A population study of ethnic variations in the angiotensin-converting enzyme I/D polymorphism: relationships with gender, hypertension and impaired glucose metabolism

Background The presence of the deletion allele of the angiotensin-converting enzyme (ACE) I/D polymorphism is associated with an excess risk of vascular disease and diabetic nephropathy, Objective To examine the importance of this polymorphism as a determinant of hypertension and impaired glucose metabolism in a population-based study of three ethnic groups and assess the potential modifying effect of gender. Design population-based cross-sectional study in South London. The population-based sample of 1577 men and women, age 40-59 years, was obtained from stratified random sampling of general practice lists where 25% of the residents were born outside the UK. The ACE I/D polymorphism was determined for 1366 individuals (86.6%): 462 whites, 462 of African descent and 442 of South Asian origin. Results The genotype frequency within each ethnic group was in Hardy-Weinberg equilibrium. The frequencies were similar in whites and those of African descent (II, ID, DD: 18.4%, 49.6%, 32.0% for whites and 18.4%, 50.5%, 30.9% for those of African descent), but there was a much higher frequency of the II genotype in those of South Asian origin (39.8%, 41.8%, 18.3%; chi(2) = 77.6; P < 0.0001). There was no association between the I/D polymorphism and impaired glucose metabolism in any ethnic group. There were also no significant associations between the I/D polymorphism and hypertension in whites and in those of South Asian origin. This contrasts with a highly significant association between the D allele and hypertension in women of African descent (OR = 2.54; 95% CI 1.38-4.65; P = 0.003) but not in men of African descent (0.79; 0.36-1.72) (test far differences between sexes P = 0.023). Conclusions These observations provide estimates of the frequency distribution of the ACE I/D polymorphism in whites, in people of African descent and in people of South Asian origin. Moreover, these results highlight the potential importance of gender-dependent interactions between genetic background and expression of hypertensive phenotype. J Hypertens 1999, 17:657-664 (C) Lippincott Williams & Wilkins.