DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE
Diagnostic performance of body mass index to identify excess
body fat in children with cerebral palsy
1 Center of Prevention and Rehabilitation, University of Cologne, Cologne; 2 Children’s and Adolescent’s Hospital, University of Cologne, Cologne, Germany.
3 Department of International, Health Maastricht University, School CAPHRI, Care and Public Health Research Institute, Maastricht, the Netherlands. 4 Cologne Center
for Musculoskeletal Biomechanics (CCMB), University of Cologne, Cologne, Germany.
Correspondence to Ibrahim Duran, Center of Prevention and Rehabilitation, University of Cologne, Lindenburger Allee 44, 50931 Cologne, Germany. E-mail: Ibrahim.firstname.lastname@example.org
This article is commented on by Stevenson on page 639 of this issue.
Accepted for publication 15th January
Published online 7th March 2018.
DAG German Society for Adiposity
DXA Dual-energy X-ray
WHO World Health Organization
WOF World Obesity Federation
To assess the diagnostic performance of body mass index (BMI) cut-off values according
to recommendations of the World Health Organization (WHO), the World Obesity Federation
(WOF), and the German Society for Adiposity (DAG) to identify excess body fat in children
with cerebral palsy (CP).
The present study was a monocentric retrospective analysis of prospectively
collected data among children and adolescents with CP participating in a rehabilitation
programme. Excess body fat was deﬁned as a body fat percentage above the 85th centile
assessed by dual-energy X-ray absorptiometry.
In total, 329 children (181 males, 148 females) with CP were eligible for analysis.
The mean age was 12 years 4 months (standard deviation 2y 9mo). The BMI cut-off values
for ‘overweight’ according to the WHO, WOF, and DAG showed the following sensitivities
and speciﬁcities for the prediction of excess body fat in our population: WHO: sensitivity
0.768 (95% conﬁdence interval [CI] 0.636–0.870), speciﬁcity 0.894 (95% CI 0.851–0.928); WOF:
sensitivity 0.696 (95% CI 0.559–0.812), speciﬁcity 0.934 (95% CI 0.898–0.960); DAG: sensitivity
0.411 (95% CI 0.281–0.550), speciﬁcity 0.993 (95% CI 0.974–0.999).
Body mass index showed high speciﬁcity, but low sensitivity in children
with CP. Thus, ‘normal-weight obese’ children with CP were overlooked, when assessing
excess body fat only using BMI.
The World Health Organization (WHO) deﬁnes over-
weight and obesity as abnormal or excessive fat accumula-
tion that may impair health.
Overweight and obesity
among children and adolescents are associated with higher
risk of adult obesity and are major risk factors for diseases
such as hypertension, early markers of cardiovascular dis-
ease, and insulin resistance.
Body mass index (BMI),
deﬁned as a person’s weight in kg divided by the square of
the person’s height in meters (kg/m
), is commonly used
to identify overweight and obesity in adults. To deﬁne
overweight and obesity in children, age-related reference
centiles are used.
Although the BMI is easy to apply and
epidemiological data showed an association between
increased BMI and cardiovascular events,
the BMI is not
able to distinguish between fat mass and lean body mass.
Thus, BMI tends to overestimate body fat in individuals
who are muscular and underestimates body fat in individu-
als with low lean body mass.
The prevalence of overweight or obesity (using BMI)
among children and adolescents in the USA was 31.8% in
2011 to 2012.
Studies indicated that the prevalence of
overweight in children with physical impairment and intel-
lectual disability may be higher than in typically develop-
A recent systematic review has shown that
BMI had high speciﬁcity but low sensitivity to detect obe-
sity and fails to identify more than 25% of otherwise typi-
cally developing children with excess body fat percentage.
Kuperminc et al. showed that BMI is a poor predictor to
detect excess body fat in children with moderate-to-severe
cerebral palsy (CP).
CP is a group of disorders caused by
permanent but not progressive damage of the developing
brain, and is the most common cause of physical impair-
ment in children (with an incidence of about two per 1000
Tools directly assessing fat mass, such as dual-energy
X-ray absorptiometry (DXA) or hydrostatic weighing, are
validated methods for measurement of obesity. However,
these methods are more invasive than BMI and their use is
limited by cost, availability, and the lack of widely accepted
cut-off values of normal body fat percentage in children.
Therefore, BMI is widely recommended by institutions
such as WHO, the World Obesity Federation (WOF), and
680 DOI: 10.1111/dmcn.13714 © 2018 Mac Keith Press