Concordance between self-reported body mass index with weight perception, self-rated health and appearance satisfaction in people living in Tehran

Self-reported BMI showed the high prevalence of overweight in men and normal weight in women compared with other BMI categories. In contradiction with previous studies which indicated the mean BMI 24.9 kg/m² (normal) and 26.5 kg/m² (overweight) in Iranian males and females, respectively [4], the current results showing the higher prevalence of overweight in men than women based on the self-reported BMI. These findings proposed that self-reported BMI may have some error in identifying the real BMI. A systematic review revealed trends of under-reporting for weight and BMI and over-reporting for height; although the degree of the trend was different for men and women [18]. Height overestimation and weight underestimation reported in women and great differences were observed between self-reporting and measuring the height and weight of several women based on the review article [19]. These findings explain the cause of the contradiction of the present results which showed that the men were more overweight than women, in spite of prevalence of overweight in Iran. The relationship between perceived weight status and self-reported BMI was moderate in another cross-sectional study [20].

The concordance between self-reported BMI and weight perception was significantly in both men and women found in our study. Weight perception in women is in line with self-reported BMI as normal, overweight, and obese category. Weight perception in overweight and obese men was according to self-reported BMI, and the women perceived their weight less than that their weight showed. Our data showed the women who had normal self-reported BMI perceived them as thin or underweight. In contrast with Japanese workers [21], the current results indicated a difference regarding the relationship between BMI categories and weight perception in men and women, i.e. based on kappa coefficient; the rate of agreement in women was greater than men (0.38 vs. 0.23). This finding is confirmed by higher percent of normal self-reported BMI in women than men, which could result in better weight perception in women. The association between BMI and weight perception of Japanese differed by age and gender [21]. A cross-sectional analysis among employees showed that the women considered themselves more to be overweight through each BMI category when compared with men [22]. Within weight perception categories, high percentage of women correctly perceived their weight as normal, overweight and obese in high degree of agreement with self-reported BMI.

It was found that the perception regarding weight, health, and appearance satisfaction varied according to gender and subjects’ age group. There were no significant differences among age groups within levels of weight perception and appearance satisfaction.

Within age groups, only women in their thirties significantly perceived themselves healthy or very healthy (p??0.0001) which probably due to absence of or little chronic diseases in this age decade. The results also revealed that the overweight men considered themselves healthy and very healthy, whereas normal weight women stated SRH as very healthy more. The correlation between SRH and self-reported BMI was examined through the agreement coefficient. Despite of the upper kappa coefficient in women compared to men, our data showed there was no significant agreement between SRH and self-reported BMI in each sex groups, which is contrast with the study of Giron’s on Spanish women who report SRH as good less than men [23]. SRH may be affected by disease status; based on self-report diseases, they suffered from a variety of disease and disabilities by gender. Thus, we explored the impact of chronic diseases’ history, as a possible confounder, on the association of self-reported BMI and SRH. This association was similar between two groups with and without chronic diseases by sex. Figure 1 showed the status of SRH in men and women who suffered from chronic disease.

Comparison between percentage of men and women in terms of association of chronic diseases and self-rated health

Using SRH provides a valid, reliable, and cost-effective health assessment; particularly, in studies lacking other forms of health information [24, 25]. In our study, the pattern of SRH was similar to weight perception, so that the observed agreement between weight perceptions with self-reported BMI in women was also seen for SRH concordance with self-reported BMI, however, this agreement was not significant. Nevertheless, the results of a research on Swedish people demonstrated that the reliability of SRH is as good as or even better than more specific questions [26].

Our data revealed more overweight men were satisfied with their appearance, and women classified themselves better than men similar to European Union [11]. The majority of overweight men were meaningfully satisfied with their appearance as much or very much. According to the findings, the discordance between the men’s appearance satisfaction and self-reported BMI was more compared to women; however, the agreement coefficient was not significant in both gender groups. As Korean people who were not satisfied with their weight in spite of the fact that they had normal or lean figures, except for 3.3 % of the total samples [27]. Body image would be valid between 30 % of men and 60 % of women approximately; however, it had high validity in underweight males and females, [11]. The rate of appearance satisfaction in women was reduced with increasing the age. Body image is the important factor to evaluate women in societies like Iran and for this reason they are pushed to pay more attention to body shape and its management. Thus, their appearance satisfaction will be lower than men [9].

This study has some limitations which the most important being the lack of direct weight and height measurement. Because of extensive data collected in the main great design and a large number of subjects, this was not possible. These data could help the researchers to compare the self-reported and real measurements to conclude the better association between variables.