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

Research Theme Park: Medland

Podcast, Volume , Number ,

Print Friendly, PDF & Email

Lithium takes on the coronavirus. Benzos impair facial recognition. Antipsychotic doses find their maximum, the profile of a lithium responder, and topiramate for compulsive shopping.

Published On: 4/21/20

Duration: 13 minutes, 51 seconds

Studies Referenced:

Got feedback? Take the podcast survey.

Rough Transcript:

Dr. Aiken: The Research Theme Park is a magical place in the Carlat Psychiatry Podcast that showcases papers that are practical, playful, and reveal a little bit about human nature. Today, we’ll take you to a quiet place in the park, Medland, where the clinical trials that inform our practice take place. It’s empty now, as the stay-at-home order has shut them down to prevent the spread of the coronavirus. All that’s left are graphs and papers, and we’ll pick up a few of the latest ones to share with you here. 

Welcome to the Carlat Psychiatry Podcast, keeping psychiatry honest since 2003. I’m Chris Aiken, the editor in chief of the Carlat Report. 

Kellie: And I’m Kellie Newsome, a psychiatric NP and a dedicated reader of every issue. 

Chloroquine and Mania
Chloroquine and hydroxy chloroquine have been used as antimalaria agents for over 70 years, and lately they’ve created a stir as potential therapies for COVID-19. In bench research, aka “in vitro” studies – they’ve shown an ability to keep the virus from entering and exiting cells. 

What’s more controversial is whether they will help in person like they do in a petri dish. Chloroquine has already failed to live up to its in vitro promise with other viral illnesses like the Ebola virus disease, chikungunya, in-fluenza, HIV infection, and dengue. In the case of the chikungunya virus, patients had more long-term complications when they took Chloroquine – possibly because Chloroquine can impair immune function, inhibiting T cell and interleukins. 

The Chloroquines have small uncontrolled studies showing benefits against COVID-19 from France and China, and RCTs are being expedited. 

While we await those results, many people are already taking Chloroquine for COVID-19 and Kellie and I have already some in practice. As psychiatrists, there’s a few things you need to know about the Chloroquines. They are associated with psychiatric side effects, most notably psychosis, mania, and mood lability. 

And here’s a little known fact for these viral times. Lithium has antiviral properties against over a dozen viruses, and has even been used successfully as a treatment for herpes virus. Lithium improves immune function, but it also impairs a virus’ ability to replicate. This month, a group of physicians in Poland put forth a paper suggesting that lithium should be studied against the current coronavirus, officially known as severe acute respiratory syndrome coronavirus 

  1. So far, lithium has proved effective against 4 other coronavirus strands in vitro and animal studies. While we await those results, remind your patients on lithium that viral infections are a risky time for lithium toxicity. They should stay hydrated, avoid NSAIDs and use acetaminophen instead for temperature control, and if severe nausea or diarrhea ensues, hold lithium until it resolves. 

Antipsychotic Dosing: How High is Too High?

Despite decades of research, we don’t actually know much about dose-response relationship of antipsychotics in schizophrenia. This meta-analysis set out to find the near-maximum effective doses for 20 antipsychotics. They wanted to know if patients got better as the dose went higher, or if their response just leveled off. 


The main result: Most antipsychoitcs follow the law of diminishing returns. As the dose stretches beyond a certain point, the average patient doesn’t get any better. They figured this out by graphing the dose-response curve for each antipyshoctic. Most of the graphs plateaued or became a bell shaped curve as the dose went higher. The specific doses they arrived at were often a little below the max dose in the PDR: 

  • aripiprazole, 11.5 mg/day 
  • aripiprazole LAI (lauroxil), 463 mg every 4 weeks 
  • asenapine, 15.0 mg/day 
  • brexpiprazole, 3.36 mg/day 
  • haloperidol, 6.3 mg/day 
  • iloperidone, 20.13 mg/day 
  • lurasidone, 147 mg/day 
  • olanzapine, 15.2 mg/day olanzapine LAI, 277 mg every 2 weeks 
  • paliperidone, 13.4 mg/day paliperidone LAI, 120 mg every 4 weeks 
  • quetiapine, 482 mg/day 
  • risperidone, 6.3 mg/day risperidone LAI, 36.6 mg every 2 weeks 
  • ziprasidone, 186 mg/day 

However, this was not a hard-and-fast rule. For 6 of the 20 meds, the benefits kept rising as the dose went up. These were:

  • Illoperidone 20
  • lurasidone 160
  • olanzapine 15
  • paliperidone 13
  • ziprasidone 200 

Topiramate tested for Compulsive Buying disorder. But we’re not buying it. 

Topiramate is often used in impulsive and compulsive disorders, like gambling, binge eating, bulimia, OCD, binge drinking, cocaine/amphetamine abuse, and irritability. This study tested it out in Compulsive Buying disorder, but the results tell us more about how to read a study than whether we should use topamax in this population. It was a 12-week randomized placebo controlled trial. Sounds good so far – 12 weeks is pretty long. Only 50 patients though, which is a bit small. We’d like to see studies with 100 or more subjects. The main problem though is that the medication only made a difference on secondary measures. 

Kellie: Yes but some of the secondary outcomes were positive, so isn’t that like the glass is half full and half empty? 

Dr. Aiken: Yes, but it’s the wrong half. Research studies are like gambling – in fact the statistical laws that we use to interpret studies have the origins in games of chance. Compulsive buying disorder is not in the DSM – the proposed criteria for it are: 

  1. Over-preoccupation with buying. 
  2. Distress or impairment as a result of the activity. 
  3. Compulsive buying is not limited to hypomanic or manic episodes.[33] 
  4. Constant obsessing with buying as well as being dissatisfied all the time. 

Who gets better on lithium? 

This retrospective study of 900 patients from the Consortium of Lithium Genetics is one of the largest studies to date of lithium response in bipolar I disorder. Patients who responded to lithium tended to have more manic episodes than depressive ones. 

Their illness often began with a mania, and they had an age of onset between 15-32. They were more likely to have bipolar disorder in their family history, suggesting a stronger genetic 

influence on their illness. This is all consistent with previous research, which has suggested that patients with the classic phenotype of bipolar – the textbook case – tend to respond better to lithium. Specifically, those who achieve full remission between episodes, and have clearly separated episodes of mania and depression rather than a lot of mixed states, tend to do better on lithium. Where this new study differs is that it found evidence that comorbid panic disorder predicted a good response to lithium. In older studies, multiple comorbidities and anxiety disorder predicted a poorer course. 

So who did worse on lithium in this study? Those were the patients with comorbid OCD, alcohol or substance abuse, and psychosis. 


This next study suggests that it’s hard to tell when people are angry at you when you’re taking a benzo There have been a handful of investigations into the effects of benzodiazepines on social cognition – aka the ability to read people’s faces. The idea makes sense. Benzodiazepines act on structures that are involved in reading faces, like the limbic system, specifically the thalamus and hypothalamus. 


This paper pulled 8 of them into a metaanalysis. They concluded that benzos do impair social cognition, but only in one area: Anger. People who took benzodiazepines were less accurate at identifying facial expressions of anger compared with those receiving placebo. They did not affect the ability to read sadness, disgust, fear, surprise, or happiness, although there was a trend toward a deficit in those negative emotions (sadness, disgust, fear). Anger is one of the most dangerous emotions, and it’s possible that by turning down the anxious reflex in the nervous system the benzos are dampening people’s reactions to real dangers. 


The deficit clocked in at a medium effect size, which means it’s big enough to notice in clinical practice, if you’re looking for it. And maybe that’s the point of this study. Few of us are looking for this, and yet it could have profound effects on our patient’s relationships. Imagine if you were slow to pick up on the fact that your spouse was angry at you. It might be like having thick skin, or it could be that your spouse gets increasingly annoyed at your non-responsiveness until it boils over into a fight. 

The studies in this review used low dosages of benzos and short-term administration. There is only one study that assessed the long term effects of benzodiazepineuse on emotion recognition, and it found that people who took benzos over the long term had difficulty reading negative emotions across the board: fear, sad-ness, and anger 

SSRIs also affect the ability to read faces, but in the studies I’m aware of – most of which were done in normal subjects – they helped people read faces more accurately. The results were palpable –in one of those studies they had people play a cooperative game in groups. The groups that took SSRIs could read each other better and were more cooperative, resulting in more wins.


Leave A Comment