Effects on pregnancy outcome of changing partner between first two births: prospective population studyBMJ 2003; 327 doi: https://doi.org/10.1136/bmj.327.7424.1138 (Published 13 November 2003) Cite this as: BMJ 2003;327:1138
- 1Department of Public Health and General Practice, Norwegian University of Science and Technology, NO-7489 Trondheim, Norway
- 2Department of Public Health and Primary Health Care, University of Bergen, NO-5021 Norway
- Correspondence to: L J Vatten
- Accepted 15 August 2003
Objective To compare the effects on pregnancy outcomes of changing partner between the first two births with having the same partner for both births.
Design Prospective population study.
Participants 31 683 women who changed partner between their first two births and 456 458 women with the same partner for both births.
Results After adjustment for maternal age and education, interval between births, and decade of birth, the risk of adverse pregnancy outcomes for the second birth was higher for women who changed partner between the first two births compared with those who had the same partner for both births: preterm birth (< 37 weeks; relative risk 2.0, 95% confidence interval 1.9 to 2.1), low birth weight (< 2500 g; 2.5, 2.3 to 2.6), and infant mortality (1.8, 1.6 to 2.1). For the first birth, the risk of these adverse pregnancy outcomes was only slightly higher for mothers who subsequently had a second birth with another partner.
Conclusion Women who change partner between their first two births are at an increased risk of delivering a preterm, low birthweight baby with an increased risk of infant mortality compared with women who have the same partner for both births.
A consequence of the increase in remarriages in Western societies is that a growing number of women are having children with different partners. The risk of adverse pregnancy outcomes may be higher in women who have changed partners between births than those who have the same partner. Women who have changed partners may have particular characteristics that increase risk, or they may adopt unfavourable behaviours with a new partner.
Pregnancy outcomes associated with change of partner have been ascribed to paternal influences such as antigens or genes.1–5 This approach is useful for conditions that may recur from one pregnancy to another—for example, pre-eclampsia or congenital malformations—but in studies of spontaneous outcomes without reference to recurrence, the biological interpretation is less clear.1–5 Factors associated with change of partner, such as characteristics or changes in lifestyle, should also be considered.6 It is also possible that these effects could be modified by certain factors. For example, women with a high level of education are less likely to have adverse pregnancy outcomes, and childbearing with different partners is also less common in this group.7–10 We compared infant mortality, preterm birth, and low birth weight between the first two births in women who changed partner between births with those who kept the same partner.
Data were derived from Norway's medical birth registry, which has records of around 1.8 million births between 1967 and 1998. Midwives and doctors notify the registry of each birth by using a standardised form.1–5 Data registered include birth weight, birth length, and length of gestation estimated from the last menstrual period. Information on infant deaths was collected from vital statistics monitored by Statistics Norway.
The identity of both parents is also recorded, enabling identification of women who have children with different partners from those with the same partner. We restricted our study to women who had had at least two births. Overall, 488 141 women were included, and 31 683 of these had a different partner for a second birth.
Educational level is collected by Statistics Norway. For our study, level of maternal education was taken as completed years of education: low (< 10 years), medium (11-14 years), or high (> 14 years), according to revised national recommendations.11
Infant deaths were defined as deaths among liveborn infants during the first year of life. Preterm birth was defined as delivery before 37 weeks of gestation, estimated from the last menstrual period, and low birth weight was defined as birth weights less than 2500 g, regardless of length of gestation. The quality of the assessment of gestational age was assured by using a method described elsewhere.12
We compared pregnancy outcomes for the women who changed partner between the first two births with those who had the same partner. Deliveries were divided into three periods (1967-76, 1977-86, and 1987-98), maternal age at delivery was divided into five year categories (< 20, 20-24, 25-29, etc), and interval between births was divided into two year categories. Stratification and logistic regression techniques were used to estimate odds ratios as estimates of relative risk and to adjust for possible confounding by maternal age, interval between births, decade of birth, and maternal level of education. We also analysed the association between change of partner and adverse pregnancy outcomes separately for each level of maternal education and tested whether change of partner had a different effect on pregnancy outcomes for different levels of maternal education.
The proportion of women who changed partner between their first two births increased from 3.1% (3941/128 692) during 1967-76 to 10.0% (21 297/213 674) during 1987-98 (1). The level of maternal education also increased: 10.3% of women had a high level of education during 1967-76 compared with 22.6% during 1987-98. A change in partner between the first two births is common among women with a low level of education compared with those with a high level of education, and this difference has increased over time (2).
Infant mortality decreased from first to second birth in women who had the same partner for both births (8.3 to 6.7 per 1000 births) but not in those who changed partner (8.3 and 8.7 per 1000: 3). After adjustment for maternal age, decade of birth, and maternal level of education, the risk of infant mortality for the first child was identical for women with the same partner and those who changed partner (relative risk 1.0, 95% confidence interval, 0.9 to 1.1). For the second child, however, the risk was higher (1.8, 1.6 to 2.1) for women who changed partner, after additional adjustment for time between births.
The risk of preterm birth between a first and second birth was reduced for women who had the same partner for both births (5.1% to 3.7%) and increased for women who changed partner (6.1% to 7.6%). After adjustment for maternal age, decade of birth, and maternal level of education, the risk was slightly higher for women who changed partner (1.18, 1.1 to 1.3), and for the second birth was higher than for women who had the same partner for both births (2.0, 1.9 to 2.1).
The risk of a low birthweight baby was reduced from the first to second birth for women who had the same partner for both births (4.4% to 2.8%) but increased for women who changed partner (5.5% to 6.8%). After adjustment for maternal age, decade of birth, and maternal level of education, the risk of having a first baby of low birth weight was slightly higher for women who had a second child with a different partner (1.27, 1.2 to 1.3). The risk was, however, higher for the second child (2.5, 2.3 to 2.6). This was not altered by additional adjustment for the interval between births.
For second births there was a consistent reduction in infant mortality related to increasing level of maternal education regardless of whether partners had changed. Preterm birth and low birth weight were, however, less common in women with a high level of education, but only if they had the same partner for both births (4). For women who changed partner, education did not seem to protect against preterm birth or low birth weight, and the risks did not differ between levels of education.
The risk of infant mortality did not differ between levels of education in women who changed partner (4). The effect on preterm birth was stronger for women with a high level of education (2.6, 2.2 to 3.0) than for those with a medium or low level of education (2.0, 1.9 to 2.1 and 1.8, 1.6 to 2.0, respectively; P < 0.001 for interaction).
The risk of a second child of low birth weight associated with change of partner was modified by maternal level of education: relative risks 3.8 (3.3 to 4.5) for high level of education, 2.5 (2.3 to 2.6) for medium level of education, and 2.0 (1.7 to 2.3) for low level of education (4).
Changing partner between the first two births increases the risk of infant mortality, preterm birth, and low birth weight for the second birth compared with having the same partner. Our observation period covered more than 30 years, and shows that childbearing with different partners is increasingly more common. Women who change partner tend to have longer intervals between births and to be slightly older during childbearing, both of which may affect pregnancy outcomes.5 To avoid confounding we therefore adjusted for differences in maternal age and interval between births. Similarly, we adjusted for level of maternal education, because a high level of education has been negatively associated with adverse outcomes in pregnancy.7 8
It is possible that adverse pregnancy outcomes occur more often among women who have changed partner between their first two births than among those who have the same partner. Women who change partner may be a selected group with characteristics that increase risk or they may change their lifestyle or behaviour in ways that are unfavourable on pregnancy outcomes.6
The “selection hypothesis” implies that women who change partner have a higher prevalence of risk taking behaviour, such as heavier smoking and alcohol consumption and poorer nutrition, and that this continues in pregnancy.6 13 14 We explored this possibility in the most recent data from Norway's medical birth registry. Among 68 427 women with at least two births, smoking was nearly three times (odds ratio 2.9, 95% confidence interval 2.7 to 3.0) as likely among those who had a different partner for the second birth than those who had the same partner.
These women would be expected to also have a higher frequency of adverse pregnancy outcomes for their first birth, but the risk was only slightly higher than for women who had the same partner. This may weaken the selection hypothesis. However, another possibility is that risk of adverse pregnancy outcomes increases with duration of smoking, so that maternal smoking will do increasingly more harm with increasing parity. Women who quit smoking while pregnant may nonetheless have a higher risk of adverse outcomes than women who have never smoked.
Changes in lifestyle or social circumstances may accompany change of partner, and these changes could be important. Women may change behaviour from one pregnancy to the next, but it remains to be shown that women who change partner between births are more likely to adopt a less healthy lifestyle than those who have the same partner.
Many studies of pregnancy outcomes related to change of partner have focused on recurrence of conditions between births.1–5 Effects have therefore been attributed to change of paternal antigens or genes. We did not consider recurrence of conditions from one pregnancy to the next but examined spontaneous pregnancy outcomes. It seems less clear that change of partner could reflect genetic changes or different paternal antigens in this situation. It is more likely that change of paternal genes or antigens would result in heterogeneous outcomes rather than unfavourable pregnancy outcomes.
What is already known on this topic
Childbearing with different partners has increased in Western societies
Little is known about the effects of this on pregnancy outcomes
What this study adds
Women who change partner between their first two births are at a higher risk of having a preterm, low birthweight baby with an increased risk of infant mortality
Changes in lifestyle or social circumstances may accompany change in partner and therefore affect pregnancy outcomes
Women with high levels of education seem particularly vulnerable to the effects of changing partner
In general, high educational level is associated with a reduced risk of adverse pregnancy outcomes, as is having the same partner for both births.7 8 For women who changed partner, however, education had no protective effect, since the absolute risk of adverse outcomes was similar between women with different levels of education. In relative terms, the strongest effects related to change of partner between the first two births were observed in women who had a high level of education. In these women, change of partner was associated with a higher relative risk of a preterm, low birthweight baby than in women with medium or low levels of education. Women with high levels of education seem particularly vulnerable to the effects of changing partner.
We thank the staff of the Medical Birth Registry for data management and the women who contributed data.
Contributors Both authors analysed the data and wrote the paper. They will act as guarantors for the paper.
Funding The study was supported by the Norwegian Medical Research Council. The guarantors accept full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.
Conflict of interest None declared.
Ethical approval The study was approved by the internal review board of the Medical Birth Registry of Norway.