Clear, engaging, and practical updates on clinical psychiatry.

Earn CME for listening to the podcast with a multimedia subscription. Listen for free here or using Apple Podcasts, Android, or Stitcher.

Previous Post
Next Post

Can Antipsychotics Worsen OCD?

Podcast, Volume , Number ,

Print Friendly, PDF & Email

An interview with Michael Poyurovsky, MD, on the paradoxical effects of antipsychotics in OCD and the many ways that OCD symptoms show up in schizophrenia.

Published On: 10/11/2021

Duration: 18 minutes, 3 seconds

Related Article: Can Antipsychotics Worsen OCD?“, The Carlat Psychiatry Report, October 2021


Nearly every treatment algorithm for OCD has antipsychotics somewhere in its branches, but watch out – these medications have a few surprises in store.

Welcome to the Carlat Psychiatry Podcast, keeping psychiatry honest since 2003. I’m Chris Aiken, the editor in chief of the Carlat Report. And I’m Kellie Newsome, a psychiatric NP and a dedicated reader of every issue.

KELLIE NEWSOME: In 1957 Donald Klein introduced a term that we’ve wrestled with on this podcast: Pharmacologic dissection. The idea was that you could define, diagnose, and generally tease apart psychiatric disorders by their response to medications. And if you’re an optimist, the glass if half full on this one, but for many of our patients that glass is half empty, particularly when it comes OCD.

Takes this example. Compulsive behavior is a well-known side effect of psychostimulants like methylphenidate and amphetamine. Watch for it – if the dose goes to high you might see compulsive and repetitive behaviors of all sorts, including nail biting, tics, and overt OCD. 1 out of 14 people who abuse stimulants develop OCD on them.

But flip the glass over, and we see a different picture. Stimulants can also treat OCD. Methylphenidate successfully augmented fluvoxamine in two small randomized controlled trials, even though there are case reports of methylphenidate inducing OCD.

This month we featured an interview with Michael Poyurovsky on the overlap of schizophrenia and OCD. It’s a fairly common syndrome – obsessive compulsive symptoms have been recognized in schizophrenia for decades, and the overlap is informally known as SchizoOCD. There’s a lot to know about this overlap, and in today’s podcast we’re going to focus on a particular paradox it brings up – that antipsychotics can both cause and treat OCD.

I’ll start with two cases from my own practice – one acute and one insidious. As always, we’ve changed some details for confidentiality. Case one: A patient with bipolar disorder came to see me in a mixed state, and I started aripiprazole (Abilify). Within a few days she developed a paralyzing fear that she was going to cause harm to someone – it was full blown OCD – she was unable to drive for fear that she might accidentally roll over someone lying on the road. At home, she feared he might pick up a knife and stab her children. The fear was so intense that she went to the emergency room and was hospitalized. It all resolved when aripiprazole was stopped, and she had no history of OCD before this or since.

Next is the case of a man with intractable psychosis. His illness came on in his early 20’s – and fortunately before it started he was able to complete college and got married to another man. But now he was inching ever closer to losing it all – his marriage and his job – because of unrelenting paranoia and hallucinations. After a few antipsychotic failures I started clozapine, and he recovered completely. But slowly something else started to creep in. He worried about doing everything just right – did he was the dishes in the right order, wash his hands correctly, lock the doors, forget something at home. The obsessive doubting was wearing down his marriage more than the psychosis was. It’s easier for a spouse to detach from psychosis – to separate the person from the illness – but OCD looks enough like normality that it can start to seem intentional, and spouses lose compassion for the behavior. There was no room for intimacy, as the patient badgered his husband day and night with obsessive questions, seeking reassurance for his unquenchable doubt. The marriage ended.

When I treated these patients the literature was full of randomized controlled trials showing that antipsychotics treated OCD – augmenting SSRIs and sometimes working on their own. It was enough to make me wonder if what I was seeing was a false association – random noise that had nothing to do with the medication. But there was an older literature, before the glossy dawn of the atypicals, suggesting that antipsychotics can do the opposite – worsen OCD. It’s clear that dopamine is involved in OCD, but in which direction? So I asked Dr. Poyurovsky. Can antipsychotics flip the switch both ways?

Dr. Poyurovsky: I agree with you that it’s actually the bidirection when you have the positive therapeutic effect of atypical antipsychotics – there is some data on olanzapine and low-dose clozapine in patients with schizo-obsessive disorder. And you have a lot of observations open-label studies showing that the same indications with clozapine and olanzapine primarily may cause or exacerbate the existence of obsessive-compulsive symptoms in patients with OCD.

TCPR: Of all the antipsychotics, clozapine is most likely to cause OCD. How often does this happen on clozapine?

Dr. Poyurovsky: I would say 10%, 10-15% is a reasonable percentage.

TCPR: That’s a lot of patients.

Dr. Poyurovsky: It’s a very big number, big number, yeah. And I’m actually considering full-blown syndrome; if you take into consideration symptoms – obsessive-compulsive symptoms – I believe the numbers are even higher

TCPR: On the other hand you suggested that – in your experience – low-dose clozapine may help OCD in schizophrenia.

Dr. Poyurovsky: Can be effective, low dose. We actually published some pretty good observations several years ago and others as well in the United States that I know who just share these observations that clozapine in a limited dose range, low dose range may be effective in the amelioration of both schizophrenic and obsessive-compulsive symptoms.

TCPR: What is the low dose range?

Dr. Poyurovsky: It is actually 150 mg, and when you keep raising the dose and you up the titration of clozapine you may have the opposite effect of provoking OCD effect. So we have to really watch the OCD eliminating effects versus OCD provoking effects of clozapine.

TCPR: Is that the general consensus that this is a dose-response relationship with most antipsychotics that if you go too high you get more OCD?

Dr. Poyurovsky: Yes, but this is an impression. We don’t have the evidence but this is the impression: the more 5-HT2A antagonism you have with an increased dose, the more OCD-provoking effect.

TCPR: Okay, so it may not be a dopamine effect, it may be serotonin 5-HT2A antagonism?

Dr. Poyurovsky: Right, right.

TCPR: Are there are other 5-HT2A antagonists besides antipsychotics that we should worry about?

Dr. Poyurovsky: It’s actually a good question. Mirtazapine, for example, is a very important 5-HT2 antagonist; there are some cases OCD provoking.

TCPR: Are some antipsychotics more prone to causing OCD?

Dr. Poyurovsky: I think that clozapine and olanzapine are more prone to induce OCD.

TCPR: That’s interesting because quetiapine is an analogue of clozapine, but there are controlled studies in bipolar disorder and OCD where quetiapine TREATED OCD. Can quetiapine also cause OCD?

Dr. Poyurovsky: Quetiapine there are some observations and case reports, but it seems that quetiapine its potential to induce OCD is lower than clozapine.

TCPR: Which antipsychotics are less likely to trigger OCD?

Dr. Poyurovsky: Aripiprazole as I mentioned with its partial dopamine agonist and very modest 5-HT2A antagonist and may be an option. And there are some data from Germany…there was a group that showed actually that the addition of aripiprazole to clozapine may be efficacious in both induction of schizophrenia and subsequent schizophrenia symptoms and obsessive-compulsive symptoms.

TCPR: You mentioned that antipsychotics might cause OCD through serotonin 5-HT2A antagonism. I understand this is all theoretical and we don’t have solid evidence, but do you have any idea how they treat OCD? Is that through dopamine?

Dr. Poyurovsky: I think that the most solid evidence of the involvement of the dopaminergic system in OCD is the therapeutic efficacy of low-dose of D2 dopamine receptor antagonist primarily risperidone, haloperidol in patients with severe OCD

TCPR: What doses are you talking about for those that help OCD?

Dr. Poyurovsky: It is the low dosages of both risperidone. I will say 2 mg of haloperidol, 2.5 to 5 mg of haloperidol,   a very low dose of these medications.

TCPR: And for risperidone, what dosage?

Dr. Poyurovsky: It’s 2-3 mg – definitely not 6.

As I talked with Dr. Poyurovsky, I realized we are getting into some confusing territory. There are a lot of unknowns. Antipsychotics affect many different neurotransmitter systems, and pharmacologic dissection with them is not going to be a straight line. He mentioned that aripiprazole is more favorable for OCD, but one of the cases I reported on here was of severe OCD on aripiprazole. He mentioned that clozapine is the most likely to cause OCD, but clozapine’s structural analogue quetiapine is among the least likely to cause it.

In our online interview Dr. Poyurovsky shares more about antipsychotics that are more or less favorable in schizoOCD, the risks and benefits of using SSRIs in this population, as well as other pharmacologic and non-pharmacologic treatments that he finds helpful for patients whose psychotic symptoms overlap with OCD.

KELLIE NEWSOME: This is a very dicey area of psychopharmacology. Schizophrenia can cause OCD even without medications on board, and antipsychotics can cause or worsen OCD. And the problem extends beyond schizophrenia. Bipolar disorder doesn’t exactly cause OCD, but OCD is much more common in the bipolar population then it is in the general public. So when you see a patient with these comorbidity, it’s hard to know if it’s a separated condition, a medication effect, or if the medication is helping it – just not all the way.

CHRIS AIKEN: Or it could be a false positive. Lots of patients have OCD-like symptoms who don’t have OCD. I’ve seen people with post-traumatic stress disorder count things to self-sooth – and they don’t do it out of any compulsion – it’s done in response to clear PTSD triggers. I’ve seen compulsive behaviors in ADHD that patients do when they are bored and distracted, and it goes away when treated with a stimulant.

KELLIE NEWSOME: Dr. Poyurovsky also mentions in his interview that some patients with schizophrenia have OCD-like symptoms that aren’t due to OCD at all – that are false positives. For example, akathisia can make people ask repetitive questions and seek reassurance in a compulsive way.


This observation just might be…if your patient comes to you and starts to ask you repetitive questions, despite your repetitive answers, be careful to question akathisia if you’ve recently started antipsychotics. Akathisia associated with marked obsessional symptoms associated with akathisia should be dealt with, and it’s not an easy clinical situation. There’s both an obsessional and a compulsive component. For example, repetitive, intrusive asking questions.

CHRIS AIKEN: And while akathisia may be a false positive, he pointed out that patients with genuine schizoOCD are more likely to have akathisia on antipsychotics. He also described cases that resemble body dysmorphic disorder in OCD:

Dr. Poyurovsky: Sometimes it is typical BDD (body dysmorphic disorder) like non-schizophrenia BDD. Sometimes again it’s a complex disorder when delusional thinking and more delusional compulsivity. But within schizophrenia you may have very complex psychological constructs associated with both delusions, hallucinations and body dysmorphic experiences. So where you have OCD you have to look at additional components; they run together. There is multimorbidity of these obsessive patients.

CHRIS AIKEN: Modern psychopharmacology was born in the 1950’s, and we’ve come a long way in sorting out which psychotropics are best for which disorders. But the dissection is not as clean as we hoped for. Antidepressants can treat anxiety, and cause anxiety; they can even trigger suicidal thinking in patients under age 25. One of the earliest examples of pharmacologic dissection was mood disorders, which used to be lumped together in one category – manic depression – regardless of whether the patient suffered from mania or depression. As doctors started to realize that some patients got better with antidepressants, while others did better on mood stabilizers, they split the category in two: bipolar and unipolar, which first appeared in the 1980 DSM-III. But that split wasn’t so clean, and within months of DSM-III’s publication articles started appearing arguing that some depressed people responded to lithium – a mood stabilizer. And while many bipolar patients got worse on antidepressants, some – a small minority – seemed to respond to them. And just because a patient’s mood gets worse on an antidepressant, it doesn’t mean they have bipolar disorder. To learn more about conditions that worsen on antidepressants, check out our February 15th podcast from earlier this year, Six Depressions that can Worsen on Antidepressants.

KELLIE NEWSOME: Michael Poyurovsky is a professor of psychiatry at the Ma’ale HaCarmel Mental Health Center, Tirat Carmel, Israel. He is the author of the textbook Schizo-Obsessive Disorder and is releasing a book on the subject soon.

And now for the word of the day….Failed trial

CHRIS AIKEN: Sometimes drugs fail, and sometimes studies fail. But when you see a disappointing outcome in a clinical trial, how do you know if the fault lies with the drug or the study? That’s where you need to understand the difference between a negative trial and a failed trial.

In a negative trial, it’s the drug that’s the problem. The drug came up negative, no better than placebo, it simply didn’t work. A good example is gabapentin (Neurontin), which failed to separate from placebo in several randomized controlled trials of mania – including one where the placebo arm worked better. Atomoxetine (Strattera) for depression, aripiprazole (Abilify) for ADHD, and vortioxetine (Trintellix) for generalized anxiety disorder are other examples that come to mind.

In a failed trial, it’s the study itself that was the problem. Maybe it didn’t enroll enough subjects, so was “underpowered” to detect a difference between drug and placebo, or maybe there were too many drop outs. Or maybe the placebo response was inflated. That can happen when the investigators provide a lot of support to the study subjects, or they enroll a lot of people with mild conditions that are highly responsive to placebo.

When you see a trial with a disappointing result, how do you know if it was the drug or the study that failed? Sometimes the investigators will try to cast blame on the study, as if they believe that there must be a grain of truth in their hypothesis that the study failed to detect. Often they argue that the placebo response was higher than expected. But that is all speculation. The true test of whether a study failed is a rare one – the trial would have to have 3 arms – a placebo, the new drug, and a similar drug that is already well-known to work. The placebo is there to test the new drug, and the other well-established drug is there to test the study. For example, one of the schizophrenia trials for lumateperone (Caplyta) included 3 arms – placebo, lumateperone, and the old stand-by risperidone. Lucky for lumateperone, both drugs surpassed placebo to about the same degree.

Lurasidone (Latuda) has a failed study in schizophrenia where neither lurasidone nor haloperidone separated from placebo. Escitalopram (Lexapro) has a failed study in geriatric depression, where neither it nor fluoxetine (Prozac) separated from placebo. One of the most famous examples of a failed study was a randomized controlled trial of St. John’s wort from 2002. In that study was large and well-designed, enrolling 360 patients with major depression, but neither St John’s wort, nor the otherwise effective sertraline, worked any better than placebo. This study is often presented as proof that St John’s wort doesn’t work, or even that sertraline doesn’t work, but the truth is much more mundane: It was the study that failed, full of sound and fury but – ultimately – signifying nothing.

This month we report on a series of negative studies of an atypical antipsychotic in mania, and in next weeks podcasts we’ll tally the best and worst antipsychotics for mania.

Got feedback? Take the podcast survey.


Leave A Comment