Goel Ruchika, Misra Anoop, Kondal Dimple, Pandey Ravindra M, Vikram Naval K, Wasir Jasjeet S, Dhingra Vibha, Luthra Kalpana
Department of Pediatrics, Children's Hospital of Pittsburgh of UPMC, Pittsburgh, PA, USA.
Clin Endocrinol (Oxf). 2009 May;70(5):717-24. doi: 10.1111/j.1365-2265.2008.03409.x. Epub 2008 Sep 5.
Biochemical measures for assessment of insulin resistance are not cost-effective in resource-constrained developing countries. Using classification and regression tree (CART) and multivariate logistic regression, we aimed to develop simple predictive decision models based on routine clinical and biochemical parameters to predict insulin resistance in apparently healthy Asian Indian adolescents.
Community based cross-sectional study.
Data of apparently healthy 793 adolescents (aged 14-19 years) were used for analysis. WHO's multistage cluster sampling design was used for data collection.
Homeostasis Model of Assessment value > 75th centile was used as cut-off for defining the main outcome variable insulin resistance. CART was used to develop the decision tree models and multivariate logistic regression used to develop the clinical prediction score.
Three classification trees and an equation for prediction score were developed and internally validated. The three decision trees were termed as CART I, CART II and CART III, respectively. CART I based on anthropometric parameters alone has sensitivity 88.2%, specificity 50.1% and area under receiver operating characteristic curve (aROC) 77.8%. CART II based on anthropometric and routine biochemical parameters has sensitivity 94.5%, specificity 38.3% and aROC 73.6%. CART III based on all anthropometric, biochemical and clinical parameters together has sensitivity 70.7%, specificity 79.2% and aROC 77.4%. Prediction score for insulin resistance = 1 x (waist circumference) + 1.1 x (percentage body fat) + 1.6 x (triceps skin-fold thickness) - 1.9 x (gender). A score cut-off of > 0 (using values marked for each) was a marker of insulin resistance in the study population (sensitivity 82.4%, specificity 56.7%, and aROC 73.4%).
These simple and cost-effective classification rules may be used to predict insulin resistance and implement population based preventive interventions in Asian Indian adolescents.
在资源有限的发展中国家，用于评估胰岛素抵抗的生化检测方法性价比不高。我们利用分类回归树（CART）和多变量逻辑回归，旨在基于常规临床和生化参数开发简单的预测决策模型，以预测表面健康的亚洲印度青少年的胰岛素抵抗情况。
基于社区的横断面研究。
对793名表面健康的青少年（年龄在14至19岁之间）的数据进行分析。数据收集采用世界卫生组织的多阶段整群抽样设计。
采用稳态模型评估值>第75百分位数作为界定主要结局变量胰岛素抵抗的临界值。使用CART构建决策树模型，并使用多变量逻辑回归构建临床预测评分。
开发了三个分类树和一个预测评分方程，并进行了内部验证。这三个决策树分别称为CART I、CART II和CART III。仅基于人体测量参数的CART I的灵敏度为88.2%，特异度为50.1%，受试者工作特征曲线下面积（aROC）为77.8%。基于人体测量和常规生化参数的CART II的灵敏度为94.5%，特异度为38.3%，aROC为73.6%。基于所有人体测量、生化和临床参数的CART III的灵敏度为70.7%，特异度为79.2%，aROC为77.4%。胰岛素抵抗预测评分 = 1×（腰围） + 1.1×（体脂百分比） + 1.6×（肱三头肌皮褶厚度） - 1.9×（性别）。评分临界值>0（使用为每个参数标注的值）是研究人群中胰岛素抵抗的一个指标（灵敏度82.4%，特异度56.7%，aROC 73.4%）。
这些简单且性价比高的分类规则可用于预测亚洲印度青少年的胰岛素抵抗，并实施基于人群的预防性干预措施。
