One cup of coffee drunk with a meal reduces non-haem iron absorption by 39 %. That figure comes from a dual-isotope study published in the American Journal of Clinical Nutrition in 1983, and it is the one coffee effect that specifically concerns vegetarians. The gastric acidity everyone talks about carries far less weight than its reputation suggests.

Coffee blocks plant iron, and the measurement is precise

The reference study gave coffee to iron-replete participants alongside meals labelled with two different isotopes, which measures real absorption instead of inferring it. The numbers are specific and instructive.

A cup of coffee taken with a meal reduced iron absorption by 39 % compared with the same meal without coffee. Tea, a known potent inhibitor, managed 64 %. With a meal built from semi-purified ingredients, absorption fell from 5.88 % to 1.64 % with drip coffee and to 0.97 % with instant coffee. Doubling the strength of the instant coffee pushed it down to 0.53 %.

The most useful practical point concerns timing. No reduction occurred when coffee was consumed one hour before the meal. Drunk one hour after, however, it inhibited absorption just as much as drinking it during the meal. The effect depends on coffee and iron being in the digestive tract together, and it persists past the end of the meal.

The mechanism runs through polyphenols, including coffee’s chlorogenic acids, which form poorly absorbable complexes with non-haem iron. It affects plant iron exclusively, which accounts for all dietary iron in a plant-based diet and the large majority of it for people who also eat dairy and eggs.

Why this matters more for vegetarians

Plant iron is not meat iron. Haem iron from meat is absorbed at a fairly stable rate and is largely indifferent to inhibitors. Non-haem iron, the kind in lentils, oats, tofu and spinach, is absorbed at a highly variable rate that depends heavily on what else is on the plate and in the cup.

A worked example from the Swiss Food Composition Database. A meal of 60 g dry red lentils, 60 g dry parboiled rice, 100 g raw red pepper and an onion supplies 4.0 mg of iron and 154 mg of vitamin C. That is a good meal from an iron standpoint, because the vitamin C in the pepper strongly boosts non-haem iron absorption, and the grain-with-pulse pairing helps too, as our page on combining grains and legumes explains. Coffee at the end of the meal works in exactly the opposite direction.

So the coffee question lands differently depending on what you eat. For someone whose iron comes mainly from red meat, an after-dinner coffee is a minor annoyance. For someone whose iron comes entirely from pulses and wholegrains, it is the easiest lever available. Our page on plant-based iron absorption covers the other enhancers and inhibitors.

Acidity: what the measurements actually show

Coffee’s pH sits around 5, which is genuinely acidic on the scale. The stomach, however, operates at a pH between roughly 1.5 and 3.5 between meals, hundreds of times more acidic. Pouring a pH 5 liquid into that environment means adding something considerably less acidic than what is already there.

Measured coffee acidity depends mostly on which acids are present. A study in Food Chemistry in 2024, covering eight different brewing methods, showed that chlorogenic acids dominate pH and titratable acidity, while perceived sourness correlates more closely with organic acid concentration. What you taste as acidic and what the instrument records as acidic are not quite the same thing.

On roasting, a study in Scientific Reports in 2024 tracked titratable acidity inside a 5 kg commercial drum roaster: it invariably peaks at first crack, then decays back to its original value by second crack. The idea that a dark roast gives a less acidic coffee is directionally right, but the dynamics are less linear than the usual telling suggests.

Reflux: the effect is real, small, and not where you think

On gastro-oesophageal reflux we have solid data, and it does not support the standard advice.

A meta-analysis in Clinical and Translational Gastroenterology in 2026, pooling 40 studies and 122,074 patients of whom 85,400 were coffee drinkers, found reflux in 34.9 % of coffee drinkers versus 30.7 % of non-drinkers, an odds ratio of 1.18 with a confidence interval of 1.03 to 1.36. The association is statistically significant, but the authors themselves write that the magnitude is of unclear clinical significance and that routine coffee avoidance as universal lifestyle advice for reflux needs further evaluation. No association was found with Barrett’s oesophagus.

A randomised crossover study in the European Journal of Gastroenterology & Hepatology in 1999 measured 24-hour oesophageal pH in seven reflux patients and eight healthy controls, comparing 280 ml of paper-filtered coffee with 280 ml of warm water. Coffee had no effect on postprandial acid reflux time or on the number of reflux episodes, in either group. The only measured effect was for coffee taken during fasting by reflux patients, which raised the percentage of acid reflux time. Lower oesophageal sphincter pressure was unchanged.

One finding matters most for anyone looking for a fix: a double-blind study in Alimentary Pharmacology & Therapeutics in 1997, with seventeen reflux patients, compared 300 ml of regular coffee against 300 ml of decaffeinated coffee taken with a standardised breakfast. Reflux fraction time fell from a median of 17.9 % to 3.1 % with the decaffeinated version. In that protocol the culprit is caffeine, not acidity.

That changes the whole approach. If you have reflux and coffee bothers you, switching to a darker roast or to cold brew targets the wrong variable. Decaffeinated coffee, or coffee with a meal rather than on an empty stomach, is what the evidence points to.

The acid-base story does not hold up

One claim deserves to be dismissed plainly, because it circulates everywhere: eating so-called alkalising foods does not neutralise coffee acidity in the stomach, and does not change blood pH.

Arterial blood pH is held between 7.35 and 7.45 by buffer systems, by breathing and by the kidneys. That control does not depend on what you had for lunch. A sustained departure from that range is a medical condition, not a dietary outcome. Spinach and cucumber are excellent foods for other reasons; they do not buffer your stomach.

What is true is that food in the stomach changes how coffee feels, through simple dilution and slower gastric emptying. Having coffee with something substantial rather than on an empty stomach is sensible advice. It just has nothing to do with any acid-base balance.

The filter: the one brewing choice with a measured health effect

If a single preparation variable deserves your attention, it is neither the grind nor the water temperature nor how long you let the cup stand. It is the paper filter.

Coffee contains two diterpenes, cafestol and kahweol, which raise blood cholesterol. Paper retains them; other methods do not. A study in Nutrition, Metabolism and Cardiovascular Diseases in 2025 measured their concentrations in Swedish workplace machine coffee and compared them with home brews.

Method	Median cafestol	Median kahweol
Paper-filtered, home brewed	12 mg/L	8 mg/L
Liquid-model machines	8 mg/L	7 mg/L
Workplace brewing machines	176 mg/L	142 mg/L
Boiled coffee	939 mg/L	678 mg/L
Boiled coffee poured through a fabric filter	28 mg/L	21 mg/L

French presses and percolators landed around 90 mg/L of cafestol and 70 mg/L of kahweol. Some espresso samples reached 2447 mg/L of cafestol. The authors’ conclusion is blunt: insufficiently filtered coffee consumed during working hours could be an overlooked cardiovascular factor.

The gap between 12 and 939 mg/L is a factor of seventy. No other brewing decision has consequences on that scale. And the same study shows a fabric filter already solves most of the problem, cutting 939 down to 28 mg/L.

How much caffeine, and from what point

The European Food Safety Authority published a scientific opinion on caffeine safety in 2015. The benchmarks that emerge are clear.

Single doses up to 200 mg, roughly 3 mg per kilogram of body weight for a 70 kg adult, raise no safety concerns for healthy adults. Daily intakes from all sources up to 400 mg, about 5.7 mg per kilogram, raise no concerns for the general adult population, with the exception of pregnant women. For them the figure is 200 mg per day from all sources. For children and adolescents, the opinion uses 3 mg per kilogram per day as a basis.

Two details matter in practice. The opinion counts all sources: tea, dark chocolate, cola drinks and energy drinks all add to the coffee. And while 400 mg roughly corresponds to four cups of filter coffee or five to six espressos, variation between beans, doses and methods is wide enough that counting precisely by the cup means little.

Which common advice survives, and which does not

Still valid: drinking coffee with food rather than fasting if you are prone to reflux, since the 1999 study found the effect precisely in the fasting condition; paper filtering if cholesterol is a concern; cutting caffeine rather than acidity if coffee triggers reflux symptoms; respecting the 400 mg daily and 200 mg pregnancy benchmarks.

Does not survive scrutiny: the idea that coffee disturbs a bodily acid-base balance; the idea that a high-fibre diet speeds transit enough to change how coffee acidity lands; the idea that letting coffee stand makes it less irritating, since organic acids do not evaporate out of a cup on the table; and the claim that cinnamon tempers acidity, which circulates without measurement behind it.

Switching to cold brew is more nuanced than it looks. Cold extraction does produce a less bitter cup that is often perceived as smoother, which is a legitimate taste argument. But if your problem is reflux, caffeine is the target, and a concentrated cold brew can contain as much as filter coffee or more.

The question that matters: when to drink it

For a vegetarian, how the day is organised matters more than which beans you buy. The rule comes straight from the 1983 iron study.

Coffee taken one hour before a meal showed no effect on iron absorption. Coffee taken during or one hour after the meal inhibited it. The window to protect is the one around iron-rich meals, not the whole day.

In practice that gives a simple pattern. Morning coffee works fine before or with a low-iron breakfast, or well away from lunch. The after-lunch coffee, the most common habit of all, is precisely the one that costs the most iron: pushing it back an hour or two does not help, since the inhibition measured at one hour matched the inhibition measured during the meal.

If you change only one habit, change that one: drink the midday coffee in the middle of the afternoon, far from the next meal. It is the only gain in this article that rests on a direct measurement in humans.

Strengthening iron on the plate

Coffee is only one factor. In a vegetarian diet, two levers work in the opposite direction.

Vitamin C eaten in the same meal markedly increases non-haem iron absorption. The Swiss database figures make the point: 100 g of raw red pepper supplies 149 mg of vitamin C, against 37 mg for 100 g of orange juice. Half a raw pepper in the lentil salad beats a glass of juice.

Soaking and sprouting pulses lowers their phytate content, the other major inhibitor of plant iron. Our page on soaking and sprouting legumes gives the method and the timings, and the page on optimising plant iron absorption gathers the moves that matter at mealtime.

Beyond that, meeting requirements comes down to variety of sources, covered in our complete guide to plant proteins and in the page on tips for better iron absorption.

Coffee itself, nutritionally

A cup of unsweetened black coffee supplies 2 kcal per 100 g according to the Swiss database, with 26 mg of potassium and traces of magnesium and calcium. Nutritionally negligible, which is good news: coffee displaces nothing.

Adding a plant drink shifts the arithmetic modestly. Unsweetened white coffee comes to 20 kcal per 100 g. The real issue lies elsewhere: if you use a plant drink, its calcium content depends entirely on fortification. A fortified plain soya drink supplies 120 mg of calcium per 100 g against 12 mg for the unfortified version, a tenfold gap. That information sits in the ingredients list, not on the front of the carton, and our page on reading a nutrition label explains where to look. For anyone avoiding dairy, daily intakes are covered in the page on calcium and magnesium without dairy.

On sugar, the Swiss dietary recommendations from the Federal Food Safety and Veterinary Office place sweetened drinks in the eat-sparingly category, at most one portion a day. Two sweetened coffees a day contribute there without anyone counting them. The same recommendations set 1 to 2 litres of unsweetened drinks daily, a topic covered in our page on hydration over twenty-four hours.

Coffee, exercise and iron on a plant-based diet

For people who train, two things overlap that rarely appear together in advice columns. Caffeine is widely taken before exercise as a performance aid, and endurance sport happens to be one of the situations where iron requirements rise and low iron status turns up more often.

The 1983 data do not call for giving anything up; they call for an order of operations. Pre-workout coffee usually sits well away from main meals anyway, which puts it in the window where no inhibition was measured. The problematic cup is the one drunk with the recovery meal straight after a session, precisely when an iron-rich plate is on the table.

The practical fix is unglamorous: keep the pre-workout coffee, and move the post-workout one back by a couple of hours. Nothing about training changes the polyphenol chemistry, and nothing about the chemistry rules out caffeine as an ergogenic aid. They simply need to sit at different points in the day.

Self-assessment of a deficiency is no substitute for blood work, and iron supplementation without a measured deficit is not harmless. If your training load is high and you eat no meat, a ferritin measurement tells you more than any dietary rule of thumb.

Common mistakes

Hunting for a less acidic coffee to fix reflux. The 1997 decaffeination study shows the decisive variable is caffeine, not acidity.

Drinking coffee right after a meal on the theory that food protects you. It protects against perceived irritation, not against iron inhibition: one hour after the meal, inhibition matched the level measured during it.

Counting on alkalising foods. Blood pH does not depend on the menu, and nothing you eat buffers the stomach in the sense these tips imply.

Treating workplace machine coffee as filter coffee. The Swedish measurements give 176 mg/L of cafestol for brewing machines against 12 mg/L for home paper-filtered coffee.

Doubling the instant coffee. In the 1983 study, that single change was what pushed iron absorption below 1 %, down to 0.53 %.

Counting only coffee cups for caffeine. The EFSA opinion works from all sources, tea, chocolate and energy drinks included.

Adapting to different situations

With diagnosed iron deficiency or low ferritin, separating coffee from meals becomes a concrete measure, worth discussing alongside any supplementation with whoever is following your case.

During pregnancy, two constraints stack up: the EFSA limit of 200 mg of caffeine a day, and markedly higher iron requirements. This is where moving coffee away from meals earns the most.

For people with heavy periods, iron losses are higher and the same logic applies.

For anyone prone to reflux, the experiment to run is decaffeinated coffee, not a darker roast, and coffee at the table rather than fasting.

For adolescents, the EFSA opinion uses 3 mg per kilogram per day, about 150 mg for a 50 kg person, energy drinks included.

If your cholesterol is being monitored, moving from boiled coffee, a French press or a workplace machine to paper filtering is the single most effective change available to you.

Limits of this evidence

The 1983 dual-isotope study used small numbers of iron-replete participants, and it measured absorption from a given meal rather than iron status over months. The direction and order of magnitude are robust; extrapolating to a predictable deficiency in any one person is not.

The 2026 reflux meta-analysis pools observational studies with high heterogeneity, reflected in an I² of 89.38. Its authors stay cautious about the clinical reach of an odds ratio of 1.18.

The 1997 and 1999 reflux trials involved seventeen and fifteen participants. These are careful pH-measurement protocols, but small ones.

The cafestol measurements come from a Swedish sample of workplace machines: absolute values depend on equipment and local habits, while the filtered-versus-unfiltered gap is the robust result.

The Swiss database values are generic averages. A real bag of lentils or carton of plant drink will differ, and calcium fortification varies between brands.

Frequently asked questions

Does decaffeinated coffee also block iron? Yes. The inhibition works through polyphenols, present in decaffeinated coffee as much as in regular. Decaffeination addresses reflux, not iron.

Is tea worse than coffee? In the 1983 study, yes: 64 % reduction for tea against 39 % for coffee, on the same meal.

How long should I wait after a meal? The study found identical inhibition one hour after the meal. Waiting longer is therefore necessary, and drinking coffee before the meal is the option documented as neutral.

Does plant milk in coffee change the iron picture? Nothing suggests it cancels the polyphenol effect. Calcium in quantity is itself a modest inhibitor of non-haem iron, so a fortified drink does not help on this specific point.

Is espresso less of a problem because it is small? For iron, what counts is the polyphenol load, lower in 30 ml than in 200 ml. For cafestol, the Swedish study measured espresso samples up to 2447 mg/L, so espresso is not filtered coffee.

Is cold brew gentler on the stomach? It is often perceived as smoother. The reflux trials did not test cold brew, and the variable those trials identified is caffeine, which a concentrated cold brew carries in quantity.

Should vegetarians give up coffee? Nothing in this evidence justifies that. Moving cups away from iron-rich meals removes most of the documented downside.

Does weaker coffee inhibit less? Yes, the effect is concentration dependent: in the 1983 study, doubling instant coffee strength reduced iron absorption further.

Sources

• Morck T.A., Lynch S.R., Cook J.D., Inhibition of food iron absorption by coffee, American Journal of Clinical Nutrition, 1983 (PMID 6402915)
• Association of coffee intake with risk of gastroesophageal reflux disease and complications: a systematic review and meta-analysis, Clinical and Translational Gastroenterology, 2026 (PMID 41677121)
• Pehl C. et al., Effect of coffee on gastro-oesophageal reflux in patients with reflux disease and healthy controls, European Journal of Gastroenterology & Hepatology, 1999 (PMID 10563539)
• Pehl C. et al., The effect of decaffeination of coffee on gastro-oesophageal reflux in patients with reflux disease, Alimentary Pharmacology & Therapeutics, 1997 (PMID 9218070)
• Cafestol and kahweol concentrations in workplace machine coffee compared with conventional brewing methods, Nutrition, Metabolism and Cardiovascular Diseases, 2025 (PMID 40089392)
• Comprehensive investigation of coffee acidity on eight different brewing methods, Food Chemistry, 2024 (PMID 38810441)
• The effect of roast profiles on the dynamics of titratable acidity during coffee roasting, Scientific Reports, 2024 (PMID 38589450)
• EFSA, Scientific opinion on the safety of caffeine, 2015 (200 mg single dose, 400 mg daily, 200 mg in pregnancy)
• Federal Food Safety and Veterinary Office, Swiss dietary recommendations
• Swiss Food Composition Database
• World Health Organization, healthy diet fact sheet