|| Checking for direct PDF access through Ovid
There is conflicting evidence of a link between circulating insulin concentrations and blood pressure. The aim of this study, therefore, was to investigate this relationship in a representative population sample of different ethnic groups.A cross-sectional study was performed, using a random cluster sample of 5080 adults aged 25-74 yeas from the Indian Ocean island of Mauritius.Glucose tolerance was assessed with a 75-g oral glucose tolerance test. Hypertension and glucose tolerance were classified using World Health Organization criteria. Subjects known to have diabetes before the survey were excluded from these analyses, leaving 3104 Asian Indian, 1256 African-origin Creole and 384 Chinese subjects eligible for this study.There were statistically significant linear correlations between fasting and 2-h insulin and mean blood pressure in all sex-ethnic subgroups, except for Chinese men. However, after controlling for age and the body mass index, partial correlations were much reduced and remained significant only for fasting insulin in Indian men (r=0.07, P<0.01) and women (r=0.09, P<0.001) and 2-h insulin in Chinese women (r=0.15, P<0.05). After controlling for age and body mass index, mean fasting and 2-h insulin concentrations were not significantly different between hypertensive and normotensive subjects in any of 36 possible subgroups defined by ethnic group, gender or glucose tolerance (normal, impaired glucose tolerance or newly diagnosed diabetes). In both men and women with normal glucose tolerance (ethnic groups combined) there was no evidence of increasing systolic or diastolic blood pressure across quartiles of either fasting or 2-h serum insulin within any quartile of the body mass index. By contrast, at any given level of serum insulin there was a clear association between an increasing body mass index and increasing blood pressure, suggesting strongly that insulin is not the final pathway by which obesity influences blood pressure.This study of three ethnic groups does not support the hypothesis that the circulating insulin concentration is an independent determinant of blood pressure in populations.