I have read this article back in July, and immediately found it interesting.  One reason is that it is (albeit a poor one) an example of the use of “Big Data” type analysis where a large database of patient biopsy specimens can be searched.

Second reason – is the actual clinical finding – less polyps than expected in those with microscopic colitis – prevalence of ALL types of polyps was reduced.

Summary:  130,000 patients with symptoms of diarrhea had a colonoscopy with biopsies.  Compared with 97,000 control cases.  So, in this CASE-CONTROL study for patients with workup of diarrhea – less polyps (Odds ratios = 0.46 for hyperplastic, 0.24for serrated adenomas, and 0.35 for tubular adenomas).

Criticism: It would have been nice if the authors have age-matched the cases one for one from the beginning, even though these were adjusted for. There is no duration of disease, so the mechanism would remain elusive. Although it would be interesting to have someone do a database review of patients whose clinical history is known. It can even be compared to the current study, if the data is available, i.e. authors had 14% of colons have polyps in normals, and 8.3% in microscopic colitis,  is there an influence of disease duration or is this an all or none phenomenon?