Aliment Pharmacol Ther 2011; 34: 724–734
Background Inflammatory bowel disease (IBD) frequently affects women during their reproductive years. Pregnancy outcome in women with IBD is well described, particularly in retrospective studies.
Aim To evaluate the pregnancy outcome in patients with IBD in a prospective European multicentre case-control study.
Methods Inflammatory bowel disease pregnant women from 12 European countries were enrolled between January 2003 and December 2006 and matched (1:1) to non-IBD pregnant controls by age at conception and number of previous pregnancies. Data on pregnancy and newborn outcome, disease activity and therapy were prospectively collected every third month using a standard questionnaire. Logistic regression analysis with odds ratio was used for statistical analyses. P value < 0.05 was considered significant.
Results A total of 332 pregnant women with IBD were included: 145 with Crohn’s disease (CD) and 187 with ulcerative colitis (UC). Median age (range) at conception was 31 years (15–40) in CD and 31 (19–42) in UC patients. No statistically significant differences in frequency of abortions, preterm deliveries, caesarean sections, congenital abnormalities and birth weight were observed comparing CD and UC women with their non-IBD controls. In CD, older age was associated with congenital abnormalities and preterm delivery; smoking increased the risk of preterm delivery. For UC, older age and active disease were associated with low birth weight; while older age and combination therapy were risk factors for preterm delivery.
Conclusion In this prospective case-control study, women with either Crohn’s disease or ulcerative colitis have a similar pregnancy outcome when compared with a population of non-inflammatory bowel disease pregnant women.
Crohn’s disease (CD) and ulcerative colitis (UC) are chronic inflammatory diseases with an increasing incidence and prevalence, often affecting young women in their reproductive years.1–4 It has been estimated that after a diagnosis of inflammatory bowel disease (IBD), approximately one quarter of female patients will conceive.5 Thus, these patients and their partners have many concerns about the impact of IBD on pregnancy.
Multiple studies have been published investigating the impact of IBD on pregnancy outcome and foetal safety of IBD-related therapy, with divergent results.6–22 While some, mainly older retrospective studies, suggested no significant impact of IBD on pregnancy outcome,7, 23, 24 others found IBD to be associated with a poor pregnancy outcome.12, 15, 16, 25
Two large population–based studies11, 25 reported an increased risk of preterm delivery and low birth weight (LBW) in women with either CD or UC. The influence of disease activity was not, however, assessed in these two studies. Nevertheless, a recent large community-based study reported disease activity to be related to an increased risk of preterm birth rate in CD female patients.19
The frequency of congenital abnormalities (CAs) does not appear to differ from healthy population both in CD (OR 0.8; 95% CI: 0.2–3.8) and UC (OR 1.3; 95% CI: 0.9–1.8) as reported in two population-based studies by Norgard et al.,18, 19 whereas an increased risk of CAs in newborns of UC mothers was found in another population-based study (OR 3.8; 95% CI: 1.5–9.8)11 and also in a recent meta-analysis (OR 3.88; 95% CI: 1.47–3.82).5
Several studies have also investigated the impact of IBD medications on pregnancy outcome, showing no association with poor pregnancy outcomes.6, 14, 17, 20
So far, the current knowledge regarding pregnancy outcome in IBD is of limited quality, as the majority of studies are retrospective,9, 11, 15, 16 small, older of date,7, 21, 23, 24 with only a few being recent10, 14, 20, 26, 27 and not even one is prospective, as concluded by the recent European Crohn’s and Colitis Organization (ECCO) consensus on reproduction in IBD.28
As the natural history of pregnancy during IBD is so clinically important, a large multicentre, prospective case-control study seems to be an appropriate method to evaluate many of the parameters concerning IBD and pregnancy.
This is a European, multicentre, prospective case-control study on pregnancy and IBD in which pregnant IBD women were matched with pregnant non-IBD women and nonpregnant IBD women 1:1:1. This study is the first part of the project dealing with outcome of pregnancy in women with IBD. The study aim and protocol were first proposed and discussed at the meeting of the ECCO in Amsterdam in 2002. After approval by the ECCO Scientific Committee, 68 centres from 12 European countries participated in the study.
The Ethical Committee of the Coordinating Centre (Rho Hospital – Italy) approved the study protocol.
All consecutive pregnancies which occurred in women with IBD, and followed by the participating centres from January 2003 to December 2006, were included. At the time of enrolment (conception/first trimester until 12th gestational week), all IBD pregnant women were intended to be matched (1:1) with non-IBD pregnant controls by age at conception ( ± 5 years) and number of previous pregnancies at the Obstetric and Gynaecology Department of each participating centre.
Each participating centre received electronic case report forms (CRFs, for index cases and controls) to record the requested data. The data were prospectively collected by trained physicians at the participating centres at entry to the study and then every 3 months until the end of pregnancy by regular personal or telephone interviews as well as a by review of patient’s medical records. All the completed forms were sent electronically to the central database storage record, to allow data extraction and statistical evaluation.
The following data were collected: (i) maternal variables, (ii) treatment for IBD, (iii) disease activity, (iv) outcome of pregnancy and (v) mode of delivery.
Maternal variables included age at conception, number of previous pregnancies, smoking status (current smoker vs. nonsmoker), alcohol use (current alcohol use vs. no alcohol use) and details on underlying IBD. Medical and surgical therapies, as well as disease activity, were recorded at conception and then every third month. Disease activity was assessed by Harvey & Bradshaw’s Activity Index (HBI)29 for CD patients and Simple Clinical Colitis Activity Index (SCCAI)30 for UC patients and classified into two categories for each trimester as remission (HBI ≤ 5; SCCAI ≤ 3) or active disease.
The outcomes of pregnancy recorded were: live births, spontaneous or therapeutic abortion, extra-uterine pregnancy, infant death in utero, preterm delivery (<37 weeks of gestation) and at term delivery. Furthermore, mode of delivery (vaginal or caesarean), occurrence and type of CAs and birth weight (LBW was defined as <2500 g) were registered.
Similarly, outcomes of pregnancy in non-IBD healthy females were followed prospectively and data on demographics, smoking, alcohol use and pregnancy outcome were recorded on CRFs in the same manner as in the IBD population.
Matched logistic regression was used to compare cases and controls for the parameters of interest. The odds ratio and 95% CI for the parameter of interest were adjusted for the following covariates measured for both cases and controls: age at conception, smoking, alcohol use. The impact of all covariates, including disease specific parameters only measured for the cases, was analysed by a standard logistic regression model for the cases only. Due to sparse data for some of the prognostic factors, the association between the parameters of interest and all prognostic factors were also investigated by Fisher’s Exact Test (for the cases only). All tests were performed with a significance level of 5%.
A total of 520 IBD pregnant women: 244 with CD, 264 with UC and 12 with indeterminate colitis (IC) were enrolled. Nevertheless, only 373 of them were matched to non-IBD pregnant controls and considered to be eligible for the study. Further 32 IBD women had to be later excluded due to missing data on pregnancy outcome in their controls. Women with IC were not included into data analysis due to the small numbers. Hence, the final population comprised 332 IBD pregnant women, 145 with CD and 187 with UC, who were matched (1:1) to non-IBD pregnant controls (Figure 1). Two hundred and fifty-eight (78%) pregnant women were enrolled from Italian centres and 74 (22%) from the other European centres. No significant difference in pregnancy outcome between Italian and non-Italian patients was observed (data not shown). The pregnancy outcome of the excluded population (98 women with CD, 78 with UC and 12 with IC) is outlined in Table S1 (on-line).
The demographic and disease characteristics of included women are summarised in Table 1.
|Pregnant||CD (n = 145)||Controls to CD (n = 145)||UC (n = 187)||Controls to UC (n = 187)|
|Age*||31 (15–40)||31 (16–45)||31 (19–42)||32 (19–42)|
|Smoking (%)||21 (14.5)||15 (10.3)||15 (8. 0)||26 (13.9)|
|Alcohol (%)||3 (2.1)||9 (6.2)||8 (4.3)||13 (7.0)|
|Disease duration (months)*||84 (2–311)||n.a.||66 (1–270)||n.a.|
|Disease localisation/extension (%)|
|Small bowel||46 (32.0)||n.a.||n.a.||n.a.|
|Small bowel + colon||69 (48.0)||n.a.||n.a.||n.a.|
|Left sided colitis||n.a.||n.a.||55 (30.0)||n.a.|
|Previous intestinal surgery (%)||47 (32.4)||n.a.||6 (3.2)||n.a.|
|Previous perianal/rectovaginal fistula (%)||18 (12.4)||n.a.||n.a.||n.a.|
|Remission at conception/1st trimester (%)||126 (86.9)||n.a.||148 (79)||n.a.|
|Onset during pregnancy (%)||1 (0.7)||n.a.||2 (1.1)||n.a.|
|Any therapy at conception/1st trimester (%)||117 (80.7)||n.a.||165 (88.2)||n.a.|
|- Mesalazine||96 (66.2)||n.a.||156 (83.4)||n.a|
|- Corticosteroids||32 (22.1)||n.a.||74 (39.6)||n.a.|
|- Azathioprine/6MP||32 (22.1)||n.a.||19 (10.2)||n.a.|
|- Infliximab||8 (5.5)||n.a.||0||n.a.|
|- Cyclosporine||0||n.a.||1 (0.5)||n.a.|
The majority of pregnant women with CD, 126 (86.9%), were in remission at conception/first trimester, while 19 (13.1%) had active disease. Of those in remission, 105 (85.8%) maintained remission during the entire pregnancy, whereas 38 (14.2%) women relapsed. As to women with active disease, 14 (73.7%) obtained remission later during the pregnancy, while five (26.3%) remained active until delivery.
At conception and/or during the first trimester, 28 (19.3%) patients were without any medical treatment, while 117 (80%) women were under treatment with mono- or combined therapy. Sixty eight (46.9%) women were on 5-aminosalicylic acid (5-ASA) monotherapy, 20 (13.8%) on immuno-monotherapy [azathioprine (AZA), 6-mercaptopurine (6-MP), cyclosporine, biologics or corticosteroids] and 28 (19.3%) on combination therapy (two or more preparations). The median daily dose of 5-ASA preparations was 2400 mg (range 800–4800) with 37 (46%) women taking a dosage of ≥ 3000 mg a day at conception, maintaining the same dose during the pregnancy in majority of them. The pregnancy outcome of these women was favourable with 35 (95%) live births, 33 at term and two preterm deliveries. Two (8%) pregnancies ended in spontaneous abortion. No CAs in newborns of women treated with high dose 5-ASA were observed. Four women required intervention during pregnancy.
One woman with colonic disease and history of perianal fistula underwent incision of a perianal abscess at gestational week 24. She had been in remission on azathioprine therapy since conception until the third trimester, when a perianal abscess occurred. She delivered an at term 3300 g baby boy. A second patient with ileo-colonic CD was in remission and received no medication until relapse at gestational week 14 and required drainage of an intra-abdominal ileo-colonic abscess at gestational week 19. This patient delivered a preterm living 1040 g baby boy at week 26 by caesarean section. The third woman, with ileo-colonic and perianal involvement, was in remission on topical corticosteroids and systemic and topical 5-ASA therapy. A relapse at gestational week 18 was complicated by intestinal perforation necessitating segmental ileo-colonic resection. This patient, however, went on to deliver a living at term 3350 g baby girl by caesarean section. The fourth woman with colonic CD was in remission on no medication at conception/first trimester and relapsed during the second trimester. Corticosteroids were initiated; nevertheless the disease remained active until the end of pregnancy. At gestational week 39, an intra-abdominal abscess was found and subsequently an ultrasonic drainage was performed. The patient delivered a 2800 g baby girl at term by caesarean section.
Demographic and clinical characteristics of all women included are summarised in Table 1.
The majority of pregnant women with UC, 148 (79%), were in remission at conception/first trimester, while 39 (21%) had active disease. Of those in remission, 109 (73.6%) maintained remission during the entire pregnancy and 39 (26.4%) relapsed. Regarding women with active disease, 26 (67%) went into remission later during the pregnancy, while 13 (33%) remained active until delivery.
At conception and during the first trimester, 22 (11.8%) were untreated, while 165 (88.2%) women were under treatment with mono- or combined therapy. Eighty eight (47.1%) women were on 5-ASA monotherapy, 14 (7.5%) on immuno-monotherapy (AZA, 6-MP, cyclosporine or corticosteroids) and 63 (33.7%) on combination therapy (two or more preparations). None of the UC patients received anti-TNF therapy. The median dose of 5-ASA preparations was 2400 mg a day (range 800–4800) with 37 (30%) women having a daily dosage of ≥ 3000 mg at conception, maintaining the same dose during the pregnancy in majority of them. Similar to CD, the pregnancy outcome of women with UC on high dose 5-ASA therapy was favourable. There were 35 (95%) live births, 31 at term and four preterm deliveries. One (3%) pregnancy ended in spontaneous abortion and one (3%) in therapeutic abortion. No CAs in newborns of mothers receiving 5-ASA dose ≥ 3000 mg a day was observed.
One woman, with extensive active UC since conception, required subtotal colectomy at gestational week 12 for steroid refractory UC. The patient delivered a healthy 2850 g baby girl at term by caesarean section.
There was no significant difference in live births, spontaneous and therapeutic abortions, infant death in utero, preterm deliveries and caesarean sections, comparing both CD and UC women with their non-IBD pregnant controls (Table 2). Likewise, no increased risk of CAs in infants of CD or UC women vs. healthy controls was seen (Table 2; Table S2 on-line).
|Outcome||CD n = 145||Controls to CD n = 145||OR (95% CI)||P||UC n = 187||Controls to UC n = 187||OR (95% CI)||P|
|Live births (%)||127 (88.2)||135 (93.1)||0.48 (0.19–1.24)||0.13||177 (94.7)||170 (91.4)||0.48 (0.19–1.24)||0.13|
|Spontaneous abortions (%)||9 (6.3)||8 (5.5)||1.30 (0.41–4.06)||0.66||9 (4.8)||15 (8.0)||0.62 (0.26–1.52)||0.30|
|Therapeutic abortions (%)||6 (4.2)||2 (1.4)||4.07 (0.45–36.85)||0.21||1 (0.5)||1 (0.5)||1.56 (0.09–28.12)||0.76|
|Infant death in utero (%)||1 (0.7)||0||–||0||0||–||–|
|Preterm delivery <37 weeks (%)||13 (9.1)||11 (7.6)||0.89 (0.38–2.08)||0.78||12 (6.5)||14 (7.5)||0.99 (0.41–2.39)||0.97|
|Caesarean section (%)||50 (40.3)||34 (26.2)||1.34 (0.74–2.43)||0.33||55 (31.3)||46 (27.7)||1.29 (0.74–2.24)||0.36|
|Congenital abnormalities (%)||5 (3.5)||1 (0.7)||5.0 (0.58–42.79)||0.14||0||3 (1.7)||–||–|
|Birth weight (g)||3212 (431)||3269 (395)||–||0.56||3190 (471)||3272 (448)||–||0.07|
|Mean ± SD|
Similarly, no significant difference in birth weight was observed between CD and UC women and their controls (Table 2).
Factors with impact on pregnancy outcome
In CD, multivariate analysis identified older age ( ≥ 35 years) to be associated with an increased risk of CAs and preterm delivery. Alcohol intake was found to decrease the probability of live birth, whereas smoking increased the risk of preterm delivery. Previous intestinal surgery was associated with a higher rate of caesarean sections. No impact of disease activity on pregnancy outcome was observed. Regarding IBD-related medication, women on 5-ASA monotherapy were less likely to have caesarean sections (27% vs. 50%, P = 0.01), while those on combination therapy were at increased risk of this mode of delivery (57% vs. 34%, P = 0.02) (Table 3).
|Factor||Live birth||Spontaneous abortion||Therapeutic abortion||Preterm delivery||Congenital abnormality||Caesarean section|
|Age ≥ 35 years||1.3 (0.3 –6.0)||1.64 (0.2–11.2)||1.6 (0.1–19.9)||9.8(1.7–55.6)||15.0(1.4–163.4)||1.1 (0.4–3.2)|
|Smoking status||0.6 (0.1–2.5)||1.0 (0.1–10.3)||1.7 (0.1–19.8)||5.6(1.1–28.9)||P = 1.00||1.0 (0.3–3.2)|
|Alcohol use||0.05 (0.003–0.6)||7.2 (0.5–113.5)||P = 0.12||P = 1.00||P = 1.00||P = 1.00|
|Previous surgery||1.4 (0.3–6.2)||1.3 (0.2–7.9)||P = 0.18||0.4 (0.04–3.7)||–||3.6(1.3–10.2)|
|Perianal disease*||2.0 (0.2–19.1)||P = 0.60||3.2 (0.2–57.9)||0.6 (0.05–7.2)||–||2.0 (0.6–7.5)|
|Disease activity**||1.0 (0.2–33.3)||0.34 (0.06–4.8)||3.9 (0.2–76.3)||0.3 (0.04–1.9)||3.2 (0.1–51.4)||0.9 (0.3–2.7)|
|any therapy||0.9 (0.2–4.7)||0.6 (0.04–8.1)||19.3 (0.5–760.8)||2.5 (0.02–409.6)||P = 0.053||P = 0.66|
|5-ASA monotherapy||2.4 (0.2–29.0)||5.0 (0.2–111.9)||P = 0.23||0.4 (0.003–55.3)||P = 0.37||P = 0.01|
|IS monotherapy||1.8 (0.1–30.7)||9.4 (0.2–393.3)||P = 1.00||0.7 (0.003–134.4)||P = 0.49||P = 0.80|
|combination therapy||2.1 (0.1–33.4)||11.0 (0.4–313.0)||P = 0.64||2.7 (0.02–417)||P = 0.58||P = 0.02|
In UC, women with older age were found to be at increased risk of preterm delivery, whereas smoking women were found to have a higher rate of caesarean sections. Older age and disease activity at any time during pregnancy were associated with a lower birth weight (mean 3146 g vs. 3343 g, P = 0.02 and 3110 g vs.3268 g P = 0.04, respectively). Regarding IBD-related medication, women on 5-ASA monotherapy were less likely to get preterm delivery (1% vs. 10%, P = 0.01) whereas patients on combination treatment were at increased risk of preterm delivery (13% vs. 1%, P = 0.03) (Table 4).
|Factor||Live birth||Spontaneous abortion||Therapeutic abortion||Preterm delivery||Caesarean section||Birth weight|
|Age ≥ 35 years||3.0 (0.3–25.9)||0.4 (0.05–3.5)||P = 1.00||3.9(1.1–14.2)||0.9 (0.4–1.9)||P = 0.02|
|Smoking status||0.5 (0.05–5.4)||2.2 (0.2–23.1)||P = 1.00||2.0 (0.3–13.0)||9.9(2.4–40.0)||P = 0.27|
|Alcohol use||P = 1.00||P = 1.00||P = 1.00||P = 1.00||0.5 (0.06–4.89)||P = 0.74|
|Previous surgery||P = 1.00||P = 1.00||P = 1.00||P = 1.00||1.8 (0.2–13.2)||P = 0.52|
|Disease activity*||5.6 (0.6–52.2)||0.2 (0.02–1.9)||P = 1.00||0.5 (0.1–1.8)||0.7 (0.3–1.7)||P = 0.04|
|Therapy for UC*:|
|any therapy||P = 0.60||P = 0.56||P = 1.00||P = 0.60||P = 1.00||P = 0.11|
|5-ASA monotherapy||P = 0.75||P = 1.00||P = 0.44||P = 0.01||P = 0.051||P = 0.83|
|IS monotherapy||P = 1.00||P = 1.00||P = 1.00||P = 1.00||P = 0.40||P = 0.27|
|combination therapy||P = 1.00||P = 1.00||P = 1.00||P = 0.004||P = 0.41||P = 0.94|
To our knowledge this is the first prospective case-control study on pregnancy outcome in women with IBD. In this study, IBD pregnant women were matched (1:1) at conception/during first trimester with non-IBD pregnant healthy controls and follow-up was performed every 3 months during the gestational period. Women with IBD, overall considered, were shown to have similar pregnancy outcomes as compared with healthy pregnant controls, with no difference in spontaneous and therapeutic abortions, preterm delivery, congenital abnormalities and birth weight. Logistic regression analysis identified prognostic factors with potential influence on pregnancy outcome: older age, smoking, disease activity, alcohol and IBD-related medication. Smoking and age ≥ 35 years were found to be risk factors for preterm delivery in CD, with the latter being associated with an increased risk of congenital abnormalities in CD women and preterm deliveries in UC patients. Disease activity during pregnancy and older age were associated with a lower birth weight in patients with UC. Women on 5-ASA monotherapy had lower probability of preterm delivery in UC and caesarean section in CD while combination therapy increased the risk of these events.
The important strength of this study is the prospective character which allows standardised follow-up of both pregnant women and matched healthy pregnant controls. This offers the high validity of the data and a solid level of evidence regarding the outcome of pregnancy in women with IBD.28, 31
Previous case-control studies have shown an increased risk of adverse pregnancy outcome in women with IBD.7, 9 In the study of Baird et al.,7 IBD women who were retrospectively matched by age, parity and smoking status, showed a two to threefold increased risk of preterm birth compared with their healthy controls. Another retrospective case-control study9 comparing pregnancy outcome of IBD women in pre- and post- diagnosis period, revealed that prior to diagnosis of IBD, CD patients carried an increased risk of preterm delivery, whereas infants of UC patients had an increased risk of LBW. After diagnosis, women with IBD were shown to have higher incidence of CAs while patients with CD had a higher risk of LBW infants.
Several other large population-based11, 15 and registry-based studies12, 25 also reported an increased risk of preterm delivery and LBW in women with both CD and UC. However, neither disease activity nor therapies were assessed in these studies.
Different results were found in women with UC in a large nationwide Danish study25 involving 1531 infants in which no differences were demonstrated in rates of preterm delivery overall considered, LBW and small for gestational age (SGA) compared with controls. Finally, a meta-analysis5 of 12 studies performed from 1980 to 2006 including 3907 IBD patients and 320 531 controls reported an increased risk of premature birth both in CD and UC patients, and higher rates of caesarean sections and LBW in infants of CD women.
In our study, pregnancy outcome in women with CD and UC did not differ significantly from that of the healthy background population. This favourable result might be attributed to the high remission rates of our patients, with 86% and 88% of CD and UC women, respectively, having quiescent disease at conception/first trimester. This is also in agreement with findings of previous studies showing a higher risk of premature birth, LBW, stillbirths and miscarriages when active disease is present at conception.7, 21, 23, 24
An increased risk of CAs in women with UC than in healthy controls (7.9% vs. 1.7%; OR 3.8, 95% CI: 1.5–9.8) was shown in a study by Dominitz et al.11. Correspondingly, higher incidence rate of CAs in patients with UC vs. controls (OR 3.8; 95% CI 1.47–3.82) was found in the recent meta-analysis by Cornish et al.5 In contrast, our study did not find an increased risk of CAs either in UC or in CD. Similarly, no significantly increased overall risk of CAs was shown in infants born to women with UC in a large population-based case-control registry study by Norgard et al.18 (OR 1.3; 95% CI: 0.9–1.8).
When looking at separate factors with impact on pregnancy outcome, maternal age above 35 years and smoking women with CD carry a risk of preterm delivery in our study. Tobacco use in women with CD was considered as a risk factor in study by Fonager et al.12 Another study,16 however, has shown that CD diagnosis and maternal age above 35 years, but not current tobacco use, are predictive of adverse pregnancy outcome. Active disease during pregnancy and maternal age above 35 years were associated with a LBW in patients with UC in our study. Our results are supported by the findings in the other studies,19, 23 showing that if the pregnancy occurs while the disease is active; there is a high risk of premature birth or LBW in infant born to women with UC.
More than 80% of CD and UC women were on medication during pregnancy with the majority of them receiving either 5-ASA preparations (66% and 83%), corticosteroids (22% and 40%) or azathioprine/6-mercaptopurine (22% and 10%). Eight women were treated with anti-TNFα medications. The finding of an association between combination therapy and increased risk of preterm delivery in UC and caesarean sections in CD, with an inverse result regarding 5-ASA monotherapy, might be just a reflection of a more or less complicated disease course in such patients requiring more or less intensive anti-inflammatory therapy. No impact of medication on occurrence of CAs was observed in our study.
To date, there is an increasing evidence that the majority of the drugs used in IBD are safe during pregnancy and that there is no need to withdraw the therapy prior to or after conception.32–34 Moreover, there is increasing evidence that biological therapy is also feasible and safe during pregnancy until week 32.35, 36 Discontinuation of the medication, may lead to disease activity which has been reported to be associated with an adverse pregnancy outcome.23, 24 Thus it is recommended today to sustain medical therapy, apart from methotrexate, during pregnancy to maintain IBD patients in remission.32–34
It has to be emphasised that 46% of CD patients and 30% of UC women on 5-ASA therapy in our study received a dose which is higher than recommended (<3000 mg) in ECCO consensus.31 This may be clinically important, mainly regarding a recent concern about dibutyl phthalate (DBP) included in 5-ASA enteric-coating preparations. High dose of DBP in animal studies37 has been shown to be associated with foetal malformations. To date, there is limited evidence of potential side effects also in humans.20, 38, 39 Since until 2005 the substantial proportion of 5-ASA preparations contained DBP, the majority of our women included and treated with 5-ASA were probably exposed to DBP. Nevertheless, the pregnancy outcome in both CD and UC women was favourable and no CAs in newborns were observed.
The study has several limitations. Firstly, there was the big drop out of enrolled women due to a lack of matching and incomplete data record which influenced the final sample size. This could have had an impact on statistical significance of some rare pregnancy events, such as CAs, therapeutic and spontaneous abortions. Thus, further prospective studies are needed to confirm or disprove our results. Secondly, the number of our women treated with immunomodulators or biologicals was too low to make any firm conclusion about the impact of the therapy on pregnancy outcome. On the other hand, our study population reflects to some extent the everyday clinical practice when the majority of women conceive while being in remission and do not require immunosuppressive or biological therapy.
In conclusion, this is the first prospective European case-control study performed in pregnant women with IBD. The study found similar pregnancy outcomes in women with IBD as compared with their age and parity matched non-IBD controls. We underline that the majority of pregnant IBD patients were in remission on maintenance therapy. Regression analysis showed that age < 35 years, inactive disease course and nonsmokers carry the least risk of adverse pregnancy outcome. Taking into consideration that the incidence of IBD has substantially increased in the past decades and the natural disease course has also changed, there is still a need for prospective evaluations.
The study was initiated by ECCO. The guarantors of the manuscript are AB, NP and PM. AB, NP and an ECCO collaborative group performed the data collection. DD, NP, JKJ (Statistician) and PM performed the statistical analyses. AB and NP drafted the manuscript, which was critically revised by all co-authors. All authors approved the final version of the manuscript.
We thank all collaborators from participating centres who contributed with inclusion of patients and controls: Alberto Prada, Rho Hospital, Rho, Italy; Colm O’Morain Dublin University, Ireland; Giovanni Casella, Desio Hospital, Italy; Giorgio Zoli, Cento Hospital, Italy; Licia Snider, Sant’Anna Hospital, Italy; Marco Daperno, Mauriziano Hospital, Torino, Italy; Pierenrico Lecis, San Martino Hospital, Belluno, Italy; Christian Felley, Lausanne, Switzerland; Mario Gatti, Giussano Hospital, Italy; Giovanni Fornaciari, Reggio Emilia, Italy; Renzo Caprilli, La Sapienza University, Roma, Italy; Maddalena Terpin, Legnano Hospital, Italy; Fabrizio Bossa, San Giovanni Rotondo, Italy; Giorgio Mortara, San Carlo Hospital, Milano, Italy; Roberto de Franchis IRCCS Policlinico, Milano, Italy; Maurizio Vecchi San Donato Hospital, San Donato, Italy; Michele Comberlato, Bolzano Hospital, Italy; Christian Folwaczny, Poliklinik, Innerstadt München, Germany; Enrico Colombo, Garbagnate Hospital, Italy; Carlo Maria Girelli, Busto Hospital, Italy; Paolo Ravelli, Bolognini Hospital, Seriate, Italy; Walter Reinisch, University Hospital, Vienna, Austria; Luca Ferraris, Gallarate Hospital, Italy; Guido Lupinacci, Crema Hospital, Italy; Daan Hommes, Leiden, The Netherlands; Gerhard Rogler, University Klinik of Regensburg, Germany; Vito Annese, San Giovanni Rotondo, Italy; Maria Carla Di Paolo, S.Giovanni Addolorata Hospital, Rome, Italy; Epameinondas V. Tsianos, Joannina, Greece; Gerhard Rogler, University Klinik of Regensburg, Germany; Konstantinos Katsanov, Joannina, Greece; Venerina Imbezi, Policlinico San Matteo, Pavia, Italy; Anna Kohn, San Camillo Forlanini, Rome, Italy. Declaration of personal interests: None. Declaration of funding interests: The authors have received funding from ECCO and from a research fund in the department of Pia Munkholm for the present study.