Introduction

Emerging evidence suggests that neurologic manifestations in patients with inflammatory bowel diseases (IBD) are more common than previously estimated [1–3]. Literature data suggest that IBD patients can manifest demyelinating events in both central and peripheral nervous system [4], while the association between multiple sclerosis (MS) and IBD has not only been suggested by their common epidemiological and immunological patterns [5, 6], but also due to observations of increased incidence of both IBD among MS patients [7] and MS among IBD patients [8].

The aim of the present systematic review and meta-analysis is to estimate the cumulative risk of concurrent IBD and MS comorbidity, using data from all available case–control studies.

Methods

This meta-analysis has adopted the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic reviews and meta-analyses [9] and was written according to the Meta-analysis of Observational Studies in Epidemiology (MOOSE) proposal [10]. Eligible case–control study protocols that reported prevalence rates of MS concurrence in patients with IBD and controls, or vice versa IBD concurrence in MS patients and controls, were identified by searching MEDLINE and SCOPUS databases. The combination of search strings that was used in both database searches included the terms: “inflammatory bowel disease”, “Crohn’s Disease”, “Ulcerative Colitis”, and “Multiple Sclerosis”. No language or other restrictions were imposed. Last literature search was conducted on June 12th, 2016. The complete search algorithm used in MEDLINE search is available in the on-line only material. Reference lists of all articles that met the criteria and of relevant review articles were examined to identify studies that might have been missed by the database search. All retrieved studies were scanned independently by 2 reviewers (AHK and KHK), while in case of disagreement regarding the literature search results between the aforementioned two authors, the remaining coauthors were consulted and disagreement was resolved with consensus. We excluded from the quantitative/qualitative analysis all observational cohort studies (with no control population), case series, case reports, and studies either not reporting either data on prevalence disease rates or reporting data on forms of demyelination other than MS. Absolute or percent numbers of patients with MS and IBD concurrence reported in either IBD or MS registries together with the absolute or percent numbers of patients with either IBD or MS within the control population were extracted independently by the same authors who performed the literature search (AHK and KHK).

Statistical analyses

We calculated the corresponding Risk ratios (RRs) in each included case–control study to express the risk of IBD and MS concurrence at a given population, with RR values greater than 1 denoting an increased risk of IBD/MS concurrence when one of them is already present. A random-effects model (DerSimonian Laird) was used to calculate the pooled RRs. The equivalent z test was performed for each pooled RR, and if p < 0.05 it was considered statistically significant. After the main analysis, we performed additional subgroup analyses according to (1) the type of registry from which the data of the cases were exported (registry with IBD patients or registry with MS patients) and (2) IBD type (Crohn’s disease or Ulcerative colitis). In both subgroup analyses the mixed-effects model was used to calculate both the pooled point estimate in each subgroup and the overall estimates. According to the mixed-effects model, a random-effects model was first used to combine studies within each subgroup and then a fixed effect model was used to combine subgroups and estimate the overall effect. We assumed the study-to-study variance (tau-squared) to be the same for all subgroups. Tau-squared was first computed within subgroups and then pooled across subgroups.

We assessed heterogeneity between studies with the Cochran Q and I 2 statistics. For the qualitative interpretation of heterogeneity, I 2 values of at least 50% were considered to represent substantial heterogeneity, while values of at least 75% indicated considerable heterogeneity, as per the Cochrane Handbook [11]. We evaluated publication bias both graphically using a funnel plot and with the Egger’s statistical test for funnel plot asymmetry [12].

Statistical analyses were conducted using Review Manager (RevMan) Version 5.3 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and Comprehensive Meta-analysis Version 2 software (Borenstein M, Hedges L, Higgins J, Rothstein H, Biostat, Englewood NJ, 2005).

Results

Study selection and study characteristics

Systematic search of MEDLINE and SCOPUS databases yielded 960 and 867 results, respectively. After removing duplicates, the titles and abstracts from the remaining 1566 studies were screened and 16 potentially eligible studies for the meta-analysis were retained. After retrieving the full-text version of the aforementioned 16 studies, 6 studies were excluded (Supplemental Table I, on-line only material). In the final presentation of the literature search results, there was no conflict or disagreement between the two reviewers and the 10 studies that met the study protocol’s inclusion criteria were included both in the qualitative and quantitative synthesis (Fig. 1) [13–22]. The characteristics of the included studies, comprising 1,086,430 patients (0.08% of them with concurrent IBD and MS) are summarized in Table 1.

Fig. 1
figure 1

Flow chart presenting the selection of eligible studies

Full size image
Table 1 Included case–control studies in the meta-analysis
Full size table

Overall and subgroup analyses

Pooled RR for IBD/MS comorbitity was 1.54 (95% CI 1.40–1.67; p < 0.0001; Fig. 2) with neither evidence of heterogeneity across included studies (I 2 = 17%, p for Cochran Q statistic: 0.29), nor differences (p = 0.91) among IBD and MS registries (RR 1.53, 95% CI 1.36–1.72, p < 0.001 for MS comorbidity in IBD patients vs. RR 1.55, 95% CI 1.32–1.81, p < 0.001 for IBD comorbidity in MS patients). Additionally, both visual inspection of the funnel plot and the Egger’s statistical test (p = 0.203) revealed no evidence of publication bias (Supplemental Figure I, on-line only material).

Fig. 2
figure 2

Subgroup analysis on the risk of multiple sclerosis and inflammatory bowel disease comorbidity, according to the registry that provided the data

Full size image

In the subsequent subgroup analysis, no difference was found on the risk of MS comorbidity among patients with CD or UC (RR 1.52, 95% CI 1.34–1.72, p < 0.001 vs. RR 1.55, 95% CI 1.38–1.74, p < 0.001; p for subgroup differences: 0.84; Fig. 3).

Fig. 3
figure 3

Subgroup analysis on the risk of multiple sclerosis in patients with Crohn’s disease or ulcerative colitis compared to controls

Full size image

Discussion

We found that both IBD and MS patients have a fifty-percent increased risk of MS or IBD comorbidity, respectively, with no apparent differences between patients with CD or UC. Our results are on the same line with a recent systematic review on the topic of autoimmune comorbidity in IBD patients, reporting a higher prevalence of IBD in MS patients compared to the general population both before (range 0.10–1.6%) and after (range 0.36–4.66%) MS diagnosis [23]. Our systematic review and meta-analysis, by including 4 additional studies, also confirms the results of a previous meta-analysis on the risk of autoimmune disease in people with MS and their relatives, reporting an increased risk of IBD in patients with MS (OR 1.56, 95% CI 1.28–1.90, p < 0.0001) and no increase in the IBD risk in relatives of people with MS (OR 1.29, 95% CI 0.92–1.82, p = 0.14) [24].

Even though our systematic review and meta-analysis confirms the accumulating literature data by employing a comprehensive meta-analytical approach on available data from over 1 million patients and is consistent with the results from previous reports, several limitations should also be taken into consideration. First, all included studies are case–control study protocols from large-scale registries which inherently suffer from the possibility of insufficient matching between cases and controls. Even though, 8 out of 10 studies reported matching for controls on potential confounders (including age and sex) [13, 14, 16–19, 21, 22], none of the studies provided the baseline characteristics of cases and controls (Table 1). More importantly, treatment status and particularly the percentage of IBD patients under treatment with anti-TNFa drugs are not clearly presented in the included studies. Since, anti-TNFa agents have been implicated for drug-induced demyelination, the use of anti-TNFa drugs should be regarded as a potential confounder of the association between MS and IBD [4]. Second, because both IBD and MS are chronic inflammatory diseases with exacerbations and remissions, there is additional difficulty in identifying with certainty which of them preceded the other, and thus determining the true prevalence of one condition in the course of the other. In our meta-analysis, no difference was detected between IBD and MS registries in the rate of patients with IBD and MS comorbidity, suggesting that presumably in those patients both entities coexisted -one in a dormant stage- at the time of diagnosis of the symptomatic one. Finally, the diversity of methods of ascertainment for cases and controls both within the setting of an individual study and also between included studies may constitute another significant source of bias, especially in the diagnosis of MS, due to the revisions of Mc Donald’s diagnostic criteria in both 2005 [25] and 2010 [26].

In conclusion, due to the increased risk of MS and IBD comorbidity that was highlighted in the present meta-analysis regular neurological examination of IBD patients and routine gastroenterological monitoring of MS patients might be warranted.