Purpose: This study explores the relationship between obesity and mortality, focusing on obesity-related diseases (ORDs), using a nationally representative sample of the U.S. civilian noninstitutionalized population from the National Health and Nutrition Examination Survey (NHANES) III public-use datasets. ORDs in our study include coronary heart disease, hypertension, type 2 diabetes mellitus, hypercholesterolemia, and stroke.
Method: Obesity has been linked to increased mortality because individuals are at a higher risk of developing obesity-related comorbidities. We used information provided in NHANES III to investigate how the sampled population's individual characteristics, including gender, race, body mass index (BMI), and status of obesity-related diseases influenced their mortality and life expectancy. Deviating from the Cox proportional hazard model adopted in the literature, our study employed the mixed proportional hazards model, which takes into account the unobserved heterogeneity that affects, not only the outcome, but also the observed confounders. Failing to account for it results in biased estimation, and, consequently, misinterpretation of the results.
Result: Our findings showed that people developing ORDs increase their hazard rate by 44%, on average, compared to the people without ORDs of the same gender, race, and BMI level. Being a male, a black, or having a higher BMI level is also associated with higher mortality. Our sensitivity analyses also provided similar results, showing that the findings in the baseline case are robust to different parametric specifications. When broken down to different BMI categories (Figure 1), we found that the hazard ratio of the people with ORDs to those without ORDs is the highest (2.0) among the class II obesity group (35 kg/m2 ≤ BMI < 40 kg/m2), indicating that ORDs increase the hazard rate by 100%. The life years lost due to ORDs are on average 4.65 years. Blacks lose nine more life months due to ORDs than whites in the class II obesity group.
Conclusion: This study provided evidence suggesting that ORDs increase mortality by at least 50% for people with BMI greater than 18.5 kg/m2, especially for those with BMI between 35 and 39.5 kg/m2, after controlling for individuals' observed and unobserved characteristics. Life years lost due to ORDs range between 2.5 to 7 years.
Figure 1 Life years lost due to ORDs and hazard ratios at all levels of BMI
See more of: The 33rd Annual Meeting of the Society for Medical Decision Making