Abstract
BACKGROUND AND OBJECTIVES: Recent studies demonstrated an association between depressive affect and higher mortality risk in incident hemodialysis patients. This study sought to determine whether an association also exists with hospitalization risk. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: All 8776 adult incident hemodialysis patients with Medical Outcomes Study Short Form 36 survey results treated in Fresenius Medical Care North America facilities in 2006 were followed for 1 year from the date of survey, and all hospitalization events lasting >24 hours were tracked. A depressive affect score was derived from responses to two Medical Outcomes Study Short Form 36 questions ("down in the dumps" and "downhearted and blue"). A high depressive affect score corresponded with an average response of "some of the time" or more frequent occurrence. Cox and Poisson models were constructed to determine associations of depressive affect scores with risk for time to first hospitalization and risk for hospitalization events, as well as total days spent in the hospital, respectively. RESULTS: Incident patients with high depressive affect score made up 41% of the cohort and had a median (interquartile range) hospitalization event rate of one (0, 3) and 4 (0, 15) total hospital days; the values for patients with low depressive affect scores were one (0, 2) event and 2 (0, 11) days, respectively. For high-scoring patients, the adjusted hazard ratio for first hospitalization was 1.12 (1.04, 1.20). When multiple hospital events were considered, the adjusted risk ratio was 1.13 (1.02, 1.25) and the corresponding risk ratio for total hospital days was 1.20 (1.07, 1.35). High depressive affect score was generally associated with lower physical and mental component scores, but these covariates were adjusted for in the models. CONCLUSIONS: Depressive affect in incident hemodialysis patients was associated with higher risk of hospitalization and more hospital days. Future studies are needed to investigate the effect of therapeutic interventions to address depressive affect in this high-risk population.