A cup of black coffee beside a small espresso cup on a cafe table

Caffeine and Anxiety

The dose curve for caffeine and anxiety is steeper than most people think, and flatter than the headlines suggest — both at once. At 150 mg, about a cup and a half of coffee, a careful 2025 trial found no increase in self-reported anxiety at all, even in people with panic disorder. At 400 mg and above, about four cups, half of the panic disorder patients tested had a panic attack. The gap between those two numbers is where most of the real-life confusion lives, and it is also where the useful advice lives: the question is not whether caffeine causes anxiety, it is how much, and in whom.

No affiliate links on this page — just the papers.

The Dose Curve, in Plain Numbers

Before the details, here is the shape of the evidence. Low doses — around 150 mg — did not raise self-reported anxiety in a randomised trial of people with panic disorder and healthy controls. High doses — 400 to 750 mg, mostly 480 mg — triggered panic attacks in about half of panic disorder patients and almost no healthy controls. In healthy people without any psychiatric diagnosis, a meta-analysis found caffeine raised anxiety overall, with a much larger effect above 400 mg than below it.

Notice what is missing from that curve: the middle. The high-dose studies mostly used one dose, 480 mg, so nobody has mapped the slope between 150 and 400 mg properly. The reviews say so themselves. That gap matters, because it means there is no proven “safe” line to quote you. What follows is the evidence at the two ends of the curve, and an honest account of the empty middle.

150 mg: What the Newest Trial Found

The most useful caffeine study in years was published in 2025 by Hoppe and colleagues at Uppsala University, and it tested the dose people actually drink. Twenty-nine patients with panic disorder and 53 healthy controls, all with low habitual caffeine use, abstained from caffeine for 36 hours, then received either 150 mg of caffeine or a placebo in a double-blind crossover design. The researchers expected caffeine to raise subjective anxiety, especially in the panic group. It did not. There was no increase in self-reported anxiety, no increase in attention to bodily signals, and no increase in anxiety from those signals. Only one panic attack occurred, in the panic group, during an emotional task after caffeine.

That does not mean 150 mg did nothing. In both groups, caffeine increased skin conductance responses to faces, increased costly avoidance behaviour, and impaired attention to external tasks. The body was more aroused and the mind was more avoidant — the self-reported anxiety just did not rise. The authors’ conclusion is worth quoting in spirit: recommendations for caffeine abstinence in panic disorder should be based on higher doses and ideally on individual assessments, not on a blanket ban. A cup and a half of coffee is not the enemy this literature describes. Four cups might be.

400 mg and Up: Where Panic Shows Up

The high end of the curve is where the evidence gets dramatic. A 2022 systematic review and meta-analysis by Klevebrant and Frick pooled ten blinded, placebo-controlled challenge studies. Across nine studies, 51.1% of 237 patients with panic disorder had a panic attack after caffeine, and none had one after placebo. In the six studies with healthy controls, panic followed caffeine in 53.9% of patients and 1.7% of controls. The doses ran from 400 to 750 mg, mostly 480 mg — roughly five cups of coffee, taken in one go.

An older study makes the same point with more colour. In 1985, Charney and colleagues gave caffeine at 10 mg per kilo of body weight — about 700 mg for a 70-kilo person — to 21 patients with panic disorder and 17 healthy volunteers. The patients reported far greater increases in anxiety, nervousness, fear, nausea, palpitations, restlessness and tremor than the healthy volunteers, and 71% of the patients said the caffeine effects felt similar to their panic attacks. The authors concluded that patients with anxiety disorders may benefit from avoiding caffeine-containing foods and beverages — a recommendation that has held up for forty years, with the dose caveat the newer trial added.

Healthy people are not exempt, either. A 2024 meta-analysis by Liu and colleagues pooled 14 studies with 546 healthy participants and found caffeine raised anxiety overall, with a standardised mean difference of 0.94 — a large effect by the usual yardstick — and a far larger effect at high doses (above 400 mg) than at low ones. So the pattern holds in both groups: the higher the dose, the bigger the anxiety effect, and people with panic disorder sit much further along the curve.

Who Is Most Sensitive

The single biggest predictor of whether caffeine will make you anxious is whether you already have panic disorder. The Klevebrant review shows it in the starkest terms: the same dose that set off panic in half the patients set it off in fewer than 2% of healthy controls. Generalised anxiety sits somewhere in between — the reviews focus on panic disorder because that is where the challenge studies were done, not because other anxiety is immune.

A few other things shift the risk. Low habitual use seems to matter: the Hoppe trial deliberately recruited people who drank little caffeine, and even they did not get anxious at 150 mg — which suggests tolerance is not the whole story. Timing matters too, through sleep: caffeine late in the day wrecks sleep, and poor sleep raises next-day anxiety, a loop our sleep and anxiety guide covers in detail. And the form matters less than the total: coffee, tea, energy drinks, pre-workout and cola all count towards the same daily number, and the total is what your nervous system feels.

Cutting Back Without the Crash

If your total is high and your anxiety is loud, the cheapest experiment on this site is a two-week caffeine cut. Do it gradually — drop by roughly a cup a day — because stopping suddenly brings headaches, fatigue and irritability that feel a lot like anxiety and will muddy the result. Keep the morning cup if you want it; the biggest gains usually come from cutting the afternoon and evening doses, which are the ones stealing your sleep.

Track it simply. Each evening, rate your anxiety from 0 to 10 and write down how much caffeine you had and when. Two weeks of that tells you more about your own dose curve than any meta-analysis can, because your curve is the one that matters. If cutting back helps, you have found a free treatment. If it does not help, you have ruled out a suspect and can move on — to the breathing work in our breathing techniques guide, or to the conversation in our guide on when to see a doctor.

The Honest Bottom Line

Caffeine and anxiety follow a dose curve, not a yes-or-no rule. Around 150 mg, a good trial found no rise in self-reported anxiety even in people with panic disorder. Above 400 mg, half of panic disorder patients tested had a panic attack, and healthy people’s anxiety rose too. The middle of the curve is unmapped, so nobody can give you a proven personal limit — but if you have panic disorder, or your anxiety spikes on heavy-caffeine days, the evidence points one way: bring the total down, cut the afternoon doses first, and test the difference for two weeks. It costs nothing, and it is the most likely free win in this whole guide series.

Frequently Asked Questions

How much caffeine causes anxiety?
There is no single proven threshold. In trials, 150 mg did not raise self-reported anxiety, while 400–750 mg triggered panic in about half of panic disorder patients. The range between those doses has not been mapped properly.

Can one cup of coffee cause a panic attack?
Very unlikely on the trial evidence. In the 2025 trial, 150 mg — about a cup and a half — caused one panic attack across 29 panic disorder patients and 53 controls. High doses are a different story.

Does caffeine make generalised anxiety worse?
The strongest evidence is for panic disorder, where the effect is large. In healthy people, a meta-analysis found caffeine raised anxiety overall, most sharply above 400 mg. For generalised anxiety specifically, the challenge-study evidence is thinner.

Should I quit caffeine completely?
Not necessarily. The evidence supports cutting back, especially if you have panic disorder or drink a lot. Try reducing gradually and cutting afternoon doses first — a blanket ban is not what the newest trial recommends.

Does decaf help?
Decaf still contains a small amount of caffeine, usually under 15 mg a cup, which is far below the doses in these studies. If you like the ritual, decaf keeps it while removing almost all of the dose.

Can caffeine withdrawal feel like anxiety?
Yes — headaches, fatigue, irritability and low mood are common for a few days after a sudden stop, and they can feel like anxiety. That is why cutting back gradually matters, and why a two-week test beats a one-day test.

Sources

This guide is general information, not medical advice, and it is not a substitute for your own doctor — especially if you take regular medication or have a health condition. If anxiety is seriously affecting your life, talk to a doctor promptly rather than trying supplements first.