Reprinted from Eating Disorders Review
March/April Volume 24, Number 2
©2013 Gürze Books
Q. I see quite a few adolescents with eating disorders in my practice and usually try to assess patients in relation to their expected body weights. I understand that different methods are used to determine expected body weight, including both weight for stature and the body mass index (BMI, kg/m2). Can you recommend which of these measures to use? (RL, Chicago)
A. Your question is interesting and has been the subject of a recent study conducted by adolescent medicine specialists at Stanford. This group wanted to compare the sensitivity, specificity and positive predictive value of the two methods you mention, weight for stature and BMI, for detecting adolescents who weighed <75% of expected body weights. Utilizing National Center for Health Statistics data sets, they calculated the expected body weights for 12,047 individual adolescents aged 12 to 19 years, using both methods. Norms for expected body weight using weight for stature was calculated as the median weight in kilograms by sex, age (in half-year increments), and height group in centimeters. Expected body weight using the BMI was defined as median BMI for age and sex obtained from data used to draw the 2000 Center for Disease Control BMI growth charts.
The authors found that the two methods for measuring expected body weights are not equivalent. For girls and boys, estimates of expected body weights were 3.5% and 3.45% higher, respectively, when using the weight for stature compared with the BMI method. Among adolescent girls, 65% had higher expected body weights using the weight for stature compared to the BMI method. Thus, BMI may underestimate the degree of malnutrition compared to the weight for stature method (or, conversely, the weight for stature method may overestimate the degree of malnutrition compared to BMI). Absolute differences in expected body weights were most pronounced at the extremes of height. (Pediatrics. 2012 Dec; 130(6):e1607-13).
However, until future studies are conducted to determine if these differences are actually meaningful with respect to outcomes, there’s no guidance to suggest that one works better than the other. At least for now, you get to pick the one you like.