Cross-sectional and longitudinal associations between serum testosterone concentrations and hypertension: Results from the Fangchenggang Area Male Health and Examination Survey in China.

BACKGROUND:Low testosterone concentrations have been suggested as a risk factor for hypertension, but their contribution to the development of hypertension is not well studied. We carried out a cohort study based on the results of an earlier cross-sectional investigation. We established the association between testosterone concentrations and hypertension. METHOD:Data on 2427 healthy male subjects, aged from 17 to 88 y, were collected for the cross-sectional study. A representative sample of 853 individuals who did not suffer from hypertension at baseline was followed up for 4 y. Differences between the tertiles groups of sex hormones were analyzed, relative risks (RR) were estimated using binary logistic regression model. RESULTS:In the cross-sectional analysis, the serum total testosterone (TT), free testosterone (FT), and bioavailable testosterone (BT) concentrations of the hypertensive population were lower than those of the non-hypertensive population. Binary logistic regression analysis showed that TT, BT, and FT were inversely associated with hypertension. Moreover, decreasing odds ratio (OR) was observed from the lowest tertile group to the highest tertile group. After multivariate adjustment, the correlation between FT, BT, and hypertension was attenuated. Statistically significant differences remained only in the middle tertile group of TT and in the highest tertile group of TT, FT, and BT. In the longitudinal analysis, the 4-y incidence of hypertension was higher in participants with lower TT than in those with higher TT. Subjects in the middle and highest tertile groups of TT had an RR of 0.35 (0.22-0.57) and 0.30 (0.18-0.50), respectively (P for trend <0.001). After further adjustments, these associations still remained statistically significant. CONCLUSIONS:Serum TT, FT, and BT concentrations were inversely associated with blood pressure in man, and TT independent of age and body mass index (BMI) influences the development of hypertension. Furthermore, TT can be employed as a risk marker for hypertension in the identification of high-risk individuals.