There is a need for increased knowledge about endogenous sex hormone levels and clinical outcomes of risk/benefit. Immunoassays have poor specificity to reliably measure low steroid concentrations in elderly. The objective of the study was to evaluate plasma steroid concentrations with regard to prevalent cardiovascular disease (CVD) in elderly, using mass spectrometry. The study was conducted at a university hospital research unit. Plasma samples were analyzed from 202 70-yr-olds as part of a large population-based study, Prospective Investigation of the Vasculature in Uppsala Seniors. Twenty-eight of these had prevalent CVD. Eleven steroids were quantified, using liquid chromatography-tandem mass spectrometry. Women with current/previous menopausal hormone therapy (n = 35) were excluded. Men without prevalent CVD had higher plasma 17beta-estradiol (E2), compared with women. Men with prevalent CVD, compared with those without, had lower 17-hydroxypregnenolone (17OHPregn), 17-hydroxyprogesterone, and higher estrone/androstenedione and E2/testosterone (T) (aromatase activity). Women with prevalent CVD had lower pregnenolone, 17OHPregn, and dehydroepiandrosterone (DHEA) but higher DHEA/17OHPregn, androstenedione/DHEA, E2/T, E2/estrone, and E2/SHBG. The aromatase index, E2/T, was higher for prevalent CVD in both sexes. Adjustment for statin use, smoking, and body mass index yielded additional significant differences in men, whereas some were lost in women. Logistic regression indicated strong associations between prevalent CVD and low 17OHPregn, adjusted odds ratio of 0.18, 95% confidence interval (0.06-0.61); P = 0.006, in women and low 17-hydroxyprogesterone, 0.45 (0.25-0.80); P = 0.007 in men, most likely caused by increased throughput (consumption) toward estrogen synthesis. Prevalent CVD was associated with indications of lower androgen precursors, increased aromatase activity, and higher estrogen levels in both sexes. Results might represent an endogenous response to a condition of developing atherosclerosis, rather than a causative relationship. Furthermore studies are needed.
Read full abstract