BMI Fails to Identify More Than One-Fourth of Obese Children

We’re not the first to have pointed out the shortcomings of the Body Mass Index (BMI), a useful but imperfect measurement of obesity. The authors of a new meta-analysis published online in Pediatric Obesity on June 24, 2014, have found that the BMI relies on body weight regardless of body composition, and fails to distinguish between lean and fat mass. They credit BMI with being “an invaluable epidemiologic tool to assess population-level risk and track national trends in obesity,” but point out that it “has limitations in diagnosing obesity at the individual level.”

In our third publication, Prediabetes: Beating the Odds of Getting Type 2 Diabetes, we point out that waist circumference is a far better indication of obesity. Obviously, it’s impractical to use for mass studies.

You’re probably familiar with the BMI. When you step on the scale in your doctor’s office, the nurse asks you your height and jots down a number on your chart. That number is your BMI—your weight in kilograms divided by your height in meters squared.

BMI Used in Children: High Specificity, Low Sensitivity

The current meta-analysis included 37 studies that evaluated more than 53,500 patients whose average ages ranged from 4 to 18 years. The authors concluded that when used in children, BMI “has a high specificity but low sensitivity to detect adiposity and fails to identify over a quarter of children with excess body fat percentage.

(Specificity pertains to how well a test or procedure identifies only the patients who actually have a given condition, such as obesity. Sensitivity has to do with how good a test is at identifying patients with the condition. So the authors were saying that if the IBM indicates that a child is obese, the diagnosis is almost certainly correct. But the IBM fails to identify a lot of children [more than 25%] who are in fact obese.)

So, what is the solution? A clinician who suspects obesity—despite a normal BMI—could use secondary tools, such as skin-fold measurements or imaging studies. The authors see a need for additional studies to develop reference values for skin-fold thickness and for ADP and DXA (imaging studies), and to determine how well these test can—independently or together—predict whether a child is likely to develop conditions associated with obesity.