Endocrine Abstracts

Background: Low testosterone levels are a common finding in men with type 2 diabetes and are also associated with low HDL cholesterol (HDL-C) levels which are an independent risk factor for cardiovascular disease. The androgen receptor CAG repeat polymorphism (AR CAG) and GGN repeat polymorphism (AR GGN) affect receptor function such that shorter AR CAG and AR GGN=23 are associated with greater transcriptional activity in vitro.

Methods and results: We assessed the relationship between HDL-C, serum total testosterone (TT), AR CAG and AR GGN in a group of 182 men with type 2 diabetes. Group means were compared using Student’s t-test and results considered significant at P<0.05. The group was split into quartiles according to total testosterone. Mean HDL-C levels increased across quartiles with significantly higher levels in quartile four (HDL-C=1.29 mmol/l, TT=20.9 nmol/l) compared with quartile one (HDL-C=1.06 mmol/l, TT=6.9 nmol/l, P=<0.01). Separately we split the group according to AR CAG (AR CAG=21 or less versus 22 or more) and AR GGN (AR GGN=23 versus others). Combining these gave four groups including one group with the most active receptors (AR CAG 21 or less, AR GGN 23) and one with the least active. HDL-C was found to be lowest in the group with the most active receptors and highest in the group with the least active receptors (HDL=1.08 vs 1.21 mmol/l, P=0.042).

Discussion: Testosterone and AR function have contradictory associations with HDL-C in men with diabetes. Low testosterone levels and an active receptor are both associated with reductions in HDL-C. The effect of testosterone on HDL-C in clinical trials has also been variable. Testosterone may influence HDL-C in conflicting ways with reductions occurring via the AR CAG but opposite effects via actions elsewhere. The dominant effect may depend on the population studied.