Is social exposure to obesity associated with weight status misperception? Assessing Australians ability to identify overweight and obesity

Australia is experiencing an obesity epidemic, directly and indirectly costing the economy a conservative estimate of $8.6 billion (2014/15), a figure forecast to reach $87.7 billion by 2025 if not addressed [1]. In 2014–15, almost two out of three Australians (11.2 million people, 63%) were overweight or obese, a significant increase since 1995 (56%) [2]. Unhealthy weight increases the risk of chronic conditions including some forms of cancer, eating disorders, poor mental health, osteoarthritis, Type 2 Diabetes Mellitus (T2DM) and Cardiovascular Disease (CVD) [3, 4]. Overweight and obesity prevalence is even greater in rural populations (69%) [5].

Early intervention and secondary prevention of an unhealthy weight relies on Australian primary healthcare professionals routinely calculating body mass predictors (height and weight to calculate a Body Mass Index (BMI) and waist circumference), on every adult [4, 6]. Criticisms have been raised regarding the use of BMI due to the inability to assess proportions of muscle mass compared with fat mass [7], though when combined with waist circumference, it is considered a strong predictor of health risk [8]. Eighty seven percent of Australians access their General Practitioner (GP) at least once a year [9], emphasising that GPs are well poised to identify an unhealthy weight and treat accordingly [4, 8]. Furthermore, patients expect their GP to initiate a weight related discussion when required [10]. Yet, contemporary evidence suggests that less than a quarter of Australian adults (22%) have a BMI documented in their GP medical record and even fewer have a waist measurement recorded [11]. Morevoer, obesity is the individual problem managed in a patient encounter in 0.8% of GP consultations (0.8%) [6].

An unhealthy weight is however, a risk factor for many of the most common presenting conditions in general practice including musculoskeletal encounters (12% of GP activity), circulatory encounters (10%) or endocrine and metabolic encounters (9%) and therefore may provide a supplementary opportunity for brief intervention [6]. Although average appointment times are brief (approximately 15 min) [6], the majority of GP encounters involve a single issue and it is important to highlight the contribution of weight to relevant presenting complaints. Any opportunity for brief intervention (30 s) to motivate weight loss is known to be helpful and acceptable to over 99% of patients with obesity, even when the encounter is unrelated to weight [12]. Incidently, Ling and colleagues report that for adolescents, health professional recommendations are more likely to influence weight loss attempts than parents [13].

In a society where more people exhibit an unhealthy weight than a healthy weight, social perception of a healthy weight is at risk of becoming obscured [14], leading to people with overweight and/or obesity perceiving their weight as healthy. If people with overweight and/or obesity do not recognise their weight as a risk factor for chronic conditions, public health messages will not resonate. Binkin et al. (2013) reported in their obesity perception study among mothers and children that ‘what is common has a greater likelihood of being perceived as normal’ [15]. Misperception develops as early as 9 years of age and is heavily influenced by peer and familial weight [16]. Among adults, males are reportedly more likely to misperceive an overweight or obese weight status than females, regardless of cultural background [17]. Any misperception among society may be cause for concern, considering that it is believed that an individual attempt to lose weight is mainly driven by a self-perception of being “too heavy”, regardless of accuracy [13]. Health professionals are not immune to weight status misperception [10, 14, 18, 19] hence, poor identification of an unhealthy weight by health professionals may mislead or contribute to misperception.

There is limited evidence investigating whether adult Australians may be becoming obscured in their perceptions of weight status by the increase in prevalence of overweight and obesity. The aim of this study was to assess whether Australians, including health professionals are challenged to identify overweight or obesity and whether rural residence, body size or gender impacts identification.

A total of 806 questionnaires were completed, a majority of responders were female (n = 656, 85.1%), over half were HPs (n = 379, 50.8%) and of those that provided a postcode (n = 770, 95.5%), 64.4%, (n = 519) lived in a regional/remote community and 32.6% (n = 251) resided in a major city. On average participants were 41.0 ± 12.5 years old, a high rate of tertiary education and employment status is evident (Table 1).

Overweight and obesity prevalence among self-reporting study respondents was lower than current Australian estimates (46.7% versus 63.4% respectively) [2]. Health literacy levels may be assumed to be high, considering the sample were well educated and mostly employed health professionals. Yet, a majority of respondents were unable to identify the lower limit of a BMI corresponding with the obese category, in both themselves and others. In fact, as weight status increased, accuracy decreased. The national obesity epidemic, where almost two out of three people are either overweight or obese [2] may have engendered a society that no longer recognises the threshold of unhealthy weight.

Men in particular lacked accuracy in self-perception of weight status using medicalised words (underweight; healthy weight; overweight; obese) and were more likely to underestimate their weight status using silhouettes than females. Men’s inability to accurately perceive their weight status using either medicalised words or visual cues has some support in the literature. A study from the United States (US), found that men with an overweight weight BMI perceived that they had a BMI corresponding with an underweight or a healthy weight status [24].

Weight status misperception using medicalised terminology was greatest among respondents corresponding with each end of the weight staus index (underweight and obese), regardless of gender. People at greatest risk of chronic ill health due to their weight therefore appear to have the poorest self-perception of their weight. In their cross-sectional survey of adolescents in Hong Kong, Cheung et al. (2007) found that females in particular are motivated to adopt weight control behaviours if they perceive themselves to be overweight [25]. Assuming active weight control behaviours resulting from weight self-perception in adolescents continue into adulthood, consistent attempts to measure weight predictors and discuss the effects of weight upon health is likely to be beneficial throughout life.

Respondents with a BMI corresponding with the obese category were however, more accurate in self-perception of weight status using silhouettes than medicalised words, a finding consistent with Harris et al. 2008. Self-perceiving an unhealthy weight is known to be most problematic among people with a BMI equating to overweight or obesity [20, 26]. Surprisingly however, regardless of weight status, almost nine out of ten respondents misperceived the lower limit of obesity using silhouettes, a finding that suggests that increasing prevalence of obesity and social exposure may indeed be influencing social perception of obesity [18].

Healthy weight people are now a minority in Australian society potentially influencing the perception of both HPs and the general population alike. HPs and non-HP respondents equally misperceived the lower limit of obesity using either medicalised terminology or silhouettes. Wong and colleagues (2016) found that medically trained professionals and the general public were mutually inaccurate in their perceptions of people identified as overweight or obese [19]. A recent US study found that parents underestimated the weight of their children with an overweight or obese BMI (96 and 72% respectively) and this pattern appeared independent of socioeconomic/educational factors [27]. The authors attributed the findings to social and cultural beliefs and attitudes that have resulted from the high incidence of children identified as overweight or obese in contemporary society in comparison to previous generations [27]. As it appears that weight related perceptions develop early in life, weight concerns ought to be addressed as early as possible, particularly in rural areas where prevalence is greater [5].

It is important to note the identified limitations of this study. Firstly, the responding adults in the study may be subject to selection bias and findings cannot be generalised to the Australian population, though may be representative of Australians with a mid to high socioeconomic status. The online survey was primarily disseminated within regional health services where the authors are situated, presumably contributing to the high representation of English speaking females with internet access, health professionals with a higher education and regional/remote people. If the sample was more representative of the Australian population and included a higher proportion of males, people with low socioeconomic status and low educational attainment, then perception of overweight and obesity may be able to be more fully explored among people at particular risk of obesity. Secondly, results are subject to social desirability bias where it is known that people may report an inaccurate height and weight in an attempt to impress and conform with social norms [28], though underestimation of weight and overestimation of height is generally insufficient to alter BMI category [29]. Respondents were additionaly asked to self-report their waist circumference as a secondary predictor of health risk [8], in an attempt to mitigate bias, particularly as it is known that BMI alone fails to differentiate muscle mass and fat mass [7]. It is also known that people poorly monitor their weight, particularly men, a contributing factor in self-reporting bias [28]. Web-based testing however, has been found to moderate socially desirable responses 1.51 times greater than comparable pen-and-pencil questionnaires when seeking a disclosure of sensitive issues, due largely to the anonymous nature of inquiry [30].

People from cities, rural areas, health professionals and non-health professionals are equally challenged to accurately perceive a BMI corresponding with obesity in themselves and others. Weight status misperception was more likely to exist among those with a BMI less than 18.5 or 30 or more (underweight and obese). Largely, as weight status increased, accurate perception decreased. Visual cues however, appear to improve accuracy, given respondent corresponding with an obese BMI were significantly more likely to accurately self-perceive their weight status using silhouettes than medicalised words. Silhouettes therefore may act as an effective visual cue in initiating weight related discussions.