Advertisement
Letters
JAMA. 2005;294(5):552. doi: 10.1001/jama.294.5.552-b

Underweight, Overweight, Obesity, and Excess Deaths

  1. Eric L. Ding, BA
  1. eding@hsph.harvard.edu
    Department of Epidemiology
    Harvard School of Public Health
    Boston, Mass

Since this article does not have an abstract, we have provided the first 150 words of the full text.

To the Editor: The study by Dr Flegal and colleagues1 showed that those persons in the overweight category (BMI 25 to <30) had a lower number of excess deaths compared with those in the normal-weight category (BMI 18.5 to <25). However, their summary results were obtained from pooling 3 separate cohorts of statistically divergent results from very heterogeneous time periods, a factor that has major methodological implications in how these data should be statistically pooled.

The cohort-specific estimates were based on the populations of NHANES I, II, and III, ranging from 1971 to 1994. Due to heterogeneity of available medical treatments over more than 20 years, the implications of long-term survival from being overweight and obese may be different across the different decades. This heterogeneity suggests that the proper method of pooling the differential estimates into a summary statistic is a random-effects model.2-3 Although the analysis …

Related articles

« Previous | Next Article »Table of Contents

More in JAMA & Archives Journals