The aim of this metaanalysis was to evaluate the risk of the development of obstetric complications in women with celiac disease. We searched electronic databases from their inception until February 2015. We included all cohort studies that reported the incidence of obstetric complications in women with celiac disease compared with women without celiac disease (ie, control group). Studies without a control group and case-control studies were excluded. The primary outcome was defined a priori and was the incidence of a composite of obstetric complications that included intrauterine growth restriction, small for gestational age, low birthweight, preeclampsia and preterm birth. Secondary outcomes included the incidence of preterm birth, intrauterine growth restriction, stillbirth, preeclampsia, small for gestational age, and low birthweight. The review was registered with PROSPERO (CRD42015017263) before data extraction. All authors were contacted to obtain the original databases and perform individual participant data metaanalysis. Primary and secondary outcomes were assessed in the aggregate data analysis and in the individual participant data metaanalysis. We included 10 cohort studies (4,844,555 women) in this metaanalysis. Four authors provided the entire databases for the individual participant data analysis. Because none of the included studies stratified data for the primary outcome (ie, composite outcome), the assessment of this outcome for the aggregate analysis was not feasible. Aggregate data analysis showed that, compared with women in the control group, women with celiac disease (both treated and untreated) had a significantly higher risk of the development of preterm birth (adjusted odds ratio, 1.35; 95% confidence interval, 1.09–1.66), intrauterine growth restriction (odds ratio, 2.48; 95% confidence interval, 1.32–4.67), stillbirth (odds ratio, 4.84; 95% confidence interval, 1.08–21.75), low birthweight (odds ratio, 1.63; 95% confidence interval, 1.06–2.51), and small for gestational age (odds ratio, 4.52; 95% confidence interval, 1.02–20.08); no statistically significant difference was found in the incidence of preeclampsia (odds ratio, 2.45; 95% confidence interval, 0.90–6.70). The risk of preterm birth was still significantly higher both in the subgroup analysis of only women with diagnosed and treated celiac disease (odds ratio, 1.26; 95% confidence interval, 1.06–1.48) and in the subgroup analysis of only women with undiagnosed and untreated celiac disease (odds ratio, 2.50; 95% confidence interval; 1.06–5.87). Women with diagnosed and treated celiac disease had a significantly lower risk of the development of preterm birth, compared with undiagnosed and untreated celiac disease (odds ratio, 0.80; 95% confidence interval, 0.64–0.99). The individual participant data metaanalysis showed that women with celiac disease had a significantly higher risk of composite obstetric complications compared with control subjects (odds ratio, 1.51; 95% confidence interval, 1.17–1.94). Our individual participant data concurs with the aggregate analysis for all the secondary outcomes. In summary, women with celiac disease had a significantly higher risk of the development of obstetric complications that included preterm birth, intrauterine growth restriction, stillbirth, low birthweight, and small for gestational age. Since the treatment with gluten-free diet leads to a significant decrease of preterm delivery, physicians should warn these women about the importance of a strict diet to improve obstetric outcomes. Future studies calculating cost-effectiveness of screening for celiac disease during pregnancy, which could be easily performed, economically and noninvasively, are needed. In addition, further studies are required to determine whether women with adverse pregnancy outcomes should be screened for celiac disease, particularly in countries where the prevalence is high.
Celiac disease is a genetically determined autoimmune condition, with an estimated worldwide prevalence of approximately 1%. It usually is diagnosed by duodenal biopsy that is performed at the time of endoscopy. Celiac disease is induced by the ingestion of gluten, and the only treatment available is the elimination of gluten from the diet.
Once considered a gastrointestinal disease of childhood, celiac disease is now recognized as a systemic disease. The most frequent signs and symptoms are weight loss and chronic diarrhea. Complications include disorders of fertility and pregnancy complications. Women with unexplained infertility or recurrent miscarriage were found to have a significantly higher risk of celiac disease than the general population, maybe because of the nutrient deficiencies and the increased level of serum autoantibodies. In 2014 a large epidemiologic study showed an increased risk for malformation among the offspring of mothers or fathers with celiac disease. Moreover, so far, the effect of a gluten-free diet on prevention of celiac disease complications in pregnancy is still a subject of debate.
The aim of this metaanalysis was to evaluate the risk of the development of obstetric complications in women with celiac disease.
Methods
Search strategy
This review was performed according to a protocol designed a priori and recommended for systematic review. Electronic databases (MEDLINE, PROSPERO, Scopus, ClinicalTrials.gov , EMBASE, Science direct, the Cochrane Library) were searched from their inception until February 2015 with no limit for language. The following search terms were used: “celiac,” “celiac disease,” “coeliac,” “coeliac disease,” “preterm birth,” “small for gestational age,” “miscarriage,” “pregnancy,” “premature,” “newborn,” “low birth weight,” “fertility,” “preeclampsia,” “recurrent,” “intrauterine growth restriction,” “stillbirth,” “pregnancy,” “obstetric,” “complications,” and “spontaneous preterm birth.” No restrictions for language or geographic location were applied. In addition, the reference lists of all identified articles were examined to identify studies not captured by electronic searches. The electronic search and the eligibility of the studies were assessed independently by 2 of the authors (G.S., V.B.). Differences were discussed, and consensus reached.
Study selection
We included all cohort studies that reported the incidence of obstetric complications in women with celiac disease compared with women without celiac disease (ie, control group). Studies without a control group and case-control studies were excluded.
Data extraction
Data abstraction was completed by 2 independent investigators (G.S., L.S.). Each investigator independently abstracted data from each study separately. Data from each eligible study were extracted without modification of original data onto custom-made data collection forms. Disagreements were resolved by consensus with a third reviewer (P.M.). Information on potential confounders that were adjusted for and adjusted risk estimates were collected when available. All authors were contacted to obtain the original databases and perform individual patient-level metaanalysis.
Two reviewers (G.S., V.B.) independently assessed the risk of bias of the included studies via the Methodological Index for Non-Randomized Studies. Seven domains that are related to risk of bias were assessed in each study: (1) aim (ie, clearly stated aim), (2) rate (ie, inclusion of consecutive patients and response rate), (3) data (ie, prospective collection of data), (4) bias (ie, unbiased assessment of study end points), (5) time (ie, follow-up time appropriate), (6) loss (ie, loss to follow-up), (7) size (ie, calculation of the study size). Review authors’ judgments were categorized as “low risk,” “high risk” or “unclear risk of bias.” Discrepancies were resolved by discussion with a third reviewer (P.M.).
The primary outcome was defined a priori and was the incidence of a composite of obstetric complications that included at least 1 of the following complications: intrauterine growth restriction (IUGR; ie, ultrasound estimated fetal weight <10th percentile for gestational age), small for gestational age (SGA; ie, birthweight <10th percentile for gestational age), low birthweight (LBW; ie, birthweight <2500 g), preeclampsia, and preterm birth (PTB; ie, PTB <37 weeks). Secondary outcomes included the incidence of PTB, IUGR, stillbirth, preeclampsia, SGA and LBW. We planned to assess the incidence of PTB in a subgroup analysis in women with treated and untreated celiac disease. Diagnosed and treated celiac disease thereafter is called, for simplicity, just “treated celiac disease”; and undiagnosed and untreated celiac disease is called, for simplicity, just “untreated celiac disease.” Women were classified as having the celiac disease diagnosis and treatment before pregnancy (treated celiac disease) or afterward (untreated celiac disease).
Data analysis
The data analysis was completed independently by 2 authors (G.S., V.B.) with the use of Review Manager software (version 5.3; The Nordic Cochrane Centre, Cochrane Collaboration, 2014, Copenhagen, Denmark) and Statistical Package for Social Sciences software (version 19.0; IBM Inc, Armonk, NY). Inconsistencies were discussed by the reviewers, and consensus was reached. Heterogeneity across studies was assessed using the Higgins I 2 test. In case of statistically significant heterogeneity (I 2 ≥0%), the random effects model of DerSimonian and Laird was used; otherwise, a fixed effect model was performed. The pooled results from the aggregate metaanalysis were reported as odds ratio (OR) with 95% confidence interval (CI). Potential publication biases were assessed graphically by the use of the funnel plot and statistically by the use of the Begg’s and Egger’s tests.
In line with other metaanalyses, no adjustment for risk estimates was made. For studies that reported both unadjusted and adjusted risk for confounders statistically proved, we performed an aggregate data metaanalysis using generic inverse variance method to obtain the adjusted odds ratio (aOR) for the incidence of PTB in the aggregate data analysis.
To use the data as best as possible, we also combined the obtained databases to assess an individual patient-level metaanalysis. Primary and secondary outcomes were assessed in both aggregate and patient-level data analysis. We expressed continuous variables as mean with standard deviation and categoric variables as number with percentage. Chi-square test and Fisher’s exact test were used for categoric variables, and the Student t test or Mann-Whitney test for normally and nonnormally distributed continuous variables, respectively. A probability value <.05 was considered statistically significant.
The metaanalysis was reported according to the Preferred Reporting Item for Systematic Reviews and Meta-analyses statement. Before data extraction, the review was registered with the PROSPERO International Prospective Register of Systematic Reviews (registration no.: CRD42015017263) according to the Preferred Reporting Item for Systematic Reviews and Meta-analyses guidelines for protocols.
Results
Study selection and study characteristics
The flow of study identification is shown in Figure 1 . Seventeen full-text articles were assessed for eligibility. Seven studies were excluded. Six studies were excluded because they evaluated the incidence of celiac disease among women with obstetric complications (ie, case-control studies), and one study was excluded because it was based on the same cohort as a more recent study.
Ten cohort studies, which included 4,844,555 women, were analyzed. All studies reported the incidence of obstetric complications in women with celiac disease compared with women without celiac disease (ie, control group). Four studies reported separate analyses and subgroup analysis for women with undiagnosed and untreated celiac disease (ie, untreated celiac disease). In all included studies, a diagnosis of celiac disease was based on either duodenal biopsy or level of serum autoantibodies.
The risk of publication bias was assessed by visual inspection of funnel plot; the symmetric plot suggested no publication bias ( Figure 2 ). Publication bias, which was assessed with the use of Begg’s and Egger’s tests, showed no significant bias ( P = .19 and P = .10, respectively). Table 1 shows the characteristics of the included studies. Most of them (9 of the 10) originated from Europe. Eight studies included only singleton gestations. The quality of the studies included in our metaanalysis was assessed by the Methodological Index for Non-Randomized Studies’ tool for assessment of the risk of bias ( Figure 3 ). All studies had low risk of bias in “aim” and most risk in “rate” and in “time.” Four of the them were retrospective cohort studies ; 3 studies were prospective ; the other 3 studies were large high-quality population-based cohort studies. Four authors kindly provided the entire databases from their study to obtain additional and unpublished data and to perform individual patient level metaanalysis.
| Study | Study location | Study period | Type of study | No. of included women a | Confounders adjusted | Outcomes assessed |
|---|---|---|---|---|---|---|
| Sher and Mayberry, 1996 | United Kingdom | 2005-2006 | Prospective cohort | 136 (68 vs 68) | None | Miscarriage, b stillbirth |
| Martinelli et al, 2000 | Italy | 1998-1999 | Prospective cohort | 218 (12 vs 206) | Maternal age, previous preterm birth, socioeconomic status, smoking | Preterm birth, b stillbirth |
| Greco et al, 2004 | Italy | 2001-2002 | Retrospective cohort | 5,076 (79 vs 4,997) | Maternal age, smoking, socio-economic status | Intrauterine growth restriction b |
| Tata et al, 2005 | United Kingdom | 1987-2002 | Retrospective cohort | 9,244 (1,521 vs 7,723) | Socio-economic status | Cesarean delivery, b stillbirth, preeclampsia, intrauterine growth restriction |
| Ludvigsson et al, 2005 | Sweden | 1964-2001 | Population-based cohort | 2,817,400 (2,071 vs 2,815,329) | Maternal age, parity, nationality | Preterm birth, intrauterine growth restriction, low birthweight c |
| Sheiner et al, 2006 | Israel | 1988-2002 | Retrospective cohort | 143,711 (48 vs 143,663) | None | Intrauterine growth restriction, b preeclampsia |
| McCarthy et al, 2009 | Ireland | N/R | Retrospective cohort | 270 (118 vs 152) | Maternal age, maternal body mass index, gestational age, infant sex and year of birth | Small for gestational age, b birthweight |
| Khashan et al, 2010 | Northern Europe | 1979-2004 | Population-based cohort | 1,504,342 (1,451 vs 1,502,891) | Maternal age, parity, paternal age, maternal chronic medical conditions | Preterm birth, b small for gestational age |
| Martinelli et al, 2010 | Italy | 2008 | Prospective cohort | 228 (49 vs 179) | None | Intrauterine growth restriction b |
| Abdul Sultan et al, 2014 | United Kingdom | 1997-2012 | Population-based cohort | 363,930 (892 vs 363,038) | Body mass index, smoking | Preterm birth, stillbirth, low birthweight, preeclampsia c |
| Total | — | — | — | 4,844,555 (6,309 vs 4,838,246) | — | — |
a Number (N) of included women: total N (N of women with celiac disease vs N of women with no celiac disease)
Synthesis of results
Because that none of the included studies stratified data for the primary outcome (ie, composite outcome), assessment of this outcome for the aggregate data analysis was not feasible.
Compared with the control group, women with celiac disease (both treated and untreated) had a significantly higher risk of the development of PTB (OR, 1.40; 95% CI, 1.18–1.6 [ Figure 4 ]; aOR, 1.35; 95% CI, 1.09–1.66; [ Figure 5 ]), IUGR (OR, 2.48, 95% CI, 1.32–4.67 [ Figure 6 ]), stillbirth (OR, 4.84; 95% CI, 1.08–21.75 [ Figure 7 ]), LBW (OR, 1.63; 95% CI, 1.06–2.51 [ Figure 8 ]), and SGA (OR, 4.52; 95% CI, 1.02–20.08 [ Figure 9 ]); no statistically significant difference was found in the incidence of preeclampsia (OR, 2.45; 95% CI, 0.90–6.70 [ Figure 10 ]).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
