Mid-term Body Mass Index increase among obese and non-obese individuals in middle life and deprivation status: A cohort study

BACKGROUND In the UK, obesity is associated with a clear socioeconomic gradient, with individuals of lower socioeconomic status being more likely to be obese. Several previous studies, using individual measures of soecioeconomic status, have shown a more rapid increase in Body Mass Index (BMI) over time among adults of lower socioeconomic status. We conducted a study to further examine whether ecologically defined deprivation status influences within-individual BMI change during middle life, as the answer to this question can help determine optimal preventive strategies both for obesity per se, and its' associated socioeconomic disparities. METHODS Anonymised records of participants to the Stockport population-based cardiovascular disease risk factor screening programme were analysed. Individuals aged 35-55 who had a first screening episode between 1989 and 1993, and a subsequent screening episode were included in the study. Deprivation status was defined using quintiles of the Townsend score. Mean annual BMI change by deprivation group was calculated using linear regression. Subsequently, deprivation group was included in the model as an ordinal variable, to test for trend. The modelling was repeated separately for individuals who were obese (BMI < 30) and non-obese at the time of first screening. In supplementary analysis, regression models were also adjusted for baseline BMI. RESULTS Of 21,976 women and 19,158 men initially screened, final analysis included just over half of all individuals [11,158 (50.8%) women and 9,831 (51.3%) men], due to the combined effect of loss to follow-up and incomplete BMI ascertainment. In both sexes BMI increased by 0.19 kg/m2 annually (95% Confidence Intervals 0.15-0.24 for women and 0.16-0.23 for men). All deprivation groups had similar mean annual change, and there was no evidence of a significant deprivation trend (p = 0.801, women and 0.892, men). Restricting the analysis to individuals who were non-obese at baseline did not alter the results in relation to the lack of a deprivation effect. When restricting the analysis to individuals who were obese at baseline however, the findings were suggestive of an association of BMI increase with higher deprivation group, which was further supported by a significant association when adjusting for baseline BMI. CONCLUSION In the study setting, the BMI of non-obese individuals aged 35-55 was increasing over time independently of deprivation status; among obese individuals a positive association with higher deprivation was found. The findings support that socioeconomic differences in mean BMI and obesity status are principally attained prior to 35 years of age. Efforts to tackle inequalities in mean BMI and obesity status should principally concentrate in earlier life periods, although there may still be scope for focusing inequality reduction efforts on obese individuals even in middle life.