The recently published article in the British Journal of Cancer entitled “Muscle-building supplement use and increased risk of testicular germ cell cancer in men from Connecticut and Massachusetts” has not surprisingly generated a lot of media coverage.
WebMD wrote up a review here: WEBMD Review.
A Note On Observational Studies: Causation vs Association
The most important thing to remember about observational studies like this one is that they do not prove a causational relationship (A causes B), so the conclusion, that some lay publications incorrectly have been putting out, that “muscle-building supplements” cause testicular cancer is without question, false.
Let me explain.
To prove causation several conditions need to be met. Many epidemiologists use the Bradford Hill criteria, which requires many studies to begin to suggest causation, which is not the case here. Therefore, all that can be said for this study is that “muscle building supplements” were associated with testicular cancer.
Did They Find Association?
There are several key issues with this study that make even association between muscle building supplements and testicular cancer questionable;
The first, and most glaring, is the grouping of over 30 types of supplements together into “muscle building supplements, including creatine, protein, and androstenedione or its booster”. These are all different chemicals that have different effects and different carcinogenic properties. Further, I would be interested to see a breakdown of how much of each of these 30 supplements were taken. Intuitively, I would think protein would be consumed at much higher levels than androstenedione and likely creatine as well. I would also be interested to see a mechanism for how whey protein could be carcinogenic. To my knowledge I do not know this to be true. The author suggested that their might be “‘hidden’ ingredients that are not listed on the label”. It is equally, if not more likely, that there is “less active ingredients than indicated on the product label” which would then actually weaken the association.
Furthermore, the references they provide to suggest a mechanism for which “muscle-building supplements” could be carcinogenic are a bit of a stretch to say the least. They do not support the biologic plausibility of “muscle-building supplements” causing testicular cancer.
The Yu et al paper, which suggests that creatine is converted into formaldehyde, is from the journal “Medical Hypotheses” which isn’t peer-reviewed. And although this claim was published nearly 15 years ago, I was unable to find another study in a peer-reviewed journal to support it. Given the prevalence of creatine, if it really was degraded into a toxic substance there would be peer reviewed evidence to support this.
The Han et al paper is on formaldehyde inducing testicular damage in male rats in a dose dependent manner. Although this type of testicular damage could certainly induce infertility, it doesn’t necessarily mean it would result in testicular cancer development. The National Center for Biotechnology Information reference also provides no support for a mechanism for which Androstenedione could result in testicular cancer. Certainly infertility and testicular atrophy with long time usage, but not cancer. So, no real evidence is provided to suggest the mechanism by which “muscle-building supplements” leads to testicular cancer.
Additionally, aspects of the study design may have lead to finding a false association;
The study sample included men up to age 55. However, testicular cancer is the most common solid tumor in males between the ages of 18-40, with a peak incidence in late 20’s for non-seminomatous GCT and early 30’s for seminoma. Very few cases (<15%) of testicular cancer are diagnosed in men older than 45 years.
Since the popularity of “muscle-building supplements” has been growing over the last several years and younger individuals are more likely to use “muscle-building supplements” including individuals up to 55 years of age introduces bias in matching the control cohort. This is demonstrated by the higher mean age for the controls.
It is also interesting that the authors presented age as a mean without giving the standard deviations to show the distribution. For a continuous variable like this the median value with interquartile range (25th and 75th percentile) would likely further highlight the differences between testicular cancer patients and controls…eg testicular cancer patients are younger and more likely to have taken “muscle-building supplements” than the older controls.
Furthermore, testicular cancer is five times more common in white males than black males, and is linked to northern European descent. For example, Scandinavia has some of the highest incidences of testicular cancer in the world. This is also demonstrated in the differences between the 95% of whites in the testicular cancer group and the 89.5% of the controls. There is evidence that whites are more likely to take supplements than any other race (Foote et al, American Journal of Epidemiology 2003).
Although their statistical modeling adjusted for age and race the introduction of these confounders are not always fully accounted for.
In short, I think the research found that young, white men are more likely to take “muscle-building supplements”. It just so happens that young white men are also the group most likely to develop testicular cancer. In my opinion, that is the only thing you can draw from this study, nothing more.
Dr. Eugene Pietzak received his medical degree from the Stony Brook University School of Medicine in New York, graduating with Alpha Omega Alpha Honors. He is currently a chief resident in urologic surgery at the Hospital of the University of Pennsylvania in Philadelphia and will be starting his fellowship in urologic oncology at Sloan Kettering in July 2015. Dr. Pietzak’s clinical and research interests are in urologic oncology, particularly prostate, kidney, testicular, and bladder cancer. He has authored several book chapters and research articles on genitourinary cancers and has presented his research at numerous professional meetings. Prior to medical school he was a 4 year member of the Stony Brook University Division 1-AA Football squad, being named captain as a senior.