Rasmussen, Svein MD, PhD; Irgens, Lorentz M. MD, PhD; Skjærven, Rolv PhD; Melve, Kari Klungsøyr MD, PhD
Stillbirth is a relatively common adverse pregnancy outcome and constitutes more than half of all perinatal mortality.1 Considerable effort has been made to identify risk factors for stillbirth, but still 25–40% of cases remain unexplained1,2 and stillbirth continues to be an important challenge in antenatal care.
Conditions such as second- and third-trimester growth restriction, preeclampsia, and abruptio placentae carry a high risk of stillbirth.1–6 Although the mechanisms leading to these conditions are not established, epidemiologic, clinical, and morphological evidence supports a hypothesis that placental dysfunction or decidual (placental bed) vasculopathy, characterized by occlusive lesions in the maternal uteroplacental (spiral) arteries, represents a shared pathogenetic pathway.7–10 The vasculopathy is likely caused by failure of fetal cells (trophoblasts) to invade the spiral arteries in early pregnancy (impaired placentation).7 Additionally, Salafia et al8 reported decidual vasculopathy in a significant proportion of preterm births consistent with clinical and epidemiologic evidence that fetuses with growth restriction tend to be spontaneously born preterm.6
It has been suggested that these decidual changes tend to recur in a woman from one pregnancy to her next.10 Thus, it might be expected that fetal growth restriction, preterm birth, abruptio placentae, and preeclampsia in one pregnancy are associated with an excess risk of stillbirth in a subsequent pregnancy in the same woman, even without recurrence of these conditions. Consistently, longitudinal studies have reported that previous preterm birth,11,12 small for gestational age (SGA)11,13 and preeclampsia12 are associated with subsequent stillbirth. However, all studies failed to assess the effects of extremely early preterm birth (less than 28 weeks of gestation). Nor did the studies include preeclampsia, abruptio placenta and fetal growth restriction severe enough to cause stillbirth in the first pregnancy. Earlier studies have indicated paternal influences on preeclampsia,14 birth weight,15 and length of pregnancy.16 However, little is known about risk factors for stillbirth mediated through the father and the proportions of stillbirths in a subsequent pregnancy attributable to these risk factors.
The aim of the present study was to estimate whether a history of fetal growth restriction, abruptio placenta, preeclampsia, or live preterm birth in one pregnancy is associated with excess risk of stillbirth and late abortion in a subsequent pregnancy. We also wanted to estimate the proportion of stillbirth and late abortion in a subsequent pregnancy attributable to these risk factors and to estimate the maternal and paternal contributions to such effects.
MATERIALS AND METHODS
Since 1967, data on all live births and miscarriages or terminations of pregnancy on medical indications at 16 weeks of gestation or more are forwarded to the Medical Birth Registry of Norway based on compulsory notification.17 The midwives transfer all data to a notification form from a standardized antenatal form, completed during antenatal care and brought by the women to the delivery unit. To ensure notification of every newborn in the country, all records are matched with the Central Population Registry.17 In addition to extracting data from the pregnancy record, completion of the notification form is also based on interviewing the mother and on the hospital records. If there are insufficient data for completion of the notification, the delivery unit requests data from the primary health services source.
From 1967 through June 2005, more than 2,200,000 births have been registered. The national identification number allows linkage of a woman’s subsequent deliveries. Stillbirth in second pregnancies was identified, and associations with gestational age, birth weight percentiles, preeclampsia, and abruptio placentae in first pregnancies were assessed. The proportion of unknown fathers amounted to 159,884 (7.1%) births. We used the mother’s national identification number to identify 677,389 pairs of first and second singleton births. We excluded 2,924 women with missing birth weight in the first or second birth and further restricted the sample to mothers with gestational age between 16 and 44 weeks in both births, leaving 611, 957 pairs of first and second births for study. To increase sample size (in analyses on maternal and paternal contributions to fetal death), we used the parents’ identification numbers to include pairs of second and third, third and fourth, and fourth and fifth pregnancies in addition to pairs of first and second pregnancies, 747,221 pairs with the same mother and father, 51,708 with the same mother and different father, and 65,602 with the same father and different mother.
Stillbirths and late abortions are reported on the notification form at birth as death before or during labor or unspecified. In the present study, deaths at 16 weeks of gestation or later were included. Clinical criteria of preeclampsia and transient hypertension in Norway have been in accordance with the recommendations by the American College of Obstetricians and Gynecologists in 197218 and are referred to in the Medical Birth Registry’s instructions for completion of the notification form. Abruptio placentae is defined as the premature separation of a normally situated placenta. Abruptio placentae is usually a clinical diagnosis based on prenatal signs and symptoms, such as antepartum hemorrhage, uterine pain or tenderness, or fetal distress, or a retroplacental blood clot or impression. In the present study, severe abruptio placentae was defined as an abruptio placentae birth before 37 weeks of gestation, with birth weight below 2,500 g or with perinatal death.19 Other cases of abruptio placentae were defined as mild. In December 1998, a revised version of the notification form was implemented to include new variables like maternal smoking habits and several chronic diseases, which are notified by checking of boxes.17 Gestational age was estimated by subtracting the date of the first day in the last menstrual period from the date of birth. Outliers in gestational age were removed using a linear regression approach in which gestational age was regressed against birth weight (g) within 100-g strata of birth weight. We excluded births with z scores (absolute values) above 3.5. From 1998, gestational age based on ultrasound dating was available and was used when data on the last menstrual period were lacking (7.2%).
The associations of gestational age categories (22–27, 28–32, 33–36, and 37 or more weeks), small for gestational age (SGA, birth weight less than the 10th percentile for each gestational age week, gender, and birth order), preeclampsia, and abruptio placentae in the first pregnancy with stillbirth and late abortion (16–21 weeks of gestation) in the second were assessed by odds ratios (ORs) obtained by logistic regression. Adjustments were made for maternal age (19 or less, 20–29, 30–34, 35–39, 40 or more years), maternal marital status (married, cohabiting, unmarried or single, other, unknown), maternal education (less than 8, 8–10, 11–12, 13–17, 18 or more years, unknown), and year of birth in the second pregnancy (1967–1979, 1980–1992, and 1993–2005). The groupings for maternal education were based on the divisions of education in Norway, and the groupings for year of birth coincide with important changes in fetal surveillance, obstetric and pediatric practices, and fetal monitoring. We did not adjust for interpregnancy interval because adverse pregnancy outcome in the first pregnancy, such as abruptio placentae, may influence the interval to the next pregnancy. To avoid assumptions of linear associations, covariates were represented by indicator variables. The associations were analyzed in pairs of births with and without birth defects in the first pregnancy.
In the assessment of the associations of SGA, preeclampsia, and abruptio placentae with subsequent stillbirth, we did not adjust for or exclude stillbirth in the first pregnancy, as it might be caused by one of these conditions and is thus a possible intermediate step in the causal pathway between these conditions and stillbirth in the second pregnancy. However, when assessing the association between preterm birth and subsequent stillbirth, we excluded stillbirth in the first pregnancy because stillbirth would often cause, rather than be a consequence of, preterm birth. Additionally, we repeated analyses with stillbirth before (n=2,650) and during labor (n=547) and with unspecified time of death (n=1,285).
In supplementary analyses, we excluded preeclampsia, SGA, and abruptio placentae in the case definition of stillbirth in the second pregnancy. This left 2,991 (67%) of the total of 4,482 cases of stillbirth for analysis.
We also restricted the analyses to second births in each pair of births that occurred from 1999 onward when the revised notification form was introduced and further adjusted for smoking; chronic renal disease; asthma; chronic hypertension; rheumatoid arthritis; gestational, insulin dependent–, and non-insulin–dependent diabetes mellitus; preeclampsia; and transient hypertension (nonproteinuric pregnancy-induced hypertension) in the second pregnancy.
In the assessments of the maternal and paternal contributions to the effects of the risk factors, the risk was assessed up to four times in a single woman or man in pairs of first to second, second to third, third to fourth, and fourth to fifth birth. Thus, the data were organized in a two-level hierarchy with clusters of level-one data (the current pregnancy of birth orders two, three, four, and five) nested within each level-two unit (the woman). To avoid underestimated standard errors caused by the nested structure of the data, we used multilevel logistic regression analysis, which estimates average relationships between outcome and predictor variables, taking into account the hierarchical nature of the data.20
Gestational age, gender, and birth order–specific birth weight percentiles were calculated by separate linear regression analyses of birth weight against gestational age (days) within strata of whole gestational weeks. The percentiles were calculated by standard deviation scores or z scores obtained by the regression models, using method of scaled absolute residuals to model standard deviation against gestational age.21 Birth weight was power-transformed to normality (birth weight [g]1.87).
To estimate the proportions of all cases of stillbirth and late abortion in the second birth attributable to a history of preterm birth, SGA, preeclampsia, or abruptio placenta, population-attributable risk percentages were estimated as 100×(incidence in the population−incidence in the exposed group)/incidence in the population.22 The statistical analyses were carried out with SPSS (SPSS Inc., Chicago, IL) and the MlWin program (Centre for Multilevel Modeling, University of Bristol, UK).
We received an exemption from the research ethics committee of the University of Bergen because our study is based on anonymous registry data.
Women with stillbirth in the second pregnancy were more often aged less than 20 years and 40 years or more, less often married, had lower education, and were more often smokers than were the women with live births. From 1999, the proportions of chronic renal disease, prepregnancy diabetes, asthma, rheumatoid arthritis, epilepsy, and chronic heart disease were statistically nondifferent (Table 1). Preterm birth, SGA, preeclampsia, and abruptio placentae in the first pregnancy were all associated with a subsequent stillbirth or late abortion. The excess stillbirth rate after a live preterm birth increased with decreasing gestational age of the previous birth. After an extremely early preterm birth (22–27 weeks of gestation) the rate was six times greater than after a term birth (OR 5.7, 95% confidence interval [CI] 4.2–7.6). In a supplementary analysis, we compared the effects of preterm birth on later stillbirth in three cohorts of first births with spontaneous onset (n=523,674, 86%), induction of labor (n=81,950, 13%), or elective cesarean delivery (n=6,333, 1%). For each cohort, the effects were similar.
Adjusted ORs of stillbirth subsequent to pregnancies with SGA, preeclampsia, and abruptio placentae were 1.7 (95% CI 1.6–1.9), 1.6 (95% CI 1.5–1.9), and 2.8 (95% CI 2.2–3.5), respectively. The associations of SGA and preeclampsia with later stillbirth were greater in preterm than in term pregnancies (Table 2). For stillbirth before and during labor and with unspecified time of death, the associations were similar (data not presented). Combinations of gestational age below 33 weeks with preeclampsia carried sixfold to ninefold increased risk of later stillbirth (OR 9.0, 95% CI 4.4–18.4 for 22–27 weeks and 5.9, 95% CI 4.1–8.5 for 28–32 weeks compared with 37 weeks or more without preeclampsia, not presented). Combinations of gestational age below 33 weeks of gestation with SGA carried a 6-fold to 13-fold effect on later stillbirth (Fig. 1). The interaction of preterm birth with SGA was statistically significant (P<.001). We also assessed the risk of stillbirth without SGA, preeclampsia, and abruptio placentae in the second pregnancy (Table 3). The excess risks associated with the risk factors persisted.
After extremely preterm birth (less than 28 weeks of gestation), ORs for a subsequent late abortion (16–21 weeks of gestation) and stillbirth (22–28 weeks of gestation) were high (8.1 and 11.8) and higher than that of subsequent later preterm stillbirth (Table 4). Preterm birth was not associated with later term stillbirth. This pattern was less evident after SGA, preeclampsia, and abruptio placentae births, although preeclampsia and abruptio placenta had strongest effects on stillbirth between 22 and 36 weeks of gestation, while the effects on late abortion and term stillbirth were less or nonsignificant. Preterm SGA also had a strong effect on late abortion (OR 4.1, 95% CI 2.6–6.7), while term SGA had a small effect on stillbirth (OR 1.3, 1.1–1.7) (Table 4).
The associations between the conditions in a first pregnancy and stillbirth in a subsequent were observed throughout the observation period (Table 5).
The proportion of birth defects in first live births were 5.2%, 6.1%, 4.7%, and 3.3% for 22–27, 28–32, 33–36, and 37 or more weeks of gestation, respectively. Exclusion of these births had negligible effects on the associations.
The excess risk of stillbirth after preterm birth, SGA, preeclampsia, and abruptio placentae persisted after change of paternity from one pregnancy to the next (Table 6). A man who fathered a preterm birth or abruptio placenta with one woman was not more likely to father a stillbirth in a pregnancy with another woman. However, for pregnancies with preterm preeclampsia, the association was significant (OR 2.4, 95% CI 1.1–5.5) (Table 6).
Restriction of second births to those that occurred from 1999 onward, when the revised notification form was introduced, allowed adjustment for smoking and several maternal conditions, which had negligible effects on the associations (data not presented).
Population-attributable risk percentages for a history of preterm birth without stillbirth, SGA, preeclampsia, and abruptio placentae, were 5.0%, 4.9%, 2.4%, and 0.9%, respectively. The population-attributable risk percentage for a history of any of the four risk factors was 12.4% compared with the population-attributable risk percentage of 5.8% for a history of stillbirth.
Women with a live preterm birth, SGA, preeclampsia, or abruptio placentae in a first pregnancy had increased risk of stillbirth in a subsequent pregnancy. Previous preterm birth had a synergistic effect with SGA. The associations were particularly strong for stillbirth before 28 weeks of gestation and remained more or less unchanged throughout the observation period 1967–2005. With the exception of the effect of preterm preeclampsia on later stillbirth, no significant paternal effects on the associations were found. Altogether 12% of all stillbirths in a subsequent pregnancy could be attributed to these four risk factors.
The strengths of the study include the prospective collection of data using a validated database23 and its population-based design. The national identification number allowed linkage of all pregnancies. Stillbirths are reported from 16 weeks of gestation. At the lower gestational ages, ascertainment of stillbirths may not be complete, but underreporting would most likely be nondifferential and not affect the relative risks. Another strength was that our database holds data on several possible confounders. Adjustment for these variables and year of birth in the subsequent pregnancy had small effects on the OR estimates. Lack of data precluded adjustment for other possible confounders such as maternal obesity and thrombophilia. It is well known that maternal obesity is associated with stillbirth and preeclampsia. Thus, some of the increased risk of stillbirth after pregnancies with preeclampia could be explained by obesity. However, because obesity tends to increase fetal weight, adjusting for obesity would increase rather than decrease the effect of SGA on later stillbirth. Besides, although obesity has become more prevalent in Norway, the population of pregnant women is still relatively lean.24 Maternal thrombophilia has been associated with stillbirth as well as the risk factors preterm birth, fetal growth restriction, preeclampsia, and abruptio placentae, and might explain some of the effect, but the associations are inconsistent or weak.25
Surkan et al11 reported similar effects of preterm birth and fetal growth restriction on stillbirth in the next pregnancy. However, their Swedish Medical Birth Registry data covered only births at 28 weeks of gestation or above. A large Scottish study assessed the associations of preterm birth, preeclampsia, and SGA in first live births with “unexplained” stillbirth (without direct causes such as preeclampsia, antepartum hemorrhage, mechanical, and maternal) in the second birth.12 The case definition of “unexplained” stillbirth in the second pregnancy included 70% of all stillbirths, which is similar to the proportion in the present study (67%) of stillbirth without SGA, preeclampsia, and abruptio placentae. Although the Scottish study excluded preeclampsia, SGA, and abruptio placentae severe enough to cause stillbirth in first pregnancies, the results were consistent with the present study.
To prevent a stillbirth in a current pregnancy, not only a previous stillbirth should alert the clinician, but also previous preterm birth, fetal growth restriction, preeclampsia, and abruptio placenta. Counseling should be offered because of the excess risk of stillbirth, but also a higher risk of other placental dysfunction related conditions.11,14,26,27 Because the complications are associated with placental dysfunction, the results of the present study provide support for closely monitoring subsequent pregnancies with respect to fetal well-being.
Although stillbirth and the risk factors studied have complex etiological backgrounds, evidence suggests that at least some cases have shared environmental or genetic mechanisms. Among possible shared environmental causes is recurrent chorioamnionitis, which has been associated with stillbirth, preterm birth, and fetal growth restriction.28 Infection has been estimated to cause 9–15% of all stillbirths and may be especially important as a cause of stillbirth occurring early in pregnancy.2
Another possible shared mechanism is placental dysfunction which has been associated with all the risk factors as well as with stillbirth. Normally, trophoblasts invade the maternal spiral arteries early in pregnancy, resulting in increased diameter; flaccid, unresponsive venous-like vessels; and greatly increased placental perfusion. Morphologic studies indicate that fetal growth restriction and preeclampsia are caused by shallow invasion of the trophoblasts.7 Later, these histologic characteristics have been extended to include about one third of cases of preterm birth and as much as 70% of births at 22–28 weeks of gestation,8 which is consistent with clinical and epidemiological evidence that fetal growth restriction is associated with excess risk of preterm birth (less than 37 weeks of gestation).6 Inadequate trophoblast invasion has also been found in abruptio placentae,9 although the women included also had preeclampsia or fetal growth restriction, which is also associated with inadequate placentation. Longitudinal studies have provided evidence that preeclampsia, preterm birth, abruptio placentae, and fetal growth restriction share an etiological factor that tends to recur in the same women from one pregnancy to the next.10,19,26,27 The results of the present study are consistent with a mechanism in which decidual vasculopathy or placental dysfunction represent such a common path.
A man who fathered a pregnancy with preterm preeclampsia in one woman was significantly more likely to father a stillborn fetus in another woman, which is consistent with the earlier reported paternal recurrence of preeclampsia.14
In conclusion, the present study indicates that a history of live preterm birth, fetal growth restriction, preeclampsia, or abruptio placenta, and particularly severe conditions, is strongly associated with later stillbirth, even without recurrence of the conditions. Because the conditions and stillbirth have been associated with placental dysfunction, our results provide additional support for closely monitoring subsequent pregnancies with respect to fetal well-being.
1. Rasmussen S, Albrechtsen S, Irgens LM, Dalaker K, Maartmann-Moe H, Vlatkovic L, et al. Risk factors for unexplained antepartum fetal death in Norway 1967–1998. Early Hum Dev 2003;71:39–52.
2. Fretts RC, Boyd ME, Usher RH, Usher HA. The changing pattern of fetal death, 1961–1988. Obstet Gynecol 1992;79:35–9.
3. Cnattingius S, Haglund B, Kramer MS. Differences in late fetal death rates in association with determinants of small for gestational age fetuses: population based cohort study. BMJ 1998;316:1483–7.
4. Basso O, Rasmussen S, Weinberg CR, Wilcox AJ, Irgens LM, Skjærven R. Trends in fetal and infant survival following preeclampsia. JAMA 2006;296:1357–62.
5. Rasmussen S, Irgens LM, Bergsjø P, Dalaker K. Perinatal mortality and case fatality after placental abruption in Norway 1967–1991. Acta Obstet Gynecol Scand 1996;75:229–34.
6. Rasmussen S, Kiserud T, Albrechtsen S. Foetal size and body proportion at 17–19 weeks of gestation and neonatal size, proportion, and outcome. Early Hum Dev 2006;82:683–90.
7. Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascular response to placentation in pregnancies complicated by preeclampsia and by small-for-gestational age infants. Br J Obstet Gynaecol 1986;93:1049–59.
8. Salafia CM, Vogel CA, Vintzileos AM, Bantham KF, Pezzullo J, Silberman L. Placental pathologic findings in preterm birth. Am J Obstet Gynecol 1991;165:934–8.
9. Dommisse J, Tiltman AJ. Placental bed biopsies in placental abruption. Br J Obstet Gynaecol 1992;99:651–4.
10. Rasmussen S, Irgens LM, Dalaker K. A history of placental dysfunction and risk of placental abruption. Paediatr Perinat Epidemiol 1999;13:9–21.
11. Surkan PJ, Stephansson O, Dickman PW, Cnattingius S. Previous preterm and small-for-gestational-age births and the subsequent risk of stillbirth. N Engl J Med 2004;350:777–85.
12. Smith GCS, Shah I, White IR, Pell JP, Dobbie R. Previous pre-eclampsia, preterm delivery and delivery of a small for gestational age infant and the risk of unexplained stillbirth in the second pregnancy: a retrospective cohort study, Scotland 1992–2001. Am J Epidemiol 2007;165:194–202.
13. Salihu HM, Sharma PP, Aliyu MH, Kristensen S, Grimes-Dennis J, Kirby RS, et al. Is small for gestational age a marker of future fetal survival in utero? Obstet Gynecol 2006;107:851–6.
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. Magnus P, Gjessing HK, Skrondal A, Skjærven R. Paternal contribution to birth weight. Epidemiol Community Health 2001;55:873–7.
16. Lie RT, Wilcox AJ, Skjærven R. Maternal and paternal influences on length of pregnancy. Obstet Gynecol 2006;107:880–5.
17. Irgens LM. The Medical Birth Registry of Norway. Epidemiological research and surveillance throughout 30 years. Acta Obstet Gynecol Scand 2000;79:435–9.
18. National High Blood Pressure Education Program Working Group. High blood pressure in pregnancy. Am J Obstet Gynecol 1990;163:1691–712.
19. Rasmussen S, Irgens LM. Occurrence of placental abruption in relatives. BJOG 2009;116:693–9.
20. Goldstein H, Browne W, Rasbash J. Multilevel modelling of medical data. Stat Med 2002;21:3291–315.
21. Altman DG. Construction of age-related reference centiles using absolute residuals. Stat Med 1993;12:917–24.
22. Miettinen OS. Proportion of disease caused or prevented by a given exposure, trait or intervention. Am J Epidemiol 1974;99:325–32.
23. Gissler M, Louhiala P, Hemminki E. Nordic Medical Birth Registers in epidemiological research. Eur J Epidemiol 1997;13:169–75.
24. Samuelson G. Dietary habits and nutritional status in adolescents over Europe. An overview of current studies in the Nordic countries. Eur J Clin Nutr 2000;54:S21–8.
25. Kist WJ, Janssen NG, Kalk JJ, Hague WM, Dekker GA, de Vries JI. Thrombophilias and adverse pregnancy outcome - A confounded problem! Thromb Haemost 2008;99:77–85.
26. Rasmussen S, Irgens LM, Albrechtsen S, Dalaker K. Predicting preeclampsia in the second pregnancy from low birth weight in the first pregnancy. Obstet Gynecol 2000;96:696–700.
27. Rasmussen S, Irgens LM. History of fetal growth restriction is more strongly associated with severe rather than milder pregnancy-induced hypertension. Hypertension 2008;51:1231–8.
28. Williams MC, O’Brien WF, Nelson RN, Spellacy WN. Histologic chorioamnionitis is associated with fetal growth restriction in term and preterm infants. Am J Obstet Gynecol 2000;183:1094–9.
Figure. No caption available.