It is difficult to imagine a health warning that would generate more fear than the announcement that masturbation leads to prostate cancer. Over 100 newspaper articles trumpeting the findings of a new paper on the subject have generated a wave of embarrassment and concern. That’s why it is important to analyze the paper very carefully to determine whether the evidence supports the authors’ claims. In my judgment, the paper has some very serious drawbacks that render its conclusions suspect.

The paper is titled Sexual activity and prostate cancer risk in men diagnosed at a young age. The paper only looked at a subset of men with prostate cancer, the 25% who are diagnosed before age 65. It is unclear why they chose to look only at this group and whether they believe there are any differences in prostate cancer occurring before age 65 and prostate cancer occurring after.

The way the study was conducted (the methodology) raises questions about the validity of the conclusions. The study is a case-control study, in which men with prostate cancer were compared to a control group of similarly aged men without cancer. According to the Oxford Centre classification of studies, a case control study is among the weakest forms of study, rating a 3 on a scale of 1-5; the two lower grades apply to descriptions that make no comparisons and to opinions.

This is also a retrospective study, meaning that the participants were asked to recall events that took place as many as 45 years before. Not surprisingly, these types of studies are often weakened by inability to properly recall details.

Finally, this is a very small study, looking at only 400 men with prostate cancer and 400 controls. A smaller study is less likely to generate valid results.

What were the authors expecting to see? The authors were hoping to determine if there is a relationship between sexual activity and the risk of developing prostate cancer before the age of 65. They looked at two types of sexual activity, masturbation and sexual intercourse.

What would such a relationship look like? In general, if a relationship exists, you would expect to see a “dose-response” effect. In other words, if increased sexual activity increased the risk of prostate cancer, you would expect that the lowest sexual activity levels would lead to the lowest risk, moderate levels of activity would be associated with moderate risk, and high levels of activity would be associated with high levels of risk.

What did the authors find? They found that there was no relationship between overall sexual activity and prostate cancer risk. The proportion of men with low, medium and high frequency of sexual activity were equal.

It’s difficult to publish a study that doesn’t show anything, so the authors went back and reworked their data. First, they looked at the relationship of sexual activity in each decade (20’s, 30’s, 40’s, 50’s) to the risk of developing prostate cancer before the age of 65. There was a suggestion that increased sexual activity in the 20’s was correlated with an increased rate of prostate cancer, but that relationship did not hold for any other decade. In fact, increased sexual activity in the 50’s was actually correlated with a decreased rate of prostate cancer

Then they further sub-divided the data into sexual intercourse frequency and masturbation frequency per decade. There was no relationship between sexual intercourse frequency in any decade and risk of prostate cancer before age 65.

Finally, they looked at the relationship between masturbation frequency per decade and the risk of prostate cancer before age 65. There was no relationship there, either BUT they noticed that low frequency of masturbation in the 20’s and 30’s was associated with a higher rate of prostate cancer, but masturbation in the 50’s was associated with a lower rate of prostate cancer.

What should we conclude from this? We should conclude that there is no overall relationship between sexual activity (whether intercourse or masturbation) and risk of developing prostate cancer before age 65. The few, random associations that appeared in the data are almost certainly anomalies, reflecting the small sample size, and the limitations inherent in any case-control studies. These findings are consistent with multiple previous studies that showed both positive and negative associations, as well as no association at all. The fact that there is no consistent pattern among all the studies taken together further reinforces the conclusion that there is no relationship between sexual activity and risk of prostate cancer.

That’s not what the authors decided to conclude, however. They chose to assume that the random associations they found were real, and not artifacts. They provide no explanation for why or how masturbation could increase the risk of prostate cancer in some decades and also decrease it in other decades, while at the very same time intercourse had no effect at all.

Does masturbation increase the risk of prostate cancer? There is no evidence in this paper that it does, and, a survey of previous papers on the topic provide no evidence of a consistent relationship of any kind. The conclusions of this paper are weak and unjustified, and the publicity campaign waged by the authors and publishers is disingenuous at best, if not outright irresponsible.