Body weight dissatisfaction by socioeconomic status among obese, preobese and normal weight women and men: results of the cross-sectional KORA Augsburg S4 population survey

Body image and, more specifically, body weight dissatisfaction have been shown to be positively associated with intending or trying to lose weight in adults [1–5]. Obviously, this does not guarantee the eventual success of such intentions or attempts. Quite the contrary, there is at least mixed evidence concerning the assumption that body weight dissatisfaction is a pre-treatment barrier to weight control [6, 7]. However, this evidence only pertains to patients already enrolled in an obesity treatment program. Thus, it has been argued that being dissatisfied with one’s body weight may not only be a risk factor for eating disorders and obesity (such as in adolescents [8] or chronic dieters [9]). It may also be a potential motivator for initiating healthy lifestyle changes [7], especially among obese adults [10], as long as it does not exceed critical levels which may actually impede motivation [11].

In any case, there is little doubt that body weight dissatisfaction is an important factor in preventing weight gain and promoting weight loss or maintenance. The distribution of body weight dissatisfaction among relevant target groups is, therefore, of particular interest. Within economically wealthy countries with Western lifestyles, these groups include those with relatively low socioeconomic status (SES) which tend to be affected by the highest obesity rates [12]. In these countries where body dissatisfaction is most common (most notably in the US [13, 14]), it has been asserted that women—particularly those with a comparatively high SES—are those most dissatisfied with their body and body weight [15]. For men, such research has traditionally been more scarce [16]. Studies from some European countries have shown, however, that higher SES is associated with higher odds of body weight dissatisfaction in men, but not women [17, 18].

Most studies on the relationship between SES and body weight dissatisfaction have tended to scrutinize this association in general, rather than looking at normal weight, preobese or obese subgroups separately [1, 3–5, 15, 17, 18]. A similar assertion holds for self-perceived weight appropriateness [19–21], a construct related to though distinct from body weight dissatisfaction [22]. Hypothetically, the positive association between SES and body weight dissatisfaction should primarily be present among overweight groups, given that those with high SES “…have more narrowly defined standards for acceptable body size” [20]. In other words, although those in the normal weight range must also meet relatively narrow standards, roughly increased odds of body weight dissatisfaction in high SES groups should be found for those who are overweight since they more strongly deviate from the defined standards. Alternatively, if the assertion holds that “…social class is not related to body shape ideals, since people from different social classes present similar ideals” [23], social inequalities in body weight dissatisfaction should be rather small, regardless of body weight status.

Against this background, this study presents further analyses of German population-based data from the KORA S4 study [18]. In specific, it examines the distribution of body weight dissatisfaction across different socioeconomic status groups among normal weight, preobese and obese women and men.

To summarize the results of the present study, no significant differences in the odds of body weight dissatisfaction were found for normal weight women and men by SES. Among the preobese, increased odds of body weight dissatisfaction were found in men with the highest SES index values, a middle or high educational level, high income, and middle or higher occupational status, while in women, odds were only increased for those with a middle educational level. In the obese groups, SES differences in body weight dissatisfaction were greater than in the non-obese groups. Among men, odds were up to 3.7 times higher in the group with the highest SES. As in the group of preobese men—though even more pronounced—a dose–response gradient in dissatisfaction was observed across SES groups. Among obese women, however, again no significant differences were found except that those with a high educational level had marginally significant higher odds of dissatisfaction than those with the lowest level of education. The analyses conducted for moderately obese participants revealed results that were essentially similar to those for the obese groups as a whole.

Among the strengths of the present study is that its data come from a survey which used rigorous quality assurance procedures [24]. Body weight and height were measured (not self-reported), and sample size allowed for stratification for both BMI status and gender. At the same time, the study also has a number of noteworthy limitations. First, body weight dissatisfaction was assessed by a single item only. The main reason for this was practicality since the survey from which data were taken was an omnibus survey which already took participants an average of three hours to complete. While comparably simple items have been used in previous epidemiological studies [3, 5], the use of a single item for measuring body weight dissatisfaction implies that the reliability of this indicator is uncertain and that there is a need for replication using a more sophisticated measure (e.g., the Body Areas Satisfaction Scale [35]). Regarding the sensitivity of this single dichotomized item, cross-checks of the logistic models with linear models using the original 4-point scale indicated that the results essentially lead to similar conclusions, with the only exception being that mean differences on the scale appear to be of less magnitude than differences in percentages of respondents being dissatisfied. A second limitation is that the survey was regionally confined. Since body weight dissatisfaction may vary systematically with regional affluence and obesity prevalence [36, 37], studies conducted in other regions or at the national level controlling for regional attributes could clarify whether similar patterns would emerge under different circumstances. Third, no measure of perception regarding weight status as such was available. In other words, lack of awareness of overweight status among men with lower SES may have contributed to their lower prevalence of body weight dissatisfaction. Finally, subsample sizes did not allow for an investigation of the individual obesity classes 1, 2 and 3 (i.e., BMI of 30–34.9, 35–40, and ≥ 40). Thus, specificities of severe obesity (classes 2–3)—e.g., in terms of disproportionally lowered health-related quality of life in men [38]—may have been overlooked. For instance, while among men the health care costs of severe obesity are known to be disproportionately higher than for other BMI groups (especially given high SES [39]), among women these costs tend to increase most when progressing from class 2 to class 3 obesity [40]. Thus, body weight dissatisfaction associated with specific (perceived or evaluated) need factors may be more prevalent in higher SES women with morbid obesity (class 3). At the same time, it can be asserted that the results for moderately obese women and men regarding the association between SES and the likelihood of being dissatisfied with one’s body weight reflected those for obese groups as a whole, with significant differences among men being even somewhat more pronounced, which, in a sense, makes the strategy to analyze moderate and severe obesity together seem rather conservative in terms of hypothesis testing. In other words, the finding that body weight dissatisfaction is more frequent in high socioeconomic groups among obese men and that there is a social gradient in body weight dissatisfaction in this group indicates that body weight satisfaction is not “merely” attributable to certain specificities (e.g., ill-health states) of the severely obese subgroup.

Bearing in mind the study’s limitations, its results statistically support previous evidence that body weight dissatisfaction is practically a normative discontent among women, not men (at least in European countries). A total of 48.3% of the women in this study were dissatisfied with their weight (men: 33%), a rate nearing the 50% which has been argued to constitute normative discontent [17]. While body weight dissatisfaction in bivariate (i.e., unstratified and unadjusted) analysis was negatively associated with women’s SES, no association was found within any of the three examined BMI categories. Thus, when compared to men, this bivariate association may be attributed to the relatively high rates of both low SES and obesity in women within the lowest SES group. By contrast, body weight dissatisfaction rates among men increased gradually with SES both overall and in the preobese and particularly the obese group. This suggests that, in this population of German adults, the premise that SES is unrelated to body size standards [23] may hold for women. Conversely, standards among overweight men in higher socioeconomic echelons may be more narrowly defined than those among their lower SES counterparts.

Nevertheless, the findings of this study should be tested against alternative interpretations since the study did not assess individual body image ideals. Using measures of abdominal obesity and muscularity, for example, which was beyond the scope of this analysis, may give hints as to whether the relatively low proportion of body weight dissatisfaction in lower SES obese men possibly reflects adherence to bodily standards other than fat-free leanness (i.e., less internalization of the thin ideal). In addition, psychosocial factors other than individual body ideals and (gender- and SES-specific) internalizations of the thin ideal may be operative [16, 41]. First, importance of appearance—known to remain important throughout the lifespan of women [41] but not men [16]—may be less SES-specific among (pre-)obese women than men. Second, self-objectivation (defined as adopting an observer’s perspective of one’s physical self due to social construal of the body as an object to be evaluated) may be a mediating factor along a similar line of argument. Third, exposure to and perception of pressure from social environments, such as the media, may vary for both sexes in different SES groups. For instance, the number of magazines with a focus on men’s appearance and health has increased substantially over recent decades [42]. It is possible that this “innovation” has diffused (i.e., magazines have been read) more among men with a comparatively high SES given that in general “… individuals’ socioeconomic status is highly related to their degree of change agent contact” [43]. By contrast, publications focusing on women’s appearance and health have been around longer giving them more time to be diffused across the whole SES continuum.

Gender differences in the associations between body weight dissatisfaction and different SES indicators also merit discussion. First of all, in this study, the SES index and the individual SES indicators (educational level, income and occupational status) produced rather similar results in men, which of course is not always the case for other health or health-related outcomes [44]. The fact that the highest odds ratios pertained to the SES index compares well to previous results on perceived weight appropriateness [19]. A different pattern emerged for women in this study, however. Here, educational level stood out from the general trend of equality in body weight dissatisfaction across SES strata. Preobese women with a middle educational level and especially obese women with a high level of education had greater odds of being dissatisfied with their body weight than those in the respective lowest stratum. Although these results were somewhat inconsistent in the preobese group (middle not high educational level) and unstable in the obese group (large confidence interval), they may reflect gendered upbringing regarding body size standards. That is, given that education is a marker of childhood social environment resulting in differences in awareness among adults [44], shared socialization in the past may have affected both men and women. By contrast, current income and occupational status seem more relevant for men than women.

Body weight dissatisfaction is a double-edged sword. It has been argued to be both a risk factor for obesity among young adults as well as a factor motivating weight loss in middle-aged or aged adults. The present finding that body weight dissatisfaction in German adults follows a dose–response gradient across socioeconomic strata in preobese and particularly in obese men (but not women) may point to obese men with a lower SES as a target group for obesity management. On the one hand, body weight satisfaction may partly explain why obesity is more frequent in groups with low socioeconomic status among men, while among women, SES seems unrelated to body size standards (even though the inverse social gradient in obesity is steeper than in men). On the other hand, combined with the fact that German men with low SES have particularly high odds of not exercising at all [45], health care and public health interventions should probably focus more on body image and, specifically, on body weight satisfaction in overweight men with relatively adverse socioeconomic backgrounds.