Microscopic Colitis

Microscopic colitis (MC) is a chronic inflammatory condition of the colon that produces chronic watery, non-bloody diarrhea in a patient whose colonoscopy looks essentially normal. That single fact is the highest-yield teaching point in this entire chapter: unlike ulcerative colitis or Crohn disease, microscopic colitis leaves no visible mucosal lesion for the endoscopist to target with a biopsy forceps — the diagnosis exists entirely at the microscopic level, in random biopsies taken from a grossly unremarkable-looking colon.^[1][2] A pharmacist reading a colonoscopy report that says "normal colon, biopsies pending" should never assume this rules out a genuine, treatable inflammatory colonic disease — it may be exactly the picture MC produces.

Microscopic colitis comprises two principal histologic subtypes — collagenous colitis (CC), defined by an abnormally thickened subepithelial collagen band, and lymphocytic colitis (LC), defined by an increased density of intraepithelial lymphocytes — plus a third, increasingly recognized category, incomplete microscopic colitis (MCi), in which changes are present but fall short of the threshold for either full subtype.^[1] All three share the same clinical presentation, the same normal-appearing colonoscopy, and, per current guidance, essentially the same first-line treatment — the distinction matters for pathology reporting and research, not for the pharmacist's practical treatment approach.^[1]

MC is not rare. Pooled incidence across population-based studies is approximately 11.4 per 100,000 person-years, and pooled prevalence is approximately 119 per 100,000 persons — genuinely comparable in scale to Crohn disease in many Western populations.^[1] It accounts for roughly 1 in 8 cases (12.8%) of otherwise unexplained chronic watery diarrhea worked up in referral settings, disproportionately affects women (pooled odds ratio 2.52 versus men), and is typically diagnosed after age 60 — though up to a quarter of cases are diagnosed before age 45, and pediatric cases are documented.^[1] No Canadian Association of Gastroenterology practice guideline for microscopic colitis exists; this chapter is built primarily from the 2021 European Microscopic Colitis Group/United European Gastroenterology guideline and the American Gastroenterological Association's 2016 guideline, cross-checked against current primary literature throughout.^[1][2]

The presentation: chronic, watery, non-bloody diarrhea

The defining symptom of MC is chronic, watery, non-bloody diarrhea, reported by 84%-100% of patients across studies, with an average of 6-7 bowel movements per day at presentation.^[1] Concomitant symptoms are common and clinically distinctive: fecal urgency (roughly 55%), nocturnal stools (roughly 35%), and fecal incontinence (roughly 26%) all occur meaningfully more often than in ordinary diarrhea-predominant irritable bowel syndrome, and their presence should raise, not lower, suspicion for MC.^[1] Mild abdominal pain and bloating are common; modest weight loss can occur, but the visible blood, marked weight loss, and systemic illness that suggest classic inflammatory bowel disease are notably absent — their presence should prompt reconsidering the diagnosis rather than assuming a more severe presentation of MC (see Red Flags & Refer).

The classic patient

Think of an older woman — median age at diagnosis in the 60s, though MC can occur at any age — presenting with weeks to months of watery diarrhea that has often already been through one or more rounds of empiric management for presumed irritable bowel syndrome or infectious diarrhea before a colonoscopy is finally obtained.^[1] This delay is common and avoidable: MC should be actively considered, not deferred to last, whenever an older patient presents with new persistent watery diarrhea, particularly if a new medication was started in the preceding months (see Risk Factors) or the patient carries a personal history of celiac disease or autoimmune thyroid disease.

Two things that make MC genuinely different from classic inflammatory bowel disease

What untreated collagenous colitis looks like under the microscope

Trichrome-stained colonic biopsy in collagenous colitis. Trichrome staining highlights the thickened subepithelial collagen band (≥10 micrometres) that defines this subtype, seen here as the pale band directly beneath the surface epithelium — a finding invisible to the endoscopist and detectable only on histology from a colon that can look completely normal at colonoscopy. (Ed Uthman, Wikimedia Commons, CC BY 2.0)

• Age over 60 and female sex — pooled odds ratio 2.52 for women versus men across major studies; disease can occur at any age but is most often diagnosed after 60.^[1]
• Current cigarette smoking — odds ratio 2.99 versus never-smokers, more strongly associated with collagenous colitis (OR 5.5) than lymphocytic colitis (OR 2.96); former smoking also carries some elevated risk (OR 1.6).^[1]
• Drug exposure — the single most pharmacist-actionable risk factor in this chapter (see Choose Initial Treatment for the clinical response). A 2023 systematic review and meta-analysis found significant pooled associations for proton pump inhibitors (OR 2.34), SSRIs (OR 2.34), NSAIDs (OR 1.79), and aspirin (OR 2.84).^[3] PPI risk is concentrated disproportionately in lansoprazole specifically, whose individual signal is markedly stronger than the PPI class average — reported as high as OR 9.57 in the same meta-analysis, and OR 4.69 for continuous 4-12 month use in the European guideline's own pooled estimate.^[3][1] Statins, ACE inhibitors, benzodiazepines, beta-blockers, diuretics, and H2-receptor antagonists were not significantly associated in the same meta-analysis, which matters for not over-attributing every drug on a long medication list.^[3]
• Celiac disease — genuinely bidirectional: roughly 3%-4% of MC patients (up to about 13% of collagenous colitis specifically, in some cohorts) also have celiac disease, several-fold higher than the general-population rate, and roughly 7% of celiac disease patients will themselves be diagnosed with MC at some point.^[1]
• Autoimmune thyroid disease — a large 2025 Swedish nationwide case-control study found autoimmune thyroid disease in 12.0% of MC patients versus 7.8% of matched controls (adjusted OR 1.65), with a substantially stronger association (OR 2.41) in patients diagnosed with MC before age 50.^[4]
• Other autoimmune disease — rheumatoid arthritis, type 1 diabetes, Sjögren disease, and psoriasis are all recognized associations; one case-control analysis found autoimmune conditions overall in 18.5% of MC patients versus 7.9% of controls.^[4]
• Concurrent inflammatory bowel disease and bile acid malabsorption are also recognized, clinically relevant associations (see Assessment & Diagnosis and Treatment Failure/Next Step for the bile-acid-diarrhea link).

Case definition

MC is diagnosed on the combination of chronic diarrhea lasting at least one month plus the histologic findings described below, in a colon that looks grossly normal or near-normal at colonoscopy.^[1]

Colonoscopy and biopsy technique — the step most often done incompletely

Because the mucosa looks normal, biopsies must be taken irrespective of endoscopic appearance, not reserved for visibly abnormal areas.^[1] Current guidance specifically recommends biopsies from at least the right and the left colon, submitted in separate, labelled containers so the pathologist knows which segment each sample came from. Concordance between right- and left-colon histology is high (95%-98%), but rectosigmoid-only biopsies are not an adequate substitute and risk a false-negative result, since disease can be patchy or, in a minority, most pronounced proximally.^[1] An endoscopist who declines to biopsy a "normal-looking" colon in a patient with a genuinely compatible history of chronic watery diarrhea has not actually completed the diagnostic evaluation.

What the pathologist is looking for

What NOT to order

Fecal calprotectin is not a useful screening or monitoring test for MC — levels overlap substantially between active disease, quiescent disease, and unaffected controls, unlike its genuinely useful role in classic inflammatory bowel disease.^[1] Routine histological monitoring or repeat biopsy in a clinically improved patient is not recommended — histologic findings correlate only weakly with clinical activity, can persist in 45%-77% of patients even after clinical remission, and 6%-18% of patients show subtype conversion over time without this changing practical management.^[1]

Screen for celiac disease

Because of the genuinely elevated bidirectional prevalence (see Risk Factors), current guidance specifically recommends screening every newly diagnosed MC patient for celiac disease (tTG-IgA plus a total IgA level) — a strong recommendation on high-quality evidence, not an optional add-on reserved for patients with additional celiac-suggestive features.^[1]

Differential diagnosis

• Achieve clinical remission — defined and validated in trials by the Hjortswang criteria: a mean of fewer than 3 stools per day and a mean of fewer than 1 watery stool per day, averaged over a 1-week symptom diary.^[1] Use this, or a simplified version of it, as an objective target rather than relying only on "the patient says they feel better."
• Actively identify and, wherever feasible, discontinue any temporally-associated causative drug before or alongside starting pharmacotherapy — a genuinely first-line, low-risk intervention this chapter treats as its own distinct step, not an afterthought (see Choose Initial Treatment).^[1]
• Use budesonide, the only agent with consistent high-quality trial evidence, at the lowest effective dose for the shortest duration that maintains remission — most patients relapse quickly off treatment, so "shortest duration" does not mean stopping after a single course regardless of outcome (see Monitoring & Follow-Up).^[1]
• Avoid unnecessary exposure to systemic corticosteroids (prednisone/prednisolone) given budesonide's comparable efficacy and substantially better systemic safety profile.^[1]
• Screen for, and correct, celiac disease and other treatable comorbidities that can perpetuate diarrhea despite otherwise adequate MC-directed therapy.
• Recognize genuine treatment failure early, and distinguish it from an inadequate biopsy protocol, an unstopped causative drug, or a coexisting diagnosis, rather than escalating therapy against a possibly incorrect assumption.

When budesonide genuinely is not feasible

Cost, access, or intolerance can occasionally make budesonide impractical. Second-line, weaker evidence bismuth subsalicylate and mesalazine both have some historical trial support and remain listed as conditional options by the AGA^[2], but the more recent 2021 European guideline is more skeptical: it explicitly recommends against mesalazine for induction (no meaningful benefit over placebo across the available trials) and judges the evidence for bismuth subsalicylate insufficient to recommend either way.^[1] This is a genuine area of evolving evidence rather than a settled contradiction — the practical takeaway is that budesonide should be pursued as the true first choice whenever accessible, and mesalazine/bismuth subsalicylate should be understood as a fallback of meaningfully weaker evidence, not an equivalent alternative chosen simply on cost.

Bile acid diarrhea overlap

Bile acid malabsorption/diarrhea coexists often enough in MC that a trial of a bile acid binder (cholestyramine) is a reasonable option, particularly in a budesonide non-responder or when formal bile-acid testing is unavailable — current guidance gives this a weak recommendation in favour, reflecting genuine but limited-quality evidence.^[1]

Symptomatic adjunct

Adjunct only loperamide has no dedicated randomized trial evidence in MC specifically, but given its well-established effect in chronic diarrhea generally, expert opinion favours its use for symptomatic relief in mild disease or as an adjunct alongside directed therapy — it does not treat the underlying inflammation and should not substitute for budesonide in a patient with at least moderate active disease.^[1]

Not recommended

Avoid antibiotics, probiotics, and Boswellia serrata all lack adequate supporting evidence and are specifically recommended against for MC treatment.^[1]

• Systematically review and, in partnership with the prescriber, stop or substitute any temporally plausible causative drug (see Choose Initial Treatment) — genuinely the single most pharmacist-actionable non-drug-therapy intervention in this chapter.^[1][3]
• Smoking cessation is reasonable general advice given the genuine association between current smoking and MC risk, though current evidence is insufficient to confirm that quitting after diagnosis changes the disease course — counsel for its many other health benefits rather than promising it will resolve the diarrhea.^[1]
• Adequate oral hydration and electrolyte attention, particularly in an older patient with high-volume diarrhea, urgency, or nocturnal stools — dehydration risk in this population is genuinely under-recognized relative to more visibly "dramatic" inflammatory bowel disease presentations.
• A trial dietary review for caffeine, dairy (if lactose intolerance is plausible), and other individually reported triggers is reasonable supportive advice, though no specific diet has been shown to alter the underlying disease course the way it does in, for example, celiac disease.
• Reassure the patient early that microscopic colitis does not increase colorectal cancer risk and does not require a special surveillance colonoscopy program — this is a genuinely common and unnecessary source of patient anxiety once any "colitis" diagnosis is mentioned.^[1]

Older adults

This is genuinely the population MC most often affects — most patients are diagnosed after age 60.^[1] High-volume watery diarrhea, urgency, and nocturnal stools carry real fall, dehydration, and functional-decline risk in frail older patients; treat these symptoms with real urgency rather than as a minor nuisance complaint. Review the full medication list with particular attention to PPIs continued without an ongoing indication (a well-recognized, broadly under-deprescribed pattern in this age group generally) and to NSAIDs used for chronic musculoskeletal pain. Watch for cumulative corticosteroid exposure — bone mineral density can decline with prolonged budesonide use even though it is substantially safer than prednisone (see Monitoring & Follow-Up).^[1]

Women

The roughly 2.5-fold female predominance holds across both major subtypes and across the age spectrum.^[1] No MC-specific reason exists to alter counseling by sex beyond the higher baseline pretest probability when a woman presents with compatible symptoms.

Patients who cannot simply stop the causative drug

Not every implicated drug can be discontinued — an SSRI may be essential for the patient's mental health, an NSAID may be the only effective option for another chronic condition, or a PPI may have a genuine, ongoing indication (confirmed erosive esophagitis, a bleeding-risk indication). In these situations, do not withhold budesonide while waiting for a deprescribing conversation to resolve — treat the MC directly, and pursue medication review as a parallel, collaborative discussion with the prescriber rather than a precondition for treatment.

Patients with celiac disease or autoimmune thyroid disease

Given the genuinely elevated bidirectional prevalence, persistent GI symptoms in a patient with known celiac disease or autoimmune thyroid disease — even one who is otherwise well-controlled — deserve real consideration of a second, coexisting MC diagnosis rather than being attributed automatically to the known condition.^[1][4] Correcting one diagnosis does not rule out the other.

Pregnancy and lactation

MC is uncommon in patients of reproductive age given its typical age of onset, but it does occur. Safety data specific to budesonide use for MC in pregnancy are limited; individualize therapy with obstetric input, favouring the lowest effective dose and involving the patient's obstetric care team in any treatment decision.

Steroid-dependent or biologic-refractory patients

A small but genuine and growing body of evidence, including a 2022 Canadian multicentre case series (7 patients across 3 Winnipeg-area centres, most either oral-steroid-dependent or with prior biologic failure/intolerance), supports vedolizumab as an option once budesonide, immunomodulators, and/or anti-TNF therapy have failed — in that Canadian series, stool frequency fell from 7 or more daily to 1-3 daily in the majority of patients by more than 6 months post-induction, with clear improvement in urgency, incontinence, and nocturnal symptoms.^[5] This remains specialist-directed, refractory-disease-only therapy (see Treatment Failure/Next Step).

• Reassess symptoms against the Hjortswang remission criteria (mean <3 stools/day and <1 watery stool/day over a week) at follow-up, not simply a "better/worse" impression.^[1]
• Do not order routine repeat colonoscopy, biopsy, or fecal calprotectin to monitor a clinically improving patient — none of these change management, and histologic findings correlate only weakly with symptoms.^[1]
• For a patient continuing budesonide beyond a single induction course: reassess at regular intervals whether maintenance therapy is genuinely still needed, aim for the lowest effective dose (6 mg daily, or an alternate-day 3 mg/6 mg regimen once stable), and periodically attempt a taper, given that roughly 20% of patients need 6 mg/day or more long-term while many others can step down successfully.^[1]
• Watch for bone-health implications with prolonged budesonide use — while its systemic corticosteroid burden is substantially lower than prednisone's, prolonged high-dose use has still been associated with measurable bone mineral density change in other budesonide-treated GI populations; apply the same general bone-health vigilance used for any patient on long-term corticosteroid-class therapy.^[1]
• Reconfirm at every visit that any causative drug identified at diagnosis was actually stopped, not merely "discussed" — this is a common, quietly incomplete step.
• Revisit the celiac disease screen if it was not completed at diagnosis, and reassess for it if new or unexplained symptoms emerge later.^[1]

1. Confirm the original diagnosis was made correctly — review whether biopsies were actually taken from both the right and left colon (not rectosigmoid-only) and whether the histologic report met formal criteria for CC, LC, or MCi; an inadequate original biopsy protocol is a genuinely underappreciated cause of apparent "non-response."^[1]
2. Confirm any identified causative drug was actually stopped, not just discussed — reconcile the current medication list directly rather than relying on the patient's recollection of an earlier conversation.
3. Confirm adequate budesonide dose (9 mg/day) and duration (a full 6-8 weeks) were actually completed before labelling the induction course a failure.^[1]
4. Screen (or re-screen) for celiac disease if not already done, and consider testing for bile acid diarrhea or empirically trialling a bile acid binder, particularly if the diarrhea pattern is compatible.^[1]
5. Genuine budesonide non-response, intolerance, or a requirement for more than roughly 6 mg/day to maintain remission: refer to gastroenterology for consideration of a thiopurine (azathioprine), an anti-TNF agent (infliximab or adalimumab), or vedolizumab — vedolizumab in particular has real, growing, specialist-directed evidence in exactly this steroid-dependent/biologic-refractory population, including genuine Canadian clinical experience.^[1][5]
6. Do not add methotrexate at any point in this pathway — current guidance specifically recommends against it for MC.^[1]
7. Rare, severe, medically refractory disease despite the above: surgery may be considered as a genuine last resort in carefully selected patients, though this is uncommon.^[1]

• Ruling out an inflammatory colonic disease, and therefore MC specifically, because "the colonoscopy was normal" — a normal-appearing colon is the expected finding in MC, not a reason to skip biopsy.^[1]
• Accepting rectosigmoid-only biopsies as an adequate substitute for biopsies from both the right and left colon.^[1]
• Starting prednisone or another systemic corticosteroid before budesonide, rather than reserving systemic steroids for the rare situation where budesonide is genuinely unusable.^[1]
• Not reviewing the medication list for a PPI (especially lansoprazole), SSRI, NSAID, or aspirin with a plausible temporal relationship to diarrhea onset before or alongside starting pharmacotherapy.^[1][3]
• Continuing a chronic PPI indefinitely in an MC patient without ever reassessing whether the original indication still applies — a broadly under-deprescribed pattern that is specifically relevant here.
• Ordering fecal calprotectin or scheduling routine repeat colonoscopy/biopsy to monitor a clinically stable or improving patient.^[1]
• Treating budesonide as a fixed, one-time course regardless of outcome — a genuine majority of patients relapse after stopping induction therapy, and maintenance dosing is a real, guideline-supported option, not "long-term steroid dependence" to be avoided reflexively.^[1]
• Skipping celiac disease screening at diagnosis.^[1]
• Telling a patient that microscopic colitis raises their colorectal cancer risk or that they need extra surveillance colonoscopies — current evidence shows no increased risk, and no special surveillance program is recommended.^[1]
• Attributing new or changing GI symptoms in a patient with known celiac disease or autoimmune thyroid disease automatically to that known condition, without considering a second, coexisting MC diagnosis.^[4]

How to use this table: Doses are adult doses and require patient-specific confirmation (renal/hepatic function, weight, allergy history). Pediatric dosing is weight-based per the current product monograph unless stated. Drug pricing is intentionally omitted. Select any drug name to open its complete Pharmasist Drug Database page.

First-Line — Budesonide (Oral)

The only agent with consistent, high-quality trial evidence in MC. Preferred over prednisone and every other systemic corticosteroid given its high first-pass metabolism and substantially lower systemic exposure.

Second-Line — When Budesonide Is Genuinely Not Feasible

Both carry meaningfully weaker current evidence than budesonide. The 2021 European guideline now recommends against mesalazine for induction and judges bismuth subsalicylate evidence insufficient either way, in contrast to the AGA's older conditional recommendations for both.

Bile Acid Binder — Overlap or Adjunct Trial

Reasonable in a patient with a bile-acid-diarrhea-compatible pattern, or as a trial in a budesonide non-responder. Weak recommendation reflecting genuine but limited-quality evidence.

Symptomatic Adjunct

Does not treat the underlying colonic inflammation and should not substitute for budesonide in at least moderate active disease.

Avoid as First-Line, or Not Recommended

Prednisone carries meaningfully more systemic exposure than budesonide for equivalent efficacy. Methotrexate is specifically recommended against for MC at any point in the treatment pathway.

Refractory Disease — Specialist-Directed Immunomodulators and Biologics

Reserved for confirmed budesonide-refractory, intolerant, or steroid-dependent disease after diagnosis reconfirmation and medication review (see Treatment Failure/Next Step). All are off-label for MC.

Medications Associated With Microscopic Colitis — Review for Discontinuation or Substitution

None of these are contraindicated outright, but each is worth a specific second look at diagnosis — see Choose Initial Treatment. Pooled odds ratios from a 2023 systematic review and meta-analysis, unless otherwise noted.