Testosterone supplementation and prostate cancer risk

In case you haven’t noticed, there has been a massive surge in the use of testosterone supplements by older men in the past 10 years or so (as we seek to maintain the bizarre illusion that we are really still in our early 30s and might still be able to win an Olympic gold medal in [pick your sport]).

Now come some far from surprising findings published in a report by Hyde et al., and based on data from a large, Australian study of healthy aging (the so-called Health in Men Study).

The goal of this specific element of the Health in Men Study was to investigate whether serum levels of testosterone and related hormones are associated with incident prostate, lung, and colorectal cancers. Between 2001 and 2004, the research team had collected a wide variety of data (including blood samples) from a cohort of 3,635 apparently healthy men living in the community and aged between 70 and 88 years of age (with an average age of 77 years). They subsequently obtained notifications of diagnoses of cancer among this group of men via linkages to their electronic medical records (through December 31, 2010).

Here are the key study findings:

  • The average (mean) follow-up period was 6.7 ± 1.8 years.
  • During that follow-up period, the researchers identified
    • 297 cases of prostate cancer
    • 104 cases of colorectal cancer
    • 82 cases of lung cancer
  • Through the use of sophisticated statistical models it was shown that
    • Each one-point increase in the standard deviation above the average in free testosterone level was associated with a 9 percent increase in risk for diagnosis of prostate cancer.
    • Other hormone levels were not significantly associated with any increase in risk for diagnosis of prostate cancer.
    • There were no significant associations between hormonal levels and risk for diagnosis of colorectal cancer.
    • Men with higher total testosterone levels appeared to be at increased risk for diagnosis of lung cancer.
    • Men with higher free testosterone levels also appeared to be at increased risk for diagnosis of lung cancer.
    • The apparent increases in risk for diagnosis of lung cancer were still evident after exclusion of data from current smokers.

Hyde and her colleagues conclude that elevated levels of free testosterone were associated with an increased risk for diagnosis of prostate cancer and that elevated testosterone levels “may also” be associated with increased risk for diagnosis of lung cancer.

Now it needs to be clearly stated that “associations” are not the same as “cause and effect”. In other words, we cannot be certain that elevated testosterone levels are a direct cause of increased levels of diagnosis of prostate and lung cancer in men > 70 years of age. However, there is a clear warning here to men about the possible risks asociated with testosterone supplementation. … Raising serum testosterone levels above normal, healthy levels may significantly increase risk for diagnosis of some cancers (prostate cancer included) as one ages.

This is not a new idea. In America, where advertising of prescription drugs on television and in magazines is commonplace, many of us will have see advertising for testosterone supplments, and (ideally) noted the warning that use of such supplements may be associated with increased risk for prostate cancer. 

Hyde is quoted in The Financial Express as saying that, while higher levels of testosterone were unlikely to actually cause cancer, they might make an existing cancer grow faster, and that “We need to conduct large-scale, long-term trials of testosterone therapy to see if this risk applies to men receiving testosterone.”

14 Responses

  1. Precautions for Androgel include the note “Can cause or worsen an enlarged prostate”. There’s no “may be associated with” in the language for this possible side effect.

  2. Dear Doug:

    Actually, if you look carefully at the listed side effects for Androgel and for Androgel 1.62%, they both include the term “Possible increased risk of prostate cancer.”

  3. I can only see the abstract, but it does not appear that they addressed the issue of clinically significant vs. not clinically significant prostate cancer.

  4. Tracy:

    That wasn’t the point of the study.

  5. Other studies (see #18 28,000 men & #19 Johns Hopkins) suggest the opposite, that high T levels are associated with lower incidence of PC. The androgen receptors are saturated just above castration level so more T seems moot.

    18) Cancer, Epidemiology, Biomarkers Preview, v.6, 1997, p967-9
    19) Prostate, v.27, 1995, p. 25-31

  6. Terry:

    Since I can’t check either of those references at present, its hard for me to make any meaningful comment … but they are based on data from nearly 20 years ago, which always makes me wonder how accurate they are.

  7. Haven’t we killed enough women with estrogen? Now we are going to kill men with testosterone. Women were promised the fountain of youth and had their breasts cut off.

    Men are being fed the same lie.



    Hi Sitemaster and group,

    As an old statistics hand I was confused by the wording “… each one-point increase in the standard deviation above the average in free testosterone level was associated with a 9 percent increase in risk for diagnosis of prostate cancer….” What the abstract actually states is
    “each one SD increase in free testosterone was associated with a 9% increase in prostate cancer risk”.

    In essence from the perspective of statistics, in any bell shaped curve of a “normal” distribution of scores (in this case testosterone and free testosterone levels), with the average (mean) in the middle, there are about three standard deviation units to the left (lower scores) and right (higher scores) of the mean. About 68% of all scores are within one SD of both sides of the mean score. About 95% are within two standard deviations, and virtually all scores (except for a small fraction of a percent) are within three standard deviations.

    The SD was not stated in the abstract, but an increase of one SD level (from the median, then to the first SD level, and then to the second SD level) often will involve large increases in score, with the increase to the third level often being relatively small. For example, in the abstract of this study the mean for total testosterone was 15 nmol/l, and the abstract related increased likelihood of lung cancer at levels of 25 nmol/l and 30 nmol/l, the latter figure being twice as large as the mean. (Neither value is necessarily at the first, second or third SD level for testosterone and obviously does not apply to free testosterone.)

    What this means regarding this paper is that the 9% increase in prostate cancer incidence associated with increases to each SD level likely involves substantial increase in free testosterone above the average for men at the first and second SD levels.

    Terry, your information is interesting but old, as Sitemaster noted. I too am under an impression that high T levels may be protective for those who do not yet have the disease. It would be helpful to know if research has sorted this out. On the other hand, as a 12.5 year veteran of androgen deprivation as my sole therapy for a challenging case, I am thoroughly aware of the relationship between testosterone and prostate cancer growth for the vast majority of us who already have prostate cancer.

  9. A thought: If, after several years of zero PSA following a prostatectomy, might it be argued that a man might have even a less chance of having any chance of prostate cancer cells in his body than a random man off the street, given that what … 1 in 5 guys or something like that will get it in their life?

    Therefore any testosterone supplementation given to someone with no prostate cells in his body would be safer than giving it to someone who still has a prostate, and therefore the 1 in 5 inherent risk of having cancer cells in his body? Follow?

  10. Dear Mark:

    If you have no prostate or no living prostate tissue (after any form of definitive treatment), then certainly you are at no risk for prostate cancer, and therefore the risk for prostate cancer as a consequence of using testosterone supplements is zero. The article discussed above is solely applicable to men who have never been diagnosed with prostate cancer.

  11. My last PSA was 0.014. It has been 8 years since I had seed implants in the prostate. My testosterone levels are considerably below normal and I have below average energy. I would like some testosterone therapy if it would help. Is that OK? I am 74 years of age.

  12. Dear George:

    The use of testosterone supplements by a man with a prior history of prostate cancer always comes with some degree of risk. No one can tell you that it is completely safe.

    However, if your PSA has been low and stable for 8 years and your original diagnosis was of a relatively low-risk form of prostate cancer (e.g., PSA less than 10; Gleason score of 6 or less; clinical stage no higher than T2a) then you could certainly discuss this with your family physician and/or with a specialist in the management of prostate cancer. The problem is that individual doctors tend to have strong feelings about whether this type of therapy is really worth the risk for the patient, and so the response you will get from your doctors could be based entirely on their gut feeling one way or the other. There are few good data that make a compelling case one way or other.

  13. I had radiation therapy for prostate cancer 2 years ago and my PSA is almost non-existant. I am now 59 years old and feel like I’m 90. I have no energy, big loss of muscle mass and strength, and my desire is gone for most of life’s pleasures. My doctor will not give me hormone supplements. But one has to weigh the effects of therapy versus living longer with no body to enjoy life with. If my PSA levels are almost no-existent, why can’t therapy be used to bring my levels up to a normal range?

  14. Dear Greg:

    This is a very controversial topic. Some physicians would be prepared to give you supplemental testosterone therapy; others are not going to be willing to do it at all; and then there are a lot of doctors who are “on the fence.” I can only suggest that you seek a second opinion. The other factor that may be important here is whether your insurance provider will cover the costs involved.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: