Section 3: Beyond First-Line: Alternative and Augmentation Agents

The agents in this section span a broad pharmacological range. Some (buspirone, hydroxyzine, pregabalin) have genuine anxiolytic evidence and occupy defined clinical niches. Others (gabapentin, propranolol, mirtazapine) are widely prescribed off-label with thinner evidence. Still others (CBD, kava, ashwagandha) are agents your patients will ask about. For each, the goal is the same: honest evidence appraisal, specific clinical positioning, and practical prescribing guidance where the evidence supports it.

Section 3 Agents at a Glance

Agent	Mechanism	FDA Anxiety Approval	Onset	Evidence Level	Primary Clinical Niche	Key Risk	Schedule
Buspirone	5-HT1A partial agonist (+ D2 partial)	Yes (GAD)	2-4 weeks	Moderate (low ES)	SUD patients; SSRI augmentation	Low efficacy as monotherapy	None
Hydroxyzine	H1 antagonist (HCl and pamoate forms)	Yes (anxiety)	30-60 min	Moderate	PRN acute anxiety (young adults)	Anticholinergic; Beers-listed; QTc	None
Pregabalin	Alpha-2-delta Ca channel	EU/UK/Canada (GAD); not US	~Day 4	Strong (manufacturer-funded)	Somatic GAD; BZD taper; elderly	Dependence/withdrawal; abuse; weight gain	V (US)
Gabapentin	Alpha-2-delta Ca channel	No	Variable	Weak (GAD); Limited (SAD)	Off-label when pregabalin unavailable increasingly state-scheduled	Nonlinear PK; abuse signal	Unscheduled federally; state-scheduled
Propranolol	Beta-1/2 antagonist	No	60-90 min PRN	Strong (performance only)	Performance anxiety	Overdose lethality; no GAD data	None
Mirtazapine	NaSSA (alpha-2/H1/5-HT2)	No	1-2 weeks	Weak (anxiety-specific)	Anxious insomnia + cachexia	Sedation; weight gain	None
Trazodone	SARI (5-HT2A/H1/alpha-1)	No (approved for MDD)	30-60 min (sleep)	Moderate (insomnia); Historical (GAD)	Anxiety-driven insomnia	Priapism; orthostasis; QTc	None
Clonidine	Non-selective alpha-2 agonist	No	30-60 min	Limited (anxiety)	PTSD hyperarousal; SUD withdrawal	Hypotension; rebound HTN	None
Guanfacine XR	Selective alpha-2A agonist	No (approved for ADHD)	1-2 weeks	Limited (pilot study)	Pediatric ADHD + comorbid anxiety	Sedation; hypotension	None
Vilazodone	SSRI + 5-HT1A partial agonist	No (approved for MDD)	2-4 weeks	Very limited (anxiety)	Intermediate step after SSRI failure	GI side effects	None
Vortioxetine	Multimodal serotonergic	No (approved for MDD)	2-4 weeks	Very limited (anxiety)	Cognitive complaints + anxiety	Nausea; long half-life (66h washout)	None
Quetiapine	Multi-receptor (D2/5-HT2/H1)	No	Days	Moderate (augmentation)	Treatment-resistant (Stage 2-3 only)	Metabolic syndrome; TD; mortality	None
Aripiprazole	D2 partial agonist / 5-HT1A	No	Days-weeks	Limited (augmentation)	TR-anxiety augmentation; anergia	Akathisia; metabolic risk	None
Risperidone	D2/5-HT2A antagonist prolactin	No	Days	Limited (augmentation)	TR-OCD augmentation	Metabolic; EPS	None

3.1 Buspirone

Buspirone is the only FDA-approved non-benzodiazepine, non-antidepressant anxiolytic. It occupies a unique pharmacological niche, and understanding both its genuine strengths and its real-world limitations is essential, because this is an agent you will prescribe, inherit, or have patients ask about repeatedly.

Mechanism

Buspirone is a partial agonist at the serotonin 5-HT1A receptor. In plain clinical terms, it partially activates the same receptor that full serotonin agonists would activate, producing a moderate anxiolytic effect without the sedation, cognitive impairment, or dependence liability of benzodiazepines. It also has clinically relevant partial agonist activity at the dopamine D2 receptor (similar mechanistically to aripiprazole), which may partly explain its utility in SSRI augmentation for anxious depression. It has zero activity at the GABA-A receptor, which means it carries no risk of physiological dependence, no tolerance, and no withdrawal syndrome upon cessation. (Stahl, 2021, Cambridge University Press)

It does not work immediately. Unlike benzodiazepines, which produce anxiolysis within 30 minutes, buspirone requires 2 to 4 weeks of consistent daily dosing before meaningful clinical effects emerge. This delayed onset is the single most important clinical fact about buspirone, because it is the primary driver of both patient dissatisfaction and premature discontinuation.

What the evidence shows

In a cited effect-size comparisons for GAD pharmacotherapy (Hidalgo et al., 2007, Journal of Psychopharmacology), buspirone produced the lowest magnitude of therapeutic benefit of any medication class studied: an effect size of 0.17, compared to SSRIs at 0.36, benzodiazepines at 0.38, SNRIs at 0.42, hydroxyzine at 0.45, and pregabalin at 0.50. (Note: these effect sizes come from a single meta-analysis now nearly two decades old; more recent network meta-analyses may produce somewhat different rank orderings, though the general pattern of buspirone's relative position has been consistent.)

That number requires honest context. An effect size of 0.17 is small by any standard, but buspirone's clinical utility is not primarily about outperforming other agents on rating scales. Its value is in specific clinical scenarios where the alternatives are contraindicated or carry unacceptable risks.

In pediatric populations, the data is disappointing. Two large abandoned clinical trials obtained via FDA Supplement 43 (Protocol CN101-124, 227 participants; Protocol CN101-125, 341 participants) failed to show statistically significant separation from placebo on primary outcome measures. A subsequent Bayesian re-analysis of these abandoned datasets suggested that the lack of efficacy signal may reflect methodological limitations rather than true absence of effect, but the practical reality is that buspirone does not have a robust pediatric evidence base. (Strawn et al., 2018, Journal of Child and Adolescent Psychopharmacology)

Recent literature from 2023 to 2025 highlights buspirone's most promising niche: augmentation in anxious depression. A 12-week trial of buspirone augmentation demonstrated significant decreases in HAM-A scores (from 25.2 to 15.4, p < 0.001) alongside concurrent reductions in HAM-D (from 19.4 to 12.7, p < 0.001), with adverse reactions in only 3.7% of participants. (Woo et al., 2025, Clinical Psychopharmacology and Neuroscience)

Dosing and practical implementation

The therapeutic range for GAD is 20 to 45 mg per day, with the FDA maximum at 60 mg per day. The recommended starting dose is 7.5 mg twice daily (15 mg/day), titrating upward by 5 mg every 2 to 3 days as tolerated to the target range. Clinical audits consistently reveal that typical real-world dosing is suboptimal, often sitting at 10 to 15 mg per day, well below the range where meaningful anxiolysis occurs.

Because buspirone has a short half-life (approximately 2 to 3 hours), the total daily dose must be divided into twice-daily or three-times-daily regimens. This dosing frequency is a significant adherence barrier. Patients must also be counseled to take it consistently with or without food, because food alters bioavailability, and inconsistent administration relative to meals produces erratic serum levels.

A practical clinical point: patients who have previously been treated with benzodiazepines often report that buspirone "doesn't work." This is a widely observed clinical phenomenon. Early work by Schweizer and Rickels (1986) suggested that prior benzodiazepine exposure may reduce buspirone's subjective anxiolytic effect, though the mechanism remains uncertain. One hypothesis involves neuroadaptive changes at the GABA-A receptor that alter the patient's internal reference point for what anxiolysis feels like. Regardless of mechanism, the practical implication is that benzodiazepine-experienced patients require explicit expectation-setting when transitioning to buspirone. (Schweizer, Rickels, and Lucki, 1986, New England Journal of Medicine)

When buspirone is the right choice

The clinical scenarios where buspirone becomes genuinely valuable are specific: (1) patients with a history of substance use disorder where any benzodiazepine exposure is contraindicated, (2) elderly patients where benzodiazepine fall risk and cognitive effects are unacceptable, (3) augmentation of SSRIs in patients with prominent residual anxiety, particularly when the anxiety co-occurs with depression, and (4) patients who refuse or cannot tolerate serotonergic antidepressants. Outside these scenarios, other agents typically offer better efficacy.

In Plain Language for Families:

Buspirone is a medication for anxiety that works differently from most others your family member might have tried. It does not calm anxiety immediately; it takes 2 to 4 weeks of daily use before it starts working, which can feel frustrating. Think of it as adjusting a thermostat rather than opening a window: the temperature change is gradual, but once it reaches the right setting, it stays there without the risks that come with faster-acting medications. It carries no risk of addiction or withdrawal, which makes it especially useful for people who cannot safely take other anxiety medications. The most important thing to know: if it seems like it is not working in the first week or two, that is expected. Give it time.

3.2 Hydroxyzine

Hydroxyzine is a first-generation antihistamine with FDA approval for anxiety (one of the few antihistamines with this indication). It is widely prescribed in both primary care and emergency departments as a rapid-acting, non-addictive anxiolytic, and clinicians need to understand both where this agent has genuine utility and where its risks outweigh its benefits.

Mechanism

Hydroxyzine produces anxiolysis primarily through potent antagonism of central histamine H1 receptors. Unlike benzodiazepines, it has no activity at the GABA-A receptor, which means it does not produce the classical physiological dependence or withdrawal syndrome associated with benzodiazepines, though rebound insomnia and anxiety after abrupt discontinuation of prolonged daily use have been reported. It also has significant anticholinergic (muscarinic receptor blocking) properties, which produce dry mouth, constipation, urinary retention, and cognitive blunting, and these off-target effects are the source of its most significant long-term risks.

What the evidence shows

Hydroxyzine has genuine short-term efficacy for GAD. In the Hidalgo et al. (2007) effect-size analysis, hydroxyzine achieved an effect size of 0.45 for GAD, though this estimate is based on a small number of included studies and should be interpreted with appropriate caution. (Hidalgo et al., 2007, Journal of Psychopharmacology)

The most rigorous study is Llorca et al. (2002), a randomized, double-blind, placebo-controlled trial comparing hydroxyzine 50 mg/day, bromazepam 6 mg/day (a benzodiazepine), and placebo in 334 GAD patients over 12 weeks. Hydroxyzine was significantly superior to placebo on HAM-A scores and was non-inferior to the benzodiazepine, with a significantly lower rate of rebound anxiety upon discontinuation. (Llorca et al., 2002, Journal of Clinical Psychiatry) A Cochrane review by Guaiana et al. (2010) confirmed hydroxyzine's short-term efficacy for GAD but noted significant heterogeneity across trials and concerns about the overall quality of the evidence base. (Guaiana et al., 2010, Cochrane Database of Systematic Reviews)

The age-dependent risk calculation

Here is where the clinical picture splits sharply by population. In younger adults without cardiovascular risk factors, hydroxyzine 25 to 50 mg as needed for acute anxiety provides rapid relief (onset within 30 to 60 minutes) without the dependence risks of benzodiazepines. For this population, it functions as a reasonable PRN option.

In older adults, the risk-benefit calculation reverses. The 2023 AGS Beers Criteria explicitly lists hydroxyzine as a medication to avoid in adults over 65, due to its strong anticholinergic burden. (AGS Beers Criteria, 2023) Cumulative anticholinergic exposure in the elderly is associated with increased risks of falls, delirium, and incident dementia (the same anticholinergic-dementia signal discussed in the diphenhydramine section below). Hydroxyzine should be actively deprescribed in geriatric populations.

An additional safety concern: hydroxyzine carries a dose-dependent risk of QTc prolongation. At standard therapeutic doses this is generally modest, but in patients with preexisting cardiac conduction abnormalities or those taking other QTc-prolonging medications, this risk becomes clinically relevant and warrants baseline and follow-up ECG consideration.

Dosing

Hydroxyzine comes in two salt forms with different clinical positioning: hydroxyzine HCl (Atarax) is more commonly used for acute anxiety and pruritus, while hydroxyzine pamoate (Vistaril) is more commonly used for chronic anxiety and insomnia due to its slightly different absorption profile. For clinical purposes, the two forms are dose-equivalent.

For acute anxiety (PRN use): 25 to 50 mg as needed, with a maximum of 100 mg per day in divided doses. For scheduled anxiolysis in GAD: 50 to 100 mg per day in divided doses (typically 25 mg two to four times daily). Onset is 30 to 60 minutes. Duration of anxiolytic effect is 4 to 6 hours.

Tolerance to the sedative effects develops rapidly, often within days, which limits its utility for chronic daily anxiety management. This tolerance development is important to communicate to patients: the drowsiness they experience on day one will typically fade within a week, but so may some of the perceived anxiolytic benefit if the effect they noticed was primarily sedation.

When hydroxyzine is the right choice

Best clinical fit: (1) young to middle-aged adults needing a PRN anxiolytic without benzodiazepine risks, (2) acute situational anxiety in settings where benzodiazepines are contraindicated (SUD history, elderly patients not appropriate for daily use), (3) comorbid anxiety and urticaria or pruritus (where the antihistamine effect is independently therapeutic). Not appropriate for: elderly patients (Beers listed), patients on QTc-prolonging medications without monitoring, chronic daily anxiolysis where tolerance undermines benefit, or pregnancy without careful risk-benefit assessment (hydroxyzine crosses the placenta; neonatal withdrawal symptoms have been reported; safety data in pregnancy is limited).

Pause and think: A 72-year-old retired teacher with GAD has failed two adequate SSRI trials. She has a history of alcohol use disorder (in sustained remission for 8 years), osteoporosis with a recent fall, and chronic low back pain. Her daughter asks whether there is "something for anxiety that isn't an antidepressant." Of the three agents covered above, which would you consider, and which would you rule out?

See a model response

Buspirone is the strongest fit. It carries no risk of physiological dependence, no sedation, and no fall risk, which matters in a 72-year-old with osteoporosis and a recent fall. Her alcohol use disorder history makes any agent with abuse potential problematic. Buspirone's effect size is modest (0.17), but in this clinical context its safety profile is the deciding factor, not its ranking on efficacy tables. The key practical requirement: she needs to reach the therapeutic range (20 to 45 mg per day in divided doses), not the subtherapeutic 10 to 15 mg that is commonly prescribed. Expectation-setting is critical since she will not feel any effect for 2 to 4 weeks.

Hydroxyzine is problematic. The 2023 AGS Beers Criteria explicitly lists hydroxyzine as a medication to avoid in adults over 65 due to its anticholinergic burden. In a patient with osteoporosis and a fall history, the sedation and anticholinergic effects compound her existing risk.

Pregabalin has favorable elderly-specific trial data and a faster onset than SSRIs (separating from placebo by approximately day 4). However, it carries dependence potential and is Schedule V, with documented euphoria at supratherapeutic doses. In a patient with a substance use disorder history, even one in sustained remission, that risk profile warrants careful consideration. Pregabalin could be considered with structured monitoring if buspirone proves insufficient, but buspirone is the safer starting point.

3.3 Pregabalin

Pregabalin occupies a paradoxical position in anxiety pharmacotherapy: it has one of the strongest evidence bases of any non-antidepressant anxiolytic, a faster onset than SSRIs, and favorable effect sizes, yet it lacks FDA approval for anxiety in the United States. Understanding why, and what the evidence actually shows, is essential for informed prescribing.

Mechanism

Pregabalin is a high-affinity ligand at the alpha-2-delta subunit of voltage-gated calcium channels. Despite being a structural analogue of GABA, it does not bind to GABA-A or GABA-B receptors and does not alter GABA reuptake. Instead, it works by selectively reducing calcium influx at presynaptic terminals in hyperexcited neurons, which decreases the pathological release of excitatory neurotransmitters (glutamate, norepinephrine, substance P) in the amygdala, hippocampus, and other anxiety-mediating circuits. (Stahl, 2021, Cambridge University Press)

In clinical terms: pregabalin dampens the neural "volume" of overactive fear circuits without the sedation profile of GABAergic agents. However, it is not dependence-free. Prolonged use at therapeutic doses can produce physiological dependence, and abrupt discontinuation may cause withdrawal symptoms including rebound anxiety, insomnia, nausea, and rarely seizures. Gradual tapering is recommended when discontinuing after chronic use. The dependence profile is generally less rapid in onset and less severe than benzodiazepines, but it is real and should be discussed with patients.

What the evidence shows

The effect size data places pregabalin at the top of the non-antidepressant anxiolytics for GAD: 0.50, outperforming SSRIs (0.36), SNRIs (0.42), benzodiazepines (0.38), hydroxyzine (0.45), and buspirone (0.17). The number needed to treat (NNT) for clinical response ranges from 5 to 8. (Hidalgo et al., 2007, Journal of Psychopharmacology)

The landmark trials supporting these numbers are substantial. Feltner et al. (2003), in a double-blind, placebo-controlled, fixed-dose trial published in the Journal of Clinical Psychopharmacology, demonstrated that pregabalin matched lorazepam's anxiolytic efficacy without the withdrawal syndrome upon discontinuation. (Feltner et al., 2003, Journal of Clinical Psychopharmacology) Rickels et al. (2005) showed clear superiority over placebo and efficacy matching alprazolam, with particular effectiveness against both psychic and somatic anxiety symptoms. (Rickels et al., 2005, Archives of General Psychiatry)

An important transparency note: the pivotal pregabalin trials (Feltner, Rickels, Montgomery) were predominantly funded by Pfizer, the manufacturer. Independent replication is limited primarily to post-marketing European data and meta-analytic pooling. This does not invalidate the findings, but the prescriber should be aware that the evidence base is manufacturer-derived.

A clinically important finding: pregabalin separates from placebo by approximately day 4 of treatment, faster than SSRIs or SNRIs (which require 2 to 6 weeks). This rapid onset was demonstrated in active-comparator trials against venlafaxine by Montgomery et al. (2006, 2008, Journal of Clinical Psychiatry; British Journal of Psychiatry), which also specifically validated pregabalin's safety and efficacy in elderly populations.

In the EU, UK, and Canada, pregabalin is formally approved for GAD and positioned as a first-line option in CANMAT and BAP guidelines. (Katzman et al., 2014, BMC Psychiatry; Baldwin et al., 2014, Journal of Psychopharmacology)

Why the FDA rejected it

The FDA rejected pregabalin's GAD applications in 2004 and again in 2009. The rejection was not based on a finding of harm or inefficacy; it reflected the FDA's determination that the submitted monotherapy trial package included too many individual trials that failed to separate from placebo, a notoriously common problem in psychiatric drug trials driven by high placebo response rates. The aggregate meta-analytic evidence strongly favors pregabalin's efficacy, but the FDA evaluates individual trial success, not pooled estimates. This regulatory discrepancy means U.S. practitioners must use pregabalin entirely off-label for anxiety.

Dosing and practical implementation

Starting dose: 75 mg twice daily (150 mg/day). Titration: increase to 150 mg twice daily (300 mg/day) after 1 week if tolerated and needed. Maximum: 600 mg/day in divided doses, though most anxiety efficacy is demonstrated at 150 to 300 mg/day.

Primary side effects are dizziness, somnolence, and dose-dependent weight gain. Compared to SSRIs, pregabalin is associated with lower rates of sexual dysfunction and does not produce initial anxiogenic activation. Compared to benzodiazepines, it demonstrates significantly fewer withdrawal symptoms upon discontinuation.

The abuse concern

This is not a negligible issue. Pregabalin can produce euphoria, dissociation, and relaxation at supratherapeutic doses. It is classified as Schedule V in the United States and as a Class C controlled substance (Schedule 3) in the United Kingdom. Toxicological data shows that 88% of pregabalin poisoning presentations involve co-ingestion (opioids in 41%, benzodiazepines in 36%). In patients with active substance use disorders, particularly opioid or polysubstance use, pregabalin should be prescribed with the same caution as other controlled substances, and PDMP review is appropriate. In patients without SUD history, the abuse liability is lower than benzodiazepines.

When pregabalin is the right choice

Strongest clinical fit: (1) GAD with prominent somatic symptoms (muscle tension, GI distress, autonomic hyperarousal) where SSRIs are inadequate or not tolerated, (2) patients who need anxiolytic effect faster than SSRIs can provide but without benzodiazepine dependence risk, (3) elderly patients with GAD (supported by specific elderly trial data), (4) facilitating benzodiazepine taper (switch studies support this application), (5) treatment-resistant GAD augmentation. Not appropriate for: patients with active SUD (especially opioid), patients where weight gain is a significant concern.

PsychHQ

Alternative Agents I: Buspirone, Hydroxyzine, and Pregabalin