Moderate agreement between body mass index and measures of waist circumference in the identification of overweight among 5-year-old children; the ‘Be active, eat right’ study
1 Department of Public Health, Erasmus MC – University Medical Center Rotterdam, Rotterdam, the Netherlands
2 Youth Policy Department, Municipal Public Health Service Rotterdam area, Rotterdam, the Netherlands
3 Institute of Health Sciences, Faculty of Earth and Life Sciences, VU University Amsterdam, Amsterdam, the Netherlands
4 Department of Public and Occupational Health, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, the Netherlands
BMC Pediatrics 2013, 13:63 doi:10.1186/1471-2431-13-63Published: 23 April 2013
Body mass index (BMI) is a common indirect method to assess weight status among children. There is evidence that BMI data alone can underestimate overweight-related health risk and that waist circumference (WC) should also be measured. In this study we investigated the agreement between BMI and WC and BMI and the waist-height ratio (WHtR) when used to identify overweight among children.
This cross-sectional population-based study uses baseline data from 5-year-olds (n = 7703) collected by healthcare professionals for the ‘Be active, eat right’ study.
According to age-specific and sex-specific cut-off points for BMI (IOTF, 2000) and WC (Fredriks et al., 2005), the prevalence of overweight (obesity included) was 7.0% and 7.1% among boys, and 11.6% and 10.1% among girls, respectively. For the WHtR the 90th percentile was used as the cut-off point. Among boys, observed proportion of agreement between BMI and WC classification was 0.95, Cohen’s kappa 0.58 (95% CI; 0.53-0.63), and proportions of positive and negative agreement were 0.61 and 0.97, respectively. Observed proportion of agreement between BMI and WHtR classification was 0.92, Cohen’s kappa 0.46 (95% CI; 0.41-0.51), and proportions of positive and negative agreement were 0.51 and 0.95. Children identified as overweight according to WC were relatively tall, and children classified as overweight according to the WHtR only were relatively short (comparable results for girls).
There is moderate agreement between BMI and measures of WC on the presence of overweight among 5-year-olds. If BMI data and cut-offs continue to be used, then part of the group of children identified as overweight according to WC and the WHtR will be omitted. Follow-up of the children classified as overweight according to BMI only, WC only, and WHtR only, will give indications whether WC should be measured in addition to BMI or whether WC should only be measured in certain subgroups (e.g. relatively tall or short children) to identify and monitor overweight in children. This may improve early identification and prevention of overweight and overweight-related health problems in children.