Twenty-nine studies were included. There were 18 RCTs (613 men), 7 double-blind crossover studies (89 men), 1 single-blind crossover study (6 men), 1 double-blind study with an unknown method of treatment allocation (36 men) and 2 non-randomised studies (67 men).
Testosterone supplementation increased lean body mass (5 out of 7 studies) and decreased fat mass (7 out of 9 studies assessing some measure of body fat) in healthy men with low to slightly decreased testosterone levels. However, these changes were not significant in all cases.
For all studies, testosterone supplementation was associated with either an improvement or no change in lower body strength (5 studies: all reported no change), upper body strength (5 studies: 3 reported no change and 2 reported an improvement), function (7 studies: 4 reported no change and 3 reported improvements in some measures), sexual function (5 studies: 4 reported no change and 1 reported improvements in some measures), and mood (10 studies: 6 reported no change and 4 reported improved mood or energy).
Testosterone supplementation had mixed effects on cognitive function (5 studies). Improvement was found on some measures: spatial and verbal memory, spatial ability and spatial cognition improved in 2 studies, while working memory improved in 1 study. No improvement was found on other measures, i.e. memory recall and verbal fluency (1 study). In addition, verbal fluency failed to improve with practice (1 study).
Testosterone supplementation improved exercise-induced coronary ischaemia seen on electrocardiography (4 studies), but the results for angina were mixed (4 studies: 2 reported reduced angina and 2 reported no change).
A few studies found that men with low testosterone levels were more likely to show improvements in the following outcomes than men with a lesser degree of testosterone deficiency: lumbar bone mineral density (2 studies reported an increase for low testosterone men; no unrestricted studies assessed this outcome), self-reported functional status (4 studies reported no change for low testosterone men; 3 unrestricted studies reported an improvement), libido (3 studies reported an improvement for low testosterone men; 1 unrestricted study reported an improvement), erectile function (1 low testosterone study reported an improvement; 2 unrestricted studies reported no improvement), and exercise-induced coronary ischaemia (2 studies reported a benefit for low-normal testosterone men; 1 study reported a benefit for markedly reduced testosterone).
Testosterone supplementation had no major adverse effects on lipids: 4 studies reported that low-density lipoprotein cholesterol levels were reduced or unchanged, while 6 studies reported that high-density lipoprotein levels were unchanged.
Most of the studies reported that the mean haematocrit increased from baseline by 2.5 to 5%. Between 6 and 25% of men developed haematocrit levels above the normal range.
Nine studies found no increase in voiding symptoms, abnormal findings on examination of the prostate, or postvoid residual urine during treatment. One study found that the prostate increased in size by 12%. The results for prostate-specific antigen were mixed: 5 studies reported no effect, while 4 studies reported a significant increase even after 6 weeks' treatment.
One study reported no change in sleep apnoea with testosterone supplementation.