Anxiety Medication Options, Compared

Home > Anxiety Psychiatrist NYC > Anxiety Medication Options

SSRIs and SNRIs are the first-line medications for generalized anxiety disorder, panic disorder, and social anxiety disorder. In a network meta-analysis of 89 randomized trials covering 25,441 adults with generalized anxiety disorder, duloxetine, pregabalin, venlafaxine, and escitalopram each lowered Hamilton Anxiety scores by roughly 2.5 to 3.1 points more than placebo while keeping study dropout close to placebo (Slee and colleagues, Lancet, 2019). Quetiapine produced the largest reduction (3.6 points), yet patients stopped it 44% more often than placebo, and paroxetine and the benzodiazepines were effective with similarly poor acceptability.

Across the three main anxiety disorders, a meta-analysis of 234 trials and 37,333 patients found pre-post effect sizes of 2.25 for SNRIs, 2.15 for benzodiazepines, and 2.09 for SSRIs, compared with 1.29 for placebo pills (Bandelow and colleagues, International Clinical Psychopharmacology, 2015). Those are within-group changes, so a large share reflects placebo response and time, and the gaps between the active drugs are small. Accordingly, the choice turns on tolerability, speed of onset, dependence risk, and what happens when the medication stops. The anxiety psychiatrist page covers the evaluation that comes first.

SSRIs and SNRIs are the first-line medications for all three disorders

Sertraline, escitalopram, paroxetine, fluoxetine, venlafaxine, and duloxetine form the first tier for all three disorders, and guideline reviews treat them that way (Bandelow and colleagues, Dialogues in Clinical Neuroscience, 2017). In the Lancet network, sertraline and fluoxetine were also efficacious and well tolerated, although those estimates rested on small samples. Paroxetine worked yet was poorly tolerated, which is broadly consistent with clinical experience of its sedation, weight gain, and discontinuation symptoms.

Onset is slow, with meaningful improvement usually appearing over several weeks, and anxiety often increases briefly during the first week or two, which is why I start anxious patients at lower doses than I would use for depression and warn them about the early bump.

Continuation matters. In a meta-analysis of 22 relapse-prevention trials, staying on the antidepressant after response lowered the odds of relapse by about 80% in generalized anxiety disorder (pooled odds ratio 0.20) and by roughly two thirds in panic disorder (0.35) (Donovan and colleagues, Journal of Affective Disorders, 2010). Guideline reviews recommend continuing for 6 to 12 months after remission before a gradual reduction. Decisions about dose and duration are individual and made with the prescriber, and the medication management page describes how I monitor them.

Buspirone works modestly in generalized anxiety and takes weeks

Buspirone is effective for generalized anxiety disorder, and the effect is modest. The Cochrane review of azapirones pooled 36 trials with 5,908 participants and found buspirone superior to placebo, with a number needed to treat of 4.4 on the Clinical Global Impression scale (Chessick and colleagues, Cochrane Database of Systematic Reviews, 2006). The same review judged it possibly less effective than benzodiazepines, found that more participants stopped buspirone than stopped a benzodiazepine, and noted that it appeared most useful in people who had never taken a benzodiazepine.

Three features define where it fits. It is non-sedating for most people, carries no dependence liability, and rarely causes the sexual side effects that SSRIs can, which makes it a judicious add-on when an SSRI has helped partially. Nearly all of its controlled evidence is in generalized anxiety. And it is slow; the Cochrane trials ran four to nine weeks, and I tell patients to judge it at six weeks rather than six days. The FDA label starts at 15 mg daily in divided doses, allows increases of 5 mg per day every two to three days, and caps the dose at 60 mg daily. The label also notes that buspirone has no cross-tolerance with benzodiazepines and will not block benzodiazepine withdrawal, so it cannot stand in for a taper.

Hydroxyzine has thin evidence and a real niche

Hydroxyzine is an antihistamine whose FDA label covers symptomatic relief of anxiety and tension, with effects usually noted within 15 to 30 minutes of a dose. The controlled evidence is thin. The Cochrane review found five trials with 884 participants in generalized anxiety disorder (Guaiana and colleagues, Cochrane Database of Systematic Reviews, 2010). Hydroxyzine beat placebo (odds ratio for non-response 0.30) and matched chlordiazepoxide and buspirone, with more drowsiness. The reviewers rated the risk of bias as high and declined to call it a reliable first-line treatment.

In practice hydroxyzine is an as-needed drug without dependence risk. It acts quickly, sedates, and carries anticholinergic effects (dry mouth, constipation, and, in older adults, confusion) that limit daytime use. I use it for worry-driven sleep-onset trouble and for occasional acute anxiety in people for whom a benzodiazepine is a poor idea.

Benzodiazepines relieve anxiety fast, and the risks make them short-term drugs

Benzodiazepines relieve anxiety within the hour, and the trial data confirm efficacy: a pre-post effect size of 2.15 across 42 study arms in the 2015 meta-analysis, and clear efficacy in the Lancet network, where they were nonetheless poorly tolerated relative to placebo.

In September 2020 the FDA required a boxed warning on every benzodiazepine covering abuse, misuse, addiction, physical dependence, and withdrawal reactions. The agency reported about 92 million benzodiazepine prescriptions dispensed from US outpatient pharmacies in 2019, that about half of patients who received one in 2018 took it for two months or longer, and that physical dependence can develop within several days to weeks even at prescribed doses. The warning advises the minimum dose and duration needed and a gradual taper when stopping.

Older adults carry added risk. A meta-analysis of 22 studies with 79,081 participants over 60 found benzodiazepine use associated with falls (adjusted odds ratio 1.41), an association antidepressants shared (1.36) (Woolcott and colleagues, Archives of Internal Medicine, 2009). An umbrella review of five meta-analyses covering 30 observational studies found benzodiazepine use associated with dementia, with effect sizes from 1.38 to 1.78, while rating the evidence weak (Wu and colleagues, Journal of Personalized Medicine, 2023). Those data cannot show causation. They do raise concerns regarding long-term use in people already at risk of cognitive decline.

Dependence risk has to be weighed against the real benefits of short or intermittent use and the risk of the untreated condition (Baldwin and colleagues, Journal of Psychopharmacology, 2013). Accordingly, I use them two ways: as a bridge for the first weeks while an SSRI takes hold, and as an occasional rescue dose for infrequent, predictable triggers. Daily use beyond the bridge, and any use in people with alcohol or sedative problems, warrants a different plan. When a patient arrives on long-term daily benzodiazepines, the reduction is slow, supervised, and never abrupt.

Propranolol is a performance-anxiety drug

Propranolol has consistent support in one setting, the discrete performance, and little beyond it. A systematic review found eight randomized trials: panic disorder (four trials, 130 patients), specific phobia (two, 37), social phobia (one, 16), and PTSD (one, 19). Pooled analyses found no difference from benzodiazepines for short-term panic treatment, and the authors concluded the evidence was insufficient to support routine use for any anxiety disorder (Steenen and colleagues, Journal of Psychopharmacology, 2016).

The performance data are older and smaller. In two double-blind trials involving 29 musicians, propranolol taken before performing reduced stage-fright symptoms and improved performance quality as judged by music critics, without sedation (Brantigan and colleagues, American Journal of Medicine, 1982). That matches what I see: the drug blunts tremor, racing heart, and shaky voice for a presentation or audition, and it does little for the anticipatory worry that defines generalized social anxiety. The use is off-label, it requires caution with asthma, low blood pressure, or cardiac conduction problems, and I prescribe it for events only.

Pregabalin, gabapentin, and mirtazapine are second-line options with uneven evidence

Pregabalin has the strongest evidence of the three and is off-label for anxiety in the United States. In the Lancet network it ranked beside duloxetine and venlafaxine (2.79 points below placebo), and a meta-analysis of eight placebo-controlled trials with 2,299 patients found a Hedges' g of 0.37 with no between-trial heterogeneity and dropout no higher than placebo (Generoso and colleagues, International Clinical Psychopharmacology, 2017). Dizziness, sedation, weight gain, and a misuse signal keep it second-line for me.

Gabapentin rests on far less. One 14-week placebo-controlled trial randomized 69 patients with social phobia to gabapentin (900 to 3,600 mg daily) or placebo and found a significant reduction in symptoms, with more dizziness and somnolence (Pande and colleagues, Journal of Clinical Psychopharmacology, 1999). That single trial remains the main controlled support for gabapentin in anxiety, and its wide off-label use has outrun the evidence.

Mirtazapine appeared efficacious and well tolerated for generalized anxiety in the Lancet network, and a 10-week placebo-controlled trial of 66 women with social phobia found it superior on the Social Phobia Inventory and the Liebowitz scale (Muehlbacher and colleagues, Journal of Clinical Psychopharmacology, 2005). It sedates and causes weight gain in many patients, which makes it reasonable when anxiety comes with insomnia and poor appetite, and a poor choice otherwise.

CBT alone matches medication, and the combination does best

Cognitive behavioral therapy carries the highest level of evidence among psychotherapies for anxiety disorders, and head-to-head data put it level with medication. The clearest trial randomized 312 patients with panic disorder to imipramine, CBT, placebo, or combinations (Barlow and colleagues, JAMA, 2000). At three months, response rates were 45.8% for imipramine, 48.7% for CBT, and 21.7% for placebo; combined treatment reached 60.3% acutely and 57.1% after six months of maintenance. Six months after treatment stopped, 41.0% of the CBT-plus-placebo group and 31.9% of the CBT-alone group still met response criteria, against 19.7% for imipramine alone. CBT held after it ended. Medication effects largely did not.

Combination has its own evidence. A meta-analysis of 52 trials with 3,623 patients, 21 of them in anxiety disorders, found combined psychotherapy and antidepressant treatment superior to the antidepressant alone, with a Hedges' g of 0.43 and a number needed to treat of 4.2, the strongest anxiety evidence in panic disorder, and effects persisting up to two years (Cuijpers and colleagues, World Psychiatry, 2014).

The medication versus therapy versus both page works through the decision, and the CBT page describes the therapy itself.

How I sequence these medications

A patient asked me recently, "If the medication takes six weeks to work, what am I supposed to do until then?" The answer differs by disorder and by severity.

For generalized anxiety disorder, I start with an SSRI or SNRI at a low dose, paired with CBT when the patient will do it, and add buspirone when the response at eight to twelve weeks is partial. Hydroxyzine covers worry-driven insomnia early on. A benzodiazepine bridge is reserved for anxiety severe enough to prevent work or sleep, with a stop date set at the first visit.

For panic disorder, the Barlow trial shapes the plan: CBT with interoceptive exposure is the treatment with the most durable effect, and an SSRI is added when attacks are frequent or agoraphobic avoidance has set in. A short-acting benzodiazepine taken during an attack likely teaches the brain that the attack was dangerous and relief came from a pill, which works against exposure, so I rarely prescribe one as rescue medication for panic.

For social anxiety disorder, the generalized form gets an SSRI such as sertraline plus CBT, and the performance-only form gets propranolol taken before the event, with no daily medication. Pregabalin, mirtazapine, and gabapentin come in after two adequate first-line trials have failed.

In children and adolescents, SSRIs plus CBT remain the evidence-supported pairing, and I do not prescribe benzodiazepines to youth outside procedural settings. Whatever the sequence, medication decisions are individual and made with the prescriber, and the plan for eventually reducing the medication belongs in the first conversation. Response is heterogeneous, and the main risk of any effective anxiolytic is that relief makes it easy to stop paying attention. Vigilance about dose, duration, and the exit plan is part of the treatment.

Who this practice fits, and when to go elsewhere

This practice fits adults, adolescents, and children with generalized anxiety, panic, or social anxiety who want a prescriber who lays out the options with the numbers attached, works with CBT, and plans the exit from medication as carefully as the entry. It also fits people on a daily benzodiazepine for months or years who want a slow, supervised reduction onto a sustainable alternative, and people whose anxiety has never fully responded to SSRIs and who may have untreated ADHD, which the ADHD versus anxiety page covers. Many first visits are second opinions on a stalled regimen.

Some situations belong elsewhere. Anxiety with active suicidal thinking needs same-day evaluation, and an emergency department if there is a plan. Anxiety driven by alcohol or sedative dependence belongs with an addiction-trained team, since unsupervised withdrawal can be dangerous. Panic-like symptoms with chest pain, fainting, or new palpitations warrant medical evaluation first.

Cost and booking

Initial evaluations run 60 to 90 minutes and include a full psychiatric history, a review of prior medication trials with doses and durations, screening for ADHD, depression, and substance use, and a treatment plan. The practice is out-of-network with all insurance plans. You pay directly and receive a superbill to submit for reimbursement; many New York plans with out-of-network benefits reimburse a meaningful share of the fee once the deductible is met. The numbers are laid out at fees and out-of-network costs, with a guide to checking your out-of-network benefits.

Bring a list of every anxiety medication you have tried, with doses and durations. Schedule a consultation or call (646) 893-4309.

Related

Anxiety psychiatrist NYC · Generalized anxiety disorder · Panic disorder · Social anxiety · CBT · Medication vs therapy vs both · Medication management · Emotional blunting on antidepressants · Antidepressant sexual side effects · Antidepressants and weight gain · Depression with ADHD: combining medications · How long antidepressants take to work · Lexapro vs Zoloft · SSRIs for teens · Stopping antidepressants · Wellbutrin vs SSRIs

Frequently Asked Questions

What is the best medication for anxiety?

For most adults with generalized anxiety, panic disorder, or social anxiety, an SSRI or SNRI is the best-supported starting point. In the 2019 Lancet network meta-analysis of 89 trials, duloxetine, pregabalin, venlafaxine, and escitalopram combined the largest benefit with acceptability close to placebo, and sertraline was also effective and well tolerated. The choice depends on side-effect profile, other conditions, and prior trials.

How long does buspirone take to work?

Weeks. The Cochrane trials of buspirone in generalized anxiety ran four to nine weeks, and I ask patients to judge it at about six weeks. The FDA label starts at 15 mg daily in divided doses with increases every two to three days to a maximum of 60 mg. Buspirone has no cross-tolerance with benzodiazepines and will not relieve benzodiazepine withdrawal.

Are benzodiazepines safe for daily anxiety treatment?

They relieve anxiety quickly, and daily use carries risks the FDA describes in a 2020 boxed warning: abuse, misuse, addiction, physical dependence, and withdrawal. Physical dependence can develop within days to weeks even at prescribed doses. In older adults, benzodiazepines are associated with falls and, in weaker observational data, with dementia. I use them as a bridge or an occasional rescue dose, with a stop plan set at the start.

Does propranolol work for social anxiety?

For performance situations, yes. In two double-blind trials of 29 musicians, propranolol taken before performing reduced stage-fright symptoms and improved performance quality as judged by critics. For generalized social anxiety disorder, the evidence is insufficient; a 2016 systematic review of eight trials found no basis for routine use in any anxiety disorder. It blunts tremor and racing heart and does little for anticipatory worry.

Can anxiety be treated with therapy alone?

Often, yes. In a 312-patient trial for panic disorder, CBT alone produced a 48.7% response rate at three months, equal to imipramine (45.8%), and CBT's effect held six months after treatment ended while medication's largely did not. Combination treatment outperformed medication alone in a 52-trial meta-analysis (Hedges' g 0.43). Severity, access to a skilled CBT therapist, and patient preference decide the starting point.

How long should I stay on an anxiety medication?

Guideline reviews recommend continuing an effective antidepressant for 6 to 12 months after remission before a gradual reduction. In a meta-analysis of 22 relapse-prevention trials, staying on the antidepressant cut relapse odds by about 80% in generalized anxiety disorder and by roughly two thirds in panic disorder. The timing and pace of any reduction is an individual decision made with the prescriber, usually with CBT skills in place first.

Sources

1. Slee A, et al. Pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. Lancet. 2019;393(10173):768–777. PMID 30712879
2. Bandelow B, et al. Efficacy of treatments for anxiety disorders: a meta-analysis. Int Clin Psychopharmacol. 2015;30(4):183–192. PMID 25932596
3. Bandelow B, Michaelis S, Wedekind D. Treatment of anxiety disorders. Dialogues Clin Neurosci. 2017;19(2):93–107. PMID 28867934
4. Donovan MR, et al. Comparative efficacy of antidepressants in preventing relapse in anxiety disorders: a meta-analysis. J Affect Disord. 2010;123(1-3):9–16. PMID 19616306
5. Chessick CA, Allen MH, Thase M, et al. Azapirones for generalized anxiety disorder. Cochrane Database Syst Rev. 2006;(3):CD006115. PMID 16856115
6. Buspirone hydrochloride tablets. Prescribing information (Accord Healthcare). DailyMed, U.S. National Library of Medicine. dailymed.nlm.nih.gov
7. Guaiana G, Barbui C, Cipriani A. Hydroxyzine for generalised anxiety disorder. Cochrane Database Syst Rev. 2010;(12):CD006815. PMID 21154375
8. Hydroxyzine hydrochloride tablets. Prescribing information. DailyMed, U.S. National Library of Medicine. dailymed.nlm.nih.gov
9. U.S. Food and Drug Administration. FDA requiring Boxed Warning updated to improve safe use of benzodiazepine drug class. September 23, 2020. fda.gov
10. Woolcott JC, Richardson KJ, Wiens MO, et al. Meta-analysis of the impact of 9 medication classes on falls in elderly persons. Arch Intern Med. 2009;169(21):1952–1960. PMID 19933955
11. Wu CC, et al. Benzodiazepine use and the risk of dementia in the elderly population: an umbrella review of meta-analyses. J Pers Med. 2023;13(10):1485. PMID 37888096
12. Baldwin DS, Aitchison K, Bateson A, et al. Benzodiazepines: risks and benefits. A reconsideration. J Psychopharmacol. 2013;27(11):967–971. PMID 24067791
13. Steenen SA, van Wijk AJ, van der Heijden GJ, van Westrhenen R, de Lange J, de Jongh A. Propranolol for the treatment of anxiety disorders: systematic review and meta-analysis. J Psychopharmacol. 2016;30(2):128–139. PMID 26487439
14. Brantigan CO, Brantigan TA, Joseph N. Effect of beta blockade and beta stimulation on stage fright. Am J Med. 1982;72(1):88–94. PMID 6120650
15. Generoso MB, et al. Pregabalin for generalized anxiety disorder: an updated systematic review and meta-analysis. Int Clin Psychopharmacol. 2017;32(1):49–55. PMID 27643884
16. Pande AC, Davidson JR, Jefferson JW, et al. Treatment of social phobia with gabapentin: a placebo-controlled study. J Clin Psychopharmacol. 1999;19(4):341–348. PMID 10440462
17. Muehlbacher M, Nickel MK, Nickel C, et al. Mirtazapine treatment of social phobia in women: a randomized, double-blind, placebo-controlled study. J Clin Psychopharmacol. 2005;25(6):580–583. PMID 16282842
18. Barlow DH, et al. Cognitive-behavioral therapy, imipramine, or their combination for panic disorder: a randomized controlled trial. JAMA. 2000;283(19):2529–2536. PMID 10815116
19. Cuijpers P, et al. Adding psychotherapy to antidepressant medication in depression and anxiety disorders: a meta-analysis. World Psychiatry. 2014;13(1):56–67. PMID 24497254