Changes in Prevalence of Depressive Symptoms in Alameda County

Abstract

Trends in depression are examined using data from the Alameda County Study, a three-wave, prospective community survey covering an 18-year period (1965-1974-1983). Age-period-cohort effects in rates of depressive symptoms were analyzed using logistic regression procedures to adjust for attrition due to loss-to-follow-up and nonresponse. The results indicate the presence of all three effects. There was a marked period effect, with rates in 1974 significantly higher than in either 1965 or 1983. There also was a definite cohort effect, with older cohorts exhibiting higher rates of depressive symptoms than younger cohorts. These findings do not provide support for an "age of melancholia" nor for an increasing cohort effect in succeeding generations. These prospective data on depressive symptoms are contrasted with those from studies that use cross-sectional data to estimate temporal trends in clinical depression, employing retrospective reports of lifetime prevalence and discussing possible reasons for the disparate findings.