Older adults and depressive symptoms

This study examined the quantitative and qualitative characteristics of depressive symptoms in influencing diagnosis and treatment for depressive disorders in older adults. The study population was sampled from adults aged 65 and older enrolled in a social health maintenance organization. Subjects were followed for up to four years following a baseline survey of DSM-III-R depressive symptoms, completed by 582 subjects. Filled prescriptions for antidepressants were assessed using pharmacy claims, and visits to mental health providers obtained from medical claims.; Chi-square statistics tested association between diagnosis and treatment. In all analyses, depressive symptoms were entered either as a count of symptoms, or as indicator variables for each specific depressive symptom. T-tests tested associations between number of depressive symptoms and diagnosis (or treatment); chi-squares/Fisher's exact tests tested associations between specific depressive symptoms and diagnosis (or treatment).; To assess differences in treatment process for patients with and without a diagnosis, bivariate probit models with selection were estimated on treatment with a selection on diagnosis. Logistic models estimated association between depressive symptoms and diagnosis, and associations between treatment, diagnosis and depressive symptoms. Multivariate methods included logistic regression with random effects, generalized estimating equations with independent and AR(1) covariance structures, and weighting to account for stratified sampling structure.; Diagnosis was significantly associated with treatment; however, other covariates influencing treatment did not differ for patients with and without a diagnosis. Number of depressive symptoms was associated with increased odds of diagnosis of at least 50% for each one symptom increase. Fatigue was consistently associated with odds of diagnosis almost 3 times higher than the odds for patients without fatigue. Number of depressive symptoms was not consistently associated with treatment, either for the whole population or among subjects without a diagnosis. Odds of treatment were 50% lower for subjects with sleep problems compared to subjects without the symptom, even after adjusting for diagnosis.; Results indicating associations of specific symptoms with diagnosis and treatment hint at more effective targeting mechanisms for screening older adults for depression that may provide a significant opportunity to improve diagnosis and treatment for depressive disorders.