full moon and lunar effects

"...spooky effects have been ascribed to the phases of the moon....But when the statistics are redone properly, the correlation with lunar phase always evaporates....Yet many sensible people—including police officers and emergency room staff—continue to believe otherwise." --Steven Strogatz

The full moon has been linked to crime, suicide, mental illness, disasters, accidents,  birthrates, fertility, and werewolves, among other things. Some people even buy and sell stocks according to phases of the moon, a method probably as successful as many others. Numerous studies have tried to find lunar effects. So far, the studies have failed to establish much of interest. Lunar effects that have been found have little or nothing to do with human behavior, e.g., the discovery of a slight effect of the moon on global temperature,* which in turn might have an effect on the growth of plants. Of course, there have been single studies here and there that have found correlations between various phases of the moon and this or that phenomenon, but nothing significant has been replicated sufficiently to warrant claiming a probable causal relationship.

Ivan Kelly, James Rotton and Roger Culver (1996) examined over 100 studies on lunar effects and concluded that the studies have failed to show a reliable and significant correlation (i.e., one not likely due to chance) between the full moon, or any other phase of the moon, and each of the following:

-the homicide rate
-traffic accidents
-crisis calls to police or fire stations
-domestic violence
-births of babies
-suicide
-major disasters
-casino payout rates
-assassinations
-kidnappings
-aggression by professional hockey players
-violence in prisons
-psychiatric admissions [one study found admissions were lowest during a full moon]
-agitated behavior by nursing home residents
-assaults
-gunshot wounds
-stabbings
-emergency room admissions [but see]
-behavioral outbursts of psychologically challenged rural adults
-lycanthropy
-vampirism
-alcoholism
-sleep walking
-epilepsy

If so many studies have failed to prove a significant correlation between the full moon and anything, why do so many people believe in these lunar myths? Kelly, Rotton, and Culver suspect four factors: media effects, folklore and tradition, misconceptions, and cognitive biases. A fifth factor should be considered, as well: communal reinforcement.

the media perpetuate lunar myths

Lunar myths are frequently presented in films and works of fiction. "With the constant media repetition of an association between the full moon and human behavior it is not surprising that such beliefs are widespread in the general public" (Kelly et al. 1996). Reporters also "favor those who claim that the full moon influences behavior." It wouldn't be much of a story if the moon was full and nothing happened, they note. Anecdotal evidence for lunar effects is not hard to find and reporters know that one good anecdote trumps ten scientific studies when it comes to reader interest, even though such evidence is unreliable for establishing significant correlations. Relying on personal experience ignores the possibility of self-deception and confirmation bias.  Such evidence may be unreliable, but it is nonetheless persuasive.

folklore and tradition

Many lunar myths are rooted in folklore. For example, an ancient Assyrian/Babylonian fragment stated that "A woman is fertile according to the moon." Such notions have been turned into widespread misconceptions about fertility and birthrates. For example, Eugen Jonas, a Slovakian psychiatrist, was inspired by this bit of folklore to create a method of birth control and fertility largely rooted in astrological superstitions. The belief that there are more births during a full moon persists today among many educated people. Scientific studies, however, have failed to find any significant correlation between the full moon and number of births (Kelly and Martens 1994; Martens et al.1988 ). In 1991, Benski and Gerin reported that they had analyzed birthdays of 4,256 babies born in a clinic in France and "found them equally distributed throughout the synodic (phase) lunar cycle" (Kelly, et al. 1996: 19). In 1994, Italian researchers Periti and Biagiotti reported on their study of 7,842 spontaneous deliveries over a 5-year period at a clinic in Florence. They found "no relationship between moon phase and number of spontaneous deliveries" (ibid.).

Despite the fact that there is no evidence of a significant correlation between phases of the moon, the menstrual cycle, and fertility, some people not only maintain that there is, they have a "scientific" explanation for the non-existent correlation.* Some think the light of the moon affects fertility in women, the way it does in corals. The light of the moon is a very minor source of light in most women's lives, and is no more likely than the moon's gravitational force to have a significant effect on a woman's ovulation. Furthermore, the average menstrual cycle is 28 days but varies from woman to woman and month to month, while the length of the lunar month is a consistent 29.53 days.* Some of us have noticed that these cycles are not identical. Furthermore, it would seem odd that natural selection would favor a method of reproduction for a species like ours that depended on the weather. Clouds are bound to be irregularly and frequently blocking moonlight, which would seem to hinder rather than enhance our species' chance for survival.

Some mythmakers believe that long ago women all bled in sync with the moon, but civilization and indoor electric lighting (or even the discovery of fire by primitive humans) have disturbed their rhythmic cycle. This theory may seem plausible until one remembers that there are quite a few other mammals on the planet that have not been affected by firelight or civilization's indoor lighting and whose cycles aren't in harmony with the moon. In short, given the large number of types of mammals on our planet, one would expect that by chance some species' estrus and menstrual cycles would harmonize with lunar cycles (e.g., the lemur). It is doubtful that there is anything of metaphysical significance in this.

What we do know is that there has been very little research on hormonal or neurochemical changes during lunar phases. James Rotton's search of the literature "failed to uncover any studies linking lunar cycles to substances that have been implicated as possible correlates of stress and aggression (e.g., serotonin, melatonin, epinephrine, norepinephrine, testosterone, cortisol, vasopressin [directly relevant to fluid content], growth hormone, pH, 17-OHCS, adrenocrotropic hormone [? adrenocorticotropic hormone?])" (Rotton 1997). One would think that this area would be well-studied, since hormones and neurochemicals are known to affect menstruation and behavior.

misconceptions

Misconceptions about such things as the moon's effect on tides have contributed to lunar mythology. Many people seem to think that since the moon affects the ocean's tides, it must be so powerful that it affects the human body as well. The lunar force is actually a very weak tidal force. A mother holding her child "will exert 12 million times as much tidal force on her child as the moon" (Kelly et al., 1996: 25). Astronomer George O. Abell claims that a mosquito would exert more gravitational pull on your arm than the moon would (Abell 1979). Despite these physical facts, there is still widespread belief that the moon can cause earthquakes.* It doesn't; nor does the sun, which exerts much less tidal force on the earth than the moon.

The fact that the human body is mostly water largely contributes to the notion that the moon should have a powerful effect on the human body and therefore an effect on behavior. It is claimed by many that the earth and the human body both are 80% water. This is false. Eighty percent of the surface of the earth is water. Furthermore, the moon only affects unbounded bodies of water, while the water in the human body is bounded.

Also, the tidal force of the moon on the earth depends on its distance from earth, not its phase. Whereas the synodic period is 29.53 days, it takes 27.5 days for the moon to move in its elliptical orbit from perigee to perigee (or apogee to apogee). Perigee (when the moon is closest to earth) "can occur at any phase of the synodic cycle" (Kelly et al. 1990: 989). Higher tides do occur at new and full moons, but not because the moon's gravitational pull is stronger at those times. Rather, the tides are higher then because "the sun, earth, and moon are in a line and the tidal force of the sun joins that of the moon at those times to produce higher tides" (ibid.: 989).

Many of the misconceptions about the moon's gravitational effect on the tides, as well as several other lunar misconceptions, seem to have been generated by Arnold Lieber in The Lunar Effect (1978), republished in 1996 as How the Moon Affects You. In The Lunar Effect, Lieber incorrectly predicted a catastrophic earthquake would hit California in 1982 due to the coincidental alignment of the moon and planets. Undeterred by the fact that no such earthquake had occurred, Lieber did not admit his error in the later book. In fact, he repeated his belief about the dangers of planet alignments and wrote that they "may trigger another great California earthquake." This time he didn't predict when.

cognitive biases and communal reinforcement

Many believe in lunar myths because they have heard them repeated many times by members of the mass media, by police officers, nurses, doctors, social workers, and other people with influence. Once many people believe something and enjoy a significant amount of communal reinforcement, they get very selective about the type of data they pay attention to in the future. If one believes that during a full moon there is an increase in accidents, one will notice when accidents occur during a full moon, but be inattentive to the moon when accidents occur at other times. If something strange happens and there is a full moon at the time, a causal connection will be assumed. If something strange happens and there is no full moon, no connection is made, but the event is not seen as counterevidence to the belief in full moon causality. Memories get selective, and perhaps even distorted, to favor a full moon hypothesis. A tendency to do this over time strengthens one's belief in the relationship between the full moon and a host of unrelated effects.

the moon, madness and suicide

Probably the most widely believed myth about the full moon is that it is associated with madness. However, in examining over 100 studies, Kelly et al. found that "phases of the moon accounted for no more than 3/100 of 1 percent of the variability in activities usually termed lunacy" (1996: 18). According to James Rotton, "such a small percentage is too close to zero to be of any theoretical, practical, or statistical interest or significance" (Rotton 1997).

Finally, the notion that there is a lunar influence on suicide is also unsubstantiated. Martin et al. (1992) reviewed numerous studies done over nearly three decades and found no significant association between phases of the moon and suicide deaths, attempted suicides, or suicide threats. In 1997, Gutiérrez-García and Tusell studied 897 suicide deaths in Madrid and found "no significant relationship between the synodic cycle and the suicide rate" (p. 248). These studies, like others which have failed to find anything interesting happening during the full moon, have gone largely unreported in the press.

postscript: There are likely to be many studies in the future that find a positive correlation between some lunar phase and some human behavior (or process affected by human behavior, such as the stock market). Remember to consider a few caveats: correlation doesn't establish causation; studies that are well designed still need to be replicated before they are accepted as not being flukes; some studies with positive results will suffer from design flaws or methodological errors. For example, in 2005 Yuan, Zheng, and Zhu found "that stock returns are lower on the days around a full moon than on the days around a new moon. The magnitude of the return difference is 3% to 5% per annum based on analyses of two global portfolios: one equal-weighted and the other value-weighted." Whether this is a lunar effect remains to be seen. The study needs to be replicated with a significant number of data points.

update Feb 1, 2000: According to Allan Hall of the Sunday-Times, German researchers Hans-Joachim Mittmeyer of the University of Tübingen and Norbert Filipp from the Health Institute of Reutlingen claim that "a study of police reports for 50 new and full Moon cycles" shows that the moon is "responsible for binge drinking."

According to Hall, Mittmeyer and Filipp claim in their paper "Alcohol Consumption and the Moon's Influence" to have studied police arrest reports and blood-alcohol tests of 16,495 people and Mittmeyer said "The results show there is a definite correlation between new and full Moons and the amount of alcohol consumed." 

Hall writes:

More of those with an excess of 2ml of alcohol per 100ml of blood inside them - drunk, according to German law - were caught by police during the five-day full Moon cycle.

On average 175 drink-drivers per day were caught in two German states two days before a full Moon, 161 were caught during the full Moon cycle and the figure dropped to about 120 per day at other times.

Jan Willem Nienhuys, a mathematician in the Eindhoven (Netherlands) University of Technology, claims that "Hall's story is a garbled version of a story by the German Press Agency DPA." According to Nienhuys, Hall invented the notion of a five-day full Moon cycle; the expression is not used by Mittmeyer and Filipp in their paper. Furthermore, 668 of the 16,495 arrested and tested were found to be sober, leaving 15,827 with alcohol in their blood, but only 4,512 with more than 0.2 percent blood alcohol (i.e., drunk). 

According to Nienhuys, the 161 figure refers to the average number of drunk drivers arrested on any given date in the lunar month; he believes this number was arrived at by dividing 4,512 by 28 (rather than 29.53, the length of a lunar month) and hence should be 153, not 161. About the only thing Hall got right, says Nienhuys, is that Mittmeyer and Filipp do claim to have found a significant correlation between the moon and excessive drinking. He notes that the pair provide graphs but no statistical analysis of their data. When such an analysis is done, says Nienhuys, one discovers that the study is "pompous pseudoscience." According to Nienhuys, a standard statistical test yields p-values which show that there is nothing to investigate.

Here are the data, according to Nienhuys. Day 0 is the day of the new moon and day 14 is the full moon.

day      drunks   drinkers, including drunks
0        145      551
1        160      528
2        162      552
3        122      527
4        162      538
5        157      531
6        156      504
7        158      560
8        140      523
9        152      540
10       150      552
11       146      477
12       173      563
13       150      545
14       150      523
15       149      498
16       145      543
17       142      539
18       143      507
19       119      508
20       157      532
21       163      552
22       156      513
23       148      530
24       154      528
25       158      536  
26       175      582
27       176      581
28       169      590
---------------------
        4437    15553
missing   75      274
       

---------------------
        4512    15827

The three big days were the 12th, 26th and 27th. You figure it out!

(Nienhuys's article, entitled "Triply garbled tripe" is being prepared for publication. He was kind enough to send me a pre-publication copy of the paper.)

See also control study and post hoc fallacy.

reader comments

further reading

books and articles

Abell, George. "The Alleged Lunar Effect" in Science Confronts the Paranormal, edited by Kendrick Frazier. (Buffalo, N.Y.: Prometheus Books, 1986). Abel provides a very critical review of psychiatrist Arnold L. Lieber's The Lunar Effect: Biological Tides and Human Emotions.

Abell, George O.  "The moon and the birthrate," Skeptical Inquirer, Summer 1979, vol. 3, no. 4.

Byrnes, Gail and I. W. Kelly. "Crisis Calls and Lunar Cycles: A Twenty-Year Review," Psychological Reports, 1992, 71, 779-785.

Campbell, David E. and John L. Beets Lunacy and the moon. Psychological Bulletin, Volume 85, Issue 5, September 1978, Pages 1123-1129. (It is concluded that lunar phase is not related to human behavior and that the few positive findings are examples of a Type I error, i.e., seeing patterns where there aren't any.)

Gutiérrez-García, J. M. and  F. Tusell. "Suicides and the Lunar Cycle," Psychological Reports, 1997, 80, 243-250.

Hines, Terence. Pseudoscience and the Paranormal 2nd ed. (Buffalo, NY: Prometheus Books, 2003).

Iosif, Alina and Bruce Ballon. 2005. Bad Moon Rising: the persistent belief in lunar connections to madness. CMAJ, 173, 1498-1500.