iNOS is known for releasing large amounts of nitric oxide as a defensive tactic. Nitric oxide is a free radical that can be used as a weapon and, therefore, sudden low testosterone essentially puts a huge oxidative load on the brain and may actually do some damage that leads to Parkinson's.  A further verification of this is the fact that, according to one study, men are times more likely to contract Parkinson's and that low testosterone levels have been associated with Parkinson's.  In addition, giving male Parkinson's patients testosterone can help with tremors. 
ATHLETES: Overtraining can seriously affect performance. Researchers have identified several short term key markers that result from overtraining, including "impaired anaerobic lactic acid perforance and a reduced time to exhaustion". [ 7 ] The same study discusses many other soon-to-follow issues including problems with uric acid levels, ammonia, creatine kinase, free testosterone/cortisol, growth hormone and so on. More subjective issues include excessively sore muscles and sleep and mood disorders. Again, overtraining is real and leads to real physiological issues.
We observed the response of serum growth hormone (GH) and testosterone (T) to a progressive resistance strength training program. Basal levels (after a 12-h fast) of GH and T were measured in young (23 years) and elderly (63 years) subjects before and after a 12-week training program. The response of GH and T to an acute bout of exercise was also measured. The exercise training, which involved all the major muscle groups, was conducted on Nautilus equipment and required 45-60 min for completion. The subjects completed three sets of lifts with 8-10 Reps/set. Blood was drawn from an anticubital vein, centrifuged (1169 g) for 15 min and the serum frozen for later analysis. The acute exercise blood samples were taken immediately before and after the exercise and at 15 min post-exercise during week 1 and 12. The hormone assay was carried out with radioimmunoassay kits for GH and T. The basal level of GH increased by % in the young and by only 3% in the elderly but neither change was significant. In response to a single exercise session GH levels in the young went from +/- to +/- ng/ml before training and from +/- to +/- after training. Each response was significant (P less than ) as were the pre-post differences (P less than ). In the elderly the response was not as great, values increasing from +/- to +/- ng/ml before training and from +/- to +/- ng/ml after training were recorded. These differences represented significant increases (P less than ) but did not demonstrate pre- to post-changes. Basal levels of T decreased in both groups, but were not significant. The T response to an acute bout of exercise was not significant but did increase in both age groups. In conclusion, the data presented here indicate that strength training can induce growth hormone and testosterone release, regardless of age, but that the elderly response does not equal that of the young.