INTRODUCTION: The relevance of low testosterone concentrations for incident coronary heart disease (CHD) and mortality has been discussed in various studies. Here, we evaluate the predictive value of low baseline testosterone levels in a large population-based cohort.
METHODS: We measured the serum levels of testosterone in 7671 subjects (3710 male, 3961 female) of the population-based FINRISK97 study.
RESULTS: The median follow-up (FU) was 13.8 years. During the FU, a total of 779 deaths from any cause, and 395 incident CHD events were recorded. The age-adjusted baseline testosterone levels were similar in subjects suffering incident events during FU and those without incident events during FU (men: 15.80 vs. 17.01 nmol L-1 ; P = 0.69, women: 1.14 vs. 1.15 nmol L-1 ; P = 0.92). Weak correlations of testosterone levels were found with smoking (R = 0.09; P < 0.001), HDL cholesterol levels (R = 0.22, P < 0.001), systolic blood pressure (R = -0.05; P = 0.011), BMI (R = -0.23; P < 0.001) and waist-hip-ratio (R = -0.21; P < 0.001) in men, and with eGFR (R = -0.05; P = 0.009) in women. Kaplan-Meier analyses did not reveal a positive association of testosterone levels with incident CHD or mortality. Accordingly, also in Cox regression analyses, testosterone levels were not predictive for incident CHD or mortality - neither in men (HR 1.02 [95%CI: 0.70-1.51]; P = 0.79 for lowest versus highest quarter regarding CHD and HR 1.06 [95%CI: 0.80-1.39]; P = 0.67 regarding mortality), nor in women (HR 1.13 [95%CI: 0.69-1.85]; P = 0.56 for lowest versus highest quarter regarding CHD and HR 0.99 [95%CI: 0.71-1.39]; P = 0.80 regarding mortality).
CONCLUSIONS: Low levels of testosterone are not predictive regarding future CHD or mortality - neither in men, nor in women.