Several newspapers reported the study of sexual practices and prostate cancer published this week in the British Journal of Urology. 

The study was a case-control study, meaning that a group of men with prostate cancer (cases) was compared with a group of men who didn't have prostate cancer.  The aim of the study was to see whether there are other differences between the groups that

Two particular findings were emphasised in the news reports:  more frequent masturbation aged in the 20s and 30s was associated with a higher risk of prostate cancer, while more frequent masturbation in the 50s was associated with a lower risk.  

Several websites picked up the very serious limitations in the study to do with:

• The nature of the study:  it cannot prove causation, only association.  There might be some other factor at work.
• The (un)reliability of self-reports of sexual activity from 30 or more years in the past
  
• The representativeness of the sample
• The size of the sample
• The fact that the study did so many comparisons that the chances of a "false positive" association are very high (i.e. the association between masturbation and prostate cancer may have been found by chance)

Get the background to this story

• Behind the Headlines analysis from NHS Choices
• Comment from The Prostate Cancer Charity
• Comment from The Prostate Cancer Research Foundation (who funded the study)
• Abstract of the paper from the British Journal of Urology