Percent Body Fat Seems to Beat BMI for Predicting Mortality
After the strong and subjective opinions we keep hearing about the uselessness of BMI, it is refreshing to see a careful analysis of objective data. A new study in Annals of Family Medicine suggests measuring percent body fat through bioimpedance is superior for predicting mortality linked to excess adiposity.
Bioimpedance is the method seca and its competitors use in the precision body composition analysis devices they sell to serious obesity medicine professionals. It’s also the tool that our smart little bathroom scale uses to give us a guess at our own body fat percentage. We grant, of course, that our cheap little scale is no match for seca precision.
15-Year Mortality Risk in 4,252 Adults
For their analysis, Arch Mainous and colleagues used data from NHANES in a sample of 4,252 adults. For all these people, they had baseline data on BMI and waist circumference, in addition to percent body fat derived from bioimpedance. They looked for a relationship between each of these measures and 15-year all-cause mortality.
They found that both waist circumference and percent body fat predicted 15-year mortality. BMI did not. Senior study author Frank Orlando explains the bottom line:
“The medical community has been aware that BMI has some limitations as a measure of body composition and disease risk. It is an indirect measure of body fat percentage.
“Our study shows that a direct measure of body fat that can be done easily, practically, and inexpensively in a doctor’s office solves the problems of BMI.”
Practical Issues
Although waist circumference performed well in this analysis, the authors point out that it has “significant issues” in routine clinical use, owing to problems with measurement error and reliability. It might seem superficially simple. But in real life, it isn’t.
On the other hand, bioimpedance for body composition analysis can be almost foolproof.
The authors of this study seem think their data should “lead to a change” in how clinicians assess the health risks of adiposity. Their thought is reasonable. But when it comes to obesity, clinicians can be stubborn. BMI is so entrenched that we suppose retirements may be more effective in changing practice patterns than compelling data from this and other studies.
Still, this is absurd. BMI is supposed to be a surrogate for health risk. But it’s a lousy one. In contrast, bioimpedance works pretty well. How can this be hard to figure out?
Click here for the study in Annals of Family Medicine, here, here, and here for further reporting.
A Stroll Through the Graveyard, photograph by Ted Kyle / ConscienHealth
Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.
August 3, 2025
