Should children with overweight or obesity be excluded from height references?

• Published in: Archives of disease in childhood 2015-11, Vol.100 (11), p.1044-1048
• Main Authors: Júlíusson, Pétur B, Brannsether, Bente, Kristiansen, Hege, Hoppenbrouwers, Karel, Bjerknes, Robert, Roelants, Mathieu
• Format: Article
• Language: English
• Subjects: Adolescent; Age; Analysis; Anthropometry - methods; Belgium - epidemiology; Body Composition; Body height; Body Height - physiology; Body Mass Index; Charts; Child; Child, Preschool; Childhood obesity; Children & youth; Cross-Sectional Studies; Female; Females; Girls; Growth - physiology; Humans; Male; Males; Methods; Norway - epidemiology; Norwegian; Obesity; Obesity - epidemiology; Obesity - physiopathology; Obesity in children; Overweight - epidemiology; Overweight - physiopathology; Puberty; Reference Values; Risk factors; Scientific Concepts; Statistical Analysis; Statistical Significance; Stature; Studies
• ISSN: 0003-9888; 1468-2044
• DOI: 10.1136/archdischild-2015-308537

Background and objectivesGrowth reference charts are usually based on measurements of children free from a medical condition that affects growth. However, samples collected during the past decades often contain a large proportion of overweight or obese children. Because obesity increases linear growth, the question arises to what extent the percentiles curves for length/height are affected by the presence of children with overweight or obesity.MethodsData from two cross-sectional samples of 2-year-old to 18-year-old children were analysed: 12 252 Belgian children, measured in 2002–2004, and 6159 Norwegian children, measured in 2003–2006. The LMS method was used to estimate height-for-age curves with and without children considered overweight or obese according to the International Obesity Task Force thresholds.ResultsThe prevalence of overweight (including obesity) and obesity was 13.0% and 2.8% in the Belgian and 13.8% and 2.3% in the Norwegian sample. Children were taller when overweight (+0.49 and 0.43 SD, in the Belgian and Norwegian sample, respectively) or obese (+0.73 and 0.72 SD in the Belgian and Norwegian sample, respectively). Effect sizes were smaller in younger and older children, which points to an advanced age of maturation as a possible cause. Excluding overweight and obese children had only a minor impact on the growth curves with largest difference in mean height SD scores −0.09 in the Belgian and −0.12 in the Norwegian sample with a corresponding increase of up to 0.5% and 1.2% in number of children >+2 SD.ConclusionsCurrent Belgian and Norwegian growth references for length/height were found to be largely unaffected by the current proportion of overweight and obese children. There is, therefore, no need for revised height charts that exclude overweight or obese children.