Are daughters a marker for prostate cancer risk?

The BBC offers this somewhat odd story (Daughters linked to prostate risk)

Men who father daughters, not sons, may be at a greater risk of developing prostate cancer, researchers have said.

The Israeli team found men with three daughters and no sons were up to 60% more likely to develop prostate cancer.

Mickey offers the following insights:

The researchers grasped at various straws to explain the finding, but "there was no evidence to support these explanations". In the interests of breaking the logjam, here is a plausible hypothesis:

In a 1983 book "Too Many Women," Marcia Guttentag and Paul Secord outlined the evidence that timing of intercourse in relation to the woman's menstrual cycle is associated with variations in sex ratio of offspring. Intercourse before ovulation is associated with more female offspring. It would not be surprising if patterns of sexual activity affected prostate health as well. For example, careful physicians will tell their patients that the PSA antigen test for prostate cancer is affected by sexual activity, and one should abstain from sexual arousal for several days before the blood test in order to get a reliable reading. According to this hypothesis, both prostate cancer and a disproportionate number of female offspring may result from particular patterns of male arousal and orgasm.

This line of reasoning may have important public health implications if one can give advice that will minimize risk of prostate cancer. It is not clear what that advice would be. It could be to reduce the frequency of intercourse, or it could be to reduce the frequency of arousal that is not followed by orgasm, since orgasm releases the pressure on the prostate. It is difficult to do studies of this sort, but the answer is worth knowing.