A longitudinal study of the prevalence of depressive symptomatology in elderly widowed and married women.

The expected duration and magnitude of elevations in depressive symptomatology following bereavement have not been fully characterized. This study describes the natural history of changes in depressive symptomatology after widowhood, using the Center for Epidemiologic Studies Depression Scale. We studied 1144 elderly married women who completed a baseline interview; 136 women who were subsequently widowed, and a subset of 409 still-married women, selected as controls, were reinterviewed at 1, 6, 12, 18, and 24 months after bereavement. Widows were more depressed than their married controls at every interview, including baseline. About 10% of married women had high scores (> 15) at each interview. One month after bereavement, the proportion of widows with high scores rose to 58%; this proportion declined but remained elevated at 6 months. By 12 months, the proportion with high scores was comparable with prebereavement levels. Only about 40% of the elevation in widows' scores at 12 months is potentially attributable to bereavement.