Date of Award

Summer 8-12-2014

Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Rodney Lyn

Second Advisor

Richard Rothenberg


BACKGROUND: Type 2 diabetes and cardiovascular disease (CVD) are among the leading causes of death in the United States. The Metabolic Syndrome, which comprises a cluster of cardiometabolic risk factors, puts individuals at increased risk for these diseases. It is therefore important that people with Metabolic Syndrome, at high risk for CVD and type 2 diabetes, are identified and treated. Since it may not often be practical to obtain the laboratory measures necessary for diagnosing the Metabolic Syndrome, simple anthropometric measures are a useful way of quickly identifying individuals at increased risk for the Metabolic Syndrome.

OBJECTIVE: The purpose of this thesis is to evaluate the utility of three of the most commonly used anthropometric measures – Body Mass Index (BMI), Waist Circumference (WC), and Waist-to-Height Ratio (WC) – for classifying individuals with and without the Metabolic Syndrome and its component risk factors in the United States. Using Receiver Operating Characteristic (ROC) curve analysis and Area Under the Curve (AUC) statistics, this thesis will assess the utility of each body measurement and compare it to BMI.

METHODS: A large, multi-ethnic, nationally representative sample from the National Health and Nutrition Examination Survey (NHANES) 2007-2010 was used for this analysis. The study sample was restricted to adults aged 20-65 with complete information on height, weight, waist circumference, blood pressure, HDL cholesterol, fasting glucose, and triglycerides (n=3,769). In order to compare the utility of different anthropometric measures for classification, weighted ROC curves were constructed for each anthropometric measure-outcome combination and AUC statistics were compared. AUC statistics were calculated by approximating the definite integral of the ROC curves with the trapezoidal rule. Variances for AUC statistics and differences in AUC statistics were estimated with jackknife repeated replication. Analyses were completed for the entire sample and separately for non-Hispanic whites, non-Hispanic blacks, and Mexican Americans.

RESULTS: For the entire sample, WC (AUC=0.752) did a better job than BMI (AUC=0.728) at classifying individuals with and without the Metabolic Syndrome (p

CONCLUSION: Waist circumference should be considered, especially over BMI, for risk stratification in clinical settings and research. Further research should attempt to identify optimum waist circumference cut points for use in the US population.