More Questions Than Answers About the Obesity Paradox

A new study published this week in Obesity raises more questions than it answers about the obesity paradox. This paradox had been seen in many studies: people with cardiovascular disease who have excess weight sometimes live longer than people with weight in the normal BMI range. In this new study, Andrew Stokes and Samuel Preston have found support for the possibility that the obesity paradox may be a reflection of reverse causation and a history of smoking in people with lower weights.

Using NHANES data from 1988 to 2010, they found that the paradox was not apparent if people who had ever had obesity were excluded from the control group of people with normal weight. The finding makes sense because people with normal weight and a history of obesity might have lost weight because of poor health. That would mean that lower prospects for survival were causing lower weight — instead of the other way around.

They also found that excluding people who had ever smoked seemingly eliminated a confounding factor and yielded a result that showed higher mortality in the group with overweight and obesity. The confounding effect of smoking makes sense because smoking helps people keep their weight down, but it causes illness and death.

In the end, this study simply raises questions and does not provide a definitive answer. A significant problem is that the analysis relied upon “recalled maximum weight” to identify people who had ever had obesity. Such self-reported data for weight is well-known to be unreliable and systematically biased. So from an analysis that blends undependable self-reported histories with measured weights, we cannot draw real conclusions.

And the obesity paradox remains a puzzle to be solved.

Click here to read the study.

Heart-Shaped Hole, photograph © Premasagar Rose / flickr

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.

October 2, 2015