Abstract
Objective. To evaluate the ability of body mass index (BMI) and fat location indices to predict clustering of risk factors for cardiovascular and metabolic disorders in children and adolescents. Methods. A total of 2 835 children in the 3rd (8.2–11.3 years) and 9th (14–17.3 years) grade were randomly selected from Denmark, Estonia and Portugal. BMI, waist-circumference (WC) and waist-to-height ratio (WHt) were evaluated. Children were considered to have clustering of risk factors if ≥3 risk factors were present: high levels of total cholesterol, triglyceride, glucose, insulin, blood pressure and low levels of aerobic fitness and HDL-C. The diagnostic accuracy for detecting children at risk was evaluated through receiver operating characteristics analyses, performed separately for boys and girls in the four age groups: 8.2–9.5; 9.5–11.3; 14.0–15.5 and 15.5–17.3 years. Cut-offs producing equal sensitivity and specificity and minimizing misclassifications were derived. Children were classified as overweight and obese based on BMI cut-offs producing equal sensitivity and specificity and minimizing misclassifications, respectively. Results. The diagnostic accuracy for identifying children at risk, measured by the area under the curve (AUC), of all anthropometric indices were significantly higher than what would be expected by chance (AUC >0.5). BMI cut-offs for overweight had similar sensitivity (58.8% to 75%) and specificity (60% to 71.2%) values, whereas specificity (94.4% to 99.7%) was markedly higher than sensitivity (9.3% to 52.6%) for obesity cut-offs. Conclusion. BMI, WC and WHt can be used to identify children and adolescents aged 8.2–17.3 years with a clustering of cardiovascular risk factors.
Acknowledgements
We acknowledge the members of the EYHS group for making their data available. The EYHS is supported by the following grants from Denmark: Danish Heart Foundation, Danish Medical Research Council, Health Foundation, Danish Council for Sports Research, Foundation of 17–12–1981, Foundation in Memory of Asta Florida Bolding neé Andersen, and Faculty of Health Sciences, University of Southern Denmark, TRYG-foundation; from Estonia: Estonian Science Foundation grant 3277 and 5209; and from Portugal: European Social Fund.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
Appendix 1: Performance of each anthropometric index for clustering of risk factors over all cut-offs according to sex and age groups. *Lower limit of 95% CI of AUC<0.50 (non-signifiant).
Appendix 2. Sensitivity analyses.