Artery elasticity indices and testosterone- longitudinal study in vara- skövde cohort

Objective: To determine the associations between testosterone, sex hormone-binding globulin (SHBG) and arterial elasticity in men in a longitudinal observational study. Design and method: 2816 randomly selected individuals were examined (Men = 1400) between 2002 and 2005. After excluding participants with known hypertension and without data for arterial elasticity and testosterone, we include 1202 men in cross-sectional analyses. After a mean follow-up of 9.7 ± 1.4 years, a representative sample of 1327 original participants was re-examined. In the longitudinal analysis, 473 men were included. Body height, weight, heart rate and blood pressure were measured. Total testosterone (TT) and SHBG were analyzed using RIA. Bioavailable testosterone (BioT) is calculated by the Vermeulen coefficient. Arterial elasticity was measured based on pulse contour analysis of arterial waveform using the HDI/Pulse WaveTM CR2000 and C1 and C2 were obtained. Linear regression models were used to investigate these associations in three models: in model 1 adjusting for age, in model 2 we added heart rate, systolic blood pressure and in model 3 we added BMI. Multivariable general linear models were used to investigate the differences in C1 and C2 based on testosterone deficiency. Results: Mean age at baseline was 46.2±10.9 years. In cross-sectional analysis, an increase in TT and BioT was associated with an increase of C1 (beta = 0.08, CI: 0.02-0,14, p = 0.01 resp beta = 0.14, CI: 0.001-0.272, p = 0.05) in model 1, but not in the other models. In model 1 participants with testosterone deficiency had significantly lower C1 (C1 in TD: 17.6±5, C1 in the normal level of TT: 18.2±5, delta C1: 0.6, CI: 0.06- 1.1; p = 0.05), but not in the other models. Other associations between TT, BioT, SHBG, TD and C1 respective C2 were not significant in the models. Conclusions: We found weak associations between artery elasticity indices and testosterone levels. Surprisingly these associations were not significant after adjustments for pulse and blood pressure, even though the sample was large enough to reveal clinically significant associations.