Prescribing Aripiprazole

Why Aripiprazole Matters

Aripiprazole is the antipsychotic you reach for when you want dopamine coverage without the metabolic and prolactin baggage that made the second-generation agents infamous. It was the first dopamine partial agonist on the market, and that one mechanistic fact accounts for nearly everything useful about it: minimal weight gain, no meaningful hyperprolactinemia, low sedation, and a low burden of extrapyramidal side effects. Among the atypicals it belongs with ziprasidone and lurasidone in the "metabolically clean" group, and unlike ziprasidone it does not meaningfully prolong the QT interval.

Keep what aripiprazole is good at separate from what its marketing once claimed. When it launched in 2002, the pitch was that a "dopamine stabilizer" would be mechanistically superior for psychosis and negative symptoms. It is not. In the pivotal Kane trial, aripiprazole 15 and 30 mg matched haloperidol 10 mg on both positive and negative symptoms, no better and no worse. Its advantage over haloperidol was entirely in tolerability, not efficacy. The old Carlat Report summed it up: "the most perfect atypical, but enough about its mechanism."

Part 1: Indications

FDA-Approved Uses

• Schizophrenia (acute and maintenance; adults and adolescents 13-17)
• Bipolar I disorder, acute manic and mixed episodes (monotherapy or adjunct to lithium/valproate; ages 10+)
• Bipolar I maintenance (monotherapy or adjunct)
• Adjunctive treatment of major depressive disorder, augmentation of an antidepressant in inadequate responders (its best-known "add-on" indication)
• Irritability associated with autism spectrum disorder (ages 6-17)
• Tourette's disorder (ages 6-18)

The Evidence-Based Clinical Uses

Schizophrenia, solid but not special. Aripiprazole is a first-line antipsychotic, but choose it for tolerability, not power. In the Kane trial (414 acutely relapsed patients), aripiprazole 15 mg and 30 mg equaled haloperidol 10 mg on positive and negative symptoms and differed only in a cleaner side-effect profile. Effect sizes across the atypicals are broadly similar, with clozapine the exception. Aripiprazole is not clozapine and is not the agent for treatment-resistant psychosis. It fits best in the young, first-episode, metabolically vulnerable patient in whom you want to avoid weight gain, sedation, and prolactin elevation from day one.

MDD augmentation, its main use, honestly appraised. General psychiatry uses aripiprazole for this more than for anything else, and the data are positive but modest. Two FDA-registration RCTs (SSRI/venlafaxine non-responders randomized to adjunctive aripiprazole vs placebo) showed remission of 25-26% vs 15-16%, about a 10% absolute advantage, NNT ~10. But MADRS separated from placebo by only ~2.8 points in one trial, and patient-rated depression scales did not separate. The VAST-D trial (1,522 VA patients failing an antidepressant) gave the most clinically useful comparison: augmenting with aripiprazole produced 29% remission vs 22% for switching to bupropion (NNT ~10), with a higher response rate, but at a metabolic and akathisia cost.

Bipolar mania, a strength. For acute mania, aripiprazole works well: meta-analysis shows ~50% response vs ~35% placebo, both as monotherapy and as adjunct when lithium or valproate alone falls short. This is one of its stronger efficacy signals.

Bipolar I maintenance, real, but read the fine print. FDA-approved on the strength of ~18-month relapse-prevention RCTs. The design was enriched: only patients who had already responded acutely to aripiprazole were randomized. That inflates apparent maintenance efficacy and limits generalizability, the same statistical caveat that applies to lithium's and lamotrigine's maintenance approvals. It delays time to manic relapse more convincingly than depressive relapse.

Where Aripiprazole Underperforms

• Bipolar depression (monotherapy): it does not work. Two acute trials and meta-analysis found aripiprazole no better than placebo for bipolar depression. It is not FDA-approved for this. Do not reach for it here; use quetiapine, lurasidone, lumateperone, or cariprazine, which have the data.
• The "happily-ever-after fallacy": the fact that aripiprazole maintains remission in bipolar I does not mean it treats acute bipolar depression. Prevention and acute treatment are pharmacologically different jobs.
• ADHD: the theory that dopamine partial agonism would help was tested and failed in two negative RCTs. Don't.

Off-Label Uses With Some Support

• Borderline personality disorder: among antipsychotics, aripiprazole has some of the more positive trial data for anger and cognitive-perceptual symptoms, though the data are modest and come from small studies.
• Bipolar II depression: post-hoc and open-label data suggest benefit specifically at low doses (≤5-10 mg/day), lower than the bipolar I mania range. Off-label; weigh carefully.
• OCD augmentation, Tourette's: low-dose antipsychotic augmentation has a role; risperidone is better studied for OCD, but aripiprazole is FDA-approved for Tourette's.

Part 2: Workup and Candidacy

Aripiprazole is metabolically gentle, so the workup is lighter than for olanzapine or quetiapine. The metabolic baseline still matters because any antipsychotic can shift glucose and lipids, and because your patient may later be switched to a dirtier agent and you'll want something to compare against.

Baseline Workup

Aripiprazole is one of the atypicals for which a screening ECG is not mandatory in a patient without cardiac history, because it does not meaningfully prolong the QT interval.

Who Is a Poor Candidate?

• Elderly patients with dementia-related psychosis: boxed warning; increased mortality (see Part 6).
• Akathisia-prone patients: anyone who developed intolerable restlessness on a prior antipsychotic is at high risk of a repeat; consider a lower-akathisia alternative (quetiapine) or commit to very slow titration plus prophylaxis.
• Patients who need acute sedation: aripiprazole is activating far more often than sedating, which makes it a poor choice when you need to calm an agitated patient quickly.
• Acute bipolar depression: wrong drug (see Part 1).

Part 3: How to Start and Dose

The most important dosing principle is that the right dose depends entirely on the indication, and lower is usually better. Depression augmentation lives at 2-5 mg; psychosis and mania at 10-30 mg. Starting a depressed patient at 15 mg gets this backwards, and it is the most common way to cause needless akathisia.

Formulations

A digital pill version (Abilify MyCite, with an ingestible sensor) exists but is rarely used in practice.

Dosing by Indication

MDD augmentation (start low; that is the key):

• Start 2-2.5 mg once daily (halve a 5 mg tablet).
• Increase to 5 mg after ~5-7 days if tolerated and no early response.
• The evidence sweet spot is 2-5 mg. A 2022 meta-analysis found the odds ratio climbing from 2 mg (1.46) to 4 mg (1.87) to 5 mg (1.91) and then plateauing, so doses beyond 5 mg add tolerability cost without added benefit.
• Ceiling for augmentation: 10 mg. If no response at 10 mg after 2-4 weeks, switch strategies.
• Dose in the morning: stimulation outnumbers sedation roughly 2:1.

Schizophrenia:

• Start 10-15 mg once daily (15 mg is the usual effective starting and target dose).
• Range 10-30 mg/day; 30 mg is not more effective than 15 mg for most and brings more akathisia.
• In a first-episode or sensitive patient, starting at 5-10 mg and titrating up is reasonable to blunt akathisia.

Bipolar mania (mono or adjunct):

• Start 15 mg once daily, range 15-30 mg. Most trial patients did well at 15 mg.
• As adjunct to lithium/valproate for partial response: 15 mg is typical.

Bipolar II depression (off-label): if used at all, stay low (≤5-10 mg); higher doses were less effective in post-hoc analyses.

Autism irritability (ages 6-17): start 2 mg/day, titrate slowly to 5-15 mg based on response and tolerability.

Intellectual disability: start 2 mg for 1 week, then 5 mg for 1 week, with careful staff monitoring (higher NMS risk and fatality in this population).

Onset and Steady State

Aripiprazole has a long half-life (~75 hours; its active metabolite dehydro-aripiprazole ~94 hours), so steady state takes about 2 weeks. That has two practical consequences. Don't chase early non-response with rapid dose escalation; give each dose ~1-2 weeks. And the drug is slow to wash out, which is forgiving of a missed dose but unforgiving when a CYP inhibitor is added (see Part 7).

Dose Adjustments and Special Populations

• CYP interactions (a defining feature): aripiprazole is a CYP2D6 and CYP3A4 substrate. Halve the dose with a strong 2D6 inhibitor (fluoxetine, paroxetine) or a strong 3A4 inhibitor (ketoconazole, clarithromycin); give one-quarter with both, or a 2D6 poor metabolizer plus a 3A4 inhibitor. Roughly double the dose over 1–2 weeks with a strong 3A4 inducer (carbamazepine). LAIs have their own reduction tables.
• Hepatic / renal impairment: no dose adjustment.
• Geriatric: no specific dose change, but the class boxed warning applies: increased mortality in elderly patients with dementia-related psychosis.
• Pediatric: approved for schizophrenia (13–17), bipolar mania (10–17), autism irritability (6–17), and Tourette (6–18) at the indication doses above.
• Impulse-control behaviors: compulsive gambling, shopping, eating, or hypersexuality can emerge at any dose. Ask about them, and reduce or stop the drug if they appear.

Stopping and Switching

• Discontinuation: the long half-life self-tapers somewhat, but taper deliberately after prolonged use to reduce rebound psychosis and withdrawal dyskinesia.

Part 4: Monitoring

Aripiprazole is a "metabolically clean" antipsychotic, so it gets the lighter of the two standard monitoring schedules. Lighter does not mean none: the boxed and class risks (metabolic drift, tardive dyskinesia) still require surveillance.

Act on These

• ≥5-7% weight gain from baseline in the first weeks: intervene early with lifestyle counseling, and consider starting metformin proactively (it works best begun early; see Part 5). Aripiprazole's weight liability is low but not zero, especially in the antipsychotic-naive.
• Any new involuntary movements on AIMS: document, consider dose reduction or switch, and discuss tardive dyskinesia (see Part 5). TD is not a benign afterthought here: the class rate is ~3.9%/year (~5%/year in the elderly).
• New/worsening restlessness: think akathisia first (see Part 5); it is the most likely and most treatable complaint.

Part 5: Side Effects and How to Manage Them

Aripiprazole's side-effect profile is the whole reason to choose it, so protect it. Most patients tolerate it well. The complaints that drive discontinuation are, in order, akathisia, activation/insomnia, and (less often) modest weight gain. Manage those three well and most patients stay on the drug.

Akathisia

Akathisia is aripiprazole's characteristic liability and the side effect that defines how you manage the drug. It is an inner sense of restlessness with an urge to move, and it can be present even when the patient isn't visibly moving, so you have to ask. Rates are ~25% in depression-augmentation trials and ~15-19% in mania trials. It is deeply distressing and independently raises suicide risk, so take it seriously.

Management, in Order

1. Prevent it. A low starting dose and slow titration "dramatically reduce" incidence, and prevention is by far the most effective intervention. Much of the drug's historical reputation for akathisia came from starting too high.
2. Reduce the dose (often the fastest fix, since akathisia is dose-related).
3. Benzodiazepine: first-line pharmacotherapy for rapid relief, clonazepam 0.5 mg BID standing ± PRN (supported by small RCTs). It bridges symptom control while you adjust.
4. Propranolol: 20 mg BID-TID, titrated up to ~80-240 mg/day as tolerated (check pulse; hold if under 60 bpm). Switch to extended-release once the effective dose is found.
5. Mirtazapine: 15 mg has RCT support, with relief comparable to propranolol in a placebo-controlled trial.
6. Do NOT reach for benztropine or other anticholinergics. They do not treat akathisia, despite a persistent and harmful belief that they do; they treat drug-induced parkinsonism.

Activation, Insomnia, and (Less Often) Sedation

Aripiprazole is more likely to activate than sedate (stimulation outnumbers sedation roughly 2:1), and activation is worst early.

• Dose in the morning by default.
• Reassure: activation often improves by around week 6 as tolerance develops.
• If insomnia persists, it's reasonable to add short-term sleep support rather than abandon the drug.
• Sedation (roughly 10% at low doses) is the minority experience; if it happens, switch the dose to evening.

Weight and Metabolic Effects

Aripiprazole is in the low-liability tier with ziprasidone and lurasidone, with minimal weight gain and minimal effect on glucose, triglycerides, and cholesterol, largely because it lacks the antihistamine (H1) activity that drives olanzapine's and quetiapine's weight gain. But low is not zero, especially in antipsychotic-naive patients: in VAST-D, 25% gained ≥7% of body weight by 36 weeks (vs 5% on bupropion).

• Track weight early; catch gain in the first 1-2 months.
• Metformin is the best-evidenced agent (superior to topiramate, which carries cognitive and renal-stone costs). Start 500 mg XR with the largest meal, increase to 500 mg BID after a week, target 750-1,000 mg BID. It works best started early, before the weight is on. Metformin also improves A1c, lipids, and insulin sensitivity. Check CMP/creatinine and B12 periodically.
• Diet and exercise. If the patient is switching from a dirtier agent, recognize that some of the weight burden may have preceded aripiprazole.

Movement Disorders (EPS and Tardive Dyskinesia)

• Acute EPS/parkinsonism: low risk, and one of aripiprazole's advantages (partial agonism preserves nigrostriatal dopamine tone). If it occurs, reduce dose or add benztropine 0.5-2 mg (parkinsonism only, not akathisia).
• Tardive dyskinesia: the class risk is ~3.9%/year (~5%/year in the elderly), and TD has been reported even at 2 mg. The risk is not zero; it requires informed consent and twice-yearly AIMS exams. If TD emerges, options include dose reduction, switching, or a VMAT2 inhibitor (valbenazine, deutetrabenazine).

Prolactin

Aripiprazole tends to lower prolactin at therapeutic doses, because of its partial agonist activity at D2. Brexpiprazole and cariprazine, the other D2 partial agonists, behave the same way, so this belongs to the class rather than to aripiprazole alone. You can put that to use: adding aripiprazole 5 mg is the standard strategy to correct hyperprolactinemia caused by another antipsychotic (e.g., risperidone/paliperidone). Prolactin falls across the usual dose range rather than rebounding at the top of it, so hyperprolactinemia is a non-issue on aripiprazole itself.

GI and Anticholinergic

Nausea, constipation, and dry mouth are modestly more common than on placebo, particularly early. They are usually mild and self-limited, and standard measures (hydration, fiber, bowel regimen if needed) suffice. Aripiprazole's anticholinergic burden is low.

Cognitive Dulling

Some patients report feeling "cloudy-headed," more so at higher doses (where the drug behaves more like a D2 antagonist). If a patient complains of brain fog, lower the dose. That often resolves it, and it is another argument for staying at the low end.

The Rare-but-Serious

• Neuroleptic malignant syndrome (NMS): rare and class-wide. Fever, rigidity, altered mental status, autonomic instability, elevated CK. Stop the drug, give supportive care, and consider dantrolene/bromocriptine. Risk and fatality are both higher in intellectually disabled patients.
• Impulse-control behaviors: a distinctive effect related to the dopamine-agonist activity: pathological gambling, compulsive shopping, hypersexuality, binge eating. Warn patients up front and ask about it; it resolves with dose reduction or discontinuation.
• Seizure threshold: modestly lowered (class effect); relevant in patients with a seizure history.

Part 6: Overdose, Toxicity, and Boxed Warnings

Overdose and Toxicity

Aripiprazole is comparatively safe in overdose, far safer than lithium or TCAs. There is no specific antidote, so management is supportive: airway protection, cardiac monitoring, IV fluids for hypotension, and treatment of agitation. Its long half-life means effects (and any sedation or tachycardia) can be prolonged. Serious cardiotoxicity is uncommon given the minimal QT effect.

Other Class Safety Points

• NMS (see Part 5).
• Orthostatic hypotension: modest alpha-1 effect, more relevant in the elderly and with rapid titration.

Part 7: Drug Interactions

Aripiprazole is a substrate of CYP3A4 and CYP2D6, and its long half-life makes interactions matter more than you'd expect. An inhibitor raises the level and also makes an already long-acting drug "ultra-long-lasting."

The Three Adjustments to Remember

Practical Management

• When starting a strong inhibitor in a patient already on aripiprazole, cut the dose promptly. CYP inhibition is immediate, and the long half-life means the level keeps climbing for days.
• Sustained interactions require attention even with the LAI formulations; the label carries specific dose adjustments for LAIs used with strong 2D6/3A4 inhibitors or inducers for more than 14 days.

What It Does NOT Meaningfully Interact With

• Most SSRIs/SNRIs other than the 2D6 inhibitors. Escitalopram and venlafaxine are the clean examples. Duloxetine is a moderate 2D6 inhibitor by its own label and sertraline inhibits 2D6 at the top of its dose range, so those two do not belong here.
• Lithium and valproate (commonly co-prescribed as adjunct in bipolar).
• Pharmacodynamically: watch additive sedation with other CNS depressants and additive orthostasis with antihypertensives, but these are minor with aripiprazole.

Part 8: Special Populations

Pregnancy

Among the atypicals, aripiprazole has a growing body of reassuring data and is generally considered one of the reasonable options when an antipsychotic is needed in pregnancy. The available cohorts show no clear signal for major malformations, though the data remain more limited than for older agents. As with all antipsychotics:

• Weigh the substantial risk of untreated psychotic or bipolar illness (relapse, self-harm, poor prenatal care) against drug risk; often the balance favors continuing.
• Third-trimester exposure carries a class risk of neonatal extrapyramidal signs and withdrawal (agitation, feeding difficulty, respiratory issues); coordinate with OB and pediatrics.
• Manage as a multidisciplinary decision; use the lowest effective dose.

Lactation

Aripiprazole passes into breast milk in small amounts. A concern specific to aripiprazole is that its dopamine-agonist activity can suppress prolactin and so impair lactation/milk supply. If a mother chooses to breastfeed, monitor the infant for sedation and poor feeding, and monitor milk supply. Decisions should be individualized with pediatrics.

Elderly

• Not for dementia-related psychosis (boxed warning).
• Otherwise, aripiprazole is one of the preferred agents in older adults for its low anticholinergic burden, low orthostasis relative to clozapine/quetiapine, and metabolic friendliness.
• Start low, go slow: titrate weekly rather than daily.
• TD risk is higher (~5%/year); do the AIMS.
• In OPTIMUM, aripiprazole augmentation carried a lower fall rate than bupropion augmentation in older depressed patients (0.33 vs 0.55 falls per patient), which counts in its favor.

Renal and Hepatic Impairment

Aripiprazole is hepatically metabolized (CYP3A4/2D6) with minimal renal excretion. The label requires no dose adjustment for hepatic impairment at any severity, up to Child-Pugh 15, or for renal impairment down to a GFR of 15 mL/minute.

CYP2D6 Poor Metabolizers

Reduce the initial dose to 50% of usual (FDA label), then titrate to response. Poor metabolizers achieve higher levels and are more prone to dose-related side effects, including akathisia.

Children and Adolescents

• FDA-approved: schizophrenia (13-17), bipolar mania (10+), autism irritability (6-17), Tourette's (6-18).
• Youth are more metabolically vulnerable, so monitor weight, glucose, and lipids closely even though aripiprazole is comparatively clean.
• Start low. For autism irritability, address non-pharmacologic triggers and milder options first, because the metabolic and neurologic risks are real even for a "clean" agent.

Part 9: Discontinuation and Switching

Stopping aripiprazole does not bring a lithium-style rebound catastrophe, but two discontinuation phenomena deserve respect.

Taper, Don't Stop Abruptly

• A reasonable taper is over 2-6 weeks; longer for patients on high doses or with a history of severe/recurrent episodes.
• Abrupt cessation risks withdrawal-emergent dyskinesia (choreiform movements, peaking 1-4 weeks after stopping) and dopamine-supersensitivity/rebound psychosis from D2 upregulation during chronic exposure. A slow taper mitigates both.
• Because of the ~75-hour half-life, aripiprazole self-tapers somewhat on washout, which is more forgiving than short-acting agents, but a deliberate taper is still preferred.

Switching TO Aripiprazole Is Where People Get Burned

• Cross-titrating from another antipsychotic to aripiprazole can trigger psychotic exacerbation. In one analysis, 27% had a meaningful PANSS worsening within 28 days, 62% of those within the first week. Risk is higher with high baseline antipsychotic doses and prior first-generation use.
• Aripiprazole's partial agonism means it occupies D2 receptors but provides only partial blockade. If you yank a full antagonist too fast, the relative "dopamine surge" at supersensitive receptors can flare psychosis.

Part 10: Aripiprazole vs the Alternatives

Differences in efficacy among the atypicals for psychosis are gradual rather than categorical, with clozapine ahead of the rest, so outside treatment resistance you choose by side-effect liability, cost, and patient factors. Aripiprazole's niche is tolerability.

So when do you pick aripiprazole? When tolerability is the priority: the metabolically vulnerable patient, the one who can't tolerate prolactin elevation or sedation, the depressed patient needing augmentation, the young first-episode patient in whom you want a clean start, or as an add-on to a dirtier antipsychotic to rescue metabolic or prolactin side effects.

Mechanism

Aripiprazole is a partial agonist at the dopamine D2 receptor, which distinguishes it from every older atypical, since those are pure D2 antagonists. A partial agonist binds D2 tightly but delivers only a fraction of dopamine's intrinsic activity. The result is context-dependent "dopamine stabilization":

• Where dopamine is excessive (mesolimbic pathway, positive psychotic symptoms), aripiprazole competes it off the receptor and acts as a net antagonist and treats psychosis.
• Where dopamine is needed (nigrostriatal pathway, movement), aripiprazole provides enough intrinsic activity to preserve tone, hence the low EPS/TD risk.
• In the tuberoinfundibular pathway (prolactin control), the agonist activity suppresses prolactin, the source of the paradoxical prolactin-lowering effect.

It is also a partial agonist at 5-HT1A (buspirone-like, and plausibly part of its anxiolytic/antidepressant effect) and an antagonist at 5-HT2A. Its long half-life (~75 h; active metabolite ~94 h) makes it a smooth, forgiving once-daily drug and the natural base for long-acting injectables.

The elegant mechanism was oversold for efficacy; aripiprazole is not superior for psychosis or negative symptoms. But it is a completely satisfying explanation for the drug's real advantage, a clean side-effect profile.

Bedside Cheat Sheet