I have heard many times before that people with mental disorders become hyperactive during the full moon. I have never believed it, but has there been any scientific study carried out to prove it one way or the other?

  • 1
    @BlueRaja Sounds like a prime example of confirmation bias. In fact, this is also debunked in the review I cited. Commented Jul 20, 2011 at 19:56
  • 1
    @BlueRaja: Do you have a theory, why they get more rowdy/hyperactiv (user21229)? Because of the light? Don't they use electric light? Do they show the same symptoms at daylight and less so in wintertime, when it is darker in general? If not the light - do you have similar questions about our Sun, Jupiter, Venus and Mars? Why not? BlueRaja: Being sceptic doesn't make you immune against hokuspokus, astrology, religion or racism. Commented Jul 21, 2011 at 0:42
  • 1
    @BlueRaja: One of the wikipedia-links talks about women, which have a 29.5 +/- 1 day cycle. Wikipedia about moon talks about 27.3 days for a moon cycle. How can we explain the 1-3 days gap, which leads to an async of a half moon phase after 5 to 15 months? Or does +/- 1 day mean, that the cycle varies per woman, once it it 27 days, the next month 29 days? Still not in sync. Commented Jul 21, 2011 at 2:48
  • 1
    A significant amount of woman can only constantly experience a full moon in a specific part of their cycle, if they are in sync with the moon, or if their cycle is a whole multiple of the moon cycle, or vice versa. If some woman experience a specific phase, and have a slightly longer or shorter cycle, they will not experience this phase the next few months, until they meet again. With an offset of 2 days, it's about 14 times, with a 3-day-offset it is 9 days. Simple, pure math. Commented Jul 21, 2011 at 19:42

1 Answer 1


Such studies actually exist. The Skepdic entry for Full moon and lunar effects gives a full list of things that have been studied, but in particular there was insufficient evidence to support that the following correlate with a full moon (that means that such a correlation doesn’t exist):

  • psychiatric admissions [one study found admissions were lowest during a full moon]
  • behavioral outbursts of psychologically challenged rural adults
  • epilepsy

All of which are different effects of mental illness. The list comes from the 1996 review by Kelly, I. W., James Rotton, and Roger Culver, The Moon was Full and Nothing Happened: A Review of Studies on the Moon and Human Behavior and Human Belief

  • Insufficient evidence to support a correlation can put limits on how large a correlation can be, but it can't rule out a small correlation. Commented Jul 21, 2011 at 2:03
  • A small correlation to what? To the moonlight? What other influence can we think of? Commented Jul 21, 2011 at 2:52
  • 1
    @David If the hypothesis is “there is a statistically significant correlation” and the numbers don’t support this hypothesis, then it is rejected: no correlation is found. There is no such thing as a “large” or “small” correlation. In particular, we have here an experiment with two possible outcomes – yes and no – not a continuous or multi-value outcome so there is only one possible correlation. If you mean something else, please motivate this statistically. Commented Jul 21, 2011 at 8:28
  • @Konrad: I agree with your statement but the rejected correlations are only those studied. There may be other correlations like (imaginary example) full moon and depression or suicide attempts. If it hasn't been studied, then we can neither approve or reject it. Commented Nov 29, 2012 at 21:50
  • @ypercube Of course not. But we can (and must!) use prior probabilities, even though we do not know their exact values (this is a direct application of Occam’s razor). The prior probability of such phenomena is lower than their non-existence: the default position is to be skeptical of them until evidence for them emerges. Like Russell’s teapot, we preliminarily reject any unfounded hypothesis. Commented Nov 29, 2012 at 21:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .