Birth month and disease risk study has nothing to do with astrology

Scientists reviewed New York medical databases for 1.7 million patients and found 55 diseases linked to birth month, according to a statement released Monday by the university.

Source: Study finds link between birth month and disease risk

It turns out your birth month might determine more than just your astrological sign. When you are born also could determine some of your health risks, according to a new study from Columbia University Medical Center.

Scientists reviewed New York medical databases for 1.7 million patients and found 55 diseases linked to birth month, according to a statement released Monday by the university.

People born in May have the lowest risk of disease overall and people born in October and November have the highest, according to the study.

This is not new, it’s nothing to pay much attention to at all, and it’s NOT related to astrology. Astrology was based on the planets and constellations in the sky when you were born. That actually changes. So the zodiac symbol that was set for you is not accurate – it was over 2000 years ago, but not now. This is due to precession. Science demolishes the idea of astrology seven ways from Sunday, but people still seem very attached to the idea. Hence, it’s confusingly mentioned in this piece.

The real story is that “changes in diet and yearly waves of infection could possibly influence the growth of developing baby, with a lingering effect on its health for decades afterwards.” We don’t quite know why but we know pretty definitely it has nothing to do with what constellation the sun was in when you were born.

But to reiterate: The effects of birth month are SMALL. It’s nothing to get excited about.

See more:

Researchers say there are sound and possibly scientific reasons to pay more attention to the month you were born in

Source: See What Diseases You’re at Risk For Based on Your Birth Month | TIME

Astrology: Why Your Zodiac Sign and Horoscope Are Wrong

  9 comments for “Birth month and disease risk study has nothing to do with astrology

  1. Haldurson
    June 11, 2015 at 10:45 PM

    I don’t remember much about the statistics I studied in college, but I do remember that when you cast a wide enough net, you can easily wind up with lots of statistical anomalies that may or may not be significant. I recall an example regarding a few studies of disease clusters — there were people who were trying to blame radiation, toxic waste, or power lines on the existence of various disease clusters. The problem was that disease clusters are going to happen regardless of cause, and that the fact that they exist may simply be a symptom of casting too wide a net. That doesn’t mean that all disease clusters are merely coincidence, just that the fact that they exist is not proof of anything.

    I think there was a similar thing about twins studies, where some researcher collected data covering hundreds of different likes, dislikes, hobbies, etc. and found lots of coincidences. And he pegged up to the fact that the people were twins, when in fact he completely ignored that there were more differences than similarities, and he had made no control group of non-twins to compare that to.

    If you have a large enough collection of completely random numbers, you are going to find unusual sequences and seeming coincidences. It may point to something to study further, but don’t be surprised if it turns out to be mere coincidence.

  2. June 11, 2015 at 10:50 PM

    Haldurson – you are correct. This is called controlling for multiple comparisons. The way I most often see this done is using the Bonferoni correction which (more or less) means you demand a lower p-value (higher standard of evidence) for more comparisons made. This method tends to limit false positives but enables false negatives.

    The study referenced here uses a method known as False Discover Rates. I personally prefer this method. It (more or less) says that you acknowledge “casting a wide net” will turn up spurious correlations, but you keep track of how many you generate, and you expect a certain number of them with predictable support.

    I’ve done a modest review of this study and it’s of high enough quality that I call it interesting. I expected to go into this as a skeptic and find errors. I found no obvious ones. The authors were mindful of good methodology, but this still demands deeper investigation.

    Yet, as Sharon points out, the effect is relatively small. Assuming this holds up to deeper scrutiny and reproducibility, it will certainly be interesting but not groundbreaking.

  3. June 12, 2015 at 1:51 AM

    Even with statistical correction for multiple comparisons, the lack of an a priori hypothesis or any attempt to assess prior probability makes this a blind statistical fishing expedition, and those almost never turn up causal associations that survive subsequent study. It is one of the main reasons so much medical research proves impossible to replicate. This kind of data dredging can be fun, and it can be useful in hypothesis generation, but it’s a pretty inefficient approach, and when the media starts reporting papers like this without any context or epidemiological background, the result is pretty misleading. I will await, without much optimism, confirmation that birth month is associated with the odds of non-venemous insect bites later in life.

  4. Mike C.
    June 12, 2015 at 10:50 AM

    They say statistics done’t lie, statisticians do.

  5. Wang-Lo
    June 13, 2015 at 2:20 PM

    I’m sorry but, although I too believe that astrology is utter billshut, I just don’t see the problem here. The article doesn’t go anywhere near indicating a relationship between “your astrological sign” and “some of your health risks”.

    The only sensible reading of that first paragraph is this : Your birth date determines your astrological sign, which is true by definition, not news, and useless except to charlatans and their feeble-minded victims. In complete contrast to this, your birth date may also determine something about your health risks, which is the result of a scientific study, news, and eminently useful to you and your doctor.

    You may have overreacted to a reference to astrology which merely acknowledges its existence and in no way endorses its veracity or utility. Overreaction to a non-threat can erode credibility.

  6. jockmcdock
    June 13, 2015 at 5:28 PM

    Bloody hell, mate. I’m a statistician and I never lied.

    Although I will confess that I’m now doing statistics with n==1.

  7. jockmcdock
    June 13, 2015 at 5:41 PM

    It would be interesting to see the analysis redone using astrological signs instead of months. I suspect the results would be similar. After all, “months” and “astrological signs” are both markers for “time of year you were born”. They are just out of phase by a couple of weeks.

    Is there a time of year effect? Quite possibly. Is it statistically significant? If it exists and n is big enough, yep. Is it medically significant. Nah.

    And, just for the record, I believe astrology is a flaming pile of BS.

    Sorry, Wang-Lo. I didn’t mean to reply to your post but to the thread in general.

  8. Cathy
    June 14, 2015 at 7:19 PM

    I find this interesting and wonder if some of the results would change in the southern hemisphere due to seasonal illnesses. That would possibly help validate the statistics.

  9. Bill T.
    June 17, 2015 at 2:41 AM

    now I have coffee coming out of my nose, thanks a ton, mate.

Comments are closed.