Transforming percentage of median body mass index (%mBMI) scores into corresponding BMI z- scores yields discrepancies for age and sex: Implications for pediatric eating disorder researchers and clinicians

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Anorexia nervosa (AN) and atypical AN are distinguished by the presence or absence of a significantly low body weight. In children and adolescents, the population-based age- and sex-adjusted weight characteristics, BMI centile, BMI z-score (BMIz), and percentage of median BMI (%mBMI) are used to diagnose underweight. Whereas BMI centile and BMIz are interconvertible, %mBMI is not. We extend previous analyses to illustrate a) age- and sex-dependent discrepancies between %mBMI and BMIz and b) the non-linear relationship between BMIz and absolute body weight. Based on Centers for Disease Control growth charts, BMIz corresponding to the 65, 75, and 85%mBMI were determined for males and females aged 10 to 18 years. Body weights (in kg) corresponding to BMIz ranging from 0 to -7 (based on average heights) were also calculated and compared. Throughout the age span, BMIz corresponding to 65%mBMI ranged from − 5.0 to -6.9 in males and − 4.5 to -5.2 in females; at 75%mBMI, from − 2.9 to -3.6 in males and − 2.6 to -2.9 in females; and at 85%mBMI, from − 1.4 to -1.7 in males and − 1.3 to -1.4 in females. Body weight increments (in kg) per 1 BMIz were non-linear and dependent on reference BMIz. Transformations between %mBMI and BMIz are not linear, and these characteristics cannot be used interchangeably. Use of both BMIz and %mBMI in eating disorder research limits meta-analyses. Field consensus for universal weight-related assessment in adolescents is necessary.

Article activity feed