The Association of Coeliac Disease and Microscopic Colitis

A Large Population-based Study

M. Stewart; C. N. Andrews; S. Urbanski; P. L. Beck; M. Storr


Aliment Pharmacol Ther. 2011;33(12):1340-1349. 

In This Article


This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary, and is in agreement with the Declaration of Helsinki.


The Calgary Health Region (CHR) is the sole provider of medical and surgical services to all of the over 1.2 million people in Calgary and the surrounding area.[17,18] The only patients who are referred for services provided outside of CHR are those requiring heart, lung or liver transplantation, although their pretransplant and post-transplant care is provided within the region. Over the past decade, Calgary has experienced significant economic and population growth, with its population expanding by nearly 27% in 10 years.[19] The CHR is comprised primarily of Caucasians (80%) with large Chinese and South Asian populations (6% and 5%, respectively). Over 270 000 residents report a first language other than English or French.[18]

Pathology Database

The CHR uses a centralised laboratory to process all histopathology specimens obtained within the region. All endoscopic biopsies within the CHR are sent to the CHR laboratory for microscopic assessment and diagnosis. Physicians submitting specimens are asked to provide specific diagnostic questions for the pathologist to address. Pathology reports describing the histological findings and the diagnostic opinion of the pathologist are generated from this laboratory and are stored in a searchable database.

Endoscopy Database

The vast majority of in-patient and out-patient endoscopic procedures are performed at one of four endoscopy units located at four separate hospitals within the CHR, one of which is a paediatric unit. A fifth dedicated colon cancer screening centre was opened in 2008. All endoscopy suites use the same report generating software (endopro; Pentax Medical, Montvale, NJ, USA) that is linked to a central database. All endoscopy entries contain basic patient information; however, additional information such as indications and endoscopic findings are not consistently entered by all endoscopists.

Diagnostic Criteria

The diagnosis of coeliac disease was based on the modified Marsh classification of histopathological features of coeliac disease. The earliest lesion (Marsh I-II) is characterised by normal villous architecture with increased intraepithelial lymphocytosis (>30 per 100 enterocytes), which is the most sensitive index of disease.[20] Hyperplasic crypts are seen in Marsh II lesions. Marsh III lesions are characterised by villous atrophy and sub-classified into three groups based on the severity of villous atrophy.[20] Response to a gluten-free diet was not used as a diagnostic criterion for the purposed of this study, as adult patients rarely undergo a postgluten free diet endoscopy. At the time of this study, there was no protocol in-place when patients should undergo duodenal biopsy.

Not all CHR pathology reports specified a Marsh stage. Reports in which the pathologist commented that the biopsy was consistent with, or diagnostic of CD were considered positive CD cases. In 34 cases, the pathologist's comments were indeterminate (i.e. Marsh stage I lesions) and medical records were reviewed to obtain serological confirmation of CD in the form of tissue transglutaminase antibodies and IgA status. Of these, 26 patients were excluded based on negative or unavailable serological markers and eight were included based on positive serological markers. Serological markers were not otherwise sought.

The diagnosis of MC was based on histopathological features of colonic biopsies that have been well described in the literature.[2,3,21] The criteria used to diagnose CC were: (i) Diffuse thickening of the subepithelial collagen layer by greater than 10 μm; (ii) Inflammation in the lamina propria dominated by lymphocytes and plasma cells; (iii) Epithelial damage, such as flattening and detachment of surface epithelium.[2,3] The LC diagnostic criteria were as follows: (i) Increased intraepithelial lymphocytosis (≥20 lymphocytes per 100 epithelial cells); (ii) Inflammation in the lamina propria dominated by lymphocytes and plasma cells; (iii) Epithelial damage, such as flattening and detachment of surface epithelium; (iv) Absence of subepithelial collagen layer (<10 μm).[2,3]

Study Protocol

The total numbers of upper endoscopy and lower endoscopy (colonoscopy and sigmoindoscopy) performed in the Calgary Health Region were obtained via a centralised endoscopy database. Information on each endoscopic procedure performed between 1 January 2004 and 31 December 2008 was retrieved. This data included patient information, type of procedure, date of procedure, indication for the procedure (if stated) and endoscopic findings.

To identify patients diagnosed with CD and/or MC during the 5-year study period, a keyword search of the pathology database was performed. Results were limited to biopsy specimens obtained between 1 January 2004 and 31 December 2008. The data abstracted from each report included patient information, date of procedure, endoscopist, pathological diagnosis and additional pathologist comments. The pathology reports did not consistently specify the number of biopsies obtained from each anatomic location. Each pathology report was reviewed to determine if histological evidence of CD or MC was present.

Patients who had multiple colonic or duodenal diagnostic pathology reports generated from biopsies obtained on separate occasions were reviewed. To ensure only incident cases were included, only one report from each site (colon or duodenum) was retained for analysis. The earliest biopsy report was retained if the patient had multiple biopsies with the same diagnosis (normal or pathological). For patients with multiple biopsies resulting in discordant results, the abnormal biopsy report was retained.

Each patient biopsy was linked to the corresponding procedure entry in the endoscopy database using patient information and the date of the procedure. This allowed each biopsy to be linked with the indication for endoscopy as well as endoscopist information that was available from the Department of Gastroenterology. The endoscopist's information included speciality (gastroenterologist, surgeon, paediatric gastroenterologist), academic or private practice and age group (< or >50 years old). Endoscopy reports for all patients who had both a duodenal biopsy demonstrating CD and a colonic biopsy demonstrating MC were reviewed to confirm the indication and endoscopic findings of the procedure.

Statistical Analysis

Descriptive analysis of both the MC and CD cohorts was performed. Crude incidence rates for MC and CD were calculated using population data for the years 2004–2007 provided by the Calgary Health Region.[17] Population data for 2008 was not available and was therefore extrapolated by averaging the annual population growth over the previous decade and adjusting the 2007 population accordingly. The crude incidence rates were then age-standardised and gender-standardised to the 2006 Canadian Census. All incidence rates are presented as standardised rates. By performing this standardisation, the effect of any regional peculiarities in age and gender distribution on disease incidence were minimised.

To assess the relationship between MC and CD, standardised incidence ratios (SIR's) were calculated by dividing the number of observed diagnoses by expected diagnoses. Expected diagnoses were determined by applying crude incidence rates in the Calgary Health Region to the specific study population.[22] A SIR of 1.0 implies perfect agreement between the number of cases in the evaluated population and the number of cases in the reference population. Interval data were analysed using two-tailed Student t-tests. Pearson's chi-squares were calculated to evaluate differences within categorical data. Where data could be organised as 2 × 2 contingency tables, the Fisher Exact Test was employed. All tests were considered statistically significant at the 0.05 level. Analyses were performed using spss v.17 ( SPSS Inc., Chicago, IL, USA).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.