1 Epidemiological Research Unit on Perinatal and Womens Health, INSERM U149-IFR69, 16 avenue Paul Vaillant-Couturier, 94807 Villejuif Cedex and 2 Université René Descartes (Paris V), Maternité Port-Royal, 123 bd de Port Royal 75679, Paris cedex 14
3 To whom correspondence should be addressed. e-mail: ancel{at}vjf.inserm.fr
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: casecontrol study/European survey/history of induced abortions/pregnancy complications/preterm delivery
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In this study, we aimed to estimate the risk of preterm birth associated with a history of induced abortions in the first trimester of pregnancy from data collected in a large casecontrol survey in Europe, the EUROPOP study. This survey included detailed information on past obstetric history, making it possible to analyse the relationship between a history of induced abortions and preterm birth in various European countries. Our hypothesis was that the risk of subsequent preterm birth was likely to be higher in Eastern European countries than in other countries. With the aim of increasing our understanding of the underlying mechanisms, we studied the effect of previous abortions on various preterm birth groups defined as a function of the complications of pregnancy leading to delivery. As induced abortion may affect the outcome of subsequent pregnancies via mechanical or infectious factors, we hypothesized that previous induced abortions would be associated with spontaneous deliveries and non-hypertensive pregnancy complications. Finally, we compared association of previous induced abortion with subsequent preterm birth for very and moderate preterm births. We expected the risk to be higher for birth at younger gestational ages.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Data from Ireland were excluded because induced abortion was not legal in this country. Data from Spain and Poland were also excluded because very few previous induced abortions were declared. Finally, we excluded data from France and The Netherlands because no distinction was made between spontaneous and induced first trimester abortions in these countries. The analysis was carried out in the following ten countries, in which information about previous first trimester induced abortions was collected correctly: Germany, Finland, Scotland, Sweden, Italy, Czech Republic, Slovenia, Romania, Russia and Hungary. In these countries, data on previous induced abortions were missing for only 2.5% of controls (n = 120) and 4.1% of cases (n = 120). The analysis included 2938 cases of preterm birth and 4781 controls at term.
Data collection
Information about previous induced abortion, social class, marital status, smoking habits, maternal age, and obstetric history was collected by interviewing women during their stay in the maternity unit. Women were assigned to one of the following three categories on the basis of self-reported history of first trimester induced abortion: no previous induced abortion, one previous induced abortion, or two or more previous induced abortions. The social class of the household was that corresponding to the occupation of the woman or the childs father (whichever was higher), and was scored using the International Labour Office classification (ILO) (ISCO-88, 1991).
We defined various subgroups of preterm births on the basis of gestational age, pregnancy complications and medical delivery information collected from medical records. Delivery onset was defined as spontaneous or indicated. Spontaneous delivery included cases with spontaneous onset of labour or preterm premature rupture of membranes. Indicated deliveries included cases of induced onset of labour or elective Caesarean section, but excluded preterm premature rupture of membranes. We also classified preterm births according to the pregnancy complications leading to delivery. Women were assigned to five mutually exclusive groups: (1) preterm premature rupture of membranes, defined as spontaneous rupture of the fetal membranes before the onset of labour, without ante-partum haemorrhage and hypertension; (2) idiopathic preterm labour, defined as spontaneous onset of labour before membrane rupture, without ante-partum haemorrhage and hypertension; (3) maternal hypertension, either proteinuric or non-proteinuric; (4) ante-partum haemorrhage, defined as blood loss before delivery, including placenta praevia, abruptio placenta, and other maternal haemorrhages; (5) women suffering from complications resulting in preterm delivery not covered by the previous definitions. Finally, moderate preterm birthsbirths at 3336 completed weeks of amenorrhoea (gestational age)were distinguished from very preterm births, at a gestational age of 2232 weeks.
Statistical analysis
We defined three groups of countries according to the frequency of induced abortion, based on national statistics from 1995 (World Health Organization, 1995). The first group included countries with a high abortion rate: Romania and Russia. The second group included countries with intermediate rates: Hungary, Slovenia and the Czech Republic. The third group included the countries with the lowest rates: Sweden, Italy, Scotland, Finland and Germany. The characteristics of the women with and without a history of previous induced abortions were compared in the control groups of the three groups of countries, using the
2-test. Multiple logistic regression was used to estimate the odds ratios (OR) of preterm birth, with 95% confidence intervals (95% CI), associated with previous induced abortions, after adjustment for confounding factors. The following factors were included in multivariate analyses because they are known risk factors for preterm delivery (Berkowitz and Papiernik, 1993
): maternal age, marital status, social class of the household, smoking during pregnancy and parity. Because adjustment slightly modifies odds ratios, only adjusted associations are shown in the Results section. Preterm births are uncommon events, so odds ratios are good estimates of relative risks. The adjusted odds ratios were compared between groups of countries using the following statistic: I = w1(lnOR1 lnaOR)2 + w2(lnOR2 lnaOR)2 + w3(lnOR3 lnaOR)2, where lnORi is estimated from the independent logistic regressions for group i (i = 1, 2, 3) of countries, wi is the inverse variance (1/variance), and lnaOR (adjusted parameter) is the mean of the parameters estimated for the three groups (Paul and Donner, 1989
). Under the null hypothesis of no difference in odds ratios, this statistic will follow a
2 distribution.
In the second step of the analysis, data were grouped because no significant variation between groups of countries was observed for the relationships between previous induced abortion and preterm births. We first investigated the relationships between previous induced abortions and very and moderate preterm births. We then estimated the odds ratios for spontaneous and indicated preterm delivery associated with previous induced abortions. Finally, we investigated the relationships between previous induced abortions and the possible causes of preterm delivery (preterm premature rupture of membranes, idiopathic preterm labour, maternal hypertension, ante-partum haemorrhage, and other). In these analyses, odds ratios were estimated by polytomous logistic regression (Hosmer and Lemeshow, 1989), using all births at term as controls, and adjusting for confounding factors. The analyses were performed with SAS software (SAS Institute, 1999
).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We cannot exclude the possibility of underreporting of previous induced abortion. In previous studies, underreporting varies from 16 to 65% (Anderson et al., 1994; Fu et al., 1998
). The validity of survey responses can be checked by comparing with external sources of information such as patient records (Anderson et al., 1994
) or national counts (Fu et al., 1998
), but the proportion of abortions reported probably depends on the characteristics of the women (Fu et al., 1998
), the mode of data collection, the wording of the questions used (Houzard et al., 2000
), and the broader context of public opinion on this topic. Underreporting may have been more frequent in Northern and Western Europe than in Eastern Europe, because abortion was legalized a long time ago in Eastern Europe and carries out a lower level of social stigma than in Western Europe. This could have led to a larger underestimation of the association between previous induced abortion and preterm birth in group 3 than in groups 1 and 2. However, differences in the reporting by women of previous abortions in the various countries (Table I) were similar to differences in the actual frequency of induced abortion based on national statistics (World Health Organization, 1995
).
Another possible problem is that underreporting may differ between cases and controls. Although selective underreporting, by cases or controls, may have occurred, no data were available to check whether this bias resulted in the under- or overestimation of odds ratios. In a Danish register-based study with <10% underreporting of previous induced abortion, the odds ratios for preterm birth were 1.9 for one previous induced abortion and >2.0 for at least two abortions (Zhou et al., 1999). In an Australian population-based study, relative risks of preterm delivery were from 1.5 to >4.0, depending on the number of previous induced abortions (Lumley, 1998
). This suggests possible slight underestimation of the associations in our study.
Our results are consistent with other studies reporting a significant increase in the risk of preterm birth with the number of previous induced abortions (de Haas et al., 1991; Lang et al., 1996
; Lumley, 1998
; Martius et al., 1998
; Zhou et al., 1999
; Henriet and Kaminski, 2001
). To our knowledge, no other study has compared pregnancy outcome after induced abortion in various countries, at the same time, using the same protocol in each country. As induced abortion has long been legal and most modern methods of contraception difficult to obtain reliably in Eastern Europe (Anderson et al., 1994
), the rates of induced abortions were high in Eastern Europe, with a high proportion of women having undergone several abortions. These women may have been at higher risk of adverse pregnancy outcomes than those from other European countries. Differences may also occur between countries because the frequency of late complications may vary depending on the procedures used to perform abortion. During the 1970s and 1980s, dilatation and curettage was the main abortion procedure in Eastern European countries whereas vacuum aspiration was the common procedure in Northern and Western Europe and in the USA (Hogue et al., 1982
). However, we found no differences in odds ratios between groups of countries, taking into account the number of previous abortions, and after appropriate adjustment.
Previous studies have identified a significant relationship between history of induced abortion and spontaneous preterm birth (de Haas et al., 1991; Kristensen et al., 1995
; Lang et al., 1996
). In our study, a history of prior induced abortion seemed to be strongly associated with spontaneous preterm delivery. Like Henriet and Kaminski (2001
), we found a significant association between a history of at least two previous abortions and indicated preterm birth. Spontaneous and indicated preterm births may not be aetiologically different entities. For example, in our study, delivery was induced in 60% of women with placenta praevia and in 30% of women with other types of ante-partum haemorrhage. This suggests that the classification of preterm birth according to the type of onset of labour may be insufficient to identify underlying pregnancy complications responsible for preterm delivery.
It has been suggested that infectious diseases following abortion may account for the relationship with subsequent preterm delivery (Stürchler et al., 1986). Women with a history of induced abortion have an increased risk of intra-amniotic infection (Krohn et al., 1998
), intra-partum infection (Stürchler et al., 1986
; Mühlemann et al., 1996
) and of infection of their newborn children (Germain et al., 1995
). Intra-amniotic infection is a known risk factor for idiopathic preterm labour and preterm premature rupture of membranes (Gomez et al., 1997
; Romero et al., 2001
). To our knowledge, no data are available concerning the risk of preterm birth according to the pregnancy complications that led to delivery. Thus, identifying subgroups of preterm births on the basis of pregnancy complications may help to increase our understanding of the mechanisms underlying the effect of previous induced abortion on subsequent pregnancy outcomes. Our results, showing a significant association between previous induced abortion and both idiopathic preterm labour and preterm premature rupture of membranes, are consistent with infectious processes. Although the association between prior abortion and intra-amniotic infection may result from latent upper genital tract infection, preceding or beginning at the time of abortion (Stürchler et al., 1986
), other mechanisms should be considered. Cervical trauma from mechanical dilatation (Hakim-Elahi et al., 1990
), by increasing the risk of cervical incompetence (Molin, 1993
) and facilitating upper genital tract infection, may account for intra-amniotic infection following abortion. Consistent with previous studies (Ananth et al., 1997
; Seidman et al., 1998
), we also found a strong relationship between previous abortion and preterm birth after placenta praevia and other types of maternal haemorrhage. It has been suggested that surgical procedures, by damaging the endometrium, may increase faulty placentation in subsequent pregnancies.
The stronger association with very preterm birth than with moderate preterm birth is consistent with results from Australia (Lumley, 1998). It suggests that the underlying mechanisms involved in the adverse effect of previous induced abortion (e.g. infection and haemorrhage) are similar to the major causes of very preterm delivery (preterm premature rupture of membranes, idiopathic preterm labour, infection, ante-partum haemorrhage, and cervico-uterine abnormalities) (Hillier et al., 1988
; Hagan et al., 1996
).
Like any surgical procedure, dilatationcurettage and dilatationsuction procedures present a number of risks of early complications. Such complications, including pelvic infection, fever, tissue retention, bleeding and cervical trauma, are recorded for 5% of women (Zhou et al., 2002
). The procedures used may also have an effect on subsequent pregnancies. The risk of preterm birth has been shown to be related to (i) the technique used for induced abortion, with higher figures for dilatation and curettage than for aspiration (Zhou et al., 1999
) and (ii) the timing of the abortion, with a higher risk for abortions after 8 weeks (Zhou et al., 1999
). In Europe, between 1980 and 1995, >90% of abortions were carried out by surgical procedures (Abortion Statistics, 1995
; Le Corre and Thomson, 2000
; Zhou et al., 2002
), suggesting that most of the abortions reported by the women included in our survey were carried out by such procedures.
Further studies are required because abortion procedures and their potential side-effects have changed. Drug-based abortion has been available since the early 1990s. The antiprogestogen mifepristone, in combination with a prostaglandin analogue, provides a suitable non-surgical method for early pregnancy termination (Ashok et al., 2002). The frequency of use of this method increased from 14% (1990) to 20% (1997) in France (Le Corre and Thomson, 2000
), and from 9% (1995) to 20% (2001) in England and Wales (Abortion Statistics, 1995
, 2001). Ashok et al. (2002
) reported that women given such drug treatments had complete abortions in 98% of cases, with only 2% requiring surgical intervention. Thus, changes in the management of induced abortion may affect the relationship between previous induced abortion and subsequent pregnancy outcomes.
In conclusion, our results provide evidence that history of induced abortion is a risk factor for preterm delivery. Although the relationship between induced abortion and preterm birth was similar in all countries studied, public health consequences, depending on the frequency of prior induced abortion, varied considerably between countries. With a mean odds ratio for preterm birth of 1.6 associated with at least two previous abortions, assuming that this relationship is causal, and using the proportion of women who had undergone induced abortion in each of the three control groups, the attributable risk fractions are 4 and 2% in groups 2 and 3, and 12% in group 1. We were also able to show differences according to the cause of delivery, with an increased risk of preterm birth following preterm labour, preterm premature rupture of membranes and ante-partum haemorrhage. Although these complications of pregnancy result from cervico-uterine abnormalities, the precise underlying mechanisms, whether post-traumatic or post-infectious, remain unknown (Zhou and Olsen, 2003). As the techniques used for induced abortion have changed during the last decade, it would be useful to examine the long-term consequences of induced abortion with respect to new medical procedures.
![]() |
Acknowledgements |
---|
![]() |
Appendix: composition of the EUROPOP group |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Members of the national staffs
Czech Republic: Kudela M, MD, Vetr M, MD, in Olomouc; Finland: Heikkilä A, MD, Erkkola R, MD, Forström J, MD, in Turku; France: Papiernik E, MD, Lucidarme P, midwife, in Paris. Tafforeau J, MD, in Brussels; Germany: Künzel W, MD, Herrero-Garcia J, MD, in Giessen; Dudenhausen J, MD, Henrich W, MD, in Berlin; Greece: Antsaklis A, MD, Haritatos G, MD, in Athens; Hungary: Kovacs L, MD, Nyari T, MD, Bartfai G, MD, in Szeged; Ireland: OHerlihy C, MD, Murphy J, MD, Stewart H, in Dublin; Italy: Di Renzo GC, MD, Bruschettini PL, MD, Moscioni P, MD, in Perugia; Cosmi E, MD, Spinelli A, MSc, Donati S, MD, in Rome; Poland: Breborowicz GH, MD, Anholcer A, MD, in Poznan; Romania: Stamatian F, MD, in Cluj; Russia: Mikhailov AV, MD, in St Petersburg; Slovenia: Pajntar M, MD, Pirc M, MD, Verdenik I, MD, in Ljubljana; Spain: Escribà-Aguir V, MD, in Valencia; Carrera JM, MD, in Barcelona; Sweden: Marsal K, MD, Stale H, MD, in Malmö; The Netherlands: Buitendijk S, PhD, van der Pal K, MsC, in Leiden; van Geijn H, MD, in Amsterdam; Turkey: Gökmen O, MD, Güler C, MD, Caglar T, MD, in Ankara; UK: Owen P, MD, in Dundee.
Members of the epidemiological analysis staff
Ancel Pierre-Yves, MD, Bréart Gérard, MD, Lelong Nathalie, MsC, Saurel-Cubizolles Marie-Josèphe, PhD, France.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Abortion Statistics (2001) Legal abortions carried out under the 1967 Abortion Act in England and Wales. Office for National Statistics Series AB no. 28, London.
Anderson BA, Katus K, Puur A and Silver BD (1994) The validity of survey responses on abortion: evidence from Estonia. Demography 31,115132.[ISI][Medline]
Ananth CV, Smulian JC, Demissie K, Vintzileos AM and Knuppel RA (2001) Placental abruption among singleton and twin births in the United States: risk factor profiles. Am J Epidemiol 153,771778.
Ananth CV, Smulian JC and Vintzileos AM (1997) The association of placenta previa with history of cesarean delivery and abortion: a metaanalysis. Am J Obstet Gynecol 177,10711078.[ISI][Medline]
Ashok PW, Templeton A, Wagaarachchi PT and Flett JM (2002) Factors affecting the outcome of early medical abortion: a review of 4132 consecutive cases. Br J Obstet Gynecol 109,12811289.
Atrash HK and Hogue CJ (1990) The effect of pregnancy termination on future reproduction. Baillières Clin Obstet Gynecol 4,391405.[ISI][Medline]
Berkowitz GS and Papiernik E (1993) Epidemiology of preterm birth. Epidemiol Rev 15,414443.[ISI][Medline]
de Haas I, Harlow BL, Cramer DW and Frigoletto FD (1991) Spontaneous preterm birth: a casecontrol study. Am J Obstet Gynecol 165,12901296.[ISI][Medline]
Demissie K, Breckenridge MB, Joseph L. and Rhoads GG (1999) Placenta previa: preponderance of male sex at birth. Am J Epidemiol 149,824830.[Abstract]
French JI and McGregor JA (1996) The pathobiology of premature rupture of membranes. Semin Perinatol 20,344368.[ISI][Medline]
Fu H, Darroch JE, Henshaw SK and Kolb E (1998) Measuring the extent of abortion underreporting in the 1995 National Survey of Family Growth. Fam Plann Perspect 30,128133.[ISI][Medline]
Germain M, Krohn MA and Daling JR (1995) Reproductive history and the risk of neonatal sepsis. Paediatr Perinat Epidemiol 9,4858.[ISI][Medline]
Gomez R, Romero R, Edwin SS and David C (1997) Pathogenesis of preterm labor and preterm premature rupture of membranes associated with intraamniotic infection. Infect Dis Clin North Am 11,135176.[ISI][Medline]
Hagan R, Benninger H, Chiffings D, Evans S and French N (1996) Very preterm birth: a regional study. Part 1: Maternal and obstetrical factors. Br J Obstet Gynecol 103,230238.[ISI][Medline]
Hakim-Elahi E, Tovell HM and Burnhill MS (1990) Complications of first-trimester abortion: a report of 170 000 cases. Obstet Gynecol 76,129135.[Abstract]
Henriet L and Kaminski M (2001) Impact of induced abortion on subsequent pregnancy outcome: the 1995 French national perinatal survey. Br J Obstet Gyneacol 108,10361042.[CrossRef][ISI]
Hillier SL, Martius J, Krohn M, Kiviat N, Holmes KK and Eschenbach DA (1988) A casecontrol study of chorioamnionitis infection and histologic chorioamnionitis in prematurity. N Engl J Med 319,972978.[Abstract]
Hogue CJ, Cates W and Tietze C (1982) The effects of induced abortion on subsequent reproduction. Epidemiol Rev 4,6694.[ISI][Medline]
Hosmer DW and Lemeshow S (1989) Applied Logistic Regression. Wiley, New York, NY.
Houzard S, Bajos N, Warszwawski J, de Guibert-Lantoine C, Kaminski M, Leridon H, Lelong N, Ducot B, Hassoun D and Ferrand M (2000) Analysis of the underestimation of induced abortions in a survey of the general population in France. Eur J Contracept Reprod Health Care 5,5260.[Medline]
ISCO-88 (1991) International Standard Classification of Occupations. International Labour Office, Geneva, Switzerland.
Kristensen J, Langhoff-Roos J and Kristensen FB (1995) Idiopathic preterm deliveries in Denmark. Obstet Gynecol 85,549552.
Krohn, MA, Germain M, Mühlemann K and Hictok D (1998) Prior pregnancy outcome and the risk of intraamniotic infection in the following pregnancy. Am J Obstet Gynecol 178,381385.[ISI][Medline]
Lang JM, Lieberman E and Cohen A (1996) A comparison of risk factors for preterm labor and term small-for-gestational-age birth. Epidemiology 7,369376.[ISI][Medline]
Le Corre M and Thomson E (2000) Les IVG en 1998. Etudes et Résultats, 69, 14.
Lumley J (1998) The association between prior spontaneous abortion, prior induced abortion and preterm birth in first singleton births. Prenat Neonat Med 3,2124.[ISI]
Martius JA, Steck T, Oelher MK and Wulf KH (1998) Risk factors associated with preterm (<37 weeks+0 weeks) and early preterm birth (<32+0 weeks): univariate and multivariate analysis of 106,345 singletons births from the 1994 statewide perinatal survey of Bavaria. Eur J Obstet Gynecol Reprod Biol 80,183189.[CrossRef][ISI][Medline]
Molin A (1993) Risk of damage to the cervix by dilatation for first-trimester-induced abortion by suction aspiration. Gynecol Obstet Invest 35,152154.[ISI][Medline]
Mühlemann K, Germain M and Krohn M (1996) Does an abortion increase the risk of intrapartum infection in the following pregnancy? Epidemiology 7,194198.[ISI][Medline]
Paul SR and Donner A (1989) A comparison of tests of homogeneity of odds ratios in k 2 x 2 tables. Stat Med 8,14551468.[ISI][Medline]
Pickering RM and Deeks JJ (1991) Risks of delivery during the 20th to 36th week of gestation. Int J Epidemiol 20,456466.[Abstract]
Romero R, Gomez R, Chaiworapongsa T, Conoscenti G, Kim JC and Kim YM (2001) The role of infection in preterm labour and delivery. Paediatr Perinat Epidemiol 15(Suppl 2),4156.[CrossRef][ISI][Medline]
SAS Institute Inc. (1999) SAS/STAT Users Guide, Version 8, SAS Institute, Inc., Cary, NC.
Saurel-Cubizolles MJ, Di Renzo JC, Breart G and the EUROPOP Group (1997) Womens work and preterm birth: epidemiological knowledge and description of a European project. Prenat Neonat Med 2,161180.
Seidman DS, Ever-Hadani P, Slater PE, Harlap S, Stevenson DK and Gale R (1998) Child-bearing after induced abortion: reassessment of risk. J Epidemiol Commun Health 42,294298.
Stürchler D, Menegoz F and Daling J (1986) Reproductive history and intrapartum fever. Gynecol Obstet Invest 2,182186.
Tulchinsky TH and Varavikova EA (1996) Addressing the epidemiologic transition in the former Soviet Union: strategies for health system and public health reform in Russia. Am J Publ Health 86,313320.[Abstract]
World Health Organization (1995) Statistical Information from the Regional Office for Europe Online Health for all Statistical Data Base (HFA-DB).
Zhang J, Zeisler J, Hatch MC and Berkowitz G (1997) Epidemiology of pregnancy-induced hypertension. Am J Epidemiol 19,218232.
Zhou W and Olsen J (2003) Are complications after an induced abortion associated with reproductive failures in a subsequent pregnancy? Acta Obstet Gynecol Scand 82,177181.[ISI][Medline]
Zhou W, Sorensen HT and Olsen J (1999) Induced abortion and subsequent pregnancy duration. Obstet Gynecol 94,948953.
Zhou W, Nielsen GL, Moller M and Olsen J (2002) Short-term complications after surgically induced abortions: a register-based study of 56117 abortions. Acta Obstet Gynecol Scand 81,331336.[CrossRef][ISI][Medline]
Submitted on May 21, 2003; resubmitted on September 26, 2003; accepted on October 29, 2003.
|