Lurasidone (Latuda)

Lurasidone is the drug you reach for when somebody walks in with bipolar depression and a body that’s already been wrecked by Seroquel. It’s an atypical antipsychotic (the newer-generation antipsychotic class, the Risperdal/Seroquel/Abilify family, used for psychosis but also at lower doses for mood stuff), brand name Latuda, FDA-approved for schizophrenia and bipolar I depression. On paper it does what every other atypical does. In practice it’s one of the few in this class that won’t add thirty pounds and a metabolic panel that looks pretty rough.

Lurasidone is the drug you reach for when somebody walks in with bipolar depression and a body that’s already been wrecked by Seroquel.

The catch, and there’s always one, is that you have to take it with food. Not a cracker, not a piece of toast. At least 350 calories of actual food. Miss that and you’ve absorbed maybe half the dose, sometimes less. Patients will tell me lurasidone “stopped working” when what actually happened is they switched from dinner-with-pill to bedtime-with-pill three weeks ago and haven’t eaten anything substantial after 6 PM since. That food rule is the single most underrated counseling point in the modern antipsychotic kit, it’s printed on the bottle, patients still don’t do it, pharmacists mention it in passing if you’re lucky. By the time somebody comes in saying the drug failed, the first question I’m asking is what they ate when they took it last night, and roughly a third of the time the answer is “nothing, I take it before bed.”

Why I pick it over Seroquel or Abilify

For bipolar depression specifically, the FDA-approved options are narrow. Quetiapine (Seroquel), olanzapine-fluoxetine (Symbyax, which almost nobody prescribes anymore because olanzapine is a metabolic disaster), lurasidone, cariprazine (Vraylar, which got the indication more recently), and lumateperone (Caplyta). That’s basically it for monotherapy with an actual depression indication, which is a short list to be working from for a condition this common.

Seroquel works, that’s not in dispute. It also sedates people into oblivion at the doses that treat bipolar depression (300mg, sometimes 600mg), and it puts weight on almost everybody. The kind of patient who shows up after a couple years on 300mg of Seroquel is usually carrying a chunk of weight gain, with an A1c (a blood test that tells you what your average blood sugar has been for the last three months, basically the diabetes screening number) that’s crept into prediabetic range, sleeping a long stretch on weekends, and going through their life at half-volume. The Seroquel is doing its job for the depression, the cost is just eating them from the other side, and at some point the trade stops making sense.

Cross-tapering somebody like that to lurasidone, starting at 20mg with dinner, working up to 60mg over three weeks, the first ten days are usually rough with akathisia (which I’ll get to in a second). After that it usually settles. The metabolic stuff drifts back in the right direction, the weight comes off without anybody trying, the A1c walks itself back. That’s the trade. You give up the sedation (which some people actually want, and I get it) and you take on a real risk of akathisia in the first month, in exchange for a metabolic profile that ends up almost looking like placebo.

Abilify is the other comparator people ask about. Abilify is mostly weight-neutral, but it isn’t FDA-approved for bipolar depression as monotherapy, only as an add-on for major depression. The data for Abilify in bipolar depression is genuinely underwhelming. Lurasidone has the actual indication and the placebo-controlled trials behind it. If bipolar depression is the target, Latuda beats Abilify on the evidence.

Akathisia is the thing that derails it

Akathisia is internal restlessness. Not anxiety exactly, more like your skin wants to crawl off your body. Patients describe needing to move, legs that won’t quit, can’t sit through a meeting, the sense that something’s wrong but they can’t tell you what. It looks like agitation from the outside and feels like torture from the inside. With lurasidone it shows up in roughly 10-15 percent of patients, usually in the first two weeks, and it’s dose-dependent.

The mistake other prescribers keep making is interpreting akathisia as worsening bipolar agitation and bumping the dose, which makes the akathisia worse, which is how patients quit the drug and add it to the list of things they’ve “tried and failed.” The fix is either to drop the dose, slow the titration, or add propranolol 10-20mg twice a day, which works well for most people and is dirt cheap. Benztropine doesn’t help akathisia much despite being lumped in with the antipsychotic-side-effect cleanup crew. Propranolol is the move. Sometimes a low-dose benzo short-term while you sort out the dose, depending on the patient.

Half the lurasidone failures I’ve watched come down to one of two things… somebody taking it on an empty stomach, or akathisia that nobody named correctly in week two.

If a patient calls me at week one saying they feel jittery and can’t sit still, I don’t tell them to push through. That’s how you lose them, and once you lose somebody on a med they’re not coming back to it. We back off to 20mg, add propranolol, revisit in a week, and most of the time the akathisia fades by week three or four and we can climb back up to where we need to be.

Dosing, food, and the math nobody does

Standard dose range is 20-120mg once daily. For bipolar depression, the trial data shows 20-60mg works about as well as 80-120mg with fewer side effects, so I usually park people at 40-60mg unless they need more. For schizophrenia, 40-160mg, with 80mg being the most common landing spot.

The food thing again, because I can’t say it enough. Below 350 calories of food, absorption tanks. Studies have shown bioavailability (how much of the dose actually gets into your bloodstream) drops by half or more on an empty stomach. So if your patient is taking it before bed with a glass of water, they’re effectively on half their dose, which usually means they’re getting the side effects of a real dose and the efficacy of a fake one.

The way I counsel it is I tell patients to pick one meal and anchor the pill to it. Dinner usually, because they’re more likely to eat enough at dinner than breakfast. Same chair, same time, same routine. The drug doesn’t care if you take it morning or night, only that there’s real food in your stomach when you do.

How it stacks against Vraylar and Caplyta

Vraylar (cariprazine) is the closest comparator. Also bipolar-depression-approved, also reasonably weight-neutral, also has akathisia as the main early side effect. Vraylar has a much longer half-life, which means side effects take longer to clear when you stop. It doesn’t require food. That’s a real advantage for the patient who can’t reliably eat at the same time every day, which honestly is a lot of people with bipolar disorder.

Caplyta (lumateperone) is the newer one, approved for bipolar I and II depression. Similar weight profile, lower akathisia rate than lurasidone or Vraylar in head-to-head-ish comparisons, no food requirement. The downside is it’s still on patent, runs about $1,500 a month without insurance, and a lot of plans don’t cover it without a prior auth fight.

Lurasidone went generic in 2023. Pre-generic it was running $1,300 a month and almost nobody could afford it without good insurance, which is one of those underrated reasons drugs that work end up unprescribed. Now it’s $30-80 a month cash at most pharmacies, which changed who I prescribe it to. Used to be a drug for people with great insurance, now it’s a drug for anybody who can commit to eating dinner. The pharmaceutical industry getting around to letting good medications go generic is the cheapest mental healthcare reform we’ve got going, which honestly says something depressing about how the rest of it is structured.

The honest decision tree: if your patient eats consistently and can tolerate akathisia management, lurasidone first because it’s cheap and the data is solid. If they can’t eat reliably, Vraylar. If money isn’t an issue and they want the lowest side-effect burden, Caplyta. If they need the sedation, Seroquel still has a role despite everything I just said about it.

What I actually tell patients on day one

Few things, in this order. Eat real food when you take it, every single day, no exceptions, because if you don’t you’ll think the drug doesn’t work and you’ll be wrong. If you feel restless or jittery in the first two weeks, that’s probably akathisia and we treat it… call me before you stop the drug. Give it six weeks before you decide whether it’s working, because at week two you’re going to feel the side effects without the benefit yet and that’s the most common quitting point.

The patient-autonomy stuff applies here as much as anywhere else. If you want a specific antipsychotic that I’m not the biggest fan of for your situation, we’ll talk about it honestly and you’ll probably get it. I’m a provider, not a parent. The most I’ll do is a disapproving yes where you walk out with the script you wanted and a clear sense of what I’d watch for and why I wasn’t thrilled. I hardly ever say no. What I will do is push back if I think the patient is making a choice they’re going to regret in six months, because that’s part of the honest take, and then it’s still their call.

What’s actually nice to hear about this drug, if it fits, is the part nobody puts in the side-effect-list at the top. People who land on lurasidone after a year or two on Seroquel almost always describe the same thing once it kicks in. They feel awake during the day for the first time in a while. Their weight starts behaving. Their labs walk themselves back to a place that doesn’t make their primary care doctor anxious. That isn’t a small thing, that’s a couple years of quality of life sliding back in the right direction. The drug has annoying requirements, the food rule is real, the akathisia risk is real, and for the patient it fits, the trade is one of the better ones in this class.

Sources

1. Leucht S, Cipriani A, Spineli L, et al. Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet. 2013;382(9896):951-962. PMID 23810019. (Network meta-analysis of 15 antipsychotics)
2. Calabrese JR, Durgam S, Satlin A, et al. Efficacy and Safety of Lumateperone for Major Depressive Episodes Associated With Bipolar I or Bipolar II Disorder: A Phase 3 Randomized Placebo-Controlled Trial. Am J Psychiatry. 2021;178(12):1098-1106. PMID 34551584. (Caplyta bipolar depression RCT)
3. Loebel A, Cucchiaro J, Silva R, et al. Lurasidone monotherapy in the treatment of bipolar I depression: a randomized, double-blind, placebo-controlled study. Am J Psychiatry. 2014;171(2):160-168. PMID 24170180.
4. Storman D, Koperny M, Styczeń K, et al. Lurasidone versus typical antipsychotics for schizophrenia. Cochrane Database Syst Rev, 2025, Issue 1, CD012429. PMID 39831535. (Cochrane review, evidence rated very-low certainty)