Impact of depressive symptoms on hospitalization risk in community-dwelling elder persons

OBJECTIVE: To determine whether depressive symptoms in older adults are ass ociated with an increased risk for hospitalization. DESIGN: A 6 month cohort study. SETTING: Five counties in the northern Piedmont of North Carolina from the Duke University site of the Established Populations for Epidemiological Stu dies of the Elderly project. PARTICIPANTS: The sample included 3486 community-dwelling adults, aged 65 a nd older. MEASUREMENTS: Crude risk ratios for the effect of depressive symptoms on 6 month risk for hospitalization were calculated, followed by a multivariable analysis controlling for demographics and health status. RESULTS: Three hundred participants were hospitalized during the 6 month fo llow-up period. The crude risk ratio for the effect of depressive symptoms on hospitalization was 1.95 (95% CI = 1.47-2.58). Subgroup analysis showed significant positive risk ratios for men aged 65 to 74 and greater than or equal to 75, and women aged 65 to 74. After a multivariable analysis, howev er, these associations remained significant only among men greater than or equal to 75 (RR = 3.43; 95% CI = 1.33-8.86). CONCLUSIONS: Depressive symptoms were independently associated with a more than threefold increased risk for hospitalization among men aged greater th an or equal to 75. This result reflects differences in the effects of depre ssive symptoms across age and gender groups, and emphasizes that symptoms o f depression influence overall health and medical utilization among, at the very least, the oldest subset of men.