Clinical characteristics of type II diabetic subjects consuming high versus low carbohydrate diets in Mexico City and San Antonio, Texas

OBJECTIVE - To compare the clinical status of type II diabetic subjects identified in two population-based surveys, one performed in Mexico City, Mexico and the other in San Antonio, Texas. RESEARCH DESIGN AND METHODS - In a low income area of Mexico City, 3,517 age-eligible (35-64 years of age) individuals were randomly selected of whom 3,319 were interviewed at home and 2,198 were examined in a clinic (response rates 62.5%). In San Antonio, 2,357 similarly aged low-income Mexican Americans were randomly selected of whom 2,076 were interviewed at home and 1,511 were examined (response rate 64.1%). Oral glucose tolerance tests were performed at both sites and diabetes was diagnosed according to the World Health Organization (WHO) criteria. In Mexico City, 288 type II diabetic individuals were identified, and 255 were identified in San Antonio. The following variables were measured: height, weight, subscapular and triceps skinfolds, waist-to-hip circumference ratios (WHR), systolic and diastolic blood pressure (random 0 sphygmomanometer), fasting and 2-h postglucose load glucose and insulin concentrations, and fasting total-cholesterol, HDL-cholesterol, and triglyceride (TG) levels. A food frequency questionnaire was used to estimate total calories and the percentage of calories derived from protein, fat, and carbohydrate. Only type II diabetic patients were included in the analyses. Age-adjustment was performed by analysis of covariance for continuous variables and by the Mantel-Haenszel procedure for discrete variables. RESULTS - The mean age, the percentage newly diagnosed cases, and the percentage of males were similar in both sites. The percentage of diabetic patients treated with oral agents was significantly higher in Mexico City (56.9 vs. 72.7% in San Antonio and Mexico City, respectively, P < 0.001), whereas the percentage treated with insulin was significantly higher in San Antonio (18.8 vs. 2.1% for San Antonio and Mexico City, respectively, P < 0.001). A significant difference was observed in the percentage of calories derived from carbohydrate (61.7-63.2 vs. 47.1-47.5% for Mexico City and San Antonio, respectively, P < 0.001) and fat (18.4-20.0 and 30.1-33.0% for Mexico City and San Antonio, respectively, P < 0.001). Body mass index (BMI) was higher in San Antonio (27.6-30.4 vs. 30.2-32.9% for Mexico City and San Antonio respectively, P < 0.05). Total serum cholesterol was similar at both sites. HDL cholesterol, however, was lower in Mexico City, both in newly and in previously diagnosed patients (30.5-35.8 vs. 39.6-43.3 mg/dl in Mexico City and San Antonio, respectively, P < 0.001). TG levels were higher in Mexico City patients (187-249 vs. 167-179 mg/dl in Mexico City and San Antonio, respectively, P < 0.001). The association between diabetes and the anthropometric and metabolic variables was similar in Mexico City and San Antonio with the following exceptions: Diabetes in Mexico City was associated with less of an elevation in BMI, WHR, and fasting insulin concentration and less of a reduction in the 2-h postoral glucose load insulin concentration compared with diabetes in San Antonio. In addition, although diabetes was associated with a lower HDL in San Antonio subjects, no association appeared between diabetes and HDL in Mexico City subjects. CONCLUSIONS- Diabetic subjects in Mexico City were more likely to be treated with oral agents and less likely to be treated with insulin compared with San Antonio patients. Previously diagnosed diabetic subjects in San Antonio had higher BMIs than diabetic subjects in Mexico City. Diabetic subjects in Mexico City ate less fat but more carbohydrate than those in San Antonio. TG levels were higher and HDL-cholesterol levels were lower in Mexico City diabetic subjects compared with those in San Antonio. San Antonio diabetic subjects had lower HDL levels than nondiabetic subjects but, in Mexico City, HDL levels were similar in diabetic subjects and nondiabetic subjects. Postchallenge insulin levels appear to be less suppressed in Mexico City than in San Antonio diabetic subjects.