Ground Truth · pointed at a thing everyone gets wrong
The Ulcer That Wasn't the Curry
For most of the twentieth century the answer was spicy food and stress. It was wrong. A peptic ulcer is, overwhelmingly, an infection (a spiral bacterium, Helicobacter pylori) or the slow erosion of long-term painkillers. Two Australians proved it by 1982; one of them drank a culture to make the point. The curry never started it.
But the myth has a kernel of truth, and honesty means keeping it. A vindaloo can make an ulcer you already have hurt; it just can't dig one. Those are two different questions, and almost every explainer blurs them. So here they are, split apart. Flip the four suspects on and off and watch which bar each one actually moves.
The suspects: tap to add each one
Only two suspects ever move the top bar: H. pylori and NSAIDs (aspirin, ibuprofen, naproxen and their kin). Spice and stress leave it flat: they cannot cause an ulcer. But flip them on and the bottom bar responds: chilli and a bad week really can make an existing ulcer hurt. That is the whole of the myth's truth, and no more of it.
How we know it's the bacterium
In 1982 the pathologist Robin Warren and the physician Barry Marshall, in Perth, found spiral bacteria colonising the inflamed stomachs of ulcer patients. The establishment didn't believe a bug could live in stomach acid. So in 1984 Marshall drank a Petri-dish culture of it; days later he had gastritis, biopsy-confirmed, and antibiotics cured him. The Koch's-postulates paper followed in 1985, and in 2005 the two shared the Nobel Prize in Physiology or Medicine. Cure the infection and the ulcer stops coming back, which no diet ever did.
The twist: the bacterium's share is falling
Here is what the fifty tidy explainers miss. The famous ">90% of ulcers are H. pylori" was a snapshot of the 1980s–90s, when the infection was everywhere. As sanitation improved and antibiotics became routine, H. pylori receded across the West, and its share of ulcers fell with it. What's left is dominated by NSAIDs and by idiopathic ulcers (neither bug nor pill), now a fifth to two-fifths of cases in North America. Drag the slider from the classic era to the modern West and watch the pie turn over.
Attributed cause of ulcers – duodenal
Endpoints are cited cohort figures; the slide between them is a straight-line interpolation; it shows the direction and size of the drift, not a specific calendar year.
The heat might even help. The single most-feared "irritant," capsaicin (the molecule that makes chilli hot) is under study as protective: at ordinary dietary doses it raises gastric mucosal blood flow, nudges up mucus and bicarbonate, and can inhibit acid, which in animal models speeds ulcer healing. (At very high doses it can aggravate damage, so this is "benefactor, in moderation," not a prescription.) The biology of the reversal hasn't changed since 1982. What drifts is the bookkeeping (which cause gets the blame), and that moves with place and time.
The check: every share recomputed in front of you
The pie's three slices always sum to 100%. Its two ends are cited cohort figures; the slider linearly interpolates between them. For the current site and era:
The two-bar board is a qualitative causal map straight from the literature, not a risk score: start-weight of spice = 0 and stress = 0 (neither can cause an ulcer), while their aggravate-weight > 0 (both can worsen one). H. pylori and NSAIDs carry weight on both bars.
Reproduce every number offline: node research/does-spicy-food-cause-ulcers/verify-does-spicy-food-cause-ulcers.mjs.
Classic vs. modern West, side by side
| share of ulcers | duodenal · classic | duodenal · modern W. | gastric · classic | gastric · modern W. |
|---|---|---|---|---|
| H. pylori | 92% | 5% | 75% | 12% |
| NSAIDs | 6% | 60% | 22% | 58% |
| Idiopathic | 2% | 35% | 3% | 30% |
Classic H. pylori shares from the historical attribution literature (duodenal 90–95%, gastric 70–90%; we state gastric at the ~70–80% lower cluster). Modern-West idiopathic shares from the 20–40% North-American range; the modern duodenal 5% is the endpoint of a 25-year single-centre decline (40%→5%). NSAID shares are the remainder and cohere with NSAID-dominated bleeding series (84% of gastric, 62% of duodenal bleeders on NSAIDs). Rows sum to 100%; a small H. pylori-and-NSAID overlap is folded in.
What's exactly true, what's a modelled path, and what's idealised
Exactly true. The etiology reversal is settled medicine: peptic ulcers are caused chiefly by H. pylori and long-term NSAID use; spicy food and ordinary psychological stress do not cause them. Marshall & Warren, 1982 discovery, 1984 self-experiment, 1985 Koch's-postulates paper, 2005 Nobel. Eradicating the infection prevents ulcer recurrence.
A dated snapshot, not a law. The ">90% of ulcers are H. pylori" figure is historical and geographic: true classically and still true in high-prevalence regions, but an overclaim for the modern West, where idiopathic and NSAID ulcers have risen sharply. That's the whole second layer: the biology is fixed, the attribution shares drift.
A modelled path. The slider's two ends are cited cohort numbers; the line between is a linear interpolation for readability, not a measured year-by-year series. Real cohorts scatter with country, decade, referral pattern and whether the series is all ulcers or only bleeding ones. Treat the endpoints as solid and the middle as illustrative.
Honest edges. "Idiopathic" means H. pylori- and NSAID-negative after testing: some are missed infection, occult NSAID/aspirin use, or rarer causes; it is a residual, not a single disease. Severe physiological stress (major burns, ICU illness) genuinely causes "stress ulcers," a distinct entity, not the everyday-worry "stress" of the myth, which is what the board's zero refers to. Capsaicin's protective effect is well-supported in animal and mechanistic studies and is dose-dependent; it is not a treatment claim.