expression reported more depressive

symptoms, even after accounting for SQ. Theoretically, ES may contribute to depression through physiological stress or unresolved negative affect, and impaired social support (Hu et al., 2014). Our results show that ES had an independent and maladaptive link to mood, suggesting that intervention programs might benefit from helping students shift away from this strategy.

Cognitive ­ behavioral and mindfulness ­ based interventions that promote emotional awareness and expression could help reduce reliance on ES and potentially alleviate depressive symptoms (Guendelman et al., 2017). Additionally, we found that men reported higher mean ES than women, which aligns with evidence that gender socialization norms discourage overt emotional expression among men (Gross & John, 2003). This finding highlights the potential importance of skills training that can help reduce habitual suppression and the need to address barriers to emotional expression among male students.

Contrary to our hypothesis, CR did not significantly predict depressive symptoms in this sample. Although CR is generally considered an adaptive strategy (Gross, 1998), its effect in our model was not significant (β = ­0.23, p = .18), and its correlation with depressive symptoms was small and nonsignificant (r = ­.10). These findings are in line with some studies showing weak or null associations between reappraisal frequency and depression, particularly in nonclinical populations (Aldao et al., 2010). One possible explanation is range restriction; participants in this study reported relatively high CR use on average, limiting the ability to detect differences. Another possibility is that frequency of CR may matter less than efficacy.

Some individuals may frequently attempt CR but do so ineffectively, or face circumstances in which CR is less helpful. Cultural differences may also moderate the effectiveness of CR (Hu et al., 2014). Ultimately, our results suggest that CR is not universally protective and that its benefits may depend on context, individual differences, or specific implementation. This may suggest that simply using CR often may not be sufficient to protect against depressive symptoms when students are also experiencing poor sleep and relying on ES. From a practical standpoint, interventions that focus solely on increasing CR may have limited impact in this context. Instead, strategies that reduce ES and improve sleep, or that help students use regulation strategies more flexibly and contextappropriately may be more clinically meaningful targets.

The results of this study have broader implications for understanding depression in college students. They underscore the need for multifactorial models that incorporate both physiological and psychological factors. Both ES and sleep disturbance predicted depressive

symptoms, which reinforces the idea that dysregulation across domains contributes to mood problems. Moreover, the fact that ES, but not CR, was a significant predictor of depressive symptoms supports findings that maladaptive strategies may have stronger associations with psychopathology than adaptive ones (Aldao et al., 2010).

This study has several limitations. Its cross­sectional nature prevents causal inference. Longitudinal research is needed to clarify whether changes in sleep or ES precede changes in depression. All variables were assessed via self­report, which may introduce bias. Additionally, the sample consisted of generally healthy college students, limiting generalizability to clinical populations. Although ES and SQ were meaningful predictors in this sample, other factors such as social support, stress, and cognitive flexibility may also play critical roles in depression and should be explored in future studies (LeMoult et al., 2023; Li et al., 2023; Zheng et al., 2024). Finally, because race and ethnicity were not recorded, we could not examine potential differences across racial or ethnic groups, which is a constraint on generality.

Future research should investigate how improving sleep and reducing ES affect mood in college populations. Randomized trials could test whether interventions that enhance sleep hygiene or promote emotional expression reduce depression symptoms. It would also be valuable to explore whether poor sleep moderates or mediates the relationship between ER strategies and depression. Further, studies could distinguish between the frequency and effectiveness of CR, potentially using experimental designs or ecological momentary assessment.

References

Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation strategies across psychopathology: A meta-analytic review. Clinical Psychology Review, 30(2), 217–237. https://doi.org/10.1016/j.cpr.2009.11.004

Biggs, J. T., Wylie, L. T., & Ziegler, V. E. (1978). Validity of the Zung Self-Rating Depression Scale. The British Journal of Psychiatry, 132(4), 381–385. https://doi.org/10.1192/bjp.132.4.381

Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28(2), 193–213. https://doi.org/10.1016/0165-1781(89)90047-4

Cutuli, D. (2014). Cognitive reappraisal and expressive suppression strategies role in the emotion regulation: An overview on their modulatory effects and neural correlates. Frontiers in Systems Neuroscience, 8, Article 175. https://doi.org/10.3389/fnsys.2014.00175

Deng, Y., Cherian, J., Khan, N. U. N., Kumari, K., Sial, M. S., Comite, U., Gavurova, B., & Popp, J. (2022). Family and academic stress and their impact on students’ depression level and academic performance. Frontiers in Psychiatry, 13, Article 869337. https://doi.org/10.3389/fpsyt.2022.869337

Dinis, J., & Bragança, M. (2018). Quality of sleep and depression in college students: A systematic review. Sleep Science, 11(4), 290–301. https://doi.org/10.5935/1984-0063.20180045

Exelmans, L., & Van den Bulck, J. (2016). Bedtime mobile phone use and sleep in adults. Social Science & Medicine, 148, 93–101. https://doi.org/10.1016/j.socscimed.2015.11.037

Gao, Z., Cui, X., Wan, W., Zheng, W., & Gu, Z. (2021). ECSMP: A dataset on emotion, cognition, sleep, and multi-modal physiological signals. Data in Brief, 39,

Article 107660. https://doi.org/10.1016/j.dib.2021.107660

Goldstein, A. N., & Walker, M. P. (2014). The role of sleep in emotional brain function. Annual Review of Clinical Psychology, 10, 679–708. https://doi.org/10.1146/annurev-clinpsy-032813-153716

Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of General Psychology, 2(3), 271–299. https://doi.org/10.1037/1089-2680.2.3.271

Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2), 348–362. https://doi.org/10.1037/0022-3514.85.2.348

Guendelman, S., Medeiros, S., & Rampes, H. (2017). Mindfulness and emotion regulation: Insights from neurobiological, psychological, and clinical studies. Frontiers in Psychology, 8, Article 220. https://doi.org/10.3389/fpsyg.2017.00220

He, J. W., Tu, Z. H., Xiao, L., Su, T., & Tang, Y. X. (2020). Effect of restricting bedtime mobile phone use on sleep, arousal, mood, and working memory: A randomized pilot trial. PLOS ONE, 15(2), e0228756. https://doi.org/10.1371/journal.pone.0228756

Hershner, S. D., & Chervin, R. D. (2014). Causes and consequences of sleepiness among college students. Nature and Science of Sleep, 6, 73–84. https://doi.org/10.2147/NSS.S62907

Hu, T., Zhang, D., Wang, J., Mistry, R., Ran, G., & Wang, X. (2014). Relation between emotion regulation and mental health: A meta-analysis review. Psychological Reports, 114(2), 341–362. https://doi.org/10.2466/03.20.PR0.114k22w4

Irwin, M. R., & Opp, M. R. (2017). Sleep health: Reciprocal regulation of sleep and innate immunity. Neuropsychopharmacology, 42(1), 129–155. https://doi.org/10.1038/npp.2016.148

Kahn, M., Sheppes, G., & Sadeh, A. (2013). Sleep and emotions: Bidirectional links and underlying mechanisms. International Journal of Psychophysiology, 89(2), 218–228. https://doi.org/10.1016/j.ijpsycho.2013.05.010

Killgore, W. D. S. (2010). Effects of sleep deprivation on cognition. Progress in Brain Research, 185, 105–129. https://doi.org/10.1016/B978-0-444-53702-7.00007-5

Krause, A. J., Simon, E. B., Mander, B. A., Greer, S. M., Saletin, J. M., Goldstein-Piekarski, A. N., & Walker, M. P. (2017). The sleep-deprived human brain. Nature Reviews Neuroscience, 18(7), 404–418. https://doi.org/10.1038/nrn.2017.55

LeMoult, J., Battaglini, A. M., Grocott, B., Jopling, E., Rnic, K., & Yang, L. (2023). Advances in stress and depression research. Current Opinion in Psychiatry, 36(1), 8–13. https://doi.org/10.1097/YCO.0000000000000831

Li, G., Li, Y., Lam, A. I. F., Tang, W., Seedat, S., Barbui, C., Papola, D., Panter-Brick, C., van der Waerden, J., Bryant, R., Mittendorfer-Rutz, E., Gémes, K., Purba, F. D., Setyowibowo, H., Pinucci, I., Palantza, C., Acarturk, C., Kurt, G., Tarsitani, L., Morina, N., … Hall, B. J. (2023). Understanding the protective effect of social support on depression symptomatology from a longitudinal network perspective. BMJ Mental Health, 26(1), e300802. https://doi.org/10.1136/bmjment-2023-300802

Lim, J., & Dinges, D. F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin, 136(3), 375–389. https://doi.org/10.1037/a0018883

Lipson, S. K., Zhou, S., Abelson, S., Heinze, J., Jirsa, M., Morigney, J., Patterson, A., Singh, M., & Eisenberg, D. (2022). Trends in college student mental health and help-seeking by race/ethnicity: Findings from the national Healthy Minds Study, 2013–2021. Journal of Affective Disorders, 306, 138–147. https://doi.org/10.1016/j.jad.2022.03.038

Liu, X., & Wang, J. (2024). Depression, anxiety, and student satisfaction with university life among college students: A cross-lagged study. Humanities and Social Sciences Communications, 11, Article 1172. https://doi.org/10.1057/s41599-024-03686-y

Lo, J. C., Ong, J. L., Leong, R. L. F., Gooley, J. J., & Chee, M. W. L. (2016). Cognitive performance, sleepiness, and mood in partially sleep-deprived adolescents: The need for sleep study. Sleep, 39(3), 687–698. https://doi.org/10.5665/sleep.5552

Lund, H. G., Reider, B. D., Whiting, A. B., & Prichard, J. R. (2010). Sleep patterns and predictors of disturbed sleep in a large population of college students.

Journal of Adolescent Health, 46(2), 124–132. https://doi.org/10.1016/j.jadohealth.2009.06.016

Malinauskas, B. M., Aeby, V. G., Overton, R. F., Carpenter-Aeby, T., & Barber-Heidal, K. (2007). A survey of energy drink consumption patterns among college students. Nutrition Journal, 6, Article 35. https://doi.org/10.1186/1475-2891-6-35

Meyer, T. J., Smeets, T., & Merckelbach, H. (2012). The efficiency of reappraisal and expressive suppression in regulating everyday affective experiences. Psychiatry Research, 200(2–3), 964–969. https://doi.org/10.1016/j.psychres.2012.07.015

Mollayeva, T., Thurairajah, P., Burton, K., Mollayeva, S., Shapiro, C. M., & Colantonio, A. (2016). The Pittsburgh Sleep Quality Index as a screening tool for sleep dysfunction in clinical and non-clinical samples: A systematic review and meta-analysis. Sleep Medicine Reviews, 25, 52–73. https://doi.org/10.1016/j.smrv.2015.01.009

Okano, K., Kaczmarzyk, J. R., Dave, N., Gabrieli, J. D. E., & Grossman, J. C. (2019). Sleep quality, duration, and consistency are associated with better academic performance in college students. NPJ Science of Learning, 4, Article 16. https://doi.org/10.1038/s41539-019-0055-z

Palmer, C. A., & Alfano, C. A. (2017). Sleep and emotion regulation: An organizing, integrative review. Sleep Medicine Reviews, 31, 6–16. https://doi.org/10.1016/j.smrv.2015.12.006

Pedrelli, P., Nyer, M., Yeung, A., Zulauf, C., & Wilens, T. (2015). College students: Mental health problems and treatment considerations. Academic Psychiatry, 39(5), 503–511. https://doi.org/10.1007/s40596-014-0205-9

Preece, D. A., Becerra, R., Robinson, K., & Gross, J. J. (2020). The Emotion Regulation Questionnaire: Psychometric properties in general community samples. Journal of Personality Assessment, 102(3), 348–356. https://doi.org/10.1080/00223891.2018.1564319

Shafer, A. B. (2006). Meta-analysis of the factor structures of four depression questionnaires: Beck, CES-D, Hamilton, and Zung. Journal of Clinical Psychology, 62(1), 123–146. https://doi.org/10.1002/jclp.20213

Yoo, S. S., Gujar, N., Hu, P., Jolesz, F. A., & Walker, M. P. (2007). The human emotional brain without sleep: A prefrontal amygdala disconnect. Current Biology, 17(20), R877–R878. https://doi.org/10.1016/j.cub.2007.08.007

Zheng, K., Liu, Z., Miao, Z., Xiong, G. L., Yang, H. H., Zhong, M., & Yi, J. (2024). Impaired cognitive flexibility in major depressive disorder: Evidences from spatial-temporal ERPs analysis. Journal of Affective Disorders, 365, 406–416. https://doi.org/10.1016/j.jad.2024.08.092

Zung, W. W. (1965). A Self-Rating Depression Scale. Archives of General Psychiatry, 12(1), 63–70. https://doi.org/10.1001/archpsyc.1965.01720310065008

Author Note

This study involved a secondary analysis of the publicly available ECSMP dataset, which was collected with ethical approval from the Institutional Ethics Committee at Zhongda Hospital, Southeast University (Approval No. 2019ZDSYLL073­ P01). All data were de­identified and publicly accessible. This secondary analysis did not require additional IRB approval. The authors have no known conflicts of interest to disclose. No funding was received for this research.

Luke Stanton played a lead role in conceptualization, research design, data analysis and interpretation, and substantive original writing and a supporting role in editorial assistance. Colleen Georges played a lead role in supervision and a supporting role in editorial assistance.

Correspondence concerning this article should be addressed to Luke Stanton, Department of Psychology, Rutgers University, 53 Avenue E, New Brunswick, NJ 08901, United States.

Email: lns84@scarletmail.rutgers.edu

“Working Out” the Relationship Between Mental Health and Exercise: A Cross-Sectional Study of United States Adults

Robert A. Bourne1, Craig A. Warlick1,2, Parker Gardner2*, and Ashley C. T. Jones1*

1School of Psychology, University of Southern Mississippi

2Department of Psychological Sciences, Texas Tech University

ABSTRACT. Exercise is often associated with mental health, though it remains unclear which aspects of physical exercise independently relate to better mental health. This study used nuanced exercise variables to assess frequency, intensity, and type of exercise to establish exercise components linked to better mental health. Participants were United States adults recruited via Mechanical Turk. Of 2,825 survey respondents, 300 met inclusion criteria. Exercise behavior was measured with a 1­item frequency measure from the HealthyMinds Network, the 7­item International Physical Activity Questionnaire – Short Form, and a 3­item exercise type measure. Mental health outcomes were assessed with the Center for Epidemiologic Studies Depression Scale (CES­D), Satisfaction With Life Scale (SWLS), Generalized Anxiety Disorder Scale (GAD­7), and Secure Flourish scale (SF). We hypothesized that greater exercise dosage would predict better mental health, and that aerobic exercisers would report more favorable outcomes than those engaging in nonaerobic or mixed activities. MANOVA analyses supported these hypotheses. High­duration exercisers reported better overall mental health (n = 300, p < .001, η²p = .06), and higher frequency exercisers reported greater life satisfaction and flourishing ( p < .001, η²p = .12). No differences emerged for combined duration and intensity measures on depression and anxiety scores. However, aerobic exercisers reported less depression and anxiety than nonaerobic exercisers (n = 208,  p = .004, η²p = .07). These findings clarify that exercise—particularly aerobic activities—is linked to improved mental health. Clinicians and healthcare providers should consider exercise as an active component of treatment planning.

Keywords: physical activity, high­intensity exercise, aerobic exercise, mental health, International Physical Activity Questionnaire­Short Form

Physical exercise is associated with improved physical and mental health (e.g. Herbert et al., 2020). Despite the evidence for this, exercise is rarely used in modern treatment practices for mental health disorders (Rith­Najarian et al., 2019). In cases where physical activity treatment was implemented, exercise intervention had great impacts on mental health

Preregistration+ Badge earned for transparent research practices. The preregistration can be viewed at https://osf.io/djqhm/overview

for a variety of populations ranging from patients with severe mental health distress (e.g., schizophrenia) to those with more common mental health issues (e.g. depression and anxiety; Firth et al., 2015; Giesen et al., 2015; Huang et al., 2018). In particular, four main mental health markers decrease (i.e., depression and anxiety) or increase (i.e, life satisfaction and flourishing) with

moderate to high levels of exercise (Murphy et al., 2018; Rith­Najarian et al., 2019; Rogowska et al., 2020; Tan et al., 2020). In a study of nonstudent athlete university students, participants who were more actively engaged in exercise exhibited a decrease in their self­reported depressive symptoms over time (Herbert et al., 2020). Regarding life satisfaction, one study found that, among university employees, those who were more physically active reported greater well­being and more satisfaction with life than those who were less active (Zayed et al., 2018). When positive mental health has been examined, studies have shown a positive relationship between physical activity and overall happiness (Zhang & Chen, 2019). In a sample of university students, students who engaged in regular exercise had better overall physical health, mental health, increased happiness, and decreased risk of suicide (Brailovskaia et al., 2022; Murphy et al., 2018). Despite the breadth of these findings, research has yet to confirm what specific exercise components lead to improved mental health outcomes. The present study addresses this gap by exploring exercise frequency and intensity with established markers of psychopathology and well­being.

Exercise and Mental Health

There is often a misconception that one must spend copious amounts of time engaging in physical activity to achieve either physical or mental health benefits, but research does not support this notion. One literature review concluded that only approximately 150 minutes per week is necessary to achieve mental health benefits (Rogowska et al., 2020), either in small doses daily or in longer workouts spread over a few different days of the week. Tan and colleagues’ study of university students (2020) suggests exercising beyond this 150 ­ minute threshold is not associated with significantly different mental health benefits. Margulis and colleagues (2023) noted that exercising 90 minutes only once per week did not decrease college student stress or anxiety during an academic semester; however, implying that dosage does matter. Though these findings give some guidance as to the components of exercise that benefit mental health, these studies do not account for components such as exercise type.

Much of the research on exercise and mental health focuses on aerobic forms of exercise. These include, but are not limited to, activities such as running, cycling, swimming, skiing, and walking (Giesen et al., 2015; Li et al., 2018). These aerobic activities have shown great benefits to major mental health markers for a variety of populations—specifically, they have been shown to benefit several markers of mental health in older adults (Yao et al., 2021), decrease depression symptoms in

emerging adult and young adult populations (Li et al., 2024), and effectively decrease depressive symptoms in individuals diagnosed with major depressive disorder (Morres et al., 2019). In one study examining individuals meeting criteria for alcohol use disorder, participants who engaged in moderate to high levels of aerobic exercise activities such as cycling, skiing, and running found an increase in their mental health and even a decrease in their alcohol use disorder symptoms (Giesen et al., 2015). In addition, aerobic exercise participation is a primary determinant of global self­esteem among adolescent females, a population at risk for depression and eating disorders (Gothe et al., 2022). This research suggests strongly that aerobic exercise can increase one’s mental health, and that exercise may be a tool to help in situations where low self­esteem or comorbid diseases, such as substance use disorders or depression, are present.

Limitations of Prior Research

There are limitations to previous research that make understanding the relationship between exercise and mental health difficult. One major limitation is the lack of agreement on measurement methods of physical activity. It is common in literature to find psychological benefits associated with aerobic exercise, but there is a lack of data surrounding how other forms of exercise, such as high­impact activities of weightlifting or highintensity interval training (HIIT), which combines a variety of exercises, are related to psychological well­being, though these high impact forms of exercise are just as common as aerobic exercise.

Additionally, physical exercise is operationalized differently across studies, limiting the broad conclusions that can be drawn from this body of literature. For example, Rogowska and colleagues (2020) accounted for frequency only in their measurement of physical activity— measured in minutes per day, and days per week exercised. Conversely, Murphy and colleagues (2018) operationalized physical activity based on frequency, specifying activity level. Another study included a component of exercise type in their measure of physical activity, but did not account for intensity of the activity (Margulis et al., 2023). Disagreements about methods for measuring physical activity are further exacerbated by the difficulty of obtaining certain measurements from large, geographically dispersed population samples, such as VO2 max levels. These examples demonstrate several approaches to measuring exercise and exhibit methodological inconsistencies that make it difficult to generalize findings.

One further limitation of previous research is the use of overly­specific populations for data collection. Using specific populations can be helpful in understanding

interactions between physical activity and mental health among populations of interest, but make it difficult to generalize findings to other populations. One notable population overrepresented in previous research is students of colleges and universities. Several studies, including those discussed above, address physical activity and mental health using samples of only students (Herbert et al., 2020; Li et al., 2018; Margulis et al., 2021; Murphy et al., 2018; Rogowska et al., 2020; Tan et al., 2020). Many learning institutions implement physical activity as part of their curriculum, giving opportunities for scheduled, structured time that those outside of these populations might not have access to. Including more diversified samples could capture broader findings with greater generalizability and make treatment interventions relevant to more individuals.

A final limitation found in previous research is an overreliance on measures of psychopathology as opposed to measures of well­being. Individuals can be both experiencing symptoms of mental illness and experiencing well­being. Assessing both provides a more well­rounded, detailed examination of the complexities of mental health (Warlick et al., 2020). This assessment is necessary to better understand the relationship between exercise and mental health.

The Present Study

In the present study, we directly addressed each of the limitations presented above. To address the first limitation, we measured as many different facets of exercise as were feasible for our sample. Specifically, we examined engagement in physical exercise using three measures of exercise. We included two different measures of exercise that account for both the duration and intensity aspects of physical exercise. We also included items measuring which specific exercise type individuals engaged in most frequently. By accounting for exercise frequency, exercise intensity, and exercise type, we encompassed several aspects of physical activity in our operationalization of exercise that have not been accounted for in previous studies. To address the limitation of the recruitment of overly specific sample populations, we included individuals of diverse backgrounds across the United States who were not enrolled in an institution of higher education. Finally, to address the third major limitation, we included measures of both positive (life satisfaction and flourishing) and pathological (depression and anxiety) mental health.

Although we note, and attempt to address many of the limitations, we acknowledge that many studies have found largely positive relationships among our constructs of interest. To build upon these findings, we implemented a MANOVA framework to compare exercise groups on

several mental health outcomes. This framework allows us to account for multiple dependent variables (e.g. exercise components) simultaneously. We split our hypotheses into two different clusters, with cluster one (H1– H4) focusing on global relationships among exercise and mental health and cluster two (H5– H8) specifying exercise type and relationship to mental health.

1. We hypothesize that higher ­ dosage exercise group(s) will report lower levels of depression (H1) and anxiety (H2), and higher levels of flourishing ( H 3 ) and life satisfaction ( H 4 ) across both the HealthyMinds (HM) measure, which assesses frequency of exercise, and the International Physical Activity Questionnaire (IPAQ) measure, which assesses exercise duration and intensity.

2. We hypothesize that individuals engaging in aerobic­focused exercise activities such as running, cycling, swimming, walking, rowing, will report lower levels of depression (H5) and anxiety (H6), and higher levels of flourishing (H7) and life satisfaction (H8) compared to those engaging in nonaerobic or mixed exercise activities.

Method

Participants

Prior to data collection, the Institutional Review Board at the University of Southern Mississippi approved our study (IRB #22­741). We collected data from participants in the United States via a questionnaire using Amazon’s Mechanical Turk (MTurk) crowdsourcing platform (Warlick et al., 2018). Participants received $4.84 (which is the federal minimum wage for the estimated time of completion). Excluded participants did not receive any compensation. We followed our preregistered analytic plan by taking careful steps to prevent bot responses and duplicate responses. More specifically, participants were excluded if any of the following criteria were not met. Criterion 1: 18 years of age or older, Criterion 2: Citizen and/or resident of the United States, Criterion 3: Not enrolled in an institution of higher education, Criterion 4: Pass a region/state verification (i.e., early in the survey, participants are asked state of current residence in an open text response format and late in the survey, participants are asked to indicate the region of their current residence using a drop­down menu), and Criterion 5: Pass 80% of data integrity checks (combination of eligibility verification questions, direct queries, logical statements, qualitative responses, and an honesty check). Survey respondents who did not meet eligibility criteria (e.g., identifying as under the age of 18 or being enrolled in a higher education institution) were thanked for their time, and the survey was closed. Participants were

Exercise and Mental Health | Bourne, Warlick, Gardner, and Jones

limited to one survey attempt. In total, 2,825 respondents began the survey. After excluding participants who did not meet the eligibility criteria, 300 participants were eligible for data analysis. This screening process occurred over 20 batches between March 9, 2023 until July 3, 2023. We present full demographics in Table 1.

Measures

The Center for Epidemiologic Studies Depression Scale 10­item (CES­D; Radloff, 1977) measures depression. We used the 10­item version of the scale to measure depressive tendencies in participants (Andresen et al., 1994). The CES­D uses a 4­point response scale ranging from 0 (rarely or none of the time) to 3 (all of the time) where participants rate themselves on items such as “I felt depressed.” Higher composite scores indicate more depressive symptoms. Total score above 10 are considered “depressed.” Instrument internal consistency for all measures are listed in Table 2.

The Generalized Anxiety Disorder 7­Item Scale (GAD­7) measures anxiety. We used the seven­item version which includes items such as, “Over the last two weeks, how often have you been bothered by feeling nervous, anxious or on edge?” (Spitzer et al., 2006). The GAD­7 uses a 4­point response scale ranging from 0 (not at all) to 3 (every day). Composite scores range from 0 to 21 with higher scores indicating more anxiety. A score of 10 or above represents moderate anxiety, and a score of 15 or greater indicates severe anxiety.

The Satisfaction With Life Scale (SWLS) measures global life satisfaction using a five­item scale (Diener et al., 1985). The SWLS uses a 7­point response scale ranging from 1 (strongly disagree) to 7 (strongly agree) where participants rate themselves on items such as, “In most ways my life is close to my ideal.” Overall scores range from 5 to 35, with higher scores indicating more life satisfaction. The SWLS has demonstrated good internal consistency in its initial validation in the United States sample (α = .87; Pavot & Diener, 2008).

We used the 12­item Secure Flourish (SF) scale to measure our sample’s reported flourishing/overall wellbeing (VanderWheele, 2017). The SF consists of two items in each of six domains: happiness and life satisfaction, mental and physical health, meaning and purpose, character and virtue, close social relationships, and financial and material stability. The SF uses an 11­point response scale ranging from 0 to 10 for each item within the domains. In each item, the scoring anchors differ. For example, in one item, 0 represents “not satisfied at all,” and in another item, 0 represents “strongly disagree.” Scoring is calculated using an overall item average that ranges between 0 and 10. Although individual domain scores can be calculated, our preregistered analyses focused on the total SF score.

TABLE 1 Participant Demographics

Note. All demographics categories may not add up to 100% due to rounding, ability to select more than one category (e.g. race, gender, sexual orientation, etc.), and to participants declining to answer demographic questions.

a Age descriptives are listed as a mean and standard deviation.

b MENA denotes Middle Eastern and/or North African Ethnicity.

TABLE 2

Psychometric Properties

Note. CESD refers to the Center for Epidemiologic Studies Depression Scale. GAD-7 refers to the Generalized Anxiety Disorder Scale. SWLS refers to the Satisfaction with Life Scale. SF refers to the Secure Flourish scale. All correlations were significant at the p < .001 level.

Exercise and Mental Health | Bourne, Warlick, Gardner, and Jones

We used two different measures to assess engagement in physical exercise. The first measure was adapted from the Healthy Minds Network Database (Healthy Minds Network, 2019). This single item measures frequency of exercise engagement. Participants were asked “In the past 30 days, about how many hours per week on average did you spend exercising?” HM Exercise uses a 4­point scoring method, ranging from 1 (less than 1 hour) to 4 (5+ hours). The second measure used to quantify exercise engagement was the 7­item International Physical Activity Questionnaire­Short Form (IPAQ), which measures the frequency of physical activity that occurred within the last seven days, assessing physical activities as part of everyday life (Craig et al., 2003; IPAQ Research Committee., 2005). Unlike the HM Exercise, the IPAQ assesses exercise using minutes, while also integrating the intensity of the exercise over the past seven days. The frequency and duration of the exercise are collected for each intensity categorization (vigorous, moderate, or walking). This number is converted from minutes to MET minutes (multiples of the resting metabolic rates). MET minutes are then used to classify participants into one of three categories: inactivity, minimal activity, and health enhancing physical activity (HEPA).

Inactivity describes individuals who are “insufficiently active,” which means the participant’s MET minutes do not reach the threshold for either the HEPA active or the minimally active categories. To be classified as minimally active, participants must have at least three days meeting 20 minutes of vigorous activity, five or more days of moderately intensive activity of 30 minutes per day, or five or more days of activity that meet 600 MET minutes per week. To be classified as HEPA, participants must have at least three days meeting at least 1500 MET minutes or seven or more days that achieve at least 3000 MET minutes per week.

The last measure of physical exercise included three researcher­derived items that distinguish aerobic activities from high­intensity activities. These items were used to assess differences in the impact various forms of exercise have on mental health. These items asked how often (in days per week) a participant engaged in strictly aerobic training, strictly high­ intensity training, and mixed aerobic/high­intensity training over the last 10 days. Participants were grouped based upon their highest response. For example, if a participant reported they engaged in mixed exercise activities five times over the past 10 days, aerobic exercise one time, and nonaerobic exercise one time, the participant would be sorted into the mixed exercise group. If a participant scored a “10” on each of the items, that participant was omitted from analyses as they did not endorse a most frequent exercise type.

All participants completed demographic information first, then information regarding mental health

(including CES­D, GAD­7, SWLS, and SF), and finished the survey with exercise­related measures (including HM and IPAQ).

Analyses

In following our pre­registration plan, we calculated Cronbach’s alpha for our mental health variables. Alpha levels for all mental health variables exceeded DeVellis’ criteria. We present alpha levels, means, and standard deviations for these variables in Table 2. If the instrument’s alpha level met DeVellis’ (1991) alpha criteria, greater than 0.599, that instrument was included in our analyses. We then ensured all assumptions were met for inferential testing.

To test our first cluster of hypotheses (H1, H2, H 3, and H4), we used a series of one­way multivariate analysis of variance (MANOVA) tests to determine the effect of exercise classification on the four mental health variables using the HM grouping variable or the IPAQ grouping variable. For the second cluster of hypotheses (H5, H6, H7, and H8), we also used a series of MANOVAs to determine the effect of exercise type on the four mental health variables using the type of exercise grouping variable (aerobic, high ­ intensity, or mixed). The MANOVA framework allowed for comparisons of groups using more than one dependent variable (e.g. the exercise measures) on several mental health outcomes. All statistical procedures were computed in IBM SPSS Statistics.

Results

Exercise Classification and Mental Health

The sample consisted of 300 adults in the United States. A complete list of demographic statistics can be found in Table 1. Correlations among each independent variable are listed in Table 2. All correlations were significant at the p < .001 level. When testing for multicollinearity, we found that none of our variance inflation factors exceeded a value of 5, so we proceeded with our analyses. To test hypotheses that higher dosage exercise groups will report lower rates of depression (H1) and anxiety (H2), and higher rates of flourishing (H 3) and life satisfaction (H 4), we first conducted a MANOVA to determine the effect of exercise groups using the HM single item grouping (less than one hour, 2–3 hours, 3–4 hours, and 5+ hours) on the dependent mental health variables of depression, anxiety, life satisfaction, and flourishing. Significant differences were found among the exercise groups on our dependent variables, Wilks Lambda = .83, F(12, 751) = 4.63, p < .001, η2p =.06.

To avoid increased risk of Type 1 error, all post­hoc analysis were interpreted using the Bonferroni method, with each p value examined at the .0125 (.05/4 = .0125)

and Jones

level. Within the overall model, all individual ANOVA tests between exercise dosage and measures of mental health were significant. There were significant differences in CES­D total scores between HM exercise groups, F(3, 287) = 6.09, p < .001, η2p = .06. The 5+ hour group specifically reported significantly less depression than the one­hour group. There were also significant differences in GAD­7 total scores between HM exercise groups, F(3, 287) = 4.18, p = .006, η2p = .04. The 5+ hour group reported significantly less anxiety than the 2–3 hour group, p = .011. There were no significant differences among any other groups. Among measures of well­being, there was a significant relationship between SWLS scores and HM exercise groups, F(3, 287) = 7.05, p < .001, η2p = .07, in that the less than one hour group reported significantly lower levels of life satisfaction than 2–3 hour group (p <. 001), as well as the 5+ hour group (p < .001). There were no other significant differences among any other group. Lastly, there were significant differences in SF scores between HM exercise groups, F(3, 287) = 13.44, p <.001, η2p = .12. The less than one hour group also reported significantly lower levels of flourishing than all three other groups (p <. 001). There were no other significant differences among groups.

We then repeated this process with a different measurement of exercise: the IPAQ groups (i.e., inactive, minimally active, and HEPA). Given the scoring procedures of this multi­item measure, only 251 participants had data complete enough to be included in the analyses. As with the first measure of exercise, we found significant differences among the IPAQ­derived exercise groups on our mental health dependent variables, Wilks Lambda = .77, F(8, 490) = 8.44, p < .001, η2p =.12.

On our pathology­focused dependent variables, there were no significant differences regarding IPAQ category and depression, F(2, 248) = 0.17, p = .85, η2p = .001, or anxiety, F(2, 248) = 1.84, p = .16, η2p = .02. However, there were significant differences among our well­being measures. First, there were differences in SWLS ratings between IPAQ categories, F(2, 248) = 20.84, p < .001, η2p = .14. The inactive active group reported significantly less life satisfaction than the minimally active group and the HEPA group (p <. 001). Second, there was a significant difference in SF scores among IPAQ categories, F(2, 248) = 34.60, p < .001, η2p = .22. The inactive group reported significantly less flourishing than the minimally active group and the HEPA group (p <. 001). Additionally, the minimally active group reported significantly less flourishing than the HEPA group (p = .002).

Overall, both measures of exercise were able to detect differences regarding levels of well­being, mostly with the higher dosage of exercise groups over the

lowest dosage of exercise groups. However, only the frequency­based Healthy Minds item was able to detect differences in measures of pathology with the least active participants reporting the most psychological distress. Descriptive statistics of both exercise variables are available in Table 3.Visual representations of our findings are available in Figures 1 and 2.

Exercise Type and Mental Health

For our second cluster of hypotheses, we initially planned to conduct a one­way MANOVA to determine the effect of exercise type (aerobic, anaerobic, and mixed) on the dependent mental health variables of depression, anxiety, life satisfaction, and flourishing. However, our nonaerobic and mixed ­ exercise type groups did not have the number of participants to meet established rules for MANOVA sample size for roughly 80% power (VanVoorhis & Morgan, 2007).

As such, we re­sorted participants into one of two groups, either an aerobic dominant group (n = 119) or a nonaerobic dominant group ( n = 89) in which participants reported either anaerobic or mixed­exercise as their most frequently used exercise activity. Significant differences were found in our dependent variables according to exercise groups, Wilks Lambda = .93, F(4, 203) = 4.0, p = .004, η2p =.07.

The MANOVA on the pathology­focused variables was significant, F(1, 206) = 6.68, p = .01, η2p = .03, with those reporting engaging in aerobic exercise having lower scores of depression than those not engaging in primarily aerobic exercise. There was also a significant difference between aerobic and anaerobic exercise on GAD­7 scores, F(1, 206) = 12.83, p < .001, η2p = .06, with individuals engaging in aerobic exercise reporting lower scores for anxiety. There were no significant differences

Note. HEPA stands for health-enhancing physical activity. CESD refers to the Center for Epidemiologic Studies Depression Scale. GAD-7 refers to the Generalized Anxiety Disorder Scale. SWLS refers to the Satisfaction with Life Scale. SF refers to the Secure Flourish scale.

between aerobic and anaerobic exercise on measures of life satisfaction, F(1, 206) = 1.36, p = .25, η2p = .007, or flourishing, F(1, 206) = 0.43, p = .52, η2p = .002. These results indicate that individuals who identify as being aerobically focused report significantly less anxiety and less depression than those who identify using other exercise types. We present descriptive statistics in Table 3.

Discussion

Summary

Broadly, these results further explored existing findings surrounding exercise duration and type and the impact these two aspects have on the mental health of individuals who engage in deliberate physical activity. Our findings are mixed in support of our hypotheses.

The first finding for the duration­based HM measure was that for those who engaged in exercise 5+ hours per week, depression scores were significantly lower than those who reported exercise in the less than 1 hour per week range. Additionally, the 5+ hour group reported significantly less anxiety than those who exercised in the 2–3 hour range weekly. These findings support our hypotheses that depression and anxiety would be higher in groups that exercised less (H1 and H2). For this HM measure of exercise on pathological mental health markers, no other significant findings were found. This finding, like those of Margulis and colleagues (2021), suggests that although duration is important, it is not the only factor that determines the impact of exercise on pathology­focused mental health measures. In the wellbeing focused measures of mental health (satisfaction with life and flourishing), the less than 1 hour exercise group reported lower levels of life satisfaction than the 2–3 hour group and the 5+ hour group but not the 3–4 hour group. For the flourishing measure, those within the less than 1 hour group, reported significantly less flourishing than all other groups. This finding of diminished life satisfaction for those engaging in very low levels of exercise is consistent with findings in previous literature (Zhang & Chen, 2019) and supports our hypothesis that higher exercise groups would report higher levels of well­being as compared to the lower groups (H3 and H4).   To continue with our investigation on classification of exercise, we assessed the same dependent variables using another measure of exercise that accounted for both duration and intensity factors of exercise—the IPAQ. Our findings were not supportive of all our hypotheses for the IPAQ group comparisons. For the pathology­focused mental health variables of depression and anxiety, no significant differences existed between participants who were inactive, minimally active, and HEPA. These findings did not support our first two hypotheses in cluster 1 (H1 and H2). However, significant

differences did exist between these groups on measures of well­being, supporting the second two hypotheses of cluster 1 (H3 and H4). With regard to life satisfaction, the inactive groups reported significantly less life satisfaction than the HEPA and minimally active groups. This was also true for the flourishing variable, as those in the inactive groups reported significantly less flourishing than the minimally active and HEPA groups. For those who were minimally active, well­being was lower than all other groups on two different measures of well­being.

After assessing exercise classification using both strictly duration and a combination of duration and intensity, our next aim was to assess the relationship among specific exercise types and mental health variables. Our findings within this cluster of hypotheses were mixed. There were significant differences between the two groups on our pathology­focused measures of mental health. The aerobic exercise groups reported significantly lower levels of both depression and anxiety compared to the nonaerobic groups, supporting our hypotheses (H5 and H6). This finding also supports current literature that has shown aerobic forms of exercise impact depression (Noetel et al., 2024). There were no significant differences between those who engaged in aerobic or nonaerobic exercise on our well­being measures, as both groups reported similar results on both flourishing and satisfaction with life, which again did not support our last two hypotheses (H7 and H8).

Clinical Implications

Given these results, medical and behavioral health personnel should include questions about exercise in their screening procedures, alongside other routine measures of mental and physical health. One review of clinician trends in the United Kingdom revealed several barriers to promoting physical activity in treatment, including lack of knowledge regarding exercise benefits and screening tools (Woodhead et al., 2023). Further work on measure inclusion in treatment and research supporting benefits of physical activity could be an effective way to promote physical activity in treatment. For routine clinical assessment, the Healthy Minds single ­ item measure may be useful for its brevity, ease in scoring, and its ease in interpretation. For more detailed information, or more behaviorally oriented clinics, including both the Healthy Minds single­item and the multi­item IPAQ would provide greater insight into patient’s exercise status and mental health condition. Additionally, the accumulation of patient exercise data and well­being outcomes as part of routine outcome monitoring would provide more insight regarding each measure’s utility and associations (Rodrigues et al., 2022).

These results generally support the notion that

“physicians can prescribe” physical activity (Fortier & Morgan, 2022, p. 8562). Our findings show that this prescription could be especially relevant information for individuals who report levels of inactivity, or less than one hour of moderate exercise per week. Understanding and exploring specific exercise habits (especially duration, intensity, aerobic versus anaerobic, etc.), would be useful information for medical and behavioral health professionals in planning effective treatment. This may have particular utility for those reporting depression as pairing exercise with psychotherapy led to decreased depression (Zhang et al., 2024). Fortier and Morgan (2022) describe a more specific pathway that integrates exercise with Motivational Interviewing and SelfDetermination Theory that improved mental health outcomes among college students and adults (Fortier et al., 2011; McFadden et al., 2017).

To sustain exercise behaviors, an integrative approach, featuring individual, social, and environmental factors would provide even more aspects for consideration in behavioral implementation (Morouço et al., 2024). At the individual level, Rogowska and Morouço (2024) recommend using the Exercise Regulations Questionnaire­3 (Cocca et al. 2024; Markland & Tobin, 2004; Wilson et al., 2006) for assessing an individual’s beliefs about exercise given its psychometrics and reach across languages. Additionally, attention to social rhythms and routines (e.g., including exercise among other behaviors such as sleep, mealtimes, and nutrition) may increase self­efficacy and mental health (Huang et al., 2023). The sense of control one experiences when engaging in physical activity may foster positive mental health during turbulent situations, such as different waves of the COVID­19 pandemic (Precht et al., 2023). In summary, health care providers should assess, monitor, and consider prescribing physical activity, especially when there are matters of well­being and flourishing.

Strengths and Limitations

Despite this study’s mixed results based on our hypotheses, there are several strengths of this study. First, this study implemented rigorous sampling procedures. MTurk is a viable source of participants for health­related studies (e.g., Warlick et al., 2018). We implemented strict protocols to ensure quality data collection, as emphasized in previous literature (Chmielewski & Kucker, 2019). Our multi­step protocol led to a rejection rate of 89.38% of all submitted data. These steps provide confidence that our sample was, in fact, our targeted sample, and not a collection of bots, repeat participants, or inattentive participants. A second strength regarding our sample is that we used a diverse, noncollege student national sample, which is different from many prior studies

Exercise and Mental Health | Bourne, Warlick, Gardner, and Jones

regarding exercise or measurement which have focused on a specific subpopulation (e.g. college students; individuals with alcohol use; Giesen et al., 2015; Johnson et al., 2024). Thus, the diversity represented by our sample may be a more accurate representation of trends across a broad range of individuals with varying lifestyles and age categories.

Another strength of this study is in the multiple measures of exercise that were implemented. By accounting for exercise activity in both duration­only and duration plus intensity, we provide a stronger estimate of the importance of exercise on mental health. This is because exercise is multi­faceted in that both intensity and duration matter when measured. A final important strength of this study is its novelty in studying the impacts of aerobic and nonaerobic exercise. Much of the previous literature focused heavily on aerobic forms of exercise. By examining results based on the types of exercise individuals engage in, individuals and clinicians can be better informed about which types of exercise are likely to improve pathology­focused mental health outcomes.

Although our study contains strengths, it also contains limitations. Using a national sample is a strength, but a limitation of this study is the lack of racial diversity. Approximately 85% of the respondents to our survey reported being White. Intentional recruitment of racial and ethnic minorities may be increasingly beneficial to understand the impact any systemic racism and health disparities have had on other groups (e.g., more likely exposure to air pollution, fewer spaces for physical activity, or increased likelihood of needing to use public transportation; Gomez et al., 2021).

Additionally, 94% identified as cisgender and 86% of our sample identified as heterosexual. Intentional recruitment of sexual and/or gender minorities (SGM) would also be beneficial given well­documented health disparities for individuals with SGM identities across undergraduate students, graduate students, and the general population within the United States (e.g. Huffman et al., 2020; Lefevor et al., 2019; Ostermiller et al., 2025). Given the overwhelming evidence for exercise, investigations of positive coping during potentially stressful times may be useful for future investigations (e.g., Skidmore et al., 2024; Skidmore et al., 2025).

Separately, the broad inclusion criteria for this study limits the conclusions that can be made about specific populations of interest, such as college students. Focusing on measurement of exercise among college students may be particularly helpful as interventions could be embedded as part of structured courses, and the dosage of exercise could be optimized (e.g., Bourne et al., 2021; Bourne et al., 2023; Margulis, 2021). Researchers should continue to examine the differences in assessment

between these two exercise variables across different samples, particularly with marginalized identities (e.g., racial and ethnic minorities, SGM) to assess if these findings generalize. Lastly, although our single­item measure of exercise has been used in national surveys, including Healthy Minds, and multi­item measures are still largely preferred unless brevity is the first consideration (Karki et al., 2024). Additionally, self­report measures used in this study may not capture the precise detail that could be captured with biometric data.

Conclusion

Our findings illustrate that exercise largely has a contributory effect on mental health and well­being. However, exercise is a multi­faceted activity encompassing duration and intensity and it needs to be explored to best identify the dosages of exercise most beneficial for mental health. We found that among those who exercise, those who engage in aerobic exercise activities report lower levels of depression and anxiety when compared to their counterparts who focus primarily on nonaerobic, high­impact activities. The findings of this study should inform individuals of the association of aerobic­focused activities and pathology­focused mental health variables.

References

Andresen, E. M., Malmgren, J. A., Carter, W. B., & Patrick, D. L. (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). American Journal of Preventive Medicine, 10(2), 77–84. https://doi.org/10.1016/S0749-3797(18)30622-6

Bourne, R. A., Jones, A. C., & Warlick, C. (2021, September). Exploring the associations of exercise and mental health in a nationwide sample of undergraduate college students.[Poster presentation]. Mississippi Psychological Association Annual Conference. Biloxi, MS.

Bourne, R. A., Jones, A. C., & Warlick, C. (April, 2023). Exercise and mental health within a national population. [Poster presentation]. Southeastern Psychological Association Annual Conference. New Orleans, LA. Brailovskaia, J., Teismann, T., & Margraf, J. (2022). Positive mental health mediates the relationship between physical activity and suicide-related outcomes: A threeyear follow-up study. Current Psychology, 41(9), 6543-6548. https://doi.org/nwnz

Chmielewski, M., & Kucker, S. C. (2019). An MTurk crisis? Shifts in data quality and the impact on study results. Social Psychological and Personality Science https://doi.org/gf92b6

Cocca, A., Kopp, M., Greier, K., Labek, K., Cocca, M., & Ruedl, G. (2024). Validity, reliability, and invariance across sex of a German version of the Behavioral Regulation in Exercise Questionnaire. Frontiers in Psychology, 15, 1355928. https://doi.org/nwnt

Craig, C. L., Marshall, A. L., Sjöström, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E., Pratt, M., Ekelund, U., Yngve, A., Sallis, J. F., & Oja, P. (2003). International physical activity questionnaire: 12-country reliability and validity. Medicine and Science in Sports and Exercise, 35(8), 1381–1395. https://doi.org/bk5h6s

DeVellis, R. F. (1991). Scale development: Theory and applications (p. 121). Sage Publications Inc.

Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49(1), 71–75. https://doi.org/fqqbmr

Firth, J., Cotter, J., Elliott, R., French, P., & Yung, A. R. (2015). A systematic review and meta-analysis of exercise interventions in schizophrenia patients. Psychological Medicine, 45(7), 1343–1361. https://doi.org/f67d3r

Fortier, M. S., Hogg, W., O’Sullivan, T. L., Blanchard, C., Sigal, R. J., Reid, R. D., Boulay, P., Doucet, É., Bisson, É., Beaulac, J., & Culver, D. (2011). Impact of integrating a physical activity counsellor into the primary health care team: Physical activity and health outcomes of the physical activity

Exercise and Mental Health | Bourne, Warlick, Gardner, and Jones

counselling randomized controlled trial. Applied Physiology, Nutrition, and Metabolism 36(4), 503–514. https://doi.org/bxvdz8

Fortier, M. S., & Morgan, T. L. (2022). How optimism and physical activity interplay to promote happiness. Current Psychology, 41(12), 8559–8567. https://doi.org/nwnx

Giesen, E. S., Deimel, H., & Bloch, W. (2015). Clinical exercise interventions in alcohol use disorders: A systematic review. Journal of Substance Abuse Treatment, 52, 1–9. https://doi.org/npwc

Gómez, C. A., Kleinman, D. V., Pronk, N., Gordon, G. L. W., Ochiai, E., Blakey, C., Johnson, A., & Brewer, K. H. (2021). Addressing health equity and social determinants of health through healthy people 2030. Journal of Public Health Management and Practice, 27(Supplement 6), S249–S257. https://doi.org/gk7zsj

Gothe, N. P., Erlenbach, E., & Engels, H. J. (2022). Exercise and self-esteem model: Validity in a sample of healthy female adolescents. Current Psychology, 41(12), 8876–8884. https://doi.org/g7kqfd

Healthy Minds Network. (2019). Healthy Minds Study among Colleges and Universities, year (HMS 2019-2020) [Data set]. Healthy Minds Network, University of Michigan, University of California Los Angeles, Boston University, and Wayne State University. https://healthymindsnetwork.org/reserach/data-for-researchers

Herbert, C., Meixner, F., Wiebking, C., & Gilg, V. (2020). Regular physical activity, short-term exercise, mental health, and well-being among university students: The results of an online and a laboratory study. Frontiers in Psychology, 11. https://doi.org/gmz996

Huang, J., Nigatu, Y. T., Smail-Crevier, R., Zhang, X., & Wang, J. (2018). Interventions for common mental health problems among university and college students: A systematic review and meta-analysis of randomized controlled trials. Journal of Psychiatric Research, 107, 1–10. https://doi.org/gfb68p

Huang, Q., Wang, X., Ge, Y., & Cai, D. (2023). Relationship between self-efficacy, social rhythm, and mental health among college students: A 3-year longitudinal study. Current Psychology, 42(11), 90562. https://doi.org/nwnw

Huffman, J. M., & Warlick, C., Frey, B. B., & Kerr, B. A. (2020). Religiosity, spirituality, gender identity, and sexual orientation of sexual minorities. Translational Issues in Psychological Science. 6(4), 356–371. https://doi.org/gkghf7

IPAQ Research Committee. (2005). Guidelines for data processing and analysis of the International Physical Activity Questionnaire (IPAQ)-short and long forms. https://www.researchgate.net/file.PostFileLoader.html?id=5641f4c3614325 0eac8b45b7&assetKey=AS%3A294237418606593%401447163075131

Johnson, D. A. L., Borgogna, N. C., Ingram, P. B., Warlick, C., Spencer, S. D., Mims, C. E., Bunnell, K. L., & Nielsen, J. A. (2024). The Scrupulosity Obsessions and Compulsions Scale: A measurement of scrupulosity within an OCD framework. Journal of Obsessive-Compulsive and Related Disorders, 43, 100918. https://doi.org/ns9z

Karki, N., Warlick, C., Baimel, A., & Jong, J. (2024). Examining linear and curvilinear relationships among religiosity, spirituality, and mental health in eight countries. Psychology of Religion and Spirituality, 16(4), 388–396. https://doi.org/m7jn

Lefevor, G. T., Boyd-Rogers, C. C., Sprague, B. M., & Janis, R. A. (2019). Health disparities between genderqueer, transgender, and cisgender individuals: An extension of minority stress theory. Journal of Counseling Psychology, 66(4), 385. https://doi.org/gg6c38

Li, C., Yin, H., Zhao, J., Shang, B., Hu, M., Zhang, P., & Chen, L. (2018). Interventions to promote mental health in nursing students: A systematic review and meta-analysis of randomized controlled trials. Journal of Advanced Nursing, 74(12), 2727–2741. https://doi.org/npwm

Li, W., Liu, Y., Deng, J., & Wang, T. (2024). Influence of aerobic exercise on depression in young people: a meta-analysis. BMC Psychiatry, 24(1), 571. https://doi.org/nx5v

Margulis, A., Andrews, K., He, Z., Chen, W. (2021). The effects of different types of physical activities on stress and anxiety in college students. Current Psychology, 42, 5385–5391. https://doi.org/gj6w3k

Markland, D., & Tobin, V. (2004). A modification of the behavioral regulation in exercise questionnaire to include an assessment of amotivation. Journal of Sport Exercise Psychology, 26, 191–196. https://doi.org/gdm3z7

McFadden, T., Fortier, M. S., & Guérin, E. (2017). Investigating the effects of physical activity counselling on depressive symptoms and physical activity in female undergraduate students with depression: A multiple baseline single-subject design. Mental Health and Physical Activity, 12, 25–36. https://doi.org/gbjtmd

Morouço, P., Rogowska, A. M., & Behzadnia, B. (2024). Motivation for Physical

Activity-Volume II. Frontiers in Psychology, 15, 1533943. https://doi.org/10.3389/fpsyg.2024.1533943

Morres, I. D., Hatzigeorgiadis, A., Stathi, A., Comoutos, N., Arpin‐Cribbie, C., Krommidas, C., & Theodorakis, Y. (2019). Aerobic exercise for adult patients with major depressive disorder in mental health services: A systematic review and meta‐analysis. Depression and Anxiety, 36(1), 39–53. https://doi.org/gff2kj

Murphy, M. H., Carlin, A., Woods, C., Nevill, A., MacDonncha, C., Ferguson, K., & Murphy, N. (2018). Active students are healthier and happier than their inactive peers: The results of a large representative cross-sectional study of university students in Ireland. Journal of Physical Activity and Health, 15(10), 737–746. https://doi.org/gfcrhz

Noetel, M., Sanders, T., Gallardo-Gómez, D., Taylor, P., del Pozo Cruz, B., Van Den Hoek, D., Smith, J., Mahoney, J., Spathis, J., Moresi, M., Pagano, R., Pagano, L., Vasconcellos, R., Arnott, H., Varley, B., Parker, P., Biddle, S., & Lonsdale, C. (2024). Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials. BMJ, 384. https://doi.org/gtp3mz

Ostermiller, L., Anderson, A. R., Warlick, C., & Dahlen, E. (2024). Lifestyle behaviors and mental health outcomes in sexual and gender minority graduate students. Journal of American College Health, 73(5), 1963–1972. https://doi.org/ntbp

Pavot, W., & Diener, E. (2008). The Satisfaction With Life Scale and the emerging construct of life satisfaction. The Journal of Positive Psychology, 3(2), 137–152. https://doi.org/dgn5p5

Precht, L. M., Margraf, J., Stirnberg, J., & Brailovskaia, J. (2023). It’s all about control: Sense of control mediates the relationship between physical activity and mental health during the COVID-19 pandemic in Germany. Current Psychology, 42(10), 8531–8539. https://doi.org/grkxb9

Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. https://doi.org/b4z

Rith-Najarian, L. R., Boustani, M. M., & Chorpita, B. F. (2019). A systematic review of prevention programs targeting depression, anxiety, and stress in university students. Journal of Affective Disorders, 257, 568–584. https://doi.org/ggjs37

Rodrigues, F., Moutão, J. O. Ã. O., Teixeira, D., Cid, L., & Monteiro, D. (2022). Examining exercise motives between gender, age and activity: A first-order scale analysis and measurement invariance. Current Psychology, 41(1), 112–125. https://doi.org/nxhn

Rogowska, A. M., & Morouço, P. (2024). Motivations for physical activity. Frontiers in Psychology, 15, 1437220. https://doi.org/nwns

Rogowska, A. M., Pavlova, I., Kuśnierz, C., Ochnik, D., Bodnar, I., & Petrytsa, P. (2020). Does physical activity matter for the mental health of university students during the COVID-19 pandemic? Journal of Clinical Medicine, 9(11), 3494. https://doi.org/gjh9cx

Skidmore, S. J., & Lefevor, G. T. (2025). Understanding how coming out goes well for sexual and gender minorities. Journal of Homosexuality, 72(11), 2153–2175. https://doi.org/p4sj

Skidmore, S. J., Lefevor, G. T., Huynh, K. D., & Berg, C. O. (2024). Development and initial validation of scales for coming out vigilance and positive coming out responses. Sexuality & Culture, 28(2), 771–793. https://doi.org/gt6gtr

Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Löwe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166(10), 1092–1097. https://doi.org/d7z8rz

Tan, S. L., Jetzke, M., Vergeld, V., & Müller, C. (2020). Independent and combined associations of physical activity, sedentary time, and activity intensities with perceived stress among university students: Internet-based crosssectional study. JMIR Public Health and Surveillance, 6(4), e20119. https://doi.org/gmz97h

VanderWeele, T. J. (2017). On the promotion of human flourishing. Proceedings of the National Academy of Sciences, 114(31), 8148–8156. https://doi.org/gbsrrc

VanVoorhis, C. W., & Morgan, B. L. (2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology, 3(2), 43–50.  https://doi.org/gdqbfc

Warlick, C., Huffman, J. M., Poquiz, J., Moffitt-Carney, K. M., Krieshok, T. S., Delong, L., Schellenger, B., Leonard, J., & Nelson, J. (2020). A pilot investigation of positive psychology scales and DBT treatment outcomes. Journal of Mental Health Counseling 42(4), 356–373. https://doi.org/pqh8

Warlick, C., Richter, K. P., Catley, D., Gajewski, B. J., Martin, L. E., & Mussulman, L. M. (2018). Two brief valid measures of therapeutic alliance in counseling for

tobacco dependence. Journal of Substance Abuse Treatment, 86, 60–64. https://doi.org/hpsj

Wilson, P. M., Rodgers, W. M., Loitz, C. C., & Scime, G. (2006). “It’s Who I Am… Really!’ The importance of integrated regulation in exercise contexts. Journal of Applied Biobehavioral Research, 11(2), 79–104. https://doi.org/dcf876

Woodhead, G., Sivaramakrishnan, D., & Baker, G. (2023). Promoting physical activity to patients: A scoping review of the perceptions of doctors in the United Kingdom. Systematic Reviews, 12(1), 104. https://doi.org/qg59

Yao, L., Fang, H., Leng, W., Li, J., & Chang, J. (2021). Effect of aerobic exercise on mental health in older adults: A meta-analysis of randomized controlled trials. Frontiers in Psychiatry, 12, 748257. https://doi.org/nx5t

Zayed, K. N., Ahmed, Md. D., Van Niekerk, R. L., & Ho, W. K. Y. (2018). The mediating role of exercise behaviour on satisfaction with life, mental well-being and BMI among university employees. Cogent Psychology, 5(1), 1430716. https://doi.org/npwn

Zhang, Z., & Chen, W. (2019). A systematic review of the relationship between physical activity and happiness. Journal of Happiness Studies, 20(4), 1305–1322. https://doi.org/gf39gs

Zhang, G., Guo, D., & Zhang, Y. (2024). The use of exercise-assisted psychotherapy in China to treat depression among elderly solitary individuals. Current Psychology, 43(31), 26052–26063. https://doi.org/nwn2

Author Note

other files associated with this project. This research was previously presented at the Southeastern Psychological Association’s 2023 convention. Data collection and analyses were conducted when the second author was affiliated with the University of Southern Mississippi. The initial draft, and all subsequent revisions of this manuscript occurred when this author was associated with Texas Tech University. We have no conflicts of interest to disclose. The Institutional Review Board at the University of Southern Mississippi approved our study (IRB #22­741). This research was supported by the Drapeau Center for Undergraduate Research at the University of Southern Mississippi (Bourne PI; Warlick Supervising PI; Gardner N/A; Jones N/A). Informed consent was obtained from all participants. The datasets are not publicly available as public storage of the datasets was not included in participant informed consent documents.

Robert A. Bourne https://orcid.org/0009­0004­5435­2439

Craig A. Warlick https://orcid.org/0000­0003­1533­3003 Parker Gardner https://orcid.org/0009­0004­1889­4484

Ashley C. T. Jones https://orcid.org/0000­0002­5608­603X We preregistered our project in the Open Science Framework prior to collecting data (OSF; https://osf.io/djqhm/overview). Our OSF page allows access to primary analyses, statistical scripts, and

Conceptualization: Robert Austin Bourne, Craig Warlick; Data curation: Robert Austin Bourne; Formal analysis: Robert Austin Bourne; Funding acquisition: Robert Austin Bourne, Craig Warlick; Investigation: Robert Austin Bourne, Craig Warlick; Methodology: Robert Austin Bourne, Craig Warlick, Ashley C.T. Jones; Project administration: Craig Warlick; Resources: Craig Warlick; Software: N/A; Supervision: Craig Warlick, Ashley C.T. Jones; Validation: Robert Austin Bourne; Visualization: Robert Austin Bourne, Parker Gardner; Writing – original draft: Robert Austin Bourne, Craig Warlick, Parker Gardner, Ashley C.T. Jones; Writing –reviewing and editing: Robert Austin Bourne, Craig Warlick, Parker Gardner, Ashley C.T. Jones.

Correspondence concerning this manuscript should be addressed to Robert Austin Bourne Robert.Bourne@usm.edu

The Impact of Tardive Dyskinesia on Quality of Life in Individuals With Severe Mental Illness: A Meta-Analysis

Madelynn D. Loring*1, Trinity Lumbard*1, Brian Drwecki**1, and Kaylyn McAnally Star**2

1Department

of Psychology and Neuroscience, Regis University

2School of Rehabilitative and Health Science, Regis University

ABSTRACT. Tardive dyskinesia (TD) is an involuntary movement disorder caused by neuroleptics, a common class of medication used to treat patients with severe mental illness (SMI). TD is a cause of concern for the field of psychiatry as it results in greater healthcare costs and worsened treatment adherence for patients with SMI. Although recent generations of neuroleptics have a lower likelihood of causing TD, and treatments for this condition are available, research continues to suggest poorer quality of life in individuals with SMI who have TD compared to those who do not. We conducted a meta­analysis examining the impact of TD on quality of life in individuals with SMI. Six studies reporting on 16 effects with a total sample size of n = 4,220 were included in the final analysis. We found a small, negative effect of TD on quality of life (Cohen’s d = ­0.26, 95% CI [­0.37, ­0.15]). This effect was not significantly different between studies published before treatments for TD were available and those published afterwards. Furthermore, the asymmetry of the resulting funnel plot indicated that publication bias could be present, but further tests are needed to determine the extent of possible bias. This research highlighted the importance of studying TD in individuals with SMI to promote better quality of life overall. These findings may prompt further research on the impact of TD to examine how this effect can be diminished for better neuroleptic treatment adherence and well­being for those with SMI.

Keywords: tardive dyskinesia, quality of life, severe mental illness, neuroleptics, antipsychotic drugs

The National Institute of Mental Health (NIMH) estimates that nearly 15.4 million adults in the United States were diagnosed with a severe mental illness (also commonly referred to as serious mental illness) in 2022, representing 6.0% of all U.S. adults (2024). NIMH defines severe mental illness (SMI) as “a mental, behavioral, or emotional disorder resulting in serious functional impairment, which substantially interferes with or limits one or more major life activities” (2024). Chronic disorders like schizophrenia and bipolar disorder fall under the broader category of SMI (Galletly & Rigby, 2013; Rands et al., 2024). Previous research has shown a decrease in quality of life in patients with SMI (Berghöfer et al., 2020; Saavedra et al., 2023; Shumye et al., 2021). Although there is not a gold standard of medications or treatment plans to cure SMI, certain

medications can help reduce symptoms and allow patients with SMI to better function in society. One common class of medications prescribed for patients with SMI are neuroleptics, which are also known as antipsychotic drugs (Cornett et al., 2017; Gossman et al., 2019).

The use of neuroleptics in the treatment of SMI began in the early 1950s; however, just a few years later, in 1957, researchers reported in the German literature that this revolutionary form of medication could result in an involuntary movement disorder (Schonecker, 1957; Waln & Jankovic, 2013). This disorder is now known as tardive dyskinesia (TD), emphasizing the delayed (or tardive) onset of abnormal movements (or dyskinesia) after the use of neuroleptic drugs (Faurbye et al., 1964). These movements can present as choreiform (i.e., rapid and unpredictable, presenting primarily

in proximal muscles), athetoid (i.e., slow, writhing, impacting distal muscles), dystonic (i.e., sustained muscle contractions), or stereotypic (i.e., predictable in pattern) or it can present as a combination of these (Jeste & Wyatt, 1982). The abnormal movements of TD fluctuate in intensity and are dependent on emotional states, showing an increase with emotional arousal, decreasing in moments of relaxation, and disappearing during sleep. Spontaneous fluctuations can occur day to day and may even be apparent over hours or even minutes (Sachdev, 2000). Continued neuroleptic use worsens TD, which is irreversible in many patients, and often persists long after stopping medication (Cornett et al., 2017; Vasan & Padhy, 2023). According to a previous review on medication­induced TD, 15–30% of individuals who receive long­term treatment with antipsychotics experience TD, with occurrence rates varying depending on antipsychotic drug class and ranging from 32.4% for first­generation antipsychotics to 13.1% for second­generation antipsychotics (Cornett et al., 2017).

The initial approach to treating TD typically involves physician ­ supervised adjustments to the antipsychotic regimen, such as modifying the dosage, switching the neuroleptic, discontinuing medication, or increasing doses to temporarily suppress TD symptoms (Ricciardi et al., 2019). Tetrabenazine, developed in 2008, was the first drug used to treat TD (Cloud et al., 2014; Kaur et al., 2016; Sharma, 2009), but has since been supplanted by valbenazine and deutetrabenazine due to their more favorable side effect profiles (Ricciardi et al., 2019; Vaidyanathan & Jaiswal, 2020). However, there are currently limited findings on the impact of these drugs on patient quality of life (Patel et al., 2019; Touma & Scarff, 2018).

Given the complexities of managing TD symptoms, it is important to consider the broader implications of TD on patient outcomes and healthcare systems. TD places significant burdens on healthcare utilization and associated costs (Carroll & Irwin, 2019) and dramatically impacts treatment adherence (Jain et al., 2023). Specifically, over one­third of patients who took part in a study conducted by Jain and colleagues (2023) reported reducing (48.5%) or stopping (39.3%) their antipsychotic medication, and 35.7% of respondents indicated that they stopped visiting their clinician to receive treatment for their underlying psychotic condition. It is paramount to further research the negative impact of TD on quality of life, or general satisfaction obtained from life, for those with SMI (American Psychological Association, n.d.). SMI alone is associated with negative physical, social, and mental well­being (Berghöfer et al., 2020), and TD further exacerbates

this worsened quality of life (Caroff, 2019; Owens, 2018; Patterson­Loomba et al., 2019). Often, TD results in adverse medical outcomes such as an increased fall risk, speech impediment, and difficulty swallowing (Owens, 2018). Furthermore, the symptomatology of TD is not only disabling, but socially stigmatizing to its sufferers, with some patients reporting depression and suicidal ideation and attempts as a result of their symptoms (Caroff, 2019; Patterson­Loomba et al., 2019).

A study conducted by McEvoy and colleagues (2019) on 416 patients with SMI found that patients with TD had lower quality of life scores when utilizing the 12­Item Short Form Health Survey (SF­12v2) and the 36 ­ Item Short Form Health Survey (SF­ 36v2). In contrast, a recent study conducted by Usman and colleagues (2023) on 80 patients with SMI reported no significant association between quality of life and TD as measured by four domains (physical, psychological, social, and environmental) of the World Health Organization Quality of Life Scale (WHOQOL­BREF). Therefore, although some evidence suggests that TD worsens quality of life in an already vulnerable group, research on this subject is not widespread, and there is considerable variability among results.

When evaluating the results of clinical research, it is important to consider whether study sponsorship and source of funding might influence reported outcomes. After reviewing 162 randomized, double­blind, placebocontrolled studies, Perlis and colleagues (2005) observed that author conflict of interest, which is prevalent in psychiatric clinical trials, was linked to an increased likelihood of reporting a drug to be more effective than placebo. Similarly, Riaz and colleagues (2015) reported in an analysis of 226 trials that industry ­ sponsored studies were nearly four times more likely than nonindustry­sponsored studies to produce positive results. Given that pharmacological treatments for TD have only recently been made available, and previous work has shown that funding source can bias reported outcomes, examining these variables as potential moderating effects is vital to understanding the true impact of TD on patient quality of life (Cloud et al., 2014; Davidson, 1986; Lexchin et al., 2003; Riaz et al., 2015).

In response, we conducted a meta ­ analysis to investigate the current state of the data measuring the direct impact of TD on the quality of life with patients with SMI to provide a bias­corrected estimate of its effect across the extant literature. Since the use of antipsychotics is a known risk factor for developing TD (Cornett et al., 2017), and has been shown to impact quality of life despite the availability of treatment (Caroff et al., 2020; Rekhi et al., 2022), we hypothesized that those with SMI and TD would report worsened quality of life compared

to their counterparts with SMI but no TD, indicating that TD uniquely contributes to reductions in quality of life beyond the known impacts of SMI on quality of life. We also examined potential moderators of this effect, including the availability of treatment for TD, and source of study funding due to concerns about publication bias. Greater understanding of this problem will provide substantial benefit for both psychologists working with clients who are experiencing worsened quality of life and the psychological science community who aim to improve the general well­being of the population at large.

Method

We searched PubMed/MEDLINE, Web of Science, and PsycINFO databases using the following search terms: (“Severe Mental Illness” OR “Bipolar Disorder” OR Schizophrenia OR Psycho* OR “Serious Mental Illness” OR Schizoaffective OR Mania OR “psychotic depressive” OR “psychotic depression” OR “psychiatric”) AND (“Tardive Dyskinesia” OR “drug­induced Tardive Dyskinesia”) AND (“quality of life” OR “life quality” OR “Health related quality of life” OR well­being OR HRQoL) to identify studies to include in the metaanalysis. We conducted the initial search in July 2024 and again in February 2025 to ensure that no new reports on the subject had been published.

We included studies that reported on the impact of having TD on quality of life in individuals with SMI, which was defined as those with schizophrenia spectrum disorders, bipolar disorders (I/II/cyclothymic), or severe mood disorders requiring treatment with antipsychotics. Studies reporting on TD in participants with Alzheimer’s or other forms of dementia were not included because these conditions are classified as progressive brain diseases, not mental illnesses (Tampi et al., 2016). We operationalized tardive dyskinesia by either the presence of a clinician diagnosis of the condition, or by an elevated Assessment of Involuntary Movement Scale (AIMS) score, which has been used extensively to identify the presence of TD (Vaidyanathan & Jaiswal, 2020). We did not include studies that reported on extrapyramidal symptoms1 generally, or if AIMS was measured, but not used to create categorical groupings of the presence of TD.

Studies were included if they reported on validated quality of life measures for both those with TD and those without. Quality of life was operationalized with the following scales: WHOQOL­BREF (Skevington et al., 2004), EuroQol 5­Dimension 5­Level (EQ­5D­5L; Feng 1Extrapyramidal symptoms are a broader category of medicationinduced movement disorders, which include Tardive Dyskinesia. Extrapyramidal symptoms are typically caused by antipsychotic medications and have both acute and chronic manifestations (D’Souza et al., 2023).

et al., 2021), Quality of Life Enjoyment and Satisfaction Questionnaire Short Form (Q­LES­Q­SF; Stevanovic, 2011), Quality of Life Scale (QLS; Heinrichs et al., 1984), SF­12 (Salyers et al., 2000), SF­36v2 (Ware & Sherborne, 1992), and Schizophrenia Care and Assessment Program Health Questionnaire (SCAP­HQ; Lehman et al., 2003). Additionally, we excluded studies that did not have an English translation. If studies examined the effect of interest, but did not report sufficient data for analysis, the authors were contacted in an attempt to include relevant data. The initial pool of articles (k = 1,348) was screened by two researchers to eliminate duplicates (k = 174). The remaining article titles and abstracts were then screened according to predetermined protocol ( k = 1,174). Articles that passed title and abstract screening (k = 80) then went through a full text review to determine whether they met inclusion criteria. In the case of disagreement, articles were discussed amongst the researchers and protocol was updated. Agreement was good amongst researchers (92.5% agreement). Articles meeting the inclusion criteria, but lacking

sufficient data for inclusion, were flagged and the authors contacted. The screening and review process identified nine articles that met inclusion criteria. Articles (k = 5) where data access was not provided or not feasible given the scope of this study following contact with the authors were excluded. One additional article was excluded because it reported on retroactive analysis on a study whose sample was already included in the analysis. Thus, this analysis examines six studies reporting on 16 total effects. The flow of articles through the selection process is outlined in Figure 1.

To the fullest extent possible, our process aligned with the PRISMA guidelines for systematic review (Page et al., 2021). As this was an analysis conducted on previously published research data, no IRB approval was required to conduct this study. This study was not pre­registered, nor was any financial support received.

Moderator Coding

Included studies were coded for two moderating variables: treatment availability and funding source. As evidence­based treatment protocols for TD were released in 2013 (Cloud et al., 2014), and current treatments for TD have been proven to reduce abnormal movements whilst demonstrating favorable safety profiles (Correll et al., 2024; Golsorkhi et al., 2024; Sajatovic et al., 2022), we hypothesized that treatment availability would lessen TD’s impact on quality of life. If study data was collected before 2013, we categorized it as ‘treatment unavailable’ (k = 5); studies collected 2013 onward were categorized as ‘treatment available’ (k = 11). Previous research shows that funding source impacts the outcomes of clinical studies (Davidson, 1986; Lexchin et al., 2003; Riaz et al., 2015), and thus we hypothesized pharmaceutical funding would have a measurable impact on TD’s effect on quality of life. Studies that reported funding from a pharmaceutical company were categorized as ‘pharmaceutical company funded’ (k = 10), and studies funded by a source other than a pharmaceutical company (e.g., independent research foundation, national grant) were coded as ‘otherwise funded’ (k = 6). In line with the aforementioned research, we decided to investigate funding source and treatment availability as moderators prior to conducting the meta­analysis.

Statistical Analysis

Multilevel Meta-Analysis

Effect sizes for the relationship between TD and quality of life were extracted from all included studies. We adopted Cohen’s d as the effect size metric in the present meta­analysis, as we were interested in finding the overall effect of TD on quality of life (Cohen, 1988). If articles did not report Cohen’s d, we derived the effect size TABLE 1 Demographics, Moderator Classifications, and Other Pertinent

et al. (2008)

Ascher-Svanu et al. (2008)

SAD, SCZF

Ascher-Svanu et al. (2008) SCZ, SAD, SCZF

Ascher-Svanu et al. (2008) SCZ, SAD, SCZF

McEvoy et al. (2019)

McEvoy et al. (2019)

McEvoy et al. (2019)

McEvoy et al. (2019)

et al. (2022)

MDD, SCZ

MDD, SCZ

MDD, SCZ

SCZ

Tanner et al. (2023) SCZ, SAD, Mood DO, Other Psychiatric Disorders

et al. (2023)

Health State VAS

SAD, Mood DO, Other Psychiatric Disorders

Usman et al. (2023) SCZ, Shizotypal and Delusional Disorders, Mood DO

-BREF Physical Health

Usman et al. (2023) SCZ, Shizotypal and Delusional Disorders, Mood DO WHOQOL -BREF Psychological

Usman et al. (2023) SCZ, Shizotypal and Delusional Disorders, Mood DO WHOQOL -BREF Social Relationships

Usman et al. (2023) SCZ, Shizotypal and Delusional Disorders, Mood DO WHOQOL -BREF Environment

from computed effect sizes (e.g., Pearson’s r) and tests of difference (e.g., t ratios). Because we had included studies which reported multiple effect sizes for the relationship between TD and quality of life, we utilized a three­level meta­analysis. The three­level meta­analytic method specifically separated variance in effect sizes between studies and variance in effect sizes within studies, which allowed for data retention of multiple effects from within the same study. The meta­analysis yielded an averaged point estimate of our main effect corrected for sampling error with its corresponding variability estimates. Our analysis was performed in R using code adapted from McAnally and Hagger (2023; see the code for the multilevel meta analysis and moderator analysis on OSF at https://osf.io/az826/overview ).

Moderator Analysis

We tested the effects of two key moderator variables, treatment availability and study source funding, on the averaged effect of TD on quality of life. The effects of the identified moderator variables on effects between TD and quality of life were tested in a series of metaregression analyses in which the effect size between TD and quality of life was regressed on the moderator variables. The analyses were implemented using random effects multilevel meta­regression using the metafor package in R (Viechtbauer, 2010). The moderator codes for each included study are summarized in Table 1.

Assessment of Bias

The presence of bias was examined using a funnel plot, in which effect sizes were plotted against standard error, which is the published recommendation for representing study size (Sterne & Harbord, 2004). Asymmetry of funnel plot spread, identified through visual examination, indicates that bias, including publication bias, may have impacted the meta­analytic results (Sterne & Harbord, 2004).

Results

In total, six studies containing 16 total effects were included in the analysis, with n = 4,220 participants (41.6% female) with an average age of 45.25. We found a negative effect of TD on quality of life in individuals with SMI (Cohen’s d = ­ 0.26, 95% CI [ ­ 0.37, ­ 0.15], p < 0.001; see Figure 2). This effect is classified as small (.2) according to common effect size indices (Sullivan & Fienn, 2012). These results showed high levels of heterogeneity (I² = 95.18%), with 69.8% reported as within­study heterogeneity.

Moderation tests for treatment availability showed no significant effect (k = 16, p = .24), indicating no difference in the effect of TD on quality of life in studies conducted before TD treatments were developed compared

to those conducted after treatments became available.

Moderation tests for study funding sources also showed no significant effect (k = 16, p = .610), indicating that the results of studies funded by pharmaceutical companies did not differ significantly from those funded otherwise.

The asymmetry of the funnel plot, particularly in the upper righthand corner, indicated that publication bias could be present (see Figure 3). However, this could also be due to the high heterogeneity of quality­of­life measures, and further tests of publication bias should be examined.

TABLE 1

Demographics, Moderator Classifications, and Other Pertinent Information on Included Effects

Note. BD, Bipolar Disorder; EQ-5D-5L, EuroQoL 5-dimension 5-level questionnaire; MCS, Mental component summary of the SF-12v2; MDD, Major Depressive Disorder; Mood DO, Mood Disorder; PCS, Physical component summary of the SF-12v2; PF, Physical functioning of the SF-36v2; Q-LES-Q-SF, Quality of Life Enjoyment and Satisfaction Questionnaire Short Form; QLS, Heinrichs-Carpenter Quality of Life Scale; SAD, Schizoaffective Disorder; SCAP-HQ, SCAP-Health Questionnaire; SCZ, Schizophrenia; SCZF, Schizophreniform; SF-12v2, Short-Form 12-Item Health Survey; SF-36v2, 36-Item Short-Form Health Survey; VAS, visual analog scale; WHOQOL-BREF, World Health Organization Quality of Life Brief.

*Indicates significance. Treatment availability was coded based on study data collection year. Studies conducted before 2013 were categorized as treatment unavailable (coded as 0), while studies conducted in or after 2013 were categorized as treatment available (coded as 1). Funding was coded based on reported funding sources: studies funded by a pharmaceutical company were categorized as pharmaceutical company funded (coded as 1), whereas studies funded by independent research foundations, national grants, or other sources were categorized as otherwise funded (coded as 0).

aIndicates enrollment values.

Forest

Plot of the Effect Sizes for the Impact of Tardive Dyskinesia on Quality of Life

Note. Shows 95% CI of Cohen’s d for each study. QLS, Heinrichs-Carpenter Quality of Life Scale; Q-LES-Q-SF, Quality of Life Enjoyment and Satisfaction Questionnaire Short Form; SCAP-HQ, SCAP-Health Questionnaire; SF-12v2, Short-Form 12-Item Health Survey; SF-36v2, 36-Item Short-Form Health Survey; VAS, visual analog scale.

Discussion

The current meta­analysis provides evidence of a small, significant, negative effect of TD on quality of life for patients with SMI. This is particularly important because individuals with SMI already face a diminished quality of life due to their underlying psychological condition (Berghöfer et al., 2020; Saavedra et al., 2023; Shumye et al., 2021). The results of this meta­analysis are consistent with previous research that indicates that the side effects produced by antipsychotics often worsen the patients’ overall quality of life (Caroff, 2019; McEvoy et al., 2019; Owens, 2018; Patterson ­ Loomba et al., 2019). This finding underscores the burden that TD places on patients, exacerbating the challenges associated with their primary mental health conditions.

We also assessed the moderating effects of treatment availability and funding source on the relationship between TD and quality of life, as we theorized that symptom management and publication bias could alter the magnitude of the analyzed effect. Interestingly, we did not find a significant moderating effect of treatment availability on the relationship between TD and quality of life for patients with SMI. This is in contradiction to the previous literature that reports on the efficacy of deutetrabenazine and valbenazine on reducing the abnormal movements associated with TD, with both drugs demonstrating favorable safety profiles (Correll et al., 2024; Golsorkhi et al., 2024; Sajatovic et al., 2022). However, most of these studies focus solely on reducing involuntary movements using these drugs without examining the overall impact on quality of life (Patel et al., 2019; Touma & Scarff, 2018). In a recent review by Golsorkhi and colleagues, deutetrabenazine had no statistically significant effect on quality of life and there are no reports on the effect of valbenazine on quality of life (2024).

Several factors may underlie this disconnect between decreased abnormal movements and improvements in quality of life. Deutetrabenazine and valbenazine, while effective, have been found to have a greater impact on those with moderate to severe TD (Fernandez et al., 2017; Touma & Scarff, 2018). Due to this differential impact of the treatment, it may be that the decreased symptoms of TD are not enough to significantly impact patients’ quality of life, as the mere presence of TD may be enough to decrease patients’ quality of life.

It is also important to recognize the limitations within the AIMS measure, the standard practice of assessing the presence and severity of TD (Stacy et al., 2019; Vaidyanathan & Jaiswal, 2020). The AIMS measure does not capture social and functional deficits associated with TD, meaning a reduction in abnormal movements may not strongly correlate with improved quality of life

(Stacy et al., 2019). Previous literature suggests a disconnect between AIMS scores and quality of life measures, as well as perceived changes reported by clinicians and patients (Fernandez et al., 2019; Stacy et al., 2019; Touma & Scarff, 2018). Despite improvements in AIMS scores, treatments may not significantly address the social and functional aspects of TD, which may have a greater impact on quality of life.

Alternatively, the literature on this topic is relatively new. Although the drugs show promise in improving quality of life and producing positive clinician and patient­reported outcomes, the differences in scores have not been statistically significant (Fernandez et al., 2017; Golsorkhi et al., 2024). So while we anticipated seeing significant moderation by treatment availability, the lack of statistical significance in these moderation tests may stem from a statistically underpowered sample. Hempel et al. (2013) have found that in instances of high heterogeneity (as was the case in our study), 200 or more effects would need to be included to reach 80% power in meta­analytic moderator studies. Not only is this unfeasible given the data available, it is also rare that other meta­analyses reach this level of power (Hempel et al., 2013). These findings underscore the need for further investigation into how these treatments influence patients’ overall well­being, as even systematic review of existing literature cannot provide a complete picture of this relationship.

We also examined funding source as a potential moderator of the relationship between TD and quality of life in patients with SMI, as funding sources have been shown to impact the outcomes of clinical studies (Davidson, 1986; Lexchin et al., 2003; Riaz et al., 2015). Studies that reported funding from a pharmaceutical company were categorized as ‘pharmaceutical company funded’ and those funded by a source other than a pharmaceutical company were coded as ‘otherwise

funded.’ However, when we ran the analyses, we did not find a significant moderating effect of funding source on the relationship between TD and patient quality of life. This may act as a promising indicator of unbiased literature on this subject, although the limited number of studies included in this meta ­ analysis could also explain the lack of statistical significance. The present analysis only assessed three studies that were funded by pharmaceutical companies and three funded otherwise, and with such a small sample size, the statistical power of our moderation analysis was low.

Due to the small sample included in this metaanalysis, future research should consider re­examining the effect of source funding and treatment availability on the impact of TD on quality of life when more data becomes available. Given the previously established effect of funding sources on the outcomes of clinical studies, it is likely that there is a moderating effect of which our small sample did not allow proper investigation (Davidson, 1986; Lexchin et al., 2003; Riaz et al., 2015). Similarly, it stands to reason that treatment availability could have an effect as well, but a large sample size would be required to determine this more concretely.

Furthermore, previous literature suggests that the effect of TD intensifies with age, due to the unique physical vulnerability of older populations to the physical consequences of TD, as well as a heightened experience of the social and emotional aspects of TD, such as the feelings of isolation and depression (Citrome et al., 2021). We were unable to examine age, along with other factors of interest, such as diagnosis, medication type, and comorbid substance use, as moderators in the present study due to our small sample size, but with greater availability of data, future research should explore these avenues as potential moderating factors.

Additionally, the asymmetry of our funnel plot indicated the potential for publication bias. While this may be due to the high heterogeneity of quality­of­life measures, previous research suggests a high risk of publication bias within clinical research (Davidson, 1986; Lexchin et al., 2003; Riaz et al., 2015). Therefore, we recommend that future systematic reviews further investigate the potential publication bias within these studies.

The present study had several limitations. Two large studies, the Schizophrenia Outpatient Health Outcomes study (Novick et al., 2010) and the Clinical Antipsychotic Trials of Intervention Effectiveness study (Swartz et al., 2008), were not included due to difficulties reaching authors and accessing the data. Similarly, we were unable to reach or receive responses from the authors to access the necessary data from three smaller studies (Gaebel et al., 2007; Lee et al., 2015; Silva de Lima et al., 2005). The exclusion of these studies resulted in a

smaller sample size and considerably reduced the power of our moderator tests. As a result, our ability to detect significant moderator effects and our observed effect size may have been affected, and future reviews on the subject should attempt to include these data which were inaccessible to us.

Additionally, studies were only included if they used AIMS to make group classifications of TD. Several studies reported AIMS scores and quality­of­life measures, but did not use AIMS to diagnose TD. Further review of available studies should request participant­level data in order to do retroactive analysis on the impact of TD on quality of life in order to make a more accurate estimate of the true population effect.

Finally, the high heterogeneity found in this study presents challenges with the generalization and applicability of this data. Likely, this high heterogeneity, especially within studies, is a result of differences in the scales used in addition to the small sample size. Quality of life, while being largely important to overall life satisfaction, can be difficult to operationalize and measure. Although the included scales are validated and widely accepted as means of measuring quality of life, they are validated internally, not necessarily in comparison to one another. Additionally, some of the included scales focused only on specific aspects of quality of life, such as social quality of life or health­related quality of life. This difference in measurements may explain much of the within­study heterogeneity. Future systematic review and analysis should consider subgrouping quality of life measures as sufficient data become available, as to more accurately define the impact of TD on specific well­being domains.

Conclusion

The present study indicates that TD significantly reduces quality of life for those with SMI, despite treatments becoming available in recent years. The available treatment methods have been found to significantly reduce the abnormal movements associated with TD; however, previous literature suggests that their impact on quality of life has been insignificant, which aligns with the moderator analysis results from the present metaanalysis (Fernandez et al., 2017; Golsorkhi et al., 2024). Although there are various reasons that may explain the disconnect, the underlying issues associated with TD remain. TD not only reduces patient quality of life, but also decreases compliance with treatment plans (Caroff, 2019; Jain et al., 2023; Owens, 2018; Patterson­Loomba et al., 2019). It is too early to make sweeping clinical recommendations based on our results. Nonetheless, providers and clinicians should ensure that their patients are educated on the potential for TD and the resulting

The Impact of Tardive Dyskinesia | Loring, Lumbard, Drwecki, and Star

reductions in quality of life so they can be informed and active participants in their care. Additional research is needed to determine how to mitigate the effect of TD on quality of life so that practitioners prioritize the patient’s quality of life as the focal point of care and improve outcomes for individuals with SMI.

References

References marked with an asterisk indicate studies included in the meta-analysis. American Psychological Association. (n.d.). APA Dictionary of Psychology. Dictionary. apa.org. Retrieved August 7, 2025, from https://dictionary.apa.org/quality-of-life

*Ascher-Svanum, H., Zhu, B., Faries, D., Peng, X., Kinon, B. J., & Tohen, M. (2008). Tardive dyskinesia and the 3-year course of schizophrenia: Results from a large, prospective, naturalistic study. The Journal of Clinical Psychiatry, 69(10), 1580–1588. https://doi.org/10.4088/jcp.v69n1008

Berghöfer, A., Martin, L., Hense, S., Weinmann, S., & Roll, S. (2020). Quality of life in patients with severe mental illness: A cross-sectional survey in an integrated outpatient health care model. Quality of Life Research, 29(8), 2073–2087. https://doi.org/10.1007/s11136-020-02470-0

*Browne, S., Roe, M., Lane, A., Gervin, M., Morris, M., Kinsella, A., Larkin, C., & Callaghan, E. O. (1996). Quality of life in schizophrenia: Relationship to sociodemographic factors, symptomatology and tardive dyskinesia. Acta Psychiatrica Scandinavica, 94(2), 118–124. https://doi.org/10.1111/j.1600-0447.1996.tb09835.x

Caroff, S. N. (2019). Overcoming barriers to effective management of tardive dyskinesia. Neuropsychiatric Disease and Treatment, 15, 785–794. https://doi.org/10.2147/NDT.S196541

Caroff, S. N., Yeomans, K., Lenderking, W. R., Cutler, A. J., Tanner, C. M., Shalhoub, H., Pagé, V., Chen, J., Franey, E., & Yonan, C. (2020). RE-KINECT: A prospective study of the presence and healthcare burden of tardive dyskinesia in clinical practice settings. Journal of Clinical Psychopharmacology, 40(3), 259–268. https://doi.org/10.1097/JCP.0000000000001201

Carroll, B., & Irwin, D. E. (2019). Health care resource utilization and costs for patients with tardive dyskinesia. Journal of Managed Care & Specialty Pharmacy, 25(7), 810–816. https://doi.org/10.18553/jmcp.2019.25.7.810

Citrome, L., Isaacson, S. H., Larson, D., & Kremens, D. (2021). Tardive dyskinesia in older persons taking antipsychotics. Neuropsychiatric Disease and Treatment, 17, 3127–3134. https://doi.org/10.2147/NDT.S328301

Cloud, L. J., Zutshi, D., & Factor, S. A. (2014). Tardive dyskinesia: Therapeutic options for an increasingly common disorder. Neurotherapeutics, 11(1), 166–176. https://doi.org/10.1007/s13311-013-0222-5

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed). L. Erlbaum Associates.

Cornett, E. M., Novitch, M., Kaye, A. D., Kata, V., & Kaye, A. M. (2017). Medicationinduced tardive dyskinesia: A review and update. Ochsner Journal, 17(2), 162–174. https://pubmed.ncbi.nlm.nih.gov/28638290/

Correll, C. U., Citrome, L., Singer, C., Lindenmayer, J.-P., Zinger, C., Liang, G., Dunayevich, E., & Marder, S. R. (2024). Sustained treatment response and global improvements with long-term valbenazine in patients with tardive dyskinesia. Journal of Clinical Psychopharmacology, 44(4), 353–361. https://doi-org/10.1097/JCP.0000000000001860

Davidson, R. A. (1986). Source of funding and outcome of clinical trials. Journal of General Internal Medicine, 1(3), 155–158. https://doi.org/10.1007/BF02602327

D’Souza, R. S., Aslam, S. P., & Hooten, W. M. (2023). Extrapyramidal symptoms (EPS). Nih.gov; StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK534115/

Faurbye, A., Rasch, P. J., Petersen, P. B., Brandborg, G., & Pakkenberg, H. (1964). Neurological symptoms in pharmacotherapy of psychoses. Acta Psychiatrica Scandinavica, 40(1), 10–27. https://doi.org/10.1111/j.1600-0447.1964.tb05731.x Feng, Y.-S., Kohlmann, T., Janssen, M. F., & Buchholz, I. (2021). Psychometric properties of the EQ-5D-5L: Systematic review of the literature. Quality of Life Research, 30(3), 647–673. https://doi-org/10.1007/s11136-020-02688-y Fernandez, H. H., Factor, S. A., Hauser, R. A., Jimenez-Shahed, J., Ondo, W. G., Jarskog, L. F., Meltzer, H. Y., Woods, S. W., Bega, D., LeDoux, M. S., Shprecher, D. R., Davis, C., Davis, M. D., Stamler, D., & Anderson, K. E. (2017). Randomized controlled trial of deutetrabenazine for tardive dyskinesia: The ARM-TD study. Neurology, 88(21), 2003–2010. https://doi.org/10.1212/WNL.0000000000003960

Gaebel, W., Riesbeck, M., Wolwer, W., Klimke, A., Eickhoff, M., Von Wilmsdorff, M., Jockers-Scherübl, M.C., Kühn, K.U., Lemke, M., Bechdolf, A., Bender, S., Degner, D., Schlösser, R., Schmidt, L.G., Schmitt, A., Jäger, M., Buchkremer, G., Falkai, P., Klingberg, S., … Moller, H. J. (2007). Maintenance treatment with risperidone or low-dose haloperidol in first-episode schizophrenia: 1-year results of a randomized controlled trial within the German Research Network on Schizophrenia. Journal of Clinical Psychiatry, 68(11), 1763–1774. https://doi.org/10.4088/jcp.v68n1116

Galletly, C., & Rigby, A. (2013). An overview of cognitive remediation therapy for people with severe mental illness. ISRN Rehabilitation, 1–6. https://doi.org/10.1155/2013/984932

Golsorkhi, M., Koch, J., Pedouim, F., Frei, K., Bondariyan, N., & Dashtipour, K. (2024). Comparative analysis of deutetrabenazine and valbenazine as VMAT2 inhibitors for tardive dyskinesia: A systematic review. Tremor and Other Hyperkinetic Movements, 14(1), 13. https://doi.org/10.5334/tohm.842

Gossman, W., Sangani, A., & Saadabadi, A. (2019, July 11). Neuroleptic medications Nih.gov; StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK459150/

Heinrichs, D. W., Hanlon, T. E., & Carpenter, W. T. (1984). The Quality of Life Scale: An instrument for rating the schizophrenic deficit syndrome. Schizophrenia Bulletin, 10(3), 388–398. https://doi.org/10.1093/schbul/10.3.388

Hempel, S., Miles, J. N., Booth, M. J., Wang, Z., Morton, S. C., & Shekelle, P. G. (2013). Risk of bias: A simulation study of power to detect study-level moderator effects in meta-analysis. Systematic Reviews, 2, 107. https://doi.org/10.1186/2046-4053-2-107

Jain, R., Ayyagari, R., Goldschmidt, D., Zhou, M., Finkbeiner, S., & Leo, S. (2023). Impact of tardive dyskinesia on physical, psychological, social, and professional domains of patient lives: A survey of patients in the United States. The Journal of Clinical Psychiatry, 84(3). https://doi-org/10.4088/JCP.22m14694

Jeste, D. V., & Wyatt, R. J. (1982). Understanding and treating tardive dyskinesia Guilford Publication.

Kaur, N., Kumar, P., Jamwal, S., Deshmukh, R., & Gauttam, V. (2016). Tetrabenazine: Spotlight on drug review. Annals of Neurosciences, 23(3), 176–185. https://doi.org/10.1159/000449184

Lee, J. Y., Kim, S. W., Lee, Y. H., Kang, H. J., Kim, S. Y., Bae, K. Y., Kim, J. M., Shin, I. S., & Yoon, J. S. (2015). Factors associated with self-rated sexual function in Korean patients with schizophrenia receiving risperidone monotherapy. Human Psychopharmacology, 30(6), 416–424. https://doi.org/10.1002/hup.2489

Lehman, A. F., Fischer, E. P., Postrado, L., Delahanty, J., Johnstone, B. M., Russo, P. A., & Crown, W. H. (2003). The Schizophrenia Care and Assessment Program Health Questionnaire (SCAP-HQ): An instrument to assess outcomes of schizophrenia care. Schizophrenia Bulletin, 29(2), 247–256. https://doi.org/10.1093/oxfordjournals.schbul.a007001

Lexchin, J., Bero, L. A., Djulbegovic, B., & Clark, O. (2003). Pharmaceutical industry sponsorship and research outcome and quality: Systematic review. BMJ, 326(7400), 1167–1170. https://doi.org/10.1136/bmj.326.7400.1167

McAnally, K., & Hagger, M. S. (2023). Health literacy, social cognition constructs, and health behaviors and outcomes: A meta-analysis. Health Psychology, 42(4), 213–234. https://doi.org/10.1037/hea0001266

*McEvoy, J., Gandhi, S. K., Rizio, A. A., Maher, S., Kosinski, M., Bjorner, J. B., & Carroll, B. (2019). Effect of tardive dyskinesia on quality of life in patients with bipolar disorder, major depressive disorder, and schizophrenia. Quality of Life Research, 28(12), 3303–3312. https://doi-org/10.1007/s11136-019-02269-10

National Institute of Mental Health. (2024, September). Mental illness. National Institute of Mental Health. https://www.nimh.nih.gov/health/statistics/mental-illness

Novick, D., Haro, J. M., Bertsch, J., & Haddad, P. M. (2010). Incidence of extrapyramidal symptoms and tardive dyskinesia in schizophrenia: Thirty-six-month results from the European schizophrenia outpatient health outcomes study. Journal of Clinical Psychopharmacology, 30(5), 531–540. https://doi-org/10.1097/JCP.0b013e3181f14098

Owens, D. C. (2018). Tardive dyskinesia update: The syndrome. BJPsych Advances, 25(1), 57–69. https://doi.org/10.1192/bja.2018.45

Page, M. J., Moher, D., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … McKenzie, J. E. (2021). PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ, n160. https://doi.org/10.1136/bmj.n160

Patel, R. S., Mansuri, Z., Motiwala, F., Saeed, H., Jannareddy, N., Patel, H., & Zafar, M. K. (2019). A systematic review on treatment of tardive dyskinesia

The

with valbenazine and deutetrabenazine. Therapeutic Advances in Psychopharmacology, 9 https://doi.org/10.1177/2045125319847882

Patterson-Lomba, O., Ayyagari, R., & Carroll, B. (2019). Risk assessment and prediction of TD incidence in psychiatric patients taking concomitant antipsychotics: A retrospective data analysis. BMC Neurology, 19(1), 1–10. https://doi-org/10.1186/s12883-019-1385-4

Perlis, R. H., Perlis, C. S., Wu, Y., Hwang, C., Joseph, M., & Nierenberg, A. A. (2005). Industry sponsorship and financial conflict of interest in the reporting of clinical trials in psychiatry. The American Journal of Psychiatry, 162(10), 1957–1960. https://doi.org/10.1176/appi.ajp.162.10.1957

Rands, A. M., Benson, D., Rapacz, A. L., & Lloyd, T. (2024). Evidence-based group therapy for severe mental illness (SMI): Treatment for schizophrenia and borderline personality disorder. Group Dynamics: Theory, Research, and Practice, 28(3), 178–198. https://dx.doi.org/10.1037/gdn0000220

*Rekhi, G., Tay, J., & Lee, J. (2022). Impact of drug-induced Parkinsonism and tardive dyskinesia on health-related quality of life in schizophrenia. Journal of Psychopharmacology, 36(2), 183–190. https://doi.org/10.1177/02698811211055812

Riaz, H., Raza, S., Khan, M. S., Riaz, I. B., & Krasuski, R. A. (2015). Impact of funding source on clinical trial results including cardiovascular outcome trials. The American Journal of Cardiology, 116(12), 1944–1947. https://doi.org/10.1016/j.amjcard.2015.09.034

Ricciardi, L., Pringsheim, T., Barnes, T. R. E., Martino, D., Gardner, D., Remington, G., Addington, D., Morgante, F., Poole, N., Carson, A., & Edwards, M. (2019). Treatment recommendations for tardive dyskinesia. Canadian Journal of Psychiatry, 64(6), 388–399. https://doi.org/10.1177/0706743719828968

Saavedra, J., Brzeska, J., Matías-García, J. A., & Arias-Sánchez, S. (2023). Quality of life and psychiatric distress in people with serious mental illness, the role of personal recovery. Psychology and Psychotherapy, 96(2), 525–541. https://doi.org/10.1111/papt.12451

Sachdev P. S. (2000). The current status of tardive dyskinesia. Australian & New Zealand Journal of Psychiatry, 34(3), 355–369. https://doi.org/10.1080/j.1440-1614.2000.00737.x

Sajatovic, M., Finkbeiner, S., Wilhelm, A., Barkay, H., Chaijale, N., Gross, N., & Gordon, M. F. (2022). Long-term safety and efficacy of deutetrabenazine in younger and older patients with tardive dyskinesia. The American Journal of Geriatric Psychiatry, 30(3), 360–371. https://doi.org/10.1016/j.jagp.2021.08.003

Salyers, M. P., Bosworth, H. B., Swanson, J. W., Lamb-Pagone, J., & Osher, F. C. (2000). Reliability and validity of the SF-12 Health Survey among people with severe mental illness. Medical Care, 38(11), 1141–1150. https://doi.org/10.1097/00005650-200011000-00008

Schonecker, M. (1957). Paroxysmale Dyskinesen als Megaphenwirung [Paroxysmal dyskinesia as the effect of megaphen]. Der Nervenarzt, 28(12), 550–553. https://pubmed.ncbi.nlm.nih.gov/13517450/

Sharma, H. (2009). Treatment of tardive dyskinesia by tetrabenazine, clonazepam and vitamin E. Indian Journal of Psychiatry, 51(2), 162–163. https://doi.org/10.4103/0019-5545.49466

Shumye, S., Amare, T., Derajew, H., Endris, M., Molla, W., & Mengistu, N. (2021). Perceived quality of life and associated factors among patients with severe mental illness in Ethiopia: A cross-sectional study. BMC Psychology, 9(1), 152. https://doi.org/10.1186/s40359-021-00664-w

Silva de Lima, M., de Jesus Mari, J., Breier, A., Maria Costa, A., Pondé de Sena, E., & Hotopf, M. (2005). Quality of life in schizophrenia: A multicenter, randomized, naturalistic, controlled trial comparing olanzapine to firstgeneration antipsychotics. The Journal of Clinical Psychiatry, 66(7), 831–838. https://pubmed.ncbi.nlm.nih.gov/16013897/

Skevington, S. M., Lotfy, M., & O’Connell, K. A. (2004). The World Health Organization’s WHOQOL-BREF quality of life assessment: Psychometric properties and results of the international field trial. A report from the WHOQOL Group. Quality of Life Research, 13(2), 299–310. https://doi.org/10.1023/b:qure.0000018486.91360.00

Stacy, M., Sajatovic, M., Kane, J. M., Cutler, A. J., Liang, G. S., O’Brien, C. F., & Correll, C. U. (2019). Abnormal involuntary movement scale in tardive dyskinesia: Minimal

clinically important difference. Movement Disorders, 34(8), 1203–1209. https://doi.org/10.1002/mds.27769

Sterne, J. A. C., & Harbord, R. M. (2004). Funnel plots in meta-analysis. The Stata Journal, 4(2), 127–141. https://doi.org/10.1177/1536867X0400400204

Stevanovic, D. (2011). Quality of Life Enjoyment and Satisfaction Questionnaire–short form for quality of life assessments in clinical practice: A psychometric study. Journal of Psychiatric and Mental Health Nursing, 18(8), 744–750. https://doi.org/10.1111/j.1365-2850.2011.01735.x

Sullivan, G. M., & Feinn, R. (2012). Using effect size––or why the p value is not enough. Journal of Graduate Medical Education, 4(3), 279–282. https://doi.org/10.4300/JGME-D-12-00156.1

Swartz, M. S., Stroup, T. S., McEvoy, J. P., Davis, S. M., Rosenheck, R. A., Keefe, R. S., Hsiao, J. K., & Lieberman, J. A. (2008). What CATIE found: Results from the schizophrenia trial. Psychiatric Services, 59(5), 500–506. https://doi.org/10.1176/ps.2008.59.5.500

*Tanner, C. M., Caroff, S. N., Cutler, A. J., Lenderking, W. R., Shalhoub, H., Véronique Pagé, Franey, E., Serbin, M., & Yonan, C. (2023). Impact of possible tardive dyskinesia on physical wellness and social functioning: Results from the real-world RE-KINECT study. Journal of Patient-Reported Outcomes, 7(1). https://doi.org/10.1186/s41687-023-00551-5

Tampi, R. R., Tampi, D. J., Balachandran, S., & Srinivasan, S. (2016). Antipsychotic use in dementia: A systematic review of benefits and risks from metaanalyses. Therapeutic Advances in Chronic Disease, 7(5), 229–245. https://doi.org/10.1177/2040622316658463

Touma, K. T. B., & Scarff, J. R. (2018). Valbenazine and deutetrabenazine for tardive dyskinesia. Innovations in Clinical Neuroscience, 15(5–6), 13–16. https://pubmed.ncbi.nlm.nih.gov/30013814/

*Usman, M., Saleem, F., & Dimitrios Adamis. (2023). Holistic assessment of patients with chronic mental disorders who attend a metabolic clinic in Sligo Town catchment area. Irish Journal of Medical Science, 192(5). https://doi.org/10.1007/s11845-023-03284-y

Vaidyanathan, S., & Jaiswal, S. (2020). Tardive dyskinesia: Prevention and newer management strategies. Annals of Indian Psychiatry, 4(1), 10. https://doi.org/10.4103/aip.aip_29_20

Vasan, S., & Padhy, R. K. (2023, April 24). Tardive dyskinesia. PubMed; StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK448207/ Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1–48. https://doi.org/10.18637/jss.v036.i03 Waln, O., & Jankovic, J. (2013). An update on tardive dyskinesia: From phenomenology to treatment. Tremor and Other Hyperkinetic Movements, 3. https://doi.org/10.7916/D88P5Z71

Ware, J. E., Jr, & Sherbourne, C. D. (1992). The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Medical Care, 30(6), 473–483. https://pubmed.ncbi.nlm.nih.gov/1593914/

Author Note

Madelynn Loring https://orcid.org/0009­0006­4855­0644

Trinity Lumbard https://orcid.org/0009­0009­3257­8548

Kaylyn McAnally Star https://orcid.org/0000­0001­6197­4253

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.

Madelynn Loring and Trinity Lumbard played equal lead roles in conceptualization, research design, data collection, data analysis and interpretation, and substantive original writing. Brian Drwecki played a supporting role in research design and editorial assistance. Kaylyn McAnally Star provided editorial assistance and minor original writing and played an equal role in data analysis and interpretation.

Correspondence concerning this article should be addressed to Madelynn Loring. Email: mloring@regis.edu

Exposing the Falseness of Instagram: The Impact of Instagram vs. Reality on Body Dissatisfaction and Body Appreciation

Isabella Cerase1, Mylene Feiler*2, Sean Dougherty*1, and Madeline Dougherty*1,

1 Department of Psychology, Florida State University

2 Department of Language Arts, International Studies Preparatory Academy

ABSTRACT. A recent trend on Instagram called “Instagram vs. Reality” consists of posting side­by­side photographs of a woman, one idealized and edited and the other in its original, unedited form, to demonstrate that what is shown on social media is not always reality. The present study aimed to investigate the effect of this trend on body dissatisfaction and body appreciation in adolescent girls. Participants included 156 high school girls aged 14–19. Participants’ trait body satisfaction was assessed using the Multidimensional Body­Self Relations Questionnaire, and visual analogue scales assessing body dissatisfaction and body appreciation were administered after viewing idealized Instagram images and again after viewing Instagram vs. Reality images. Viewing Instagram vs. Reality images resulted in decreased body dissatisfaction, F(1, 154) = 16.04, p < .001, ηp² = .09, and this effect remained after controlling for trait levels of body satisfaction (p = .04). Furthermore, body appreciation increased after viewing images of Instagram vs. Reality, F(1, 154) = 7.91, p = .006, ηp² = .05; however, this effect no longer remained after controlling for trait levels of body satisfaction (p = .16). These results demonstrate that viewing Instagram vs. Reality images may help adolescent girls feel better about their bodies in the short­term. More research is needed on the long­term effects of viewing Instagram vs. Reality images, and its possible utilization in eating disorder prevention programs.

Keywords: body image, body dissatisfaction, body appreciation, Instagram, social media, idealized images, adolescents

RESUMEN. Una tendencia reciente de Instagram es llamada “Instagram vs. Reality”; consiste en publicar dos fotos de la misma mujer una a cada lado, en la cual la primera foto es una imagen idealizada y editada y en la segunda foto se muestra la imagen original sin editar, demostrando que lo que se publica en las redes sociales no es siempre real. El presente estudio tiene como objetivo investigar los efectos de esta tendencia en la apreciación e insatisfacción de la imagen corporal en mujeres adolescentes. Las participantes son 156 estudiantes mujeres adolescentes de Escuela Secundaria entre 14–19 años. En el estudio el rasgo de apreciación corporal fue medido utilizando el Cuestionario Multidimensional de la Imagen Corporal (Multidimensional Body Shape Relation Questionnaire, MBSRQ; Cash, 1990), y Visual Analogue Scales que miden la insatisfacción de la imagen corporal y la apreciación corporal. Ambos constructos fueron suministrados después de que las participantes observarán primero la imagen de Instagram idealizada y sucesivamente observaran la imagen de la tendencia Instagram vs. Reality. La visualización de imágenes tipo Instagram vs. Reality produjo una disminución de la insatisfacción corporal, F(1, 154) = 16.04, p < .001, ηp² = .09, y este efecto se mantuvo después de controlar los niveles de satisfacción corporal, (p = .04).

Además, la apreciación corporal aumentó tras la visualización de imágenes de Instagram vs. Reality, F(1, 154) = 7.91, p = .006, ηp² = .05; sin embargo, este efecto dejó de mantenerse después de controlar los niveles de satisfacción corporal, (p = .16). . Los resultados demuestran que observar las imágenes de Instagram vs. Reality, puede ayudar a las jóvenes adolescentes a sentirse mejor sobre la propia imagen corporal a breve plazo. Pero más investigaciones son necesarias en el análisis de los efectos a largo plazo de la visualización de las imágenes Instagram vs. Reality; y su posible uso en programas de prevención de trastornos alimenticios.

Palabras clave: imagen corporal, insatisfacción corporal, apreciación corporal, Instagram, redes sociales, imagen idealizada, adolescentes

Social media has infiltrated much of the modern world in recent years. As social media has become more prominent, there is an increasing concern on how social media affects body image and risk for body­related issues, including risk for eating disorders. Social media platforms typically revolve around showcasing oneself to a variety of audiences ranging from close friends to complete strangers by posting pictures, text, and videos. Because users are constantly being perceived on social media, they are likely motivated to manage the way they present themselves to others, often trying to present themselves in a positive light (Hjetland et al., 2022; Zheng et al., 2020). Since popular social media platforms are mostly image ­ based and facilitate appearance­ based focus and interactions (Rodgers, 2015), presenting well to others on social media is typically manifested as conforming to unrealistic appearance and beauty standards (Vandenbosch & Eggermont, 2016). Chua and Chang (2016) reported that adolescent girls altered their appearance on social media to achieve certain standards of beauty that would match the predicted expectations of their audiences and therefore gain “likes” and followers (i.e., social approval). Indeed, research demonstrates that likes become a form of social approval, attractiveness, and value to teen users on social media (Dumas et al., 2017; Rosenthal­von der Putten et al., 2019), and people choose to present their idealized self for selfenhancement motives and avoid sharing potential flaws with others (Zheng et al., 2020). This increased preoccupation with self­presentation on social media is associated with increased mental health problems and reduced quality of life (Skogen et al., 2021).

Social media’s link to negative mental health outcomes has conflicted findings in the literature; a meta ­ analysis conducted by Ferguson et al. (2025) did not find a clinically relevant link between time

spent on social media and mental health outcomes in youth, emphasizing the importance of studying the type of content and behaviours that users are engaging in as correlating to negative outcomes. Wright et al. (2021) found that video­based and professional­based platforms were associated with greater wellness, while image­based platforms like Snapchat were associated with poorer mental health. It has been theorized that image­based communication may enable social comparison as individuals post the most positive version of themselves (Yang, 2016), and women were found to be likely to report greater exposure to image­based platforms (Wright et al., 2020). In fact, Nienstedt et al. (2023) found that women reported worse mental health than nonusers of another popular social media app, TikTok, which is both video and image based. These studies, while conflicting, emphasize the need to study adolescent girls as a vulnerable population, and image­based platforms like Instagram and TikTok for the content they may be promoting.

As social media users are encouraged to focus on how they will be perceived, especially in photo­based apps, appearance and beauty standards become increasingly important and valued. An extensive literature has examined the impact that social media has on body image through a sociocultural lens. Sociocultural theory posits that body­related messages portrayed in the media lead to greater body dissatisfaction mediated by two processes: internalization of body ideals and comparison to these ideals (Cohen et al., 2017; Fardouly et al., 2017; Mahon & Hevey, 2021; Rodgers, 2015). This theory of body image explains the literature linking social media to increased pressure on appearance, body disturbances, and disordered eating (Holland & Tiggemann, 2016; Hu, 2018; Reel et al., 2015; Saiphoo & Vahedi, 2019; Smith et al., 2013). Instagram and TikTok have consistently remained popular apps among teenagers, with six in ten using them every day (Faverio

& Sidoti, 2024). During adolescence, the increase in adipose tissue that occurs in puberty moves girls away from the sociocultural thin­ideal of beauty and attractiveness (Clay et al., 2005; Reel et al., 2015). The relation between pubertal change in girls and body dissatisfaction is mediated by how strongly cultural ideals of beauty are internalized (Hayward, 2003). Adolescence for all genders is a difficult period determined by many conflicts; according to Erikson (1968), this stage of life is characterized by the struggle of social roles and sense of self and identity. These two processes (pubertal body changes and identity vs. role confusion) may make specifically adolescent girls especially susceptible to the negative impacts of media exposure on body image and eating disorder risk. Importantly, social media provides constant exposure to beauty standards, allowing young girls to be constantly aware of the discrepancy between their body changing and how they are expected to look to be considered attractive and beautiful by others.

As adolescent girls find that they do not meet the thin­ideal but want to achieve social approval, they may be especially likely to edit or curate their content on social media to maintain an appealing self­presentation. Most social media platforms facilitate these processes, providing tools to users on the platform to edit their body shape and size in photos to hide “imperfections” and highlight only their “best” features. Although more research is needed, photo­editing has been found to be protective for certain populations like cancer patients or non­western samples (McGovern et al., 2022). However, this kind of editing can have deleterious effects on the mental health of its consumers, especially in maintaining negative body image attitudes. For instance, Wick and Keel (2020) found that editing photos was linked to an immediate decrease in weight/shape concerns but posting edited photos caused increased weight/shape concerns, overall finding a direct link between posting edited photos and eating disorder risk factors.

Furthermore, social media allows users to both create content and consume content. Kleemans et al. (2016) found that exposure to manipulated Instagram photos led to lower body dissatisfaction. Importantly, girls viewing altered Instagram photos believed they were a representative view of reality and did not notice any reshaping of the bodies (Kleemans et al., 2016).

A vicious cycle is formed: girls view idealized and unrealistic images, internalize strict societal ideals of body image, and compare themselves to these standards resulting in decreased body satisfaction. To alleviate this dissatisfaction, girls will post only idealized images of themselves to get approval and validation, which then may increase others’ body dissatisfaction and drive to edit their own photos, thus perpetuating the cycle.

In response to the heavily curated and idealized nature of social media, body positive content has started to appear on social media platforms with the intention of fighting unrealistic and unattainable appearance ideals found on social media (Cohen et al., 2020). Body positivity is rooted in the concept of positive body image. Positive body image includes multiple facets such as body appreciation, body acceptance and love, inner positivity, and adaptive appearance investment (Tylka & Wood­Barcalow, 2015). Research on positive body image shows that it is associated with greater psychological and emotional well­being (Swami et al., 2017). At the same time, body positivity has been criticized for putting a new pressure on women to love their bodies and thus making them feel worse if they do not. Moreover, although the message of the body positivity movement is to appreciate one’s body, the content still focuses on appearance, which some argue is problematic (Cohen et al., 2020). To analyze whether body positivity content promotes what it claims to promote, Cohen et al. (2019) conducted a content analysis of 640 posts made by the most popular body positivity accounts on Instagram and found that the vast majority of posts contained messages in line with at least one of the core features of positive body image. However, the posts also contained attributes which conflict with societal beauty ideals, such as the presence of cellulite, stomach rolls, and stretch marks. These findings were supported by Lazuka et al. (2020), who found that body positivity posts made by the broader Instagram community often contained contradictory messages such as weight loss content alongside body positive messages.

One popular trend within the body positivity movement is Instagram vs. Reality. Instagram vs. Reality posts aim to discourage appearance comparison and dissatisfaction by showing that the images found on Instagram are not reality. The trend consists of a sideby­side image of the same woman, one where she is in a flattering pose and lighting, and the other where she is showing the original image, in which her appearance is typically less consistent with the thin­ideal. Research has shown that exposure to Instagram vs. Reality results in decreased body dissatisfaction in women aged 18–35 (Tiggemann & Anderberg, 2019). However, little is known about the effects of exposure to Instagram vs. Reality images on adolescent girls.

The present study aimed to analyze the impact of Instagram vs. Reality on the body image of adolescent girls. As adolescent girls are especially susceptible to negative body image due to societal standards of beauty and the importance of appearance to their self­worth (Thompson et al., 1999), this study examined if exposure to Instagram vs. Reality impacted body dissatisfaction

and body appreciation, two factors relevant to body image (Andrew et al., 2015; Sztainer et al., 2006). We hypothesized that, after viewing Instagram vs. Reality images, adolescent girls would experience a decrease in body dissatisfaction and an increase in body appreciation, consistent with the findings of Tiggemann and Anderberg (2019).

Method

Participants

Participants were 156 female students recruited from a high school in South Florida whose ages ranged from 14 to 19. Of the participants, 41.03% were ages 14—15, 51.92% were ages 16–17, and 7.05% were ages 18–19, with a mean age of 15.87 (SD = 1.23). Racial/ethnic identification of participants was as follows: 81.41% Hispanic/Latino, 15.38% White, 1.28% Asian American, and 1.92% two or more races/ethnicities.

Measures

Demographic Questionnaire

Participants were administered a demographic questionnaire assessing age, race/ethnicity, grade level, if they used social media, and for how many hours they used social media platforms on average each day (2+ hours, 1–2 hours, 30–60 minutes, or less than 10 minutes).

Multidimensional Body-Self Relations Questionnaire (MBSRQ)

Participants’ trait body satisfaction was assessed using the Appearance Evaluation subscale of the Multidimensional Body­Self Relations Questionnaire (MBSRQ­AE; Cash, 2000). Participants responded to 7 items on the MBSRQ­AE using a 5­point Likert­type scale ranging from 1 (definitely disagree) to 5 (definitely agree). The Appearance Evaluation subscale assesses one’s feelings of attractiveness and satisfaction with physical appearance (Cash, 2000). The MBSRQ assesses specific and well­defined elements of body image, and its subscales demonstrated acceptable internal consistency and stability, and strong construct validity in samples of both men and women (Brown et al., 1990; Thompson & Schaefer, 2019). Additionally, it has been found to have strong psychometric properties when tested among younger age groups (Marco et al., 2017). Internal consistency reliability of the MBSRQ­AE in our sample was good (a = .89).

Visual Analogue Scales (VAS)

VAS assessing momentary body dissatisfaction (a = .93 to .94) and body appreciation (a = .63 to .65) were administered to participants after viewing idealized images on Instagram and again after viewing Instagram

vs. Reality images. Participants were instructed to mark from 0 (none) to 100 (very much) to indicate how they felt for each item. Specific items that made up the visual analogue scales were selected from previous measures on relevant constructs, such as the Body Appreciation Scale (Avalos et al., 2005) and the Body­Esteem Scale for Adolescents and Adults (Mendelson et al., 2001) or were developed based on the existing literature surrounding body satisfaction and body appreciation (e.g., Heinberg & Thompson, 1995; Tiggemann & Anderberg, 2019; Tylka & Wood­Barcalow, 2015a). The Appendix includes all VAS items included in the present study. VAS are more sensitive to changes over short amounts of time (Aitken, 1969) and have been used successfully in previous experimental studies examining momentary changes in body image (e.g., Tiggemann & Anderberg, 2019; Wick & Keel, 2020).

Stimulus Materials

Participants were exposed to 10 Instagram vs. Reality images. Images were identified by searching the hashtags #instavsreality, #instagramvsreality, and #igvsreality on Instagram and selecting the first 10 images that met inclusion criteria. Inter­rare reliability was ensured by having a second researcher review the images. Inclusion criteria required images to consist of side­by­side photos of the same woman, one where she was in a flattering pose and lighting, and the other where she shows how her body really looks, including “imperfections” such as stomach rolls or cellulite. Moreover, all images were required to display the full body of the women. Some of the photos also included text such as “Posted/Edited,” “Edited/Original,” or “Posed/Relaxed” to emphasize and give context to the side­by­side differences. We did not include the original text during the first portion of the study as the purpose was to show participants only the idealized part of the trend. Any original text present in the pictures (e.g., “Posted/Deleted,” “Edited/Original,” “Posed/Relaxed”) was kept during the second part of the study, in which participants viewed the Instagram vs. Reality images.

Procedures

All study procedures were approved by the appropriate institutional review board in both the high school and Florida State University. Participants were asked to sign a consent form, which advised of the purpose of the study and the data collection process; for participants who were minors, the assent of their legal guardians was required. Participants received community service hours as compensation for participation in the study. All research procedures were administered online. Participants who did not have Instagram were excluded

from the study. Participants were first asked to complete the MBSRQ­AE items. Rather than presenting the pictures one at the time, participants were presented with all the images for each set (10 idealized images and their respective 10 Instagram vs. Reality images), giving them the possibility to scroll through the photos the same way they would be able to on Instagram. Participants were first shown only the idealized portion of the selected Instagram vs. Reality images, excluding any text already present. Participants were then asked to express how they felt about their body image and appearance using the VAS. Participants were then shown the Instagram vs. Reality images in the same order as their corresponding idealized portion. Participants responded to the VAS a second time to detect any change in levels of body dissatisfaction and body appreciation after short­term exposure to the trend.

Results

Sample Characteristics

The majority (98%) of participants were social media users. Participants’ daily social media time was as follows: 60% spent 2+ hours on social media daily, 29% spent 1–2 hours, 6% spent 30–60 minutes, 1% spent 10–30 minutes, and 4% spent less than 10 minutes (see Table 1).

The mean trait body satisfaction score for our sample, as measured by the Appearance Evaluation subscale of the MBSRQ (Cash, 2000), was 23.08 (SD = 6.23). In our sample, MBSRQ­AE subscale scores ranged from 7 to 35, thus capturing the full range of possible scores for this measure. Trait body satisfaction was not correlated with time spent on social media in our sample (p = .10).

Effect of Experimental Manipulation

Viewing Instagram vs. Reality images resulted in decreased body dissatisfaction, F (1, 154) = 16.04, p < .001, ηp² = .09 , and this effect remained after controlling for trait levels of body satisfaction (p = .04). Furthermore, body appreciation increased after viewing images of Instagram vs. Reality, F (1, 154) = 7.91, p = .006, ηp² = .05; however, this effect no longer remained after controlling for trait levels of body satisfaction (p = .16). Table 2 and Figure 1 includes average body dissatisfaction and body appreciation scores for participants after viewing idealized Instagram images and after viewing Instagram vs. Reality images.

The main effect of body dissatisfaction was significant, F(1, 154) = 16.04, p < .001, ηp² = .09 (Cohen’s f ≈ 0.32, observed power ≈ 0.99). The interaction between body dissatisfaction and baseline trait body satisfaction was also significant, F(1, 154) = 4.37, p = .038, ηp² = .03

(Cohen’s f ≈ 0.17, observed power ≈ 0.61), indicating a small moderating effect. For body appreciation, the main effect of body appreciation was significant, F(1, 154) = 7.91, p = .006, ηp² = .05 (Cohen’s f ≈ 0.23, observed power ≈ 0.87). The interaction between body appreciation and baseline trait body satisfaction was not significant, F(1, 154) = 1.98, p = .161, ηp² = .01 (Cohen’s f ≈ 0.12, observed power ≈ 0.27), indicating that baseline trait body satisfaction did not moderate the effect.

There was no interaction between time spent on social media and body dissatisfaction or body appreciation. The correlation between baseline trait body satisfaction and time spent on social media was negative,

TABLE 1

Descriptive Statistics and Frequencies

Note. N = number of participants; SD = standard deviation. “Time on social media” was measured on a 5-point scale, higher values indicate more time spent online.

TABLE 2

Average Scores for VAS Measures of Body Dissatisfaction and Appreciation Across Timepoints

Instagram vs. Reality and Body Image

| Cerase, Feiler, S. Dougherty, and M. Dougherty

but not statistically significant (p = .102). There was also no statistically significant correlation between ethnicity trait body satisfaction levels.

A one ­ way ANOVA examined age differences based on the grade reported in the survey and trait body satisfaction. The overall effect of age was marginally significant, F(3, 152) = 2.66, p = .050, η² = .05. Post­hoc comparison revealed that seniors (17–19 years of age) reported significantly higher body satisfaction compared to freshmen (14–15 years of age; Mean difference = 3.78, SE = 1.35, p = .035). No other grade comparison was statistically significant (all ps > .05).

Discussion

The aim of the study was to analyze the impact of the Instagram vs. Reality trend on adolescent girls’ body dissatisfaction and body appreciation. Results demonstrated preliminary evidence that viewing Instagram vs. Reality images after viewing edited images resulted in decreased body dissatisfaction and increased body appreciation in our sample of adolescent girls. As Instagram vs. Reality points out the unrealistic nature of idealized images widely shared on social media, it may help counter the body dissatisfaction that often results from viewing these same idealized images (Mahon & Hevey, 2021; Rodgers, 2015). These results were similar to Tiggemann and Anderberg (2019), who also found that exposure to Instagram vs. Reality images resulted in decreased body dissatisfaction in women aged 18–35. Moreover, the effect of viewing Instagram vs. Reality images on body dissatisfaction remained when controlling for trait levels of body satisfaction suggesting that, regardless of one’s general satisfaction with their body, they could still be positively impacted by exposure to Instagram vs. Reality on social media.

Although viewing Instagram vs. Reality images after viewing idealized images on social media resulted in increased body appreciation, that effect no longer remained after controlling for trait body satisfaction. It is interesting that trait body satisfaction accounted for changes in state body appreciation but not state body dissatisfaction in our sample. It is possible that the items included in the Appearance Evaluation subscale of the MBSRQ (Cash, 2000) tapped into body appreciation more so than body satisfaction, or maybe state and trait body satisfaction are distinct constructs. Although the trend decreased the levels of body dissatisfaction regardless of trait body satisfaction, those with higher trait body satisfaction most likely already appreciated their bodies and might have been less susceptible to the possible effects of the trend in increasing body appreciation. Another result found was that older adolescent girls had a marginally significant higher trait body

satisfaction compared to younger adolescent girls in our sample. This suggests that, as adolescent girls grow, they felt more positively about their bodies as well. These results are consistent with other longitudinal studies on adolescents and body satisfaction that showed that body image worsened between ages 10–16, and improved between 16–24 (Lacroix et al., 2023). Pubertal changes, social pressures, and social media comparisons may be especially salient in impacting body image during early to middle adolescence.

No relationship was found between time spent on social media and trait body satisfaction. This was consistent with other studies that found that rather than time spent on social media, the type of content consumed is more strongly associated with body and eating disturbances (Sanzari et al., 2023; Sun et al., 2025). Future research should analyze the impact of Instagram vs. Reality alongside the type of social media usage (e.g., passive or active) and content consumed.

To our knowledge, our study was the first to examine the exposure to Instagram vs. Reality images, a popular body positivity trend, in a sample of adolescent girls. As social media becomes more prevalent, it is essential to understand how it impacts users, especially younger generations of users who may be more susceptible to its deleterious consequences. For this reason, our study provides a novel understanding of body positivity’s impact on adolescents. Users are no longer just passively being exposed to body image ideals and beauty standards; instead, they now take part in enacting these ideas and reinforcing them through curating how they look on social media (Chua & Chang, 2016; Yau & Reich, 2018). However, even when users post highly curated and idealized content on their on social media, they are not always

aware that they are also viewing idealized portrayals of others’ images (Kleemans et al., 2016). The Instagram vs. Reality trend utilizes this mechanism in its favour, and our study provides initial evidence that it may help adolescent girls feel better about their bodies in the short term. Our findings align with those of the broad body positivity literature (e.g., Cohen et al., 2019b; Tiggemann & Anderberg, 2019; Tylka & Wood­Barcalow, 2015a) and research that suggest that exposure to body diversity may help reduce body dissatisfaction and lessen the impact of exposure to the thin­ideal (Bould et al., 2018; Ogden et al., 2020). Overall, our findings suggest that an important protective factor for body image concerns could be exposing individuals to the idealized nature of self­presentation on social media. This not only adds to the extensive literature on body image and social media, but also gives insight to possible future research on social media and adolescence in relation to appearance and self­presentation. Nevertheless, this study demonstrates that divulging some of these Instagram backstage processes—such as crafted and unreal appearances—can result in improved body appreciation and satisfaction in the short term. These results should be considered for prevention of body image issues among adolescent girls. This study is not without its own limitations. First, our sample only included adolescent female participants. This decision was made because of the increased risk for teenage girls to be impacted by social media in relation to body image and eating disorder risk factors (Clay et al., 2005; Reel et al., 2015; Thompson & Heinberg, 1999). However, because we limited our sample to participants who were female high school students, our findings cannot be generalized to other genders or age groups. According to Faverio and Sidoti (2024), Instagram use presents no gender difference among its adolescent users. Future research should try to analyze the possible benefits of body positivity and body positivity trends on a male sample. Second, our study design allowed us to determine the immediate effects of exposure to Instagram vs. Reality but not any long-term effects to consistent exposure. More longitudinal research should be conducted to analyze the long-term impact of the trend. In addition, it is important to consider that the amount of time participants spent viewing each image or the attention dedicated to the images was not measured and might have impacted our findings. Furthermore, all participants in our study underwent the same study procedure and thus we did not utilize random assignment or an experimental design with a control group. For this reason, it is possible that the decrease in body dissatisfaction and increase in body appreciation was due to habituation to the images rather than the effect of Instagram vs. Reality images.

We also only focused on Instagram as that is where the trend started and was found; better understanding of this trend across other photo-based apps would be beneficial in understanding its impact. More research is needed to determine whether exposure to Instagram vs. Reality images truly contribute to decrease in body dissatisfaction and increase body appreciation. For instance, experiments that compare the effect of exposure of Instagram vs. Reality images to other body positive trends or other nonvisual methods (such as reading about idealized images on social media) would provide more information about how helpful viewing Instagram vs. Reality images is in protecting users from body image concerns.

Another limitation is related to the body positivity movement in general, which Instagram vs. Reality images are a part of. Although body positive trends typically depict a broad range of body sizes and appearances (Cohen et al., 2019), these trends have also been criticized for including primarily thin, attractive women in the images and being anti­fat, as the women depicted still fit societal standards of beauty (Lazuka et al., 2020; Weiss, 2018). In our study, the images included were all of White and thin women, which were the most popular images found under the Instagram vs. Reality hashtag when we were selecting our stimulus materials. This discrepancy highlights the importance of prioritizing the inclusion of diverse body sizes and a range of races and ethnicities in future similar studies to better generalize these findings to individuals who are not White or thin.

Despite these limitations, our study provides preliminary evidence for the positive impact of exposure to Instagram vs. Reality images on adolescent girls. The findings on the present study have clinical implications for body image interventions among adolescent girls; this trend may serve as an effective tool to reduce thin internalization by providing a visual demonstration that many images online are edited, curated, and unrealistic, potentially reducing appearance comparison and promoting body positivity. We encourage future research to explore how different social media mechanisms relate to body image and appearance to test ways to inoculate from the negative impact of social media use. A growing literature has explored the use of media literacy programs to decrease social comparisons on social media and thus improve body satisfaction and body image (Bell et al., 2021; Burnette et al., 2017; McLean et al., 2017; Paxton et al., 2022). This study adds to the literature by providing preliminary evidence that exposing adolescent girls to Instagram vs. Reality images may be helpful for increasing and maintaining a positive body image.

References

Aitken, R. C. (1969). Measurement of feelings using visual analogue scales. Proceedings of the Royal Society of Medicine, 62, 989–993.  https://doi.org/10.1177/003591576906201005

Andrew, R., Tiggemann, M., & Clark, L. (2015). The protective role of body appreciation against media-induced body dissatisfaction. Body Image, 15, 98–104. https://doi.org/10.1016/j.bodyim.2015.07.005

Bell, B. T., Taylor, C., Paddock, D., & Bates, A. (2021). Digital bodies: A controlled evaluation of a brief classroom‐based Intervention for reducing negative body image among adolescents in the digital age. British Journal of Educational Psychology, 92(1), 280–298. https://doi.org/10.1111/bjep.12449

Bould, H., Carnegie, R., Allward, H., Bacon, E., Lambe, E., Sapseid, M., Button, K. S., Lewis, G., Skinner, A., Broome, M. R., Park, R., Harmer, C. J., Penton-Voak, I. S., & Munafò, M. R. (2018). Effects of exposure to bodies of different sizes on perception of and satisfaction with own body size: Two randomized studies. Royal Society Open Science, 5(5). https://doi.org/10.1098/rsos.171387

Brown, T., Cash, T., & Mikulka, P. (1990). Attitudinal body-image assessment: Factor analysis of the Body-Self Relations Questionnaire. Journal of Personality Assessment, 55(1), 135–144. https://doi.org/10.1207/s15327752jpa5501&2_13

Burnette, C. B., Kwitowski, M. A., & Mazzeo, S. E. (2017). “I don’t need people to tell me I’m pretty on social media:” A qualitative study of social media and body image in early adolescent girls. Body Image, 23(23), 114–125. https://doi.org/10.1016/j.bodyim.2017.09.001

Cash, T. F. (2000). Multidimensional Body-Self Relations Questionnaire: MBSRQ user’s manual. Old Dominion University.

Chua, T. H. H., & Chang, L. (2016). Follow me and like my beautiful selfies: Singapore teenage girls’ engagement in self-presentation and peer comparison on social media. Computers in Human Behavior, 55(A), 190–197. https://doi.org/10.1016/j.chb.2015.09.011

Clay, D., Vignoles, V. L., & Dittmar, H. (2005). Body image and self-esteem among adolescent girls: Testing the influence of sociocultural factors. Journal of Research on Adolescence, 15(4), 451–477. https://doi.org/10.1111/j.1532-7795.2005.00107.x

Cohen, R., Irwin, L., Newton-John, T., & Slater, A. (2019). #bodypositivity: A content analysis of body positive accounts on Instagram. Body Image, 29(29), 47–57. https://doi.org/10.1016/j.bodyim.2019.02.007

Cohen, R., Newton-John, T., & Slater, A. (2017). The relationship between Facebook and Instagram appearance-focused activities and body image concerns in young women. Body Image, 23(1), 183–187. https://doi.org/10.1016/j.bodyim.2017.10.002

Cohen, R., Newton-John, T., & Slater, A. (2020). The case for body positivity on social media: Perspectives on current advances and future directions. Journal of Health Psychology, 26(13), 135910532091245 https://doi.org/10.1177/1359105320912450

Dumas, T. M., Maxwell-Smith, M., Davis, J. P., & Giulietti, P. A. (2017). Lying or longing for likes? Narcissism, peer belonging, loneliness and normative versus deceptive like-seeking on Instagram in emerging adulthood. Computers in Human Behavior, 71, 1–10. https://doi.org/10.1016/j.chb.2017.01.037

Fardouly, J., Pinkus, R. T., & Vartanian, L. R. (2017). The impact of appearance comparisons made through social media, traditional media, and in person in women’s everyday lives. Body Image, 20(1), 31–39. https://doi.org/10.1016/j.bodyim.2016.11.002

Faverio, M., & Sidoti, O. (2024, December 12). Teens, social media, and technology 2024. Pew Research Center. https://www.pewresearch.org/ internet/2024/12/12/teens-social-media-and-technology-2024/

Ferguson, C. J., Kaye, L. K., Branley-Bell, D., & Markey, P. (2025). There is no evidence that time spent on social media is correlated with adolescent mental health problems: Findings from a meta-analysis. Professional Psychology: Research and Practice, 56(1), 73–83. https://doi.org/10.1037/pro0000589

Hayward, C. (Ed.). (2003). Gender differences at puberty. Cambridge University Press.

Hjetland, G. J., Finserås, T. R., Sivertsen, B., Colman, I., Hella, R. T., & Skogen, J. C. (2022). Focus on self-presentation on social media across sociodemographic variables, lifestyles, and personalities: A crosssectional study. International Journal of Environmental Research and Public Health, 19(17), 11133. https://doi.org/10.3390/ijerph191711133

Heinberg, L. H., & Thompson, J. K. (1995). Body image and televised images of thinness and attractiveness: A controlled laboratory investigation. Journal of Social and Clinical Psychology, 14(4), 325–338. https://doi.org/10.1521/jscp.1995.14.4.325

Holland, G., & Tiggeman, M. (2016). A systematic review of the impact of the use of social networking sites on body image and disordered eating outcomes. Body Image, 17(17), 100–110. https://doi.org/10.1016/j.bodyim.2016.02.008

Hu, Y. (2018). Exploration of how female body image is presented and interpreted on Instagram. Advances in Journalism and Communication, 06(04), 95–120. https://doi.org/10.4236/ajc.2018.64009

Kleemans, M., Daalmans, S., Carbaat, I., & Anschütz, D. (2016). Picture perfect: The direct effect of manipulated Instagram photos on body image in adolescent girls. Media Psychology, 21(1), 93–110. https://doi.org/10.1080/15213269.2016.1257392

Lacroix, E., Smith, A. J., Husain, I. A., Orth, U., & von Ranson, K. M. (2023). Normative body image development: A longitudinal meta-analysis of mean-level change. Body Image, 45, 238–264. https://doi.org/10.1016/j.bodyim.2023.03.003

Lazuka, R. F., Wick, M. R., Keel, P. K., & Harriger, J. A. (2020). Are we there yet? Progress in depicting diverse images of beauty in Instagram’s body positivity movement. Body Image, 34(34), 85–93. https://doi.org/10.1016/j.bodyim.2020.05.001

Mahon, C., & Hevey, D. (2021). Processing body image on social media: Gender differences in adolescent boys’ and girls’ agency and active coping. Frontiers in Psychology, 12. https://doi.org/10.3389/fpsyg.2021.626763

McLean, S. A., Wertheim, E. H., Masters, J., & Paxton, S. J. (2017). A pilot evaluation of a social media literacy intervention to reduce risk factors for eating disorders. International Journal of Eating Disorders, 50(7), 847–851. https://doi.org/10.1002/eat.22708

Nienstedt, C., Smith, N., Braithwaite, H., Gilbert, B., Wright, R. W. (2023). Swiping away your well-being? Examining well-being indicators among TikTok account holders. Psi Chi Journal of Psychological Research, 28(2), 96–106.  https://doi.org/10.24839/2325-7342.JN28.2.96

Ogden, J., Gosling, C., Hazelwood, M., & Atkins, E. (2020). Exposure to body diversity images as a buffer against the thin-ideal: An experimental study. Psychology, Health & Medicine, 25(10), 1–14. https://doi.org/10.1080/13548506.2020.1734219

Paxton, S. J., McLean, S. A., & Rodgers, R. F. (2022). “My critical filter buffers your app filter”: Social media literacy as a protective factor for body image. Body Image, 40, 158–164. https://doi.org/10.1016/j.bodyim.2021.12.009

Reel, J., Voelker, D., & Greenleaf, C. (2015). Weight status and body image perceptions in adolescents: current perspectives. Adolescent Health, Medicine and Therapeutics, 6(6), 149–158. https://doi.org/10.2147/ahmt.s68344

Rodgers, R. F. (2015). The relationship between body image concerns, eating disorders and internet use, part II: An integrated theoretical model. Adolescent Research Review, 1(2), 121–137. https://doi.org/10.1007/s40894-015-0017-5

Rosenthal-von der Pütten, A. M., Hastall, M. R., Köcher, S., Meske, C., Heinrich, T., Labrenz, F., & Ocklenburg, S. (2019). “Likes” as social rewards: Their role in online social comparison and decisions to like other people’s selfies. Computers in Human Behavior, 92, 76–86. https://doi.org/10.1016/j.chb.2018.10.017

Saiphoo, A. N., & Vahedi, Z. (2019). A meta-analytic review of the relationship between social media use and body image disturbance. Computers in Human Behavior, 101(1), 259–275. https://doi.org/10.1016/j.chb.2019.07.028

Sanzari, C. M., Gorrell, S., Anderson, L. M., Reilly, E. E., Niemiec, M. A., Orloff, N. C., Anderson, D. A., & Hormes, J. M. (2023). The impact of social media use on body image and disordered eating behaviors: Content matters more than duration of exposure. Eating Behaviors, 49(101722), 101722. https://doi.org/10.1016/j.eatbeh.2023.101722

Skogen, J. C., Hjetland, G. J., Bøe, T., Hella, R. T., & Knudsen, A. K. (2021). Through the looking glass of social media. Focus on self-presentation and association with mental health and quality of life. A cross-sectional survey-based study. International Journal of Environmental Research and Public Health, 18(6), 3319. https://doi.org/10.3390/ijerph18063319

Smith, A. R., Hames, J. L., & Joiner, T. E. (2013). Status update: Maladaptive Facebook usage predicts increases in body dissatisfaction and bulimic symptoms. Journal of Affective Disorders, 149(1–3), 235–240. https://doi.org/10.1016/j.jad.2013.01.032

Sun, X., Molaib, K., Xu, T., Ram., N., Reeves, B., Desai, M., & Robinson, T. N. (2025). Objectively measured social media use and weight concerns and dieting in adolescents. JAMA Pediatrics. Published online December 01, 2025. https://doi.org/10.1001/jamapediatrics.2025.4744

Swami, V., Weis, L., Barron, D., & Furnham, A. (2017). Positive body image is positively associated with hedonic (emotional) and eudaimonic (psychological and social) well-being in British adults. The Journal of Social Psychology, 158(5), 541–552. https://doi.org/10.1080/00224545.2017.1392278

Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-Dunn, S. (1999). Exacting beauty: Theory, assessment, and treatment of body image disturbance. American Psychological Association.  https://doi.org/10.1037/10312-000

Thompson, J. K., & Heinberg, L. J. (1999). The media’s influence on body image disturbance and eating disorders: We’ve reviled them, now can we rehabilitate them? Journal of Social Issues, 55(2), 339–353. https://doi.org/10.1111/0022-4537.00119

Thompson, J. K., & Schaefer, L. M. (2019). Thomas F. Cash: A multidimensional innovator in the measurement of body image; Some lessons learned and some lessons for the future of the field. Body Image, 31, 198–203. https://doi.org/10.1016/j.bodyim.2019.08.006

Tiggemann, M., & Anderberg, I. (2019). Social media is not real: The effect of “Instagram vs reality” images on women’s social comparison and body image. New Media & Society, 22(12), 2183–2199. https://doi.org/10.1177/1461444819888720

Tylka, T. L., & Wood-Barcalow, N. L. (2015). What is and what is not positive body image? Conceptual foundations and construct definition. Body Image, 14(14), 118–129. https://doi.org/10.1016/j.bodyim.2015.04.001

Vandenbosch, L., & Eggermont, S. (2016). The interrelated roles of mass media and social media in adolescents’ development of an objectified selfconcept. Communication Research, 43(8), 1116–1140. https://doi.org/10.1177/0093650215600488

Weiss, S. (2018). 4 problematic trends I see on body positive Instagrams. The Establishment. https://medium.com/the-establishment/4-problematictrends-i-see-on-body-positive-instagrams-6ee52b7f339b

Wick, M. R., & Keel, P. K. (2020). Posting edited photos of the self: Increasing eating disorder risk or harmless behavior? International Journal of Eating Disorders, 53(6), 864–872. https://doi.org/10.1002/eat.23263

Wright, R. R., Evans, A., Schaeffer, C., Mullins, R., & Cast, L. (2021). Social networking site use: Implications for health and wellness. Psi Chi Journal of Psychological Research, 26(2), 165–175. https://doi.org/10.24839/2325-7342.jn26.2.165

Wright, R. R., Schaeffer, C., Mullins, R., Evans, A., & Cast, L. (2020). Comparison of student health and well-being profiles and social media use. Psi Chi Journal of Psychological Research, 25(1), 14–21. https://doi.org/10.24839/2325-7342.jn25.1.14

Yang, C. (2016). Instagram use, loneliness, and social comparison orientation: Interact and browse on social media, but don’t compare. Cyberpsychology, Behavior, and Social Networking, 19(12), 703–708. https://doi.org/10.1089/cyber.2016.0201

Yau, J. C., & Reich, S. M. (2018). “It’s just a lot of work”: Adolescents’ selfpresentation norms and practices on Facebook and Instagram. Journal of Research on Adolescence, 29(1), 196–209. https://doi.org/10.1111/jora.12376

Zheng, A., Duff, B. R. L., Vargas, P., & Yao, M. Z. (2020). Self-presentation on social media: When self-enhancement confronts self-verification. Journal of Interactive Advertising, 20(3), 1–35. https://doi.org/10.1080/15252019.2020.1841048

Author Note

Isabella Cerase

https://orcid.org/0000­0002­2273­0536

The authors report there are no competing interests to declare. We are thankful for the high school girls who served as participants on this research, as this project would not have been possible without their data.

Isabella Cerase played a lead role in conceptualization and substantive original writing, research design and data collection, analysis, and interpretation. Mylene Feiler played a supporting role in conceptualization, data collection, and original writing. Sean Dougherty played a supporting role in minor original writing. Madeline Dougherty played a supporting role in data analysis and interpretation, as well as original writing.

Correspondence concerning this article should be addressed to Isabella Cerase, Department of Psychology at Florida State University, 1107 West Call Street, Tallahassee, FL, 32304, United States. Email: isc22b@fsu.edu

APPENDIX

Adapted Measures of Body Dissatisfaction

1. I wish I looked better.

2. I am satisfied with how I look.

3. My body makes me feel insecure.

4. I think I have a good body.

5. I do not feel good about my body.

Adapted Measures of Body Appreciation

6. I do not allow unrealistically thin images of women presented in the media to affect my attitudes toward my body.

7. When I look at social media images, I remind myself that the models’ images are altered.

8. I feel like I am beautiful even if I am different from media images of attractive people (e.g., models, actresses/actors).

9. My self-worth is independent of my body shape or weight.

10. I do not spend a lot of energy and time worrying about my shape or weight

Note: Questions were adapted from the Body Appreciation Scale (Avalos et al., 2005), the Body-Esteem Scale for Adolescents and Adults (Mendelson et al., 2001), and Body Appreciation Scale- 2 (Tylka & Wood-Barcalow, 2015).

SThe Relationships of Perceived Burdensomeness, Loneliness, and Religious Strain With Suicidal Ideation in LGBTQ+ and Non-LGBTQ+ College Students

Skyler Woolley, Abby Allen, Grace Collier, Jada Nagel, Jacob Schultz, Sarah Loertscher, Daniel Hatch*, Kirsten L. Graham*, and Bryan L. Koenig*

Department of Psychology, Southern Utah University

ABSTRACT. Suicide is a growing problem in the United States and a leading cause of death in individuals between the ages 10 and 34 years old. College students who identify as part of the LGBTQ+ community are 2 to 3 times more likely to complete suicide than those outside the community. Many factors predict suicidal ideation. This study focuses on perceived burdensomeness, loneliness, and religious strain as predictors of suicidal ideation in LGBTQ+ and non­LGBTQ+ students. Prior studies have found that these 3 variables contribute to suicidality, but this is the first study looking at the 3 variables together. It was hypothesized that all 3 variables would predict suicidal ideation in both LGBTQ+ students and non­LGBTQ+ students, but given the religious societal context of our participants, that religious strain would have the largest effect size and explain why LGBTQ+ students have higher suicidality compared to non­LGBTQ+ students. A final sample comprised 815 college students who answered questionnaires online. Contrary to expectations, in a multivariate regression model, religious strain did not predict suicidality among LGBTQ+ students. Instead, perceived burdensomeness had the largest effect size, for both LGBTQ+ and non­LGBTQ+ students. Perceived burdensomeness also best explained the difference in suicidality between the groups. These findings suggest that burdensomeness is a key predictor of suicidal ideation for both LGBTQ+ students and non­LGBTQ+ students. Prioritizing interventions aimed at preventing and alleviating feelings of perceived burdensomeness in individuals potentially struggling with suicidal ideation may be particularly beneficial.

Keywords: perceived burdensomeness, loneliness, religious strain, LGBTQ+, college students

uicide is the second leading cause of death in the United States among individuals between the ages of 10 and 34 years old (Centers for Disease Control and Prevention, 2025). Suicidality is especially common among young members of the LGBTQ+ community; indeed, estimates suggest that one individual between the ages of 13 and 24 who identifies with the LGBTQ+ community will attempt suicide every 45 seconds (The Trevor Project, 2021). Out of the 712,990 suicide attempts made during a one­year period by LGBTQ+ individuals, about 29% were college­aged (The Trevor Project, 2021). Compared to cisgender heterosexual

college students, LGBTQ+ students are 2–3 times more likely to complete suicide (Ohio State University Suicide Prevention Program, 2021). Numerous factors have been found to predict suicidal ideation. The present study examined three: perceived burdensomeness, loneliness, and religious strain.

Perceived Burdensomeness

Perceived burdensomeness is a mental state in which individuals believe that others would be better off without them (Van Orden et al., 2012). The interpersonal theory of suicidality (Van Orden et al., 2010) proposes that suicidal ideation is more likely to occur when someone feels hopelessness, resulting from perceived burdensomeness combined with thwarted belonging. The theory conceptualizes thwarted belonging as a mental state brought on when the need for connection is unmet. Building on this, Van Orden et al. (2012) found that factors such as family discord and unemployment may be associated with an increase in suicidal ideation because these circumstances may increase feelings of perceived burdensomeness.

These interpersonal factors might disproportionately affect members of the LGBTQ+ community. Supporting this possibility, a study focused on LGBTQ+ identity and perceived burdensomeness found that a significantly higher number of individuals who identified as gender nonconforming were homeless, and those who were homeless were more likely to have told their parents about their identity (Semborski et al., 2021). Notably, homelessness and nondisclosure of identity to parents each correlated with higher levels of perceived burdensomeness, whereas more social connection predicted lower perceived burdensomeness. Another study found that gay, lesbian, or bisexual college students who face rejection due to their sexual orientation experience heightened levels of perceived burdensomeness (Hill & Pettit, 2012). Such findings suggest that when individuals from the LGBTQ+ community encounter such rejection they may develop beliefs about failing to meet societal expectations, or that others are ashamed of their sexual identity, both of which could increase the individual’s feelings of perceived burdensomeness.

Religious Strain

The perceived burdensomeness that contributes to suicidal ideation in LGBTQ+ individuals may also result from feelings of rejection associated with having a religious upbringing (Semborski et al., 2022). Teen suicide rates doubled in Utah over four years. Simultaneously, the majority religion, the Church of Jesus Christ of Latterday Saints (LDS), increased discourse against same­sex couples ( Annor et al. , 2018). Although teen suicide

increased drastically in Utah, the rest of the country did not have the same increase in suicide rates. According to a study that examined the impact of religious conflict, LGBTQ+ individuals raised in religious environments and who experienced conflicts between their religious beliefs and their sexuality were more likely to experience higher rates of internalized homophobia compared to LGBTQ+ individuals who did not have a religious upbringing (Gibbs & Goldbach, 2015). This internalized homophobia was associated with higher rates of chronic suicidal thoughts. Skidmore (2021) found that religious commitment can be a protective factor, but suicidal ideation increased with exposure to minority stress—defined as stress from stigma, perceived or actual discrimination, or expectations of rejection due to minority status (Meyer, 1995). Moreover, feelings of perceived belonging were associated with decreased suicidal ideation among sexual minorities and may help with minority stressors (Skidmore, 2021).

A study looking at religious conflict experienced by LGBTQ+ individuals found that one of the most severe consequences of anti­LGBTQ+ teachings was a formed belief that God had rejected them (Shuck & Liddle, 2001). The emotional effects were often severe, including feelings of guilt and shame about one’s own identity. Many of the participants reported depression or suicidal ideation. Dahl and Galliger (2009) found that, for individuals with both LGBTQ+ and religious identities, two­thirds experience conflict between the identities. Two factors were commonly reported for members of the LGBTQ+ community who have religious affiliations to integrate their identities successfully: accepting one’s identity and having a sense of completeness. Dahl and Galliger reported that many LGBTQ+ young adults choose to disengage from their religious affiliation, but when LGBTQ+ individuals stay in a religious group, their focus on their relationship with God is crucial to integrating their identities. Having a perception that God made them LGBTQ+ is beneficial with this integration.

Loneliness

The social conflict experienced by LGBTQ+ youth resulting from prejudice—religious or otherwise—could not only increase perceived burdensomeness (Hill & Pettit, 2012) and undermine relationships with the divine (Skidmore, 2021), but also increase social isolation (Aranmolate et al., 2017). Loneliness can result from loss of important social groups and social identities, it is negatively correlated with social connection and psychological resources, and it is positively correlated with a history of mental health struggles (Hayes et al., 2022).

A study comparing rates of loneliness in LGBTQ+ and cisgender heterosexual groups found that the LGBTQ+

group reported significantly higher rates of loneliness and depression (Herrmann et al., 2022). These results indicate that loneliness is especially prevalent for LGBTQ+ individuals. Another study looking at gay, lesbian, and bisexual college students found similar results, with LGBTQ+ individuals having higher rates of depression and loneliness and fewer reported reasons for living compared to their heterosexual counterparts (Westefeld et al., 2008).

Aranmolate (2017) found that negative attitudes from family and peers toward LGBTQ+ individuals predicted feelings of isolation, rejection, and loss of support. These outcomes further predicted worse mental health, including stress, depression, substance use, and suicidal ideation. In contrast, having a sense of belonging was preventative regarding depression, which helped protect against suicidal ideation and self­harming behaviors. These findings suggest that providing a sense of belonging for LGBTQ+ individuals could involve an active display of support from one’s family and community.

Purpose

Burdensomeness, loneliness, and religious strain are leading predictors of suicidal ideation in LGBTQ+ students. Prior studies have found that these three variables contribute to suicidality, but this was the first study looking at the three variables together. The current study aimed to compare the relative impact of these predictors in LGBTQ+ students as compared with their non­LGBTQ+ peers. Given that our participants were students in the highly religious state of Utah, being the sixth most religious state, where 48% attend religious services regularly (Rotolo et al., 2025) and 78% have religious affiliations (Kiprop, 2018), we hypothesized that religious strain would be the strongest predictor of suicidal ideation in LGBTQ+ participants, but that all three variables would predict suicidal ideation within both the LGBTQ+ and non­LGBTQ+ samples. Moreover, given the prior evidence that LGBTQ+ students have higher perceived burdensomeness, loneliness, and religious strain, we wanted to see if these three variables can explain the greater levels of suicidality among LGBTQ+ students. Understanding how these variables affect suicidal ideation may help to improve interventions and resources to prevent suicide among marginalized and under researched populations such as LGBTQ+ students.

Method

Procedure

The current report uses data that were collected as part of a larger project on college student mental health. Prior to data collection, approval was received from the university’s Institutional Review Board. Participants

were recruited from a research pool of general psychology students and through the login portal announcement system at a regional university in Utah, a state in the western United States. After providing informed consent, participants completed an online survey, which included 12 questionnaires (only relevant results are reported here) and ended with demographics questions.

Participants

After excluding participants due to missing demographics regarding their gender identity or sexual orientation, the final sample was 815 students. Their mean age was 21.10 years old (SD = 5.27), with a minimum age of 17 and a maximum age of 62. Seventy percent identified as cisgender and heterosexual, and the other 30.0% identified as LGBTQ+. Similar to the demographics at the university, most participants were White (83.2%), with 3.4% Hispanic, 2.7% Asian American, 1.1% African American, 0.5% American Indian, 0.4% Hawaiian, and 8.7% selected other race, not specified Regarding biological sex as determined at birth, 72.1% were women and 27.9% were men. Of nonLGBTQ+ participants, most were LDS (65.6%), followed by 10.0% non­LDS Christian, 5.1% agnostic, 4.0% atheist, 3.9% Catholic, 2.8% self­identified, 0.4% Buddhist, and 8.3% indicated no preference. Of LGBTQ+ participants, 25.4% were agnostic, 23.4% LDS, 13.9% atheist, 13.9% self­identified, 6.1% non­LDS Christian, 2.9% Catholic, 0.4% Buddhist, 0.4% Jewish, and 13.5% indicated no preference. Of participants, 43.7% were first­year students, 20.2% sophomores, 15.6% juniors, 15.6% seniors, and 4.9% graduate students. Most were single and never married (66.7%); 10.8% were married, not separated; and 22.5% were in a relationship and not married.

Measures

Perceived Burdensomeness

Perceived burdensomeness was assessed using the Interpersonal Needs Questionnaire’s Burdensomeness Subscale (Van Orden et al., 2012). It is 6 questions about recent feelings, focusing beliefs and experiences about being a burden to others. An example item is, “These days the people in my life would be better off if I were gone.” Responses were rated on a 7­point Likert scale (1 = not at all true for me, 7 = very true for me). These scores are averaged to make each participant’s score for perceived burdensomeness. The scale has shown good internal reliability previously (Marty et al., 2012) and in the current study (α = .94).

Loneliness

Loneliness was assessed using the UCLA Loneliness Scale (Hughes et al., 2004). It is 6 items about emotional Suicidal

and Non-LGBTQ+ Students | Woolley, Allen, Collier, Nagel, Schultz, Loertscher, Hatch, Graham, and Koenig

and social loneliness, with positive and negative wording. An example item is, “I lack companionship.” Responses had four options (1 = never, 4 = often). After reverse scoring, scores are summed to calculate each participant’s total score. The psychometric properties of this scale have been supported (Neto, 2014; current study: α = .86).

Religious Strain

Religious strain was assessed using the Divine Subscale of the Religious and Spiritual Struggles Scale (Exline et al., 2014). It has 5 items about religious and spiritual struggles over the last month related to one’s relationship with God or the divine. An example item is, “Felt as though God had let me down.” Items were rated on a 5­point Likert scale (1 = not at all/does not apply, 5 = a great deal). The scores are averaged for each participant. This scale has good reliability (Wüthrich­Grossenbacher et al., 2023; current study: α = .91).

Suicidality

Suicidality was assessed using the Suicide Behaviors Questionnaire ­ Revised (Osman et al., 2001). It has 4 items about suicide ideation and attempts. An example item is, “How likely is it that you will attempt suicide someday?” Response options vary across items, and each person’s final score is their sum on the individual items, indicating overall suicidality. It has good internal reliability (Osman et al, 2001; current study: α = .82).

Analysis Plan

The data was analyzed using SPSS v28. To test our hypotheses, multivariate regressions examined how the three variables (burdensomeness, loneliness, and religious strain) predicted suicidality separately in the LGBTQ+ and non­LGBTQ+ groups. Means were compared across LGBTQ+ and non­LGBTQ+ students for burdensomeness, loneliness, divine religious strain, and suicidality. Forward regression then added one variable at a time to a series of multivariate regression models, adding the variable that is the strongest predictor of suicidality among the mental health variables of burdensomeness, loneliness, and religious strain. This allowed the evaluation of which variables could best explain the greater suicidality of LGBTQ+ students compared to non­LGBTQ+ students, and whether these variables together could fully explain the greater suicidality of LGBTQ+ students.

Results

We calculated descriptive statistics for LGBTQ+ students: suicidal ideation ( M = 10.30, SD = 3.89), perceived burdensomeness ( M = 3.06, SD = 1.64), loneliness ( M = 17.77, SD = 3.89), and divine religious strain

TABLE 1

Correlations Amoung Key Variables

Note. Correlations among theoretical variables. Correlations for LGBTQ+ students (n = 237) are above the diagonal. Correlations for non-LGBTQ+ students (n = 586) are below the diagonal.

1

Regression for the LGBTQ+ Sample

Note. Multiple regression models showing burdensomeness has a larger effect size than loneliness and religious strain when predicting suicidality in the LGBTQ+ sample.

Multiple Regression for the Non-LGBTQ+ Sample

Note. Multiple regression models showing burdensomeness has a larger effect size than loneliness and religious strain when predicting suicidality in the non-LGBTQ+ sample.

(M = 2.00, SD = 1.24); and for non­LGBTQ+ students: suicidal ideation ( M = 6.70, SD = 3.44), perceived burdensomeness ( M = 2.05, SD = 1.25), loneliness ( M = 15.40, SD = 4.28), and divine religious strain (M = 1.65, SD = 0.98). See Table 1 for correlations among key variables.

Frequency statistics were calculated on the suicidal variable in the LGBTQ+ students, 76.2% being at an elevated risk for suicidal ideation and attempts (scoring validated, Osman et al., 2001), and non ­ LGBTQ+ students, with 34.4% at an elevated risk for suicidal ideation and attempts.

Multivariate regression analyses predicted suicidality from burdensomeness, divine religious strain, and loneliness in the LGBTQ+ students, R2 = .44, F(3, 233) = 61.27. Burdensomeness was the strongest predictor, β = .57, b = 1.27, SE = 0.14, t(233) = 9.38, p < .001. Loneliness was the next strongest predictor, β = .14, b = 0.13, SE = 0.06, t(233) = 2.36, p = .019. Divine religious strain was nonsignificant, β = .03, b = 0.10, SE = 0.16, t(233) = 0.62, p = .536. For the regression predicting suicidality for the non­LGBTQ+ students, R2 = .41, F(3, 584) = 132.27, burdensomeness was the strongest predictor, β = .51, b = 1.40, SE = 0.11, t(564) = 12.41, p < .001, loneliness the next strongest predictor, β = .15, b = 0.12, SE = 0.03, t(564) = 3.82, p = < .001, and divine religious strain was the weakest predictor, β = .09, b = 0.32, SE = 0.13, t(564) = 2.52, p = .012. Overall, burdensomeness was the strongest predictor of suicidality in both groups, followed by loneliness, and then divine religious strain (see Figures 1 and 2). R2 values were similar in the two groups.

LGBTQ+ students were worse off than non­LGBTQ+ students on all of the mental health variables. LGBTQ+ students scored higher in suicidal ideation on average compared to non­LGBTQ+ students, t(809) = 13.35, p < .001, d = 1.02 (see Figure 3). Similar patterns occurred for the other variables. Compared to their non­LGBTQ+ counterparts, LGBTQ+ students experienced more divine religious strain, t(806) = 4.22, p < .001, d = 0.33; loneliness, t(810) = 7.44, p < .001, d = 0.57; and burdensomeness, t(811) = 9.65, p < .001, d = 0.74 (see Figures 4–6).

Forward stepwise regression was then used to evaluate whether the higher levels of struggles with religious strain, loneliness, and perceived burdensomeness experienced by LGBTQ+ students could explain their higher levels of suicidality. An initial regression model predicting suicidality had only LGBTQ+, which was dummy coded (1 = LGBTQ+, 0 = non­LGBTQ+), whose β = .42, p < .001. This beta indicated greater suicidality among the LGBTQ+ students compared to the nonLGBTQ+ students. Forward regression showed, when burdensomeness was accounted for in Model 2, the beta for LGBTQ+ decreased to .23, p < .001, indicating

Suicidal Ideation in LGBTQ+ and Non-LGBTQ+ Students | Woolley, Allen, Collier, Nagel, Schultz, Loertscher, Hatch, Graham, and Koenig

that perceived burdensomeness explained a substantial amount of why LGBTQ+ students had higher suicidality than non­LGBTQ+ students. However, the LGBTQ+ status variable was still significant and positive, indicating that LGBTQ+ students were still higher in suicidality even after controlling for burdensomeness. In Model 3, when loneliness was added, the beta for LGBTQ+ status changed from .23 to .22, suggesting that loneliness did little to explain the remaining greater suicidality among LGBTQ+ students when already controlling for burdensomeness. Similarly, adding divine religious strain in Model 4 did not notably affect the beta of LGBTQ+, which stayed at .22 (see Table 2).

Discussion

Originally, the hypothesis was that burdensomeness, loneliness, and religious strain would predict suicidality in both the LGBTQ+ and non­LGBTQ+ groups. Religious strain was also predicted to be the strongest predictor of suicidality and best explain why LGBTQ+ students have higher suicidality than non­LGBTQ+ students. However, the results indicated that—when controlling for burdensomeness—loneliness and religious strain had small to no association with suicidality. Results also showed that, compared to non­LGBTQ+ students, LGBTQ+ students reported more suicidality, perceived burdensomeness, loneliness, and divine religious strain. The higher mean suicidality in LGBTQ+ students was partially explained by their greater levels of perceived burdensomeness but not by their greater levels of loneliness or divine religious strain. Overall, the findings indicate that perceived burdensomeness is a strong predictor of suicidality and that it explains much but not all of why LGBTQ+ students were more suicidal than non­LGBTQ+ students.

Our hypothesis that religious strain would have the largest effect on suicidal ideation was not supported. Perhaps this is because prior studies on which we based our hypothesis did not control for burdensomeness. In addition, in our LGBTQ+ sample, 33.2% of individuals reported a religious affiliation and another 13.9% selfidentified while 79.9% of the non­LGBTQ+ participants reported religious affiliation. Skidmore (2021) found that religion can be a protective factor when there are not minority stressors. Perhaps our sample had a weaker relationship of religious strain with suicidality due to protective factors that can be found in religion. This could partially explain why religious strain has small to no effect on suicidal ideation, while controlling for the other variables. Another possible explanation is the measure of religious strain. We used the Divine Subscale of the Religious and Spiritual Struggles Scale (Exline et al., 2014), which examines one’s relationship with

God or the divine, but not other types of religious or spiritual struggles such as those pertaining to a religious community. Using only this subscale limits the ability to evaluate other types of religious strain as a predictor of suicidal ideation.

These findings have important clinical and outreach implications. They suggest that interventions focusing on perceived burdensomeness might be especially valuable for college students generally, and LGBTQ+ students in particular. The causal processes leading to suicidality are

FIGURE 6

Perceived Burdensomeness

TABLE 2

Predictors of Suicidality in LGBTQ+ College Students

Note. Forward regression models with suicidal ideation as the outcome variable, used to evaluate the degree to which perceived burdensomeness, loneliness, and divine religious strain can explain the greater suicidality among LGBTQ+ students compared to non-LGBTQ+ students.

Suicidal Ideation in LGBTQ+ and Non-LGBTQ+ Students | Woolley, Allen, Collier, Nagel, Schultz, Loertscher, Hatch, Graham, and Koenig

complex, but interventions targeting psychological states that are proximal to suicidality might be especially likely to impact suicidality. Indeed, Semborski and colleagues (2022) suggested that LGBTQ+ youth may benefit from suicide prevention efforts and that having strong familial support can help with rates of suicidality, especially in situations like homelessness, where they experience higher rates of perceived burdensomeness. This is in agreement with Hill and Pettit (2012), who proposed that interventions for suicidal prevention programs should focus on perceived burdensomeness, specifically for students in the LGBTQ+ community who have high expectations of themselves or perceived rejection.

The current study has some important limitations. Most importantly, the use of correlational methods limits the inference of causal relationships among the variables. Our sample was also self­selected and recruited at a single university in the western United States, which limits our ability to generalize to other demographic groups in other locations. Our self­selected sample consisted of 72.1% women, but the national average for undergraduates was 57.3% women in 2022 (Hanson, 2025). This is a significant limitation as some mental health conditions are more often reported by women, such as diagnosed depression rates, which are 1.8 times higher in women than men with 1.5 times higher rates of frequent mental distress (United Health Foundation, 2023). The sample was also predominantly White (83.2%), which is higher than the national average of college students (54.77%) in 2022 (Hanson, 2025). These sample characteristics limit the generalizability of our results because mental health conditions are more commonly reported by White individuals compared to ethnic minorities (Panchal, 2022). Additionally, White adults with mental illnesses are more likely to receive mental health services (46.3%) compared to minority groups, such as Black adults (30%), and Hispanic adults (27%). When ethnic minority individuals receive mental health services, it is often of a poorer quality (National Alliance on Mental Illness, 2021). This is a limitation because our sample might have a higher percentage of participants experiencing mental health conditions but at a lower severity, compared to the average college student. The current study also had notable strengths. The sample size was relatively large, and the measures are well­established and demonstrated strong statistical associations. Future research might benefit by recruiting participants from a variety of universities in order to increase the diversity of the sample. Furthermore, it would be valuable to also recruit younger people to see if the results generalize to American adolescents. Given the overall increase observed in suicide across decades in the United States (Garnett & Curtin, 2023),

and the greater suicide rates of LGBTQ+ youth (Johns et al., 2020), enhanced understanding of predictors and causes of suicidality and related interventions to decrease suicidality are of critical importance.

References

Annor, F. B., Zwald, M. L., Wilkinson, A., Friedrichs, M., Fondario, A., Dunn, A., Nakashima, A., Gilbert, L., & Ivey-Stephenson, A. Z. (2018). Characteristics of and precipitating circumstances surrounding suicide among persons aged 10–17 years: Utah, 2011–2015. Center for Disease Control and Prevention, MMWR Supplements, 67(11), 329–332. https://dx.doi.org/10.15585/mmwr.mm6711a4

Aranmolate, R., Bogan, D. R., Hoard, T., & Mawson, A. R. (2017). Suicide risk factors among LGBTQ youth. Journal of Scientific Medicine Schizophrenia, 2(2), Article 1011. https://www.jscimedcentral.com/jounal-article-pdfd/JSMSchizophrenia/schizophrenia-2-1011.pdf

Centers for Disease Control and Prevention. (2025). Suicide prevention Retrieved November 1, 2025, from https://www.cdc.gov/suicide/facts/ index.html#cdc_facts_stats_intro-leading-cause-of-death

Dahl, A., & Galliher, R. (2009). LGBQQ young adult experiences of religious and sexual identity integration. Journal of LGBT Issues in Counseling, 3(2), 92–112. https://dx.doi.org/10.1080/15538600903005268

Exline, J. J., Pargament, K. I., Grubbs, J. B., & Yali, A. M. (2014). The Religious and Spiritual Struggles Scale: Development and initial validation. Psychology of Religion and Spirituality, 6(3), 208–222. https://dx.doi.org/10.1037/a0036465

Garnett, M. F., & Curtin, S.C. (2023, April). Suicide mortality in the United States, 2001–2021. NCHS Data Brief, no 464. Hyattsville, MD: National Center for Health Statistics. Retrieved September 21, 2025, from https://dx.doi.org/10.15620/cdc:125705

Gibbs, J. J., & Goldbach, J. (2015). Religious conflict, sexual identity, and suicidal behaviors among LGBT young adults. Archives of Suicide Research, 19(4), 472–488. https://doi.org/10.1080/13811118.2015.1004476

Hanson, M. (2025). College enrollment & student demographic statistics Education Data Initiative. Retrieved February 19, 2026, from https://educationdata.org/college-enrollment-statistics

Hayes, S., Carlyle, M., Haslam, A., Haslam, C., & Dingle, G. (2022). Exploring links between social identity, emotion regulation, and loneliness in those with and without a history of mental illness. British Journal of Clinical Psychology, 61(3), 701–734. https://doi.org/10.1111/bjc.12358

Herrmann, W. J., Oeser, P., Buspavanich, P., Lech, S., Berger, M., & Gellert, P. (2022). Loneliness and depressive symptoms differ by sexual orientation and gender identity during physical distancing measures in response to COVID‐19 pandemic in Germany. Applied Psychology: Health and WellBeing, 15(1), 80–96. https://doi.org/10.1111/aphw.12376

Hill, R. M., & Pettit, J. W. (2012). Suicidal ideation and sexual orientation in college students: The roles of perceived burdensomeness, thwarted belongingness, and perceived rejection due to sexual orientation. Suicide and Life-Threatening Behavior, 42(5), 567–579. https://doi.org/10.1111/j.1943-278X.2012.00113.x

Hughes, M. E., Waite, L. J., Hawkley, L. C., & Cacioppo, J. T. (2004). A short scale for measuring loneliness in large surveys: Results from two populationbased studies. Research on Aging, 26(6), 655–672. https://doi.org/10.1177/0164027504268574

Johns, M. M., Lowry, R., Haderxhanaj, L. T., Rasberry, C. N., Robin, L., Scales, L., Stone, D., & Suarez, N. A. (2020). Trends in violence, victimization, and suicide risk by sexual identity among high school students: Youth Risk Behavior Survey, United States, 2015–2019. Center for Disease Control and Prevention, MMWR Supplements, 69(1), 19–27. https://dx.doi.org/10.15585/mmwr.su6901a3

Kiprop, V. (2018). Major religions practiced in Utah. WorldAtlas. Retrieved October 27, 2025 from https://www.worldatlas.com/articles/what-is-thereligious-composition-of-the-adult-population-of-utah.html

Marty, M. A., Segal, D. L., Coolidge, F. L., & Klebe, K. J. (2012). Analysis of the psychometric properties of the Interpersonal Needs Questionnaire (INQ) among community-dwelling older adults. Journal of Clinical Psychology, 68(9), 1008–1018. https://dx.doi.org/10.1002/jclp.21877

Meyer, I. H. (1995). Minority stress and mental health in gay men. Journal of Health and Social Behavior, 36(1), 38–56. https://doi.org/10.2307/2137286

National Alliance on Mental Illness. (2021). Mental health inequities: Racism and racial discrimination. Retrieved February 19, 2026, from https://www.nami.org/advocacy-at-nami/policy-positions/supportingcommunity-inclusion-and-non-discrimination/mental-health-inequities-

and

Students | Woolley, Allen, Collier, Nagel, Schultz, Loertscher, Hatch, Graham, and Koenig

racism-and-racial-discrimination/ Neto, F. (2014). Psychometric analysis of the short-form UCLA Loneliness Scale (ULS-6) in older adults. European Journal of Ageing 11, 313–319. https://dx.doi.org/10.1007/s10433-014-0312-1

Ohio State University Suicide Prevention Program. (2021). LGBTQIA+ students Retrieved November 1, 2025, from https://web.archive.org/web/20240222172225/https://suicideprevention. osu.edu/prevention-and-intervention/special-populations/glbt/ Osman, A., Bagge, C. L., Gutierrez, P. M., Konick, L. C., Kopper, B. A., & Barrios, F. X. (2001). The Suicidal Behaviors Questionnaire-Revised (SBQ-R): Validation with clinical and nonclinical samples. Assessment, 8(4), 443–454. https://doi.org/10.1177/107319110100800409

Panchal, N., Saunders, H., & Ndugga, N. (2022). Five key findings on mental health and substance use disorders by race/ethnicity. Kaiser Family Foundation. Retrieved February 19, 2026, from https://www.kff.org/mental-health/five-key-findings-on-mental-healthand-substance-use-disorders-by-race-ethnicity/

Rotolo, C., Wormald, B., Webster, B., & Coleman, J. (2025). How religious is your state? Comparing states on key measures of religiousness. Pew Research Center. Retrieved October 26, 2025 from https://www.pewresearch.org/ short-reads/2025/09/16/how-religious-is-your-state/?selectedState=UT Ryan, R., & Deci, E. (2000). The darker and brighter sides of human existence: Basic psychological needs as a unifying concept. Psychological Inquiry, 11(4), 319–338. https://doi.org/10.1207/S15327965PLI1104_03

Shuck, K. D., & Liddle, B. J. (2001). Religious conflicts experienced by lesbian, gay, and bisexual individuals. Journal of Gay & Lesbian Psychotherapy, 5(2), 63–82. https://doi.org/10.1300/J236v05n02_07

Semborski, S., Srivastava, A., Rhoades, H., Fulginiti, A., & Glodbach, J. T. (2021). Burden, belonging, and homelessness: Disclosure and social network differences among LGBTQ youth recruited from a suicide crisis service provider. Journal of Homosexuality, 69(5), 894–910. https://doi.org/10.1080/00918369.2021.1898801

Skidmore, S. (2021). Religious sexual minorities and suicide risk: The moderating role of LGBQ and religious belongingness. All Graduate Theses and Dissertations, 8151. https://doi.org/10.26076/3895-9a45

The Trevor Project. (2021). Estimate of how often LGBTQ youth attempt suicide in the U.S. Retrieved March 29, 2023, from https://www.thetrevorproject.org/research-briefs/estimate-of-how-often-

lgbtq-youth-attempt-suicide-in-the-u-s United Health Foundation. (2023). 2023 Mental and behavioral health data brief. Retrieved February 19, 2026 from https://www.americashealthrankings.org/publications/reports/mentaland-behavioral-health-data-brief/key-findings-gender

Van Orden, K. A., Cukrowicz, K. C., Witte, T. K., & Joiner, T. E. (2012). Thwarted belongingness and perceived burdensomeness: Construct validity and psychometric properties of the Interpersonal Needs Questionnaire. Psychological Assessment, 24(1), 197–215. https://doi.org/10.1037/a0025358

Van Orden, K. A., Witte, T. K., Cukrowicz, K. C., Braithwaite, S., Shelby, E. A., & Joiner, T. E. (2010). The interpersonal theory of suicide. Psychology Review, 117(2), 575–600. https://doi.org/10.1037/a0018697

Westefeld, J. S., Maples, M. R., Buford, B., & Taylor, S. (2008). Gay, lesbian, and bisexual college students: The relationship between sexual orientation and depression, loneliness, and suicide. Journal of College Student Psychotherapy, 15(3), 71–82. https://doi.org/10.1300/J035v15n03_06

Wüthrich-Grossenbacher, U., Mutsinze, A., Wolf, U., Maponga, C. C., Midzi, N., Mutsaka-Makuvaza, M. J., & Merten, S. (2023). A validation of the religious and spiritual struggles scale among young people living with HIV in Zimbabwe: Mokken scale analysis and exploratory factor analysis. Frontiers in Psychology, 14, 1–10. https://doi.org/10.3389/fpsyg.2023.1051455

Author Note.

This project was presented at Southern Utah Universities Festival of Excellence in Cedar City, Utah, and the Rocky Mountain Psychological Association in Albuquerque, New Mexico. We have no conflicts of interest with this project.

Skyler Woolley played a lead role in substantive original writing and data analysis and interpretation. Abby Allen played a lead role in conceptualization and data analysis. Grace Collier played a lead role in conceptualization and research design. Jada Nagel played a supporting role in original writing. Jacob Schultz and Sarah Loertscher played a supporting role in data analysis. Daniel Hatch and Kirsten L. Graham played a supporting role in supervising. Bryan L. Koenig played a lead role in supervising, data collection, and data analysis.

Correspondence concerning this article may be addressed to Skyler Woolley at woolley.sky@gmail.com

Publish Your Research in Psi Chi Journal

Undergraduate, graduate, and faculty submissions are welcome year round. Only one author (either first author or coauthor) is required to be a Psi Chi member. All submissions are free. Reasons to submit include

• a unique, doctoral­level, peer­review process

• indexing in PsycInfo, EBSCO, Web of Science's Emerging Sources Citation Index, and Crossref databases

• free access of all articles at psichi.org

• our efficient online submissions portal

View Submission Guidelines and submit your research at https://www.psichi.org/page/JN_Submissions

Become a Journal Reviewer

Doctoral­level faculty in psychology and related fields who are passionate about educating others on conducting and reporting quality empirical research are invited become reviewers for Psi Chi Journal. Our editorial team is uniquely dedicated to mentorship and promoting professional development of our authors—Please join us!

To become a reviewer, visit https://www.psichi.org/page/JN_BecomeAReviewer

Resources for Student Research

Looking for solid examples of student manuscripts and educational editorials about conducting psychological research? Download as many free articles to share in your classrooms as you would like.

Search past issues, or articles by subject area or author at https://www.psichi.org/journal_past

Learn About Psi Chi

Psi Chi is the International Honor Society in Psychology. Membership is primarily open to undergraduates, graduate students, transfer students, full­time and part­time faculty members, and alumni.

See membership benefits and a link to apply at https://www.psichi.org/page/member_benefits

Register an account: https://pcj.msubmit.net/cgi-bin/main.plex