Serum testosterone levels and clinical outcomes in male hemodialysis patients

Background: Studies linking low serum testosterone concentration to adverse clinical outcomes in hemodialysis patients have been relatively small. We investigated the role of testosterone in adverse outcomes and quality of life in an incident cohort of male Canadian hemodialysis patients.

Study design: A prospectively designed multicenter observational study using data from the Canadian Kidney Disease Cohort Study (CKDCS).

Setting & participants: Male patients initiating hemodialysis therapy since February 14, 2005, in 3 Canadian centers serving ethnically diverse populations were studied (N = 623).

Predictor: Serum testosterone levels using the International Society of Andrology, International Society for the Study of the Aging Male, and European Association of Urology cutoffs (low, <231 ng/dL; borderline, 231-346 ng/dL; normal, >346 ng/dL).

Outcomes: All-cause mortality, fatal and nonfatal cardiovascular (CV) events, and Health Utility Index (HUI)-assessed health-related quality of life.

Measurements: Participants completed a structured interview on demographics and medical history and an HUI questionnaire (version 3). Routine laboratory test results captured into the study database, and serum testosterone measured within 3 months after initiation of the baseline hemodialysis session.

Results: During a median follow-up of 20 (range, 1-81) months, 166 (27%) died and 98 (20%) had a CV event. Mean serum testosterone level was 234.1 ± 146.1 (SD) ng/dL. Higher serum testosterone levels were associated with significantly decreased unadjusted risk of death (HR per 10-ng/dL increase, 0.58; 95% CI, 0.37-0.90). There was a statistically significant trend for higher all-cause mortality with low serum testosterone levels in adjusted analyses (P < 0.001). Higher levels of log-transformed testosterone were associated with significantly higher HUI scores (P for trend <0.001), and low levels of serum testosterone were associated significantly with lower HUI scores (P for trend <0.001). Although there was a significant trend in the unadjusted risk of CV events among participants with low serum testosterone levels (P < 0.001), the risk was no longer significant after adjustment for age. There was no significant interaction with age and serum testosterone level tested as continuous variables (P = 0.07).

Limitations: A short follow-up period and serum testosterone measured on a single occasion.

Conclusions: Low serum testosterone concentration may be a modifiable risk factor for adverse outcomes and poor quality of life in male hemodialysis patients. This hypothesis should be tested in randomized controlled trials.