Weight status misperceptions among UK adults: the use of self-reported vs. measured BMI

A number of studies have examined whether personal perceptions of weight status correspond with objective weight status. A significant proportion of overweight males and females are thought to underestimate their weight status and believe that their weight ‘is about right’ [1, 2]. Likewise, some studies suggest that a concerning number of healthy weight females believe that their body weight is too heavy [1, 3]. However, estimates of how common under and over estimation of weight status is can vary substantially across studies [4–7].

One explanation of weight status misperceptions is based on body weight norms. For example, Burke, Heiland & Nadler [2] found that fewer US adults identified themselves as being overweight in 2004 than 1994 despite large increases in national rates of obesity during this time. This could be indicative of a generational shift in terms of what is considered to be a normal weight [2]. Moreover, it has been suggested that perception of weight status is more likely to be determined by using those around us as a reference rather than by using clinical recommendations [8] and in line with this overweight and obese teenagers are less likely to think of themselves as being overweight if they have overweight classmates and parents [9]. Furthermore, in recent studies visual exposure to obese males led participants to rate an overweight man as being a healthier [10] and a more normal weight [11] than those exposed to healthy weight males. Thus, it may be that frequent exposure to heavier bodies has led to an upwards shift in terms of what people think a normal weight looks like [12]. This may cause heavier body weights to appear more normal and result in weight status misperceptions [13].

An important unexplored factor that may influence estimates of weight misperception is the use of self-reported vs. measured BMI to categorise participant weight status. Some studies have used objectively measured BMI to draw inferences about weight misperceptions [14–17]. However, there are practical constraints associated with this method and a large number of studies have used self-reported weight to determine the prevalence of weight status misperceptions [6, 7, 13, 18, 19]. This could be problematic as a substantial body of research suggests that BMI is often underestimated when using self-reported data, which can result in underestimation of overweight/obesity [20–22] and overestimation of the proportion of the population who are of a healthy weight. Thus, a potential consequence of using self-reported BMI to examine weight misperception is that it may result in systematic underestimation of the percentage of overweight individuals who misperceive their weight status as being ‘about right’, when they are in fact overweight. Likewise, because self-reported BMI will erroneously categorise some individuals as being of healthy weight, when their weight actually places them in the ‘overweight’ range, it may also result in overestimation of the prevalence of weight misperception amongst healthy weight individuals. Moreover, some studies have indicated that the measurement error associated with self-reported BMI has declined over the last 20 years [23, 24]. This could result in apparent (but erroneous) increases in the prevalence of weight misperceptions. For example, one study reports an increase in underestimation of weight status in women between 2007 and 2012 [6]. However, this may have been in part caused by a larger measurement error in self-reported BMI masking the true prevalence of weight misperceptions at the earlier time point (i.e. making underestimation of weight status among overweight individuals look less common than it actually was).

Although there are general concerns about the accuracy of using self-reported BMI (as opposed to objectively measured BMI) in obesity research, to date no study has examined whether the use of self-reported BMI biases prevalence estimates of weight status misperceptions. Thus, the aim of the present study was to test this hypothesis. In the present study we made use of recently collected data from a large UK study (2013 Health Survey for England; HSE) in order to estimate the prevalence of weight misperceptions among UK adults. As HSE includes measures of self-reported and researcher measured BMI, this allowed us to also examine the hypothesis that the use of self-reported BMI may result in biased estimates of weight misperceptions. We predicted that the use of self-reported BMI may result in an underestimation of the number of overweight individuals who misperceive their weight. Because HSE also includes measured waist circumference, which may be a better indicator of disease risk than BMI [25], we also examined weight perceptions as a function of waist circumference. This allowed us to more thoroughly estimate the prevalence of weight status misperceptions in individuals whose body composition (in terms of waist circumference) places them at an elevated disease risk.

In the present study we made use of recently collected data to examine the prevalence of weight status misperceptions in a sample of UK adults. Using measured BMI, we found that a large proportion of overweight males (55 %) and females (31 %) perceived their weight as being ‘about right’, although underestimation of weight status was less common among obese participants (≤10 %). Moreover, using waist circumference as a measure of risk of future ill health, around half of males and females at increased risk and almost one fifth of those at high risk, perceived their weight status as being ‘about right’. Even when ill health risk profiles were combined (e.g. participants with overweight or obese BMI and raised weight circumference) a substantial percentage of individuals underestimated their weight as being ‘about right’. These findings may have public health implications, as weight status misperceptions are thought to be associated with a variety of health relevant outcomes, such as weight loss intentions, physical activity and mental well-being [3, 5, 18, 19]. The present findings are consistent with the notion that heavier body weights may have now become ‘normalised’ as a result of increased obesity prevalence. It has been previously suggested that this process of normalisation may have increased the prevalence of weight status misperceptions among overweight and obese individuals [2, 11]. Moreover, it is plausible that larger body sizes may have now become more acceptable due to the increased prevalence of obesity [12].

In the present study we also examined whether the use of self-reported BMI, as opposed to objectively measured BMI, produces measurement error when estimating the prevalence of weight status misperceptions. The use of self-reported BMI resulted in an underestimation of the number of overweight individuals misperceiving their weight (‘about right’) and overestimated the number of healthy weight individuals who misperceived their weight status as being ‘too heavy’. In some instances this measurement error was sizeable. For example, the use of self-reported BMI resulted in approximately only 19 % of overweight females underestimating their weight status, when in reality 31 % did so. This is of importance as a large number of studies have used only self-reported BMI data to estimate both the prevalence of weight status misperceptions and their possible consequences [6, 7, 13, 18, 19]. Moreover, because the degree of underestimation of BMI caused by self-reported weight and height may change over time [23, 24], it will be important for studies which attempt to track longitudinal changes in weight misperceptions to adjust for this potential confound or rely on objective measures of adiposity, which has not always been the case to date [6, 7]. Our findings also support recent suggestions that objective measures of adiposity are required when examining weight status misperception [28]. There are of course practical constraints associated with collecting objective measures of adiposity and this may have contributed to the reliance on self-report measures of BMI in studies examining weight perceptions [6, 7, 13, 18, 19]. Nonetheless, where possible we strongly suggest that future research should make use of objectively measured BMI. If this is not at all feasible, then it will be important to consider the likelihood that the estimates of weight misperceptions produced using self-reported BMI will be biased and underestimate the frequency by which overweight individuals misperceive their weight status.

A large proportion of UK adults who are overweight misperceive their weight status. The use of self-reported BMI data is likely to produce biased estimates of weight status misperceptions. The use of objectively measured BMI is preferable as it will provide more accurate estimates of weight misperception.