Association of serum selenium with selected cardiovascular risk factors

The association of serum selenium (S-Se) with selected cardiovascular risk factors has been studied in 3404 (1520 men, 1884 women) of 4699 subjects aged 20-73 years, who underwent a comprehensive examination between April 1986 and December 1987 within the framework of the Di.S.Co. Research Project of the National Research Council. Mean S-Se concentrations were 1.163 (SD 0.170) μmol/liter, 1.156 (0.163) μmol/liter, and 1.171 (0.179) μmol/liter in the total group, and female and male subjects, respectively. The difference by sex was statistically significant (F = 6.97, P = 0.0083). In male subjects S-Se levels were inversely associated to age (simple correlation coefficient, r = −0.2135, P < 0.001), alcohol consumption (Alcohol, r = −0.0688, P < 0.01), smoking habit (Smoke, r = −0.0663, P < 0.01), body mass index (BMI, r = −0.0596, P < 0.02), and lognormal transformation of triglycerides (Ln-Trig, r = −0.0767, P < 0.005). In multiple regression analysis Age, Smoke, Ln-Trig, and Alcohol remained significantly and inversely related to S-Se; a significant direct association of S-Se to high-density lipoprotein cholesterol (HDL) and non-HDL cholesterol (non-HDL) was also pointed out. In female subjects, S-Se was directly related to HDL (r = 0.1436, P < 0.001) and non-HDL (r = 0.0967, P < 0.001). In multiple regression analysis S-Se was directly related to both HDL and non-HDL and an inverse significant association of S-Se to Age and Alcohol was evidenced. In multivariate regression models analyzing systolic (SEP), diastolic (DBP), and mean blood pressure (MBP) as dependent variables, S-Se was a weak significant positive predictor in male but not in female subjects.