Date of Award

Fall 1-8-2016

Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Matthew Magee, Ph.D.

Second Advisor

Lawrence Phillips, M.D.

Third Advisor

Mary Rhee, M.D.


Background: The Centers for Disease Control and Prevention estimates that 30 million Americans are living with diabetes and by 2050, 1 out of 3 U.S. adults could develop diabetes. Current screening and diagnostic methods for diabetes mellitus include the oral glucose tolerance test (OGTT), hemoglobin A1c (HbA1c), and fasting plasma glucose (FPG) levels, but each test when used alone will miss a portion of patients who have the disease. It has been suggested that the 50-g glucose challenge test (GCT) could also serve as a useful diabetes screening test particularly in light of its greater convenience compared to the FPG and OGTT. Given the expected increases in the prevalence of diabetes, it is particularly important to identify convenient and efficient screening tests which can effectively predict those individuals at highest risk for the development of diabetes.

Objective: To determine the predictive utility of the GCT in identifying incident diabetes over five years and to determine whether other important risk factors, including age, gender, race, body mass index, waist circumference and lipid levels, modify the predictive utility of the GCT.

Methods: We performed a prospective observational study at the Atlanta Veterans Affairs Medical Center (2009-2012). Eligible participants were those who did not have diabetes at baseline based on an OGTT. Data on incident diabetes was collected during 5 years of follow-up. Incident diabetes was defined by patient medical records and ICD-9 codes. Receiver operating characteristics were used to assess the predictive utility of the GCT, HbA1c, and components of the OGTT (fasting, 1-hour, and 2-hour measures). The GCT consisted of a 50 gram glucose load with plasma and capillary (finger stick) glucose measurements performed one hour later.

Results: Of 1384 eligible participants, 94% were male, 73.4% Black, and median age and body mass index was 56.0 years and 29.5 kg/m2, respectively. Among those with follow up within 5 years, 133 (9.6%) participants had a new diagnosis of diabetes. The area under the receiver operating characteristics curves (AROC) for the GCTplasma, GCTcap, HbA1c, and the grouped measures from the OGTT (fasting, 1-hour, and 2-hour) were 0.634, 0.643, 0.725, 0.657, 0.691 and 0.662, respectively. When comparing AROC values in predicting incident diabetes at five years, the GCTplasma performed as well as the GCTcap, 2hr-OGTT, and FPG (p>0.05), but had a significantly lower predictive utility than that of HbA1c and 1hr-OGTT (p

Conclusion: The GCT performed as well as the currently used 2hr-OGTT and FPG tests at predicting future incidence of diabetes. Given its convenience as a non-fasting, 1-hour test, the GCT may provide a practical and efficient approach in identifying those at risk of incident diabetes in a clinical setting.