Inflammatory Bowel Disease

Meet the patient

A 24-year-old postgraduate presents with four months of diarrhoea, cramping right-lower-quadrant pain worse after meals, a 6 kg weight loss, and — almost as an afterthought — a perianal skin tag his GP treated as haemorrhoids. His stools are sometimes blood-streaked. His CRP is raised, his faecal calprotectin is 480, and a stool culture is negative.[1][2]

Two questions now sit live and they govern everything that follows: is this Crohn's or ulcerative colitis? (the ileocolonoscopy, the histology and the MRI enterography decide) and is he high-risk enough to start a biologic now rather than climb a ladder? (the phenotype — stricturing, penetrating, perianal, young — decides).[4]

Two diseases, one umbrella — the face-off that earns the marks

IBD is not two diseases fought in parallel — it is one immune-driven process read through three instruments: the endoscope, the biopsy, and the cross-sectional scan. The split between Crohn's and UC is the single most reproduced fact in a gastroenterology viva, so learn it as a face-off with one discriminator line beneath.[1][3]

[1]

The discriminator line: rectal sparing plus perianal fistulae plus granulomas points to Crohn; continuous rectal-origin bloody diarrhoea with pseudopolyps points to UC. When the picture is mixed and the work-up incomplete, the honest label is IBD-unclassified (10 to 15 percent) — not a fudge, a working diagnosis that settles after imaging and histology mature.[1]

The three classification axes you reproduce at viva

Classify along three axes — disease type, anatomy (Montreal), and activity — and never blur them. Examiners ask for all three, in that order, because each changes the management sentence.[1][2][3]

Axis 1 — disease type: Crohn versus UC versus IBD-unclassified (above).[1]

Axis 2 — anatomy (the Montreal classification). This is the language of every IBD letter and trial, reproduced verbatim:[1]

[1]

Axis 3 — activity scores. These set the target in treat-to-target and the threshold for hospitalisation in acute severe disease:[1]

• Truelove and Witts (acute severe UC): more than 6 bloody stools per day plus at least one systemic feature — fever above 37.8 degrees C, pulse above 90, haemoglobin under 105 g/L, or ESR above 30. Hospitalise for IV steroids.[1][3]
• Mayo score (UC): stool frequency, rectal bleeding, endoscopic subscore (MES 0 normal, 1 erythema or reduced vascular pattern, 2 marked erythema or erosions, 3 spontaneous bleeding or ulceration), physician global assessment. Endoscopic remission = MES 0.[1]
• CDAI (Crohn): remission under 150, severe over 450 — now mainly a trial tool; in clinic, CRP and faecal calprotectin are faster.[2]
• SES-CD (endoscopic Crohn): remission under 3; moderate 7 to 15; severe over 15.[4]
• Rutgeerts (post-operative Crohn after ileocolonic resection): i0 none to i4 diffuse ileitis — i0 to i1 is remission, i2 to i4 is recurrence and drives prophylaxis.[2]

The smoking paradox — the single most examinable fact in IBD

Smoking protects ulcerative colitis and destroys Crohn disease. Memorise it once, in both directions, because it is the one risk factor an examiner will twist.[1]

In Crohn's, smoking roughly doubles the risk of stricturing and penetrating complications, raises the surgical and re-operation rate, and accelerates post-operative recurrence — quitting is a genuine disease-modifying intervention. In UC, current smokers have about half the risk, and quitting raises it — so UC often declares itself in the year after a patient stops. The mechanism is imperfectly understood (mucosal mucus, immune modulation, microbiome) but the clinical fact is rock-solid.[1]

The second paradox — appendectomy. Done for true appendicitis before age 20, it is protective against UC (about a 50 percent reduction). In Crohn's it is associated with more severe ileal disease. Both paradoxes are viva gold.[1]

How common, and who gets it

IBD is no longer a Western disease. Prevalence is now above 0.3 percent in North America, Northern Europe and Australasia — and the "second wave" is sweeping newly industrialised regions (South Asia, East Asia, the Middle East, South America) as diets urbanise. The global burden is above 7 million people, onset is bimodal (peak 15 to 30 years; a smaller second peak over 60), and the aetiology is the classic gene-environment-microbe-immune quartet.[1]

Risk factors to run on autopilot: family history (first-degree relatives carry a 10- to 15-fold risk; monozygotic twin concordance is about 50 percent for CD, 15 percent for UC), the smoking split above, a Western low-fibre high-processed diet, early-life antibiotics, Caesarean birth, and urban living. The named susceptibility loci — NOD2/CARD15 (ileal, stricturing Crohn's), ATG16L1 and IRGM (autophagy, ileal Crohn's), IL23R (both diseases), HLA-DRB1 star 0103 (severe UC) — account for roughly a quarter of heritability.[2]

Why it happens — four factors, one inflamed gut

IBD is the prototype polygenic disease where genes, microbiome, barrier and immunity conspire. Each factor earns its place because it names a drug target or a bedside act.[1][2]

1. Genes — over 240 risk loci; the functional clusters (bacterial sensing via NOD2, autophagy via ATG16L1/IRGM, the IL-23/Th17 axis) explain both the inflammation and the biologic targets.[2]
2. Microbiome (dysbiosis) — reduced diversity, loss of the butyrate-producer Faecalibacterium prausnitzii, expansion of adherent-invasive E. coli in Crohn's and Ruminococcus gnavus in UC. Dysbiosis is both cause and consequence of inflammation.[2]
3. Barrier dysfunction — the mucus layer (MUC2), tight junctions (claudins, occludin, ZO-1) and epithelial renewal are impaired, letting luminal antigen translocate to the lamina propria.[1]
4. Immune dysregulation — in Crohn's, lamina propria CD4 T cells polarise to Th1/Th17 under IL-12 and IL-23 (the targets of ustekinumab and risankizumab); in UC the signature is atypical Th2 with NK-cell IL-13. TNF-alpha is the final common effector in both — the rationale for anti-TNF therapy.[5]

Etymology for viva gold: the disease was named for Burrill Bernard Crohn, the New York physician whose 1932 paper with Ginzburg and Oppenheimer described "regional ileitis". Crohn was first author only because the paper was presented alphabetically — his colleagues contributed as much. The eponym survived because the disease kept looking the same wherever it was found.[2]

Clinical presentation — the two fingerprints

Crohn's creeps; ulcerative colitis bleeds. Hold that contrast and the rest slots in.[1]

Crohn disease — variable, segmental, often insidious. Crampy right-lower-quadrant pain after meals, chronic diarrhoea (non-bloody in pure small-bowel disease; bloody only when the colon is involved), weight loss and fatigue from malabsorption and cytokine-driven cachexia, low-grade fever, a palpable right-lower-quadrant mass of thickened ileum, aphthous mouth ulcers, and — in a third — perianal disease (skin tags, fissures, fistula-in-ano, abscess) that may precede the bowel symptoms by years.[2]

Ulcerative colitis — bloody diarrhoea, urgency, tenesmus. The cardinal triad, with left-lower-quadrant cramping and rectal pain in proctitis. Systemic features — fever, tachycardia, weight loss — escalate with severity. The emergency end of the spectrum is acute severe UC and toxic megacolon (below).[1][3]

The extraintestinal manifestations — sort them as parallel or independent

Every IBD patient gets a full-body look, because the disease spills outside the gut in ways that track — or do not track — the bowel activity. Sort them into two buckets and the management follows.[1]

[1]

The discriminator: uveitis, pyoderma gangrenosum, axial arthritis and PSC run their own course — controlling the gut does not control them. Treating the bowel alone and assuming the skin or the eye will follow is the recurring trainee error.[1]

The diagnosis — three instruments, one answer

Diagnosis is a correlation, never a single test. You assemble the clinical picture, the ileocolonoscopy with biopsies, the cross-sectional imaging (MRE for Crohn's small bowel), the biomarkers, and — critically — the exclusion of infection.[1][2][3]

Ileocolonoscopy with segmental biopsies is the cornerstone. Take multiple biopsies from the terminal ileum and every colonic segment (5 or more to maximise granuloma yield in Crohn's); record the macroscopic pattern (continuous vs skip, cobblestone vs friability, pseudopolyps) and send for histology and CMV PCR in severe disease. A normal colonoscopy does not exclude Crohn's — the disease may live in the small bowel, which is why magnetic resonance enterography is mandatory when the suspicion is high.[1]

Cross-sectional imaging. MRE is preferred for small-bowel Crohn's (mural inflammation, strictures, fistula, abscess, the "comb sign" of mesenteric hypervascularity) and is radiation-free — vital in young patients. Pelvic MRI is the gold standard for mapping perianal fistula anatomy. CT is reserved for the acute abdomen (abscess, perforation, obstruction).[1]

Biomarkers. CRP tracks Crohn's better than ESR. Faecal calprotectin above 150 to 250 micrograms per gram strongly suggests active intestinal inflammation in the right setting, and a normal calprotectin has a high negative predictive value for IBD in primary care — the single best non-invasive triage test. Serology (p-ANCA positive in 60 to 70 percent of UC, ASCA in 50 to 70 percent of Crohn's) helps separate indeterminate colitis but is not diagnostic.[1]

[1]

The differential — exclude infection before you immunosuppress

The single non-negotiable rule: prove the diarrhoea is not infection before you reach for steroids or a biologic. Infection mimics IBD, triggers IBD flares, and turns immunosuppression into a catastrophe if missed.[1][2][3]

[1]

The classic trap — ileocaecal tuberculosis versus ileocaecal Crohn's. This is the highest-stakes differential in endemic regions. TB gives caseating granulomas, transverse (not longitudinal) ulcers, a patulous ileocaecal valve, and ascites or nodes; Crohn's gives non-caseating granulomas, aphthous and longitudinal ulcers, and skip lesions. Send IGRA, a chest radiograph, and an ascitic ADA where relevant — because starting anti-TNF in undiagnosed TB is a preventable death.[1]

Acute severe ulcerative colitis — the gastrointestinal emergency

Acute severe UC (ASUC) is an emergency. The mortality is 1 to 3 percent even in expert centres, and every hour of delay costs colon and sometimes life. Humour is off here; this is a resuscitation problem with a strict protocol.[1][3]

The threshold is the Truelove and Witts criteria: more than 6 bloody stools per day plus at least one of fever above 37.8 degrees C, pulse above 90, haemoglobin under 105 g/L, or ESR above 30. The moment the criteria are met, the patient is admitted, made nil by mouth, and the clock starts.[1][3]

[1] [3]

The mantra for ASUC: IV steroids on day 1, Travis on day 3, rescue on day 3 to 5, surgery by day 7 if rescue fails — and the surgeon is in the conversation from the start.[1]

Toxic megacolon — transverse colon above 6 cm on plain film with systemic toxicity — is a surgical emergency on top of the medical one: nil by mouth, IV fluids and electrolytes, IV steroids, broad-spectrum antibiotics, and urgent surgical review for colectomy. Perforation, uncontrolled haemorrhage, and refractory obstruction are immediate indications for laparotomy.[1]

Treat-to-target — STRIDE-II, the framework examiners now quote

The days of "control the symptoms and see" are over. STRIDE-II sets objective targets — clinical remission plus biochemical remission (CRP and faecal calprotectin) plus endoscopic remission — reassessed at named time-points, with therapy escalated whenever the target is missed.[4][6]

• Week 0 — baseline clinical activity, CRP, faecal calprotectin, endoscopy (Mayo for UC, SES-CD for Crohn's); set the target.[4]
• Week 10 to 14 — reassess clinical and biochemical response (CRP, calprotectin).[4]
• Months 6 to 12 — endoscopic reassessment, the new standard. Target: Mayo endoscopic subscore 0 to 1 (ideally 0) in UC; SES-CD under 3 in Crohn's.[4]
• Long-term — dysplasia surveillance in UC; MRE every 6 to 12 months in Crohn's small-bowel disease; ileocolonoscopy at 6 to 12 months after Crohn's resection (Rutgeerts), escalating prophylaxis if i2 to i4.[4]

The classic trap: treating to a falling symptom score while the mucosa stays inflamed. Symptoms lie; the endoscope tells the truth. A patient who feels well but has persistent ulceration is heading for stricture, surgery and cancer — escalate on the endoscopy, not the symptom diary.[4]

The drug ladder — name the drug, name the dose

The ladder is real, but it bends. In low-risk disease you climb it (5-ASA, then steroids, then immunomodulator, then biologic). In high-risk Crohn's — young, stricturing, penetrating, perianal, extensive small-bowel — you go top-down: an early biologic with an immunomodulator prevents the irreversible structural damage that step-up allows. The prognostic argument now overrides the historical sequence.[4][6]

5-aminosalicylates (mesalazine). First-line for mild-moderate UC and maintenance after remission: mesalazine 2.4 to 4.8 g daily orally (active), 2 g daily (maintenance); mesalazine 1 g rectally nightly (suppository) for proctitis, 2 g enema nightly for left-sided disease; combination oral plus rectal beats either alone. Sulphasalazine 1 g two to four times daily is cheaper but limited by sulfa side-effects (rash, haemolysis, oligospermia). The awkward truth examiners test: 5-ASA is largely ineffective in Crohn's — budesonide is the first-line for mild-moderate ileocaecal Crohn's.[1]

Corticosteroids — induction only, never maintenance. Prednisolone 40 to 60 mg daily for 1 to 2 weeks then taper by 5 mg weekly for active flares; budesonide 9 mg daily for mild-moderate ileocaecal Crohn's (high first-pass metabolism limits systemic toxicity); IV hydrocortisone 100 mg every 6 hours for ASUC. Re-treating chronic steroid-dependent disease with longer steroids is the error — add a steroid-sparing immunomodulator instead.[1]

Thiopurines (azathioprine, mercaptopurine). Azathioprine 2 to 2.5 mg/kg daily (or mercaptopurine 1 to 1.5 mg/kg daily) for maintenance of steroid-induced remission in UC and Crohn's, steroid-sparing, and post-operative Crohn's prophylaxis. Check TPMT activity first (low activity means fatal myelosuppression); NUDT15 genotyping in East Asian patients; baseline and weekly FBC and LFTs for 4 weeks then monthly. Onset is 6 to 12 weeks — overlap with steroids. Watch for marrow suppression, hepatotoxicity, pancreatitis, and EBV-driven lymphoma.[1]

Methotrexate. Subcutaneous or intramuscular, 25 mg weekly for induction then 15 mg weekly (oral bioavailability is unreliable); add folic acid 5 mg weekly on a non-methotrexate day. Stronger evidence in Crohn's than thiopurines; teratogenic — strict contraception in women and men for at least 3 months after stopping.[1]

Anti-TNF biologics. The class that transformed the field after Targan's 1997 infliximab trial proved anti-TNF worked in refractory Crohn's.[5]

• Infliximab 5 mg/kg IV at weeks 0, 2, 6 then every 8 weeks — moderate-severe CD or UC refractory to steroids and immunomodulators, and rescue in ASUC. Combination with azathioprine beats either alone for steroid-free remission (SONIC, UC SUCCESS), at the cost of higher infection risk in older patients.[5]
• Adalimumab 160 mg subcutaneously at week 0, 80 mg at week 2, then 40 mg every 2 weeks — self-administered at home; effective in CD and UC. Golimumab (UC) and certolizumab (CD) are alternatives.[2][3]

The newer biologics and small molecules. Vedolizumab 300 mg IV at 0, 2, 6 then every 8 weeks — gut-selective anti-alpha-4-beta-7 integrin with an excellent safety profile; first-line biologic in older or infection-prone patients. Ustekinumab (anti-IL-12/23) and risankizumab (anti-IL-23) target the Th17 axis. Tofacitinib 10 mg twice daily for 8 weeks then 5 mg twice daily — an oral JAK inhibitor for moderate-severe UC, with a boxed warning for VTE, major cardiac events and malignancy in patients over 50 with cardiovascular risk factors. Upadacitinib (JAK1) and ozanimod (S1P modulator) extend the oral options.[1]

Before any biologic or immunosuppressant: exclude latent TB (IGRA and chest radiograph), hepatitis B and HIV; verify and update vaccinations (varicella, hepatitis B, MMR, HPV, influenza, pneumococcal); counsel on the methotrexate and JAK-inhibitor contraception rules.[1]

Surgery — when the drugs stop working

Surgery in UC is curative of the colonic disease; surgery in Crohn's is palliative of the segment. That one-sentence contrast is viva gold and it reframes the whole conversation with the patient.[1]

UC — restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA). Usually staged (subtotal colectomy with end ileostomy, then pouch construction, then ileostomy closure). It cures the colitis and removes most of the cancer risk, at the price of chronic pouchitis in 30 to 50 percent and pouch failure in 5 to 10 percent, and a real fertility cost in young women (consider a staged procedure). An end ileostomy alone is the safer choice for frail or steroid-dependent patients.[1]

Crohn's — segmental resection of the diseased bowel (ileocolonic, segmental colonic, or strictureplasty to spare bowel length); perianal disease needs seton drainage, fistulotomy, LIFT or advancement flap. The line every candidate must reproduce: recurrence at the anastomosis is the rule (Rutgeerts i0 to i4) — 70 to 90 percent endoscopic recurrence within a year — which is why post-operative prophylaxis (thiopurine, anti-TNF, or anti-IL23 in high-risk patients) is now standard, not optional.[2]

Complications — and the preventable-harm list

The complications separate into the acute structural, the chronic systemic, and the long-term malignant — and most are preventable.[1][3]

[1]

Colorectal cancer is the long shadow of colonic IBD. Risk rises with duration, extent, concomitant PSC, prior dysplasia, and a family history of colorectal cancer. Surveillance colonoscopy with chromoendoscopy begins at 8 to 10 years of extensive colitis (sooner and yearly if PSC), because early dysplasia is flat and easy to miss with white light alone.[1]

How IBD patients come to harm — the preventable list:[1]

• Death from unrecognised ileocaecal TB treated as Crohn's with anti-TNF — the preventable death.
• Delayed colectomy in fulminant ASUC because "the steroids might still work" on day 7.[1]
• A pelvic abscess given anti-TNF without drainage — sepsis on immunosuppression.[1]
• A patient with 10 years of extensive UC who never had a surveillance colonoscopy, presenting with cancer.
• A young woman left on long steroids for "steroid-dependent disease" because no one added a thiopurine — osteoporosis, diabetes, adrenal suppression.
• A flare loaded with immunosuppression without a stool test for C. difficile or CMV.
• VTE in a bed-bound flare because LMWH was forgotten.

Special populations

Pregnancy. Outcome is best when the disease is in remission at conception — active disease raises preterm birth and low birth weight. Continue maintenance therapy through pregnancy (5-ASA, azathioprine and infliximab are acceptable; give the last infliximab at 16 to 20 weeks to minimise neonatal exposure). Methotrexate is teratogenic — stop before conception in women and men. Vaginal delivery is acceptable except with active perianal disease or an IPAA (pouch-vaginal fistula risk). Plan mode of delivery in a joint obstetric-gastroenterology clinic.[1]

The elderly. More left-sided or distal disease, more comorbidity, higher infection risk on immunosuppression — vedolizumab is favoured for its gut-selective safety profile. Watch polypharmacy, falls, and colorectal cancer screening.[1]

Post-operative Crohn's. Recurrence is the rule. High-risk patients (smoking, penetrating disease, prior resection, two or more predictors) get biologic prophylaxis within 4 weeks of surgery; lower-risk patients get ileocolonoscopy at 6 to 12 months, with prophylaxis started if Rutgeerts i2 or higher.[2]

Children. Phenotype is more severe — extensive UC, ileocolonic Crohn's with upper GI involvement, rapid progression. Growth failure and delayed puberty may be the first presentation. Exclusive enteral nutrition for 6 to 8 weeks is as effective as steroids for inducing remission in mild-moderate paediatric Crohn's — and it preserves growth. Consider monogenic IBD in infantile-onset disease.[1]

The evidence — guidelines and the one trial to name

The framework is built on the ACG UC guideline (Rubin 2019), the ECCO Crohn's and UC therapeutics guidelines (Torres 2020; Raine 2022), and the STRIDE and STRIDE-II treat-to-target initiatives (Peyrin-Biroulet 2015; Turner 2021).[1][2][3][4][6]

The single trial examiners expect you to cite is Targan et al., NEJM 1997 — the original infliximab induction trial in moderate-severe Crohn's, the study that proved anti-TNF worked and opened the entire biologic era.[5]

[1]

The mantra: Crohn is transmural, skip and not cured by the knife; UC is mucosal, continuous and cured by it — and the smoking goes the opposite way in each.[1][2]

Ward-round test — three stems

Stem 1 — the young man from the top of the topic (answer)

A 24-year-old with four months of diarrhoea, right-lower-quadrant pain, weight loss, a perianal skin tag, raised CRP and calprotectin 480, negative stool culture. What is the next step, and how do you decide between a ladder and a biologic? Model: This is suspected Crohn's until proven otherwise. The next step is ileocolonoscopy with segmental biopsies, magnetic resonance enterography, and pelvic MRI for the perianal disease — plus exclusion of infection (including TB: IGRA and chest radiograph) and CMV. The phenotype decides the strategy: if imaging shows stricturing, penetrating or extensive small-bowel disease in a young patient, go top-down with an anti-TNF plus an immunomodulator — the prognostic argument overrides the step-up ladder. If it is mild inflammatory ileocaecal disease, budesonide 9 mg daily with a steroid-sparing thiopurine is reasonable. Treat to a target: clinical plus biochemical (CRP, calprotectin) plus endoscopic (SES-CD under 3) remission at 6 to 12 months.[2][4]

Stem 2 — the UC flare that turns severe (answer)

A 32-year-old with known extensive UC presents with 9 bloody stools in 24 hours, pulse 108, temperature 38.1, Hb 98 g/L. What do you do in the first 12 hours? Model: She meets Truelove and Witts criteria for acute severe UC — above 6 bloody stools per day plus fever, tachycardia and anaemia. Admit, nil by mouth, two large-bore cannulae, FBC, U&E, CRP, LFTs, albumin, group and save, blood cultures, stool C. difficile PCR and enteric culture; exclude CMV on colonic biopsy PCR; plain abdominal film to exclude toxic megacolon. Start IV hydrocortisone 100 mg every 6 hours, IV crystalloid and electrolytes (potassium, magnesium), VTE prophylaxis with LMWH (enoxaparin 40 mg SC daily), transfuse to Hb above 100 g/L. Reassess on day 3 with the Travis index (CRP above 45 mg/L AND stools above 8/day predicts steroid failure) — if positive, rescue with infliximab 5 mg/kg IV the same day, with ciclosporin 2 mg/kg/day IV as the alternative. The colorectal surgeon is informed from admission; failure of rescue by day 5 to 7 means subtotal colectomy with end ileostomy.[1][3]

Stem 3 — the smoking paradox as a viva trap (answer)

An examiner asks: "A 30-year-old smoker is diagnosed with Crohn's ileitis. What single non-pharmacological intervention has the biggest effect on his disease course, and why is the answer the opposite in ulcerative colitis?" Model: Smoking cessation. In Crohn's, smoking roughly doubles the risk of stricturing and penetrating complications, raises the surgical and re-operation rate, and accelerates post-operative recurrence — quitting is a genuine disease-modifying intervention. In ulcerative colitis the relationship inverts: current smokers have about half the risk, and quitting raises it, so UC often declares itself in the year after a patient stops. Treat smoking cessation as a therapy in Crohn's, not a lifestyle footnote.[1]

References

1. [1]Rubin DT, Ananthakrishnan AN, Siegel CA, et al. ACG Clinical Guideline: Ulcerative Colitis in Adults Am J Gastroenterol, 2019.PMID 30840605
2. [2]Torres J, Bonovas S, Doherty G, et al. ECCO Guidelines on Therapeutics in Crohn's Disease: Medical Treatment J Crohns Colitis, 2020.PMID 31711158
3. [3]Raine T, Bonovas S, Burisch J, et al. ECCO Guidelines on Therapeutics in Ulcerative Colitis: Medical Treatment J Crohns Colitis, 2022.PMID 34635919
4. [4]Turner D, Ricciuto A, Lewis A, et al. STRIDE-II: An Update on the Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE) Initiative of the International Organization for the Study of IBD (IOIBD): Determining Therapeutic Goals for Treat-to-Target strategies in IBD Gastroenterology, 2021.PMID 33359090
5. [5]Targan SR, Hanauer SB, van Deventer SJ, et al. A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn's disease. Crohn's Disease cA2 Study Group N Engl J Med, 1997.PMID 9321530
6. [6]Peyrin-Biroulet L, Sandborn W, Sands BE, et al. Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining Therapeutic Goals for Treat-to-Target Am J Gastroenterol, 2015.PMID 26303131