Reproducibility of Glucose Tolerance in 101 Nondiabetic Women

W ITH THE increasing use of the oral glu- cose tolerance test for diabetes screening purposes, due in part to automated techniques, the reproducibility of present oral glucose tolerance test screening methods has been questioned.The present method of testing each subject once, and then giving a glucose tolerance test to those whose screening level appears abnormal, often yields a normal GTT on the retest, even when the persons tested have characteristics associated with diabetes, such as obesity or a family history of the disease.This paper concerns i-tself primarily with defining oral glucose tolerance test reproducibility by examining a select population of nondia- betic women.The study, conducted on female prisoners between July and December 1964, is a followup of a previous study of male prisoners (1).Both studies were concerned with repro- ducibility of test results, and so far as possible, were based upon the same experimental design. Description of StudyThe study population was composed of a group of female prisoners who volunteered for the project.Subjects were not accepted for the project if they were expected to leave the prison within 6 months, were known to have diabetes, Dr. McDonald is chief, and Mr. Burnham and Mr. Lewis are statisticians, Diabetes and Arthritis Control Program, Health Services and Mental Health Administration, Public Health Service.or if they had other chronic conditions which might affect glucose tolerance.