Lie, Rolv T. PhD1,2; Wilcox, Allen J. MD, PhD3; Skjærven, Rolv PhD1,2
OBJECTIVE: Biological evidence suggests that both mother and fetus are involved in triggering a normal delivery. A tendency of a child to have a gestational age at birth similar to the father’s could represent the effect of genes passed from the father to the fetus. Similar tendencies between mother and child could represent maternal genes passed to the fetus, as well as genes to the mother received from the grandmother that affect a woman’s capacity to carry a pregnancy.
METHODS: The Medical Birth Registry of Norway contains data on all births in Norway from 1967 onward. We identified 77,452 pairs of boys and girls born at term who later became parents and linked their birth data to the birth records for their first child.
RESULTS: Gestational age of the child at birth increased on average 0.58 days for each additional week in the father’s gestational age (95% confidence interval 0.48–0.67) and 1.22 days for each additional week in the mother’s gestational age (1.21–1.32). Gestational age was, however, 0.65 days reduced for each additional kilogram in the father’s birth weight, presumably due to more rapid growth of the fetus triggering delivery.
CONCLUSION: Initiation of delivery has a fetal component that is heritable (passed from father and mother to child) and an additional maternal component that is also heritable. In addition, a more rapid rate of fetal growth appears to trigger delivery at earlier gestation.
LEVEL OF EVIDENCE: II-2
The timing of delivery is a crucial aspect of fetal and infant health, although the events that trigger natural onset of labor and delivery are poorly understood. Hormonal changes in the fetus and mother just before delivery indicate a role for both parties.1–5 Less is known about possible contributions of the father. In a study of dairy cattle, bulls that produced faster-growing fetuses also produced calves that delivered earlier.6 Such effects through the father are difficult to study in humans.
We use linked data for two generations of births in Norway to study the tendency of an offspring to repeat the gestational age at delivery of the parents. An effect of the father’s gestational age would be of particular interest, because it would indicate that genes passed on from father to offspring were involved in determining time of delivery. Parents who had a high birth weight produce larger and faster-growing babies. We investigate the hypothesis that these parental predictors of fetal growth also may have an effect on the offspring’s gestational age at delivery.
MATERIALS AND METHODS
This study was based on data from the Medical Birth Registry of Norway, which holds information on all 2.3 million births in the country from 1967 to 2003. Throughout the study period, the registry has recorded first day of the last menstrual period. Gestational age was calculated as number of days from this day to the day of delivery. Our study was based on anonymized Norwegian registry data and therefore considered exempt from ethics review board approval.
We linked all available birth records of parents to the birth records of their first child, using unique national identification numbers. We excluded 8% of babies who were missing information on fathers. This left 132,253 complete family triads in which both parents were born as singletons in 1967 or later and had a singleton delivery by 2003. We excluded an additional 14% of families in which either parent or their first child was missing data on birth weight or gestational age. We then excluded 17% of families in which any of the three members had been delivered by cesarean, leaving 94,468 families. We further restricted the sample to parents born at term (ie, whose natural vaginal delivery occurred between 37 and 42 completed weeks of gestation). The final sample of 77,452 families is the basis for all analyses.
Most analyses were conducted first by graphing offspring gestational age by the parent’s gestational age and birth weight. Mean values were compared by t tests. We then used regression analyses to quantify these relationships and to estimate the independent effects of parent’s gestational age and birth weight on their offspring’s gestational age. In the regression analyses, variables containing gestational age of parents in whole days were divided by 7 to represent gestational age in weeks with decimals. Birth weights of parents in grams were divided by 1,000 to represent birth weights in kilograms with decimals. Child’s gestational age (the dependent variable) was kept in days. We adjusted in the analyses for parent’s age, year of birth of the child, the couple’s marital status and the mother’s level of education. Analyses were performed using the STATA software 8.0 (StataCorp LP, College Station, TX).
Table 1 shows demographic characteristics of the couples included in the analyses. Fathers were slightly older than mothers, and more babies were born toward the end of the study period. The distribution of gestational age for mothers and fathers and the mean birth weight at each gestational age are shown in Table 2. As expected, the mean birth weight of fathers and mothers increased with increasing gestational age.
The father’s gestational age at birth was linearly related to that of his offspring (Fig. 1A). Compared with fathers born at 37 weeks, fathers born at 42 weeks had offspring born, on average, 2 days later (283.2 days vs 281.3 days, P < .001). A mother’s gestational age at birth was even more strongly related to that of her offspring (Figure 1B). Children whose mother’s gestational age was 42 weeks were delivered, on average, 4 days later than children whose mother’s gestational age was 37 weeks (284.7 days vs 280.0 days, P < .001).
We hypothesized that a rapidly growing fetus might trigger delivery earlier. We first examined the parent’s birth weight as a predictor of the growth of the child (Fig. 2). Offspring birth weight increases by 145 g for every kilogram increase in father’s birth weight (95% confidence interval [CI] 140–150). As with gestational age, the effect was even stronger between mother and child, with the child’s birth weight increasing by 251 g per kilogram of the mother’s birth weight (95% CI 246–256).
Using the parent’s birth weight as a predictor of the fetal growth of the child, we explored the independent role of fetal growth on time of delivery. Since the parent’s gestational age could confound this association, we restricted this analysis to parents born at 40 or 41 completed weeks of gestation. The children of fathers with higher birth weights were born earlier than children of fathers with lower birth weights (Fig. 3A), and the association appears to be linear. The child’s gestational age was reduced by 0.79 days per kilogram increase in father’s birth weight (95% CI 0.45–1.12, P < .001). An opposite effect was found with mother and child (Fig. 3B). The child’s gestational age increased by 0.86 days per kilogram increase in mother’s birth weight (95% CI 0.51–1.21, P < .001).
We also used multiple linear regression analysis to adjust for the simultaneous effects of parents’ birth weight and gestational age on the gestational age of their child. After adjustment, a 1-kg increase in father’s birth weight was associated with a 0.65-day earlier delivery of his child (95% CI 0.40–0.90) (Table 3). Independently, the father’s gestational age was still positively associated with the gestational age of his child (0.58 days per each week of the father’s gestational age, 95% CI 0.48–0.67). The effect of mother’s birth weight on gestational age of her child was unchanged after adjustment (0.44 day increase per kilogram birth weight of the mother, 95% CI 0.18–0.69), with the effect of the mother’s gestational age on the child’s gestational age (1.22 days per week, 95% CI 1.12–1.32) twice the father’s. These estimates were also adjusted for the mother’s and father’s age and year of birth of the child. Further adjustment for the couple’s marital status and the mother’s education had virtually no effect on the estimates. When analyses were performed separately for boy and girl offspring, we found no indication of difference in estimated effects.
There are few studies of determinants of gestational age across generations. This study confirms previously reported effects of the mother’s gestational age7,8 and her birth weight9 on the offspring’s gestational age. However, those earlier studies did not provide particular insights on the biological mechanisms involved. Our additional data on fathers provide useful contrasts with the mothers’ effects on gestational age at birth and suggest possible mechanisms for the triggering of delivery in humans.
An association of the father’s own gestational age with the gestational age of his offspring has also been seen in retrospective data from England.8 The association presumably reflects alleles passed from the father to fetus and involved in onset of delivery. This supports other evidence showing that the fetus can help initiate the onset of labor and delivery. Although the triggering mechanisms themselves are unknown, a cascade of physiological changes takes place in the fetus and the mother before the delivery.1–5,10,11 There is experimental evidence from both sheep and horses that endocrine changes in the fetus (especially cortisol) may play a key role in triggering delivery. Evidence of a similar fetal regulation of cortisol before delivery has been described in humans.1,4 A recent study in mice has suggested that surfactant protein-A secreted in fetal lungs initiates delivery.12 Genetic regulation of similar mechanisms in humans may contribute to the heritable component of length of pregnancy between father and offspring in our data.
However, the association in gestational age between mother and child is much stronger than the association between father and child. This suggests that the mother contributes in additional ways to her offspring’s gestational age. One added component could be alleles passed from the grandmother to the mother that affects the maternal phenotype. A woman’s capacity to carry a baby for longer may itself be heritable and perhaps related to her body size or structure. Genetic variation in the length of the follicular phase could also contribute (given that gestational age in these data are defined from the onset of the last menstrual period). If so, a woman’s gestational age at birth would, to some degree, be a product of her own mother’s reproductive capacity, which the woman in turn may inherit.
The association of gestational age between mother and child is a result of maternal, as well as fetal, contributions, whereas the association between father and child should result only from a contribution through the fetus. Because the observed association between mother and child was almost exactly twice that of father and child, we might infer that the contribution of the mother’s heritable phenotype is similar in magnitude to the contribution of the fetus’ heritable phenotype.
In contrast, there is a striking difference between parents in the association of their own birth weight (adjusted for their gestational age) with the gestational age of their offspring. Mothers with higher birth weights have children with longer gestational age. Fathers with higher birth weights have children with shorter gestational age. How can this be explained? The fathers provide the simpler case, because their contribution is solely through the genotype of the fetus. We know that fathers with higher birth weights tend to have larger offspring (Fig. 2) and thus offspring who grow faster in utero. The fact that these fathers also tend to have offspring with shorter gestational age suggests that faster fetal growth triggers delivery sooner. Indirect evidence from studies of twins is consistent with this conclusion.13 More directly, a study of half siblings in dairy cattle showed that bulls producing larger fetuses also produced calves that delivered earlier.6
If faster-growing fetuses trigger delivery earlier, why do we not see the same association between the mother’s birth weight and her offspring’s gestational age? This may be because the mother’s birth weight is also associated with maternal factors that improve her capacity to carry a pregnancy. A mother’s weight at birth may be associated with such adult characteristics as pelvic size or uterine size. If so, these aspects of maternal capacity apparently counterbalance the expected negative correlation between the mother’s size at birth and the length of her pregnancies. The opposing effects of the mother’s birth weight make the net effect of her birth weight the weakest association in our analysis (Table 2). However, this relationship emerges clearly in the simpler setting of father and offspring.
Obstetric management may also contribute to shorter gestation of larger babies. However, it is not clear how this could explain the opposing effects of the father’s and mother’s birth weights. The possibility that more rapid fetal growth contributes to shorter gestational age deserves further consideration.
Our study was restricted to parents born at term from normal, spontaneous deliveries. Pathologic mechanisms such as preeclampsia14 (which have both a maternal and a fetal component) are more likely to be involved in preterm or postterm deliveries. Such pathologies may have their own patterns of familial recurrence.15–17 By restricting parents to those delivered vaginally at term (and their offspring who were delivered vaginally), our analysis focuses on natural regulation processes. Our results were not altered when we removed all families having a preeclamptic pregnancy.
Environmental and social conditions that persist across generations may, in principle, contribute to the associations we observe, although there are, in fact, few environmental factors known to affect gestational age. Even cigarette smoking, which can increase the risk of preterm delivery, has only a weak effect on mean gestational age.18 Such environmental factors are therefore not likely to represent major confounding in our analysis.
Registration of Norwegian mothers and children in our sample should be close to 100%. Information on fathers is missing for 8% of births in the Registry,14 and paternity may be incorrect for a small percent of others. Such errors would tend to obscure the associations between father and offspring. Similarly, errors in gestational age estimated from last menstrual period or birth weight would tend to reduce the estimated effects. It is therefore unlikely that the effects we see could result from systematic bias. Improvements and changes in obstetric care may also have affected the associations we see between parents’ and offspring length of pregnancy. This should, however, also have the effect of weakening the observed associations.
In conclusion, our data suggest that both the fetal and the maternal system influence the length of pregnancy through mechanisms that are subject to genetic influence. An association between the father’s and child’s gestational age supports a role of the fetus and placenta in triggering delivery. Furthermore, the link between a higher weight at birth for the father and a shorter gestation for his offspring suggests an effect of fetal growth on the duration of pregnancy.
1. Chan EC, Smith R, Lewin T, Brinsmead MW, Zhang HP, Cubis J, et al. Plasma corticotropin-releasing hormone, beta-endorphin and cortisol inter-relationships during human pregnancy. Acta Endocrinol 1993;128:339–44.
2. Silver M. Placental progestagens in the sheep and horse and the changes leading to parturition. Exp Clin Endocrinol 1994;102:203–11.
3. McLean M, Bisits A, Davies J, Woods R, Lowry P, Smith R. A placental clock controlling the length of human pregnancy. Nat Med 1995;1:460–3.
4. McLean M, Smith R. Corticotrophin-releasing hormone and human parturition. Reproduction 2001;121:493–501.
5. Challis JRG, Matthews SG, Gibb W, Lye SJ. Endocrine and paracrine regulation of birth at term and preterm. Endocr Rev 2000;21:514–50.
6. Bourdon RM, Brinks JS. Genetic, environmental and phenotypic relationships among gestation length, birth weight, growth traits and age at first calving in beef cattle. J Anim Sci 1982;55:543–53.
7. Magnus P, Bakketeig LS, Skjærven R. Correlations of birth weight and gestational age across generations. Ann Hum Biol 1993;20:231–8.
8. Hennessy E, Alberman E. Intergenerational influences affecting birth outcome. II. Preterm delivery and gestational age in the children of the 1958 British birth cohort. Paediatr Perinat Epidemiol 1998;12, S1:61–75.
9. Klebanoff MA, Yip R. Influence of maternal birth weight on rate of fetal growth and duration of gestation. J Pediatr 1987;111:287–92.
10. Thomson AJ, Telfer JF, Young A, Campbell S, Stewart CJ, Cameron IT, et al. Leukocytes infiltrate the myometrium during human parturition: further evidence that labour is an inflammatory process. Hum Reprod 1999;14:229–36.
11. Slater DM, Astle S, Bennett PR, Thornton S. Labour is associated with increased expression of type-IIA secretory phospholipase A2 but not type-IV cytosolic phospholipase A2 in human myometrium. Mol Hum Reprod 2004;10:799–805.
12. Condon JC, Jeyasuria P, Faust JM, Mendelson CR. Surfactant protein secreted by the maturing mouse fetal lung acts as a hormone that signals the initiation of parturition. Proc Natl Acad Sci U S A 2004;101:4978–83.
13. Loos RJ, Derom C, Eeckels R, Derom R, Vlietinck R. Length of gestation and birthweight in dizygotic twins. Lancet 2001;358:560–1.
14. Lie RT, Rasmussen S, Brunborg H, Gjessing HK, Lie-Nielsen E, Irgens LM. Fetal and maternal contributions to risk of pre-eclampsia: population based study. BMJ 1998;316:1343–7.
15. Olesen AW, Basso O, Olsen J. Risk of recurrence of prolonged pregnancy. BMJ 2003;326:476.
16. Laursen M, Bille C, Olesen AW, Hjelmborg J, Skytthe A, Christensen K. Genetic influence on prolonged gestation: a population-based Danish twin study. Am J Obstet Gynecol 2004;190:489–94.
17. Klebanoff MA, Schulsinger C, Mednick BR, Secher NJ. Preterm and small-for-gestational-age birth across generations. Am J Obstet Gynecol 1997;176:521–6.
18. Oyen N, Haglund B, Skjærven R, Irgens LM. Maternal smoking, birthweight and gestational age in sudden infant death syndrome (SIDS) babies and their surviving siblings. Paediatr Perinat Epidemiol 1997;11, S1:84–95.
Figure. No caption available.