Exercise adherence and 10-year mortality in chronically ill older adults.

OBJECTIVES: To compare mortality of adherents and nonadherents of an exercise program. DESIGN: Prospective intervention study. SETTING: Supervised geriatric fitness program called Gerofit. PARTICIPANTS: One hundred thirty-five adults aged 65 and older who enrolled in Gerofit between January 1, 1990, and November 30, 1999. All participants had a baseline medical screen and exercise test. They were classified as adherent (n = 70) if they participated in Gerofit for more than 47 sessions or nonadherent (n = 65) if they did not complete 47 sessions within the first 6-month period. INTERVENTION: Program participation was voluntary and consisted of aerobic, strength, flexibility, and balance exercises. The program met three times week for 90 minutes. MEASUREMENTS: All-cause mortality. RESULTS: Twenty-six deaths occurred within the 10-year follow-up period. Using proportional hazards, time to death was not related to adherence group. However, in multivariate analyses controlling for age, sex, race, baseline risk/health status, history of heart disease, cancer, diabetes mellitus, and baseline smoking status, there was significant group-by-time interaction (P =.004), indicating a crossover in mortality risk. The initial survival benefit observed in nonadherers changed over time, resulting in a long-term protective survival effect on mortality for the adherent group (hazard rate = 0.75, 95% confidence interval = 0.61-0.91 for the interaction term). CONCLUSIONS: Older adults with chronic diseases experience a long-term beneficial mortality effect from participation in exercise programs. Physicians should strongly encourage their patients, including those with comorbidities, to maintain a regular exercise program.

Duke Scholars

Author Carl F. Pieper Biostatistics & Bioinformatics, Division of Biostatistics

Author Miriam C. Morey Medicine, Geriatrics

Author Robert Joseph Sullivan Jr. Family Medicine and Community Health, Family Medicine