Employment status and depressive symptoms in middle-aged women: a longitudinal investigation.

OBJECTIVES: This study investigated the psychological effects of paid employment and nonemployment, cross-sectionally and longitudinally, in 541 middle-aged women.
METHODS: Women were seen for an extensive baseline evaluation and were reevaluated in an identical manner 3 years later. They were divided into four "synthetic" cohorts, based upon employment status at these two assessments.
RESULTS: At baseline, nonemployed women reported higher levels of depressive symptoms than employed women, with nonemployed women with less education, low support from family and friends, or low marital satisfaction the most symptomatic. Symptoms decreased in women who were nonemployed initially but employed at the later exam; in all other groups, symptoms increased. Unlike the cross-sectional findings, however, the longitudinal findings showed that women who were nonemployed at baseline and had low levels of education, social support, or marital satisfaction did not benefit more from subsequent employment than did their counterparts who had high levels of these factors.
CONCLUSIONS: New paid work may have a positive effect on mood for middle-aged women. However, given the contradictions between our cross-sectional and longitudinal results, caution should be exercised in drawing conclusions from cross-sectional studies that likely confound employment status, symptoms, and moderating variables.