Wednesday February 16 2011
There are conflicting results over hair loss and cancer risk
“Going bald early ‘doubles prostate cancer risk’,” The Daily Telegraph reported. The newspaper said a study has found that men who develop a pronounced “widow’s peak” by the age of 20 should be vigilant about the cancer in later life.
This study has suggested that male-pattern baldness at the age of 20 is associated with prostate cancer. However, other studies have not found such an association, and some have concluded that men with early hair loss have a lower risk of prostate cancer. The Daily Mail reported on one such study in March 2010, saying that hair loss “almost halves the risk of prostate cancer”. Clearly, the issue is not straightforward.
Bald men should not worry about the findings of this research, which are far from conclusive. Baldness itself is unlikely to determine whether a man will develop prostate cancer. It is far more likely that both baldness and prostate cancer share a common risk factor, such as genetics or raised testosterone levels. As the researchers said themselves, more research is needed to clarify this confusing relationship.
Where did the story come from?
The study was carried out by researchers from several medical and academic institutions in France. There is no information about the funding of this research, which was published in the peer-reviewed medical journal Annals of Oncology.
The newspapers reported the study accurately, and the Daily Mail made it clear that there has been conflicting evidence about the relationship between male-pattern hair loss and cancer. However, the accompanying headlines suggest too strongly that there is a firm link between baldness and prostate cancer, particularly as the researchers themselves do not make such claims. In fact, they say that the link between hair loss and cancer is unclear and they encourage further work on the subject. The research also has shortcomings that were not mentioned in the press coverage.
What kind of research was this?
This case-control study assessed the relationship between early-onset male-pattern baldness and the development of prostate cancer. This same study design was used by American researchers who concluded in 2010 that baldness was associated with a reduced risk of prostate cancer. Case-control studies compare a past event or the history of people with and without a condition or disease to determine whether a particular exposure is more common in those with the disease.
One key limitation of case-control studies is that they rely on people to correctly remember their exposures, in this case their level of hair loss several years earlier. People do not always remember the past accurately, and this can introduce bias into the study.
What did the research involve?
The researchers recruited 669 men, 388 of whom had prostate cancer (the cases). The patients with prostate cancer were recruited through radiation clinics in France, which they were attending as part of their treatment. The researchers selected control patients who did not have prostate cancer from the same hospital and matched them with the cases according to their date of birth.
All participants were sent a questionnaire asking whether they had a family history of prostate cancer or baldness. They were then asked to score their baldness at ages 20, 30 and 40 based on pictures showing no balding, frontal hair loss, vertex hair loss (at the top of the head) and full hair loss (both frontal and vertex). The patients’ doctors also completed a questionnaire providing further details about their prostate cancer, including age at diagnosis, severity and treatment history.
The researchers then compared the occurrence of hair loss at the different ages between men with prostate cancer and men who did not have the disease. In their analysis, they took into account the participants’ age and family history of disease, and adjusted their results for these potential confounders. From their analyses, they were able to calculate how much more likely men with prostate cancer were to have hair loss at different ages.
What were the basic results?
The study found that having any balding at the age of 20 was linked to an increased likelihood of having prostate cancer. Compared to the control group, people with prostate cancer were twice as likely to have had signs of baldness at the age of 20. There was no apparent link between prostate cancer and the development of baldness at the age of 30 or 40. Signs of baldness at 20 were not linked to the age at which tumours developed, nor was there any link between early baldness and the severity of prostate cancer.
How did the researchers interpret the results?
The researchers did not draw strong conclusions from their study. They acknowledged that other studies have not found the same association and that another study of a similar design came to the opposite conclusion. Given that there is much disparity on this issue in the literature, the researchers say that the link between male-pattern baldness and prostate cancer is not clear, and they encourage further work on the subject.
Although this case-control study found an association between early male-pattern baldness and prostate cancer, the findings are far from conclusive. There are several limitations to consider:
- There are some inherent problems with a case-control design, including the recall bias that can occur if participants fail to remember the details of their exposures. In this case, participants may not have accurately remembered exactly what baldness pattern they had at what age.
- Previous studies, even those of a similar design, did not find the same link between early-onset baldness and prostate cancer, and some found the opposite (that baldness is linked to a reduced risk of prostate cancer).
- Another problem with case-control studies is establishing the relationship in time between exposure and outcome. Any study which claims that exposure A causes outcome B must prove that A occurred before B. While the researchers captured data about the date of diagnosis of prostate cancer, they do not appear to have used this information in their analyses. The researchers could have used the date of diagnosis to limit their analyses to only those men whose baldness happened well before their cancer was diagnosed.
This is clearly not a straightforward subject, and even studies with near-identical designs have produced conflicting results. As such, no firm conclusions can be drawn from the literature about the link between baldness and prostate cancer. It is not clear why the results from these studies differ as they do, but it is possibly because complex genetic and environmental risk factors underpin both baldness and prostate cancer. Men with early signs of male-pattern baldness should not worry about this research.
Baldness itself is unlikely to cause cancer, and it is more likely that the relationships observed are the result of some undetermined factor that links both baldness and prostate cancer, such as genetics or testosterone levels. The confusion in findings can only be clarified by more robust research.