Background Previous research has suggested that body image partially mediates the association between weight and depressive symptoms, but few longitudinal studies have been conducted. Traditional mediation methods, which use multiple regression models, cannot appropriately account for other potential pathways from weight to depression, such as bullying, which can result in distorted estimates of the direct effect. Furthermore, most studies use BMI to indicate weight, which can be elevated due to increased muscle mass, so may underestimate the association with depressive symptoms. We aimed to explore the extent to which body dissatisfaction at 14 years mediated the association between childhood weight trajectory and depressive symptoms at 17 years.

Methods We used data from the UK Millennium Cohort Study, a nationally representative longitudinal study of children born around the year 2000. Using Latent Class Growth Analysis, we derived trajectories to age 14 of BMI and of adiposity (relative proportion of Fat Mass to Fat-Free Mass). Participants rated their body dissatisfaction at age 14 (scale 1–7) and completed the Kessler Psychological Distress Scale at age 17. Counterfactual mediation analysis was conducted to estimate total, direct, and indirect effects.

Results Latent Class Growth Analysis revealed three trajectories of BMI and of adiposity: low, moderate, and high. Counterfactual mediation analysis of participants with valid trajectory data (N=13,025) suggested that compared to those with low BMI trajectories, depressive symptom scores among those with moderate trajectories were 0.29 points higher (95% confidence interval (CI) 0.10, 0.49) and in the high trajectory group were 0.60 points higher (95% CI 0.21, 0.99). Using adiposity trajectories, depressive symptom scores in the moderate trajectory group were 0.46 points higher (95% CI 0.25, 0.67) and the high trajectory group were 0.93 points higher (95% CI 0.51, 1.35), compared to those with low adiposity trajectories. The proportion of the association between BMI and depressive symptoms that was mediated by body dissatisfaction was 45%. For adiposity and depressive symptoms, the proportion mediated was 30%.

Conclusion A large proportion of the association between childhood weight trajectory and depressive symptoms at 17 years old is mediated by body dissatisfaction. Using BMI as a proxy for adiposity may underestimate the association with depressive symptoms. Interventions to improve adolescent body dissatisfaction could reduce depressive symptoms.