Parce que ma moitié, dont le scepticisme au sens scientifique du terme n'est pas la qualité première, pense qu'il y a p'têt ben davantage de naissances les jours de pleine lune, je suis parti à la recherche de publis sur le sujet. Sur les onze articles listés ci-dessous, quatre concluent positivement quant à une corrélation statistique entre les phases lunaires et l'évolution du nombre de naissances. Sur ces quatre là, les articles 5 et 7 sont du même auteur (G. Ghiandoni), et l'article 5 est le seul à mentionner explicitement un lien avec le jour de pleine lune ... Le 9e article, qui ne voit pas d'influence, pèse un peu plus lourd que les autres puisqu'il s'agit d'une méta-étude sur 21 publications.
Le 11e article, qui évoque un lien avec les phases lunaires, semble le plus convaincant, non parce qu'il a tenu compte de ma naissance, mais bien par l'étendue de l'échantillon puisque portant sur près de 6 millions. Il met en évidence un rythme hebdomadaire (davantage de naissances le mardi, et moins le dimanche ... un lien avec les dispos des médecins qui accouchent, peut-être ? ), un rythme annuel (un creux en septembre-octobre, un pic en mai ... ah tiens, un rapport éventuel avec la date de conception, les périodes de vacances estivales seraient favorables ?), et enfin un rapport statistiquement significatif avec les quartiers de la lune (un creux dans le premier quartier, un pic entre le dernier quartier et la pleine lune).

1. The effect of the lunar cycle on frequency of births and birth complications.
Am J Obstet Gynecol. 2005 May;192(5):1462-4.
OBJECTIVE: The purpose of this study was to examine the influence of the lunar cycle on the frequency of deliveries and/or delivery complications.
STUDY DESIGN: This was a retrospective cohort, secondary analysis of 564,039 births across 62 lunar cycles that were identified from North Carolina birth certificate data from 1997 to 2001.
CONCLUSION: An analysis of 5 years of data demonstrated no predictable influence of the lunar cycle on deliveries or complications.

2.A lunar effect on fertility
Soc Biol. 1981 Spring-Summer;28(1-2):75-80.
Data from individual birth records for 140,000 live births occurring in New York City in 1968 provide evidence that the folkloric beliefs in the moon's influence on human reproduction found in many societies may have a factual basis.
A small but systematic variation of births over a period of 29.53 days, the length of the lunar cycle, with peak fertility at 3rd quarter, was found in the 4 independent time series and remained after weekly variation was removed. The finding itself provides little guidance as to the timing or nature of the moon's influence, but recent research suggests a connection between menstrual regularity and light. The timing of the fertility peak at 3rd quarter suggests that the period of decreasing illumination immediately after full moon may precipitate ovulation.

3. Lunar phases and incidence of spontaneous deliveries. Our experience
Minerva Ginecol. 1994 Jul-Aug;46(7-8):429-33.
STUDY DESIGN : we examined 7842 spontaneous deliveries at Obstetric and Gynaecologic Clinic of University of Florence, between January 1988 and November 1992, covering 58 synodic lunar months.
RESULTS : non significant differences were found in the incidence of spontaneous birth throughout the lunar cycle.

4. The influence of lunar phases on the frequency of deliveries
Minerva Ginecol. 1991 Jul-Aug;43(7-8):359-63.
We studied the frequency of home deliveries of Maputo (Mozambique), without any medical assistance according to lunar cycle. A study of 5226 births in 37 lunar cycles didn't show a significative increase of deliveries during specific lunar phases and week days.

5. Does lunar position influence the time of delivery? A statistical analysis
Eur J Obstet Gynecol Reprod Biol. 1998 Mar;77(1):47-50.
STUDY DESIGN: Retrospective analysis of 1248 spontaneous full-term deliveries in three-year period (36 lunar months), setted at Department of Obstetrics and Gynaecology, Civil Hospital, Fano (Marche, Italy), using circular statistics techniques.
RESULTS: A connection between the distribution of spontaneous full-term deliveries and the lunar month was found. The effect of the phases of the moon seems to be particularly relevant in multiparae and plurigravidae; in these cases, the mean day of delivery corresponds to the first or second day after the full moon.
CONCLUSIONS: In this paper the effect of lunar phases on the time of delivery is shown. This influence seems to be especially relevant in the case of multiparae and plurigravidae. Nevertheless, it is too weak to allow for prediction regarding the days with the highest frequency of deliveries.

6. Human births and the phase of the moon
N Engl J Med. 1979 Jan 11;300(2):96.
The distribution of all births during 51 lunar cycles, from March 17, 1974, to April 30, 1978, was analyzed by the authors at the University of California, Los Angeles, Hospital. There were 11.691 live births, of which 8142 were natural, 141 multiple, and 168 stillbirths. In none of the 4 samples was the mean number of births occurring on the date of the full moon above average, showing that the birthrate during the period surveyed did not in any way correlate with the cycle of lunar phases.

7. Some unexpected results in time distribution analysis of spontaneous deliveries
Gynecol Obstet Invest. 1998 Aug;46(2):88-90.
METHODS: A retrospective analysis of 1,791 spontaneous deliveries over a 4-year period (1993-1996) was made at the department of Obstetrics and Gynaecology, Civil Hospital, Fano, Italy.
RESULTS: Some unexpected infradian disuniformities (with 60- and 90-hour periods) were observed. Moreover, a correlation between the distribution of spontaneous deliveries and the lunar phases was found.

8. The moon and delivery
Rev Enferm. 2004 Nov;27(11):7-9, 11-2.
The differences found in the distribution of deliveries over the four lunar phases, along with the comparison of the means and the comparison of the number of deliveries on the central day in each phase are not statistically significant. The different phases in the lunar cycle and especially the full moon do not appear to have any influence over the distribution of deliveries in this study.

9. Lunar phase and birthrate: an update
Psychol Rep. 1994 Aug;75(1 Pt 2):507-11.
Martens, Kelly, and Saklofske in 1988 examined 21 studies considering the possible relationship between lunar periodicities and birthrate. They reported that the majority of studies uncovered no relationship and that the positive studies were inconsistent in their findings. The present update reports on six additional studies on birthrate and lunar periodicities from five different countries. None of these studies produced evidence of lunar periodicities consistent with folklore or some previous studies.

10. Birth rate and its correlation with the lunar cycle and specific atmospheric conditions
Am J Obstet Gynecol. 2005 Jun;192(6):1970-3.
STUDY DESIGN: Birth records were limited to spontaneous vaginal deliveries, 37 to 40 weeks' gestation, in Phoenix, between 1995 and 2000 (n = 167,956). Daily birth counts were merged with daily surface weather statistics from the National Weather Service for Sky Harbor Airport, and records of lunar phase for the same period.
RESULTS: The analyses revealed no significant correlates of birth rate.

11. Births, fertility, rhythms and lunar cycle. A statistical study of 5,927,978 births
J Gynecol Obstet Biol Reprod (Paris). 1986;15(3):265-71.
We have looked at 5,927,978 French births occurring between the months of January 1968 and the 31st December 1974. Using Fourier's spectral analysis we have been able to show that there are two different rhythms in birth frequencies:
-a weekly rhythm characterised by the lowest number of births on a Sunday and the largest number on a Tuesday
-an annual rhythm with the maximum number of births in May and the minimum in September-October.
A statistical analysis of the distribution of births in the lunar month shows that more are born between the last quarter and the new moon, and fewer are born in the first quarter of the moon. The differences between the distribution observed during the lunar month and the theoretical distribution are statistically significant.