Psychological, Social, and Behavioural Mediators of the Link Between Obesity and Depression in Childhood and Adolescence: A Systematic Review and Meta-Analysis

Abstract

Young people with obesity are significantly more likely to experience depression than their lower weight peers (Quek et al., 2017; Sutaria et al., 2019). This is problematic, as both conditions track into adulthood (Alberga et al., 2012; Thapar et al., 2012) and their comorbidity is associated with poor health and treatment outcomes (McElroy, 2015). The mechanisms through which obesity may contribute to depression are poorly understood. Uncovering these processes could provide targets for treatment and prevention that would improve outcomes for children and youth in larger bodies. The purpose of this systematic review and meta-analysis was to identify these mechanisms, by synthesizing and critically appraising the scientific evidence examining psychological, social, and behavioural mediators of the link between weight status and depressive symptoms in this population. Methods: MEDLINE, CINAHL, PsycINFO, and PubMed were systematically searched using a peer-reviewed strategy. Eligible reports used formal mediation analysis or causal modelling to assess the impact of any psychological, social, and/or behavioural variable(s) on the association between weight status and depressive symptoms in children and youth between five and 18 (22) years. Clinical samples and intervention trials were excluded. Quantitative (random-effects meta-regressions and subgroup analysis) and qualitative (statistical vote counting) methods were employed to synthesize the body of evidence for each outcome, the strength of which was rated using a structured framework. Results: 19 reports were included, yielding 43 effects eligible for inclusion in quantitative and/or qualitative synthesis. Results identified a small, statistically significant average indirect effect of all investigated mediators. Gender did not significantly moderate mediation pathways, with many mechanisms showing similar strength of effect for boys and girls. Insufficient evidence was available with which to evaluate developmental stage and culture as moderators of mediation pathways. Partial support was found for the validity of the gender additive model of depression (GAM; Stice & Bearman, 2001) in explaining the link between weight status and depression: a small, statistically significant effect was found for GAM variables, though no effects were reported for 'pressure to be thin' or 'thin ideal internalization' and effects were not significantly different than those of non-GAM variables. Cognitive, social, and behavioural mediators were found to have similar magnitudes of effect, suggesting that clinical assessment and intervention target all three of these areas. Body dissatisfaction, teasing, and dieting were highlighted as individual mediators warranting clinical attention, though the strength of this evidence was generally rated low, signalling a need for further research in this area.