Olmesartan-associated enteropathy presenting as severe malnutrition

Summary

A woman in her late 70s presented with 14 months of refractory diarrhoea and 32% unintentional weight loss, meeting Global Leadership Initiative on Malnutrition criteria for severe malnutrition. Olmesartan, a component of fixed-dose combination antihypertensive therapy, went unrecognised during five encounters at three institutions. Targeted treatment of hypercalcaemia-related diarrhoea, pancreatic insufficiency, small intestinal bacterial overgrowth and suspected coeliac disease, together with empiric budesonide, failed. At the sixth encounter, component-level medication reconciliation identified olmesartan and revealed repeated improvement when olmesartan had not been administered during prior hospitalisations. Biopsies obtained during push enteroscopy were consistent with olmesartan-associated enteropathy, prompting permanent withdrawal (Naranjo score +9, definite adverse drug reaction). At 2-month follow-up, she had gained 4.8 kg with symptom resolution. Two obstacles contributed to the delay. Brand-name documentation of fixed-dose combination products obscured olmesartan on medication lists, and drug-induced diarrhoea was not considered during the evaluation.

Background

Olmesartan medoxomil (Benicar; Daiichi Sankyo) is an angiotensin II receptor blocker (ARB) used to treat hypertension. In 2013, the US Food and Drug Administration mandated a label change after reports of severe sprue-like enteropathy.¹ A landmark series of 22 patients had characterised the association the previous year.² Olmesartan-associated enteropathy presents with chronic diarrhoea and weight loss. Biopsy shows villous atrophy histologically indistinguishable from coeliac disease, but coeliac serologies are negative, symptoms do not respond to a gluten-free diet, and they resolve when the drug is stopped.^{2 3} Diagnosis remains delayed. Patients often undergo broad evaluations for infectious, pancreatic and immune-mediated causes, and reported symptom duration before identification ranges from 3 to 53 months.^2–5

The true incidence is difficult to establish. A systematic review of 248 cases of ARB-associated enteropathy found olmesartan implicated in 94%, with remission in 97% after drug discontinuation⁶; a more recent systematic review of 183 patients similarly found olmesartan responsible for approximately 90% of cases, with villous atrophy in 89% and complete clinical recovery after drug withdrawal.⁵ A French national survey captured 36 biopsy-confirmed cases over 4 years at referral centres, a figure that almost certainly underestimates the real burden.⁴ Large pharmacoepidemiologic cohorts add population-level perspective: in a multidatabase study of more than 1.9 million ARB initiators, olmesartan carried a higher relative rate of enteropathy-related outcomes than other ARBs, though the absolute incidence was low and the excess risk was concentrated among older patients and those with longer or higher-dose exposure.⁷

Current evidence points towards a cell-mediated mechanism. It implicates overexpression of interleukin-15, expansion of activated CD8-positive intraepithelial lymphocytes and disruption of the tight-junction protein zonula occludens-1, producing mucosal injury and loss of epithelial barrier function. An earlier hypothesis of impaired transforming growth factor beta signalling has not been consistently supported.⁸ HLA-DQ2/DQ8 is present in over 70% of cases but, unlike coeliac disease, is not required for pathogenesis.^{5 6}

We present a case in which olmesartan-associated enteropathy remained unrecognised through five encounters at three institutions over 14 months and was identified during a sixth, after Global Leadership Initiative on Malnutrition (GLIM) stage 2 severe malnutrition developed. Two obstacles contributed. Olmesartan was not readily apparent in fixed-dose combination antihypertensives recorded by brand name, and drug-induced diarrhoea was not considered during the preceding evaluation.^{9 10} The case illustrates the value of component-level medication reconciliation and structured adverse drug reaction review.

Case presentation

A woman in her late 70s with hypertension, hyperlipidaemia and primary hyperparathyroidism status post partial parathyroidectomy presented with 14 months of watery, non-bloody diarrhoea (up to 10 times daily, including nocturnal episodes), abdominal cramping, fatigue and profound anorexia, with a weight of 56.1 kg representing a 32% loss from her baseline of 82.6 kg. Her antihypertensive regimen was amlodipine/olmesartan (Azor), transitioned from amlodipine/olmesartan/hydrochlorothiazide (Tribenzor) when hydrochlorothiazide was discontinued for hypercalcaemia management approximately 14 months before admission; she had been taking a combination antihypertensive for approximately 3 years. There was no family history of coeliac disease, inflammatory bowel disease, or malignancy and no recent non-steroidal anti-inflammatory drug use, sick contacts, travel or substance use. The clinical timeline is summarised in table 1.

Table 1. Clinical timeline.

Time point	Event	Details
~3 years before admission	Olmesartan initiated	Primary care visit for hypertension (outside institution; records not available). Combination antihypertensive started.
~2.5 years before admission	Triple combination confirmed	Amlodipine/olmesartan/hydrochlorothiazide (Tribenzor) confirmed at outside institution.
14 months before admission	Emergency department presentation (encounter 1 at outside hospital)	Watery diarrhoea up to 10 times daily and weight loss from a baseline of 82.6 kg. Hypercalcaemia identified, parathyroidectomy recommended and hydrochlorothiazide discontinued.
13 months before admission	Outside hospital admission (encounter 2)	Coeliac panel negative on a gluten-containing diet; faecal elastase 103 µg/g; pancreatic enzyme replacement started; upper endoscopy and colonoscopy with sprue-like pathology; empirically started on gluten-free diet; olmesartan-associated enteropathy not identified.
9–11 months before admission	Outpatient gastroenterology evaluation (encounter 3)	Small intestinal bacterial overgrowth breath test positive; rifaximin without benefit. Empiric gluten-free diet without benefit. Parathyroidectomy performed approximately 10 months before admission; diarrhoea persisted. Olmesartan-associated enteropathy not identified.
2 months before admission	First hospitalisation, our institution (encounter 4)	Olmesartan not administered (formulary unavailability, normotension). Gastroenterology consulted; olmesartan-associated enteropathy not identified. Repeat coeliac serologies drawn after several months of gluten avoidance. Symptoms improved; recurred after discharge (rechallenge 1).
1 month before admission	Second hospitalisation, our institution (encounter 5)	Olmesartan not administered. Gastroenterology consulted; olmesartan-associated enteropathy not identified. Empiric budesonide 9 mg daily initiated; symptoms initially improved for 2 weeks then recurred (rechallenge 2).
Index admission (encounter 6)	Index hospitalisation	Severe malnutrition (56.1 kg; 32% loss). Primary team identified olmesartan as a possible cause; gastroenterology recommended push enteroscopy and colonoscopy.
Hospital day 2	Push enteroscopy and colonoscopy	Macroscopically normal throughout; biopsies obtained from stomach, duodenum, jejunum and colon (ascending, descending and sigmoid).
~1 week after discharge	Biopsy findings consistent with olmesartan-associated enteropathy	Olmesartan formally and permanently discontinued.
~2 weeks after discharge	Subjective improvement	Marked reduction in diarrhoea; appetite improved.
~6 weeks after discharge	Capsule endoscopy cancelled	Patient-directed cancellation following symptom resolution.
2 months after discharge	Cardiology follow-up	Weight 60.9 kg (BMI 23.8); symptoms fully resolved; adverse event documented.
2 months after discharge	Telephone follow-up	Ongoing resolution confirmed; gastroenterology follow-up scheduled.

BMI, body mass index.

Over 14 months, the patient was evaluated by her primary care provider, hospitalists and gastroenterologists across 6 encounters. The emergency department evaluation at symptom onset was encounter 1. At an outside hospital admission 13 months before the index admission (encounter 2), while she was consuming an unrestricted gluten-containing diet, a coeliac panel was negative and duodenal biopsy demonstrated villous blunting with a sprue-like pattern. Faecal elastase was 103 µg/g, prompting empiric pancreatic enzyme replacement. Olmesartan-associated enteropathy was not identified. She was empirically started on a gluten-free diet. Hypercalcaemia was considered a contributing cause, and parathyroidectomy was performed approximately 10 months before admission without symptomatic benefit. During outpatient gastroenterology evaluation at a second outside institution approximately 9 to 11 months before admission (encounter 3), rifaximin given for a positive small intestinal bacterial overgrowth breath test provided no benefit, and an empiric gluten-free diet was similarly ineffective.

During two hospitalisations at our institution (approximately 2 months before admission, encounter 4 and 1 month before admission, encounter 5), gastroenterology was consulted on both occasions. Olmesartan was not administered during either admission because of formulary unavailability and normotension, and it was resumed on discharge. Only after the diagnosis was made did the patient recognise the pattern. She had consistently felt better during each admission and worse after discharge, reflecting two unintentional rechallenges. During encounter 5, empiric budesonide 9 mg daily was initiated for presumed small bowel disease. Symptoms improved for approximately 2 weeks before recurring, leading to the index admission.

On admission for the index hospitalisation (encounter 6), the patient appeared cachectic and fatigued. Vital signs were heart rate 88 bpm, blood pressure 104/62 mm Hg, temperature 36.8°C, and oxygen saturation 97% on room air. Weight was 56.1 kg (body mass index (BMI) 21.9 kg/m²). Formal nutrition-focused physical examination documented moderate muscle wasting at the temples, interosseous muscles, clavicle, shoulders and scapula, and moderate subcutaneous fat loss at the triceps. Mild diffuse abdominal tenderness and 1+ bilateral lower extremity oedema were present.

Investigations

Admission investigations demonstrated severe hypoalbuminaemia (2.3 g/dL), total protein 5.1 g/dL, hypokalaemia (2.8 mEq/L), hypomagnesaemia (0.7 mg/dL), hypophosphataemia (1.9 mg/dL), hypocalcaemia (ionised calcium 0.71 mmol/L) and hyponatraemia (133 mEq/L), consistent with high-output diarrhoea and malabsorption. Haemoglobin was 110 g/L. Fat-soluble vitamin deficiencies included vitamin D (11 ng/mL), vitamin A (<5 µg/dL) and vitamin K (<50 pg/mL); zinc was low. C reactive protein was borderline elevated, and erythrocyte sedimentation rate was elevated; thyroid-stimulating hormone was normal.

Coeliac serologies (tissue transglutaminase IgA, tissue transglutaminase IgG and deamidated gliadin peptide IgG) were negative, and total IgA was 642 mg/dL, excluding IgA deficiency. Notably, the coeliac serologies were obtained at encounter 4, by which time the patient had been following a gluten-free diet for several months, so a negative result at that point cannot by itself exclude coeliac disease. The earlier panel obtained at encounter 2, while she was consuming an unrestricted gluten-containing diet, was also negative and obtained at an outside institution. Antinuclear antibody was positive at 1:320; its clinical significance is discussed below. Anti-enterocyte antibody was ordered, but the specimen was not collected prior to discharge, as clinical suspicion for autoimmune enteropathy had diminished given the overall constellation of findings. HLA-DQ2/DQ8 was not obtained because clinical suspicion for coeliac disease was low. Detailed results are shown in table 2.

Table 2. Selected laboratory values.

Laboratory test	Result	Reference range	Interpretation
Albumin	2.3 g/dL	3.4–5.4 g/dL	Severely low
Total protein	5.1 g/dL	6.0–8.3 g/dL	Low
Potassium	2.8 mEq/L	3.5–5.0 mEq/L	Low
Magnesium	0.7 mg/dL	1.7–2.4 mg/dL	Low
Phosphorus	1.9 mg/dL	2.5–4.5 mg/dL	Low
Sodium	133 mEq/L	136–145 mEq/L	Mildly low
Calcium	5.7 mg/dL (ionised 0.71 mmol/L)	8.5–10.5 mg/dL (ionised 1.13–1.32 mmol/L)	Low
Haemoglobin	110 g/L	120–150 g/L	Low
Vitamin D (25-OH)	11 ng/mL	30–100 ng/mL	Deficient
Vitamin A	<5 µg/dL	38–98 µg/dL	Deficient
Vitamin E (alpha-tocopherol)	8.6 mg/L	>5.0 mg/L	Within normal limits
Vitamin E (beta-gamma-tocopherol)	<1.0 mg/L	–	Low fraction; clinical significance uncertain
Vitamin K	<50 pg/mL	130–1500 pg/mL	Deficient
Zinc	57 µg/dL	60–120 µg/dL	Low
C reactive protein	1.1 mg/dL	<1.0 mg/dL	Borderline elevated
Erythrocyte sedimentation rate	39 mm/hour	<20 mm/hour	Elevated
Thyroid-stimulating hormone	2.27 mIU/L	0.4–4.0 mIU/L	Normal
Antinuclear antibody	1:320	Negative	Positive; significance uncertain (see Discussion section)
Tissue transglutaminase IgA*	1.0 U/mL	<4.0 U/mL	Negative
Tissue transglutaminase IgG*	1.1 U/mL	<6.0 U/mL	Negative
Deamidated gliadin peptide IgG*	<0.6 U/mL	<20 U/mL	Negative
Total IgA*	642 mg/dL	70–400 mg/dL	Elevated; IgA deficiency excluded
HIV	Negative	Negative	–
Faecal elastase (outside, unverified)	103 µg/g	>200 µg/g	Prompted empiric enzyme replacement; unverified and likely a dilutional artefact of high-output diarrhoea
Faecal elastase (index admission)	229 µg/g	>200 µg/g	Normal; pancreatic exocrine insufficiency excluded
Stool cultures and clostridioides difficile	Negative	Negative	Infectious aetiology excluded
Anti-enterocyte antibody	Not collected	–	Ordered; specimen not collected prior to discharge
HLA-DQ2/DQ8	Not obtained	–	Not performed; see Discussion section

Values are from the index admission (encounter 6) unless otherwise indicated.

^*Celiac serologies and total IgA were drawn at Encounter 4, after several months of gluten avoidance. An earlier celiac panel obtained at Encounter 2, at an outside institution, while the patient was consuming an unrestricted gluten-containing diet, was also negative.

Push enteroscopy and colonoscopy were performed on hospital day 2. Endoscopic inspection was macroscopically normal throughout, consistent with the 52% rate of normal duodenal appearance reported in a national survey,⁴ and biopsies were obtained from the stomach, duodenum, jejunum and colon. The colonoscopy reached the cecum, and biopsies from the ascending, descending and sigmoid colon showed normal colonic mucosa, excluding collagenous and lymphocytic colitis.

Histopathology demonstrated a malabsorptive pattern confined to the small bowel, consistent with medication-associated small bowel injury.¹¹ The jejunum and duodenum showed villous attenuation, mildly increased intraepithelial lymphocytes, an irregular subepithelial collagen table and paucity of goblet and Paneth cells; the duodenum additionally showed focal acute inflammation. The stomach demonstrated a reactive chemical gastritis pattern, which can be associated with bile reflux. No recent non-steroidal anti-inflammatory drug exposure was reported, and Helicobacter pylori was negative. Periodic acid-Schiff and Congo red stains and Tropheryma whipplei immunohistochemistry were negative. Findings are detailed in table 3 and figure 1.

Table 3. Histopathologic findings by site.

Site	Histopathologic findings
Jejunum	Villous attenuation, mildly increased intraepithelial lymphocytes, irregular subepithelial collagen table, paucity of goblet and Paneth cells
Duodenum	Villous attenuation, mildly increased intraepithelial lymphocytes, irregular subepithelial collagen table, paucity of goblet and Paneth cells, focal acute inflammation
Stomach	Reactive chemical gastritis pattern, compatible with bile reflux. No recent non-steroidal anti-inflammatory drug exposure reported. Helicobacter pylori negative
Colon (ascending, descending, sigmoid)	Normal colonic mucosa; no subepithelial collagen band and no increase in intraepithelial lymphocytes, excluding collagenous and lymphocytic colitis
Special stains and immunohistochemistry	Periodic acid-Schiff negative; Congo red negative; Tropheryma whipplei immunohistochemistry negative

Figure 1. Histopathologic features of olmesartan-associated enteropathy. Original histopathology images provided by EDY, Department of Pathology, Johns Hopkins Hospital. (A) H&E, low power. Duodenal mucosa demonstrating villous attenuation with broadened, blunted villi and increased lamina propria cellularity. (B) H&E, medium power. Villous attenuation with a visible pale subepithelial collagen band and diffuse inflammatory infiltrate in the lamina propria. (C) Trichrome stain, high power. Irregular, thickened subepithelial collagen table (blue) consistent with evolving collagenous sprue. (D) H&E, high power. Villous epithelium demonstrating paucity of goblet cells and patchy, increased intraepithelial lymphocytes. (E) H&E, high power. Crypts demonstrating patchy preservation of a Paneth cell (arrow) alongside goblet cells (asterisks), a pattern arguing against diffuse autoimmune enteropathy.

The patient met GLIM criteria for severe (stage 2) malnutrition on the basis of the severity of her weight loss.¹² She had lost 32% of her body weight over 14 months, well above the greater than 20% threshold that defines severe grading beyond 6 months. Additional supporting phenotypic findings were reduced muscle mass on formal nutrition-focused physical examination and BMI 21.9 kg/m², below the age-adjusted threshold of less than 22 kg/m² for patients over 70 years. Aetiological criteria were met by chronic reduced intake and malabsorption from prolonged diarrhoeal illness. Weight trajectory is summarised in table 4.

Table 4. Weight trajectory.

Time point	Weight	BMI (kg/m²)	Clinical context
Baseline (preillness)	82.6 kg (182.1 lb)	32.2	Outside hospital initial presentation; patient-reported weight
~11 months before admission	73.0 kg (160.9 lb)	28.5	~10 kg loss from baseline
~4 months before admission	66.2 kg (145.9 lb)	25.9	Ongoing decline
2 months before admission	61.0 kg (134.5 lb)	23.8	First hospitalisation at our institution (encounter 4)
1 month before admission	58.1 kg (128.0 lb)	22.7	Second hospitalisation at our institution (encounter 5)
Index admission (nadir)	56.1 kg (123.7 lb)	21.9	32% total weight loss; GLIM stage 2 criteria met
2 months after discharge	60.9 kg (134.2 lb)	23.8	4.8 kg regained, BMI increased to 23.8 and symptoms fully resolved

BMI, body mass index; GLIM, Global Leadership Initiative on Malnutrition.

Differential diagnosis

The differential diagnosis for chronic malabsorptive diarrhoea with villous atrophy was worked through systematically. Coeliac disease was excluded on the basis of a negative coeliac serology panel obtained at encounter 2 while the patient was consuming an unrestricted gluten-containing diet, failure to respond to an adequate trial of a gluten-free diet, which she had abandoned before the index admission and complete resolution of symptoms after drug withdrawal alone while she was consuming gluten (from index admission onwards). The second negative panel, drawn at encounter 4 after several months of gluten avoidance, is consistent with this conclusion but does not independently exclude the diagnosis. Seronegative coeliac disease was considered unlikely on the same grounds, particularly the failure of an adequate gluten-free diet and complete resolution after olmesartan withdrawal while the patient was consuming gluten. Microscopic colitis, both collagenous and lymphocytic, was excluded by normal biopsies from three colonic segments, obtained despite a macroscopically normal colonoscopy. Infectious aetiologies including Tropheryma whipplei were excluded by negative stains and immunohistochemistry. Autoimmune enteropathy was considered unlikely because goblet and Paneth cell loss was patchy rather than diffuse, there was no hypogammaglobulinaemia, and symptoms resolved after olmesartan withdrawal without immunosuppression. Pancreatic exocrine insufficiency, suggested by a low outside faecal elastase, was excluded when repeat testing at our institution returned within normal limits. Drug-induced enteropathy was not raised until the index admission, when component-level medication reconciliation identified olmesartan as an ingredient of the combination antihypertensive and prompted recognition of its association with enteropathy.

Treatment

At the index admission, olmesartan was again not administered because of normotension and formulary unavailability. Once olmesartan was suspected as the cause, the primary team instructed the patient to continue avoiding it as an outpatient. She was transitioned to amlodipine monotherapy and counselled that olmesartan appears under multiple brand names including Benicar, Benicar HCT, Azor and Tribenzor. She received intravenous fluid and electrolyte repletion, vitamin supplementation and nutritional monitoring with refeeding precautions. Enteral and parenteral nutrition were not required. Pancreatic enzyme replacement was discontinued, as pancreatic exocrine insufficiency was effectively excluded when repeat faecal elastase at our institution returned within normal limits (229 µg/g), and the prior outside-hospital value was interpreted as a dilutional artefact from high-output diarrhoea. Biopsy results returned approximately 1 week after discharge and were consistent with olmesartan-associated enteropathy, at which point olmesartan was formally and permanently discontinued.

Outcome and follow-up

Within approximately 2 weeks of stopping olmesartan, the patient reported marked improvement in diarrhoea and appetite. At 2-month cardiology follow-up, she had gained 4.8 kg from her nadir (56.1 to 60.9 kg, BMI 23.8), with complete symptom resolution and the adverse event was formally documented. A prescheduled capsule endoscopy was cancelled by the patient after her symptoms resolved. Recovery without enteral or parenteral support was consistent with prior reports of rapid improvement following drug withdrawal.^{2 3 5 13–17}

Discussion

That olmesartan-associated enteropathy remained unrecognised through five encounters before identification during a sixth, despite specialist consultation and nearly one-third body-weight loss, suggests a process failure rather than a single oversight.^2–4

The fixed-dose combination formulation was central to the delay. At encounter 1, the patient transitioned from Tribenzor to Azor after hydrochlorothiazide was discontinued for hypercalcaemia management. Both products contained olmesartan, but documentation under different brand names reduced continuity in recognising the shared exposure across encounters.^{3 9 10}

Hydrochlorothiazide was recognised and discontinued at encounter 1 because hypercalcaemia prompted a targeted review for thiazide exposure. That review did not extend to systematic assessment of the remaining components as possible causes of the patient’s diarrhoea. Identifying one relevant component of a fixed-dose product therefore did not ensure comprehensive medication review.

Identifying olmesartan by name, however, still requires recognising its association with enteropathy. Although olmesartan is catalogued among drug causes of chronic diarrhoea,⁹ the presentation can closely mimic coeliac disease, pancreatic insufficiency and inflammatory enteropathy.^{2 4} The association did not surface during the first five encounters.

Structured adverse drug reaction review against a pharmacopoeia or adverse event database can make recognition reproducible rather than dependent on prior experience. Applied during any of the first five encounters, it could have identified olmesartan before severe malnutrition developed. The approach is generalisable to unexplained gastrointestinal presentations but must be targeted.^{9 10} Because diarrhoea is listed for more than 700 medications,⁹ indiscriminate review could generate marginal associations and alert fatigue. Reserving it for unexplained or treatment-refractory presentations may improve its usefulness. These process observations from a single case are hypothesis-generating and require prospective evaluation.

The hospitalisation pattern provided additional evidence: olmesartan was not administered for independent reasons, with symptomatic improvement followed by recurrence after discharge and readministration, constituting an unintentional dechallenge and rechallenge.^{2 3} The apparent budesonide response during encounter 5 began during the admission and was likely confounded by concurrent olmesartan withdrawal.^{3 5} The Naranjo Adverse Drug Reaction Probability Scale score was +9 (definite): previous reports (+1), onset after exposure (+2), improvement after withdrawal (+1), recurrence after readministration (+2), absence of an alternative cause (+2) and objective confirmation (+1); the remaining items scored 0.¹⁸ Alternative explanations were further weakened by persistent diarrhoea despite parathyroidectomy for hypercalcaemia, rifaximin for small intestinal bacterial overgrowth and a gluten-free diet.

The biopsy findings deserve comment. The irregular subepithelial collagen table points towards evolving collagenous sprue, a variant present in roughly one-third of cases.^{2 15 19} The paucity of goblet and Paneth cells raised the possibility of autoimmune enteropathy. It was considered unlikely because cell loss was patchy rather than diffuse, there was no hypogammaglobulinaemia, and symptoms resolved after olmesartan withdrawal without immunosuppression. The positive antinuclear antibody at 1:320 lacked corroborating clinical or serologic features and was interpreted as an incidental finding. Both push enteroscopy and colonoscopy were macroscopically normal, yet biopsies demonstrated small bowel injury, underscoring that a normal-appearing endoscopy does not exclude the diagnosis.^{4 20}

Reported cases span a wide clinical spectrum (table 5). Against this published experience, the present case is notable for its severity: 32% body-weight loss (26.5 kg) meeting GLIM stage 2 criteria sits at the upper end of the reported range and well above the 13 kg median reported across 183 cases, reflecting the cumulative toll of a 14-month delay.^{5 12}

Table 5. Comparison of the present case with published reports of olmesartan-associated enteropathy.

Study (year)	Age/ sex	Drug and exposure	Weight loss	Biopsy findings	Endoscopic appearance	Rechallenge	Outcome
Present case	F, late 70s	Olmesartan (fixed-dose combination), ~3 years	26.5 kg (32%), GLIM stage 2	Villous atrophy, intraepithelial lymphocytosis, irregular subepithelial collagen, goblet/Paneth cell paucity	Macroscopically normal	Yes, unintentional after discharge from 2 admissions (Naranjo+9)	Full recovery, 4.8 kg regained at 2 months
Gurgel 202413	M, 63	Olmesartan 20 mg, ~2 years	~10 kg	Villous atrophy, intraepithelial lymphocytosis. Follow-up biopsy normal	Duodenal atrophy, loss of folds	No formal rechallenge	Full recovery. Follow-up biopsy normal
Goshgarian 202415	F, 74	Olmesartan (dose not reported), ~6 months	~27 kg (60 lb)	Collagenous sprue with marked villous atrophy and thickened collagen	Superficial erosions	Not performed	Symptom resolution, BMI restored
Schiepatti 2024 (review, n=183)5	Mean 69 years, 55% F	Olmesartan in ~90%	Median 13 kg	Villous atrophy 89%; intraepithelial lymphocytosis 95% (79/83 patients with reported data)	Often normal or non-specific	Recurrence on rechallenge where reported	100% clinical recovery on withdrawal (178/178 patients with reported outcomes)
Suardi 202614	F, 60s	Olmesartan 40 mg, ~1 year	Not reported	Severe villous atrophy, crypt hyperplasia, intraepithelial lymphocytosis	Duodenal oedema, villous atrophy, erosions	Yes, inadvertent after discharge	Full recovery after ARB-class withdrawal
Nikolaeva 202616	M, 53	Olmesartan/amlodipine, ~2 years	~8 kg over 4 weeks	Focal villous blunting, CD3+/CD8+exocytosis	Macroscopically normal	Yes, unintentional recurrence <48 hours	Full recovery after switch to non-ARB
Sugiura 202617	M, 82	Olmesartan (long term)	~5 kg	Normal colon, CT duodenal-ileal wall thickening	Colonoscopy normal	Yes, twice, with acute kidney injury and dialysis-requiring acidosis	Full recovery, no recurrence >1 year

Original synthesis compiled by the authors from the cited reports. –

ARB, angiotensin II receptor blocker; BMI, body mass index; F, female; GLIM, Global Leadership Initiative on Malnutrition; M, male.

These observations align with current management principles. Drug-induced diarrhoea should be considered early in unexplained chronic diarrhoea, making careful medication review essential.^{9 10} When olmesartan-associated enteropathy is suspected, early drug withdrawal is both diagnostic and therapeutic. Symptoms often improve within 1 to 2 weeks.^{3 5 20} An unambiguous response to early withdrawal may make endoscopy and colonoscopy unnecessary. These procedures can be reserved for patients who fail to improve or in whom diagnostic uncertainty persists. If endoscopy or colonoscopy is performed, biopsies should be obtained even when the mucosa appears macroscopically normal. Follow-up endoscopy and immunosuppression are generally unnecessary when clinical recovery is clear, and patients requiring continued antihypertensive therapy should be switched to a different drug class rather than another ARB.⁵ Our patient’s sustained recovery after she cancelled the planned capsule endoscopy was consistent with this approach.

Patient’s perspective.

I had been sick for over a year and no one could tell me what was wrong. I kept getting weaker and losing more weight. I stopped going out and stopped going to my church events because I had to wear Depends as I could not be far from a bathroom. I felt like I could not live a normal life anymore. When I was finally told it was a blood pressure pill causing all of this, I was relieved but also frustrated, as it seemed like such a simple answer after so much suffering. After stopping the medication, I started feeling better within a few weeks. Now I feel like I can live a normal life again. Written by the patient in her own words; grammar lightly edited by the authors.

Learning points.

• Drug-induced diarrhoea is often overlooked during the evaluation of chronic diarrhoea. Olmesartan can cause a sprue-like enteropathy severe enough to produce life-threatening malnutrition and may be overlooked when included in a fixed-dose combination product documented by brand name.

• Reviewing every component of combination products, and applying structured adverse drug reaction review to the complete medication list, makes recognition a reproducible process rather than a matter of prior experience.

• Stopping the suspected drug is a reasonable first diagnostic and therapeutic step and may avoid invasive testing. If endoscopy is performed, take small bowel and colonic biopsies even when the mucosa appears normal.

• Improvement during hospitalisation when a drug is not administered, followed by recurrence after discharge, is an unintentional rechallenge and can be the first clue to the diagnosis.