Evaluation of the association of autoantibodies with mortality in the very elderly: a cohort study.

Abstract

To evaluate whether autoantibodies in the absence of rheumatic diseases increase the risk of mortality among very elderly subjects who are otherwise in good functional condition. Autoantibodies were measured in 1987 in 156 elderly nursing home residents (median age 84 yr) who were followed subsequently over 14.6 yr. Eleven subjects had anticardiolipin antibodies, 30 had rheumatoid factor and 19 had antibodies to single-stranded DNA (ssDNA). Other autoantibodies were more rare. During follow-up, 144 subjects died. Adjusting for age as a time-dependent covariate, the hazard ratio for death was 0.71 [95% confidence interval (CI) 0.38-1.32] for anticardiolipin antibodies, 0.93 (95% CI 0.60-1.41) for rheumatoid factor, 1.08 (95% CI 0.65-1.79) for antibodies to ssDNA, and 0.99 (95% CI, 0.70-1.41) for any autoantibody. Hazard ratios were similar when adjusted also for sex and clinical conditions. Our results exclude the possibility that the autoantibodies evaluated increase substantially the risk of death among very elderly subjects in good functional condition.