The Anger-Depression Connection: Between-Persons and Within-Person Associations from Late Adolescence to Midlife

Six waves of data from 944 Canadian high school seniors surveyed for 25 years (age 18 to 43) were analyzed to compare three models of the temporal associations between expressed anger and depressive symptoms (anger [right arrow] depression; depression [right arrow] anger; anger [left right arrow] depression) and a between-persons differences only model ([up arrow] anger = [up arrow] depression). Perceived availability of social support was examined as a mediator of the anger-depression association. Random intercept cross-lagged panel analyses supported the between-persons differences only model, controlling for sociodemographic variables, within-time covariances between construct residuals, and autoregressive stabilities: Individuals who were higher on expressed anger tended to be higher on depressive symptoms and, within individuals, neither fluctuations in anger nor depression appeared to influence the other across intervals ranging from one year to 18 years. Although social support did not mediate an anger-depression connection, intraindividual increases in expressed anger predicted future intraindividual increases in perceived support (i.e., age 18 anger to age 19 support, and age 20 anger to age 22 support). A similar path from age 19 depressive symptoms to age 20 perceived support was observed. There were also some gender differences in the paths. The results point to stable interindividual differences in negative emotionality in place by age 18 and maintained up to age 43.