Testosterone therapy, association with age, initiation and mode of therapy with cardiovascular events: a systematic review.

BACKGROUND

Although male hypogonadism is associated with increased cardiovascular events (CVE), recent concerns are that testosterone supplementation may increase CVE. The purpose was to determine associations with age, initiation or mode of therapy to explain these discrepancies.

DATA SYNTHESIS

Meta-analyses were supplemented through Scopus and PubMed with search terms 'testosterone', 'random' and 'trial'. CVE, defined before data extraction, were death, myocardial infarction, acute coronary syndrome, percutaneous coronary intervention, coronary bypass, syncope, arrhythmia, hospital admission for congestive heart failure or cerebrovascular event.

RESULTS

There were 45 trials with 5328 subjects evaluated, with a mean age of 63·3 (SD ±7·9) years, followed for mean study duration of 10·6 (± 8·6) months. Overall, testosterone supplementation was not associated with increased CVE risk ratio (rr = 1·10 (95% CI 0·86; 1·41, P = 0·45)). However, there was an increase event rate during the first 12 months (rr = 1·79 (1·13;2·83, P = 0·012)), predominantly in those ≥65 years, (rr = 2·90 (1·35;6·21, P = 0·006)). Within studies with lipid data, CVE were associated with fall in HDL, P = 0·002. Intramuscular testosterone appeared neutral for CVE (rr = 0·96 (0·462;1·98, P = 0·91)) compared with oral testosterone (rr = 2·28 (95% CI 2·28;8·59, P = 0·22)) and transdermal testosterone (rr = 2·80 (1·38;5·68, P = 0·004)). Intramuscular testosterone had the least effect of lowering HDL and non-HDL cholesterol (both P < 0·001).

CONCLUSIONS

Testosterone supplementation may be associated with increased CVE in those ≥65 years especially during the first year. Biological actions may differ depending upon mode of testosterone administration with intramuscular testosterone having less cardiovascular risk.