[Skip to Content]
[Skip to Content Landing]
April 9, 2003

Obesity and Years of Life Lost

Author Affiliations

Letters Section Editor: Stephen J. Lurie, MD, PhD, Senior Editor.

JAMA. 2003;289(14):1777. doi:10.1001/jama.289.14.1777-a

In Reply: Dr Beasley notes that our title implied demonstration of causation whereas, in fact, we found an association that may or may not indicate causation. He is correct and we regret our linguistic imprecision. Beasley further suggests that past research may have shown fitness to be a better predictor of mortality rate (MR) than BMI. He also suggests that there are potential causal interconnections among BMI, sedentariness, dietary intake, physiological variables, and MR. We are not certain that fitness has been shown to be a better predictor of MR than BMI or obesity, nor are we clear what it means to be a "better" predictor. A more critical question, in our view, is whether each predictor has independent causative influences on MR. Some data do suggest that BMI, fitness, and activity levels all have independent associations with MR.1 If independent associations of these variables imply independent effects on MR, then the fact that one of them is a predictor in no way obviates the importance of the others. Clearly more research will be needed to elucidate the complex and interactive relations among body weight, body composition, fitness and physical activity levels, physiological variables, and MR. As Fisher wrote, "In no case, however, can one judge whether or not it is profitable to eliminate a certain variate unless we know, or are willing to assume, a qualitative scheme of causation."2

First Page Preview View Large
First page PDF preview
First page PDF preview