Anxiety Treatment: The Options, Honestly Compared
There is more evidence behind anxiety treatment than almost any other area of mental health, and it points fairly clearly in some directions and away from others.
Anxiety disorders are among the best-studied conditions in psychiatry, which means the evidence points somewhere fairly clearly. It points towards structured therapy and a small number of medication classes, and away from the drug most people are still handed first.
The short version
- Cognitive behavioural therapy, particularly with an exposure component, has the strongest evidence base of any anxiety treatment. Its effects tend to persist after treatment ends.
- SSRIs and SNRIs are the first-line medications. They take weeks to work, and anxiety often worsens for a period first — which is why they are started low and why people abandon them too early.
- Benzodiazepines work fast and are a poor long-term answer: tolerance, dependence, a genuinely difficult withdrawal, and cognitive effects. They have narrow legitimate short-term uses.
- Propranolol is useful for the physical symptoms of performance anxiety and does not treat the anxiety disorder itself.
- Sleep, alcohol and caffeine have real effects here, and are not a substitute for treatment.
One thing worth establishing first. "Anxiety" covers several distinct diagnoses — generalised anxiety disorder, panic disorder, social anxiety disorder, specific phobias, and related conditions like OCD and PTSD that are classified separately but share features. The treatments overlap substantially but not entirely, and which one you have changes the best first move.
Psychological treatment
Cognitive behavioural therapy
CBT is the most thoroughly tested psychological treatment for anxiety, across every major anxiety diagnosis, in trials run by many different groups over several decades. That consistency is unusual in mental health research.
What it involves is more concrete than the phrase suggests. You identify the specific thoughts that fire during anxious episodes, test them against what actually happens, and change the behaviours holding the anxiety in place — principally avoidance and safety behaviours. It is structured, time-limited, and involves work between sessions, typically over eight to twenty sessions rather than open-ended.
The most important feature is that the benefit tends to hold after treatment stops. Medication generally works while you are taking it; therapy teaches something that remains available afterwards. That is why it is usually the recommended starting point where it is accessible.
Exposure
Exposure-based therapy does most of the work in phobias, panic disorder and social anxiety, and it is the component most often left out. The principle: anxiety is maintained by avoidance. Every time you avoid the feared situation you get short-term relief and the fear is reinforced. Exposure means approaching the feared thing deliberately, in a graded and planned way, staying with it until the anxiety comes down on its own, and repeating until the situation stops carrying the same charge.
It is uncomfortable by design and should never be improvised. Done properly, with a therapist, at a pace you have agreed, it is one of the more reliably effective interventions in mental health. Done badly — too fast, without preparation, alone — it can entrench the fear instead. Ask a prospective therapist directly whether they practise CBT with exposure for your specific diagnosis; the labels are used loosely.
Medication
SSRIs and SNRIs
These are the first-line medications for most anxiety disorders. Several SSRIs and SNRIs carry FDA approvals for specific anxiety diagnoses; others are used off-label within the same classes, which is standard practice. Which one a clinician chooses depends on your diagnosis, your other conditions, interactions, and side-effect profile.
Three things are worth knowing before you start, because not knowing them is the main reason people stop.
They are slow. Some benefit may appear within two or three weeks, but the fuller picture generally takes six to eight weeks at a therapeutic dose. Judging one at week two is judging nothing.
Anxiety often gets worse first. The first week or two can bring restlessness, jitteriness, worse sleep and heightened anxiety. This is well recognised, it is why doses in anxiety are started lower than in depression and raised slowly, and it typically settles. It is also why a great many people conclude the drug is making them worse and stop. Knowing it is coming makes it survivable, and telling your prescriber rather than stopping unilaterally gives them the chance to slow the titration.
Stopping needs a plan. These are not addictive the way benzodiazepines are, but stopping abruptly can produce discontinuation symptoms — dizziness, electric-shock sensations, irritability, flu-like feelings, disturbed sleep. Tapering under supervision avoids most of that.
Other medications appear in specific situations: buspirone for generalised anxiety, hydroxyzine short-term, pregabalin in some settings. Clinical decisions rather than menu choices.
Be careful here: benzodiazepines
Diazepam, lorazepam, alprazolam, clonazepam and related drugs work within an hour and work well. That is precisely the problem. Tolerance develops, so the dose that worked stops working. Physical dependence can develop within weeks of regular use — the body adapts and reacts to the drug's absence. Withdrawal is genuinely dangerous: abrupt cessation after sustained use can cause seizures, and the psychological withdrawal can be prolonged, with rebound anxiety worse than the original. Never stop a benzodiazepine abruptly; a taper has to be managed by a prescriber. There are also cognitive effects with ongoing use — memory, attention, reaction time — and risk rises with age and sharply in combination with alcohol or opioids.
They are not banned drugs and there are legitimate short-term uses: a brief course during an acute crisis, a single dose for a procedure, bridging the first weeks of an SSRI. The problem is almost never the two-week prescription. It is the two-week prescription still being refilled three years later. If you are on one long-term, that is a conversation to have with your prescriber, not a reason to stop on your own.
Propranolol
Propranolol is a beta blocker. It blunts the physical machinery of anxiety — racing heart, tremor, shaking hands, the voice going — without touching the mental component. That makes it genuinely useful for performance anxiety: a presentation, a viva, a musical performance, anything where the physical symptoms themselves become the thing you dread.
It is used situationally rather than daily for this purpose, and it is not a treatment for generalised anxiety disorder or panic disorder. It is not suitable for people with asthma, certain cardiac conditions, or low blood pressure, which is why it needs a prescriber rather than a friend's spare tablets.
The lifestyle factors, without the platitudes
These get listed everywhere in a way that makes them easy to dismiss. A few of them genuinely move the needle, and it is worth being specific about which.
Sleep. The relationship runs both ways, and insufficient sleep measurably raises next-day anxiety in healthy people. If your sleep is broken, treating it is not an adjunct to anxiety treatment — it may be the treatment. Why some clinicians now treat sleep first covers this properly.
Alcohol. It reduces anxiety for a few hours and increases it the following day, through the rebound as it clears. For someone with an anxiety disorder that is a reliable way to build a cycle. It also interacts badly with benzodiazepines and blunts the effect of antidepressants.
Caffeine. Pharmacologically it produces a state that closely resembles anxiety: raised heart rate, jitteriness, hypervigilance. In people prone to panic it can precipitate attacks. Reducing it is free, immediate and occasionally dramatic. Taper rather than stop dead if you drink a lot.
Exercise. Aerobic exercise has a real, if moderate, effect in trials. It is not a replacement for treatment, and telling an acutely anxious person to go for a run is unhelpful. As part of a plan, it earns its place.
Worth knowing
Anxiety and depression co-occur so often that treating one without assessing for the other is a mistake. It also runs alongside physical conditions that produce identical symptoms — thyroid disease being the obvious one, and worth excluding with a blood test if it has never been checked. A proper assessment covers both directions.
Combining treatments, and how long to give things
For most anxiety disorders, therapy and medication together tend to outperform either alone, particularly where symptoms are severe enough that engaging with therapy is hard without some pharmacological help first. That is a reasonable sequence: start medication, start therapy a few weeks later once the early jitteriness has settled.
Give an SSRI or SNRI a proper trial — adequate dose, six to eight weeks — before concluding it hasn't worked. If it hasn't, a switch within class or to a different mechanism is the usual next step, and there are several to work through. What counts as an adequate trial is worth reading before you decide a drug has failed. And if the anxiety arrived alongside exhaustion tightly tied to one situation, usually work, read burnout or depression as well.
When to get seen sooner rather than later
Panic attacks becoming frequent. Avoidance that is narrowing your life — not going to work, not leaving the house, not driving. Anxiety affecting your ability to eat or sleep. Alcohol or medication use that has crept up as a way of managing it. None of those are wait-and-see situations.
And if you are having thoughts of harming yourself, that is an urgent conversation, not one to have online. Call or text 988 for the Suicide & Crisis Lifeline, or go to your nearest emergency room. Telehealth is not appropriate for a psychiatric emergency.
Otherwise, our stress and anxiety service covers assessment and ongoing management, and you can book a consultation. If your anxiety sits alongside depression that has already resisted several treatments, our guide to ketamine for treatment-resistant depression covers what does and does not open up at that point.
Talk to a licensed clinician
Reading about a treatment is not the same as knowing whether it fits your history. A consultation is a conversation about your own situation — not a sales call, and not a promise of any outcome.
Book a consultation
