Stillbirths complicate 3.5–6.6 of 1,000 pregnancies in the United States, United Kingdom, and Israel,1–3 and they make up about half of all perinatal deaths. Since the 1960s, a decrease in stillbirths has been documented and has been attributed to improvements in prenatal screening and treatments.4 The exact mechanism underlying most stillbirths is unclear, and a substantial proportion of stillbirths remains unexplained, even after thorough investigation.5,6 Antepartum fetal death has been associated with maternal, fetal, and pregnancy-related factors, many of which are interrelated. About 10% of stillbirths are associated with maternal diseases such as diabetes, hypertension, systemic lupus erythematosus, and renal diseases.7 Fetal factors include congenital malformations and intrauterine growth restriction (IUGR).8,9 Pregnancy-related conditions include infections, placental abruption, and preeclampsia.8
There is growing evidence that both preeclampsia and low birth weight are associated with increased cardiovascular morbidity and mortality of mothers.10,11 A recent study12 demonstrated an increased risk for cardiovascular hospitalizations among women who have had either placental abruption or preeclampsia; these were doubled when the women also experienced stillbirths. Little is known about the long-term sequelae of mothers whose infants were stillborn. Although maternal conditions are expected to be associated with maternal outcomes, it is not clear whether factors related to fetal outcome, such as stillbirths, have long-term implications for the health of mothers. The aim of this study was to assess the association between stillbirths and long-term maternal all-cause and cause-specific mortality.
PARTICIPANTS AND METHODS
The Jerusalem Perinatal Study is a population based research cohort of all births to residents of western Jerusalem and its surroundings between 1964 and 1976. The database includes demographic, obstetric, and neonatal information about 92,408 births, including 949 stillbirths. When the cohort was initiated, stillbirth was defined as fetal loss in a pregnancy beyond 28 weeks of gestation, or if gestational age was not known, a loss of a fetus weighing more than 1,000 g, corresponding with the current definition of late stillbirths.8
Detailed information on data collection has been previously described.13,14 Briefly, data on all births were copied from birth notifications, and for 91% of participants, data were also abstracted from maternity ward logbooks. Additional information on lifestyle and maternal medical conditions was collected by antenatal and postnatal interviews of subsets (31%) of the study population, selected according to year of delivery. All data were collected by using structured questionnaires, completed according to precise instructions and checked by the nurse supervisor and statistician of the survey team. Medical information was verified, where necessary, with the obstetric nurses and medical staff of the hospitals. Data from various sources were checked for consistency and logical errors. Information was recorded per birth with mother's and offspring's identity numbers as their unique identifiers. These identity numbers enabled data linkage with the Israel Population Registry to ascertain vital status until December 31, 2004, as well as dates and causes of deaths. These linkages with the Israeli Population Registry enabled further validation of demographic data. Altogether, we were able to verify the identities and vital status of 41,206 mothers (95.9% of the mothers in the original cohort). Because fetuses who died in utero did not receive identity numbers and stillbirths were not recorded in the Israeli Population Registry, a mother could not be traced if her only delivery in 1964–1976 happened to be a stillbirth.
Mothers to stillbirths were traced in the Israeli Population Registry only if they had in addition a live birth between 1964 and 1976. Therefore, we restricted the study to the 25,118 mothers who had at least two deliveries during data collection, with at least one live offspring. We followed mothers since their first birth in the cohort and compared the characteristics and survival of mothers who had a stillbirth (n=595) with mothers who did not experience any stillbirth (n=24,523). The study was approved by the institutional review boards at Hadassah-Hebrew University and Columbia University.
Maternal demographic and social variables included age at first birth in the cohort and social class according to the father's occupation. Ethnic origin was classified according to religion as Jewish or non-Jewish. Jewish mothers were further classified according to their fathers’ country of birth as: Israel, West Asia, North Africa, and Europe/America. Parity was entered into the models as the maximal number of births a mother had up to the end of data collection, including births which took place before the cohort inception (categories: 2, 3–4, and 5 or more births).
Maternal medical conditions such as heart disease, diabetes, or other medical conditions were analyzed separately and grouped together as a dichotomous variable, defined as a mother having any of the above conditions. Pregnancy-related conditions, including gestational diabetes, preeclampsia, and placental abruption at any pregnancy, were assessed separately. Placental syndrome was defined as having either preeclampsia or abruption at any pregnancy. Adverse obstetric outcomes included having at least one offspring with birth weight below 2,500 g (excluding stillborn offspring) and having at least one offspring with congenital malformations (excluding stillborn offspring). Data were 99.5% complete for all variables except smoking; data on smoking (ever/never) were available for 64.4% of the mothers.
Underlying causes of death were available in the Population Registry for deaths in 1978–1998. Causes were coded according to the International Classification of Diseases, 9th Revision (ICD-9) codes until 1997, and in 1998, according to International Classification of Diseases, 10th Revision (ICD-10) codes.15,16 Causes of deaths were categorized into the following groups: coronary heart disease (ICD-9 codes 410–414, 427.4, and 427.5; ICD-10 codes I20–I25, I46, and I49); all circulatory conditions, including coronary heart diseases (ICD-9 codes 390 through 459; ICD-10 codes I0–I99); all neoplasms (ICD-9 codes 140–239; ICD-10 codes C0–C99 and D0-D48), death due to all other causes taken together and death due to renal causes (ICD-9 codes 580–629; ICD-10 codes N0–N99).
The association between stillbirths and all-cause mortality of mothers was estimated by using the Kaplan-Meier method. The difference in survival has been evaluated by the log-rank test. Time was counted from first birth in the cohort until death or conclusion of follow-up (April 30, 2005). Cox proportional hazards models with time-to-event as the dependent variable were used to calculate the hazard ratio after controlling for study covariates. The association between stillbirths and cause-specific mortality was estimated until December 31, 1998, by using the Cox proportional hazards model.
We examined these models among the entire study population of 25,118 women and in strata of maternal age categories, socioeconomic levels, ethnic origin categories, presence or absence of maternal medical conditions, and placental syndrome, number of stillbirths, and birth order of stillborn, to explore the possibility of effect modifications.
Between 1964 and 1976, 595 women had 639 stillbirths, representing a stillbirth rate of 8.7 per 1,000 births. Compared with mothers who had only live births during data collection, mothers who had stillbirths were older, had higher parity, were less educated and of lower socioeconomic status, had overall larger families, and were less likely to be of European/American origin (Table 1). Mothers who had stillbirths were more likely to have suffered from heart disease, type 1 diabetes or other medical conditions before pregnancy, and to develop preeclampsia, gestational diabetes, and placental abruption during pregnancy. They were also more likely to have had low birth weight offspring compared with mothers who had only live births.
During a median follow up of 36.5 years (interquartile range 33.2–39.5), the 25,118 mothers contributed 900,926 person years of observation. Altogether, as of April 30, 2005, 1,596 mothers had died at a median age of 56.5 years (interquartile range 48.6–63.5), 78 (13.1%) mothers who had had stillbirths and 1,518 (6.2%) mothers who had no stillbirth, yielding mortality rates of 36.5 (95% confidence interval [CI] 28.35–44.65) and 17.26 (95% CI 16.39–18.13) per 10,000 person years, respectively. Figure 1 presents the Kaplan-Meier survival curves of women with and those without stillbirths and demonstrates a reduced survival for women who experienced stillbirths (P<.001, log rank test). The crude hazard ratio for death during follow-up was 2.08 (95% CI 1.65–2.61) for mothers who had stillbirths compared with mothers who had no stillbirth.
Table 2 presents the crude and age-adjusted association of stillbirth with mortality of mothers stratified by the study covariates. The association between stillbirths and mortality was strongest among women in the lowest socioeconomic category (age adjusted hazard ratio 2.04, 95% CI 1.39–3.00) and among women of North-African origin (age adjusted hazard ratio 3.16, 95% CI 2.19–4.56). Mothers whose first birth was a stillbirth had an age adjusted hazard ratio of 1.80 (95% CI 1.02–3.19), similar to the hazard ratio of mothers who had stillbirth in subsequent births (age adjusted hazard ratio 1.79, 95% CI 1.40–2.29).
After controlling for age, socioeconomic status, parity, maternal conditions at cohort assembly, and placental syndrome, mothers who had stillbirths had a hazard ratio of 1.40 (95% CI 1.11–1.77) for all-cause mortality (Table 3). Introduction of maternal education, ethnic origin, smoking, gestational diabetes, congenital malformations, or having low birth weight offspring into the models did not materially change the results. Restriction of the analysis to mothers who had underlying causes of death had trivial effect on the results.
Among mothers to stillbirths, 39 had more than one stillbirth. Excluding these women from the analysis did not materially change the association between stillbirth and mortality (adjusted hazard ratio 1.45, 95% CI 1.14–1.84).
Mothers who had stillbirths were at increased risk for death due to coronary heart diseases (adjusted hazard ratio 2.00, 95% CI 1.02–3.93), all circulatory causes (hazard ratio 1.70, 95% CI 1.02–2.84), and renal causes (hazard ratio 4.70, 95% CI 1.47–15.0) causes (Table 3). They had a nonsignificantly increased risk of death due to cancer (adjusted hazard ratio 1.29, 95% CI 0.78–2.15) and no increased risk of death due to other causes (Table 3).
Although the introduction of origin into the models did not change the association between stillbirth and mortality, there was an interaction between origin and stillbirth in the association with mortality (P for heterogeneity=.011). This heterogeneity was attributed to the association of stillbirth and mortality among women of North African origin. Only among women of North African origin was the adjusted hazard ratio for all-cause mortality after stillbirth significantly increased (hazard ratio 2.47, 95% CI 1.69–3.63). Jewish women of West-Asian and Israeli origin who had stillbirth had a nonsignificantly increased risk of death (hazard ratios of 1.14 and 1.51, respectively), and a nonsignificantly decreased risk was evident among women of European-American origin (hazard ratio 0.82, 95% CI 0.42–1.60). Among non-Jewish women who had stillbirth, only one had died, precluding meaningful analysis of the effect of stillbirth.
Mothers of North African origin who had stillbirth had increased risk of death from coronary heart diseases (adjusted hazard ratio 5.16, 95% CI 2.06–12.91), all circulatory causes (hazard ratio 2.69, 95% CI 1.19–6.09), cancer (hazard ratio 2.90, 95% CI 1.31–6.41), and renal causes (hazard ratio 9.26, 95% CI 1.63–52.44).
Our study demonstrated an increased risk of long-term mortality for women who experienced late fetal deaths. This study joins a growing body of evidence on the association between obstetric complications and events, especially preeclampsia and low birth weight, and maternal morbidity and mortality later in life. We and others have demonstrated that preeclampsia is associated with long-term maternal mortality, especially cardiovascular mortality.10,17 Smith et al19 found additive associations between birth weight and preeclampsia and the mother's risk of mortality associated with ischemic heart disease. Ray et al12 demonstrated an increased risk for cardiovascular hospitalizations in a median of 8.7 years of follow-up for mothers with placental syndrome (hazard ratio of 1.9–3.0). Having, in addition, an offspring with IUGR was associated with a hazard ratio of 3.1, whereas having the placental syndrome was associated with a hazard ratio of 4.4 (95% CI 2.4–7.9) for intrauterine fetal death. There are reports on the association between low birth weight of an offspring and long-term maternal mortality, especially cardiovascular mortality.11,19 In our cohort, we found a U-shaped association between birth weight and maternal mortality (Friedlander Y, Paltiel O, Manor O, Deutsch L, Yanetz R, Calderon-Margalit R, et al. Birthweight of offspring and mortality of parents: The Jerusalem Perinatal Study Cohort. Am Epidemiol. In press).
Given that they are associated with both stillbirth and long-term morbidity and mortality of mothers, preeclampsia, placental abruption, and low birth weight may serve either as intervening events or confounders in the association of stillbirth with long-term mortality. In this study, introduction of preeclampsia and placental abruption into the analysis attenuated the association of stillbirths with mortality, but the association remained significantly increased. Having a low birth weight offspring did not change the studied association. Therefore, we assume that the association of stillbirths with mortality is partially mediated through the same mechanisms that underlie the association of preeclampsia with long-term mortality of mothers.
Such mechanisms could include prepregnancy obesity, because increased body mass index (BMI) is associated with several adverse outcomes of pregnancy, as well as with long-term mortality. Women with high BMI have been found to be at increased risk for IUGR and stillbirths.20 Obesity is associated with the development of the metabolic syndrome—ie, dyslipidemia, hypertension, and type 2 diabetes,21 the latter being the leading cause for renal insufficiency,22 end-stage real disease,23 and associated mortality.22 Diabetic patients were shown to be at increased risk for deaths due to cardiovascular diseases and renal insufficiency.24 Our study limitations include lack of information regarding obesity, incidence of diabetes, or incidence of renal diseases after the cohort assembly. We considered gestational diabetes as a proxy for metabolic changes associated with prepregnancy weight25 and with the development of type 2 diabetes later in life.26 Controlling for gestational diabetes in the multivariable analysis did not materially change the association between stillbirths and mortality.
Our study suggests that stillbirth is a risk marker for premature mortality in fertile women. However, one must keep in mind that this study is based on stillbirths occurring in the 1960s and 1970s. The rates of stillbirths have been dropping since the 1960s in Israel,3 as well as in other developed countries.8 It is reasonable to assume that at least some of the stillbirths that occurred during our cohort assembly would be prevented today with improvements in antenatal diagnosis and care. Nevertheless, we assume that the association between fetal deaths and late maternal mortality is not causal but rather may represent maternal metabolic abnormalities that underlie both stillbirths and the development of fatal illnesses later in life.
The Jerusalem Perinatal Study cohort provides a unique opportunity to study long-term implications of obstetric events and complications in a population of diverse ethnicities, including a large proportion of individuals of West Asian and North African origin. This population has already provided novel findings on the association between preeclampsia and cancer,27 findings that were contrary to those of Northern European cohorts. Indeed, in our study, stratification by country of origin demonstrated a significant association between stillbirths and long-term mortality, especially among mothers of North African origin. A previous study in Israel demonstrated that socioeconomic status was a major confounder in the association between country of origin and mortality,28 but in our study, country of origin was not a confounder but rather a modifier.
Women of North African origin who had stillbirth experienced an increased risk of death from cancer, as well as from cardiovascular and renal causes. The modification by origin seen in our study could be attributed to genetic factors unique to North African Jews, such as the ataxia telangiectasia (ATM) gene. Jews of Moroccan origin have a carrier rate of mutated ATM alleles of 0.6–12.5%.29–31 Although no association has been described between carriership of ATM mutations and stillbirth, it is interesting to mention that, although carriers of ATM are considered to be asymptomatic, they have been found to have increased risk of deaths due to all causes combined as well as deaths from cancer and from ischemic heart diseases.32
Our study's limitations include information on stillbirths only during data collection (1964–1976) and not throughout the entire reproductive history of the study population. Therefore, there might be events of stillbirth among women considered to be nonexposed either before or after data collection. However, when data analysis was restricted to women who had a stillbirth in their first birth, the association remained practically unchanged, and there was no effect modification by birth order of stillbirth.
Our study suggests that women who experienced stillbirths might be at increased risk for long-term mortality. Additional research is required to establish whether this is a universal phenomenon or one that is specific to certain ethnic groups. If our findings are confirmed, women who experience stillbirths make up a population at risk who will require careful follow-up and perhaps screening for cardiovascular and renal diseases. Knowledge of this possible risk factor may have implications for the primary care of women in the setting of the obstetric and gynecology clinic.33,34
We performed a PubMed search. Search terms used included “fetal death(s),” “stillbirth(s),” “mortality,” “risk,” “mothers,” “maternal,” “long-term,” and “outcomes.” We also searched via references of papers on stillbirths and papers on mortality of parents after preeclampsia or low birth weight. In addition, we looked through the “related articles” search of these papers. There was no time or language limit on the search.
This study is the first report on the association between stillbirths and long-term maternal mortality. Further studies are needed to support our findings and to elucidate the mechanisms underlying the associations between reproductive complications and events and premature death in women.
1. Salihu HM, Kinniburgh BA, Aliyu MH, Kirby RS, Alexander GR. Racial disparity in stillbirth among singleton, twin, and triplet gestations in the United States. Obstet Gynecol 2004;104:734–40.
2. Bell R, Glinianaia SV, Rankin J, Wright C, Pearce MS, Parker L. Changing patterns of perinatal death, 1982–2000: a retrospective cohort study. Arch Dis Child Fetal Neonatal Ed 2004;89:F531–6.
3. Marriages, divorces, live births, deaths, natural increase, infant deaths and stillbirths, by population group and religion. Table 3.1. Statistical Abstract of Israel, No. 56, 2005. Available at: http://www1.cbs.gov.il/shnaton56/st03_01.pdf
. Retrieved March 20, 2007.
4. Fretts RC, Boyd ME, Usher RH, Usher H. The changing pattern of fetal death, 1961–1988. Obstet Gynecol 1992;79:35–9.
5. Cnattingius S, Stephansson O. The epidemiology of stillbirth. Semin Perinatol 2002;26:25–30.
6. Huang DY, Usher RH, Kramer MS, Yang H, Morin L, Fretts RC. Determinants of unexplained antepartum fetal deaths. Obstet Gynecol 2000;95:215–21.
7. Simpson LL. Maternal medical disease: risk of antepartum fetal death. Semin Perinatol 2002;26:42–50.
8. Fretts RC. Etiology and prevention of stillbirth. Am J Obstet Gynecol 2005;193:1923–35.
9. Froen JF, Gardosi JO, Thurmann A, Francis A, Stray-Pedersen B. Restricted fetal growth in sudden intrauterine unexplained death. Acta Obstet Gynecol Scand 2004;83:801–7.
10. Funai EF, Friedlander Y, Paltiel O, Tiram E, Xue X, Deutsch L, et al. Long-term mortality after preeclampsia. Epidemiology 2005;16:206–15.
11. Smith GD, Harding S, Rosato M. Relation between infants’ birth weight and mothers’ mortality: prospective observational study. BMJ 2000;320:839–40.
12. Ray JG, Vermeulen MJ, Schull MJ, Redelmeier DA. Cardiovascular health after maternal placental syndromes (CHAMPS): population-based retrospective cohort study. Lancet 2005;366:1797–803.
13. Harlap S, Davies AM, Grover NB, Prywes R. The Jerusalem perinatal study: the first decade 1964–73. Isr J Med Sci 1977;13:1073–91.
14. Davies AM, Prywes R, Tzur B, Weiskopf P, Sterk VV. The Jerusalem Perinatal Study: 1. Design and organization of a continuing, community-based, record-linked survey. Isr J Med Sci 1969;5:1095–106.
17. Irgens HU, Reisaeter L, Irgens LM, Lie RT. Long term mortality of mothers and fathers after pre-eclampsia: population based cohort study. BMJ 2001;323:1213–7.
18. Smith GC, Pell JP, Walsh D. Pregnancy complications and maternal risk of ischaemic heart disease: a retrospective cohort study of 129,290 births. Lancet 2001;357:2002–6.
19. Smith GD, Whitley E, Gissler M, Hemminki E. Birth dimensions of offspring, premature birth, and the mortality of mothers. Lancet 2000;356:2066–7.
20. Cnattingius S, Bergstrom R, Lipworth L, Kramer MS. Prepregnancy weight and the risk of adverse pregnancy outcomes. N Engl J Med 1998;338:147–52.
21. Schunkert H. Obesity and target organ damage: the heart. Int J Obes Relat Metab Disord 2002;26:S15–20.
22. Evans M, Fryzek JP, Elinder CG, Cohen SS, McLaughlin JK, Nyren O, et al. The natural history of chronic renal failure: results from an unselected, population-based, inception cohort in Sweden. Am J Kidney Dis 2005;46:863–70.
23. Stengel B, Billon S, Van Dijk PC, Jager KJ, Dekker FW, Simpson K, Briggs JD. Trends in the incidence of renal replacement therapy for end-stage renal disease in Europe, 1990-1999. Nephrol Dial Transplant 2003;18:1824–33.
24. Tseng CH. Mortality and causes of death in a national sample of diabetic patients in Taiwan. Diabetes Care 2004;27:1605–9.
25. Seghieri G, De Bellis A, Anichini R, Alviggi L, Franconi F, Breschi MC. Does parity increase insulin resistance during pregnancy? Diabet Med 2005;22:1574–80.
26. Kim C, Newton KM, Knopp RH. Gestational diabetes and the incidence of type 2 diabetes: a systematic review. Diabetes Care 2002;25:1862–8.
27. Paltiel O, Friedlander Y, Tiram E, Barchana M, Xue X, Harlap S. Cancer after pre-eclampsia: follow up of the Jerusalem perinatal study cohort. BMJ 2004;328:919.
28. Manor O, Eisenbach Z, Israeli A, Friedlander Y. Mortality differentials among women: the Israel Longitudinal Mortality Study. Soc Sci Med 2000;51:1175–88.
29. Gilad S, Bar-Shira A, Harnik R, Shkedy D, Ziv Y, Khosravi R, et al. Ataxia-telangiectasia: founder effect among north African Jews. Hum Mol Genet 1996;5:2033–7.
30. Kreiss Y, Barak F, Baruch RG, Levy-Lahad E, Pras E, Friedman E. The founder mutations in the BRCA1, BRCA2, and ATM genes in Moroccan Jewish women with breast cancer. Genet Test 2000;4:403–7.
32. Su Y, Swift M. Mortality rates among carriers of ataxia-teleangiectasia mutant alleles. Ann Intern Med 2000;133:770–8.
33. Scholle SH, Kelleher K. Assessing primary care performance in an obstetrics/gynecology clinic. Women Health 2003;37:15–30.
34. Henderson JT, Weisman CS. Women's patterns of provider use across the lifespan and satisfaction with primary care coordination and comprehensiveness. Med Care 2005;43:826–33.