journal articles
PREDICTORS OF 4-YEAR MORTALITY AMONG NONAGENARIANS AND CENTENARIANS IN DUJIANGYAN, CHINA
J.H. Flaherty, B. Dong, H. Wu, G. Liu, Q. Yuan, J. Deng, Y. Zhang, J. Wu, G. Zeng, X. Ren, J. Hu, W. Wu, T. K. Malmstrom
J Aging Res Clin Practice 2012;1(2):125-130
Objectives: To identify predictors of mortality in a long-lived Chinese population. Design: Four-year follow up mortality data of a previously performed cross-sectional observational study. Setting: Dujiangyan, China. Participants: Eight-hundred eight Han Chinese aged 90-108 years. Measurements: Trained researchers performed face-to-face interviews and physical and geriatric assessments to obtain information on sociodemographic factors, self-reported medical diseases, geriatric-specific conditions, number of hospitalizations and biomedical measurements (systolic/diastolic blood pressure, body mass index (BMI) and blood tests for albumin, fasting glucose, creatinine clearance, hemoglobin and lipid panel). Results: Of 808 participants, 424 (52.5%) died during the 4-year period. In univariate analyses, age, 3 out of 10 medical diseases (osteoarthritis, respiratory illnesses and diabetes), increasing number of medical diseases (comorbidities), 4 out of 8 geriatric-specific conditions (ADL impairment, hearing problems, cognitive impairment and weight loss), and two biomedical measurements (albumin <40.0 g/L; creatinine clearance <1200 mL/s) were significantly associated with mortality. In multivariate Cox regression analysis, none of the medical diseases that were significant in the univariate analyses, nor comorbidities, were associated with mortality. Three geriatric conditions were significant: ADL impairment [hazard ratio (HR) = 1.42, 95% confidence interval (CI) = 1.14-1.78, P=.002], cognitive impairment (HR = 1.51, 95% CI = 1.18-1.92, P=.001) and weight loss (HR = 1.41, 95% CI = 1.05-1.90, P=.022). Female gender (HR =.75, 95% CI =.59-.95, P=.018) also reached statistical significance. Two biomedical measures were significantly associated with mortality: albumin <40.0 g/L (HR=1.39, 95% CI=1.02-1.88) and CrCl <1200 mL/s (HR=1.41, 95% CI=1.06-1.88). Conclusion: Among a long-lived cohort of Chinese, several geriatric conditions (functional disability, cognitive impairment and weight loss) predict mortality. Comorbidities and specific medical diseases do not. Adding objective biomedical measures does not seem to add substantial information to the risk profile for mortality.