Births at Christmas are different: population based survey of 2 million deliveries

BMJ 1997; 315 doi: (Published 20 December 1997) Cite this as: BMJ 1997;315:1654
  1. Mads Melbye (mme{at}, professor and heada,
  2. Jan Wohlfahrt, statisticiana,
  3. Tine Westergaard, research fellowa,
  4. Anne Kristine Valeur Jensen, research fellowa,
  5. Anders Koch, research fellowa,
  6. Henrik Hjalgrim, research fellowa,
  7. Annemette Kristensen, research assistanta,
  8. Peter Aaby, professora
  1. a Department of Epidemiology Research, Danish Epidemiology Science Centre, Statens Serum Institut, Artillerivej 5, DK-2300 Copenhagen S, Denmark
  1. Correspondence to: Professor Melbye


The aim of the present study was to evaluate whether births occurring at Christmas are different from births taking place on other days of the year. We specifically wanted to test whether the pressure on the pregnant woman to have everything ready for the most important family feast of the year might increase her risk of premature labour. In Denmark, that feast takes place on the eve of 24 December.

Material and methods

For a previous study we obtained, from the Danish civil registration system, details of all births to women born between 1935 and 1978, including information on date of births, sex of children, and multiple births.1 For the present study we added information from the Danish national birth registry on gestational age and whether delivery was by caesarean section. Preterm births were defined as those with a gestational age of less than 37 weeks. All 2 005 096 births in the period 1 January 1960 to 30 September 1994 were considered in the statistical analysis.


The risk of giving birth on 24 December (Christmas Eve) in comparison to an average day of the year was reduced by 26% (observed=4068, expected= 2 005 096 births/365.25days; O/E=0.74, 95% confidence interval 0.72 to 0.76). Rates for the 3 days before Christmas Eve and the following 3 days were similar (0.84 and 0.76, respectively). This seemed to be due to a significantly lower number of term deliveries as the number of preterm deliveries was as expected (O/E=1.01; 0.84 to 1.21). As shown in the 1, this led to an overall 44% increased risk of a preterm compared to term delivery on Christmas Eve when comparing with other days of the year

The frequency of caesarean sections during the three days before Christmas Eve was as expected (OR=0.99, 0.91 to 1.09), but it was significantly lower than expected on Christmas Eve (OR=0.72, 0.60 to 0.87) and on the 3 following days (OR=0.76, 0.68 to 0.84).

Women under 20 years of age had close to the expected proportion of births at Christmas (0.25%), but this proportion fell to 0.22% in those aged 20–24 and was between 0.18% and 0.19% in older age groups (P for trend <0.001). The number of liveborn children was inversely associated with risk of a Christmas birth; this was explained primarily by a drop in risk between women without a previous birth and women with a history of one or more previous births (1). This association remained significant after age was adjusted for (P for trend <0.001). Births at Christmas and at other times did not differ with respect to the sex of the child and the proportion of multiple births.

Characteristics of births in the week around Christmas Eve and on Christmas Eve compared with births on other days of the year

View this table:


The increase in the proportion of preterm births around Christmas was entirely explained by a lower rate of other births during this period. Therefore, women in late pregnancy can safely take part in the often stressful preparations for Christmas.

Compared with what was expected, older women were significantly less likely than younger women to give birth around Christmas. This resulted in a particularly high proportion of firstborn children around Christmas but, as we note, without any sex difference.

It is not obvious to us that changes in the medical procedures around this holiday should explain the peculiar age distribution among delivering women. Both caesarean sections and induction are more often used in young rather than older women, and if anything the rate of caesarean sections decreased during Christmas. We lacked information on induction of labour, but it seems reasonable to assume that the pattern is similar. Could it be that older women who are already mothers more often than very young women plan their births to take place at other times of the year? We are eager to hear from readers with personal or professional experience regarding this matter.


Other members of the CPS Group were A M Nybo Andersen, M Andersen, P K Andersen, K Begtrup, B Høgh, H Jensen, K Mølbak, N M Nielsen, L L Winther Nielsen, S Olsen, and J B Pedersen.

Funding: Danish National Research Foundation.

Conflict of interest: None.


  1. 1.
View Abstract