Association between weight bias internalization and metabolic syndrome among treatment-seeking individuals with obesity

Authors

  • Rebecca L. Pearl,

    Corresponding author
    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Thomas A. Wadden,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Christina M. Hopkins,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    2. Department of Psychology, Duke University, Durham, North Carolina, USA
    Search for more papers by this author
  • Jena A. Shaw,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Matthew R. Hayes,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    2. Department of Biobehavioral Health Sciences, School of Nursing, University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Zayna M. Bakizada,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Nasreen Alfaris,

    1. Department of Medicine Gastrointestinal Unit, MGH Weight Center, Massachusetts General Hospital, Boston, Massachusetts, USA
    Search for more papers by this author
  • Ariana M. Chao,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    2. Department of Family and Community Health, School of Nursing, University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Emilie Pinkasavage,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Robert I. Berkowitz,

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    2. Department of Psychiatry and Behavioral Sciences, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Naji Alamuddin

    1. Department of Psychiatry, Center for Weight and Eating Disorders, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    2. Department of Medicine, Institute for Diabetes, Obesity, and Metabolism, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA
    Search for more papers by this author

  • See Commentary, pg. 280.

  • Funding agencies: This study was supported by an investigator-initiated grant from Eisai Pharmaceutical Co. (TAW). AMC was supported by an NRSA postdoctoral fellowship from the NINR/NIH #T32NR007100.

  • Disclosure: TAW discloses serving on advisory boards for Novo Nordisk, Nutrisystem, and Weight Watchers, as well as receiving grant support, on behalf of the University of Pennsylvania, from Eisai Pharmaceutical Co. The other authors declared no conflict of interest.

  • Author contributions: RLP held primary responsibility for this specific study design, data analysis, and manuscript preparation. TAW was responsible for the design and oversight of the clinical trial and assisted with manuscript preparation. CMH, ZMB, AMC, and EP assisted with data collection and manuscript preparation. JAS and NAla assisted with data collection, data analysis, and manuscript preparation. MRH assisted with manuscript preparation. NAlf and RIB assisted with design and oversight of the clinical trial and manuscript preparation.

  • Clinical trial registration: ClinicalTrials.gov identifier NCT02388568.

Abstract

Objective

Weight stigma is a chronic stressor that may increase cardiometabolic risk. Some individuals with obesity self-stigmatize (i.e., weight bias internalization, WBI). No study to date has examined whether WBI is associated with metabolic syndrome.

Methods

Blood pressure, waist circumference, and fasting glucose, triglycerides, and high-density lipoprotein cholesterol were measured at baseline in 178 adults with obesity enrolled in a weight-loss trial. Medication use for hypertension, dyslipidemia, and prediabetes was included in criteria for metabolic syndrome. One hundred fifty-nine participants (88.1% female, 67.3% black, mean BMI = 41.1 kg/m2) completed the Weight Bias Internalization Scale and Patient Health Questionnaire (PHQ-9, to assess depressive symptoms). Odds ratios and partial correlations were calculated adjusting for demographics, BMI, and PHQ-9 scores.

Results

Fifty-one participants (32.1%) met criteria for metabolic syndrome. Odds of meeting criteria for metabolic syndrome were greater among participants with higher WBI, but not when controlling for all covariates (OR = 1.46, 95% CI = 1.00–2.13, P = 0.052). Higher WBI predicted greater odds of having high triglycerides (OR = 1.88, 95% CI = 1.14–3.09, P = 0.043). Analyzed categorically, high (vs. low) WBI predicted greater odds of metabolic syndrome and high triglycerides (Ps < 0.05).

Conclusions

Individuals with obesity who self-stigmatize may have heightened cardiometabolic risk. Biological and behavioral pathways linking WBI and metabolic syndrome require further exploration.

Introduction

Across the world, individuals with obesity are viewed as stereotypically lazy, lacking willpower, incompetent, unattractive, and to blame for their excess weight [1]. Due to these negative, prejudicial beliefs (known as weight bias), individuals with obesity face public derogation, devaluation, and discrimination (known as weight stigma) [2]. Weight stigma is a prominent psychosocial consequence of obesity that is associated with increased risk for depression, anxiety, body dissatisfaction, and low self-esteem [3]. In addition to experiencing weight stigma perpetrated by others, some individuals with obesity may internalize weight-biased attitudes by applying negative stereotypes to themselves and devaluing themselves due to their weight [4]. This self-directed stigma, also known as weight bias internalization (WBI), is associated with similar psychological outcomes as experiences of weight-based stigmatization enacted by others [3]. WBI is higher among individuals with more frequent experiences of weight stigma [5, 6], although experiencing and internalizing stigma are considered related yet independent constructs [6, 7]. Consistent with research that cognitive appraisal of a stressful life event is more predictive of negative psychological outcomes than the event itself [8], recent findings suggest that WBI (a cognitive process) may be a more robust predictor of psychological distress than the experience of weight stigma alone [9].

Additional research has begun to establish that experiencing (or perceiving) weight stigma, principally weight discrimination, is associated with physical health consequences. In a study of two large, representative samples that followed participants in the U.S. longitudinally for 4 to 10 years, Sutin et al. [10] reported that perceiving weight discrimination increased risk of mortality by 57% over and above the effects of mental and physical health risk factors. Other studies have also demonstrated that perceived weight discrimination is associated with increases in weight, waist circumference, and, for those in the overweight range, the development of obesity over time [11, 12]. Among individuals with and without type 2 diabetes, weight-based discrimination is also associated with poorer glycemic control, even when controlling for body mass index (BMI), depressive symptoms, and other relevant health variables [13, 14].

A combination of biological and behavioral pathways may explain the associations between weight stigma and poor physical health [15]. Stigma broadly is considered a chronic stressor [16], which can elicit a biochemical stress response marked by changes in autonomic response, hypothalamic-pituitary-adrenocoritcal axis activation, oxidative stress, and inflammation [15]. For example, experimental studies that exposed participants (predominantly white women) to weight stigma have demonstrated heightened blood pressure and increased cortisol reactivity [17, 18]. Increased appetite and/or physiological arousal resulting from these biochemical changes, combined with emotional responses, may contribute to observed increases in calorie consumption and binge eating in individuals who have been directly exposed to or experienced weight stigma [19, 20]. Individuals who report weight-stigmatizing experiences also tend to avoid physical activity [21]. In addition, individuals who perceive or experience weight stigma in health care settings are more likely to avoid or delay preventive care, thus increasing risk for disease progression [22].

These studies examining the relationship between weight stigma and physical health have not included assessments of WBI, and generally less is known about the relationship between WBI and cardiometabolic markers of health. In several studies, correlations have been found between WBI and poorer self-reported health, increased binge eating, and reduced physical activity [6, 20, 23-27], suggesting that similar pathways to poor health may exist for WBI as with weight stigma perpetrated by others. Several studies have demonstrated that WBI is a partial mediator between experiences of weight stigma and poor health behaviors (such as uncontrolled eating and reduced exercise) [5, 6]. Additionally, one study found that BMI was only associated with poor health-related quality of life among individuals with high versus low WBI [28]. Thus, it is possible that activation of internalized, negative beliefs following or independent of stigmatizing experiences may lead to heightened stress and subsequent maladaptive coping behaviors (e.g., binge eating).

Although limited research has investigated cardiometabolic risk associated with other forms of internalized stigma (e.g., racism) [29], some studies have reported associations between internalized racism and greater risk of abdominal obesity, insulin resistance, and other indices of obesity-related health [30]. However, no study to date has investigated the relationship between WBI and cardiometabolic risk factors commonly associated with obesity. Metabolic syndrome, which refers to a cluster of risk factors for cardiovascular disease and type 2 diabetes [31], is particularly relevant to investigate. Approximately 35% to 40% of adults in the U.S. have metabolic syndrome, and 60% of adults with obesity meet criteria for this diagnosis [32]. Individuals with metabolic syndrome have a twofold increased risk for cardiovascular disease and a fivefold risk for type 2 diabetes [31, 33].

This study examined the relationship between WBI and metabolic syndrome. We hypothesized that in persons with obesity, higher levels of WBI would be associated with increased odds of having metabolic syndrome. We also explored the relationships between WBI and the individual cardiometabolic risk factors that constitute metabolic syndrome, in order to determine which risk factors may be more strongly linked to WBI.

Methods

Participants

Participants were 178 adults with obesity recruited from the community to participate in a weight-loss trial. The primary purpose of the study was to test the effects of a 1-year weight-loss maintenance program, to which participants were randomly assigned if they lost ≥5% of initial weight during a 14-week diet run-in period. Eligible participants had a BMI ≥33 kg/m2 (or ≥30 kg/m2 with comorbidities) and ≤55 kg/m2 and were: age 21 to 65 years; weight stable with no history of bariatric surgery; not on medications that would affect weight or on medications contraindicated with the weight-loss medication used in the 1-year maintenance study; and under the care of a primary care physician. They also were free of: current, severe depressive episodes; suicidal ideation; type 1 or type 2 diabetes; uncontrolled hypertension; cardiovascular, valvular heart, hepatic, renal, or uncontrolled thyroid disease; or other medical conditions that could compromise the participant's ability to complete the weight-loss program. Only baseline data were analyzed in the current study. The study was approved by the University of Pennsylvania Institutional Review Board.

Procedures

Participants were recruited from the community through print, radio, and Web advertisements and screened over the phone for eligibility. Following the initial phone screen, interested participants were evaluated in person by trained clinicians for eligibility and appropriateness. All participants underwent a medical exam with a physician or nurse practitioner before enrollment, which included assessment of current medications. At this initial screening visit, blood pressure and waist circumference were measured, and blood draws completed to obtain measures of fasting blood glucose, triglycerides, and high-density lipoprotein (HDL) cholesterol. Height and weight also were measured at this screening visit, and weight was measured again during the first week of the study (prior to any weight-loss intervention). Questionnaires were administered online (via REDcap) or in hard copy form via mail up to 2 weeks prior to the start of the program. Data for the present study are based on these screening measures, obtained before the 14-week run-in diet.

Measures

Anthropometric measures

Weight was measured (with participants dressed in light clothing and without shoes) on a digital scale (Detecto, model 6800A). Height was measured using a wall-mounted stadiometer (Veeder-Root, Elizabethtown, NC).

Cardiometabolic risk factors

Blood samples were collected following an 8-hour fast and analyzed by Quest Diagnostics (Horsham, PA) for levels of triglycerides, HDL cholesterol, and glucose (and other tests to assess patient safety). Blood pressure and pulse were measured in duplicate, at 1-minute intervals (after a 5-minute seated rest), using an automated monitor (Dinamap, model 9300, Johnson & Johnson), following methods described previously [34]. Waist circumference was measured with a flexible tension-controlled measuring tape midway between the iliac crest and lowest rib to the nearest 0.1 cm [34]. Metabolic syndrome was defined by the presence of three or more of the following five risk factors: central or abdominal obesity (i.e., waist circumference ≥35 inches for women, ≥40 inches for men); elevated triglycerides (i.e., ≥150 mg/dL) and/or use of medication for dyslipidemia (i.e., statins); low HDL cholesterol (i.e., ≤50 mg/dL for women, ≤40 mg/dL for men); elevated blood pressure (i.e., ≥130/85 mm Hg) and/or antihypertensive medication; and elevated fasting blood glucose (i.e., ≥100 mg/dL) and/or medication to treat prediabetic levels of blood glucose [31].

Weight Bias Internalization (WBI) Scale

The 11-item WBIS [4] is a widely used, validated measure of WBI (i.e., self-directed weight stigma), including in treatment-seeking samples [23-25]. The scale assesses the extent to which people apply weight-based stereotypes to themselves (e.g., “Because of my weight, I don't understand how anyone attractive would want to date me.”) and evaluate themselves negatively due to weight (e.g., “I hate myself for being overweight.”). Responses are rated on a scale of 1 (strongly disagree) to 7 (strongly agree), with higher scores signifying greater WBI. The scale demonstrated good internal validity in the current sample (α = 0.84).

Depression

The Patient Health Questionnaire (PHQ-9) [35] is a nine-item, self-report measure of depression. Items assess the frequency of symptoms of depression (e.g., “little interest or pleasure in doing things”) over the past 2 weeks, and responses range from 0 (not at all) to 3 (nearly every day). Scores are summed, with higher scores indicating greater severity of depressive symptoms. The PHQ-9 is currently recommended by the American Psychiatric Association as the optimal depression severity measure for DSM-5 [36]. The PHQ-9 had good internal validity in the current study (α = 0.85).

Statistical analyses

Natural log transformations were performed to adjust for skewed values of triglycerides, waist circumference, HDL cholesterol, and PHQ-9 scores. All tests were two-tailed and used a P < 0.05 level of significance. Analyses of variance and correlations were tested to determine associations between WBI and participant characteristics, BMI, and depressive symptoms.

We constructed a binary logistic regression model to test our primary hypothesis that participants with higher WBI would exhibit increased odds of having metabolic syndrome. To receive this diagnosis, participants had to meet criteria for three of the five cardiometabolic risk factors described above. We controlled for age, sex, race, ethnicity, and BMI in all analyses. Given some findings showing a relationship between metabolic syndrome and depression [37], we also controlled for PHQ-9 scores in all analyses. Thus, we tested the relationship between WBI and metabolic syndrome above and beyond the effects of demographics, BMI, and depression.

Following the primary analysis, separate logistic regression models were constructed for each risk factor (defined categorically, as described previously) to test their associations with WBI. We also calculated partial correlations (controlling for demographics, BMI, and PHQ-9 scores) between WBIS scores and continuous variables of each cardiometabolic risk factor. Additionally, we included supplementary analyses in which we constructed the same six logistic regression models, but with WBI defined categorically as “high” versus “low” based on tertiles. This dichotomization of WBI is consistent with prior research [28, 30].

Missing data

At baseline, eight participants were missing waist circumference measurements. These participants were excluded from analyses that included waist circumference or metabolic syndrome diagnosis as outcome variables.

Results

Participant characteristics

Completed study questionnaires were obtained from 159 participants, whose characteristics are summarized in Table 1. The majority of participants were women and black. Table 1 also presents a summary of the number of participants meeting criteria for metabolic syndrome, as well as participants’ mean weight, height, BMI, and scores on the WBIS and PHQ-9.

Table 1. Sample characteristics (N = 159)
VariableN (%) or M ± SD
  1. a

    Waist circumference data were missing for eight participants. Percentages for the waist circumference criterion and metabolic syndrome diagnosis were calculated from the total number of participants (N = 159).

  2. SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; HDL, high-density lipoprotein cholesterol.

Sex 
Women140 (88.1)
Men19 (11.9)
Race and ethnicity 
Black or African American, non-Hispanic107 (67.3)
White, non-Hispanic36 (22.6)
Asian4 (2.5)
American Indian or Alaska Native1 (.6)
Other1 (.6)
More than one race4 (2.5)
Hispanic or Latino/a9 (5.7)
Metabolic syndrome 
SBP ≥130, DBP ≥85 mm Hg, or medication treating hypertension101 (63.5)
Triglycerides ≥150 mg/dL or medication treating dyslipidemia36 (22.6)
Waist circumference ≥40 in for men, ≥35 in for womena150 (94.3)
FBG ≥100 mg/dL or medication treating prediabetes8 (5.0)
HDL cholesterol <40 mg/dL for men, <50 mg/dL for women46 (28.9)
Diagnosis (three of five criteria)a51 (32.1)
Age (y)43.8 ± 11.3
Weight (kg)115.2 ± 20.7
Height (cm)167.2 ± 8.5
Body mass index (kg/m2)41.1 ± 5.9
SBP (mm Hg)129.8 ± 16.1
DBP (mm Hg)75.5 ± 10.0
Triglycerides (mg/dL)100.7 ± 52.2
Waist circumference (cm)a114.4 ± 14.2
FBG (mg/dL)85.8 ± 8.1
HDL (mg/dL)57.5 ± 18.7
Weight Bias Internalization Scale3.6 ± 1.1
Patient Health Questionnaire-94.8 ± 4.8

Race/ethnicity was coded into three categories of non-Hispanic white, non-Hispanic black, and all other racial/ethnic groups. Analysis of variance revealed significant differences in WBIS scores by race (P < 0.001). Post hoc paired t-tests revealed that black participants scored significantly lower on the WBIS than white participants: mean (±SD) scores for black participants = 3.36 ± 1.08, and for white participants = 4.22 ± 1.15 (P < 0.001). WBIS scores correlated significantly with depressive symptoms (r = 0.38, P < 0.001). Age, sex, and BMI were not significantly associated with WBIS scores.

Primary analysis: Metabolic syndrome

Table 2 presents the results from the logistic regression analysis that examined the association between WBI and metabolic syndrome, controlling for demographics, BMI, and depressive symptoms. When controlling for BMI and depressive symptoms, participants with higher WBIS scores had greater odds of meeting criteria for metabolic syndrome (OR = 1.41, P = 0.042). However, when demographics were included in the model, this relationship was no longer statistically significant (OR = 1.46, P = 0.052). Participant age was the only covariate associated with metabolic syndrome in the model (OR = 1.05, P = 0.004).

Table 2. Odds ratios (95% confidence intervals) for weight bias internalization as a continuous measure
 Metabolic syndromeHigh triglycerides and/or medication
  1. Block 1: Weight bias internalization, body mass index, and depressive symptoms. Block 2: Block 1 plus demographics (age, sex, race/ethnicity). Due to missing waist circumference data, N = 151 for metabolic syndrome. N = 159 for triglycerides/medication.

  2. +P = 0.052; *P < 0.05; **P < 0.01.

Block 11.41 (1.01–1.97)*1.79 (1.21–2.64)**
Block 21.46 (1.00–2.13)+1.88 (1.14–3.09)*

Secondary analyses: Cardiometabolic risk factors

To explore the factors responsible for heightened odds of metabolic syndrome, we examined the associations between WBI and each component of the condition via five separate logistic regression models (all controlling for demographics, BMI, and PHQ-9 scores) and partial correlations (controlling also for respective medication use). Of the logistic regression models, only the model that tested the relationship between WBI and high triglycerides was significant (see Table 2). Participants with higher WBIS scores had greater odds of having high triglycerides and/or taking medication for dyslipidemia, even when controlling for all covariates (OR = 1.88, P = 0.013). Only one participant did not meet the waist circumference criterion, and no associations between this risk factor and WBIS scores emerged. Associations were not significant between WBIS scores and blood pressure or glucose. Only eight participants met the glucose/medication criterion (due to the participant exclusion criterion of type 2 diabetes). WBIS scores correlated positively with triglycerides (P = 0.021) and negatively with HDL cholesterol (P = 0.009). However, these correlations were not statistically significant when controlling for all covariates (see Table 3).

Table 3. Correlations (and partial correlations) between weight bias internalization and cardiometabolic risk factors
VariableHDL cholesterolaFBGWaist circumferenceTriglyceridesDPBSBP
  1. Partial correlations control for age, sex, race/ethnicity, body mass index, depression, and respective medication use (for blood pressure, triglycerides, and glucose).

  2. a

    Variable was transformed using the natural logarithm.

  3. b

    N = 151 for waist circumference due to missing data.

  4. *P < 0.05; **P < 0.01; ***P < 0.001.

Weight bias internalization−0.21** (−0.13)0.04 (0.02)0.10 (0.02)0.18* (0.13)0.08 (0.07)0.06 (0.09)
Systolic blood pressure (SBP)0.08 (0.06)0.18* (0.11)0.12 (0.06)0.07 (0.04)0.58*** (0.59***) 
Diastolic blood pressure (DPB)−0.03 (−0.02)0.10 (0.06)0.10 (0.05)0.09 (0.06)  
Triglyceridesa−0.41*** (−0.40***)0.15 (0.08)−0.03 (−0.01)   
Waist circumferencea, b−0.07 (0.06)0.14 (0.05)    
Fasting blood glucose (FBG)−0.13 (−0.13)     

Supplementary analyses

Consistent with prior research in which internalized stigma was dichotomized to determine effects of “high” versus “low” values [28, 30], we examined the relationship between WBI, metabolic syndrome, and the individual risk factors in six additional logistic regression models (controlling for all aforementioned covariates), in which WBI was categorized as high versus low based on tertiles. Cutoff scores of 4.18 (n = 57) and 3.09 (n = 52) were used to define high and low scores, respectively. Logistic regression analyses controlling for all covariates demonstrated that participants high in WBI had three times greater odds of meeting criteria for metabolic syndrome (OR = 3.19, P = 0.039), and six times greater odds of having high triglycerides and/or taking medication (OR = 6.13, P = 0.018) than participants low in WBI (see Table 4).

Table 4. Odds ratios (95% confidence intervals) for high versus low weight bias internalization
 Metabolic syndromeHigh triglycerides and/or medication
  1. Block 1: Weight bias internalization, body mass index, and depressive symptoms. Block 2: Block 1 plus demographics (age, sex, race/ethnicity). Categorization of high versus low weight bias internalization was based on tertiles: high ≥4.18, low ≤3.09. Due to missing waist circumference data, N = 103 for metabolic syndrome. N = 109 for triglycerides/medication.

  2. *P < 0.05 **P < 0.01.

Block 12.68 (1.05–6.88)*5.64 (1.72–18.54)**
Block 23.19 (1.06–9.56)*6.13 (1.37–27.46)*

Discussion

This study of treatment-seeking individuals with obesity found mixed results concerning the relationship between metabolic syndrome and WBI. This relationship was not significant when all covariates were included in a regression model in which WBI was represented continuously. However, the odds of having metabolic syndrome were significantly heightened among participants categorized as having high (versus low) levels of WBI. WBI has been shown previously to be associated with binge eating [20], reduced physical activity motivation and engagement [6, 23], and poorer self-reported physical health [24, 25]. However, this is the first report of which we are aware to demonstrate a possible association between WBI and metabolic syndrome.

The only individual component of metabolic syndrome that was significantly associated with WBI was high triglycerides/use of medication for dyslipidemia. Internalized weight bias could elicit a chronic stress response similar to that observed in reaction to experiences of weight stigma, such as heightened levels of oxidative stress and cortisol secretion [15]. The act of self-stigmatizing may lead to a state of physiological arousal that itself increases risk for metabolic abnormalities through biological pathways (e.g., cortisol secretion). This state of physiological and affective stress may also lead individuals to cope by eating unhealthy food or binge eating [38]. Additionally, individuals with high WBI exhibit diminished self-efficacy to exercise due to endorsing negative stereotypes, such as laziness [6, 23], and thus are more likely to avoid physical activity [21]. Further research is needed to determine the specific biological and/or behavioral pathways that could account for unfavorable lipid levels. For example, advanced lipoprotein testing could help to determine whether the observed effects are diet-induced versus metabolically endogenous. In-depth dietary analysis could also examine whether individuals with high WBI consume more high-fat foods, which could potentially account for the elevated triglyceride levels observed in this study. The correlation between WBI and triglycerides was not significant when controlling for covariates, and waist circumference and blood glucose control were largely homogeneous in this sample. Thus, replication is required, and further research with a more heterogeneous representation of these risk factors (e.g., including individuals with type 2 diabetes) is needed.

This study was cross-sectional in design, precluding conclusions about causality. Thus, a converse relationship could be present, in which poorer physical health (e.g., metabolic syndrome) may lead individuals with obesity to be more prone to internalizing weight bias. Individuals with obesity frequently experience weight stigma in health care settings [22]; increased need for health care services due to poor health may consequently increase exposure to stigma and heighten vulnerability to internalizing it. Heightened disease burden may also increase susceptibility to self-blame due to the perceived controllability of weight [2]. A longitudinal study of individuals with varying levels of WBI at baseline, but without obesity-related comorbidities, could help to clarify the temporal relationship between WBI and risk of metabolic syndrome.

This sample was unique in its predominant composition of black women, which allowed us to examine WBI among a group of individuals with obesity that is often not well represented in research addressing this topic. Black participants had lower levels of WBI on average, which may reflect reduced vulnerability to body dissatisfaction among black women [39]. Black children and adults also generally have lower triglyceride levels and are thus less likely to be diagnosed with metabolic syndrome [40]. Given these racial differences, it is possible that the relationship between WBI, triglycerides, and metabolic syndrome may differ by race as well. We may have been underpowered in our analyses to detect significant racial differences in the regression models, as well as sex differences due to the relatively small number of men in the present sample. Prior studies examining the relationship between weight discrimination and physical health outcomes have used large, nationally representative datasets [10-12], in which small effects may be better detected. We recommend that future large-scale investigations of obesity and health—particularly cardiovascular disease risk—incorporate the WBIS into assessment batteries in order to accumulate more data and test for potential moderators such as sex and race.

Our findings pertain only to individuals with obesity seeking treatment for weight loss and may not generalize to the broader population of individuals with obesity. Prior evidence suggests that treatment-seeking individuals with obesity exhibit heightened rates of psychopathology (including depression) in comparison with nontreatment seeking individuals [41]. However, given this study's exclusion criteria of severe depression or use of antidepressant medication, our clinical sample had relatively low levels of depressive symptoms. The mean WBIS score in this sample was lower than in other recent studies of treatment-seeking samples [42], which may reflect the relatively good mental and physical health of this sample. Research assessing associations between WBI and health among individuals with higher levels of WBI, and potentially depressive symptoms to test for mediation, would be informative to more fully understand the extent to which self-stigma may affect cardiometabolic risk.

Conclusion

The present findings contribute to the growing body of literature demonstrating the relationship between weight stigma and adverse physical health outcomes, once again contradicting a persistent argument that stigma motivates behavior change and improves health. Replication is required to determine the relationship between WBI and metabolic syndrome above and beyond relevant demographics and health-related covariates. Nevertheless, these findings highlight the importance of informing the public that weight stigma—including self-directed stigma—is stressful and may contribute to poor health. Further research is needed to develop and test the effects of interventions that aim to reduce WBI among individuals with obesity. To inform these interventions, future research efforts should focus on identifying specific biological and behavioral pathways—such as lipoprotein profiles, dietary intake, and engagement in physical activity—between WBI and cardiometabolic risk.