RESUMEN
Non-insulin-dependent diabetes mellitus (NIDDM) is a chronic disabling disease, that shortens length of life and implies a high burden for a community. Its prevalence goes from 0 per cent in Papua, New Guinea to 34 per cent in Pima Indians. There are very few prevalence studies in Mexico, and the strength of association of the known risk factors with the occurrence of the disease is not established. A prevalence cross sectional study was carried out with users of a first level medical care unit, with a meter measure of capillary glucose levels. Those with a previous diagnosis of diabetes or whose capillary glucose level were 200 mg or over were considered diabetics. Hyperglycemia was when the levels were recorded between 121 and 199 mg. The crude prevalence of NIDDM was 5.6 per cent (CI 95% 4.5-6.8), With almost no sex difference. Hyperglycemia prevalence was 2.9 per cent (CI 95% 2.0-3.7). Age was the main risk factor for the development of NIDDM. Those between 40 and 59 years showed a high risk (OR 10.8; CI 95% 5.4-22.0; p < 0.0001), and it was greater for the 60 years or elder (OR 20.6; CI 95% 9.8-44.1; p < 0.0001). Weight was also an important risk factor, with a 2.7 fold greater risk for obese persons (CI 95% 1.6-4.6; p < 0.0001). Other, risk factors were familiar history of diabetes (OR 1.5; CI 95% 0.9-2.3; p = 0.096), and overcrowding (OR 1.9; CI 95% 1.0-3.4; p = 0.03). In order to analyze independently each variable, a logistic regression model was applied, and a similar strength of association was observed for the crude model, but for obesity whose effect was modified by age. When only new cases were analyzed in the former model, the association with obesity was maintained. There is a need to develop prevalence studies of NIDDM in Mexico and to measure the strength of association with the known and the not jet well known risk factors of this disease in order to establish health policies according to the Mexican reality.