Associations between weight/shape overvaluation, sociodemographic features and BMI: 10-year time trends

To investigate the prevalence of overvaluation across sociodemographic features and weight status over time.

The data included sequential cross-sectional surveys with representative samples of the adolescent and adult (15 years or older) population in South Australia. Five surveys that assessed overvaluation were conducted in the years 2005 (n = 3047), 2008 (n = 3034), 2009 (n = 3007), 2015 (n = 3005) and 2016 (n = 3047). Overvaluation was assessed by structured interview based on the Eating Disorder Examination. To examine unique effects of demographic variables on the likelihood to report overvaluation, and also to examine whether this varied as a function of time, a multivariate binary logistic regression was computed.

Across survey years, participants who were more likely to endorse overvaluation were female (2005: OR 2.85, CI 2.04-3.99; 2008/9: OR 1.74, CI 1.50-2.01; 2015/6: OR 1.54, CI 1.34-1.76), had a BMI > 30 (2005: OR 3.93, CI 1.49-10.34; 2008/9: OR 2.22, CI 1.31-3.78; 2015/6: OR 2.09, CI 1.19-3.67), had left school (2015/6: OR 1.36, CI 1.14-1.63), and lived in the country (2015/6: OR 1.95, CI 1.69-2.24). Being in the oldest age group was protective against endorsing overvaluation in each survey year. There was also a main effect of survey year, with participants in the 2015/6 survey more likely to endorse overvaluation (p < 0.001).

Female, young and obese people were more likely to endorse overvaluation; however, the prevalence of overvaluation increased significantly in all sociodemographic and BMI groups in since 2005-2016.

Level IV, evidence obtained from multiple time series with or without the intervention, such as case studies.