Abstract

Background/objectives To identify predictors of obesity in adults and investigate to what extent these predictors are independent of other major confounding factors. Subjects/methods Data collected at baseline from 1441 participants from the Food4Me study conducted in seven European countries were included in this study. A food frequency questionnaire was used to measure dietary intake. Accelerometers were used to assess physical activity levels (PA), whereas participants self-reported their body weight, height and waist circumference via the internet. Results The main factors associated (p 0.05) with higher BMI per 1-SD increase in the exposure were age (beta:1.11 kg/m(2)), intakes of processed meat (beta:1.04 kg/m(2)), red meat (beta:1.02 kg/m(2)), saturated fat (beta:0.84 kg/m(2)), monounsaturated fat (beta:0.80 kg/m(2)), protein (beta:0.74 kg/m(2)), total energy intake (beta:0.50 kg/m(2)), olive oil (beta:0.36 kg/m(2)), sugar sweetened carbonated drinks (beta:0.33 kg/m(2)) and sedentary time (beta:0.73 kg/m(2)). In contrast, the main factors associated with lower BMI per 1-SD increase in the exposure were PA (beta:-1.36 kg/m(2)), intakes of wholegrains (beta:-1.05 kg/m(2)), fibre (beta:-1.02 kg/m(2)), fruits and vegetables (beta:-0.52 kg/m(2)), nuts (beta:-0.52 kg/m(2)), polyunsaturated fat (beta:-0.50 kg/m(2)), Healthy Eating Index (beta:-0.42 kg/m(2)), Mediterranean diet score (beta:-0.40 kg/m(2)), oily fish (beta:-0.31 kg/m(2)), dairy (beta:-0.31 kg/m(2)) and fruit juice (beta:-0.25 kg/m(2)). Conclusions These findings are important for public health and suggest that promotion of increased PA, reducing sedentary behaviours and improving the overall quality of dietary patterns are important strategies for addressing the existing obesity epidemic and associated disease burden.