Examining the temporal relationship between depression and obesity: meta-analyses of prospective research

Abstract

Cross-sectional research has linked depression with obesity, leading to prospective assessments of the two variables. We conducted two meta-analyses aimed at (1) clarifying the temporal sequence of obesity and depression development; and (2) identifying moderator variables. Meta-analysis 1 involved studies assessing depression as a risk factor for weight gain, and included 26 effect sizes, based on the responses of 31,189 individuals. The weighted average correlation between depression and weight gain was small, but statistically significant (r=0.08). Moderator analyses showed more substantial effect sizes for females (r=0.12) and for studies that used as their dependent variable becoming obese (r=0.17). Meta-analysis 2 involved studies assessing elevated weight as a risk factor for depression, and included 13 effect sizes (15 for moderator analyses) based on the responses of 24,120 participants. The weighted average correlation between baseline overweight/obesity and increased depression at follow-up was also small (r=0.05), but significant. No significant moderators were found. The findings are consistent with a bi-directional model of depression and obesity, with small effect sizes.