Few questions in modern diet culture are asked more often, and few get vaguer answers. Does black coffee break a fast? The usual reply is that it is fine because it has almost no calories, sometimes followed by a warning that any amount of anything spikes insulin. Both answers skip the step that decides the question: what you are fasting for. A fast before a blood test, a fast for weight control and a fast for the metabolic effects people read about online are different things with different thresholds, and coffee affects them differently. The research below cannot tell you what your goal is, but it can tell you what coffee measurably does.
First, what fasting is being measured
In clinical research, fasting usually means no food or caloric drink for a set period, typically 8 to 12 hours before a blood sample. In intermittent fasting practice, it means confining eating to a window, often 8 hours, with the remaining 16 hours free of calories. In the laboratory work on fasting biology, researchers are interested in insulin staying low, fat being used for fuel, and cellular maintenance processes that are mostly studied in animals and cell cultures rather than people.
Black coffee lands differently on each. Brewed black coffee contains a very small amount of energy from trace compounds, well under 5 calories a cup, which is nutritionally negligible. That settles the clinical fasting question and most of the weight-control one. It does not settle whether coffee changes the metabolic state a faster is aiming for, and that is where the real evidence sits.
What coffee does to insulin and glucose on its own
The most direct test comes from Reis and colleagues[1], who ran a randomised, single-blind cross-over study in 17 healthy men. Each man completed five trials, consuming decaffeinated coffee, regular coffee, regular coffee with sugar, water, or water with sugar. An hour later they were given an oral glucose tolerance test with a labelled glucose tracer, allowing glucose metabolism to be modelled rather than just measured at the surface.
The result was the opposite of the usual assumption. Decaffeinated coffee improved insulin sensitivity compared with water, without changing incretin hormone levels. Glucose, insulin, C-peptide and the incretin hormones GLP-1 and GIP were not statistically different after regular or decaffeinated coffee compared with water, and regular coffee with sugar did not differ from water with sugar. The authors concluded there was no short-term adverse effect on glucose homeostasis.
So on its own, on an empty stomach, coffee did not produce the insulin response that fasting advice warns about. If anything the decaffeinated version improved the measure people are trying to protect.
What coffee does to the meal that follows
Then the picture changes, and this is the part almost no fasting article covers. Moisey and colleagues[2] had 10 healthy men complete four trials in random order, taking caffeinated or decaffeinated coffee before either a high or low glycaemic index meal. Compared with decaf, caffeinated coffee with the high glycaemic index meal produced 147% greater area under the curve for glucose, 29% greater for insulin and 40% greater for C-peptide. With the low glycaemic index meal the gaps were larger still, at 216%, 44% and 36%. Insulin sensitivity was 40% lower after caffeinated coffee with the high glycaemic index meal than after decaf.
Beaudoin and colleagues[3] pushed the same design further, combining an oral lipid challenge with caffeinated coffee and finding that the two additively decreased glucose tolerance in healthy men. The direction is consistent across this group's work: caffeine taken before carbohydrate makes the body handle that carbohydrate worse for a few hours.
This matters for time-restricted eating in a way that is rarely discussed. If you break a 16 hour fast with a large carbohydrate meal shortly after a strong coffee, the evidence suggests the glucose response to that meal will be larger than it would have been otherwise.
Short-term harm, long-term benefit
The apparent contradiction with coffee's well-known association with lower type 2 diabetes risk was addressed directly by Reis and colleagues[4] in a systematic review of 8 clinical trials covering 247 participants. Splitting the trials by duration, short-term studies of 1 to 3 hours showed caffeinated coffee may increase the area under the curve for glucose, while long-term studies of 2 to 16 weeks showed caffeinated coffee may improve glycaemic metabolism by reducing the glucose curve and increasing the insulin response.
Their summary is worth keeping in mind whenever you read a single coffee-and-glucose study: caffeinated coffee may lead to unfavourable acute effects, with an improvement over longer periods. A person drinking coffee daily for years is living in the long-term column, not the acute one. The topic is covered at length in our article on coffee and type 2 diabetes.
Does coffee suppress hunger enough to matter?
The other reason people drink coffee while fasting is appetite. This has been tested properly. Schubert and colleagues[5] ran a cross-over trial comparing coffee, decaffeinated coffee, caffeine and placebo, measuring appetite sensations, gastric emptying and energy intake at a subsequent meal. Energy intake did not differ significantly between the trials, at roughly 2,118 kilojoules for placebo, 2,128 for decaf, 2,287 for caffeine and 2,016 for coffee. Appetite ratings and plasma glucose showed no significant differences between treatments beyond the effect of time, and gastric emptying was not significantly different either.
In other words, in this trial coffee did not measurably blunt hunger or reduce what people ate later. The subjective sense that coffee holds off breakfast is real for many people, but it did not translate into eating less when it was measured. Longer-term effects on energy balance were not tested and the authors said so.
Caffeine, fat burning and the claim about metabolism
Fasting advice often pairs coffee with fat burning, and there is genuine older evidence behind the metabolic part. Acheson and colleagues[6] and Dulloo and colleagues[7] measured caffeine's effect on energy expenditure and substrate use, finding increases in metabolic rate and fat oxidation. Hursel and colleagues[8] pooled the caffeine and thermogenesis literature.
What those studies show is a modest, short-lived increase in energy expenditure, not a mechanism that empties fat stores. The weight research, covered in our article on coffee and weight loss, is the better guide to whether any of that adds up over time, and the short answer there is that it does not add up to much.
So does milk break a fast?
This is where the answer becomes arithmetic rather than biology. A splash of milk adds a small number of calories and a small amount of lactose and protein. Whether that breaks your fast depends on the threshold you are using, and there is no research establishing a calorie threshold at which the benefits of fasting stop, because the trials that show benefits from time-restricted eating were testing eating windows rather than exact calorie cut-offs.
There is one relevant measurement. Wong and colleagues[9] gave healthy adults decaffeinated coffee with milk and sugar added before a high glycaemic index meal and found it improved postprandial glycaemic and insulinaemic responses. That is the opposite of what a strict reading of fasting rules would predict, and it is a reminder that the body's response to a small pre-meal drink is not as simple as calories in equals fast broken.
Why the studies do not all agree
Not every study finds coffee affecting the next meal. Hätönen and colleagues[10] found coffee did not modify postprandial glycaemic and insulinaemic responses induced by carbohydrates. Gavrieli and colleagues[11] found that sex and body mass index modified the effect of increasing doses of caffeinated coffee on postprandial glucose and insulin, which is a clue as to why results differ between small trials of a handful of participants.
Dose, timing, what was eaten, who was tested and whether they were habitual coffee drinkers all move these numbers. Tolerance matters too: much of the acute impairment literature tested people given a substantial caffeine dose in controlled conditions, which is not the same as a habitual drinker's usual cup.
What about bulletproof coffee?
Coffee with butter or coconut oil blended in is popular in fasting circles on the argument that fat does not trigger insulin, so the fast continues. Two things in the research above bear on that. First, the drink plainly contains calories, often 200 or more, which ends any fast defined by energy intake. Second, Beaudoin and colleagues[3] found that an oral lipid challenge and caffeinated coffee additively decreased glucose tolerance in healthy men, meaning fat and caffeine together worsened the response more than either alone.
That was a lipid challenge in a laboratory rather than a mug of blended butter, so it is suggestive rather than a test of the drink. But it points the opposite way from the claim usually made for it, and no trial we found has tested fat-added coffee against black coffee during a fast for any outcome that matters.
Timing is the practical lever
Pulling the findings together, the actionable detail is not whether coffee is permitted but when it sits relative to food. The impairment measured by Moisey and colleagues[2] was in a meal taken shortly after coffee. Reis and colleagues[1] gave their glucose tolerance test an hour after coffee and found no adverse effect. The acute effect of caffeine is largest in the first hour or two, as described in our article on how long caffeine lasts.
For someone eating in a window, that suggests a coffee early in the fasting period is on much safer ground than a coffee immediately before breaking it. It also suggests that if you want coffee with your first meal, the decaf option is the one the evidence treats more kindly, since decaf was the comparison arm that consistently looked better in these trials.
The limits of this evidence
Almost every trial here is small, often 10 to 20 people, usually young healthy men, and almost always measuring hours rather than months. None of them tested intermittent fasting as practised, meaning a 16 hour fast with coffee in it, against the same fast without coffee. The autophagy and cellular-maintenance claims that circulate in fasting communities were not tested in any of these studies and are not assessed here, because the human evidence for them is thin. Nothing here measured weight, body composition or any disease outcome.
Reading this from the Philippines
Time-restricted eating has become popular in Philippine cities, and it runs into a specific local obstacle: the default coffee here is often sweet. Three-in-one sachets, bottled coffee drinks and café orders built on condensed milk or syrup all carry sugar, and the trials above that found no adverse effect were testing black coffee or decaf, not those. A sweetened coffee during a fasting window is a small meal by any definition used in this research. That is not an argument for or against fasting, which none of these studies evaluated, just a note that the answer to whether coffee breaks a fast depends far more on what is in the cup than on the coffee itself. For the unsweetened end of the local range, our guide to kapeng barako covers how it is traditionally brewed.
The practical answer
If you are fasting for a blood test, black coffee is almost certainly fine, but ask the clinic, since some tests are affected by caffeine rather than by calories. If you are fasting to control calories, black coffee is close to calorie free and does not, on the evidence, make you eat more later. If you are fasting for metabolic reasons, the finding worth knowing is not that coffee breaks a fast but that caffeine before a carbohydrate meal makes the glucose response to that meal considerably larger, so the more useful change may be spacing your coffee and your first meal apart rather than giving up either. For related reading see our article on coffee on an empty stomach and our article on coffee and blood sugar. Nothing here is medical advice, and anyone fasting with diabetes or on glucose-lowering medicine should take direction from their own clinician.