The Protective Effect of High Body Mass on Mortality Risk. Obesity Paradox or Data Artifact?

Ryan K. Masters, University of Colorado, Boulder
Sammy Zahran, Colorado State University

An “obesity paradox” is often reported in mortality analyses, wherein high body mass is found to confer a survival advantage at older ages. Some researchers have cautioned that obesity paradoxes should be met with skepticism because age-specific estimates of the obesity-mortality association are likely biased by (1) reverse causality and/or (2) age-related selection processes. We test the intuitions behind these claims by analyzing the age-specific associations between high body mass and mortality risk in two U.S. samples, NHANES 1988-1994, 1999-2004 and NHIS 1986-2004. We examined all-cause mortality risk by body mass index (BMI) in both data sets, BMI 10 years prior to survey in NHANES data, and waist-to-thigh ratio (WTR) in NHANES data. Cause-specific mortality risk (circulatory diseases and accidents) was also assessed by BMI in NHIS data. Results from all analyses indicate that “obesity paradoxes” are likely data artifacts that reflect biased estimates from age-related selection processes.