High body mass index does not predict mortality in older people: analysis of the Longitudinal Study of Aging

J Am Geriatr Soc. 2001 Jul;49(7):968-79. doi: 10.1046/j.1532-5415.2001.49189.x.


Objective: To determine the excess mortality associated with obesity (defined by body mass index (BMI)) in older people, with and without adjustment for other risk factors associated with mortality and for demographic factors.

Design: Retrospective cohort analysis of the Longitudinal Study of Aging (LSOA).

Setting: Nationally representative sample of community-dwelling older people.

Participants: Seven thousand five hundred and twenty-seven participants age 70 and older in 1984.

Measurements: We used Cox regression to calculate proportional hazards ratios for mortality over 96 months. We tested the hypothesis that increased BMI (top 15%) increased mortality rates in older people.

Results: Death occurred in 38% of the cohort: 54% of the thin (lowest 10% of the population, BMI <19.4 kg/m(2)), 33% of the obese (highest 15%, BMI> 28.5 kg/m(2)), and 37% of the remaining participants (normal) died. Adjustment for demographic factors, health services utilization, and functional status still demonstrated reduced mortality in obese older people (hazard ratio 0.86, 95% confidence interval (CI) = 0.77-0.97) compared with normal. After adjustment, thin older people remained more likely to die (hazard ratio 1.46, 95% CI = 1.30-1.64) than normal older people. Sensitivity analyses for income, mortality during the first two years of follow-up, and medical comorbidities did not substantively alter the conclusions.

Conclusion: Obesity may be protective compared with thinness or normal weight in older community-dwelling Americans.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Activities of Daily Living
  • Age Distribution
  • Aged
  • Body Mass Index*
  • Causality
  • Cause of Death*
  • Comorbidity
  • Death Certificates
  • Educational Status
  • Female
  • Geriatric Assessment
  • Health Services / statistics & numerical data
  • Humans
  • Income / statistics & numerical data
  • Longitudinal Studies
  • Male
  • Multivariate Analysis
  • Obesity / mortality*
  • Predictive Value of Tests
  • Proportional Hazards Models
  • Retrospective Studies
  • Sensitivity and Specificity
  • Sex Distribution
  • Survival Analysis
  • United States / epidemiology