ORIGINAL ARTICLE
Birth Outcomes in Women with Inflammatory Bowel Disease: Effects of Disease Activity and Drug Exposure Gabriella Bröms, MD,* Fredrik Granath, PhD,* Marie Linder, PhD,* Olof Stephansson, MD, PhD,* ,† Maria Elmberg, MD, PhD,* and Helle Kieler, MD, PhD*
Background: Ulcerative colitis (UC) and Crohn’s disease (CD) have been associated with increased risks of adverse birth outcomes. Disease activity and drug exposure may contribute to the association.
Methods: A cohort from the Swedish health registers including 470,110 singleton births in Sweden from July 2006 to December 2010; 1833 to women with UC and 1220 to women with CD. Birth outcomes for women with UC and CD were compared with outcomes among those without disease. Diseased women were categorized by drug exposure, need of surgery, and hospital admissions as (1) no disease activity and (2) stable or (3) flaring disease. Logistic regression was used to calculate odds ratios with adjustments (aOR) for maternal age, parity, smoking status, body mass index, and comorbidity.
Results: There were increased risks of preterm birth for both UC (aOR, 1.78; 95% confidence interval [CI], 1.49–2.13) and CD (aOR, 1.65; 95% CI, 1.33–2.06). Risks were more pronounced in women with flaring disease during pregnancy. Risks of small for gestational age, low Apgar score, and hypoglycemia were also increased. The risk of stillbirth was elevated in women with CD, particularly among those with flaring disease (aOR, 4.48; 95% CI, 1.67–11.90). Thiopurine exposure increased risks for preterm birth, both in women with stable (aOR, 2.41; 95% CI, 1.05–5.51) and with flaring disease (aOR, 4.90; 95% CI, 2.76–8.69). Conclusions: Women with UC and CD are at increased risk of adverse birth outcomes, such as stillbirth, growth restriction, and preterm birth, particularly when they suffer from flares throughout pregnancy. Thiopurine exposure seems to further increase risks, independently of disease activity. (Inflamm Bowel Dis 2014;20:1091–1098) Key Words: pregnancy, Crohn’s disease, ulcerative colitis
I
nflammatory bowel disease (IBD) includes ulcerative colitis (UC) and Crohn’s disease (CD). These chronic conditions often affect women in their reproductive years, and prevalence is rising.1 Questions about consequences of IBD and its associated medical therapy for the offspring are often brought up when physicians are counseling women of childbearing age. A higher frequency of voluntary childlessness among women with IBD has been observed, and concerns for fetal exposure to medical therapy seem to be part of the reason.2,3 The natural course of UC and CD involves fluctuating disease activity, where periods of clinical remission are interrupted
Received for publication March 17, 2014; Accepted March 31, 2014. From the *Unit of Clinical Epidemiology, Centre for Pharmacoepidemiology, Karolinska Institutet, Stockholm, Sweden; and †Department of Women’s and Children’s Health, Karolinska University Hospital and Institutet, Stockholm, Sweden. Centre for Pharmacoepidemiology receives financial support from several pharmaceutical companies, including Janssen Biotech. This study was independently developed from a project initiated by Janssen Biotech. M. Elmberg has been a consultant (none above 1000 USD) for Tillotts Pharma, AbbeVie, Ferring. Grant support (not above 1500 USD) from MSD. The remaining authors have no conflicts of interest to disclose. Reprints: Gabriella Bröms, MD, Centre for Pharmacoepidemiology, Building T2, Karolinska University Hospital, Stockholm 171 76, Sweden (e-mail: gabriella. [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000060 Published online 7 May 2014.
Inflamm Bowel Dis Volume 20, Number 6, June 2014
by flares of increased bowel inflammation. Management of IBD aims at maintaining remission, which often requires medical treatment. Typically, systemic 5-aminosalicylates, sulfasalazine, thiopurines, and TNFa-inhibitors are used for this purpose. Flares are usually treated with corticosteroids, rectal administration of 5-aminosalicylates, or bowel surgery. Previous studies have addressed birth outcomes in women with IBD and in women with UC and CD separately; however, results remain conflicting. Increased risks of preterm birth and low birth weight have been reported for both UC and CD.4–6 A higher proportion of infants born small for gestational age (SGA) has been observed in some studies, whereas others have failed to show such an association.6–9 Distinguishing effects of medical therapy and effects of disease activity remain a challenge. The aim of this study was to assess risks of adverse birth outcomes in women with UC and CD by disease activity and drug exposure.
MATERIALS AND METHODS We conducted a population-based cohort study in which all singleton births identified in the Medical Birth Register (MBR) between July 1, 2006 and December 31, 2010, were included. The personal identification number assigned to all residents in Sweden enabled merging of data between the MBR, the National Patient Register, and the Prescribed Drug Register on an individual level. www.ibdjournal.org |
1091
Bröms et al
Women with IBD starting from 1997 and before childbirth were identified using the National Patient Register that holds nationwide information on both inpatient and outpatient hospital visits (outpatient visits from 2001 and onwards). Data for each visit include discharge diagnoses based on the International Classification of Diseases, Tenth Revision (ICD-10), date of admission, and discharge and codes for surgical procedures. We used ICD-codes K51 for UC and K50 for CD, respectively. In cases where the patient had received both a diagnosis of UC and of CD, the most recent diagnosis was used. To be considered as having IBD, patients were required to have had at least 2 recorded diagnoses of UC or CD or 1 visit or hospitalization in combination with a medical or surgical treatment for IBD. The National Patient Register was also used to obtain data on comorbidities within 5 years before pregnancy including diabetes (E10-E11, O240-O241), hypertension (I10-I15, O10), chronic autoimmune disease (primary sclerosing cholangitis K83; primary biliary cirrhosis K743; rheumatoid arthritis M05-06; vasculitis, SLE and other systemic disease M30-35; and psoriasis L40), and asthma (J45). Filled drug prescriptions are recorded in the Prescribed Drug Register by date of dispensing, drug name, and preparation according to the Anatomical Therapeutic Chemical classification. In general, prescriptions are filled for a maximum of 3 months. Exposure to drugs used in the treatment of IBD was defined as having filled at least 1 prescription of sulfasalazine (A07EC01), aminosalicylates (A07EC02-03), and corticosteroids for systemic use: betamethasone, methylprednisolone, prednisolone, prednisone, hydrocortisone (H02AB), rectal prednisolone (A07EA01), rectal hydrocortisone (A07EA02), rectal budesonide (A07EA06), azathioprine (L04AX01), 6-mercaptopurine (L01BB02), infliximab (L0AB02), and adalimumab (L04AB04) during pregnancy. The MBR includes information from standardized records that are completed by physicians and midwives throughout the antenatal and perinatal care program. We used the variables of maternal age when giving birth, parity, smoking status, and body mass index in early pregnancy to establish maternal characteristics. Adverse birth outcomes were assessed as rates and risks of preterm birth (,37 wk of gestation), spontaneous preterm birth, preterm premature rupture of the membranes, low birth weight (,2500 g), SGA, Apgar score ,7 at 5 minutes, stillbirth, hypoglycemia, and neonatal hyperbilirubinemia. For SGA, we assessed birth weights by gender ,2 SD below the mean and birth weights in the lower 10th percentile. Stillbirth was classified as intrauterine death occurring from gestational week 28 and onwards for births between 2006 and 2007. From 2008, stillbirths from gestational week 22 and onwards were recorded in the MBR and thus included in the cohort. In post hoc analyses, we tested whether restricting the analyses to only include stillbirths occurring from gestational age 28 weeks and onwards would influence the risk estimates for stillbirth. Spontaneous preterm birth was defined as spontaneous start of labor excluding induction or elective cesarean delivery. Preterm premature rupture of the membranes was defined by ICD-10 code O42.
1092
| www.ibdjournal.org
Inflamm Bowel Dis Volume 20, Number 6, June 2014
We used logistic regression to analyze risks of each birth outcome in infants to women with UC and CD, respectively, by comparing them with infants to women without these diseases. Risks were computed as crude and adjusted odds ratios (aORs) with 95% confidence intervals (CI). Women with UC and with CD were categorized by disease activity according to a proxy model as women with: (1) No disease activity, no medication, surgery, or hospitalization, (2) Stable disease: systemic 5-aminosalicylates, sulfasalazine, thiopurines, or TNFa-inhibitors throughout pregnancy, or (3) Flaring disease: corticosteroids, rectal 5-aminosalicylates after endoscopic examination, surgery, or hospital admission with IBD as primary diagnosis. For women with flares, we assessed risks of adverse outcome by timing of the flares as early, late, or both. Women with flares occurring from 3 months before conception and during the first trimester of pregnancy only were classified as having early pregnancy flares. Late pregnancy flares were those occurring in the second or third trimester. For women with stable or flaring disease, we also assessed risks of adverse birth outcomes associated with exposure to thiopurines. Adjustments were made for maternal age, parity, smoking, body mass index, and comorbidity (a diagnosis of diabetes, hypertension, chronic autoimmune disease, or asthma). We also performed analyses stratified by smoking status. Because observations were not independent in women who gave birth more than once during the study period, we used the generalized estimation equation method for clustered data. All analyses were conducted using SAS software (version 9.3; SAS Institute Inc, Cary, NC). The proxy model for disease activity was validated by scrutinizing medical charts of women with UC or CD from the cohort who gave birth in 2006. To be eligible, they had records of either at least 1 inpatient visit or 2 or more outpatient visits because of IBD during pregnancy implying flaring disease. The charts of 88 women were thoroughly examined concerning information on clinical events indicating disease activity. Frequent bowel movements, blood or mucus in stools, abdominal pain, and fever indicated flaring disease. Assessment was performed by one of the authors (G.B.) who was blinded to our disease activity proxy.
Ethical Considerations The study was conducted in accordance to ethical review by the Regional Ethical Review Board, Stockholm (Record number 2006/5:7).
RESULTS We included 1833 singleton births to 1492 women with UC and 1220 births to 1026 women with CD. All 467,057 births to women without UC or CD occurring during the study period constituted the comparison group. Women with UC and with CD were somewhat older and those with CD were more often smokers and had lower body mass index (Table 1). Among women with UC, there was 30% with no disease activity, 34% with stable disease, and 36% with flaring disease.
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Births Outcomes in Inflammatory Bowel Disease
TABLE 1. Descriptive Data on Singleton Births Among Women with and Without IBD
Maternal age at birth, yr #24 25–34 $35 Missing Parity Primiparous Multiparous Smoking in first trimester Nonsmoker Smoker Missing Body mass index #19 20–24 25–29 $30 Missing Comorbidity prepregnancya Yes Disease activity during pregnancy No disease activity Stable disease Flaring disease Timing of flare Early pregnancy Late pregnancy Early and late pregnancy TNFa-inhibitor exposure Thiopurine exposureb Yes No
UC N ¼ 1833, n (%)
CD N ¼ 1220, n (%)
No IBD N ¼ 467,057, n (%)
143 (7.80) 1256 (68.52) 434 (23.68) —
131 (10.74) 813 (66.64) 276 (22.62) —
68,592 (14.69) 296,334 (63.45) 102,128 (21.87) 3
827 (45.12) 1006 (54.88)
545 (44.67) 675 (55.33)
210,404 (45.05) 256,653 (54.95)
1703 (92.91) 59 (3.22) 71 (3.87)
1058 (86.72) 122 (10.00) 40 (3.28)
416,556 (89.19) 31,043 (6.65) 19,458 (4.17)
177 (9.66) 954 (52.05) 423 (23.08) 138 (7.53) 141 (7.69)
148 (12.13) 575 (47.13) 286 (23.44) 110 (9.02) 101 (8.28)
42,724 225,457 108,005 51,453 39,418
198 (10.80)
167 (13.69)
31,434 (6.73)
542 (29.57) 625 (34.10) 666 (36.33)
551 (45.16) 347 (28.44) 322 (26.39)
— — —
207 247 212 6
113 92 117 19
— — — —
159 1132
262 407
— —
(9.15) (48.27) (23.12) (11.02) (8.44)
a
Diabetes, hypertension, autoimmune disease, and asthma. Thiopurine exposure among women with stable and flaring disease.
b
The corresponding proportions for CD were 45%, 28%, and 26%, respectively. There were 438 women with UC and CD who gave birth more than once during the study period, and 66% of these had similar disease activity during their pregnancies. There were 109 women eligible for the validating chart review. We were able to collect charts of 88 women, corresponding to a response rate of 81%. Of these, 59% had signs of flaring disease from 3 months before and throughout pregnancy, and 41% had stable disease or no disease activity. When we compared our proxy model with the medical records, the sensitivity for flaring disease was 83% and the specificity was 81%.
Risks of adverse birth outcomes in association with UC or CD are shown in Table 2 and risks by disease activity in Table 3. Preterm births were more common (8.4%) in women with UC or CD as compared with women without IBD (5.0%). The aORs for preterm births were 1.78 (95% CI, 1.49–2.13) for UC and 1.65 (95% CI, 1.33–2.06) for CD, respectively. For CD, restricting the analyses to spontaneous preterm birth only, the aOR was 1.43 (95% CI, 1.07–1.91). The highest risks for preterm birth and for spontaneous preterm birth were found among those with flaring disease. Analyses of low birth weight exhibited a similar pattern with risks for UC in general at aOR 1.64 (95% CI, 1.31–2.04) and www.ibdjournal.org |
1093
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
TABLE 2. Birth Outcomes in Women with UC and CD Compared with Women Without IBD UC, N (%) Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
153 71 39 99 54 196 6 41 46 86
(8.35) (3.87) (2.13) (5.40) (2.95) (10.69) (0.33) (2.24) (2.51) (4.69)
CD, N (%) Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
103 57 28 78 49 127 11 30 33 60
(8.44) (4.67) (2.30) (6.39) (4.02) (10.41) (0.90) (2.46) (2.70) (4.92)
No IBD, N (%)
23,492 14,689 6548 16,014 10,954 46,637 1558 8960 9445 17,858
No IBD, N (%)
23,492 14,689 6548 16,014 10,954 46,637 1558 8960 9445 17,858
for CD at aOR 1.86 (95% CI, 1.46–2.38). For flaring UC, the risk estimate in the adjusted analysis was doubled, and for CD, it was tripled. Women with CD had increased risks of SGA births, with the highest risks among those with flaring disease. For flaring CD, there were increased risks of low 5-minute Apgar score (aOR, 2.21; 95% CI, 1.24–3.93) and hypoglycemia (aOR, 2.38; 95% CI, 1.40–4.04). Systemic corticosteroid treatment was more common in women whose infants had hypoglycemia, 29% versus 18%. When adjusting for SGA, risks for hypoglycemia in infants born to women with flaring CD was lowered (aOR, 1.97; 95% CI, 1.11–3.50). For women with UC, an increased risk of hyperbilirubinemia was observed in women with flaring disease (aOR, 1.58; 95% CI, 1.12–2.22), but the risk increase did not persist when adjusting for preterm birth. Risks of stillbirth were increased in women with CD, with risk estimates rising across disease activity groups to an aOR in flaring disease of 4.48 (95% CI, 1.67–11.90). These results were based on 11 cases among births to women with CD. Six of these were preterm. One was SGA below 2 SD, and 4 were SGA below the 10th percentile. Restricting the analyses to only include stillbirths occurring in gestational week 28 or later and stratifying for smoking had only minor effects on the risk estimates (results not shown in tables).
1094
| www.ibdjournal.org
(5.03) (3.15) (1.40) (3.43) (2.35) (9.99) (0.33) (1.92) (2.02) (3.82)
(5.03) (3.15) (1.40) (3.43) (2.35) (9.99) (0.33) (1.92) (2.02) (3.82)
Crude OR (95% CI)
1.79 1.35 1.67 1.63 1.23 1.08 1.17 1.16 1.19 1.25
(1.50–2.14) (1.06–1.73) (1.21–2.30) (1.31–2.03) (0.92–1.64) (0.92–1.27) (0.52–2.61) (0.84–1.62) (0.87–1.63) (1.00–1.57)
Crude OR (95% CI)
1.76 1.48 1.60 2.02 1.86 1.04 2.98 1.33 1.45 1.35
(1.42–2.19) (1.11–1.98) (1.07–2.40) (1.58–2.58) (1.37–2.52) (0.85–1.26) (1.60–5.54) (0.91–1.94) (1.01–2.09) (1.03–1.77)
aOR (95% CI)
1.78 1.37 1.63 1.64 1.26 1.11 1.22 1.19 1.16 1.26
(1.49–2.13) (1.07–1.75) (1.18–2.25) (1.31–2.04) (0.95–1.69) (0.94–1.30) (0.54–2.71) (0.85–1.65) (0.85–1.59) (1.01–1.58)
aOR (95% CI)
1.65 1.43 1.49 1.86 1.72 0.97 2.93 1.30 1.32 1.32
(1.33–2.06) (1.07–1.91) (1.00–2.25) (1.46–2.38) (1.27–2.34) (0.80–1.18) (1.57–5.47) (0.88–1.88) (0.89–1.91) (1.00–1.74)
For women with UC or with CD, the risks of adverse birth outcomes were, with some exceptions, higher when flares occurred in late pregnancy than in early pregnancy. The highest risks were in women with flares in both early and late pregnancy (Table 4). In general, women with flares treated with thiopurines had the highest risks of preterm birth and SGA. The influence of thiopurines on risks of preterm birth seemed to be independent of disease activity because higher risks were found both among women with stable disease and those with flaring disease (Table 5).
DISCUSSION In this study, we observed increased risks of preterm birth and low birth weight among women with UC and CD. Adverse birth outcomes were, generally associated with flaring disease during pregnancy, particularly in women with CD, who had a 4-fold increased risk of stillbirth. For preterm birth and the growth-related outcomes, we found strong associations with both disease activity and with thiopurine treatment.
Stillbirth Elevated risks of stillbirth in women with IBD have, to our knowledge, not been reported previously.5 In our previous studies
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Births Outcomes in Inflammatory Bowel Disease
TABLE 3. Birth Outcomes by Disease Activity: aORs, Compared with Women Without IBD No Disease Activity UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
30 15 6 20 12 42 3 11 15 19
aOR (95% CI)
1.30 1.06 0.90 1.28 1.12 0.91 2.02 1.24 1.52 1.06
(0.89–1.90) (0.61–1.83) (0.40–2.02) (0.82–2.01) (0.63–1.99) (0.66–1.25) (0.65–6.28) (0.68–2.25) (0.90–2.55) (0.67–1.68)
Stable Disease N
35 18 8 27 15 67 2 13 10 24
No Disease Activity CD
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
26 14 4 21 14 43 3 10 7 25
aOR (95% CI)
1.05 0.92 0.57 1.24 1.18 0.84 1.95 1.08 0.66 1.36
(0.70–1.58) (0.54–1.57) (0.21–1.52) (0.79–1.97) (0.67–2.10) (0.61–1.15) (0.63–6.05) (0.58–2.00) (0.31–1.41) (0.90–2.07)
including women with UC and CD giving birth in Denmark and Sweden, we found no increased risks of stillbirth, except for women with UC who had undergone previous surgery.8,9 Different time periods and different approaches to identify women with IBD may to some extent explain these inconsistent findings. With increasing opportunities to treat CD over time and the stricter definition of IBD, requiring at least 2 recorded diagnoses or 1 diagnosis together with an adjacent medical or surgical treatment instead of 1 diagnosis, it is possible that a higher proportion of women included in this study had a more severe disease. That women with flaring disease had the highest risks for stillbirth supports that disease severity might have had a major impact on the results. Also, in the previous study, we only included primiparous women and although we, in this study, adjusted for parity and performed cluster analyses, factors relating to parity might still have influenced the results. This study is, however, more representative of pregnant women in general, than our previous study. Although the age limit for reporting intrauterine fetal deaths as stillbirth to the MBR changed in 2008 from gestational week 28 to 22, post hoc analyses including stillbirths from gestational week 28 had only minor effects on our risk estimates for stillbirth. In a study
aOR (95% CI)
1.26 1.03 0.98 1.45 1.15 1.20 1.18 1.24 0.84 1.10
(0.89–1.79) (0.65–1.65) (0.49–1.97) (0.98–2.14) (0.68–1.92) (0.93–1.56) (0.29–4.73) (0.72–2.15) (0.45–1.57) (0.73–1.65)
Flaring disease N
75 35 24 41 21 71 1 12 15 36
Stable Disease N
26 15 8 16 12 24 3 5 9 16
aOR (95% CI)
1.68 1.55 1.73 1.47 1.55 0.74 3.06 0.84 1.30 1.30
(1.12–2.52) (0.92–2.61) (0.85–3.53) (0.89–2.43) (0.87–2.75) (0.49–1.11) (1.00–9.49) (0.35–2.01) (0.66–2.55) (0.78–2.16)
aOR (95% CI)
2.72 1.92 2.82 2.10 1.49 1.17 0.57 1.09 1.19 1.58
(2.12–3.48) (1.37–2.70) (1.87–4.26) (1.51–2.90) (0.96–2.31) (0.91–1.50) (0.08–4.06) (0.62–1.94) (0.71–1.99) (1.12–2.22)
Flaring Disease N
39 20 12 34 21 47 4 12 16 15
aOR (95% CI)
2.66 2.13 2.74 3.30 2.75 1.46 4.48 2.21 2.38 1.28
(1.89–3.74) (1.35–3.37) (1.53–4.92) (2.29–4.74) (1.72–4.38) (1.06–2.02) (1.67–11.9) (1.24–3.93) (1.40–4.04) (0.76–2.15)
from the United States, Mahadevan et al10 reported increased risks of adverse birth outcomes in women with IBD, when combining stillbirth, SGA, and preterm birth as one entity, although no separate risk estimates were calculated for each outcome. The rate of stillbirth in the study by Mahadevan and in an Italian study by Bortoli was around 1%, which is in accordance with what we observed for women with CD.10,11 The observed increase in risks of stillbirth across disease activity groups in CD suggests that the extent of the inflammation is important. Whether the association is linked to the disease, the treatment, or both remains, however, unclear. Because of the limited sample size, we could not assess the impact of thiopurine exposure on stillbirth. Infection is an important risk factor for stillbirth, and because bowel inflammation and immunosuppressive therapy may increase the sensitivity to infections, it is plausible that a secondary infection could be the underlying cause to the increased risk of stillbirth observed here.12 It is important to note that although the relative risks were clearly elevated, stillbirth is a rare event in developed countries.13 Nevertheless, the observation suggests that women with CD require close surveillance during pregnancy and delivery. www.ibdjournal.org |
1095
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
TABLE 4. Birth Outcomes by Timing of Flare Indicator, Compared with Women Without IBD Activity in Early Pregnancy UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile
22 11 9 7 18
Activity in Late Pregnancy
aOR (95% CI)
2.60 1.96 1.46 1.59 0.92
N
(1.66–4.07) (1.07–3.61) (0.75–2.83) (0.75–3.41) (0.56–1.50)
aOR (95% CI)
23 9 15 7 30
2.12 1.27 1.98 1.28 1.32
Activity in Late Pregnancy
CD
N
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile
7 5 6 5 7
1.26 1.48 1.54 1.79 0.55
(0.58–2.72) (0.60–3.65) (0.67–3.50) (0.75–4.27) (0.26–1.17)
N
(1.38–3.26) (0.66–2.46) (1.18–3.35) (0.60–2.72) (0.89–1.95)
Activity in Early Pregnancy aOR (95% CI)
Activity in Both Early and Late Pregnancy
30 15 17 7 23
2.41 1.55 3.96 4.03 2.22
3.62 2.71 2.90 1.65 1.25
(2.40–5.45) (1.60–4.58) (1.70–4.95) (0.77–3.53) (0.81–1.94)
Activity in Both Early and Late Pregnancy
aOR (95% CI)
10 4 11 8 17
aOR (95% CI)
N
(1.23–4.73) (0.60–4.24) (2.07–7.55) (1.94–8.38) (1.31–3.78)
aOR (95% CI)
22 11 17 8 22
4.43 3.23 4.67 2.79 1.95
(2.76–7.11) (1.73–6.03) (2.77–7.89) (1.32–5.87) (1.21–3.15)
TABLE 5. Birth Outcomes by Thiopurine Exposure No Thiopurine Exposure Stable Disease UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2SD SGA, below 10th percentile
28 14 25 15 65
aOR (95% CI)
1.12 0.90 1.52 1.30 1.34
(0.77–1.65) (0.53–1.54) (1.01–2.28) (0.78–2.18) (1.03–1.75)
Thiopurine Exposure
Flaring Disease N
58 24 32 16 60
aOR (95% CI)
2.40 1.51 1.88 1.30 1.15
(1.83–3.16) (1.01–2.28) (1.32–2.69) (0.79–2.14) (0.87–1.50)
Stable Disease N
aOR (95% CI)
N
7 4 2 0 2
2.41 (1.05–5.51) 2.07 (0.79–5.45) 0.91 (0.24–3.47) — 0.26 (0.06–1.09)
17 11 9 5 11
No Thiopurine Exposure Stable Disease CD
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2SD SGA, below 10th percentile
9 5 7 9 15
1096
| www.ibdjournal.org
aOR (95% CI)
1.09 1.01 1.26 2.41 0.98
(0.56–2.15) (0.41–2.48) (0.59–2.69) (1.23–4.72) (0.54–1.65)
25 15 20 11 29
aOR (95% CI)
2.25 2.18 2.53 1.88 1.17
aOR (95% CI)
4.90 4.64 3.50 1.58 1.30
(2.76–8.69) (2.48–8.68) (1.63–7.54) (1.04–6.41) (0.69–2.47)
Thiopurine Exposure
Flaring Disease N
Flaring Disease
(1.48–3.42) (1.29–3.69) (1.59–4.02) (0.97–3.62) (0.77–1.76)
Stable Disease N
17 10 9 3 9
aOR (95% CI)
2.34 2.12 1.70 0.74 0.51
(1.40–3.90) (1.12–4.03) (0.88–3.30) (0.24–2.32) (0.26–1.02)
Flaring Disease N
14 5 14 10 14
aOR (95% CI)
3.92 2.01 5.74 5.52 2.39
(2.19–7.01) (0.81–5.00) (3.18–10.4) (2.83–10.80) (1.42–4.03)
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Preterm Birth Being born preterm is associated with significant morbidity.14 Increased risks of preterm birth among women with IBD in general, and UC and CD separately, have been reported previously.5,6,8,9,15 In contrast to this study, none of the previous studies did, however, specifically assess spontaneous preterm birth. We found increased risks for spontaneous preterm birth, with the risk estimates being only slightly lower than those found when assessing preterm birth in general, which suggests that other than iatrogenic factors might explain the previously reported associations. Inflammation is part of the physiology of birth and infection, subclinical, or clinical, is often present in spontaneous preterm birth.16,17 Accordingly, both the immunological similarities between IBD and the initiation and progression of labor, and the increased risks of a secondary infection in women with IBD could pose possible explanations to the association between IBD and preterm birth.18
Fetal Growth and Hypoglycemia Low birth weight has been reported to be more prevalent in women with IBD.5 However, low birth weight is a less accurate measure of fetal growth than SGA. SGA might be a sign of intrauterine growth restriction and is associated with adverse perinatal outcomes and long-term morbidity.19 In this study, 2 cutoffs for SGA were used—birth weights below 2 SD according to fetal gender and birth weights below the 10th percentile. For CD, we found increased risks of SGA for the stricter cutoff, whereas for UC, there was no increased risk. In our previous study, risks were less elevated, whereas an American study reported a 2-fold increase in risks and presumably most of these women exhibited some disease activity.6,8 Hypoglycemia is the most common metabolic disorder in the neonatal period and growth-restricted infants are at a higher risk.20 We observed increased risks of hypoglycemia in women with flaring CD. Being SGA confounded this result, although the risk remained increased even after adjustment for SGA. Hypoglycemia has been linked to intrauterine growth restriction, and the higher risks of neonatal hypoglycemia and of SGA suggest an association between intrauterine growth restriction and CD.21 The associations between SGA and hypoglycemia and CD should be investigated in future studies.
Thiopurine Exposure and Timing of Flare Treatment with thiopurines was generally associated with increased risks of both growth restriction and preterm birth. The risk estimates associated with thiopurine treatment were mostly higher than the estimates for no treatment, both among women with stable and flaring disease, although the CIs overlapped. The highest risk estimates were found for flaring disease. The increased risks of preterm birth associated with thiopurines have been observed in some previous studies but not in others.22–25 Because thiopurine treatment had been used in both stable and flaring disease, the impact of the drug could be evaluated indirectly irrespective of disease activity, and our data suggest that disease activity mainly seems to affect fetal growth, whereas thiopurine treatment might be associated with preterm birth.
Births Outcomes in Inflammatory Bowel Disease
In this study, we also assessed risks according to timing of flare activity. Having flare events in both early and late pregnancy was associated with 3-fold risk increases of preterm birth and low birth weight for UC and a 4-fold increase for CD. Flare in late pregnancy was associated with higher observed risk estimates than activity in early pregnancy. To our knowledge, the impact of timing of flare activity on adverse birth outcomes has not been reported previously.
Strengths and Limitations This study includes one of the largest study samples of women with UC and CD assessing risks of adverse birth outcomes and one of the few to include data on clinical visits and pharmacological treatment. The population-based design generates high generalizability for women with UC and CD seen at hospitals. Medical care in Sweden is public, and data in the national health registers have been found to be accurate and of high validity.26–28 A limitation is the lack of first-hand clinical data such as frequency of bowel movements, blood in stool, and biochemical markers of inflammation, which would have further improved the assessment of disease activity during pregnancy. In line with previous studies, we used data on clinical events and drug prescriptions, which are available in the national health registers, as proxies for disease activity.9,29 The proxy model was validated by scrutinizing a sample of medical charts and found to have high sensitivity and specificity for flaring disease. We were not able to assess the actual intake of drugs, but it should be noted that the agreement between self-reported use and 1 or more filled prescriptions for drugs used for IBD during pregnancy has been found to be high.30
CONCLUSIONS Women with UC and CD had increased risks of adverse birth outcomes, such as low birth weight and preterm birth. Women with CD also had increased risks of stillbirth. For most of the evaluated outcomes, the risk increase was found among women with flaring disease during pregnancy. High disease activity throughout pregnancy and thiopurine treatment denoted further increased risks of adverse birth outcome. The results confirm that maintaining remission and obstetric surveillance is imperative, but the optimal treatment is yet to be established in further comparative studies.
ACKNOWLEDGMENTS Author contributions: G. Bröms had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. G. Bröms, F. Granath, O. Stephansson, M. Elmberg, and H. Kieler conceived and designed the study. G. Bröms and H. Kieler drafted the manuscript. G. Bröms and M. Linder were responsible for the statistical analysis. All authors participated in interpreting the data and critically revising the manuscript. www.ibdjournal.org |
1097
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
REFERENCES 1. Molodecky NA, Soon IS, Rabi DM, et al. Increasing incidence and prevalence of the inflammatory bowel diseases with time, based on systematic review. Gastroenterology. 2012;142:46 e42–54 e42; quiz e30. 2. Mountifield R, Bampton P, Prosser R, et al. Fear and fertility in inflammatory bowel disease: a mismatch of perception and reality affects family planning decisions. Inflamm Bowel Dis. 2009;15:720–725. 3. Selinger CP, Eaden J, Selby W, et al. Inflammatory bowel disease and pregnancy: lack of knowledge is associated with negative views. J Crohns Colitis. 2013;7:e206–e213. 4. Norgard B, Fonager K, Sorensen HT, et al. Birth outcomes of women with ulcerative colitis: a nationwide Danish cohort study. Am J Gastroenterol. 2000;95:3165–3170. 5. Cornish J, Tan E, Teare J, et al. A meta-analysis on the influence of inflammatory bowel disease on pregnancy. Gut. 2007;56: 830–837. 6. Dominitz JA, Young JC, Boyko EJ. Outcomes of infants born to mothers with inflammatory bowel disease: a population-based cohort study. Am J Gastroenterol. 2002;97:641–648. 7. Kornfeld D, Cnattingius S, Ekbom A. Pregnancy outcomes in women with inflammatory bowel disease–a population-based cohort study. Am J Obstet Gynecol. 1997;177:942–946. 8. Stephansson O, Larsson H, Pedersen L, et al. Crohn’s disease is a risk factor for preterm birth. Clin Gastroenterol Hepatol. 2010;8:509–515. 9. Stephansson O, Larsson H, Pedersen L, et al. Congenital abnormalities and other birth outcomes in children born to women with ulcerative colitis in Denmark and Sweden. Inflamm Bowel Dis. 2011;17:795–801. 10. Mahadevan U, Sandborn WJ, Li DK, et al. Pregnancy outcomes in women with inflammatory bowel disease: a large community-based study from Northern California. Gastroenterology. 2007;133:1106–1112. 11. Bortoli A, Saibeni S, Tatarella M, et al. Pregnancy before and after the diagnosis of inflammatory bowel diseases: retrospective case-control study. J Gastroenterol Hepatol. 2007;22:542–549. 12. Smith GC, Fretts RC. Stillbirth. Lancet. 2007;370:1715–1725. 13. Cousens S, Blencowe H, Stanton C, et al. National, regional, and worldwide estimates of stillbirth rates in 2009 with trends since 1995: a systematic analysis. Lancet. 2011;377:1319–1330. 14. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet. 2008;371:261–269. 15. Nguyen GC, Boudreau H, Harris ML, et al. Outcomes of obstetric hospitalizations among women with inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol. 2009;7:329–334.
1098
| www.ibdjournal.org
16. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med. 2000;342:1500–1507. 17. Romero R, Espinoza J, Goncalves LF, et al. Inflammation in preterm and term labour and delivery. Semin Fetal Neonatal Med. 2006;11:317–326. 18. Nasef NA, Ferguson LR. Inflammatory bowel disease and pregnancy: overlapping pathways. Transl Res. 2012;160:65–83. 19. Longo S, Bollani L, Decembrino L, et al. Short-term and long-term sequelae in intrauterine growth retardation (IUGR). J Matern Fetal Neonatal Med. 2013;26:222–225. 20. Garg M, Devaskar SU. Glucose metabolism in the late preterm infant. Clin Perinatol. 2006;33:853–870; abstract ix-x. 21. Engineer N, Kumar S. Perinatal variables and neonatal outcomes in severely growth restricted preterm fetuses. Acta Obstet Gynecol Scand. 2010;89:1174–1181. 22. Francella A, Dyan A, Bodian C, et al. The safety of 6-mercaptopurine for childbearing patients with inflammatory bowel disease: a retrospective cohort study. Gastroenterology. 2003;124:9–17. 23. Norgard B, Pedersen L, Fonager K, et al. Azathioprine, mercaptopurine and birth outcome: a population-based cohort study. Aliment Pharmacol Ther. 2003;17:827–834. 24. Akbari M, Shah S, Velayos FS, et al. Systematic review and meta-analysis on the effects of thiopurines on birth outcomes from female and male patients with inflammatory bowel disease. Inflamm Bowel Dis. 2013;19: 15–22. 25. Coelho J, Beaugerie L, Colombel JF, et al. Pregnancy outcome in patients with inflammatory bowel disease treated with thiopurines: cohort from the CESAME Study. Gut. 2011;60:198–203. 26. Wettermark B, Hammar N, Fored CM, et al. The new Swedish Prescribed Drug Register–opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf. 2007;16:726–735. 27. Cnattingius S, Ericson A, Gunnarskog J, et al. A quality study of a medical birth registry. Scand J Soc Med. 1990;18:143–148. 28. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. 29. Norgard B, Pedersen L, Christensen LA, et al. Therapeutic drug use in women with Crohn’s disease and birth outcomes: a Danish nationwide cohort study. Am J Gastroenterol. 2007;102:1406–1413. 30. Stephansson O, Granath F, Svensson T, et al. Drug use during pregnancy in Sweden—assessed by the Prescribed Drug Register and the Medical Birth Register. Clin Epidemiol. 2011;3:43–50.
Birth Outcomes in Women with Inflammatory Bowel Disease: Effects of Disease Activity and Drug Exposure Gabriella Bröms, MD,* Fredrik Granath, PhD,* Marie Linder, PhD,* Olof Stephansson, MD, PhD,* ,† Maria Elmberg, MD, PhD,* and Helle Kieler, MD, PhD*
Background: Ulcerative colitis (UC) and Crohn’s disease (CD) have been associated with increased risks of adverse birth outcomes. Disease activity and drug exposure may contribute to the association.
Methods: A cohort from the Swedish health registers including 470,110 singleton births in Sweden from July 2006 to December 2010; 1833 to women with UC and 1220 to women with CD. Birth outcomes for women with UC and CD were compared with outcomes among those without disease. Diseased women were categorized by drug exposure, need of surgery, and hospital admissions as (1) no disease activity and (2) stable or (3) flaring disease. Logistic regression was used to calculate odds ratios with adjustments (aOR) for maternal age, parity, smoking status, body mass index, and comorbidity.
Results: There were increased risks of preterm birth for both UC (aOR, 1.78; 95% confidence interval [CI], 1.49–2.13) and CD (aOR, 1.65; 95% CI, 1.33–2.06). Risks were more pronounced in women with flaring disease during pregnancy. Risks of small for gestational age, low Apgar score, and hypoglycemia were also increased. The risk of stillbirth was elevated in women with CD, particularly among those with flaring disease (aOR, 4.48; 95% CI, 1.67–11.90). Thiopurine exposure increased risks for preterm birth, both in women with stable (aOR, 2.41; 95% CI, 1.05–5.51) and with flaring disease (aOR, 4.90; 95% CI, 2.76–8.69). Conclusions: Women with UC and CD are at increased risk of adverse birth outcomes, such as stillbirth, growth restriction, and preterm birth, particularly when they suffer from flares throughout pregnancy. Thiopurine exposure seems to further increase risks, independently of disease activity. (Inflamm Bowel Dis 2014;20:1091–1098) Key Words: pregnancy, Crohn’s disease, ulcerative colitis
I
nflammatory bowel disease (IBD) includes ulcerative colitis (UC) and Crohn’s disease (CD). These chronic conditions often affect women in their reproductive years, and prevalence is rising.1 Questions about consequences of IBD and its associated medical therapy for the offspring are often brought up when physicians are counseling women of childbearing age. A higher frequency of voluntary childlessness among women with IBD has been observed, and concerns for fetal exposure to medical therapy seem to be part of the reason.2,3 The natural course of UC and CD involves fluctuating disease activity, where periods of clinical remission are interrupted
Received for publication March 17, 2014; Accepted March 31, 2014. From the *Unit of Clinical Epidemiology, Centre for Pharmacoepidemiology, Karolinska Institutet, Stockholm, Sweden; and †Department of Women’s and Children’s Health, Karolinska University Hospital and Institutet, Stockholm, Sweden. Centre for Pharmacoepidemiology receives financial support from several pharmaceutical companies, including Janssen Biotech. This study was independently developed from a project initiated by Janssen Biotech. M. Elmberg has been a consultant (none above 1000 USD) for Tillotts Pharma, AbbeVie, Ferring. Grant support (not above 1500 USD) from MSD. The remaining authors have no conflicts of interest to disclose. Reprints: Gabriella Bröms, MD, Centre for Pharmacoepidemiology, Building T2, Karolinska University Hospital, Stockholm 171 76, Sweden (e-mail: gabriella. [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000060 Published online 7 May 2014.
Inflamm Bowel Dis Volume 20, Number 6, June 2014
by flares of increased bowel inflammation. Management of IBD aims at maintaining remission, which often requires medical treatment. Typically, systemic 5-aminosalicylates, sulfasalazine, thiopurines, and TNFa-inhibitors are used for this purpose. Flares are usually treated with corticosteroids, rectal administration of 5-aminosalicylates, or bowel surgery. Previous studies have addressed birth outcomes in women with IBD and in women with UC and CD separately; however, results remain conflicting. Increased risks of preterm birth and low birth weight have been reported for both UC and CD.4–6 A higher proportion of infants born small for gestational age (SGA) has been observed in some studies, whereas others have failed to show such an association.6–9 Distinguishing effects of medical therapy and effects of disease activity remain a challenge. The aim of this study was to assess risks of adverse birth outcomes in women with UC and CD by disease activity and drug exposure.
MATERIALS AND METHODS We conducted a population-based cohort study in which all singleton births identified in the Medical Birth Register (MBR) between July 1, 2006 and December 31, 2010, were included. The personal identification number assigned to all residents in Sweden enabled merging of data between the MBR, the National Patient Register, and the Prescribed Drug Register on an individual level. www.ibdjournal.org |
1091
Bröms et al
Women with IBD starting from 1997 and before childbirth were identified using the National Patient Register that holds nationwide information on both inpatient and outpatient hospital visits (outpatient visits from 2001 and onwards). Data for each visit include discharge diagnoses based on the International Classification of Diseases, Tenth Revision (ICD-10), date of admission, and discharge and codes for surgical procedures. We used ICD-codes K51 for UC and K50 for CD, respectively. In cases where the patient had received both a diagnosis of UC and of CD, the most recent diagnosis was used. To be considered as having IBD, patients were required to have had at least 2 recorded diagnoses of UC or CD or 1 visit or hospitalization in combination with a medical or surgical treatment for IBD. The National Patient Register was also used to obtain data on comorbidities within 5 years before pregnancy including diabetes (E10-E11, O240-O241), hypertension (I10-I15, O10), chronic autoimmune disease (primary sclerosing cholangitis K83; primary biliary cirrhosis K743; rheumatoid arthritis M05-06; vasculitis, SLE and other systemic disease M30-35; and psoriasis L40), and asthma (J45). Filled drug prescriptions are recorded in the Prescribed Drug Register by date of dispensing, drug name, and preparation according to the Anatomical Therapeutic Chemical classification. In general, prescriptions are filled for a maximum of 3 months. Exposure to drugs used in the treatment of IBD was defined as having filled at least 1 prescription of sulfasalazine (A07EC01), aminosalicylates (A07EC02-03), and corticosteroids for systemic use: betamethasone, methylprednisolone, prednisolone, prednisone, hydrocortisone (H02AB), rectal prednisolone (A07EA01), rectal hydrocortisone (A07EA02), rectal budesonide (A07EA06), azathioprine (L04AX01), 6-mercaptopurine (L01BB02), infliximab (L0AB02), and adalimumab (L04AB04) during pregnancy. The MBR includes information from standardized records that are completed by physicians and midwives throughout the antenatal and perinatal care program. We used the variables of maternal age when giving birth, parity, smoking status, and body mass index in early pregnancy to establish maternal characteristics. Adverse birth outcomes were assessed as rates and risks of preterm birth (,37 wk of gestation), spontaneous preterm birth, preterm premature rupture of the membranes, low birth weight (,2500 g), SGA, Apgar score ,7 at 5 minutes, stillbirth, hypoglycemia, and neonatal hyperbilirubinemia. For SGA, we assessed birth weights by gender ,2 SD below the mean and birth weights in the lower 10th percentile. Stillbirth was classified as intrauterine death occurring from gestational week 28 and onwards for births between 2006 and 2007. From 2008, stillbirths from gestational week 22 and onwards were recorded in the MBR and thus included in the cohort. In post hoc analyses, we tested whether restricting the analyses to only include stillbirths occurring from gestational age 28 weeks and onwards would influence the risk estimates for stillbirth. Spontaneous preterm birth was defined as spontaneous start of labor excluding induction or elective cesarean delivery. Preterm premature rupture of the membranes was defined by ICD-10 code O42.
1092
| www.ibdjournal.org
Inflamm Bowel Dis Volume 20, Number 6, June 2014
We used logistic regression to analyze risks of each birth outcome in infants to women with UC and CD, respectively, by comparing them with infants to women without these diseases. Risks were computed as crude and adjusted odds ratios (aORs) with 95% confidence intervals (CI). Women with UC and with CD were categorized by disease activity according to a proxy model as women with: (1) No disease activity, no medication, surgery, or hospitalization, (2) Stable disease: systemic 5-aminosalicylates, sulfasalazine, thiopurines, or TNFa-inhibitors throughout pregnancy, or (3) Flaring disease: corticosteroids, rectal 5-aminosalicylates after endoscopic examination, surgery, or hospital admission with IBD as primary diagnosis. For women with flares, we assessed risks of adverse outcome by timing of the flares as early, late, or both. Women with flares occurring from 3 months before conception and during the first trimester of pregnancy only were classified as having early pregnancy flares. Late pregnancy flares were those occurring in the second or third trimester. For women with stable or flaring disease, we also assessed risks of adverse birth outcomes associated with exposure to thiopurines. Adjustments were made for maternal age, parity, smoking, body mass index, and comorbidity (a diagnosis of diabetes, hypertension, chronic autoimmune disease, or asthma). We also performed analyses stratified by smoking status. Because observations were not independent in women who gave birth more than once during the study period, we used the generalized estimation equation method for clustered data. All analyses were conducted using SAS software (version 9.3; SAS Institute Inc, Cary, NC). The proxy model for disease activity was validated by scrutinizing medical charts of women with UC or CD from the cohort who gave birth in 2006. To be eligible, they had records of either at least 1 inpatient visit or 2 or more outpatient visits because of IBD during pregnancy implying flaring disease. The charts of 88 women were thoroughly examined concerning information on clinical events indicating disease activity. Frequent bowel movements, blood or mucus in stools, abdominal pain, and fever indicated flaring disease. Assessment was performed by one of the authors (G.B.) who was blinded to our disease activity proxy.
Ethical Considerations The study was conducted in accordance to ethical review by the Regional Ethical Review Board, Stockholm (Record number 2006/5:7).
RESULTS We included 1833 singleton births to 1492 women with UC and 1220 births to 1026 women with CD. All 467,057 births to women without UC or CD occurring during the study period constituted the comparison group. Women with UC and with CD were somewhat older and those with CD were more often smokers and had lower body mass index (Table 1). Among women with UC, there was 30% with no disease activity, 34% with stable disease, and 36% with flaring disease.
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Births Outcomes in Inflammatory Bowel Disease
TABLE 1. Descriptive Data on Singleton Births Among Women with and Without IBD
Maternal age at birth, yr #24 25–34 $35 Missing Parity Primiparous Multiparous Smoking in first trimester Nonsmoker Smoker Missing Body mass index #19 20–24 25–29 $30 Missing Comorbidity prepregnancya Yes Disease activity during pregnancy No disease activity Stable disease Flaring disease Timing of flare Early pregnancy Late pregnancy Early and late pregnancy TNFa-inhibitor exposure Thiopurine exposureb Yes No
UC N ¼ 1833, n (%)
CD N ¼ 1220, n (%)
No IBD N ¼ 467,057, n (%)
143 (7.80) 1256 (68.52) 434 (23.68) —
131 (10.74) 813 (66.64) 276 (22.62) —
68,592 (14.69) 296,334 (63.45) 102,128 (21.87) 3
827 (45.12) 1006 (54.88)
545 (44.67) 675 (55.33)
210,404 (45.05) 256,653 (54.95)
1703 (92.91) 59 (3.22) 71 (3.87)
1058 (86.72) 122 (10.00) 40 (3.28)
416,556 (89.19) 31,043 (6.65) 19,458 (4.17)
177 (9.66) 954 (52.05) 423 (23.08) 138 (7.53) 141 (7.69)
148 (12.13) 575 (47.13) 286 (23.44) 110 (9.02) 101 (8.28)
42,724 225,457 108,005 51,453 39,418
198 (10.80)
167 (13.69)
31,434 (6.73)
542 (29.57) 625 (34.10) 666 (36.33)
551 (45.16) 347 (28.44) 322 (26.39)
— — —
207 247 212 6
113 92 117 19
— — — —
159 1132
262 407
— —
(9.15) (48.27) (23.12) (11.02) (8.44)
a
Diabetes, hypertension, autoimmune disease, and asthma. Thiopurine exposure among women with stable and flaring disease.
b
The corresponding proportions for CD were 45%, 28%, and 26%, respectively. There were 438 women with UC and CD who gave birth more than once during the study period, and 66% of these had similar disease activity during their pregnancies. There were 109 women eligible for the validating chart review. We were able to collect charts of 88 women, corresponding to a response rate of 81%. Of these, 59% had signs of flaring disease from 3 months before and throughout pregnancy, and 41% had stable disease or no disease activity. When we compared our proxy model with the medical records, the sensitivity for flaring disease was 83% and the specificity was 81%.
Risks of adverse birth outcomes in association with UC or CD are shown in Table 2 and risks by disease activity in Table 3. Preterm births were more common (8.4%) in women with UC or CD as compared with women without IBD (5.0%). The aORs for preterm births were 1.78 (95% CI, 1.49–2.13) for UC and 1.65 (95% CI, 1.33–2.06) for CD, respectively. For CD, restricting the analyses to spontaneous preterm birth only, the aOR was 1.43 (95% CI, 1.07–1.91). The highest risks for preterm birth and for spontaneous preterm birth were found among those with flaring disease. Analyses of low birth weight exhibited a similar pattern with risks for UC in general at aOR 1.64 (95% CI, 1.31–2.04) and www.ibdjournal.org |
1093
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
TABLE 2. Birth Outcomes in Women with UC and CD Compared with Women Without IBD UC, N (%) Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
153 71 39 99 54 196 6 41 46 86
(8.35) (3.87) (2.13) (5.40) (2.95) (10.69) (0.33) (2.24) (2.51) (4.69)
CD, N (%) Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
103 57 28 78 49 127 11 30 33 60
(8.44) (4.67) (2.30) (6.39) (4.02) (10.41) (0.90) (2.46) (2.70) (4.92)
No IBD, N (%)
23,492 14,689 6548 16,014 10,954 46,637 1558 8960 9445 17,858
No IBD, N (%)
23,492 14,689 6548 16,014 10,954 46,637 1558 8960 9445 17,858
for CD at aOR 1.86 (95% CI, 1.46–2.38). For flaring UC, the risk estimate in the adjusted analysis was doubled, and for CD, it was tripled. Women with CD had increased risks of SGA births, with the highest risks among those with flaring disease. For flaring CD, there were increased risks of low 5-minute Apgar score (aOR, 2.21; 95% CI, 1.24–3.93) and hypoglycemia (aOR, 2.38; 95% CI, 1.40–4.04). Systemic corticosteroid treatment was more common in women whose infants had hypoglycemia, 29% versus 18%. When adjusting for SGA, risks for hypoglycemia in infants born to women with flaring CD was lowered (aOR, 1.97; 95% CI, 1.11–3.50). For women with UC, an increased risk of hyperbilirubinemia was observed in women with flaring disease (aOR, 1.58; 95% CI, 1.12–2.22), but the risk increase did not persist when adjusting for preterm birth. Risks of stillbirth were increased in women with CD, with risk estimates rising across disease activity groups to an aOR in flaring disease of 4.48 (95% CI, 1.67–11.90). These results were based on 11 cases among births to women with CD. Six of these were preterm. One was SGA below 2 SD, and 4 were SGA below the 10th percentile. Restricting the analyses to only include stillbirths occurring in gestational week 28 or later and stratifying for smoking had only minor effects on the risk estimates (results not shown in tables).
1094
| www.ibdjournal.org
(5.03) (3.15) (1.40) (3.43) (2.35) (9.99) (0.33) (1.92) (2.02) (3.82)
(5.03) (3.15) (1.40) (3.43) (2.35) (9.99) (0.33) (1.92) (2.02) (3.82)
Crude OR (95% CI)
1.79 1.35 1.67 1.63 1.23 1.08 1.17 1.16 1.19 1.25
(1.50–2.14) (1.06–1.73) (1.21–2.30) (1.31–2.03) (0.92–1.64) (0.92–1.27) (0.52–2.61) (0.84–1.62) (0.87–1.63) (1.00–1.57)
Crude OR (95% CI)
1.76 1.48 1.60 2.02 1.86 1.04 2.98 1.33 1.45 1.35
(1.42–2.19) (1.11–1.98) (1.07–2.40) (1.58–2.58) (1.37–2.52) (0.85–1.26) (1.60–5.54) (0.91–1.94) (1.01–2.09) (1.03–1.77)
aOR (95% CI)
1.78 1.37 1.63 1.64 1.26 1.11 1.22 1.19 1.16 1.26
(1.49–2.13) (1.07–1.75) (1.18–2.25) (1.31–2.04) (0.95–1.69) (0.94–1.30) (0.54–2.71) (0.85–1.65) (0.85–1.59) (1.01–1.58)
aOR (95% CI)
1.65 1.43 1.49 1.86 1.72 0.97 2.93 1.30 1.32 1.32
(1.33–2.06) (1.07–1.91) (1.00–2.25) (1.46–2.38) (1.27–2.34) (0.80–1.18) (1.57–5.47) (0.88–1.88) (0.89–1.91) (1.00–1.74)
For women with UC or with CD, the risks of adverse birth outcomes were, with some exceptions, higher when flares occurred in late pregnancy than in early pregnancy. The highest risks were in women with flares in both early and late pregnancy (Table 4). In general, women with flares treated with thiopurines had the highest risks of preterm birth and SGA. The influence of thiopurines on risks of preterm birth seemed to be independent of disease activity because higher risks were found both among women with stable disease and those with flaring disease (Table 5).
DISCUSSION In this study, we observed increased risks of preterm birth and low birth weight among women with UC and CD. Adverse birth outcomes were, generally associated with flaring disease during pregnancy, particularly in women with CD, who had a 4-fold increased risk of stillbirth. For preterm birth and the growth-related outcomes, we found strong associations with both disease activity and with thiopurine treatment.
Stillbirth Elevated risks of stillbirth in women with IBD have, to our knowledge, not been reported previously.5 In our previous studies
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Births Outcomes in Inflammatory Bowel Disease
TABLE 3. Birth Outcomes by Disease Activity: aORs, Compared with Women Without IBD No Disease Activity UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
30 15 6 20 12 42 3 11 15 19
aOR (95% CI)
1.30 1.06 0.90 1.28 1.12 0.91 2.02 1.24 1.52 1.06
(0.89–1.90) (0.61–1.83) (0.40–2.02) (0.82–2.01) (0.63–1.99) (0.66–1.25) (0.65–6.28) (0.68–2.25) (0.90–2.55) (0.67–1.68)
Stable Disease N
35 18 8 27 15 67 2 13 10 24
No Disease Activity CD
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth PPROM Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile Stillbirth Apgar ,7 at 5 min Hypoglycemia Hyperbilirubinemia
26 14 4 21 14 43 3 10 7 25
aOR (95% CI)
1.05 0.92 0.57 1.24 1.18 0.84 1.95 1.08 0.66 1.36
(0.70–1.58) (0.54–1.57) (0.21–1.52) (0.79–1.97) (0.67–2.10) (0.61–1.15) (0.63–6.05) (0.58–2.00) (0.31–1.41) (0.90–2.07)
including women with UC and CD giving birth in Denmark and Sweden, we found no increased risks of stillbirth, except for women with UC who had undergone previous surgery.8,9 Different time periods and different approaches to identify women with IBD may to some extent explain these inconsistent findings. With increasing opportunities to treat CD over time and the stricter definition of IBD, requiring at least 2 recorded diagnoses or 1 diagnosis together with an adjacent medical or surgical treatment instead of 1 diagnosis, it is possible that a higher proportion of women included in this study had a more severe disease. That women with flaring disease had the highest risks for stillbirth supports that disease severity might have had a major impact on the results. Also, in the previous study, we only included primiparous women and although we, in this study, adjusted for parity and performed cluster analyses, factors relating to parity might still have influenced the results. This study is, however, more representative of pregnant women in general, than our previous study. Although the age limit for reporting intrauterine fetal deaths as stillbirth to the MBR changed in 2008 from gestational week 28 to 22, post hoc analyses including stillbirths from gestational week 28 had only minor effects on our risk estimates for stillbirth. In a study
aOR (95% CI)
1.26 1.03 0.98 1.45 1.15 1.20 1.18 1.24 0.84 1.10
(0.89–1.79) (0.65–1.65) (0.49–1.97) (0.98–2.14) (0.68–1.92) (0.93–1.56) (0.29–4.73) (0.72–2.15) (0.45–1.57) (0.73–1.65)
Flaring disease N
75 35 24 41 21 71 1 12 15 36
Stable Disease N
26 15 8 16 12 24 3 5 9 16
aOR (95% CI)
1.68 1.55 1.73 1.47 1.55 0.74 3.06 0.84 1.30 1.30
(1.12–2.52) (0.92–2.61) (0.85–3.53) (0.89–2.43) (0.87–2.75) (0.49–1.11) (1.00–9.49) (0.35–2.01) (0.66–2.55) (0.78–2.16)
aOR (95% CI)
2.72 1.92 2.82 2.10 1.49 1.17 0.57 1.09 1.19 1.58
(2.12–3.48) (1.37–2.70) (1.87–4.26) (1.51–2.90) (0.96–2.31) (0.91–1.50) (0.08–4.06) (0.62–1.94) (0.71–1.99) (1.12–2.22)
Flaring Disease N
39 20 12 34 21 47 4 12 16 15
aOR (95% CI)
2.66 2.13 2.74 3.30 2.75 1.46 4.48 2.21 2.38 1.28
(1.89–3.74) (1.35–3.37) (1.53–4.92) (2.29–4.74) (1.72–4.38) (1.06–2.02) (1.67–11.9) (1.24–3.93) (1.40–4.04) (0.76–2.15)
from the United States, Mahadevan et al10 reported increased risks of adverse birth outcomes in women with IBD, when combining stillbirth, SGA, and preterm birth as one entity, although no separate risk estimates were calculated for each outcome. The rate of stillbirth in the study by Mahadevan and in an Italian study by Bortoli was around 1%, which is in accordance with what we observed for women with CD.10,11 The observed increase in risks of stillbirth across disease activity groups in CD suggests that the extent of the inflammation is important. Whether the association is linked to the disease, the treatment, or both remains, however, unclear. Because of the limited sample size, we could not assess the impact of thiopurine exposure on stillbirth. Infection is an important risk factor for stillbirth, and because bowel inflammation and immunosuppressive therapy may increase the sensitivity to infections, it is plausible that a secondary infection could be the underlying cause to the increased risk of stillbirth observed here.12 It is important to note that although the relative risks were clearly elevated, stillbirth is a rare event in developed countries.13 Nevertheless, the observation suggests that women with CD require close surveillance during pregnancy and delivery. www.ibdjournal.org |
1095
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
TABLE 4. Birth Outcomes by Timing of Flare Indicator, Compared with Women Without IBD Activity in Early Pregnancy UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile
22 11 9 7 18
Activity in Late Pregnancy
aOR (95% CI)
2.60 1.96 1.46 1.59 0.92
N
(1.66–4.07) (1.07–3.61) (0.75–2.83) (0.75–3.41) (0.56–1.50)
aOR (95% CI)
23 9 15 7 30
2.12 1.27 1.98 1.28 1.32
Activity in Late Pregnancy
CD
N
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2 SD SGA, below 10th percentile
7 5 6 5 7
1.26 1.48 1.54 1.79 0.55
(0.58–2.72) (0.60–3.65) (0.67–3.50) (0.75–4.27) (0.26–1.17)
N
(1.38–3.26) (0.66–2.46) (1.18–3.35) (0.60–2.72) (0.89–1.95)
Activity in Early Pregnancy aOR (95% CI)
Activity in Both Early and Late Pregnancy
30 15 17 7 23
2.41 1.55 3.96 4.03 2.22
3.62 2.71 2.90 1.65 1.25
(2.40–5.45) (1.60–4.58) (1.70–4.95) (0.77–3.53) (0.81–1.94)
Activity in Both Early and Late Pregnancy
aOR (95% CI)
10 4 11 8 17
aOR (95% CI)
N
(1.23–4.73) (0.60–4.24) (2.07–7.55) (1.94–8.38) (1.31–3.78)
aOR (95% CI)
22 11 17 8 22
4.43 3.23 4.67 2.79 1.95
(2.76–7.11) (1.73–6.03) (2.77–7.89) (1.32–5.87) (1.21–3.15)
TABLE 5. Birth Outcomes by Thiopurine Exposure No Thiopurine Exposure Stable Disease UC
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2SD SGA, below 10th percentile
28 14 25 15 65
aOR (95% CI)
1.12 0.90 1.52 1.30 1.34
(0.77–1.65) (0.53–1.54) (1.01–2.28) (0.78–2.18) (1.03–1.75)
Thiopurine Exposure
Flaring Disease N
58 24 32 16 60
aOR (95% CI)
2.40 1.51 1.88 1.30 1.15
(1.83–3.16) (1.01–2.28) (1.32–2.69) (0.79–2.14) (0.87–1.50)
Stable Disease N
aOR (95% CI)
N
7 4 2 0 2
2.41 (1.05–5.51) 2.07 (0.79–5.45) 0.91 (0.24–3.47) — 0.26 (0.06–1.09)
17 11 9 5 11
No Thiopurine Exposure Stable Disease CD
N
Adverse birth outcomes Preterm birth ,37 wk Spontaneous preterm birth Low birth weight ,2500 g SGA, below 2SD SGA, below 10th percentile
9 5 7 9 15
1096
| www.ibdjournal.org
aOR (95% CI)
1.09 1.01 1.26 2.41 0.98
(0.56–2.15) (0.41–2.48) (0.59–2.69) (1.23–4.72) (0.54–1.65)
25 15 20 11 29
aOR (95% CI)
2.25 2.18 2.53 1.88 1.17
aOR (95% CI)
4.90 4.64 3.50 1.58 1.30
(2.76–8.69) (2.48–8.68) (1.63–7.54) (1.04–6.41) (0.69–2.47)
Thiopurine Exposure
Flaring Disease N
Flaring Disease
(1.48–3.42) (1.29–3.69) (1.59–4.02) (0.97–3.62) (0.77–1.76)
Stable Disease N
17 10 9 3 9
aOR (95% CI)
2.34 2.12 1.70 0.74 0.51
(1.40–3.90) (1.12–4.03) (0.88–3.30) (0.24–2.32) (0.26–1.02)
Flaring Disease N
14 5 14 10 14
aOR (95% CI)
3.92 2.01 5.74 5.52 2.39
(2.19–7.01) (0.81–5.00) (3.18–10.4) (2.83–10.80) (1.42–4.03)
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Preterm Birth Being born preterm is associated with significant morbidity.14 Increased risks of preterm birth among women with IBD in general, and UC and CD separately, have been reported previously.5,6,8,9,15 In contrast to this study, none of the previous studies did, however, specifically assess spontaneous preterm birth. We found increased risks for spontaneous preterm birth, with the risk estimates being only slightly lower than those found when assessing preterm birth in general, which suggests that other than iatrogenic factors might explain the previously reported associations. Inflammation is part of the physiology of birth and infection, subclinical, or clinical, is often present in spontaneous preterm birth.16,17 Accordingly, both the immunological similarities between IBD and the initiation and progression of labor, and the increased risks of a secondary infection in women with IBD could pose possible explanations to the association between IBD and preterm birth.18
Fetal Growth and Hypoglycemia Low birth weight has been reported to be more prevalent in women with IBD.5 However, low birth weight is a less accurate measure of fetal growth than SGA. SGA might be a sign of intrauterine growth restriction and is associated with adverse perinatal outcomes and long-term morbidity.19 In this study, 2 cutoffs for SGA were used—birth weights below 2 SD according to fetal gender and birth weights below the 10th percentile. For CD, we found increased risks of SGA for the stricter cutoff, whereas for UC, there was no increased risk. In our previous study, risks were less elevated, whereas an American study reported a 2-fold increase in risks and presumably most of these women exhibited some disease activity.6,8 Hypoglycemia is the most common metabolic disorder in the neonatal period and growth-restricted infants are at a higher risk.20 We observed increased risks of hypoglycemia in women with flaring CD. Being SGA confounded this result, although the risk remained increased even after adjustment for SGA. Hypoglycemia has been linked to intrauterine growth restriction, and the higher risks of neonatal hypoglycemia and of SGA suggest an association between intrauterine growth restriction and CD.21 The associations between SGA and hypoglycemia and CD should be investigated in future studies.
Thiopurine Exposure and Timing of Flare Treatment with thiopurines was generally associated with increased risks of both growth restriction and preterm birth. The risk estimates associated with thiopurine treatment were mostly higher than the estimates for no treatment, both among women with stable and flaring disease, although the CIs overlapped. The highest risk estimates were found for flaring disease. The increased risks of preterm birth associated with thiopurines have been observed in some previous studies but not in others.22–25 Because thiopurine treatment had been used in both stable and flaring disease, the impact of the drug could be evaluated indirectly irrespective of disease activity, and our data suggest that disease activity mainly seems to affect fetal growth, whereas thiopurine treatment might be associated with preterm birth.
Births Outcomes in Inflammatory Bowel Disease
In this study, we also assessed risks according to timing of flare activity. Having flare events in both early and late pregnancy was associated with 3-fold risk increases of preterm birth and low birth weight for UC and a 4-fold increase for CD. Flare in late pregnancy was associated with higher observed risk estimates than activity in early pregnancy. To our knowledge, the impact of timing of flare activity on adverse birth outcomes has not been reported previously.
Strengths and Limitations This study includes one of the largest study samples of women with UC and CD assessing risks of adverse birth outcomes and one of the few to include data on clinical visits and pharmacological treatment. The population-based design generates high generalizability for women with UC and CD seen at hospitals. Medical care in Sweden is public, and data in the national health registers have been found to be accurate and of high validity.26–28 A limitation is the lack of first-hand clinical data such as frequency of bowel movements, blood in stool, and biochemical markers of inflammation, which would have further improved the assessment of disease activity during pregnancy. In line with previous studies, we used data on clinical events and drug prescriptions, which are available in the national health registers, as proxies for disease activity.9,29 The proxy model was validated by scrutinizing a sample of medical charts and found to have high sensitivity and specificity for flaring disease. We were not able to assess the actual intake of drugs, but it should be noted that the agreement between self-reported use and 1 or more filled prescriptions for drugs used for IBD during pregnancy has been found to be high.30
CONCLUSIONS Women with UC and CD had increased risks of adverse birth outcomes, such as low birth weight and preterm birth. Women with CD also had increased risks of stillbirth. For most of the evaluated outcomes, the risk increase was found among women with flaring disease during pregnancy. High disease activity throughout pregnancy and thiopurine treatment denoted further increased risks of adverse birth outcome. The results confirm that maintaining remission and obstetric surveillance is imperative, but the optimal treatment is yet to be established in further comparative studies.
ACKNOWLEDGMENTS Author contributions: G. Bröms had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. G. Bröms, F. Granath, O. Stephansson, M. Elmberg, and H. Kieler conceived and designed the study. G. Bröms and H. Kieler drafted the manuscript. G. Bröms and M. Linder were responsible for the statistical analysis. All authors participated in interpreting the data and critically revising the manuscript. www.ibdjournal.org |
1097
Inflamm Bowel Dis Volume 20, Number 6, June 2014
Bröms et al
REFERENCES 1. Molodecky NA, Soon IS, Rabi DM, et al. Increasing incidence and prevalence of the inflammatory bowel diseases with time, based on systematic review. Gastroenterology. 2012;142:46 e42–54 e42; quiz e30. 2. Mountifield R, Bampton P, Prosser R, et al. Fear and fertility in inflammatory bowel disease: a mismatch of perception and reality affects family planning decisions. Inflamm Bowel Dis. 2009;15:720–725. 3. Selinger CP, Eaden J, Selby W, et al. Inflammatory bowel disease and pregnancy: lack of knowledge is associated with negative views. J Crohns Colitis. 2013;7:e206–e213. 4. Norgard B, Fonager K, Sorensen HT, et al. Birth outcomes of women with ulcerative colitis: a nationwide Danish cohort study. Am J Gastroenterol. 2000;95:3165–3170. 5. Cornish J, Tan E, Teare J, et al. A meta-analysis on the influence of inflammatory bowel disease on pregnancy. Gut. 2007;56: 830–837. 6. Dominitz JA, Young JC, Boyko EJ. Outcomes of infants born to mothers with inflammatory bowel disease: a population-based cohort study. Am J Gastroenterol. 2002;97:641–648. 7. Kornfeld D, Cnattingius S, Ekbom A. Pregnancy outcomes in women with inflammatory bowel disease–a population-based cohort study. Am J Obstet Gynecol. 1997;177:942–946. 8. Stephansson O, Larsson H, Pedersen L, et al. Crohn’s disease is a risk factor for preterm birth. Clin Gastroenterol Hepatol. 2010;8:509–515. 9. Stephansson O, Larsson H, Pedersen L, et al. Congenital abnormalities and other birth outcomes in children born to women with ulcerative colitis in Denmark and Sweden. Inflamm Bowel Dis. 2011;17:795–801. 10. Mahadevan U, Sandborn WJ, Li DK, et al. Pregnancy outcomes in women with inflammatory bowel disease: a large community-based study from Northern California. Gastroenterology. 2007;133:1106–1112. 11. Bortoli A, Saibeni S, Tatarella M, et al. Pregnancy before and after the diagnosis of inflammatory bowel diseases: retrospective case-control study. J Gastroenterol Hepatol. 2007;22:542–549. 12. Smith GC, Fretts RC. Stillbirth. Lancet. 2007;370:1715–1725. 13. Cousens S, Blencowe H, Stanton C, et al. National, regional, and worldwide estimates of stillbirth rates in 2009 with trends since 1995: a systematic analysis. Lancet. 2011;377:1319–1330. 14. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet. 2008;371:261–269. 15. Nguyen GC, Boudreau H, Harris ML, et al. Outcomes of obstetric hospitalizations among women with inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol. 2009;7:329–334.
1098
| www.ibdjournal.org
16. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med. 2000;342:1500–1507. 17. Romero R, Espinoza J, Goncalves LF, et al. Inflammation in preterm and term labour and delivery. Semin Fetal Neonatal Med. 2006;11:317–326. 18. Nasef NA, Ferguson LR. Inflammatory bowel disease and pregnancy: overlapping pathways. Transl Res. 2012;160:65–83. 19. Longo S, Bollani L, Decembrino L, et al. Short-term and long-term sequelae in intrauterine growth retardation (IUGR). J Matern Fetal Neonatal Med. 2013;26:222–225. 20. Garg M, Devaskar SU. Glucose metabolism in the late preterm infant. Clin Perinatol. 2006;33:853–870; abstract ix-x. 21. Engineer N, Kumar S. Perinatal variables and neonatal outcomes in severely growth restricted preterm fetuses. Acta Obstet Gynecol Scand. 2010;89:1174–1181. 22. Francella A, Dyan A, Bodian C, et al. The safety of 6-mercaptopurine for childbearing patients with inflammatory bowel disease: a retrospective cohort study. Gastroenterology. 2003;124:9–17. 23. Norgard B, Pedersen L, Fonager K, et al. Azathioprine, mercaptopurine and birth outcome: a population-based cohort study. Aliment Pharmacol Ther. 2003;17:827–834. 24. Akbari M, Shah S, Velayos FS, et al. Systematic review and meta-analysis on the effects of thiopurines on birth outcomes from female and male patients with inflammatory bowel disease. Inflamm Bowel Dis. 2013;19: 15–22. 25. Coelho J, Beaugerie L, Colombel JF, et al. Pregnancy outcome in patients with inflammatory bowel disease treated with thiopurines: cohort from the CESAME Study. Gut. 2011;60:198–203. 26. Wettermark B, Hammar N, Fored CM, et al. The new Swedish Prescribed Drug Register–opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf. 2007;16:726–735. 27. Cnattingius S, Ericson A, Gunnarskog J, et al. A quality study of a medical birth registry. Scand J Soc Med. 1990;18:143–148. 28. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. 29. Norgard B, Pedersen L, Christensen LA, et al. Therapeutic drug use in women with Crohn’s disease and birth outcomes: a Danish nationwide cohort study. Am J Gastroenterol. 2007;102:1406–1413. 30. Stephansson O, Granath F, Svensson T, et al. Drug use during pregnancy in Sweden—assessed by the Prescribed Drug Register and the Medical Birth Register. Clin Epidemiol. 2011;3:43–50.
