Prevalence and disease associations of certain autoantibodies in elderly patients.

OBJECTIVE

To determine the prevalence and association with various diseases of certain autoantibodies among elderly patients, in order to challenge the hypothesis that these autoantibodies are elevated generally in these patients as a result of immunosenescence.

DESIGN

Prospective prevalence study.

PATIENTS

A total of 399 elderly patients: 63 aging successfully (without chronic illness), 301 with a variety of chronic general illnesses (frail elderly) and 35 with a clinical diagnosis of rheumatoid arthritis. These were compared with 250 healthy adult blood donors.

INTERVENTIONS

Measurement of autoantibodies to rheumatoid factor, antinuclear antibody, double-stranded (native) DNA (nDNA), extractable nuclear antigens and anticardiolipin antibodies.

OUTCOME MEASURES

Prevalence of these autoantibodies and correlation with disease states.

RESULTS

Antibodies to rheumatoid factor and antinuclear antibody were significantly more prevalent in the elderly patients with chronic illness or rheumatoid arthritis but were not disease-specific. The prevalence of nDNA and extractable nuclear antigens was not increased in either the healthy or frail elderly groups. Anticardiolipin antibodies were significantly more prevalent in the frail elderly group when compared with normal controls and the healthy elderly group. The prevalence of anticardiolipin antibodies correlated with clinical features of cerebrovascular disease, in particular multi-infarct dementia and stroke, but not with Alzheimer's disease.

CONCLUSIONS

The prevalence of the autoantibodies measured was not elevated in healthy elderly subjects, and autoantibodies such as nDNA and extractable nuclear antigens are specific to disease states in all groups of elderly patients. Anticardiolipin antibodies correlate with cerebrovascular events. Therefore, the clinical significance of autoantibodies in elderly patients is related more to global health status than to the effects of aging.