Intramuscular testosterone esters and plasma lipids in hypogonadal men: a meta-analysis.

PURPOSE It is unclear whether intramuscular administration of testosterone esters to hypogonadal men is associated with changes in plasma lipids. We therefore analyzed 19 studies published between 1987 and 1999 that focused on male subjects with nonexperimental hypogonadism, treated subjects with an intramuscular testosterone ester and reported pretreatment and post-treatment concentrations of total cholesterol, low-density lipoprotein (LDL), high-density lipoprotein (HDL) cholesterol, or total triglyceride. METHODS We calculated study-specific, post-treatment minus pretreatment differences in each plasma lipid concentration (mean [95% confidence interval]). After testing of between-study homogeneity, we combined the study-specific differences. We then determined whether heterogeneity of differences could be explained in models of the differences on study and patient characteristics (mean +/- SE) before and after excluding extreme values using a multiple outlier procedure. RESULTS The studies represented 272 hypogonadal men (age 44 +/- 4 years; 20% with hypergonadotropic hypogonadism; total testosterone 0.5 +/- 0.2 ng/mL) who received, on average, 179 +/- 13 mg intramuscular testosterone ester every 16 +/- 1 days for 6 +/- 1 months. Fixed-effects estimates of post-treatment minus pretreatment differences were -14 [-17 to -11] mg/dL (total cholesterol), -5 [-8 to -1] mg/dL (LDL cholesterol), -4 [-5 to -2] mg/dL (HDL cholesterol), and -1 [-6 to + 4] mg/dL (triglyceride). Decreases in HDL cholesterol were larger at lower dosages of testosterone ester (r = -0.54, P = 0.055), but were not explained by attrition, regression to the mean, dosing frequency or duration, concomitant elevation of plasma total testosterone, aromatization of testosterone to estradiol, or other study and patient characteristics. CONCLUSION Intramuscular administration of testosterone esters to hypogonadal men is associated with a small, dosage-dependent decrease in HDL cholesterol and concomitant declines in total cholesterol and LDL cholesterol. The aggregate effect of these changes on cardiovascular risk remains unknown but deserves further study.