Methods: Data were collected from six elementary schools in Argentina between April and August 2007. ... obesity (OB) in children is one of the most alarming.
Original Article
DIABETES TECHNOLOGY & THERAPEUTICS Volume 11, Number 9, 2009 ª Mary Ann Liebert, Inc. DOI: 10.1089=dia.2009.0026
Comparison of Different Anthropometric Indices for Identifying Insulin Resistance in Schoolchildren Valeria Hirschler, M.D.,1 Amalia Ruiz, M.D.,2 Teresa Romero, M.D.,2 Ricardo Dalamon, M.D.,2 and Claudia Molinari, M.S.3
Abstract
Objectives: Anthropometric indices have been associated with insulin resistance in children. This study (1) determined the association between insulin resistance and anthropometric indices, including body mass index (BMI), waist circumference (WC), WC=height, weight=(sitting height)2, and WC=sitting height, and (2) compared the abilities of these five indices to identify children with insulin resistance. Methods: Data were collected from six elementary schools in Argentina between April and August 2007. Anthropometric data and Tanner staging were obtained. Fasting serum concentrations of glucose, lipids, and insulin were measured. Results: Six hundred twenty-five children (318 boys) between 6 and 14 years old were examined. The mean age of the children was 9.6 2.0 years. Ninety-six (15.4%) of the children were obese, 91 (14.6%) were overweight, and 438 (70.1%) were normal weight using Centers for Disease Control and Prevention norms. Sixty percent, 23.0%, 14.0%, and 3.0% were Tanner stage I, II, III, and IV, respectively. The areas under the receiver operator characteristic curves were as follows: WC ¼ 0.78 0.021 (95% confidence interval [CI] 0.74–0.82), BMI ¼ 0.77 0.022 (95% CI 0.73–0.82), weight=(sitting height)2 ¼ 0.76 0.022 (95% CI 0.72–0.81), WC=height ¼ 0.67 0.027 (95% CI 0.62–0.72), and WC=sitting height ¼ 0.67 0.27 (95% CI 0.62–0.72), indicating that BMI, WC, and weight=(sitting height)2 were acceptable predictors for insulin resistance, whereas WC=height and WC=sitting height were fair predictors as the areas under the curve were 90th percentile was between four and five times higher in the 4th versus the 1st quartile (P < 0.001). Only 16% (70 of 438) of the normal weight children were insulin-resistant (HOMA-IR 4th quartile), whereas 46% (86 of 187) of the children who were OW or obese (P < 0.001) were insulin-resistant.
HIRSCHLER ET AL. WC=height, weight=(sitting height)2, and WC=sitting height with HOMA-IR >3rd quartile as the dichotomous variable. The area under the ROC curve was significantly different from 0.5 in BMI, WC, WC=height, weight=(sitting height)2, and WC=sitting height and ranged from 0.77 to 0.67: BMI ¼ 0.77 0.022 (95% CI 0.73–0.82), WC ¼ 0.78 0.021 (95% CI 0.74–0.82), weight=(sitting height)2 ¼ 0.76 0.022 (95% CI 0.72–0.81), WC=height ¼ 0.67 0.027 (95% CI 0.62– 0.72), and WC=sitting height ¼ 0.67 0.27 (95% CI 0.62–0.72) (Fig. 2), indicating that BMI, WC, and weight=(sitting height)2 were acceptable predictors for insulin resistance, whereas WC=height and WC=sitting height were fair predictors as the areas under the curve were 3rd quartile in children.
ANTHROPOMETRIC INDICES–INSULIN RESISTANCE
619
Table 3. Optimal Anthropometric Measures
WC (cm) BMI (kg=m2) WC=height WC=(sitting height)2 Weight=(sitting height)2 (kg=m2)
Area under the curve (95% CI)
Threshold
0.782 (0.74–0.82) 0.77 (0.73–0.82)
66 cm 18.4 kg=m2
0.67 (0.62–0.72) 0.67 (0.62–0.72) 0.76 (0.72–0.81)
0.46 88.7 66 kg=m2
Thresholds for optimal sensitivity and specificity varied in different measures, ranging from the 50th to the 60th percentile. Sensitivity and specificity for all groups ranged from 0.58 to 0.73. The WC=height ratio cutoff point was 0.46. Discussion The association between insulin resistance and morbidity justifies looking for various markers of adiposity in all children. Regression analysis showed that the five anthropometric measures we studied were associated with insulin resistance. Most studies have focused on adolescents, whereas in our study nearly 60% of the subjects were prepubertal (Tanner stage I). To our knowledge only a few studies have examined the association between sitting height as a component of anthropometric measures and insulin resistance in children. The present study does not show any significant difference in the prediction of insulin resistance between sitting height and total height. This study also shows that there was little difference in the abilities of BMI, WC, and weight=(sitting height)2 to identify children with insulin resistance, even if WC showed slightly stronger associations. In contrast, the areas under the ROC curves for WC=height and WC=sitting height were fair (
