Abstract

The number of births varies markedly by season, but the causes of this variation are not well understood. The proposed explanations include temperature or photoperiod (affecting hormonal concentrations, sperm quality or sexual activity), seasonal variation in pregnancy loss, or cultural factors. In this paper we examined whether birth seasonality is influenced by socio-demographic factors. We used data on all live births registered in the Czech Republic in 1989–1991 (n = 387 496). Differences in the degree of seasonality between socio-demographic groups (defined by maternal age, marital status, education and birth order) were examined by inspection of curves, by comparing coefficients of variations of monthly numbers of births, and by calculating the ratios of the number of births in the 3 peak months (March to May) to the number of births in the 3 lowest months (October to December). We found large differences in the size of the seasonal variation in births by socio-demographic factors. The seasonal variation was highly pronounced in mothers who were 25–34 years old, had higher education, were married, and were pregnant with their second or third child. By contrast, birth seasonality was weak in mothers who were ≤19 years or ≥35 years old, unmarried, had low education, and expected their first or fourth or higher order birth. In a multivariate model, all four socio-demographic variables contributed significantly to seasonal variation. These results suggest that the seasonality of births is, at least in this population, strongly influenced by socio-demographic factors.

Introduction

In most populations with reliable data, birth rates vary by the season of the year. The seasonality of births is not identical in all populations. In Northern Europe, for example, most births occur in Spring (say March to May) while birth rates are lowest in the Autumn (October to November); in the United States, by contrast, most births occur in Summer and early Autumn (July to September), and the minimum is in Spring (March to May) (Lam and Miron, 1994).

Although a number of explanations have been put forward, the causes of these seasonal variations are not fully understood. Perhaps most attention is given to physical environmental characteristics, such as temperature and light (Roenneberg and Aschoff, 1990a, b; Lam and Miron, 1991, 1996). It has been suggested that temperature and light, possibly via hormonal changes, may affect the quality of semen, frequency of coitus, or length of menstrual cycle, and thus the ability to conception (Jongbloet, 1983; Rojansky et al., 1992; Centola and Eberly, 1999; Gyllenborg et al., 1999). It has also been proposed that fetal loss, if seasonally dependent, may underlie the seasonality of births (Jongbloet, 1983; Weinberg et al., 1994). Among cultural factors, seasonal changes of the rate at which women enter (or leave) the population at risk may play a role. For example, seasonality in marriages would affect the number of women at risk of conception, and thus result in seasonal changes in birth rates. This has been shown for a number of traditional populations (Fialova, 1995; Stolwijk et al., 1996). In modern societies, however, it is likely that the probability of conception depends more on the choice of the time of pregnancy (probably related to the use of contraception) than on climate or duration of marriage.

The vast majority of studies of birth seasonality so far examined aggregate data for whole populations. Several early studies used data aggregated by a social indicator or race within one population; the results suggested that seasonality of birth differs by social class (Pasamanick et al., 1960; Zelnik, 1969; Erhardt et al., 1971; James, 1971; Chaudhury, 1972; Warren and Tyler, 1979), although the pattern of these differences was not uniform. However, individual data may prove more useful to study the independent effects of different socio-demographic factors. If the intentional timing of pregnancies is responsible for a part of the seasonal variation, it is plausible to assume that some social groups within populations would be more successful than others; this would lead to differently pronounced seasonality in different social and demographic groups within a given population. We tested this hypothesis on data collected by the Czech national birth register in 1989–1991.

Materials and methods

We used data on all 387 496 live births registered in the Czech Republic in the period 1989–1991. The register contains individual records of all births, with indicators of socio-economic status of the mother and basic information on the newborn infant (Koupilova et al., 1998). The following characteristics of the mother were used in this study: education, marital status, and age. Mothers were classified into four categories of education: primary, vocational, secondary (corresponding to A-levels), and university. We used two categories of marital status: married and unmarried (which included never married, divorced, and widowed). Maternal age was classified into the following age groups: ≤19, 20–24, 25–29, 30–34, and ≥35 years. Birth order was categorized into four groups: 1st, 2nd, 3rd and 4th or higher.

Seasonality in births by socio-demographic group was first examined by visual inspection of the seasonal trends in the monthly ratios of observed/expected numbers of births (the numbers were corrected for different number of days by month). As the results were identical in all calendar years, the data were pooled. We quantified the extent of the seasonal variation as follows. First, we calculated the coefficient of variation by the month of birth. For each month, the number of births was recorded and the variation coefficient between the 12 months was calculated, for all the births and separately according to socio-demographic characteristics. Second, in each socio-demographic category, we calculated the ratio of the number of births in the month with the most and least births, and the ratio of the numbers of births in 3 months with most births (March to May) with the 3 months with the smallest number of births (October to December). Finally, we examined the independent contribution of the individual factors to the seasonal variation by building a logistic regression model on data restricted to high and low birth rates periods (March to May and October to December). The outcome variable was period of birth (being born in March to May was coded as 1 and being born in October to December as 0), and maternal age, education, marital status and birth order were the independent variables.

Results

The monthly ratios of observed/expected numbers of birth by calendar year are shown in Table I and by socio-demographic characteristics in Table II. There was a clear seasonal variation in all births, with most births occurring in March to May and minimum birth rates in October to December (Table I). Figures 1–4 show the numbers of births by month by maternal characteristics. The seasonal curve was relatively flat for births of mothers ≤19 years and ≥35 years old, and it was highly pronounced in mothers aged 25–29 and 30–34 years (Figure 1). There was a large difference in seasonality by marital status: it was highly pronounced among married but almost absent among unmarried women (Figure 2). Maternal education had a large influence on the month of births; the seasonal variation was minimal among mothers with primary education and highly pronounced among mothers with university education (Figure 3). There were also large differences by birth order—seasonal variation was smallest for first-born infants and largest for children born as second or third (Figure 4). In some groups there was a minor secondary peak in September.

The seasonal variation by socio-demographic group is quantified in Table III. Both indicators of the seasonal variation, the coefficient of variation and the ratio of the numbers of birth in the trimester with the largest versus the lowest number of births, confirm the pattern seen in Figures 1–4. For example, while among the youngest mothers the numbers of births in the trimesters with the least and most births were similar, among mothers aged 25–34 years there were ~30% more births in March to May than in October to December. Differences of a similar magnitude were observed by maternal marital status, maternal education and birth order. In the multivariate logistic regression model, all variables were significantly associated with seasonality when simultaneously entered in one model, but all coefficients were smaller than in bivariate models (Table III).

Discussion

We found that the magnitude of the seasonal variation in births was strongly associated with maternal socio-demographic characteristics. The seasonality of births was most pronounced in mothers who were 25–34 years old, married, better educated, and expected their second or third child. The birth rate peaked in March to May, which roughly corresponds to conceptions in summer, during the main holidays. There was only a weak seasonal variation among mothers who were either very young or ≥35 years, unmarried, with low education, and were pregnant with their first or fourth or higher order child.

The results are unlikely to be due to random error. We used complete data for the whole population; the numbers of births in the analyses were large, the seasonal differences were clearly pronounced and formed a consistent pattern across different socio-demographic characteristics. The results are also unlikely to be due to low quality data. The Czech national birth register is considered complete and reliable (Koupilova et al., 1998); more importantly, it is difficult to envisage a systematic error in birth registration that would produce the observed pattern of seasonality. Confounding by physical environmental factors is unlikely, as the whole population was exposed to the same degree. The pronounced seasonality in mothers pregnant with their second or third child suggests that timing of the marriage did not play a major role.

Our results are consistent with data from Britain in the 1960s in which birth seasonality was more pronounced in higher social classes (James, 1971). Seasonality pattern in France was also similar to our results but data from Italy and The Netherlands were not (Prioux, 1988). The US data are largely inconsistent with the pattern found in the Czech Republic. A study in Georgia found that social variation in births increased with lower social status (Warren and Tyler, 1979). In New York seasonality was more pronounced in non-whites and in illegitimate births (Erhardt et al., 1971). Several analyses of data from Baltimore produced mixed results (Pasamanick et al., 1960; Zelnik, 1969; Chaudhury, 1972), perhaps because of the relatively small number of subjects in these analyses. The differences in seasonal patterns and their relation to socio-demographic factors both between areas with similar climatic conditions and over time within populations (Hoffmann and Kawiani, 1976; Prioux, 1988; Lam and Miron, 1994b) seem to suggest that birth seasonality is influenced by place- and time-specific factors. Erhardt et al. suggested that such factors include psychological, behavioural and social influences (Erhardt et al., 1971).

This proposition may appear at odds with the body of evidence suggesting that there is a biological basis for birth seasonality (Jongbloet, 1983; Kallan and Udry, 1989; Smits et al., 1998). However, the two explanations, biological and social, may be compatible. First, it is important to distinguish between different concepts related to childbearing: fecundability (the physiological ability to conceive), fetal loss, and fertility (the realization of the potential to reproduce) (Golden and Millman, 1993). By analysing the distribution of births we study fertility. Factors that influence fecundability and fetal loss can influence fertility, but some factors that influence fertility, such as factors related to personal choice and behaviour, do not affect fecundity or fetal loss. Fecundability in humans appears to fluctuate by season (Kallan and Udry, 1989; Smits et al., 1998), although this finding may not be universal (Stolwijk et al., 1996).

Factors that influence fecundity (e.g. temperature or light) can thus play a role in birth seasonality. However, it is likely that the effect of biological factors on seasonality in birth in developed countries is weaker, because in these countries fertility is largely determined by factors related to individual choice (Bongaarts, 1978). In populations with fertility levels much lower than natural fertility (because of the wide use of contraception), and at the time when most of the world population experience fertility below the replacement level (2.1 children per women), it is evident that individual choices are particularly important. At the population level, factors related to pregnancy choice and planning would override factors that influence the physiological ability to reproduce (Erhardt et al., 1971). Although there is some seasonal variation in fetal loss, it is probably too small to be a major cause of seasonality of birth (Warren et al., 1986; Weinberg et al., 1994).

Second, there may be an interaction between physical environmental factors and socio-demographic factors. Although climatic conditions in the Czech Republic are mild, and there are no known socio-economic differences in the quality of housing or exposure to outdoor factors, it is possible that the effect of meteorological factors are modified by the social environment.

The observation that seasonality is differently pronounced in different population groups may be attributed to a combination of several factors (that may interact with biological influences on fecundability). The differences in birth seasonality between socio-demographic groups can be due to differences in family planning. The weak seasonal variation in women who were ≤19 or ≥35 years old, or unmarried, suggests less well or less successfully planned pregnancies. This proposition is plausible. Young women are more vulnerable to poor family planning. The vast majority of Czech unmarried women who gave birth were, at the time of the study, without a stable male partner; this situation is, again, suggestive of poor family planning. The absence of seasonality among mothers ≥35 years old may be related to smaller importance given to the time of conception (or birth) and the higher risk of fetal loss at higher maternal ages (Weinstein et al., 1993; Nybo Andersen et al., 2000). By contrast, better educated women in stable marriages, with one or two children already, are more likely to plan their next pregnancy. We do not have data on contraception use by all maternal characteristics used in this study, but there is a pronounced educational gradient in Czech women (unpublished data).

Summer holidays may be the best time to attempt a conception (James, 1971); Basso et al. reported that summer was the preferred time for starting pregnancy in several European countries (Basso et al., 1995). Consistent with this explanation, the major peak in births in our data corresponds to conceptions in the summer and the minor peak (in September) to conceptions around the Christmas holiday. Preference of the date of birth/conception may have also played a role. It is possible that different social groups may have different preferences of the date of conception or birth. Our data do not allow firm conclusions on the contribution of different aspects. In either case, the social and demographic factors seem to play an important role in shaping the seasonal variation in births in this population.

It is possible that in different countries different socio-demographic factors predict birth seasonality differently, perhaps because the interactions between the biological and social variables are not constant across time and place. This question can be addressed by birth registration data that are available in most developed countries. The association between socio-demographic factors and birth seasonality, if confirmed, may be useful for further studies of seasonality in reproductive outcomes, but can also provide the link between the month of birth and subsequent fertility (Smits et al., 1999) or mortality (Doblhammer, 1999).

Table I.

Absolute and observed/expecteda numbers of births by month and year

Month of birth All births 1989 1990 1991 
 Abs. No. Obs/Expa Abs. No. Obs/Expa Abs. No. Obs/Expa Abs. No. Obs/Expa 
aThe observed/expected monthly numbers of births (×100) were calculated after correction for the unequal numbers of days. 
Abs. = Absolute; Obs/Exp = Observed/Expected. 
January 32 533  98.8 10 622  97.6 10 492  94.8 11 419 104.1 
February 30 774 103.5 10 062 102.4 10 069 100.7 10 643 107.4 
March 35 416 107.6 11 859 109 11 691 105.6 11 866 108.2 
April 34 515 108.3 11 555 109.7 11 218 104.7 11 742 110.6 
May 35 255 107.1 11 782 108.3 11 600 104.8 11 873 108.2 
June 33 651 105.6 11 229 106.6 11 133 103.9 11 289 106.3 
July 33 651 102.2 11 312 103.9 11 196 101.1 11 143 101.6 
August 32 924 100 11 012 101.2 11 229 101.4 10 683  97.4 
September 31 457  98.7 10 223  97.1 10 711 100 10 523 99.1 
October 30 072  91.3  9797  90 10 620  95.9  9655  88 
November 28 222  88.6  9115  86.6 10 063  93.9  9044  85.2 
December 29 026  88.2  9539  87.7 10 300  93  9187  83.8 
Month of birth All births 1989 1990 1991 
 Abs. No. Obs/Expa Abs. No. Obs/Expa Abs. No. Obs/Expa Abs. No. Obs/Expa 
aThe observed/expected monthly numbers of births (×100) were calculated after correction for the unequal numbers of days. 
Abs. = Absolute; Obs/Exp = Observed/Expected. 
January 32 533  98.8 10 622  97.6 10 492  94.8 11 419 104.1 
February 30 774 103.5 10 062 102.4 10 069 100.7 10 643 107.4 
March 35 416 107.6 11 859 109 11 691 105.6 11 866 108.2 
April 34 515 108.3 11 555 109.7 11 218 104.7 11 742 110.6 
May 35 255 107.1 11 782 108.3 11 600 104.8 11 873 108.2 
June 33 651 105.6 11 229 106.6 11 133 103.9 11 289 106.3 
July 33 651 102.2 11 312 103.9 11 196 101.1 11 143 101.6 
August 32 924 100 11 012 101.2 11 229 101.4 10 683  97.4 
September 31 457  98.7 10 223  97.1 10 711 100 10 523 99.1 
October 30 072  91.3  9797  90 10 620  95.9  9655  88 
November 28 222  88.6  9115  86.6 10 063  93.9  9044  85.2 
December 29 026  88.2  9539  87.7 10 300  93  9187  83.8 
Table II.

Absolute and observed/expecteda monthly numbers of births by month and socio-demographic group, 1989–1991

Month of birth All births Maternal age Birth order Maternal education Maternal marital status 
 Abs. No. Obs/Expa <20 20–24 25–29 30–34 35+ 1st 2nd 3rd 4th+ Prim. Voc. Second. Univ. Married Not married 
aThe observed/expected monthly numbers of births (×100) were calculated after correction for the unequal numbers of days. 
Abs. = Absolute; Obs/Exp = Observed/Expected; Prim. = Primary; Voc. = Vocational; Second. = Secondary; Univ. = University. 
January 32 533 98.8 99.8 99.3 95.9 101.2 101.7 97.1 99.0 104.8 101.0 102.7 99.7 97.4 95.1 99.0 97.3 
February 30 774 103.5 102.4 104.3 102.9 104.8 99.6 100.7 107.0 104.1 105.7 105.0 103.5 103.6 100.5 103.7 101.6 
March 35 416 107.6 103.4 107.1 110.8 105.7 109.0 103.2 112.9 108.3 109.2 104.8 107.0 109.8 104.8 108.2 100.7 
April 34 515 108.3 102.2 107.7 112.8 113.0 103.1 104.1 115.4 109.7 98.1 104.4 106.6 110.3 113.3 109.2 99.9 
May 35 255 107.1 97.7 106.9 112.0 110.0 101.6 103.6 112.7 106.4 98.4 102.3 104.4 109.7 115.0 107.8 99.7 
June 33 651 105.6 99.6 105.0 109.2 107.1 106.1 103.1 109.8 103.0 104.1 100.2 105.3 106.8 110.4 106.0 102.0 
July 33 651 102.2 97.4 101.6 103.5 106.2 105.2 101.9 101.5 104.5 104.1 99.9 102.5 102.3 104.3 102.1 103.7 
August 32 924 100.0 101.8 100.0 99.6 99.3 101.9 102.3 97.1 100.8 100.1 100.2 100.7 99.2 100.2 100.0 100.1 
September 31 457 98.7 104.0 100.1 95.7 95.1 95.5 104.4 92.7 93.3 101.5 99.9 98.9 98.5 97.3 98.6 100.1 
October 30 072 91.3 96.8 92.2 88.8 87.8 91.5 95.1 86.4 91.7 93.6 94.2 91.7 90.4 89.8 90.9 96.2 
November 28 222 88.6 97.3 88.2 84.6 85.2 93.9 93.0 82.2 87.9 94.2 92.0 90.0 86.7 85.2 87.8 96.8 
December 29 026 88.2 97.5 87.6 84.3 84.6 91.0 91.7 83.7 85.7 90.1 94.6 89.6 85.4 84.1 86.8 102.0 
Month of birth All births Maternal age Birth order Maternal education Maternal marital status 
 Abs. No. Obs/Expa <20 20–24 25–29 30–34 35+ 1st 2nd 3rd 4th+ Prim. Voc. Second. Univ. Married Not married 
aThe observed/expected monthly numbers of births (×100) were calculated after correction for the unequal numbers of days. 
Abs. = Absolute; Obs/Exp = Observed/Expected; Prim. = Primary; Voc. = Vocational; Second. = Secondary; Univ. = University. 
January 32 533 98.8 99.8 99.3 95.9 101.2 101.7 97.1 99.0 104.8 101.0 102.7 99.7 97.4 95.1 99.0 97.3 
February 30 774 103.5 102.4 104.3 102.9 104.8 99.6 100.7 107.0 104.1 105.7 105.0 103.5 103.6 100.5 103.7 101.6 
March 35 416 107.6 103.4 107.1 110.8 105.7 109.0 103.2 112.9 108.3 109.2 104.8 107.0 109.8 104.8 108.2 100.7 
April 34 515 108.3 102.2 107.7 112.8 113.0 103.1 104.1 115.4 109.7 98.1 104.4 106.6 110.3 113.3 109.2 99.9 
May 35 255 107.1 97.7 106.9 112.0 110.0 101.6 103.6 112.7 106.4 98.4 102.3 104.4 109.7 115.0 107.8 99.7 
June 33 651 105.6 99.6 105.0 109.2 107.1 106.1 103.1 109.8 103.0 104.1 100.2 105.3 106.8 110.4 106.0 102.0 
July 33 651 102.2 97.4 101.6 103.5 106.2 105.2 101.9 101.5 104.5 104.1 99.9 102.5 102.3 104.3 102.1 103.7 
August 32 924 100.0 101.8 100.0 99.6 99.3 101.9 102.3 97.1 100.8 100.1 100.2 100.7 99.2 100.2 100.0 100.1 
September 31 457 98.7 104.0 100.1 95.7 95.1 95.5 104.4 92.7 93.3 101.5 99.9 98.9 98.5 97.3 98.6 100.1 
October 30 072 91.3 96.8 92.2 88.8 87.8 91.5 95.1 86.4 91.7 93.6 94.2 91.7 90.4 89.8 90.9 96.2 
November 28 222 88.6 97.3 88.2 84.6 85.2 93.9 93.0 82.2 87.9 94.2 92.0 90.0 86.7 85.2 87.8 96.8 
December 29 026 88.2 97.5 87.6 84.3 84.6 91.0 91.7 83.7 85.7 90.1 94.6 89.6 85.4 84.1 86.8 102.0 
Table III.

Numbers of births, and indicators of the seasonal variation in births by socio-demographic characteristics

 No. of births (%) Coefficient of variation (%) Ratio of peak versus lowest birth rate month Ratio of peak versus lowest birth rate trimesters Odds ratios (95% confidence intervals) of being born in the peak birth rate trimestera 
     Crude Multivariate 
aOnly births in March to May and October to December were included in calculating the odds ratios (see Materials and methods). 
bAll maternal characteristics in one model. 
All births 387 496 (100)  7.2 1.23 1.2  b 
Maternal age group       
–19  55 817 (14)  2.7 1.07 1.04 
20–24 174 403 (45)  7.1 1.23 1.2 1.16 (1.13–1.19) 1.07 (1.04–1.10) 
25–29 103 181 (27) 10.3 1.34 1.3 1.26 (1.22–1.29) 1.12 (1.08–1.16) 
30–34  38 399 (10)  9.8 1.4 1.28 1.22 (1.18–1.27) 1.11 (1.06–1.16) 
35+  15 696 (4)  5.8 1.21 1.14 1.09 (1.03–1.14) 1.03 (0.97–1.09) 
Maternal marital status       
Unmarried  33 931 (9)  2.3 1.08 1.02 
Married 353 365 (91)  7.8 1.26 1.22 1.20 (1.17–1.24) 1.12 (1.09–1.16) 
Maternal education       
Primary  53 656 (14)  4.3 1.14 1.11 
Vocational 150 753 (39)  6.3 1.19 1.17 1.06 (1.03–1.09) 1.03 (1.00–1.06) 
Secondary 148 722 (38)  8.8 1.29 1.26 1.13 (1.10–1.17) 1.08 (1.04–1.11) 
University  34 365 (9) 10.3 1.37 1.29 1.16 (1.12–1.20) 1.07 (1.02–1.11) 
Birth order       
1st 187 613 (48)  4.6 1.14 1.11 
2nd 142 126 (37) 11.9 1.4 1.35 1.21 (1.19–1.24) 1.17 (1.14–1.20) 
3rd  42 298 (11)  8.2 1.28 1.22 1.11 (1.07–1.14) 1.08 (1.04–1.12) 
4th and higher  15 459 (4)  5.5 1.21 1.1 0.99 (0.94–1.03) 1.00 (0.94–1.05) 
 No. of births (%) Coefficient of variation (%) Ratio of peak versus lowest birth rate month Ratio of peak versus lowest birth rate trimesters Odds ratios (95% confidence intervals) of being born in the peak birth rate trimestera 
     Crude Multivariate 
aOnly births in March to May and October to December were included in calculating the odds ratios (see Materials and methods). 
bAll maternal characteristics in one model. 
All births 387 496 (100)  7.2 1.23 1.2  b 
Maternal age group       
–19  55 817 (14)  2.7 1.07 1.04 
20–24 174 403 (45)  7.1 1.23 1.2 1.16 (1.13–1.19) 1.07 (1.04–1.10) 
25–29 103 181 (27) 10.3 1.34 1.3 1.26 (1.22–1.29) 1.12 (1.08–1.16) 
30–34  38 399 (10)  9.8 1.4 1.28 1.22 (1.18–1.27) 1.11 (1.06–1.16) 
35+  15 696 (4)  5.8 1.21 1.14 1.09 (1.03–1.14) 1.03 (0.97–1.09) 
Maternal marital status       
Unmarried  33 931 (9)  2.3 1.08 1.02 
Married 353 365 (91)  7.8 1.26 1.22 1.20 (1.17–1.24) 1.12 (1.09–1.16) 
Maternal education       
Primary  53 656 (14)  4.3 1.14 1.11 
Vocational 150 753 (39)  6.3 1.19 1.17 1.06 (1.03–1.09) 1.03 (1.00–1.06) 
Secondary 148 722 (38)  8.8 1.29 1.26 1.13 (1.10–1.17) 1.08 (1.04–1.11) 
University  34 365 (9) 10.3 1.37 1.29 1.16 (1.12–1.20) 1.07 (1.02–1.11) 
Birth order       
1st 187 613 (48)  4.6 1.14 1.11 
2nd 142 126 (37) 11.9 1.4 1.35 1.21 (1.19–1.24) 1.17 (1.14–1.20) 
3rd  42 298 (11)  8.2 1.28 1.22 1.11 (1.07–1.14) 1.08 (1.04–1.12) 
4th and higher  15 459 (4)  5.5 1.21 1.1 0.99 (0.94–1.03) 1.00 (0.94–1.05) 
Figure 1.

Seasonal variation in births by maternal age group.

Figure 1.

Seasonal variation in births by maternal age group.

Figure 2.

Seasonal variation in births by maternal marital status.

Figure 2.

Seasonal variation in births by maternal marital status.

Figure 3.

Seasonal variation in births by maternal education.

Figure 3.

Seasonal variation in births by maternal education.

Figure 4.

Seasonal variation in births by birth order.

Figure 4.

Seasonal variation in births by birth order.

3
To whom correspondence should be addressed. E-mail: martinb@public-health.ucl.ac.uk

We thank Miroslav Simek for providing us with the birth registration data; Hynek Pikhart for computing assistance; and Luc Smits for important comments at an early stage of these analyses.

References

Basso, O., Olsen, J., Bisanti, L. et al. (
1995
) Are seasonal preferences in pregnancy planning a source of bias in studies of seasonal variation in reproductive outcomes? The European Study Group on Infertility and Subfecundity.
Epidemiology
 ,
6
,
520
–524.
Bongaarts, J. (
1978
) A framework for analysing the proximate determinants of fertility.
Popul. Dev. Rev.
 ,
4
,
105
–132.
Centola, G.M. and Eberly, S. (
1999
) Seasonal variations and age-related changes in human sperm count, motility, motion parameters, morphology, and white blood cell concentration.
Fertil. Steril.
 ,
72
,
803
–808.
Chaudhury, R.H. (
1972
) Socioeconomic and seasonal variations in births: a replication.
Soc. Biol.
 ,
19
,
65
–68.
Doblhammer, G. (1999) Longevity and month of birth: evidence from Austria and Denmark. Demographic Research, 1/3 (www.demographic-research.org).
Erhardt, C., Nelson, F.G. and Pakter, J. (
1971
) Seasonal patterns of conception in New York city.
Am. J. Public Health
 ,
11
,
2246
–2258.
Fialova, L. (
1995
) Seasonality of demographic events in the Czech countries during 17–20th century [in Czech].
Demografie
 ,
37
,
9
–21.
Golden, M.L. and Millman, S.R. (1993) Models of fecundability. In Grey, R., Leridon, H. and Spira, A. (eds), Biomedical and Demographic Determinants of Reproduction. Clarendon Press, Oxford, pp. 183–208.
Gyllenborg, J., Skakkebaek, N.E., Nielsen, N.C. et al. (
1999
) Secular and seasonal changes in semen quality among young Danish men: a statistical analysis of semen samples from 1927 donor candidates during 1977–1995.
Int. J. Androl.
 ,
22
,
28
–36.
Hoffmann, F. and Kawiani, D. (
1976
) Seasonal variations in the birth rate and conception rate within the last 200 years.
Geburtshilfe Frauenheilkunde
 ,
36
,
780
–785.
James, W.H. (
1971
) Social class and season of birth.
J. Biosoc. Sci.
 ,
3
,
309
–320.
Jongbloet, P.H. (
1983
) Menses and moon phases, ovulation and seasons, vitality and month of birth.
Dev. Med. Child. Neurol.
 ,
25
,
527
–531.
Kallan, J.E. and Udry, J.R. (
1989
) Demographic components of seasonality of pregnancy.
J. Biosoc. Sci.
 ,
21
,
101
–108.
Koupilova, I., Vagero, D., Leon, D.A. et al. (
1998
) Social variation in size at birth and preterm delivery in the Czech Republic and Sweden, 1989–1991.
Paediatr. Perinatal. Epidemiol.
 ,
12
,
7
–24.
Lam, D.A. and Miron, J.A. (
1991
) Temperature and the seasonality of births.
Adv. Exp. Med. Biol.
 ,
286
,
73
–88.
Lam, D.A. and Miron, J.A. (
1994
) Global patterns of seasonal variation in human fertility.
Ann. N.Y. Acad. Sci.
 ,
709
,
9
–28.
Lam, D.A. and Miron, J.A. (
1996
) The effects of temperature on human fertility.
Demography
 ,
33
,
291
–305.
Nybo Andersen, A.M., Wohlfahrt, J., Christens, P. et al. (
2000
) Maternal age and fetal loss: population based register linkage study.
Br. Med. J.
 ,
320
,
1708
–1712.
Pasamanick, B., Dinitz, S. and Knoblock, H. (
1960
) Socio-economic and seasonal variations in birth rates.
Millbank Mem. Fund Quart.
 ,
38
,
248
–254.
Prioux, F. (
1988
) Mouvement saisonnier des naissances: influence du rang et de la legitimite dans quelques pays d'Europe occidentale.
Population
 ,
3
,
587
–610.
Roenneberg, T. and Aschoff, J. (
1990
) Annual rhythm of human reproduction: I. Biology, sociology, or both?
J. Biol. Rhythms
 ,
5
,
195
–216.
Roenneberg, T. and Aschoff, J. (
1990
) Annual rhythm of human reproduction: II. Environmental correlations.
J. Biol. Rhythms
 ,
5
,
217
–239.
Rojansky, N., Brzezinski, A. and Schenker, J.G. (
1992
) Seasonality in human reproduction: an update.
Hum. Reprod.
 ,
7
,
735
–745.
Smits, L.J., Zielhuis G.A., Jongbloet, P. et al. (
1998
) Seasonal variation in human fecundability.
Hum. Reprod.
 ,
13
,
3250
–3524.
Smits, L., Zielhuis, G., Jongbloet, P. et al. (
1999
) The association of birth interval, maternal age and season of birth with the fertility of daughters: a retrospective cohort study based on family reconstitutions from nineteenth and early twentieth century Quebec.
Paediatr. Perinat. Epidemiol.
 ,
13
,
408
–420.
Stolwijk, A.M., Straatman, H., Zielhuis G.A. et al. (
1996
) Seasonal variation in the time to pregnancy: avoiding bias by using the date of onset.
Epidemiology
 ,
7
,
156
–160.
Warren, C.H. and Tyler, C.W. (
1979
) Social status and season of birth: a study of a metropolitan area in the Southeastern United States.
Soc. Biol.
 ,
26
,
275
–288.
Warren, C.W., Gwinn, M.L. and Rubin, G.L. (
1986
) Seasonal variation in conception and various pregnancy outcomes.
Soc. Biol.
 ,
33
,
116
–126.
Weinberg, C.R., Moledor, E., Baird, D.D. et al. (
1994
) Is there a seasonal pattern in risk of early pregnancy loss.
Epidemiology
 ,
5
,
484
–489.
Weinstein, M., Wood, J. and Ming-Cheng, C. (1993) Age patterns of fecundability. In Gray, R. et al. (eds), Biomedical and Demographic Determinants of Reproduction. Clarendon Press, Oxford, pp. 209–227.
Zelnik, M. (
1969
) Socio-economic and seasonal variation in birth: a replication.
Millbank Mem. Fund. Quart.
 ,
47
,
159
–165.