Sources of Weight Discrimination and Health

Abstract

Individuals who perceive weight discrimination tend to be less healthy. The correlates of weight discrimination, however, may vary by source of the discrimination (e.g., from a healthcare provider, stranger, family member, etc.). The present research used a large national sample (N=4,990) to examine the prevalence and demographic predictors of sources of weight discrimination and whether subjective health and depressive symptoms associated with discrimination varied by source. Approximately 13% of the respondents reported experiencing discrimination based on their body weight. Of these participants, the prevalence across sources ranged from 4% from educators to 52% from strangers. Women and younger participants were more likely to experience bias from their families; participants with higher BMIs were more likely to experience discrimination from a service industry professional or a stranger. Participants had lower subjective health when the source of weight discrimination came from either a stranger or a family member, whereas depressive symptoms were higher across all sources of discrimination except for from a service professional or a stranger. These results support previous research that indicates weight discrimination is common and comes from many sources and further suggests that the health correlates may differ depending on who is the source of the unfair treatment.

Unfair treatment based on body weight is a common form of discrimination (Puhl, Andreyeva, & Brownell, 2008). The health correlates of such discrimination are well documented. Those who experience discrimination based on body weight have more depressive symptoms (Schmitt, Branscombe, Postmes, & Garcia, 2014), are at greater risk for weight gain and associated comorbidities over time (Sutin, Stephan, Carretta, & Terracciano, 2015), and are ultimately at risk of premature mortality (Sutin, Stephan, & Terracciano, in press).

Weight bias comes from many sources. Healthcare providers, educators, friends, and family may hold negative attitudes about obesity that can lead to discrimination (Puhl & Heuer, 2009). Individuals are most likely to experience unfair treatment on the basis of body weight from strangers and family members and least likely to experience it from their colleagues or bosses at work (Vartanian, Pinkus, & Smyth, 2014). Less is known, however, about whether sources of discrimination are differentially related to health-related outcomes; weight discrimination from specific sources may be more harmful than from other sources. For example, applying the label “fat” to an adolescent is associated with greater weight gain, an association that is stronger when the label is applied by the family than by others (Hunger & Tomiyama, 2014). Among adults, weight discrimination is associated with greater weight gain (Sutin & Terracciano, 2013) and declines in health (Sutin et al., 2015), but the sources of discrimination are generally not differentiated. To that end, the present research uses a large national sample to address the relation between sources of discrimination and health. Specifically, we aim to identify (1) the prevalence of common sources of weight discrimination, (2) demographic predictors of who experiences discrimination from specific sources, and (3) whether subjective health and depressive symptoms differ by source of discrimination.

Method

Participants and Procedure

Participants were recruited to participate in an online study of the psychological correlates of health and well-being. Survey Sampling International (SSI) recruited participants through their proprietary panel and directed them to a Qualtrics survey administered by the Florida State University (FSU) College of Medicine that was completed online. Participants had to be 18 years or older and living in the United States. The sample was stratified by age with a roughly equal percentage of participants (20%) recruited across five age bands: 18–29, 30–39, 40–49, 50–59, and 60+. African Americans were oversampled to ensure adequate sample size for the analysis. The FSU Institutional Review Board approved this study and participants provided informed consent and were compensated up to $5 for their participation (all participants received monetary compensation, but exact payment depended on the participant’s arrangement with SSI).

A total of 6,303 individuals clicked on the link provided by SSI, 6,040 consented to participate, and 4,990 completed the discrimination and relevant health measures (79% participation rate). The sample was 50% female, had a mean age of 44.69 (SD=15.32), and was 55% non-Hispanic White, 19% African American, 15% Hispanic, and 11% other or unknown.

Measures

Demographics

Participants reported their age, race/ethnicity, sex, education, and height and weight.

Perceived discrimination

Participants completed the 6-item version of the perceived experiences with everyday discrimination scale (Williams, Yu, Jackson, & Anderson, 1997). Participants were asked, “In your day-to-day life, how often have any of the following things happened to you?” and given items such as “You are treated with less courtesy or respect than other people.” Participants rated these items on a scale from 1 (never) to 6 (almost everyday). The alpha reliability was .90. Participants then made a single rating (yes/no) as to whether those experiences could be attributed to weight.

Sources of discrimination

After reporting their experiences and attributions, participants specified who was the source of the unfair treatment based on weight. Participants were asked, “If any of the above has happened to you, who was the source of those experiences (check all that apply)?” Options were a health care professional (e.g., doctor), a boss or co-worker, a service industry professional (e.g., store clerk), a public safety officer (e.g., police), an educator (e.g., teacher), a stranger, an acquaintance, a close friend, and/or a family member. Participants could choose (yes/no) as many or few as applicable.

Subjective health

Participants reported on their subjective health on a standard single item measure (Idler & Benyamini, 1997), “In general, would you say your health is…? (check one).” Response options ranged from 1 (poor) to 5 (excellent).

Depressive symptoms

Participants reported their symptoms of depression with the Patient-Health Questionnaire (PHQ; Spitzer, Kroenke, & Williams, 1999). Participants rated how often they experienced a number of symptoms (e.g., “little interest or pleasure in doing things”) they have been bothered by over the last two weeks. Response options ranged from 1 (not at all) to 4 (nearly everyday). The alpha reliability was .92.

Analytic Strategy

We used logistic regression to identify predictors of weight discrimination and each source of discrimination. Specifically, we tested age (scaled in decades), sex, ethnicity, education, and body mass index (BMI; derived as kg/m² and z-scored) as predictors. To examine whether there were mean differences in subjective health and depressive symptoms, we used Analysis of Covariance (ANCOVA) with subjective health/depressive symptoms as the dependent variable and each source of discrimination as the independent variable, controlling for the demographic factors and BMI. Cohen’s d was calculated as a measure of effect size.

Results

Across the entire sample, 13% of participants (n=642) reported discrimination based on their weight. Older (OR=.74; 95% CI=.69–.79) and African American (OR=.63; 95% CI=.49–.81) participants were less likely to report weight discrimination, whereas higher BMI was associated with having experienced it (OR=2.03; 95% CI=1.88–2.20). Men and women were equally likely to have experienced unfair treatment based on their weight (OR=1.15, 95% CI=.96–1.38); education was also unrelated to it (OR=.99, 95% CI=.92–1.06).

All sources of discrimination were endorsed; percentages ranged from 4% from educators to 52% from strangers (Table 1). The logistic regression analysis indicated that African American and Hispanic participants were more likely to experience weight discrimination from a public safety officer (OR=2.50; 95% CI=1.10–5.67 and OR=2.46; 95% CI=1.11–5.46, respectively). Participants with more years of education were more likely to report discrimination from a service industry professional (OR=1.15, 95% CI=1.01–1.32). Younger participants were more likely to report discrimination from an educator (OR=.58, 95% CI=.39–.85) or family member (OR=.71, 95% CI=.60–.84), whereas older participants were more likely to report discrimination from a service industry professional (OR=1.17, 95% CI-1.01–1.36). Higher BMI was associated with a greater likelihood of experiencing discrimination from a service industry professional (OR=1.17; 95% CI=1.01–1.36) or stranger (OR=1.25, 95% CI=1.09–1.42) and not from a public safety officer (OR=.75; 95% CI=.57–.98) or a friend (OR=.80; 95% CI=.65–.99). Finally, women were more likely than men to be treated unfairly because of weight within their family (OR=1.86, 95% CI=1.20–2.88).

Table 1.

Prevalence of Weight Discrimination and Differences in Health by Source

	%	Subjective Health			Depressive Symptoms
		Discrimination/Source		d	Discrimination/Source		d
		Yes	No		Yes	No
Weight discrimination	12.9	3.33 (.04)	3.55 (.01)	−.23*	2.16 (.03)	1.73 (.01)	.56*
Source
Healthcare professional	23.5	3.19 (.07)	3.21 (.04)	−.02	2.46 (.06)	2.19 (.04)	.33*
Boss or coworkers	16.0	3.37 (.09)	3.17 (.04)	.19*	2.47 (.08)	2.22 (.03)	.31*
Service professional	26.3	3.25 (.07)	3.19 (.04)	.06	2.36 (.06)	2.22 (.04)	.17
Public safety officer	8.1	3.23 (.13)	3.20 (.04)	.03	2.60 (.11)	2.23 (.03)	.45*
Educator	4.5	3.47 (.17)	3.19 (.04)	.28	2.77 (.15)	2.23 (.03)	.69*
Stranger	51.9	3.12 (.05)	3.30 (.05)	−.18*	2.20 (.04)	2.32 (.04)	−.15*
Acquaintance	14.5	3.14 (.09)	3.21 (.04)	−.08	2.41 (.08)	2.23 (.03)	.22*
Close friend	11.5	3.23 (.11)	3.20 (.04)	−.03	2.55 (.09)	2.22 (.03)	.40*
Family member	20.4	3.06 (.08)	3.24 (.04)	−.18*	2.41 (.07)	2.22 (.04)	.26*

Note. N=4,990 for weight discrimination; n=642 for each source of discrimination. % is the percentage of the sample that reported weight discrimination/source of discrimination. Means are estimated marginal means (standard errors) from ANCOVA that compared the subjective health and depressive symptoms from participants who did (yes) or did not (no) report weight discrimination and each source of weight discrimination controlling for age, sex, ethnicity, education, and body mass index.

^*p<.05.

Across the sample, the mean rating for subjective health was 3.52 (SD=.97, range 1–5) and 1.79 (SD=.78, range 1–4) for depressive symptoms. Consistent with previous research (Sutin et al., 2015), participants who experienced discrimination based on their weight had lower subjective health and more depressive symptoms (Table 1). Among participants who experienced weight discrimination, the impact of that discrimination for health was not distributed equally across the sources of unfair treatment. Participants had lower subjective health when the perpetrator of the discrimination was either a stranger or a family member. Discrimination by a boss or coworker was associated with better health. There was no difference in subjective health between participants who reported one source of discrimination versus having experienced it from two or more sources.

A somewhat different pattern emerged for depressive symptoms (Table 1). For nearly every source of discrimination, depressive symptoms were higher among those who endorsed the source than those who did not. Interestingly, one of the exceptions to this pattern was for discrimination by a stranger. In contrast to subjective health, in which discrimination from a stranger had the strongest effect, depressive symptoms were slightly lower among those who reported weight discrimination from a stranger than those who did not. Finally, participants who reported experiencing weight discrimination from two or more sources had more depressive symptoms than participants who reported it from one source (M=2.45 versus M=2.12, p<.01).

Discussion

The present research suggests that the source of unfair treatment on the basis of body weight matters for subjective health and depressive symptoms. Of the participants who reported weight discrimination, approximately 1 in 5 reported that a family member was the perpetrator. Being treated unfairly by a family member is particularly harmful for the individual’s health. This finding is consistent with previous research that suggests discrimination from a family member has a stronger correlation with health than from other sources (Hunger & Tomiyama, 2014). The family is typically a source of support and protection, but this protective effect may not to extend to weight. Family members may think comments about body weight will motivate their loved ones to diet and lose weight. Such strategies do not work and may be associated with declines in health.

Strangers are the most common source of discrimination; approximately 50% of the current sample reported unfair treatment from a stranger. Interestingly, discrimination from a stranger mattered less for depressive symptoms than for subjective health. Individuals often report feeling more negative affect immediately after being stigmatized because of their weight by a stranger (Vartanian et al., 2014). The current findings suggest that this negative affect does not translate into more depressive symptoms over time. Rather, such interactions are associated with lower subjective health. Experimental evidence suggests that the stress of discrimination and the anticipation of such experiences happening again have physiological consequences (Sawyer, Major, Casad, Townsend, & Mendes, 2012) that may impair overall physical health.

Individuals who are overweight or obese commonly report unfair treatment by healthcare professionals (Puhl & Brownell, 2006). And, indeed, in the present sample, unfair treatment by a healthcare professional was among the most common sources of discrimination reported. Similar to other sources of discrimination, unfair treatment from a healthcare professional was associated with more depressive symptoms but was surprisingly unrelated to subjective health. There may be more lasting emotional than subjective health correlates of such encounters.

There were demographic differences in who reported discrimination and from the various sources. Consistent with previous research (Dutton et al., 2014), African Americans were less likely and participants with higher BMI were more likely to report weight discrimination. In contrast to this previous work, men and women were equally likely to report any discrimination based on their weight. Women, however, were more prone to unfair treatment based on their weight from family members than other sources, as were younger participants. Of note, although body weight was a significant predictor of several sources, different demographic groups were more likely to experience discrimination from specific sources.

Finally, it is of note that the sources of discrimination were more consistently associated with depressive symptoms than subjective health. In addition, attributing discrimination to multiple sources was associated with more depressive symptoms. Individuals who perceive discrimination from multiple sources may be more vulnerable to experiencing depressive symptoms. It may be, however, that those who are currently experiencing more depressive symptoms may be more likely to perceive ambiguous interactions as discriminatory. It was not possible to disentangle these two possibilities with the current cross-sectional data. Other limitations of this study include the use of subjective measures of both discrimination and health. Still, the present research suggests that there are differential effects of weight discrimination on health depending on the source of the unfair treatment.