OMG … The finger length/prostate cancer risk hypothesis is back again


In July 2010 we commented on an article in The Daily Mail that was based on data from a team of Korean researchers who’d been busily measuring the lengths of the second and fourth fingers of the right hands of men presenting at their urology clinic. They were convinced that there was a correlation between relative finger length and risk for prostate cancer.

If you want to see our original article, please click here. We wrote a follow-up article later (in December 2010). The original paper by the Korean research team didn’t actually get published until February 2011. It was based on the finger lengths of 366 patients.

Now the same research team has published a new paper — based on the finger lengths of 770 patients. (We assume that these 770 patients include the first 366 on which the authors reported the first time.)

In this series of 770 men, all of whom were 40 years or older and presented at the author’s clinic with LUTS (lower urinary tract symptoms), the authors

  • Measured the lengths of their second and fourth fingers of the right hand (before they did anything else)
  • Measured their PSA levels
  • Gave them a digital rectal examination (DRE)
  • Gave them a transrectal ultrasound exam

They then categorized the patients into two groups:

  • Patients in Group A (n = 420) had to have a digit ratio of < 0.95. (In other words, their second finger — the one nearest to their thumb — was shorter than their fourth or ring finger and had to be less than 0.95 times the length of their ring finger.)
  • Patients in Group B (n = 350) had to have a digit ratio of ≥ 0.95. (In other words, their second finger had to be equal to or more than 0.95 times the length of their ring finger.)

All patients who had either a PSA ≥ 3 ng/ml or an abnormal DRE result or both were given a prostate biopsy.

Here is what the authors found:

  • 166/770 men (21.6 percent) met the criteria for biopsy.
    • 94 men came from Group A (22.4 percent of 420).
    • 72 men came from Group B (20.6 percent of 350).
  • The rates of prostate cancer detection on biopsy were
    • 44/94 men (46.8 percent) among men from Group A.
    • 17/72 men (23.6 percent) among men from Group B.
    • This difference is statistically significant (p = 0.002).
  • The rates of positive biopsy cores were
    • 282 positive cores in the 94 men from Group A.
    • 126 positive cores in the 72 men from Group B.
  • The authors calculated that
    • Biopsied men in Group A had cancer in 46.7 percent of their core volume.
    • Biopsied men in Group B had cancer in 37.1 percent of their core volume.
  • They also found that
    • Biopsied men in Group A had a Gleason score of 9 or 10 in 18/282 cores (6.4 percent).
    • Biopsied men in Group B had a Gleason score of 9 or 10 in 1/126 cores (0.8 percent).
    • Biopsied men in Group A had a primary Gleason grade of 5 in 12/282 cores (4.3 percent).
    • Biopsied men in Group B had a primary Gleason grade of 5 in 0/126 cores (0.0 percent).
    • Both these differences are statistically significant.

The authors use these data to conclude that, “A lower digit ratio is related to an increased detection rate of prostate cancer, a high percentage of core cancer volume, and a high Gleason score.”

We continue to find this hypothesis bizarre. If the authors are correct, then one ought to be able to prove the hypothesis in reverse … by measuring the relative finger lengths of a quite different group of men already actually diagnosed with prostate cancer. One should then find that, out of 1,000 men diagnosed with prostate cancer:

  • About 720 would have a digit ratio of < 0.95.
  • About 280 would have a digit ratio of ≥ 0.95.
  • Men with a digit ratio of < 0.95 would be eight times more likely to have had a Gleason score of 9 or 10 on biopsy than men with a digit ratio of ≥ 0.95.

If we did this study using men in Denmark and we were unable to demonstrate this outcome, then it would be clear that there was “something rotten in the state of” Korea.

On the other hand (assuming I am measuring the lengths of my fingers correctly), my right-hand digit ratio is 0.97, placing me in Group B, and to date (at age 64+) I have no sign other than my age indicating risk for prostate cancer. …

3 Responses

  1. The Sitemaster’s comments don’t address a huge literature on digit length ratio and masculine traits in males. There have now been a fair number of other digit ratio papers on prostate cancer patients. One out of Canada failed to find any correlation. But several others, from different countries, have reported the same pattern. One can get to all these papers through PubMed searching on: prostate digit ratio.

    This study should be seen in the context of other work on prostate cancer risk and early exposure to high levels of testosterone, which correlate with digit ratios. I can think of a few other developmental traits that are testosterone dependent and correlate with prostate cancer risk decades later.

    This stuff needs to be taken seriously — not as definite predictors of prostate cancer risk — but as clues to how androgens or androgen receptors promote promote prostate cancer.

  2. Here’s a recent study by US researchers at Tulane on the same issue – http://www.nature.com/pcan/journal/vaop/ncurrent/full/pcan201246a.html . The lead researcher, a young friend of mine, is a 5th-year medical student with an interest in prostate cancer currently interviewing for a radiation oncology residency, hopefully at UCSF.

    The basis of the theory relates to the levels of androgens in the uterus. While the Tulane study found some correlation it was statistically inconclusive.

  3. Dear Richard:

    The reason I did not address the underlying principle of a hypothetical correlation between androgen levels and digit length ratio is because I have discussed this before in one of the commentaries linked above. The “huge” literature on this topic is far from definitive (at least in my miserable opinion).

    I certainly don’t deny that androgen levels in very young males may have a correlation to prostate cancer risk later in life … right along with diet, environment, and who knows what else over a 50- to 60-year timeframe. However, as you well understand, the existence of an association is by no mean a demonstration of real clinical probability. My problem with papers like the one referred to above is that they actually add little to the existing database. Perhaps you and your colleagues would like to do the “reverse” study that I outlined above (or some other study that would help to clarify the hypothesis). It shouldn’t be that difficult.

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: