Heartburn is miserable and common, and coffee is near the top of almost every list of foods to avoid if you have it. Many people with reflux say that coffee makes their symptoms worse, and that experience is worth taking seriously. But when researchers have tried to measure whether coffee actually causes reflux across whole populations, the results have been inconsistent enough that at least one recent review questioned whether routinely telling everyone with reflux to give up coffee is justified. This article lays out what the studies found and why they keep disagreeing.
What reflux actually is
Gastro-oesophageal reflux happens when stomach contents, including acid, flow back up into the oesophagus, the tube that carries food from the mouth to the stomach. A ring of muscle at the bottom of the oesophagus, the lower oesophageal sphincter, normally acts like a one-way valve. Occasional reflux is normal. Gastro-oesophageal reflux disease, or GERD, is diagnosed when reflux causes troublesome symptoms, typically heartburn or regurgitation, at least weekly, or when it damages the lining of the oesophagus. Over many years, a minority of people with chronic reflux develop Barrett's oesophagus, a change in the lining that is a risk factor for oesophageal cancer.
The two main theories for why coffee might worsen reflux are that it increases stomach acid, and that it relaxes the sphincter so acid escapes more easily. Both have been tested directly.
The 1975 experiment that complicated the caffeine story
One of the most cited experiments on this question was published in the New England Journal of Medicine by Cohen and Booth[1]. At the time, the advice to avoid coffee was based partly on the known effects of caffeine, which stimulates stomach acid and had been reported to weaken the lower oesophageal sphincter. The researchers compared caffeine, regular coffee and decaffeinated coffee head-to-head in healthy volunteers.
Decaf produced a maximum acid output of 16.5 mEq per hour, similar to regular coffee at 20.9. Both were much higher than pure caffeine, at 8.4 mEq per hour, when compared on a cup-equivalent basis. So coffee raised stomach acid far more than its caffeine content alone would, and removing the caffeine did not remove much of the effect. The sphincter results were even more surprising: caffeine barely changed sphincter pressure, while both regular and decaf coffee significantly increased it, which would, if anything, make reflux less likely. The authors concluded that recommendations based on caffeine's known effects might bear little relation to what coffee or decaf actually do.
Measuring reflux directly, with and without caffeine
Acid output and sphincter pressure are indirect measures. A more direct approach is to put a small pH probe in the oesophagus and count how often acid actually flows back up. Wendl and colleagues, in Alimentary Pharmacology and Therapeutics[2], did this for three hours in 16 healthy volunteers who drank 300 ml of regular coffee, decaffeinated coffee or tap water with a standardised breakfast. Smaller groups also tested tea, decaffeinated tea, water with added caffeine, and coffee diluted to the caffeine concentration of tea.
Regular coffee caused significantly more reflux than tap water or tea, which did not differ from each other. Decaffeinating the coffee significantly reduced the reflux it caused. That might look like proof that caffeine is to blame, but the other arms of the study said otherwise: adding caffeine to water did not increase reflux, decaffeinating tea did not change it, and coffee diluted to tea's caffeine level still significantly increased reflux. The authors concluded that caffeine does not seem to be responsible, and that the effect must come from other components of coffee, which the decaffeination process partly removes.
This was a small study in healthy people over a single morning, but it measured the thing people actually care about, acid entering the oesophagus, and it is one of the few that tested caffeine and coffee side by side under the same conditions. It also helps explain why decaf helps some people with reflux even though decaf still stimulates stomach acid: the two effects are not the same.
Does the roast matter?
If compounds other than caffeine drive coffee's effect on the stomach, then the way beans are roasted, which changes their chemistry, might matter. Rubach and colleagues, in Molecular Nutrition and Food Research[3], compared two coffees with similar caffeine content in nine healthy volunteers: a standard medium-roast market blend and a dark roast blend. The dark roast contained more N-methylpyridinium, a compound formed during roasting, at 87 mg per litre compared with 29, and less chlorogenic acid, at 323 compared with 1,126 mg per litre. Gastric acid secretion was stimulated less effectively by the dark roast blend.
The study was very small and measured stomach acid rather than reflux or heartburn, so it cannot say whether a dark roast actually causes fewer symptoms. The authors themselves called for further studies. It is a plausible lead rather than a recommendation, and an interesting counterpoint to the common assumption that darker, stronger-tasting coffee must be harsher on the stomach.
That was a small physiological study, and later work on the sphincter has produced conflicting results. A review of coffee and gut function by Boekema and colleagues[4] summarised the picture two decades later: coffee does promote reflux and stimulates acid secretion, heartburn is the symptom most often reported after coffee, but studies of its effect on sphincter pressure disagree. It also found no association between coffee and dyspepsia, the general term for upper abdominal discomfort.
The first big meta-analysis: no clear link
If coffee reliably caused reflux, large population studies should show it. A meta-analysis by Kim and colleagues in Diseases of the Esophagus[5] combined 15 case-control studies published up to 2012. Overall, there was no significant association between coffee intake and reflux disease, with an odds ratio of 1.06. Breaking the studies down by how much coffee people drank, by study quality or by how coffee intake was measured did not reveal a link either.
There was one exception. When studies were split by how reflux was diagnosed, the group that used endoscopy, a camera examination of the oesophagus, did show a significantly higher odds ratio, while studies diagnosing reflux by symptoms alone did not. That is a curious result that the authors could not fully explain, and it is one reason the question has stayed open.
The prospective study that did find a link
Case-control studies compare people who already have a condition with people who do not, and they are vulnerable to recall bias. A stronger design follows people who are healthy at the start. Mehta and colleagues, writing in Clinical Gastroenterology and Hepatology[6], did this with the Nurses' Health Study II, following 48,308 women aged 42 to 62 who had no regular reflux symptoms and were not taking acid-suppressing medication.
Over 262,641 person-years, 7,961 women developed reflux symptoms at least once a week. After adjusting for other factors, women drinking more than six servings of coffee a day had a 34% higher risk than women drinking none, a hazard ratio of 1.34, and the trend across intake levels was significant. Tea was associated with a 26% higher risk and soda with a 29% higher risk at the same intakes. Crucially, the researchers found similar results when they split drinks by caffeine status, which again suggests caffeine is not the whole explanation. Milk, water and juice showed no association.
The study also estimated what swapping would do. Replacing two servings a day of coffee with two servings of water was associated with a hazard ratio of 0.96, a small reduction in risk. That is a modest effect, and it is worth noting how high the top intake category is: the clearest increase in risk was at more than six servings a day, far above what most people drink.
The largest and most recent pooled analysis
A 2026 meta-analysis by Muftah and colleagues in Clinical and Translational Gastroenterology[7] gathered 40 studies covering 122,074 people, of whom 85,400 drank coffee and 36,674 did not. Reflux disease was more common among coffee drinkers, 34.9% compared with 30.7%, giving an odds ratio of 1.18 that was statistically significant. There was, however, no significant association between coffee and Barrett's oesophagus.
The authors were careful about how to interpret this. They described the link as a small increase of unclear clinical significance, and said that routinely telling everyone with reflux to avoid or reduce coffee as a universal lifestyle change needs further evaluation. They also reported very high heterogeneity between studies, meaning the individual results varied a great deal, which limits how much weight a single pooled number can carry.
What genetic evidence adds
Observational studies struggle to separate coffee from everything else about coffee drinkers, including body weight and smoking, both of which are strongly linked to reflux. Yuan and Larsson[8] used Mendelian randomisation to test several suspected causes at once, using genetic data on 71,522 people with reflux disease and 261,079 without.
The clearest causal signals were for body weight and smoking. The odds of reflux were 1.49 times higher for each standard deviation increase in body mass index, and 1.41 times higher for each standard deviation increase in the likelihood of taking up smoking. Genetically predicted caffeine consumption showed a suggestive association, an odds ratio of 1.08 for each 80 mg increase, while coffee consumption itself did not show a significant association in their adjusted analyses. The authors described the caffeine finding as a possible role, not an established one.
Why the studies keep disagreeing
Several things make this question genuinely hard. Reflux symptoms are subjective and fluctuate. People who get heartburn after coffee tend to cut down, which can make coffee drinkers look less affected than they would be, and people with reflux may remember coffee as a trigger more readily than people without it. Coffee is often drunk with other known triggers, such as large meals, cigarettes or pastries. And coffee is not one exposure: strength, roast, temperature, milk and serving size vary enormously.
There is also a real possibility that the effect varies between people. A population average of no effect or a small effect is entirely compatible with some individuals having a strong, reliable trigger. That is the most sensible way to reconcile the lived experience of many people with reflux and the modest numbers in the large studies.
The genetic result also fits with the laboratory studies above. If caffeine played a large causal role, the pH-probe experiment should have found reflux rising when caffeine was added to water, and it did not. A small possible caffeine effect, sitting alongside larger effects from other coffee compounds, weight and smoking, is consistent with every study in this article.
The limits of the evidence
Almost everything here is observational, apart from small physiological studies of acid, reflux and sphincter pressure. The observational studies differ in how they define reflux, how they measure coffee, and which populations they study, and the recent meta-analysis had very high heterogeneity. The prospective cohort was made up of middle-aged American nurses. No long-term randomised trial has tested whether cutting coffee reduces reflux over months or years in people who have it.
What to take from this
The evidence does not support the idea that coffee is a major cause of reflux disease across the population. The most recent pooled analysis found a small association, the best prospective study found higher risk mostly at very high intakes, and genetic evidence suggests weight and smoking matter far more. Decaf is not automatically gentler on acid, because the early experiments found it stimulated acid almost as much as regular coffee.
An endoscopy study of 8,013 Japanese adults by Shimamoto and colleagues[9] points the same way for visible damage: coffee drinking was not associated with reflux oesophagitis or with non-erosive reflux disease once other factors were accounted for. For an individual with reflux, the population evidence leaves room for personal observation rather than blanket avoidance: noticing whether coffee reliably brings on symptoms, and whether timing, size of cup, decaf, or what you eat with it makes a difference. If you want to read about coffee's other effects on the stomach, see our article on coffee on an empty stomach and the article on how acidic coffee really is. Persistent heartburn, difficulty swallowing or weight loss need a doctor, not a coffee experiment.