Objective. Sex hormone status has been proposed as an important determ
inant of low bone mineral density (BMD) in both men and women. Our obj
ective was to study the relationship between sex hormones and BMD in p
atients with rheumatoid arthritis (RA), and how steroid therapy affect
s both. Methods. We studied 99 men with RA to assess their BMD and sex
hormone status. A comparative group of 68 age paired men was used. We
made comparative tests, linear correlations, and multiple regression
analysis. Results. We found significant reductions in lumbar BMD (p =
0.0005), femoral BMD (p < 0.0005), salivary testosterone (p = 0.01), a
ndrostenedione (p = 0.007), and dehydroepiandrosterone sulfate (DHEAS)
(p = 0.03) in patients with RA. In contrast, serum testosterone conce
ntrations were normal. Salivary testosterone showed correlation with f
emoral BMD (r = 0.36; p < 0.001). By multiple regression analysis, wei
ght, serum testosterone concentrations, and the cumulative dose of cor
ticosteroids were significant predictors of lumbar BMD (r = 0.41; p =
0.001). Weight, age, androstenedione concentrations, and the cumulativ
e dose of corticosteroids were the significant predictors of femoral B
MD (r = 0.79; p <0.0005). Conclusion. We confirm there is reduced BMD
in men with RA; corticosteroids contribute, but are not the only facto
r in the pathogenesis of low BMD in patients with RA; we found decreas
ed levels of androstenedione, DHEAS, and salivary testosterone in men
with RA; and salivary testosterone, as with free testosterone, is corr
elated with BMD in patients with RA, with lower levels contributing to
low BMD.