It is a misunderstanding Chris, which has become over-hyped by some.
For men with a prostate who are being closely monitored (for example, on AS trying to pinpoint when to move to radical treatment) it is worth noting that orgasm and rectal stimulation of the prostate can raise the PSA for up to 24 hours and cycling may raise the PSA for up to 4 hours. Even then, the estimate is that orgasm may raise the PSA by 10% and cycling by less than that. So it makes sense, if the trend is crucial, to try to behave in the same way before each test ... have sex or don't have sex, cycle or don't cycle but try to be consistent so that any small changes in PSA aren't misinterpreted.
The jury is out on whether cycling can raise the PSA in a man with a fried prostate but even if it could, would it really matter to you? If your current PSA is 1.0, you cycle this morning and have your PSA test this afternoon and the result is 1.05 could you be absolutely certain that you hadn't done anything before the previous test that could affect the reading?
John tries to be consistent about his behaviour at the time of each test; he cycles regularly, goes to the gym (well, he would do usually), has his blood test early in the morning, etc. If the next PSA is the same as last time, should we worry that it should have been lower because he hasn't been to the gym recently? Do the things that you enjoy and are good for your physical & mental health and at test time, remember that very small variations are not significant whereas a solid upward trend might be.