Design:  Cross-sectional epidemiological study of type 2 diabetes and heart disease risk factor prevalence among Native Hawaiians.

Subjects:  A total of 578 Native Hawaiians residing in two rural communities were examined between 1993 and 1996. Sample sizes in statistical analyses varied due to missing data and selection criteria based on ethnic ancestry.

Measurements:  Percentage of Hawaiian ancestry and non-Hawaiian ethnic admixture, assessed by self-report. Fasting and 2 h post glucose challenge plasma glucose levels. Anthropometric measures (height, weight, waist and hip circumferences). Self-report of diet and physical activity. Medical history.

Results:  Increased Hawaiian blood quantum was significantly associated with increased fasting glucose (P=0.0047), increased body mass index (BMI; P<0.0001), waist-to-hip ratio (WHR; P=0.0103), and age (P<0.0001), but not with leisure time physical activity and total dietary caloric intake. This association was attenuated after adjusting for BMI and WHR, but not by age-adjustment alone. However, when the effects of descent from other ethnic groups was examined in a subset of participants, full-Hawaiians had significantly higher fasting glucose concentrations (7.28 mmol/l) than part-Hawaiians after adjustments for age, gender, BMI and WHR. In contrast, part-Hawaiians of predominantly Asian ancestry had the highest 2 h glucose concentrations (7.62 mmol/l).

Conclusion:  These results suggest that ethnic admixture may be an important, but extremely complex, factor concerning the high prevalence of type 2 diabetes observed among this population. The complexity of this relationship may have obscured the relationship between ancestry and glucose tolerance in earlier observations of this population.