Lykke, Jacob Alexander MD; Dideriksen, Katrine Lehrmann MD; Lidegaard, Øjvind MD, DMSc; Langhoff-Roos, Jens MD, DMSc
First-trimester vaginal bleeding is one of the most common complications in pregnancy with an incidence of 15–25%. About half of these will end in miscarriage within 20 weeks of gestation,1,2 and those women who remain pregnant have an increased risk of developing other complications later in pregnancy.3–13
Local hemostatic factors in the uterus during implantation, decidualization, and early pregnancy, for example, tissue factor expressed in cytotrophoblasts,14 and systemic factors in the women during the ongoing pregnancy15 seem to play distinct roles in a successful pregnancy; dysfunction of any of these factors could lead to an adverse outcome16; for example, local formation of thrombin and soluble fms-like tyrosine kinase-1. Both of these seem to be involved in development of placental abruption and preeclampsia.14
We hypothesize that first-trimester bleeding is a marker of a general proclivity to other pregnancy complications surfacing later in the pregnancy. Also, this proclivity will resurface in the next pregnancy as either first-trimester bleeding or other pregnancy complications or both.
Several studies have investigated the consequences of bleeding on the risk of complications later in the same pregnancy but not the association between two pregnancies. Therefore, we have designed a registry-based, retrospective cohort study investigating 1) the association between first-trimester bleeding and complications later in the first pregnancy and 2) the association between first-trimester bleeding in the first pregnancy and later pregnancy complications in the second pregnancy as well as the opposite association.
MATERIALS AND METHODS
The National Patient Registry has collected information on all discharge diagnoses and complications during pregnancy and delivery in Denmark since 1978.17 We extracted information on all singleton deliveries in Denmark from January 1, 1978, to October 1, 2007, which accrued 1,795,806 deliveries of 965,475 women. From this population, we defined two cohorts free of cardiovascular morbidity, such as chronic hypertension and previous ischemic heart disease. Cohort 1 was defined as women aged 15 to 50 years with a first delivery (n=796,915), excluding those with a preceding cardiovascular diagnosis (n=11,605; 1.5%) or insulin-dependent and non–insulin-dependent diabetes (n=2,387; 0.3%) and women who died or emigrated within 3 months of delivery (n=65 and n=571, respectively). Cohort 1 consists of 782,287 women. Cohort 2 (a subpopulation of cohort 1) was defined as women with a second singleton delivery. Cohort 2 consists of 536,419 women.
The primary exposure was first-trimester bleeding defined as vaginal bleeding verified by a physician, that is, women referred to a clinic and diagnosed by a physician after having vaginal bleeding before 12 completed weeks of gestation. Gestational age was calculated from date of bleeding, date of delivery, and the reported gestational age at delivery. Of note, women with subsequent miscarriage or induced abortion before 20 completed weeks were not included in the study population. Women with bleeding and unknown gestational age were considered controls because we could not classify them as first-trimester bleedings. Likewise, many women with missing values of fetal growth were considered controls because of missing values in gestational age.
The outcomes (and secondary exposures) were preterm delivery, premature rupture of membranes (PROM), hypertensive pregnancy disorders, fetal growth, placental abruption, and stillbirth after 20 weeks of gestation. At delivery, gestational age and birth weight were recorded routinely. Initially, gestational age was assessed by the last menstruation period; gradually from 1978 to 2007, gestational age was determined by early second-trimester ultrasonography. A priori, we stratified preterm delivery into four groups by gestational age: 20–27, 28–31, and 32–37 completed weeks; we used deliveries after 37 completed weeks as the reference group. Fetal growth was measured by the birth weight standardized for sex and gestational age.18 Small for gestational age (SGA) and large for gestational age (LGA) were defined as fetal growth 2 standard deviations (SDs) below and above the mean, respectively. Of note, we did not standardize for parity, which produces a higher proportion of SGA in the first pregnancy compared with the second pregnancy.
Analyzing preterm delivery and PROM, we excluded pregnancies complicated by hypertensive pregnancy disorders, SGA, placental abruption, and stillbirth to accrue a surrogate for spontaneous delivery.
Implausible values of birth weight, gestational age, fetal growth, and the combination of these were reassigned as missing values (n=963). Missing values were analyzed as a separate group. These occurred more frequently in the earlier years of the Registry.
The presence of vaginal bleeding, PROM, hypertensive pregnancy disorders, placental abruption, stillbirth, and cardiovascular diseases were assessed by the specific International Classification of Diseases-8 and -10 codes for these diagnoses (see the Appendix, available online at http://links.lww.com/A171). Given the binomial nature of these variables in the Registry, no missing values were recorded. The hypertensive pregnancy disorders were stratified into gestational hypertension, mild preeclampsia, and severe preeclampsia (including eclampsia and HELLP syndrome). The definition of preeclampsia has changed little during the 30-year study period,19 and the frequency of preeclampsia in the Registry has remained almost stable. The accuracy of the diagnoses of the hypertensive pregnancy disorders in the Registry has in subpopulations been manually validated several times, accruing specificities above 99% for all types, but sensitivities at 10% for gestational hypertension and 69% for preeclampsia or positive predictive values of 56% for gestational hypertension, 100% for severe preeclampsia, and 74% for mild and severe preeclampsia combined.19
We used multivariable logistic regression to calculate the associations. We included maternal age and year of delivery in all models and years between pregnancies in associations across two pregnancies. Initially we calculated stratified results, some of which are presented in the article. All odds ratios (ORs) are presented with 95% confidence intervals (CIs). SPSS 16.0 for Macintosh (SPSS Inc., Chicago, IL) was used for all calculations. The study was approved by the Danish Data Protection Agency (2005–41–5262 and 2007–41–1544).
The mean age at delivery was 26.8 (SD 4.6) years in cohort 1, and 25.9 (SD 4.0) and 29.5 (SD 4.2) years in cohort 2, the first and second deliveries, respectively. In cohort 1, 18,311 (2.3%) women had bleeding; in cohort 2, 11,563 (2.2%) and 12,262 (2.3%) women had first-trimester bleeding in the first and second deliveries, respectively.
In cohort one, women with first-trimester bleeding had an increased risk of delivering preterm later in the same pregnancy. The proportion of women experiencing first-trimester bleeding increased by decreasing gestational age at delivery (Table 1). First-trimester bleeding increased the risk of preterm delivery in weeks 32–36 from 3.6% to 6.1% (OR 1.65; 95% CI 1.57–1.77) and in weeks 28–31 from 0.3% to 0.9% (OR 2.98; 95% CI 2.50–3.54). In addition, women with first-trimester bleeding had a 1.19-fold (95% CI 1.11–1.28) increased risk of PROM and 1.18-fold (95% CI 1.01–1.37) increased risk of preterm PROM, thus illustrating no effect modification by gestational age of PROM. Also, women with first-trimester bleeding had a 1.48-fold (95% CI 1.30–1.68) increased risk of placental abruption (Table 1).
In cohort 2, first-trimester bleeding in the first pregnancy increased the risk from 2.2% to 8.2% (OR 4.05; 95% CI 3.78–4.34) of first-trimester bleeding in the second pregnancy (Table 2). Also, first-trimester bleeding in the first pregnancy increased the risk of preterm delivery in the second pregnancy from 2.7% to 4.8% (OR 1.83; 95% CI 1.67–2.00), PROM from 3.1% to 4.1% (OR 1.40; 95% CI 1.27–1.54), placental abruption from 0.9% to 1.0% (OR 1.29; 95% CI 1.07–1.56), and stillbirth from 0.4% to 0.5% (OR 1.34; 95% CI 1.02–1.76) (Table 3). Only in the association with preterm delivery did we detect an effect modification by first-trimester bleeding in the second pregnancy and by preterm delivery in the first pregnancy (Tables 4 and 5); these factors diminished the effect of first-trimester bleeding in the first pregnancy. We also investigated the reverse association: women delivering preterm at gestational ages of 32–36 weeks in the first pregnancy had a 1.41-fold (95% CI 1.29–1.55) increased risk of first-trimester bleeding in the second pregnancy. Women delivering in gestational age 28–31 weeks had a 2.57-fold (95% CI 2.02–3.28) increased risk (Table 6). No effect modification by bleeding in the first pregnancy was observed in the stratified analyzes (Table 7).
Previous studies have linked first-trimester vaginal bleeding to preterm delivery,7–12 PROM,9,10,20 and placental abruption8,10,21 in the same pregnancy. We found that first-trimester bleeding in the first pregnancy tends to recur and also carries a risk of complications over to the second pregnancy. Also, preterm delivery, PROM, and placental abruption in the first pregnancy carry a risk over to the second pregnancy of first-trimester bleeding.
The association between SGA and first-trimester bleeding—in the first pregnancy and across first and second pregnancies—was small and inconsistent, and previous studies have had difficulty in demonstrating the association9–11; in our study, first-trimester bleeding did not increase the risk of LGA. Also, preeclampsia was weakly associated with first-trimester bleeding in the first pregnancy; Weiss et al previously have reported an increased risk of preeclampsia after “light bleeding,” but strangely not after “heavy bleeding,” and, in contrast, Eskild and Vatten found a protective effect of early bleeding on the risk of developing preeclampsia.10,22 Stillbirth was not associated with first-trimester bleeding in the first pregnancy; however, stillbirth in the first pregnancy increased the risk of first-trimester bleeding in the second pregnancy, although this could be ascribed to bias; Johns and Jauniaux9 and Williams et al7 did not find an association in their studies.
Impaired invasion of cytotrophoblasts and remodeling of the spiral arteries in early placentation have been demonstrated in pregnancies ending in miscarriage23 and also those pregnancies complicated by preterm delivery, PROM, and placental abruption24–26 and preeclampsia.27 The same impaired placentation could be the link between first-trimester bleeding and complications later in pregnancy.13 Another model, coined by Johns et al,13 centers on the actual bleeding in the placental bed: An iron deposit may provoke a production of excessive oxidative stress, which has been linked to preterm delivery and PROM28 and to preeclampsia.15 A nidus will make an infection more likely, which also has been linked to preterm delivery.28–30 Also, decidual bleeding will generate excess amount of thrombin from decidual-cell expressed tissue factor, which again could impede the ongoing implantation.14 These models are not mutually exclusive; we found that first-trimester bleeding in the first pregnancy carries the risk of complications over into the second pregnancy and vice versa. This suggests a common proclivity of first-trimester bleeding and preterm delivery, PROM, and placental abruption.
The strengths and weaknesses of the study are based on the nature of the employed National Patient Registry, a large population-based cohort representing a census of delivering women. An incidence of 2.3% of first-trimester bleeding seems low in comparison with previous studies1; for example, Mulik et al reported a 7.1% incidence.8 Although there was underreporting, this difference may be ascribed to the inclusion of only viable pregnancies beyond 20 weeks of gestation.2 Furthermore, the registry will include information only on women who sought medical assistance and were referred to a physician. In consequence, the woman herself could act as a bias: If a woman had serious complications in her first pregnancy, this experience could lower her threshold for seeking medical attention in second pregnancy. However, because of the prospective registration of the data, recall bias is not present.
The validity of the hypertensive diagnoses in the National Patient Registry, demonstrating positive predictive values of 56% of gestational hypertension and 74% for preeclampsia,19 warrants caution when interpreting the associations to these complications: The low-to-moderate predictive values will tend to underestimate the associations. Validation of PROM, placental abruption, and stillbirth remains to be performed.
Potential confounders such as socioeconomic status, assisted reproductive technology or fecundability, smoking, and body mass index could not be assessed in the present study. These could be risk factors for both vaginal bleeding and later pregnancy complications. Also, we could not distinguish between light and heavy bleeding,10,20 the presence of ultrasound-verified hematomas,13 or inflammation,28,30 all of which have been associated with adverse pregnancy outcomes. Thus, the weaker associations should be interpreted with caution. However, the potential bias affects the biological interpretation of the associations, not the statistical predictive value of first-trimester vaginal bleeding for later pregnancy complications.
In conclusion, first-trimester vaginal bleeding is a clinical, relevant event for the obstetrician as a marker for preterm delivery, PROM, and placental abruption in the index pregnancy as well as in the subsequent pregnancy; these findings add to the evidence of the linkage and recurrence of pregnancy complications.31 The linkage to these pregnancy complications may provide a basis for selective increased pregnancy surveillance as well as insight into the etiology of miscarriage, vaginal bleeding, and other pregnancy complications.
1. Everett C. Incidence and outcome of bleeding before the 20th week of pregnancy: prospective study from general practice. BMJ 1997;315:32–4.
2. Hasan R, Baird DD, Herring AH, Olshan AF, Jonsson Funk ML, Hartmann KE. Association between first-trimester vaginal bleeding and miscarriage. Obstet Gynecol 2009;114:860–7.
3. van Oppenraaij RH, Jauniaux E, Christiansen OB, Horcajadas JA, Farquharson RG, Exalto N, et al. Predicting adverse obstetric outcome after early pregnancy events and complications: a review. Hum Reprod Update 2009;15:409–21.
4. Saraswat L, Bhattacharya S, Maheshwari A, Bhattacharya S. Maternal and perinatal outcome in women with threatened miscarriage in the first trimester: a systematic review. BJOG 2010;117:245–57.
5. Ananth CV, Savitz DA. Vaginal bleeding and adverse reproductive outcomes: a meta-analysis. Paediatr Perinat Epidemiol 1994;8:62–78.
6. Hertz JB, Heisterberg L. The outcome of pregnancy after threatened abortion. Acta Obstet Gynecol Scand 1985;64:151–6.
7. Williams MA, Mittendorf R, Lieberman E, Monson RR. Adverse infant outcomes associated with first-trimester vaginal bleeding. Obstet Gynecol 1991;78:14–8.
8. Mulik V, Bethel J, Bhal K. A retrospective population-based study of primigravid women on the potential effect of threatened miscarriage on obstetric outcome. J Obstet Gynaecol 2004;24:249–53.
9. Johns J, Jauniaux E. Threatened miscarriage as a predictor of obstetric outcome. Obstet Gynecol 2006;107:845–50.
10. Weiss JL, Malone FD, Vidaver J, Ball RH, Nyberg DA, Comstock CH, et al. Threatened abortion: a risk factor for poor pregnancy outcome, a population-based screening study. Am J Obstet Gynecol 2004;190:745–50.
11. Wijesiriwardana A, Bhattacharya S, Shetty A, Smith N, Bhattacharya S. Obstetric outcome in women with threatened miscarriage in the first trimester. Obstet Gynecol 2006;107:557–62.
12. Hossain R, Harris T, Lohsoonthorn V, Williams MA. Risk of preterm delivery in relation to vaginal bleeding in early pregnancy. Eur J Obstet Gynecol Reprod Biol 2007;135:158–63.
13. Johns J, Hyett J, Jauniaux E. Obstetric outcome after threatened miscarriage with and without a hematoma on ultrasound. Obstet Gynecol 2003;102:483–7.
14. Lockwood CJ, Krikun G, Rahman M, Caze R, Buchwalder L, Schatz F. The role of decidualization in regulating endometrial hemostasis during the menstrual cycle, gestation, and in pathological states. Semin Thromb Hemost 2007;33:111–7.
15. Redman CW, Sargent IL. Pre-eclampsia, the placenta and the maternal systemic inflammatory response: a review. Placenta 2003;24:S21–7.
16. Norwitz ER. Defective implantation and placentation: laying the blueprint for pregnancy complications. Reprod Biomed Online 2006;13:591–9.
17. Andersen TF, Madsen M, Jørgensen J, Mellemkjoer L, Olsen JH. The Danish National Hospital Register: a valuable source of data for modern health sciences. Dan Med Bull 1999;46:263–8.
18. Marsál K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr 1996;85:843–8.
19. Klemmensen AK, Olsen SF, Osterdal ML, Tabor A. Validity of preeclampsia-related diagnoses recorded in a national hospital registry and in a postpartum interview of the women. Am J Epidemiol 2007;166:117–24.
20. Yang J, Hartmann KE, Savitz DA, Herring AH, Dole N, Olshan AF, et al. Vaginal bleeding during pregnancy and preterm birth. Am J Epidemiol 2004;160:118–25.
21. Ananth CV, Oyelese Y, Prasad V, Getahun D, Smulian JC. Evidence of placental abruption as a chronic process: associations with vaginal bleeding early in pregnancy and placental lesions. Eur J Obstet Gynecol Reprod Biol 2006;128:15–21.
22. Eskild A, Vatten LJ. Abnormal bleeding associated with preeclampsia: a population study of 315,085 pregnancies. Acta Obstet Gynecol Scand 2009;88:154–8.
23. Hustin J, Jauniaux E, Schaaps JP. Histological study of the materno-embryonic interface in spontaneous abortion. Placenta 1990;11:477–86.
24. Kim YM, Chaiworapongsa T, Gomez R, Bujold E, Yoon BH, Rotmensch S, et al. Failure of physiologic transformation of the spiral arteries in the placental bed in preterm premature rupture of membranes. Am J Obstet Gynecol 2002;187:1137–42.
25. Kim YM, Bujold E, Chaiworapongsa T, Gomez R, Yoon BH, Thaler HT, et al. Failure of physiologic transformation of the spiral arteries in patients with preterm labor and intact membranes. Am J Obstet Gynecol 2003;189:1063–9.
26. Lyall F. Priming and remodelling of human placental bed spiral arteries during pregnancy: a review. Placenta 2005;26:S31–6.
27. Redman CW, Sargent IL. Latest advances in understanding preeclampsia. Science 2005;308:1592–4.
28. Lockwood CJ. Testing for risk of preterm delivery. Clin Lab Med 2003;23:345–60.
29. Romero R, Espinoza J, Kusanovic JP, Gotsch F, Hassan S, Erez O, et al. The preterm parturition syndrome. BJOG 2006;113:17–42.
30. French JI, McGregor JA, Draper D, Parker R, McFee J. Gestational bleeding, bacterial vaginosis, and common reproductive tract infections: risk for preterm birth and benefit of treatment. Obstet Gynecol 1999;93:715–24.
31. Lykke JA, Paidas MJ, Langhoff-Roos J. Recurring complications in second pregnancy. Obstet Gynecol 2009; 113:1217–24.
Figure. No caption available.