Total homocysteine and estrogen status indicators in the Third National Health and Nutrition Examination Survey.

The possibility that estrogen status modulates total homocysteine concentration, a risk factor for vascular occlusion, was examined in a representative sample of the US population, the Third National Health and Nutrition Examination Survey (phase 2), 1991-1994. The geometric mean serum total homocysteine concentration was compared among population subgroups differing on inferred estrogen status, after adjusting for potential confounding by age, race-ethnicity, smoking, and the serum concentration of creatinine, folate, and vitamin B-12. Premenopausal women aged 17-54 years had a lower mean serum total homocysteine concentration (8.1 micromol/liter, 95% confidence interval (CI): 7.9, 8.2) than men in the same age range (8.9 micromol/liter, 95% CI: 8.6, 9.3). In the age range 17-44 years, pregnant women (6.0 micromol/liter, 95% CI: 5.4, 6.8), but not oral contraceptive users (7.9 micromol/liter, 95% CI: 7.6, 8.2), had a lower mean serum total homocysteine concentration than nonpregnant, non-oral-contraceptive-using women (8.1 micromol/liter, 95% CI: 7.9, 8.2). The mean serum total homocysteine concentration of estrogen-using women aged > or = 55 years (9.5 micromol/liter, 95% CI: 8.9, 10.1) was significantly decreased relative to nonestrogen users (10.7 micromol/liter, 95% CI: 10.3, 11.1) and men (10.4 micromol/liter, 95% CI: 9.8, 11.0) in the same age range. These findings suggest that higher estrogen status is associated with a decreased mean serum total homocysteine concentration, independent of nutritional status and muscle mass, and that estrogen may explain the previously reported male-female difference in total homocysteine concentration.