"Sleeping with more than 20 women protects men against prostate cancer, academics find," The Daily Telegraph reports.
The study in question included more than 1,500 men diagnosed with prostate cancer and a matched group of men without prostate cancer from the general population. Researchers then interviewed the men about their sexual activity.
The study concluded that having more than 20 female partners over a man’s lifetime was associated with a decreased risk of prostate cancer, while having several male partners seemed to increase the risk.
However, studies like this one cannot prove that having several female partners reduces the risk of prostate cancer, or that having more male partners increases risk.
If having multiple female partners reduced prostate cancer risk, it would be expected that the more partners you had, the lower your risk would be. However, this study did not find such a relationship between the number of female partners and associated risk. The protective effect was only seen in men in the upper category of having 21 or more female partners, and strangely, no other category.
The findings do not give a clear or coherent picture, and there may be many other factors influencing the relationship.
Where did the story come from?
The study was carried out by researchers from the Université du Québec, the University of Montreal and the University of Montreal Hospital Research Centre. It was funded by the Canadian Cancer Society, the Cancer Research Society, the Fonds de la recherche du Québec-Santé (FRQS), the FRQS-RRSE, and the Ministére du Développement économique, de l’Innovation et de l’Exportation du Québec.
The study was published in the peer-reviewed medical journal Cancer Epidemiology.
The media coverage of this story was almost universally poor, with news articles reporting non-significant findings or omitting the fact that some associations were only found in a sub-type of prostate cancer, or omitting the comparator.
For example, The Independent, the Mail Online and the Telegraph report that men who slept with more than 20 women were 19% less likely to develop an aggressive type of cancer than those who had only one female sexual partner, whereas this association was actually not statistically significant.
They go on to say that having more than 20 male partners doubled the risk of prostate cancer. In the Mail and Telegraph, this is compared to men who had never slept with a man. This is again incorrect. Having more than 20 male partners compared to having one was associated with an increased risk of less-aggressive prostate cancer, but it was not significantly associated with prostate cancer risk overall or risk of aggressive prostate cancer. Having more than 20 male partners compared to only having female partners was not significantly associated with any type of prostate cancer.
It would appear that most media outlets simply took the accompanying press release at face value, without actually bothering to read the study or subject it to any kind of critical analysis – a practice that is sadly all too common these days.
What kind of research was this?
This was a case-control study comparing sexual activity and STIs in men diagnosed with prostate cancer (the cases) and men matched by age from the electoral registry (the controls).
Case-control studies like this one cannot prove that the number of sexual partners, or their gender, is directly associated with risk of prostate cancer. There may be many factors (confounders) influencing the relationships observed in this study.
What did the research involve?
The researchers compared 1,590 men diagnosed with prostate cancer in a Montreal hospital between 2005 and 2009 (the cases) and 1,618 randomly selected age-matched men from the electoral list (the controls).
The men were interviewed to obtain information on sociodemographic factors (for example, highest level of education, family income, marital status), lifestyle (including sexual activity and STIs), medical and environmental factors.
Men were asked whether they had ever had sexual intercourse, the age they first had sexual intercourse, the number of female partners they had and the number of male partners they had.
Men were also asked to describe their sexual orientation in terms of the following categories:
- bisexual, with a preference for women
- bisexual, with a preference for men
- bisexual, with no preference for women or men
Men were also asked whether they had ever had the following STIs:
- genital herpes
- genital warts or condylomas (anal warts)
- human papillomavirus (HPV)
- human immunodeficiency virus (HIV) and acquired immune deficiency syndrome (AIDS)
- and/or other diseases
The researchers looked to see if there was an association between sexual activity and STIs and developing prostate cancer. The researchers also divided prostate cancer into aggressive (Gleason score =>7) and less-aggressive (Gleason score <7) forms of prostate cancer to see if there were different associations. (See Diagnosing prostate cancer for more information about Gleason scores)
The researchers adjusted their analyses for age, whether the men were of European, Black, Asian or other ancestry, family history of prostate cancer, and prostate cancer screening history.
What were the basic results?
Men with prostate cancer, compared to controls, were more likely to be Canadian-born and to be of European or Black ancestries, and less often of Asian ancestry. Twice as many men with prostate cancer had a first-degree relative (for example, brother or father) with prostate cancer than controls. Almost all men with prostate cancer reported being screened for prostate cancer within two years before being diagnosed with prostate cancer, whereas 76% of controls reported being screened in the preceding two years. Cases and controls had similar family income, marital status, smoking history and alcohol use.
Never having sexual intercourse was not associated with a difference in risk of prostate cancer overall. The age men first had intercourse was also not associated with a difference in risk of prostate cancer.
Compared with having one female sexual partner, having more than 20 female sexual partners was associated with a decreased risk of prostate cancer (odds ratio [OR] 0.72, 95% confidence interval [CI] 0.56 to 0.94). Men who have sex with men were included in this analysis.
Compared with having one male sexual partner, having more male sexual partners did not alter the risk of prostate cancer.
There was no association found between individual types of STIs, or with all STIs combined, and prostate cancer. However, the researchers state that this might have been because few men had an STI.
When aggressive and less-aggressive prostate cancer were analysed separately, the researchers observed no associations with aggressive prostate cancer. Never having sex, having had two to three male partners compared to never having had one, and having two to three or 21 male partners or more compared to having had one were associated with increased risk of less-aggressive prostate cancer; having more than 20 partners of either gender or more than 20 female partners compared to having one was associated with reduced risk of less-aggressive prostate cancer.
How did the researchers interpret the results?
The researchers conclude that “our findings are in support of a role for the number of sexual partners in prostate cancer development.”
This case-control study has suggested that having several female partners over a man’s lifetime is associated with a protective effect against prostate cancer, whereas having several male partners increased the risk.
However, case-control studies like this one cannot prove that having several female partners reduces the risk of prostate cancer, or that having more male partners increases risk. When it comes to complex issues such as lifestyle, sexuality and cancer outcomes, there could be a wide range of contributing factors.
If having multiple female partners reduced prostate cancer risk, it would be expected that the more partners you had the lower your risk would be. However, this study did not find such a relationship between number of female partners and associated risk. The protective effect was confined to men in the upper category of having more than 20 female partners.
You would reasonably expect to see a kind of dose-dependent relationship, so, for example, having 15 or 16 partners would also have a protective effect. No such relationship was seen, raising the possibility that the “21 or more” outcome was a statistical fluke; a confidence interval of 0.56 to 0.94 is of borderline statistical significance.
Similarly, the results are confusing for men who had male partners. Having two to three, or more than 20, male partners increased risk of less-aggressive prostate cancer compared to one male partner. However, between four and 20 was not associated with increased risk.
It seems possible that the associations seen in this study were the result of performing multiple comparisons. That is, where performing an increasing number of comparisons increases the chances that you will find some significant associations, even if there is not truly a significant link. If there were truly significant links, you would expect to see more consistent trends.
If these are real links, there could still be other factors that are influencing the relationship. It is difficult to know whether face-to-face interviews of sexual activity will have given reliable results, and also possible that a man’s knowledge of his cancer status could have influenced his recall.
Further work is required to determine whether having multiple female partners really is associated with a reduced risk of prostate cancer, and to see whether the gender of partners really makes a difference.
The most important thing to remember – regardless of the number or gender of partners – is to practice safe sex with a condom to reduce the risk of STIs.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Daily Telegraph, 28 October 2014
The Independent, 28 October 2014
Mail Online, 28 October 2014
Daily Mirror, 28 October 2014
Links to the science
Cancer Epidemiology. Published online September 29 2014