Longitudinal patterns of depressive symptoms in midlife women

Abstract

Many people assume that women in midlife, approaching and completing menopause, will experience increased depression. Yet most studies show the best predictor of depression in midlife is a history of previous depression. The aims of this study were to describe and characterize longitudinal patterns of depressive symptoms in midlife women before, during, and after the transition to menopause and to explore relationships between these patterns and menopausal transition stage (MTS), history of abuse/assault, and general life events and stressors.Women were sampled from the Seattle Midlife Women's Health study (SMWHS), a descriptive longitudinal study of women who at entry to the study in 1990--92 were aged 35--55 years, resided in Seattle, spoke and read English, had at least some high school education, still had their uterus and at least one ovary, and had a menstrual period within the previous 12 months. Depressive symptoms were measured approximately annually by the Center for Epidemiologic Studies-Depression scale (CES-D), with total scores considered a continuous variable. For determination of longitudinal patterns, a sample of SMWHS participants with at least 6 (or 5 for hierarchical cluster analysis) annually repeated scores was used. Two cluster analyses and a phenomenological analysis were conducted on samples of 216, 249, and 205 women, respectively.The important parameters of longitudinal patterns were level or severity of depressive symptoms (non-depressed, low, moderate, and high), duration of symptoms at that level (yearly or slower change), range and fluctuation (e.g., change in CES-D score >15 yearly), and trend over time (stable, increasing, or decreasing). The single largest group was stable non-depressed at all time points. MTS and life events were not consistently related to CES-D score or longitudinal pattern. History of sexual abuse/assault was significantly related to a high chronic pattern of depressive symptoms (Pearson chi2 = 15.23; df 6; p = .019).Clinical implications were recommendations to improve detection of history of depression and abuse/assault, and anticipatory guidance for long-term plans of care, regardless of MTS.