Clinical and biochemical predictors of extreme longevity would be useful in geriatric practice but have still not been clearly defined.
To identify the best nongenetic predictors of survival in centenarians, we examined 340 individuals aged 100+ years. A detailed questionnaire was completed, and comprehensive geriatric assessment and blood analyses were performed. Survival of study participants was checked annually over the period of 10 years.
In the univariate Cox proportional hazards model, a longer survival of centenarians was correlated with a higher adjusted Mini-Mental State Examination (MMSE(adj)) score (p < .000001), higher Activities of Daily Living (ADL) and adjusted Instrumental Activities of Daily Living (IADL(adj)) scores (p < .000001 and p = .00008, respectively), and younger age at the time of testing (p = .005). Blood pressure, lipid profile, and C-reactive protein and hemoglobin concentrations were not associated with survival. Multivariate analysis including age, sex, and the MMSE(adj) and ADL scores showed that both MMSE(adj) and ADL predicted survival (HR = 0.978 per each MMSE(adj) point, 95% CI: 0.964-0.993, p = .004; HR = 0.900 per each ADL point, 95% CI: 0.842-0.962, p = .002, respectively). In multivariate analysis with the ADL score substituted by the IADL(adj) score, the only predictor of survival was MMSE(adj) (HR = 0.973 per each MMSE(adj) point, 95% CI: 0.958-0.988, p = .0006).
Cognitive and functional performances are predictors of survival in centenarians.