Major Depressive Episodes and Employment Earnings Trajectories over the Following Decade among Working-aged Canadian Men and Women

Abstract

Objective: Quantify the association between experiencing a major depressive episode (MDE) and employment earnings over the following decade among working-aged Canadian men and women.

Methods: Ten national Canadian Community Health Survey (CCHS 2003-2014) cycles were linked to administrative tax records (1997-2016) for individuals 18-54 years old with information on past year MDE defined by the Composite International Diagnostic Interview Short Form (n=85,155) measured at baseline. Subsequent earnings were measured annually two to ten years post-MDE. MDE cases were matched with adult controls without MDE using 1:1 greedy nearest-neighbour without replacement propensity score matching for women and men.

Results: Using random-effects longitudinal modelling, among women (n=6,974) in the matched cohort (average initial earnings ~$52,119/annum, 2016 Canadian real dollars), experiencing an MDE was associated with average earnings of $4,473 less in the year the MDE was reported (95%CI=-$3,215,-$5,731) and an additional $363 less/year over the following decade (95%CI=-$132,-$594). Among men (n=3,620, average initial earnings ~$76,110/annum), an MDE was associated with an initial reduction in earnings of $5,023 (95%CI=-$2,453,-$7,593) followed by an earnings decline starting with $730/year, increasing to ~$1,810/year at ten years post-MDE.

Limitations: Residual confounding by measures excluded from the propensity score. Only one MDE measurement limited exploring earnings trajectories of those with multiple episodes.

Conclusions: Consistently lower earnings were seen for ten years after experiencing one MDE. The magnitude of earnings difference immediately after the MDE was similar for men and women. Findings reveal the critical period surrounding an MDE where effective clinical treatment and labor policy may help buffer longstanding earnings loss.

Keywords: earnings attainment; employment earnings; linked data; major depressive episodes; matched cohort; propensity score.