All correspondence concerning this article should be addressed to Jennifer L. Warnick, PhD, Department of Psychiatry & Human Behavior, Alpert Medical School of Brown University, 196 Richmond St, Providence, RI 02903, USA. E-mail: ude.nworb@kcinraw_refinneJ
Received 2020 Nov 29; Revised 2021 Sep 23; Accepted 2021 Sep 29.Copyright © The Author(s) 2021. Published by Oxford University Press on behalf of the Society of Pediatric Psychology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Over the past two decades, there has been a steady increase in research focused on the association between weight-based stigma and mental health outcomes in children and adolescents. The present study is a systematic review and meta-analysis of the associations between weight stigma and mental health in youth.
A systematic search of PubMed, PsychInfo, and Embase databases was conducted in January 2020. Inclusion criteria included the following: (a) examined an association between weight stigma and a mental health outcome, (b) mean sample age
Overall, meta-analytic findings using a random-effects model indicated a statistically significant moderate association between weight stigma and poorer mental health outcomes (r = .32, 95% confidence interval [0.292, 0.347], p < .001). Age and study quality each moderated the association between weight stigma and mental health. Generally, the study quality was fair to poor, with many studies lacking validated measurement of weight stigma.
Although there was a significant association between weight stigma and mental health in youth, study quality hinders the current body of literature. Furthermore, findings highlight the lack of consideration of internalized weight stigma in child populations, the importance of using validated measures of weight stigma, and the need for increased awareness of how these associations affect populations of diverse backgrounds.
Keywords: adolescents, meta-analysis, preschool children, psychosocial functioning, school-age children, systematic review
Obesity is a national epidemic and is associated with increased risk for morbidity and early mortality in both children and adults (Hales et al., 2017; Ligibel et al., 2014; National Heart, Lung, Blood Institute, National Institute of Diabetes, Digestive, & Kidney Diseases (US), 1998; Skinner et al., 2018). Weight status is often inaccurately associated with negative stereotypes in western culture, specifically that persons with overweight or obesity are lazy, lack discipline, or are not motivated (Pont et al., 2017). Some have also incorrectly suggested that shaming persons with overweight or obesity (i.e., weight stigma) will lead to motivation to lose weight (Callahan, 2013; Puhl & Heuer, 2010; Stuber et al., 2008). Weight stigma, also called weight bias or weight-based discrimination, is the societally constructed rejection or marginalization of people due to negative perceptions of overweight (Puhl & Brownell, 2006). Prior research overwhelmingly supports that these inaccurate beliefs and behaviors from others (i.e., weight stigma) lead to increased food intake and decreased physical activity, which only exacerbates the overweight and obesity epidemic (Puhl & Heuer, 2010). Combined, these beliefs lead to discrimination against persons with overweight or obesity (Puhl & Heuer, 2009).
The formation of self-concept from childhood into adolescence is highly influenced by outside individuals (e.g., peers, parents, teachers; Blakemore & Mills, 2014; Van Harmelen et al., 2017). Experiencing weight stigma from others and/or internalizing those stigmatizing ideals may have harmful effects for social and emotional development in youth. Weight stigma in youth has been related to increased risk for developing numerous psychological and emotional problems, such as depression, anxiety, low self-esteem, and body dissatisfaction (Bucchianeri et al., 2014; Jensen & Steele, 2009; Puhl & Latner, 2007). Furthermore, adolescents who experience weight stigma are at higher risk for self-harm behaviors and suicidal ideation (Eaton et al., 2005; Eisenberg et al., 2003). Suicide is one of the leading causes of death in U.S. youth (Heron, 2019), and mental health problems can lead to severe morbidity and early mortality. Given the devasting potential risks to youth resulting from mental health concerns, a more thorough review of weight stigma and mental health outcomes specific to youth is warranted to better understand how these variables are related, which can then inform future prevention and intervention efforts.
Weight stigma can be experienced both externally (i.e., public stigma) or internally (i.e., weight bias internalization [WBI] and self-stigma). In youth, external weight stigma, defined as prejudice or negative attitudes occurring between two people due to perceived overweight or obesity, is often experienced by peers in a school setting. Youth also report experiencing weight stigma by caregivers, coaches, health-care professionals, educators, and the media (Pont et al., 2017; Puhl et al., 2013). External weight stigma encompasses many concepts including weight-bullying, weight teasing, and weight-based victimization; from here on, each of these will be subsumed within the term external weight stigma. Weight status, particularly in persons with overweight/obesity, is the most common reason for bullying in youth (Puhl et al., 2011); and children with obesity are more likely to experience external weight stigma compared to their healthy weight peers (Harriger & Thompson, 2012; Pont et al., 2017; Puhl & Latner, 2007). However, weight stigma has been reported across the weight spectrum, including in individuals at a healthy weight status among adults (Jackson et al., 2014). External weight stigma has been associated with increased weight over time (Dutton et al., 2014; Jackson et al., 2014; Schvey et al., 2019) as well as unhealthy weight control behaviors and disordered eating (Puhl et al., 2020).
Beyond weight stigma from others, weight stigma can also be experienced internally. Internal weight stigma—referred to as WBI from here on—is when an individual internalizes the experienced stigma and believes the negative attitudes are true for oneself (Corrigan et al., 2001; Durso & Latner, 2008). Specifically, WBI is conceptualized as the personal belief that social stereotypes relating to obesity are true for oneself (e.g., being lazy, unmotivated, “less than”) and includes negative self‐evaluation due to one’s weight (Durso & Latner, 2008). Internalizing society’s inaccurate views of weight can lead to shame, low self-esteem, and unhealthy behaviors such as excess eating and physical inactivity (Jackson et al., 2014; Pila et al., 2015).
Although several studies have explored associations between weight stigma and mental health in youth, no prior work has quantitatively synthesized this body of literature to better understand these associations. Puhl et al. (2020) published a narrative synthesis of the literature on how weight stigma has the potential to contribute to the obesity epidemic, highlighting the relationship between weight stigma and health outcomes broadly (e.g., maladaptive eating behaviors, increased weight gain, and psychological stress) across the lifespan. The authors called for future research to determine the links between weight stigma and psychological outcomes in youth specifically. Emmer et al. (2020) published a meta-analysis examining the association between weight stigma and mental health outcomes across the lifespan. Within this meta-analysis, higher reported weight stigma was associated with poorer mental health outcomes; an association which was stronger as body mass index increased. In examining differences between WBI and external weight stigma, the authors identified that WBI had a stronger association with mental health problems as compared to external weight stigma. This review, however, was limited as it included few studies with youth samples. Furthermore, a key limitation of this study was the aggregation of child and adolescent study results with adult studies. This methodology precludes the ability to quantify the association between weight stigma and mental health during formative developmental years.
Childhood is a key developmental period for understanding the impact of weight stigma as children and adolescents are highly susceptible to judgments from themselves and others as they grow and begin to develop identities (Neumark-Sztainer & Eisenberg, 2005). Furthermore, this is a developmental period associated with the onset of mental illness (Kessler et al., 2007). Thus, understanding the associations between weight stigma and mental health outcomes specific to youth is vital to better understand this relationship during a key developmental period.
Beyond the association between weight stigma and mental health in youth, examination of the impact of age, sex, and weight status on this association is needed. Prior work has identified evidence of stronger associations between weight stigma and body dissatisfaction in girls compared to boys (Puhl et al., 2020). However, other mental health outcomes such as depression, anxiety, and conduct disorder have not been examined. In a large meta-analysis across the lifespan, age was not a significant moderator of the association between weight stigma and mental health outcomes across the lifespan (Emmer et al., 2020). However, this study did not focus on youth and included minimal focus on developmental differences between childhood and adolescence. Given the significant developmental changes that occur throughout childhood and adolescence, there may be a differential association between weight stigma and mental health across childhood (Emmer et al., 2020). Finally, although weight stigma is sometimes thought of as only affecting persons with overweight or obesity, prior research has contradicted this assumption in adults (Jackson et al., 2014). Examination of weight status as a moderator of the association between weight stigma and mental health will elucidate any potential differences in the harmful effects of weight stigma across the weight spectrum in youth (Puhl & Heuer, 2009).
A systematic and meta-analytic exploration of how weight stigma is associated with multiple mental health outcomes in child and adolescent populations is key to better understanding and advancing both research and clinical agendas. Weight stigma is one of the last forms of stigma still considered by some to be socially acceptable (Puhl et al., 2020). The associations between weight stigma and mental health have not been well studied, particularly as relates to the moderating effects of age, sex, and weight status (Emmer et al., 2020; Puhl et al., 2020), given that prior work has not quantitatively examined these relationships in youth. The current systematic review and meta-analysis sought to identify and describe the relationships between weight-based stigma and mental health in youth. In addition, the current review aimed to examine the potential moderating effects of age, sex, and weight status on the association between weight stigma and mental health outcomes in youth. Study quality was also examined as a moderator of the effect between weight-based stigma and mental health in youth. Results highlight areas for future intervention development, future research directions, and clinical indicators of the association between weight stigma and mental health in youth.
Standard protocol and guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) were followed (Moher et al., 2015; Shamseer et al., 2015). The current review was not registered nor was an accessible protocol prepared. Systematic searches were conducted in three databases (PubMed, PsychINFO, and Embase). No limits were placed on the date of publication. Search results were imported into Endnote for screening. Search terms were selected based on previous reviews in adults and relevant primary studies (Supplementary Table 1; Emmer et al., 2020; Pont et al., 2017; Puhl & Heuer, 2009). Database searches were conducted in January 2020. Following initial screening, reference lists of included articles were searched to identify any additional articles for inclusion.
Studies were eligible for the current review based on the following criteria: (a) examined the association between weight stigma (i.e., external weight stigma, WBI) and a mental health outcome, (b) focused on children or adolescents (mean age was not more than 1 SD above 18 years old), (c) were written in English, and (d) were peer reviewed. Studies were excluded if they met criteria for one or more of the following: (a) did not explicitly measure weight stigma independent of other constructs (e.g., if weight and appearance stigma were combined, the study was excluded), (b) did not include a mental health variable (e.g., dieting not specific to mental health outcome such as disordered eating, quality of life measure without a subscale measuring emotional function), (c) primary focus of the study was a separate underlying medical condition (e.g., asthma, type 1 diabetes) or focused on children with an underlying medical condition that caused them to gain weight (e.g., Prader–Willi syndrome, thyroid disease), (d) did not provide enough information to calculate effect sizes, or (e) did not include original data (i.e., was a review article). All articles were screened to remove duplicate reporting on the same sample, which was assessed through multiple indicators (e.g., sample size, authors, methods section, study title). All authors participated in review of the articles in the current study, rotating articles to review such that two authors independently reviewed each article. Any discrepancies were discussed between authors, and when consensus could not be reached, an additional (i.e., third) author assessed the articles to obtain consensus. The following data were extracted from eligible articles for Table I: author name(s), year of publication, study design (i.e., cross-sectional or longitudinal), weight stigma type (i.e., external weight stigma or WBI) and measure, sample size, mean and standard deviation of sample age, percent sample identified as female, percent of sample who met criteria for overweight or obesity (body mass index ≥ 85 th percentile), mental health measure, key statistical findings, and brief description of the outcome. Further explanation of the review process is detailed in the Results section below.
Overview of Included Studies
| Authors (year) | Design | Weight stigma | N | Demographics (mean age [SD], % female) | % With OV/OB | Quality | Mental health outcome | Brief description of outcome | |
|---|---|---|---|---|---|---|---|---|---|
| Measure | Type | ||||||||
| Balantekin et al. (2018) | Cross-sectional | 1 question (Project EAT family) | External | 166 | 15 (NA) years, 100% female | NR | Poor | UWCBs (list of UWCBs) | Maternal and friend weight stigma positively associated with unhealthy weight control behaviors. Paternal weight stigma not statistically significant |
| Blanco et al. (2019) | Cross-sectional | POTS-S | External | 100 | 10.14 (1.35) years, 60% female | 50.0 | Fair | Anxiety (STAI-C), Depression (CDI), Self-esteem (LAWSEQ) | Weight stigma correlated with poorer psychological well-being, including greater depression and anxiety and lower self-esteem |
| Ciupitu-Plath et al. (2018) | Cross-sectional | WBIS | Internal | 191 | 15.06 (1.49) years, 51.31% female | 100 | Fair | Quality of life (KIDSCREEN), Self-esteem (RSE) | WBIS negatively correlated with quality of life and self-esteem |
| Eisenberg et al. (2019) a | Cross-sectional | 1 question (Project EAT family) | External | 1,577 | 14.5 (NR) years, 49.5% female | NR | Poor | Body dissatisfaction (Likert response to list of body parts), Depression (Depressive Mood Scale), Self-esteem (RSE), UWCB (endorsement of behaviors via list) | Those who reported experiencing weight stigma reported poorer outcomes across all mental health outcomes compared to those who did not report experiencing weight stigma |
| Gayes and Steele (2015) | Cross-sectional | POTS, WCA | External | 307 | 9.85 (0.72) years, 50.64% female | NR | Fair | Quality of life (PedsQL) | Both weight stigma scales were negatively correlated with HRQoL |
| Gerke et al. (2013) | Cross-sectional | POTS | External | 92 | 13.92 (1.84) years, 67.4% female | 100 | Fair | Disordered eating (ChEDE-Q), Depression (CDI), Self-esteem (CSEI) | Weight stigma positively correlated with depression and disordered eating, negatively correlated with self-esteem |
| Goldfield et al. (2010) | Cross-sectional | MRSF-III (parent and peer subscales) | External | 1,491 | 14.7 (1.8) years. 57% female | 24.0 | Poor | Anxiety (MASC-10), Depression (CDI), Disordered eating (DEBQ) | Weight stigma positively correlated with anxiety, depression, and disordered eating |
| Greenleaf et al. (2014) | Cross-sectional | 1 author-created question (yes/no) | External | 1,419 | 12.41 (0.97) years, 55% female | NR | Poor | Depression (CES-DC), Self-esteem (SDQII-GSE) | Those who reported experiencing weight stigma reported higher depression and lower self-esteem compared to those who did not report experiencing weight stigma |
| Greenleaf et al. (2017) | Cross-sectional | WST/PARTS | External | 343 | 12.3 (0.8) years, 41.7% female | 100 | Fair | Body satisfaction (BPSS-R), Depression (CES-DC), Self-esteem (SDQII-GSE) | Weight stigma positively correlated with depression and negatively correlated with body satisfaction and self-esteem |
| Guardabassi et al. (2018) | Cross-sectional | POTS | External | 600 | 9.64 (0.88) years, 51.8% female | 35.8 | Fair | Quality of life (PedsQL™4.0) | Weight stigma negatively correlated with PedsQL |
| Himmelstein and Puhl (2019) | Cross-sectional | 1 author-created question of frequency of weight teasing | External | 148 | 15.97 (1.25) years, 50% female | 82.0 | Poor | Author-created questions related to negative emotions and unhealthy eating behaviors | Weight stigma from peers positively correlated with negative emotions and eating, weight stigma from teachers negatively correlated with negative emotions and eating. Weight stigma from family and friends not statistically significant |
| Himmelstein et al. (2019) | Cross-sectional | 2 questions (Project EAT family and peers and frequency of teasing) | External | 9,838 | 15.6 (1.26) years, 44% cisgender female | 37.2 | Poor | Binge eating (4 yes/no questions), UWCB (9 questions), Stress (1 question) | Frequency of weight stigma correlated with more severe binge eating, UWCB, and stress |
| Hutchinson and Rapee (2007) | Cross-sectional | POTS | External | 1,094 | 12.3 (0.52) years, 100% female | NR | Fair | Body image concern (BAQ); Disordered eating (EDI-Bulimia subscale, DEBQ-Restraint subscale), UWCB (EWLB) | Weight stigma positively correlated with disordered eating behaviors and body image concern |
| Ievers-Landis et al. (2019) | Cross-sectional | POTS | External | 334 | 14.59 (1.38) years, 63.7% female | NR | Fair | Depression (CES-DC); Self-esteem (RSE) | Weight stigma negatively correlated with self-esteem and positively with depression |
| Jensen and Steele (2012) | Longitudinal | POTS | External | 93 | 11.7 (2.6) years, 59% female | 100 | Good | Quality of life (PedsQL™4.0) | Weight stigma negatively correlated with quality of life |
| Lin et al. (2020) | Longitudinal | WBIS | Internal | 934 | 15.7 (1.2) years, 52.5% female | 100 | Poor | Psychological dysfunction (DASS-21), Quality of life (PedsQL) | WBIS positively correlated with psychological dysfunction and negatively correlated with quality of life |
| López-Guimerà et al. (2012) | Cross-sectional | POTS | External | 1,501 | 14.3 (1.4) years, 48% female | 15.5 | Fair | Body dissatisfaction (EDI-2), Disordered eating (ChEAT), Self-esteem (RSE) | Weight stigma positively correlated with disordered eating and body dissatisfaction, negatively associated with self-esteem |
| Maïano et al. (2019) | Cross-sectional | WSSQ | Internal | 156 | 16.31 (0.8) years, 48.1% female | 100 | Poor | Anxiety (HADS), Depression (HADS), Disordered eating (EAT-26), Self-esteem (SDQ-12) | WSSQ negatively correlated to self-esteem and positively correlated to anxiety, depression, and disordered eating |
| McCormack et al. (2011) | Cross-sectional | 2 author-created questions | External | 148 | 11.2 (NR) years, 52.4% female, | 45.2 | Poor | Body dissatisfaction (5 yes/no questions) | Those who reported weight stigma had lower body satisfaction than those did not report weight stigma |
| Modi and Zeller (2008) | Cross-sectional | Sizing Them Up Teasing/Marginalization subscale (3 items) | External | 220 | 11.6 (3.3) years, 68% female | 100 | Fair | Quality of life (PedsQL-Kids) | Weight stigma negatively correlated to quality of life |
| Nelson et al. (2011) | Cross-sectional | POTS | External | 382 | 10.8 (0.6) years, 46% female | 32.8 | Fair | Body dissatisfaction (figure scales), Self-esteem (5 items modified from CSEI and RSE) | Weight stigma positively correlated to body dissatisfaction and negatively correlated to self-esteem |
| Neumark-Sztainer et al. (2002) b | Cross-sectional | 2 questions (Project EAT family and peers) | External | 4,746 | 14.9 (1.7) years, 49.8% female | NR | Poor | Disordered eating and unhealthy weight control behaviors (yes/no unstandardized questionnaire) | Weight stigma associated with greater disordered eating and unhealthy weight control behaviors |
| Neumark-Sztainer et al. (2010) | Cross-sectional | 1 author-created question | External | 356 | 15.8 (1.2) years, 100% female | 46.0 | Poor | Disordered eating (2 questions), UWCB (Project EAT questions) | Weight stigma associated with greater unhealthy and extreme weight control behaviors, and binge eating |
| Olvera et al. (2013) | Cross-sectional | MRFS-IV | External | 141 | 11.1 (1.5) years, 100% female | 100 | Poor | Disordered eating and UWCB (McKnight Risk Factor Survey-IV) | Parental weight stigma positively associated with binge eating |
| Pearlman et al. (2019) | Cross-sectional | 2 author-created questions based upon Puhl et al. (2013) | External | 128 | 14.35 (1.5) years, 54% female | 100 | Poor | Depression (BDI), Disordered eating (EDE-Q), Self-esteem (RSE) | Weight stigma positively correlated with disordered eating and depression, and negatively correlated with self-esteem |
| Porter et al. (2013) | Cross-sectional | POTS | External | 119 | 13.9 (1.7) years, 72% female | 100 | Fair | Body satisfaction (BDS, MBSRQ), Depression (CDI), Self-esteem (CSEI) | Weight stigma negatively correlated with self-esteem and body satisfaction and positively correlated with depression |
| Puhl and Himmelstein (2018) | Cross-sectional | Modified WBIS | Internal | 148 | 15.97 (1.2) years, 50% female | 71.6 | Poor | Disordered eating (4 validated questions regarding binge eating symptoms) | WBIS correlated with more severe binge eating |
| Puhl and Luedicke (2012) | Cross-sectional | 28 author-created items assessing forms of teasing and 8 items assessing location of teasing | External | 394 | 16.4 (1.0) years, 56% female | 35.0 | Poor | Affect and disordered eating responses to bullying (author-created questions) | Weight stigma correlated with negative affect for boys only |
| Puhl et al. (2019) c | Cross-sectional | 2 author-created questions (peers and family) | External | 9,839 | 15.6 (1.2) years, 44% cisgender female | 37.2 | Poor | Depression (KADS), Self-esteem (RSE) | Weight stigma negatively correlated with self-esteem and positively correlated with depression |
| Rojo-Moreno et al. (2013) | Cross-sectional | POTS | External | 57,997 | Age NR (13–16 years), 49.1% female | ∼14% | Fair | Body dissatisfaction (EDI), Disordered eating (ChEAT, EDI), Low self-esteem (EDI) | Weight stigma positively correlated to all mental health variables |
| Stice and Whitenton (2002) | Longitudinal | 2 author-created questions | External | 496 | Age NR (11–15 years), 100% female | NR | Good | Body dissatisfaction (BPSS-R), Depression (K-SADS) | Weight stigma positively correlated with body dissatisfaction and depression |
| Suisman et al. (2008) | Cross-sectional | POTS | External | 265 | 12.5 (1.4) years, 100% female | NR | Fair | Disordered eating (MEBS), Negative affect (EATQ-R) | Weight stigma positively correlated with disordered eating and negative affect |
| Thompson et al. (1995) | Cross-sectional and longitudinal | 2 author-created questions on frequency of teasing and how upset were by teasing; PARTS | External | 210 | Age NR (10–15 years), 100% female, | NR | Fair | Body image (Body Figure Ratings; EDI), Disordered Eating (EDI, EAT-26) | Weight stigma history positively correlated with disordered eating and negatively with body image |
| Thompson et al. (1995) | Cross-sectional | 1 author-created question on frequency of teasing; PARTS | External | 169 | 16 (NR) years, 100% female | NR | Fair | Body image (MBSEQ-AE, BIATQ-PSS), Global Psychological Function (CDI, STAI, COOP), Disordered Eating (BULIT-B, EDI, EAT-26) | Weight stigma positively correlated with disordered eating, worse global psychological functioning, and poorer body image |
| Turker et al. (2012) | Cross-sectional | 3 author-created items | External | 945 | 14.7 (1.2) years, 100% female | 21.5 | Fair | Body dissatisfaction (BSQ), Depression (BDI), Disordered eating (EAT-26), Self-esteem (RSE) | Weight stigma positively correlated with body dissatisfaction, depression, and disordered eating |
| Van Dale et al. (2014) | Cross-sectional | PARTS | External | 651 | Age NR (14–18 years), 54.8% female, | NR | Fair | Self-esteem (RSE) | Weight stigma positively correlated with low self-esteem |
| Van den Berg et al. (2002) | Cross-sectional | 5 author-created items regarding frequency | External | 470 | 15.5 (0.2) years, 100% female | NR | Poor | Body dissatisfaction (BAQ), Global psychological function (BDI, RSE) | Weight stigma positively associated with body dissatisfaction |
| Wertheim et al. (2001) | Longitudinal | POTS | External | 130 | 12.82 (NR), 100% female | NR | Fair | Depression (BDI), Disordered eating (DEBQ) | Weight stigma positively correlated with depression and disordered eating; weight stigma predicted disordered eating |
| Wertheim et al. (2001) | Longitudinal | POTS | External | 174 | 13.71 (NR), 100% female | NR | Fair | Body dissatisfaction (EDI), Disordered eating (EDI) | Weight stigma positively correlated with disordered eating and body dissatisfaction |
| Wertheim et al. (2001) | Longitudinal | POTS | External | 131 | 15.75 (NR), 100% female | NR | Fair | Body dissatisfaction (EDI), Depression (BDI), Disordered eating (EDI), Self-esteem (RSE) | Weight stigma positively correlated with depression, disordered eating, body dissatisfaction, and negatively correlated with self-esteem |
| Wojtowicz and Von Ranson (2012) | Longitudinal | POTS | External | 393 | 15.8 (0.6) years, 100% female | NR | Fair | Body dissatisfaction (BPS), Self-esteem (RSE) | Weight stigma positively correlated to body dissatisfaction and negatively to self-esteem |
| Zeller and Modi (2009) | Cross-sectional | Sizing Me Up Teasing/Marginalization Subscale | External | 141 | 9.2 (2.2) years, 67% female | 100 | Fair | Quality of Life (PedsQL™4.0) | Weight stigma negatively correlated with quality of life |
| Zuba and Warschburger (2018) | Cross-sectional | WBIS-C | Internal | 1,484 | 10.55 (0.9) years, 51.9% female | 19.0 | Fair | Anxiety (Kinder-Angst-Test II), Body dissatisfaction (1 item), Depression (German Depression Test for Children), Disordered eating (DEBQ), Psychosocial/emotional problems (Strengths and Difficulties Questionnaire), Self-esteem (CHQ) | WBIS positively correlated with depression, anxiety, body dissatisfaction, psychosocial/emotional problems, and disordered eating; WBIS negatively correlated with self-esteem |
Note. NA = Not applicable; NR = Not reported; BAQ = Body Attitudes Questionnaire; BESAA = Body Esteem Scale for Adolescents and Adults; BDI = Beck Depression Inventory; BIATQ = Body Image Automatic Thoughts Questionnaire; BPSS-R = Body Parts Satisfaction Scale Revised; BSQ = Body Shape Questionnaire; BUILT-B = Bulimia Test Binge Eating Subscale; CDI = Children’s Depression Inventory; CES-DC = Center for Epidemiological Studies-Depressive Scale for Children; ChEAT = Children’s Eating Attitude Test; ChEDE-Q = Children's Eating Disorder Examination Questionnaire; CHQ = Child Health Questionnaire; COOP = Coopersmith Self-Esteem Inventory; CSEI = Coopersmith Self-Esteem Inventory; DEBQ = Dutch Eating Behaviour Questionnaire; EATQ-R = The Early Adolescent Temperament Questionnaire—Revised; EDDS = Eating Disorder Diagnostic Scale; EDE-Q = Eating Disorder Examination Self-Report Questionnaire; EDI = Eating Disorder Inventory; EWLB = Extreme Weight Loss Behaviors Checklist; FRS = Figure Rating Scale; HADS = Hospital Anxiety and Depression Scale; HRQoL = Health-Related Quality of Life; KADS = Kutcher Adolescent Depression Scale; KIDSCREEN = KIDSCREEN 10-Index; K-SADS = Schedule for Affective Disorders and Schizophrenia for School-Age Children; LAWSEQ = Lawrence’s Self-Esteem Questionnaire; MASC-10 = Multidimensional Anxiety Scale for Children; MBSRQ = Multidimensional Body-Self Relations Questionnaire; MEBS = Minnesota Eating Behavior Survey; NPS = Negative Physical Self Scale; MRSF = McKnight Risk Factor Survey; PANAS = Positive and Negative Affect Scale; PARTS = Physical Appearance Related Teasing Scale, weight-related only; POTSWeight Teasing Subscale of the Perception of Teasing Scale; POTS-S = Spanish version of the POTS; RSE = Rosenberg Self-Esteem Scale; SDQII-GSE = Self-Description Questionnaire II General Self-Esteem Scale; STAI = State-Trait Anxiety Inventory; STPI = Spielberger State-Trait Personality Inventory; UWCB = Unhealthy weight control behaviors; WBIS = Weight Bias Internalization Scale; WCA = Weight Criticism Scale; WSSQ = Weight Self-Stigma Questionnaire.
a Both used Project Eat 2010 dataset. b Both used Project Eat 1 dataset. c Both used LGBTQ Teen Study dataset.Quality of studies included in the systematic review and meta-analysis was assessed using the Observational Cohort and Cross-Sectional Studies checklist (Study Quality Assessment Tools | NHLBI, NIH, n.d.). This 14-item checklist was designed to guide reviewers to focus on factors impacting the internal validity of a study (e.g., “Was the study population clearly specified and defined?”). All items in the assessment tool reveal an aspect of potential for study bias, with high rate of bias indicating lower-quality studies. A yes or no response is rendered for each of the 14 items to guide reviewers to an overall quality rating (good, fair, or poor), which is assigned to each article. For each study included in the current review, two authors independently assessed the quality of the study and resolved discrepancies through discussion with the input of a third author. Coders came to 100% agreement.
A random-effects model within Comprehensive Meta-Analysis (CMA), Version 3 was used, as the sampling and methodological factors were different between studies (Borenstein et al., 2005). An effect size of zero was assigned for any findings reported as nonsignificant without statistical values (Lipsey & Wilson, 2001). Effect sizes were all coded to indicate that higher values on mental health indicated poorer mental health outcomes (e.g., higher depressive symptoms, poorer self-esteem) to ensure that all scales functioned in the same direction. If effect sizes were not calculated, relevant data were inputted into CMA, which calculated the relevant effect size. Analyses were conducted such that one overall effect size was calculated for articles reporting multiple eligible analyses on the same sample (e.g., inclusion of multiple mental health outcomes). However, for articles that reported multiple independent samples (e.g., study A, study B) both samples were included and treated as independent samples. Correlational effect size, r, was used to reflect the association between the constructs of interest. Guidelines for interpretation of correlation as an effect size are as follows: 0.10 is a small effect, 0.30 is a medium effect, and 0.50 is a large effect. Meta-regression analyses were used to examine potential moderators, including average child age, percent female, percent with overweight/obesity, and study quality. The potential of publication bias was tested using Egger’s test (Egger et al., 1997) and fail-safe N (Rosenthal, 1979). Descriptive statistics (i.e., count variables) were conducted to describe the number of studies that measured specific mental health and weight stigma constructs, as well as those that included the moderation variables examined.
In total, 29,162 articles were identified across PubMed, PsychInfo, and Embase databases, with an additional (nonduplicate) five articles identified from a backward search of included articles. Duplicates were removed (n = 7,484). Titles and abstracts of articles were first screened by two reviewers, resulting in the exclusion of 21,449 articles. The full text of the remaining 234 articles was assessed and evaluated for eligibility. As indicated in Figure 1 , the most common reason for exclusion following full-text review was not measuring weight stigma, followed by the sample not being limited to children. The authors of this review reached out to the primary and/or senior authors of 17 articles requesting additional information for the meta-analysis. They received the requested information for six articles. Articles of authors that did not respond or were unable to provide the requested information were excluded.
PRISMA flow diagram of screened articles for inclusion in systematic review. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analysis.
Forty articles met the final eligibility criteria for the current review ( Table I ), which represented 43 effect sizes (due to independent samples within the same article). Of those, 35 of the studies involved cross-sectional research designs and 5 involved longitudinal designs. However, only cross-sectional data were included (e.g., baseline data from longitudinal studies), given limitations in combining cross-sectional and longitudinal data for the purposes of meta-analysis. No studies included treatments. Sample sizes ranged from 92 to 57,997 participants. Mean participant ages ranged from 9 to 16 years. Percent of participants who were female ranged from 42% to 100%. Fifteen studies did not report what percent of their participants met the criteria for overweight or obesity status. In the studies that reported percent of participants with overweight or obesity (n = 25), this ranged from 14% to 100%.
The majority of articles included in the current review measured external weight stigma (n = 35), while only five measured WBI. Of those that measured external weight stigma, the majority used either the Weight Teasing Subscale from the Perception of Teasing Scale (POTS; n = 14) or created their own questions (n = 10). Four studies used questions from Project EAT (Eating Among Teens; Balantekin et al., 2018; Eisenberg et al., 2019; Himmelstein et al., 2019; Neumark-Sztainer et al., 2002), two used questions from the McKnight Risk Factor Survey (Goldfield et al., 2010; Olvera et al., 2013), and three used the weight subscale of the Physical Appearance Related Teasing Scale (PARTS; Greenleaf et al., 2014; Thompson et al., 1995; van Dale et al., 2014). Articles asked about instances of weight stigma directed from the following categories: family only (n = 3); family and peers (n = 6); family and friends (n = 2); family, peers, and friends (n = 1); and family, peers, friends, and teachers (n = 1). Twenty-two articles did not specify the source of weight stigma. Four of the five articles measuring WBI utilized the Weight Bias Internalization Scale (WBIS; Ciupitu-Plath et al., 2018; Lin et al., 2020; Puhl & Himmelstein, 2018; Zuba & Warschburger, 2018) while one used the Weight Self-Stigma Questionnaire (WSSQ; Maïano et al., 2019).
Studies included in this review measured a wide range of mental health constructs, with the majority of studies examining multiple mental health outcomes. The specific outcomes measured included: self-esteem (n = 19), disordered eating (n = 17), body dissatisfaction/body image (n = 16), depression (n = 15), unhealthy weight control behaviors (n = 8), negative affect (including stress and global psychological functioning; n = 8), quality of life (n = 7), and anxiety (n = 4). Eight studies looked at the associations between weight stigma and one mental health construct, while the remainder (n = 28) looked at multiple mental health constructs. More of the articles included in the current review measured mental health outcomes with validated, standard measurements, while few used validated measures to examine the weight stigma constructs. Three studies in the current review evaluated mental health constructs using either author-created questions or it was unclear where the questions were derived from Himmelstein and Puhl (2019), Neumark-Sztainer et al. (2002), and Puhl and Luedicke (2012).
Using a random-effects model, the overall association between weight stigma and mental health was statistically significant (r = .32, 95% confidence interval [0.292, 0.347], p < .001, I 2 = 97.47) for the 40 studies representing 43 effect sizes ( Figure 2 ). This represents a medium effect size. There was no indication of publication bias based on Egger’s test (t = 1.07, p = .291). Fail-safe N indicated that 9,842 nonsignificant effect sizes would be required to invalidate the overall association, indicating a robust association between weight stigma and mental health in youth samples. All studies included in the present study indicated a positive or nonsignificant association between weight stigma and mental health indicating that experiences of weight stigma were associated with poorer mental health outcomes.
Forest plot of statistical effect sizes for studies included in the meta-analysis. CI = confidence interval.
Meta-regression analyses were conducted to examine age, sex, study quality, and overweight/obesity as moderators of the overall effect. There was a significant moderating effect of age (B = −.03, SE = 0.01, z = −4.92, p < .001, I 2 = 94.36), and there was also a significant moderating effect of study quality (B = .14, SE = 0.02, z = 6.60, p < .001, I 2 = 97.53). No other moderators were significant.
As reported above, age was a significant moderator of the overall effect size with 37 effect sizes included in this analysis. Younger age was associated with a more robust relationship between weight stigma and mental health. Most studies examined these variables among adolescents between the ages of 13 and 17 (n = 19), and six studies examined these variables in school-aged youth between the ages of 6 and 12. Fifteen studies evaluated youth across child and adolescent groups, though most of these (n = 12) extended the adolescent age group to 11–12 years. The lowest mean age in the current review was 9.2 (SD = 2.2) years (Zeller & Modi, 2009). No studies included in the current review looked at the effect of age on the association between weight stigma and mental health or compared between age groups.
Final quality codes are listed for each article included in Table I . Quality significantly moderated the overall effect size, such that higher-quality studies were associated with a stronger relationship between weight stigma and mental health outcomes with all 43 effect sizes included. However, only two studies were rated as good quality in the current review (Jensen & Steele, 2012; Stice & Whitenton, 2002). The rest were rated as poor (n = 17) or fair (n = 21). Study quality was often limited by lack of reliable or valid measures, cross-sectional design, and not controlling for potentially confounding variables such as demographics or weight status.
While two studies measured gender identity (Himmelstein et al., 2019; Puhl et al., 2019), we chose to use the word “sex” to describe the outcomes related to males and females as biological sex best represents what was reported in the majority of articles included in the current review. Forty studies were included in this analysis. Eleven studies sampled female participants only, with the other studies reporting between 41.7% and 72% female participants. No studies examined sex as a moderator of the association between weight stigma and mental health, although many studies examined differences in both weight stigma and mental health outcomes by sex.
Weight status was not a significant moderator within the present study, with 25 studies included. Ten of the studies included in the meta-analysis included only individuals with overweight or obesity. Fifteen of the remaining studies reported the percentage of participants with overweight/obesity, with a range from 14% to 82%. Eight studies examined the impact of weight status on weight stigma and mental health outcomes. While weight status was not found to be a significant moderator in the current review, these eight studies found that higher weight status and weight stigma interacted such that they were associated with worsened mental health outcomes (Goldfield et al., 2010; Guardabassi et al., 2018; Himmelstein & Puhl, 2019; Ievers-Landis et al., 2019; Lin et al., 2020; Nelson et al., 2011; Thompson et al., 1995; Zuba & Warschburger, 2018).
The purpose of the current systematic review and meta-analysis was to synthesize the published information regarding associations between weight stigma and mental health outcomes in children and adolescents. Forty articles (representing 43 effect sizes) met inclusion criteria for the present review, highlighting the relevance of examining the impact of weight stigma on mental health outcomes. The number of published studies on this topic has steadily increased over the past two decades, with 37.5% of the articles captured in this review published in the last 5 years. Results highlight the vast array of mental health outcomes considered, types of weight stigma assessed (e.g., WBI and external), and challenges related to measure validity and study quality.
Broadly, the compiled literature suggests that there is a clear association between weight stigma and mental health problems in youth as meta-analytic results indicated a robust medium effect size. These findings are consistent with Emmer and colleagues (2020) meta-analysis examining the associations of weight stigma and mental health across the lifespan, with the current review finding a similar sized association (Emmer et al., 2020). Mental health variables examined in the current review included depression, anxiety, disordered eating, unhealthy weight control behaviors, poor self-esteem, body dissatisfaction, stress, negative affect, and reduced quality of life. These mental health outcomes were associated with both external experiences of weight stigma and WBI.
Conceptualization of weight stigma varied amongst studies. Notably, several published articles discussed “weight stigma” but were excluded from the present review due to measurement of broader appearance-related stigma, rather than specific measurement of weight stigma. The term weight stigma was used to describe both external weight stigma as well as WBI, at times interchangeably, amongst studies included. Yet, it is more likely that these two types of weight stigma are discrete concepts. Only five studies included in the present review focused on WBI. None of the studies provided information on the relationships between WBI and external stigma. Future researchers would benefit from (a) specifying the type of weight stigma measured and (b) measuring both external weight stigma and WBI to further clarify the relationships between these two variables and subsequent mental health outcomes.
In adults, WBI is consistently associated with poor mental and physical health outcomes (Emmer et al., 2020; Pearl & Puhl, 2018). WBI was only included in five studies in the current review, and four of those studies were with adolescent samples. This result highlights the limited study of WBI in the child and adolescent literature and warrants future research to better understand this construct in youth. Pearl and Puhl (2018) conducted a thorough review of WBI across the lifespan. Consistent with findings from the present study that review highlighted the lack of developmental research on this topic in children and adolescents (Pearl and Puhl, 2018). Furthermore, the WBIS has only recently been validated for use in adolescents and has not yet been validated in children. Only one study thus far, and included in this review, used the WBIS in a child sample, and the researchers of that study changed the questions slightly to meet the developmental level of participants (Zuba & Warschburger, 2018). Future research is needed to determine whether certain social and cognitive capabilities, which develop at different times across the lifespan, are associated with experiences of weight stigma. Younger children develop stigma/outgroup evaluations based upon instruction from others, while older children tend to develop these evaluations based upon personal experience (Kang & Inzlicht, 2012).
Age and study quality were the only significant moderators in the present study. Specifically, age moderated the overall association between weight stigma and mental health such that there was a stronger association in younger samples. This result differs from the meta-analysis of Emmer and colleagues’ (2020), where they determined that age was not a significant moderator of the association between weight stigma and mental health, although that analysis was among individuals across the age spectrum with considerable focus on studies with adults. One possible explanation for this difference may be the cognitive and emotional development throughout childhood and adolescence as opposed to more stable processes in adulthood. It may be that the experience of weight-based stigma has a larger impact on mental health for young children. During formative years, younger children may have fewer coping strategies to handle weight-based stigma. This may be especially true for children without prior experience dealing with victimization, leading to a stronger impact on mental health. However, it may also be that the current tools for measurement of weight-based stigma are not accurately capturing the experiences of younger children, who may be less likely to recognize and report more subtle experiences than older adolescents. Future research is needed to better understand these developmental processes over time. Specifically, future research should explore whether the effect of age differs across external and internal weight stigma, especially given potential developmental changes in the experiences and reporting of external and internal weight stigma across childhood.
Study quality also significantly moderated the association between weight stigma and mental health in youth. Higher-quality studies were associated with stronger relationships between weight stigma and mental health. A large majority of studies were either poor quality (n = 17) or fair quality (n = 21) due mostly to cross-sectional designs, not including relevant control variables (i.e., weight status, gender, age), and not including validated measures. Although the reviewed studies were strengthened through the use of validated measures of mental health outcomes, measurement of weight stigma was limited by a strong reliance on retrospective self-report measures and limited use of validated measures. Twelve studies used investigator-created questions to inquire about external weight stigma with associated lack of reliability and challenges to construct validity. Five used questions from Project EAT (Eisenberg et al., 2006), which decreases concerns about replication; yet these measures have not been tested for validity or reliability. Only one high-quality longitudinal study was included in the present review (Stice & Whitenton, 2002), pointing to the need for increased longitudinal, laboratory-based, and prospective studies to better elucidate the associations between weight stigma and mental health in children and adolescents. Given the lack of longitudinal studies, this review is unable to address the question of causal relationship between weight stigma and mental health outcomes.
Findings from the present study highlight significant gaps in the literature and suggestions for future researchers. It is imperative that future researchers use validated measures of weight stigma, specifically those tested with child and/or adolescent populations. Future studies should also consider novel strategies that utilize technology to collect real-time data, such as ecological momentary assessment or electronically activated recorder data, to reduce the potential bias inherent with self-report data and better understand the fluid associations between weight stigma and mental health. The association between mental health and weight stigma may be bidirectional, and future research would benefit from determining how mental health variables and weight stigma interact over time. Results indicated that, among the studies examined in this review, age was the only statistically significant demographic moderator of the relationship between weight stigma and mental health. There is a paucity of literature examining the relationship between weight stigma and mental health by age, and there is a clear gap in the studies assessing the experience of weight stigma in school-age children. More research is needed to determine how these experiences develop and change over time from childhood to adolescence and into young adulthood. Neither sex nor weight status was statistically significant moderators, though this may be due to lack of heterogeneity in the included studies. A quarter of articles in this review sampled only females and/or only youth with overweight or obesity.
Increased exploration is needed of potential differences in the relationship between weight stigma and mental health by other environmental and demographic variables such as family income, racial and ethnic identity, gender identity, and geographic regions. This may provide greater context for risk factors exacerbating the association between weight stigma and mental health. While this review excluded studies specific to childhood medical conditions, no studies were actually excluded based on this criterion, indicating the lack of literature and need for work in this area. Few of the studies included in this review examined potential differences between race or ethnicity. Not only might there be differences regarding experienced weight stigma amongst youth with differing racial identities but it is also unknown in child and adolescent literature whether weight stigma and mental health are associated differently amongst youth of different racial and ethnic identities. It is possible that the relationships between weight stigma and mental health outcomes could be compounded by racial stigma, but further research is needed. Cumulative racial stigma and weight stigma have been associated with increased substance use and self-harm in adolescents (Bucchianeri et al., 2014).
Findings from the present review also indicate that most of the external weight stigma questions focused on stigma from a child’s family, peers, and/or friends. Only one study included stigma from teachers, and 22 studies did not specify the source of the experienced stigma. This is limiting and makes generalization difficult in that literature suggests other potentially influential causes of weight stigma experiences such as health-care providers, store-clerks, authority figures, and the media (Heuer et al., 2011; Puhl & Brownell, 2006; Tomiyama et al., 2018), and these experiences may have differing associations with mental health in youth. The Lancet recently published a call for media to refrain from publishing stigmatizing or discriminatory images and information related to persons with obesity (Flint et al., 2018). Future research on how the media may be disseminating weight stigma to youth, how youth may internalize weight bias from media outlets, and the potential mental health sequelae are crucial to study as it may empower social media creators and policymakers to create safeguards to protect youth from hurtful messages.
While this study improves the understanding of the relationships between weight stigma and mental health outcomes in youth, the results should be considered within the context of several limitations. Due to the large variety of mental health outcomes, the current review collapsed across all mental health outcomes in the analysis. The authors recognize that there are conceptual differences between each (i.e., depression, disordered eating, quality of life). However, there were far too many outcome types and measures to appropriately examine by type of mental health outcome. Similarly, the study combined biological sex and gender identity in the moderation analyses, in accordance with limitations in how data were collected and reported in the included studies. The authors strongly encourage future researchers to clearly report both biological sex and gender identity so that future reviews can elucidate potential differences based on gender and sex. The review also includes many studies rated poor and fair quality, which limits the impact of the results, though the moderation analyses were completed to account for study quality. Additionally, due to the cross-sectional nature of most of the research examined, findings from the present study are based upon associations and therefore causality cannot be inferred. Strengths of the current review include the methodological rigor, inclusion of meta-analysis with demographic and study quality moderators, and the process of reviewing potentially eligible manuscripts and analyzing data with multiple authors to reduce bias.
Any provider working in a health-related field should be thoughtful of how they discuss weight with child and adolescent patients, especially with consideration to the potential mental health implications. A cautious rule is to first ask the patient and/or caregiver whether they feel comfortable discussing weight and if there are specific words they prefer regarding weight (Dutton et al., 2010; Eneli et al., 2007). Given the finding that there may be a stronger association between weight stigma and mental health in younger youth, it is important to be thoughtful when discussing weight with young patients and consider developmentally appropriate language to ask about both weight and mental health. Furthermore, providers treating youth with mental health diagnoses should consider whether weight stigma—either external experiences or WBI—could play a role in their patients’ symptoms and treatments. Educators and other care providers for youth should be mindful of weight stigma communication amongst peers and within environments they manage to appropriately intervene upon or educate youth on the harms associated with weight stigma. Finally, this review adds to the mounting evidence discrediting the idea that weight stigma may encourage improved health. Although findings from the present study are cross-sectional and do not provide evidence that weight stigma causes mental health problems or vice versa, the strength and associations from this and other reviews provide evidence that there is an association between weight stigma and poor mental health (Emmer et al., 2020; Pearl & Puhl, 2018; Puhl et al., 2020).
In sum, the current review provides a synthesized report of the literature regarding relationships between weight stigma and mental health outcomes in youth. Results indicate that weight stigma, both WBI and external weight stigma, are moderately associated with negative mental health outcomes in youth. Age and study quality moderated these associations, such that younger age and higher study quality amplified the association between weight stigma and mental health, while neither biological sex nor weight status moderated the association. The results highlight suggested areas for improvement in future studies, such as the need for longitudinal and laboratory-based studies in this area, more attention to individual factors, and the use of objective and validated measurements. Results also suggest clinical implications for anyone who provides care to children and adolescents, specifically that younger children may be more at risk for poorer mental health associations with weight stigma. Results make clear that this is a critical topic of study in the field of child and adolescent psychology, and this review offers suggestions for researchers to build upon the existing literature.
This work was supported by the National Heart, Lung, and Blood Institute (T32HL076134) and the Agency for Healthcare Research and Quality (F32HS02707). The funders had no explicit role in the current systematic review and meta-analysis.
Conflicts of interest: Dr. E. Jelalian is a Consultant for Weight Watchers International. There are no other potential conflicts of interest to disclose.
