Low plasma testosterone and elevated carotid intima-media thickness: importance of low-grade inflammation in elderly men.

CONTEXT AND OBJECTIVE: An inverse correlation between plasma testosterone levels and carotid intima-media thickness (IMT) has been reported in men. We investigated whether this association could be mediated or modified by traditional cardiovascular risk factors as well as inflammatory status.

METHODS: In the Three-City population-based cohort study, 354 men aged 65 and over had available baseline data on hormones levels and carotid ultrasonography. Plasma concentrations of testosterone (total and bioavailable), estradiol and sex hormone-binding globulin (SHBG), together with cardiovascular risk factors were measured. IMT in plaque-free site and atherosclerotic plaques in the extracranial carotid arteries were determined using a standardized protocol. Multiple linear regression models were used to analyze this association and interaction study.

RESULTS: Analyses with and without adjustment for cardiovascular risk factors showed that carotid IMT was inversely and significantly correlated with total and bioavailable testosterone levels but not with SHBG and estradiol levels. This association depended on C-reactive protein (CRP) levels (p for interaction <0.05). Among men with low-grade inflammation (CRP ≥2 mg/L), mean IMT was higher in subjects with bioavailable testosterone ≤ 3.2 ng/mL than in those with bioavailable testosterone > 3.2 ng/mL (0.76 mm and 0.70 mm respectively, p < 0.01). By contrast, among men with CRP ≤ 2 mg/L, mean IMT was similar in both groups (0.72 mm and 0.71 mm respectively, p = 0.77). Similar results were found for total testosterone although not significant. No association was found between plasma hormones levels and atherosclerotic plaques.

CONCLUSION: In elderly men, low plasma testosterone is associated with elevated carotid intima-media thickness only in those with low-grade inflammation. Traditional risk factors have no mediator role.