Resolving the obesity paradox

It's not news that Body Mass Index (BMI) is not a perfect predictor of health or longevity. Indeed, we see at least one article a week reporting the demise of BMI's clinical relevance with a gleefulness usually reserved for reports of celebrities embarrassing themselves in some way. The fact is that way back in 2013 we reported on a study that found that those with a BMI in the overweight range were actually 8% less likely to die from any cause than those with a normal BMI. This is what's known as "the obesity paradox" - when being overweight is associated with a lower risk of death from all causes.

The problem with Body Mass Index is that it is an indirect measure of body fat. It's cheaper, quicker, and requires far less training to do a mathematical calculation than to measure someone's body fat with a fair degree of accuracy. The second problem is that it was invented to be used on a large scale: Ancel Keys, who coined the term "Body Mass Index" in the early 70's, found it useful for large populations, but not so for individuals (he himself was fairly short but muscular, so he had a higher BMI than what would be considered "normal").

The good news is that recently a team of researchers in Canada were able to look at the relationship between Body Mass Index, actual body fat, and all-cause mortality (Ann Int Med 2016: doi:10.7326/M15-1181). They made use of population health data maintained by the Manitoba Centre for Health Policy of the University of Manitoba, and focused on those patients (about 49,500 women and 5,000 men) who had undergone Bone Mineral Density tests, which are the same types of tests used to measure body fat. The authors were then able to compare both the BMI and actual body fat measurements of those who died during the followup period with those who did not.

After taking into account the presence of such conditions as chronic obstructive pulmonary disease, diabetes, kidney disease, and congestive heart failure, the authors found that those women with a low BMI (less than 22.5) were 44% more likely to die of any cause than those with a BMI between 25 and 27.5 (usually considered slightly overweight). On the other hand, those with a percentage of body fat in the highest quintile (more than 38%) were 19% more likely to die of any cause. Results were similar for men, with low BMI and high body fat being associated with greater risk of mortality.

What this means for you

The authors note that the vast majority of their subjects were white, middle-aged, and female, so they suggest caution in generalizing these results to other ethnicities or age groups. They do point out that their results suggest that BMI might be a better indicator of lean body mass rather than body fat, so it "may be an inappropriate surrogate for adiposity." As we said in that 2013 Health & Nutrition Bite, we doctors deal in percentages: higher Body Mass Index still means an increased risk of chronic diseases such as diabetes and heart disease. Your doctor should be looking not just at your Body Mass Index, but at you as a whole person, including your lab test scores, diet, physical activity, and personal and family history.

First posted: March 9, 2016