Abstract
Depression is well known to share a negative cross-sectional relationship with personality constructs defined by positive emotion (positive affect, extraversion, behavioral activation). These Positive Emotionality (PE) constructs have been proposed to represent stable temperamental risk factors for depression, not merely current mood state. These constructs have also been proposed to increase risk specifically for depression, relative to anxiety. We performed a meta-analysis of longitudinal studies to examine the relationship of PE to depression (59 effect sizes) and anxiety (26 effect sizes). In cross-sectional analyses, PE constructs were negatively associated with depression (r = -.34) and anxiety (r = -.24). PE constructs also prospectively predicted depression (r = -.26) and anxiety (r = -.19). These relationships remained statistically significant, but were markedly attenuated, when baseline levels of depression (β = -.08) and anxiety (β = -.06) were controlled. Moreover, depression and anxiety were equally strong predictors of subsequent changes in PE (β = -.07 and -.09, respectively). These findings are consistent with theoretical accounts of low PE as a temperamental vulnerability for depression, but suggest that the prospective relationship of PE to depression may be weaker and less specific than previously assumed. (PsycINFO Database Record