The natural history of diabetic neuropathy and its risk factors are not well understood, apart from the recognition that prevalence increases with duration and, in many studies, degree of glycemia. The role of potential risk factors was therefore evaluated in a cross-sectional analysis from the baseline examination of the Pittsburgh Epidemiology of Diabetes Complications Study. We present results from the first 400 subjects seen at baseline examination. Neuropathy was determined by a trained internist with a standardized examination and was defined as the presence of at least two of three criteria: abnormal sensory or motor signs, symptoms consistent with neuropathy, and decreased tendon reflexes. The prevalence of neuropathy in this cohort was 34% (18%, 18-29 yr old, 58% greater than or equal to 30 yr old) with no difference by sex. By focusing on subjects greater than or equal to 18 yr old, all significant univariate variables (e.g., duration, glycosylated hemoglobin [HbA1]) were analyzed in 3 multiple logistic regression models: all subjects greater than or equal to 18 yr old and separating the same subjects into two groups based on age (18-29 and greater than or equal to 30 yr). Duration, HbA1, smoking status, and high-density lipoprotein cholesterol were found to be associated with neuropathy in the models for the greater than or equal to 18-yr-old group and the greater than or equal to 30-yr-old group. In the 18- to 29-yr-old group, duration, HbA1, and hypertension status were found to be significantly associated with neuropathy.(ABSTRACT TRUNCATED AT 250 WORDS)