If you're in crisis right now, this page is the wrong place. In Ireland and the UK, Samaritans is 116 123 — free, 24 hours. If you're at immediate risk, go to A&E or call 999 or 112.
"Anxiety" covers two quite different things, and using one word for both muddles what to do about either. One is a normal function working correctly. The other is a condition with specific criteria and specific treatments. Telling them apart is the most useful thing you can do before deciding on any next step.
Anxiety, the feeling
A normal, useful response to something uncertain or threatening. It has an object — the interview, the scan result, the conversation you're dreading. It rises, it peaks, and once the thing has happened it subsides.
This version is not a malfunction. It is the system working. Anxiety mobilises attention and energy toward something that matters, and people who feel none of it do worse, not better, at the things that warrant it. The relationship between arousal and performance is a curve rather than a line: too little and nothing happens, too much and performance falls apart, and somewhere in between is the state most people would describe as being nervous but sharp.
Three features mark it out. It is proportionate — roughly matched to what is actually at stake. It is attached to something you can name. And it resolves when the situation does.
Anxiety, the disorder
Clinically, what is looked for is different in kind rather than simply in degree. For generalised anxiety disorder, the diagnostic criteria require:
- Excessive anxiety and worry occurring more days than not for at least six months, about a number of different things
- The worry being difficult to control — not merely unpleasant, but hard to stop or redirect
- At least three of the following six: restlessness or feeling on edge · being easily fatigued · difficulty concentrating or mind going blank · irritability · muscle tension · disturbed sleep
- Clinically significant distress or impairment in work, social life or daily functioning
- And it is not better explained by a medical condition, a substance, or another disorder
That fourth criterion is doing most of the work. The question a clinician is really asking is not "how anxious are you" but "what is it stopping you doing" — which is also the most useful question to ask yourself before booking an appointment, and the most useful thing to be able to answer once you're in one.
The questionnaire clinicians actually use
The GAD-7 is the standard brief screen, and knowing it is useful because it is probably what you'll be handed. The stem is: over the last two weeks, how often have you been bothered by the following problems?
- Feeling nervous, anxious or on edge
- Not being able to stop or control worrying
- Worrying too much about different things
- Trouble relaxing
- Being so restless that it's hard to sit still
- Becoming easily annoyed or irritable
- Feeling afraid as if something awful might happen
Each is scored 0 for not at all, 1 for several days, 2 for over half the days, 3 for nearly every day, giving a total out of 21. The interpretation thresholds sit at 5 for mild, 10 for moderate and 15 for severe, and a score of 10 or more is the point at which further evaluation is recommended.
Two caveats on using it yourself. It measures the past two weeks, so a bad fortnight during a genuinely difficult period will score high without indicating a disorder — the six-month duration criterion above is what separates those. And it is a severity measure, not a diagnosis: a score is a reason to have the conversation, not the conclusion of it.
Several distinct conditions underneath
"Anxiety disorder" is a family, not a single thing, and the members respond to different approaches.
Generalised anxiety disorder — the worry moves from topic to topic and doesn't settle. Resolve one concern and it relocates rather than stopping.
Panic disorder — discrete attacks of intense fear with strong physical symptoms, peaking within minutes, plus persistent worry about the next one or behaviour changed to avoid it. Frequently mistaken for a cardiac event, and — importantly — the reverse also happens. Chest pain should be assessed rather than assumed, particularly a first episode.
Social anxiety disorder — specific to being observed, evaluated or judged. Often mislabelled as shyness for years, which delays treatment that works well for it.
Agoraphobia — fear of situations where escape would be difficult or help unavailable. Commonly develops alongside panic disorder, and the avoidance is what does the damage to someone's life.
Specific phobia — a marked fear of a particular object or situation. Among the most treatable things in this entire field.
Health anxiety — persistent fear of illness that reassurance resolves only briefly. The reassurance-seeking itself maintains the cycle, which is why more tests rarely help.
And one that is regularly filed here and shouldn't be: OCD is not an anxiety disorder in current classification, and it needs a specific form of treatment rather than generic anxiety management. It's also badly misunderstood in everyday use of the word.
The things a decent assessment rules out first
Several medical and pharmacological causes produce something indistinguishable from an anxiety disorder, and missing them means treating the wrong thing for years.
An overactive thyroid. The single most important one. It produces anxiety, agitation, tremor, palpitations, insomnia, heat intolerance and weight loss, and it is routinely treated as an anxiety disorder before anyone checks. It is one blood test — details here. If your anxiety arrived suddenly in adulthood with physical symptoms attached, ask for this specifically.
Caffeine. Genuinely underrated as a cause, not merely an aggravator. High intake produces racing heart, jitteriness and a sense of dread that is physiologically close to anxiety, and people vary enormously in how quickly they clear it — so the same three coffees are a different exposure for different people.
Alcohol, and withdrawal from it. Alcohol lowers anxiety for a few hours and raises it the following day. Regular heavy use produces a persistent anxious state that improves on stopping, and withdrawal from alcohol or from benzodiazepines can present as severe anxiety.
Medication and stimulants. Some asthma medications, decongestants, thyroid replacement at too high a dose, steroids, and stimulant medication can all do it.
Anaemia and arrhythmias both produce palpitations, breathlessness and a sense of something being wrong. A short blood panel covers a fair amount of this ground.
Why the difference changes what you do
For the feeling: the useful moves are sleep, exercise, cutting caffeine, and addressing the thing you're actually anxious about. Ordinary, and often sufficient. Nothing on that list needs a prescription or a purchase. Two that are better supported than they sound: which type of exercise actually helps, and — if the pattern is stress wrecking your nights and the lost sleep then worsening the stress — how that loop works.
For the disorder: there are treatments with substantial evidence behind them, and for most people the wait to access them is a bigger obstacle than the treatment itself. We've ranked them by how well supported they are — and the ranking is not flattering to what we sell.
The mistake worth avoiding is treating the second as though it were the first — spending two years on breathing apps, magnesium and a calming tea for something that would respond to a course of therapy. That is the specific error this article exists to prevent, and it is extremely common, partly because the first approach is available immediately and the second involves a waiting list.
The reverse error is worth naming too: pathologising an appropriate reaction to a genuinely difficult situation. Being frightened before surgery, or unable to sleep during a redundancy process, is not a disorder. It is a person responding to circumstances.
Where we sit
We sell supplements. For the feeling, a small number of the things we sell have a modest, real evidence base, and we've written up exactly what those trials found, at what doses, and how far our own doses fall short of them. On magnesium specifically — the one most often recommended for anxiety — the evidence is thinner than the marketing, and we've put a number on how thin.
For the disorder, a supplement is not the treatment — ours included — and the useful step is a GP conversation. We would rather say that here than take the order.
This article is not medical advice and cannot diagnose an anxiety disorder. Diagnostic criteria summarised from DSM-5; the GAD-7 is reproduced as a widely used screening instrument and is not a diagnostic test. If any of the above describes you, speak to a doctor.