LAUSR

ABSTRACT Individual variability may contribute to the modest and inconsistent results reported in obesity-management interventions. This study aimed to investigate the impact of non-modifiable as well as modifiable factors on body mass index (BMI) and body fat variance in adolescents with obesity followed in a clinical obesity-management programme, in order to better understand individual variability. Non-modifiable factors (i.e. socio-economic status, pregnancy BMI, weight progression across pregnancy, BMI at time of delivery, way of delivery, birth weight, breastfeeding duration, age at overweight onset, overweight duration, and FTO rs9939609 polymorphism) and modifiable factors data (i.e. self-determination level, self-efficacy and perception of importance to lose weight, energy intake, physical activity, and sedentary behaviours) from 63 adolescents (93.7% Caucasian, 55.6% girls), with a median age of 15.0 (2.5) years, and a median BMI z-score of 2.88 (0.70), followed for 6 months were analyzed. BMI z-score variance was predicted by vigorous physical activity (VPA) (F(1,57) =  4.55, p = .039), overweight duration (F(1,59) =  5.61, p = .022), way of delivery (F(2,58) =  6.55, p = .003) and self-determination level (F(1,59) =  4.75, p = .034). VPA further predicted body fat mass (%) (F(1,57) =  9.99, p = .003) as well as trunk fat mass variance (F(1,57) =  8.94, p = .006). This study suggests that although both non-modifiable and modifiable factors influence BMI and body fat variance to some extent, in adolescents with obesity, VPA (modifiable factor) stands out as the factor with the best association with both outcomes. VPA may be a potential ally in the success of clinical obesity management in adolescents, and so should be emphasised in this population. Highlights There is a huge individual variability within studies in response to adolescent obesity-management interventions. Both non-modifiable and modifiable factors may influence body mass index (BMI) and body fat variance, influencing interventions’ outcomes. The predictive value of both non-modifiable and modifiable factors largely overlaps, making lighter the burden of the former and highlighting the value of lifestyle changes. Among modifiable factors, vigorous physical activity standouts as the factor with the best (negative) association with BMI and body fat variance.