Training at the Maudsley was a rewarding experience. Along the way, had the privilege of working with talented clinicians (not just doctors, but pharmacists, occupational therapists, nurses and psychologists). Most of all, I learned from people experiencing severe and enduring mental illnesses, as well as those going through acute crises in their lives.

Here are some tips, that I often learned the hard way. Let me know in the comments if there’s anything you’d add or change.

1. Pay attention to suicide risk in older people with depression

2. Admitting someone to hospital doesn’t eliminate their risk of self-harm and suicide

3. If you are wondering whether to call your senior, just call them

4. Always ask about psychotic symptoms in depressed patients

5. Don’t copy and paste parts of someone else’s clinical notes

6. Always write an overall impression - it might be the only part of your assessment that is read

7. Patients with borderline personality disorder in crisis are not at their best - seeing these patients in psychotherapy shows a different side

8. Don’t put immobility/weakness/‘off legs’ down to a psychiatric cause until physical causes have been excluded

9. Discharge is a time of increased risk, make sure a follow-up plan is in place and the patient’s next contact with the mental health service is arranged

10. Remember a standard working day for you could be one of the most important days of your patient’s life; it could be one of the worst days of their relatives’ lives too

11. Stop benzodiazepines cautiously in patients recovering from mania

12. Always ask about sexual side-effects from medications (patients may not volunteer this information)

13. Share decisions on risk but ultimately take responsibility for your own judgement

14. Don’t take insults personally

15. Treat patients who are acutely psychotic/agitated with respect and compassion, they may remember this during their recovery

16. Get to know everyone on your ward, from the cleaning team, administrators, healthcare assistants and nurses

17. Psychiatry liaison nurses often have a wealth of experience in dealing with patients in crisis - utilise it

18. Don’t be afraid to say you don’t know

19. Beware EPSEs when prescribing potent dopamine antagonists - deal with them promptly and effectively, IM procyclidine can provide immediately relief from acute dystonia

20. Patients who stop clozapine will usually deteriorate quickly and severely

21. Ensure patients on clozapine have regular bowel movements, prescribe laxatives proactively

22. You can document a mental state examination in a patient who refuses to speak to you

23. Don’t collude in delusions - if appropriate, gently challenge these

24. Psychiatric services have little to offer patients whose primary problem antisocial personality disorder

25. Don’t discharge patients in the middle of the night, even from the ED or place of safety

26. New-onset psychotic symptoms in elderly medical patients is rarely psychosis (usually delirium)

27. Hearing voices is common in borderline personality disorder

28. Sexual abuse is more common than you might expect

29. Warn patients about initiation effects of SSRIs and that they will pass

30. Depression is highly treatable

31. For patients with poor antipsychotic compliance, think of depots early, particularly if there is risk involved

32. See patients with their care coordinator or someone who knows them well

33. Keep ward rounds to time

34. Don’t do routine CKs (a psychiatric d-dimer)

35. Don’t do routine MRI for first episode psychosis

36. Instil hope - recovery is the norm

37. Don’t require patients in crisis to guarantee their safety

38. Do a UDS if a patient deteriorates after leave

39. If you are making a difficult decision, document your rationale

40. Get collateral history

41. Don’t stop lithium lightly in stable patients

42. Don’t expect patients will take medication as prescribed

43. You don’t have to wait until a patient in the ED is ‘medically cleared’ to start your assessment

44. Have a low risk threshold for patients who are postpartum, mental states can change quickly, psychopathology may be subtle at first

45. Stop antidepressants in patients who are manic

46. Stop ADHD meds in patients who are psychotic

47. Seeing patients in their home environment can provide valuable information

48. Know the signs of rare psychiatric emergencies, serotonin syndrome, neuroleptic malignant syndrome

49. The times you miss organic causes of mental illness are usually out of hours, conducting a difficult assessment, when you’re not at your best

50. Beware being asked to ‘lay eyes’ on a patient prior to their discharge from ED

1. Bipolar Disorder Incidence During Pregnancy and in the Postpartum Period: A Population-Based Study

   The timeline of postpartum psychosis is striking; it is the first few weeks following birth which are high risk, with many women experiencing onset of symptoms in the days or even hours following parturition. For me, this is a big clue towards a hormonal trigger.

   A record-linkage study led by Professors Veerle Bergink and Trine Munk-Olsen examined new bipolar diagnoses in relation to childbirth. They found a spike in new cases the first month following birth, which drops back below the background population risk by month two. This peak was found in inpatients but not outpatients (I guess because postpartum psychosis is a condition that generally requires hospitalisation).

1. Escitalopram and brain reactivity to aggressive stimuli in premenstrual dysphoric disorder

   An elegant neuroimaging study from a team including Manon Dubol and Erika Comasco examined a curious finding in PMDD - why do SSRIs act so much faster (within hours to days) compared to their usual action as antidepressants (days to weeks). Their work implicates serotonergic functioning in corticolimbic circuits, suppressing the key symptom of irritability/anger.

2. Hormone blindness in precision psychiatry

   It amazes me that people can talk endlessly about precision psychiatry without mentioning hormones - surely some of the most tractable biological processes relevant to mental health. This BJPsych editorial makes the case for taking reproductive transitions and hormone dynamics into account for psychiatric research.

3. Efficacy and target engagement of dopamine agonist pramipexole for anhedonic depression: a randomized placebo-controlled trial

   You may remember me getting excited about pramipexole as a treatment for depression. This trial takes a novel approach of targeting anhedonia, whether this was in the context of major depressive disorder, dysthymia or bipolar disorder. The effect on (a) anhedonic symptoms and (b) light physical activity, shown below, are impressive.

   

4. Temporal course of symptom domain improvement in schizophrenia: individual participant data analysis of six antipsychotic drug trials

   Another clinically important Stefan Leucht meta-analysis showing symptom improvement over time following antipsychotic prescription. Irritability and excitement improve first, then positive symptoms, and finally negative and cognitive symptoms. Improvements plateau from week 15 (~3 months) onward - much longer than I would have expected! Take-home message: don’t give up on a medication prematurely.

5. Brain lithium temporospatial kinetics in bipolar disorder: a repeated-measures 7Li MRI study of dosing regimens in Germany

   Yes, it’s 2026 and we still don’t know how lithium works in bipolar disorder. This study showed lithium uptake into the brain correlates with serum levels. Twice-daily dosing resulted in more stable brain lithium exposure over the day, whereas once-daily and twice-daily regimens showed similar brain and serum measurements at standard 12 hour sampling.

6. Pharmacological Interventions for Weight Reduction in Patients With Schizophrenia Treated With Antipsychotics

   All prescribers of antipsychotics should be mindful of their metabolic side-effects. There is a lot of excitement about GLP-1 analogues to reduce antipsychotic associated weight gain. Where I work, it isn’t possible to prescribe these in routinely (£££) but there’s nothing stopping us using metformin (cheap as chips), which here showed an average reduction of 3.9 kg.

7. Concordance between target trial emulation and randomised controlled trials: systematic review and meta-analysis

   Emulated target trials are in vogue, but they only show moderate concordance when replicating their randomised controlled trials.

   The RCT still reigns supreme!

8. Tracking claim changes from preprint to publication across 72,644 biomedical studies using large language models

   This is a pre-print so obviously we need to wait until the findings are peer-reviewed and published in a journal…or do we? A belief of mine that pre-prints don’t usually change substantially in their finalised published form is confirmed here. It refocusses the question of why we go through such time and effort to publish in prestigious journals.

9. Shared Genetic Architecture Between Endometriosis and Psychiatric Conditions May Explain Comorbidity

   Another pre-print (no need to wait for it to be published) suggests that the comorbidity between endometriosis and mental disorders is explained by genetic overlap.

Legacy

1. Shut up about your disability

   I’m not sure how to feel about this Spectator article by Michael Simmons. On one hand I can see how someone who grew up with a visible, stigmatised disability is frustrated by so many people describing themselves as disabled. On the the other, mental disorders, particularly severe mental illness, can be genuinely disabling.

2. Antidepressants stole my orgasm

   Important first person account of sexual dysfunction following SSRI use. As this side effect gets more attention, I wonder if we’ll see increasing prescriptions of alternative antidepressants like vortioxetine and, dare I say, bupropion.

3. Inside the NHS unit helping mothers with postpartum psychosis

   Good coverage on the importance of mother and baby units.

4. Richard Dadd review – can his mental illness help explain these arresting, bizarre images?

   A new exhibition from a former Bethlem and Broadmoor patient, who was admitted after killing his father in response to psychotic symptoms. Many of his noted works, including those of fairies, were painted as an inpatient.

   

Stacks

1. Michael Halassa - The Dawn of the Muscarinics

   We are incredibly fortunate to witness a new class of medications for schizophrenia (a generation of psychiatrists never had this). Mike gets this more than anyone else.

   For seventy years, psychiatry had exactly one lever for psychosis: block dopamine, and vary the side effects. There are now, demonstrably, at least two levers that work, built independently by different people. That is what a watershed looks like in this field: the moment a mechanism stops depending on whether a single company’s molecule happens to be good, and starts existing as a category clinicians can build on. The muscarinics are here to stay. What they will ultimately be able to do for patients is being written.

2. Dr Oli Delgaram-Nejad - When Preston Begins to Feel Again

   Excellent personal account of stopping antipsychotics. A lot of subtlety and nuance which you can’t get from clinical research, demonstrating the value of a first person view. Also a reminder of the cognitive side-effects of dopamine-blocking medication. The author has also published personal account of antipsychotic medication in Schizophrenia Bulletin.

3. Endopsychoscope - Paul Janssen’s radical stance on depression and its treatment

   Turns out Paul Janssen, yes of Janssen Pharmaceuticals, yes who discovered haloperidol, didn’t believe in antidepressants.

4. Erik Svenssons Substack - The coming demise of the hard problem of consciousness and the death of panpsychism? via Samei Huda

   Will the hard problem of consciousness go the same way as vitalism?

5. Greg Katz, MD - What Longevity Doesn’t Understand About Medicine

   Many of these points are applicable to the wider biotech field.

6. UK Research and Innovation - “Don’t spread the jam”: Lord Jim O’Neill on where the UK should back R&D

   Counterpoint to the drive for redistribution of research funding across the country.

7. Anamnesis - Historical timeline of developments in psychiatry

   Very nice timeline which puts everything into perspective.

Tweets

1. Go big, big pharma

2. My money is against biological clocks speeding up and slowing down across the lifespan

3. Highly relevant, especially for mania

4. Is there already a term for this?

5. Sad how this story rings true - ultimately academia is a feudal system with certain individuals holding power over others

6. Dark humour

In short, they do not carve nature at its joints.

This has been a longstanding concern. Professor Sir Robin Murray is fond of dropsy as an analogy for our current classification of psychosis. Dropsy was a syndrome in which fluid accumulated in the extremities and abdomen, sometimes accompanied by shortness of breath. We no longer use this term. We now know there are various causes of this syndrome, such as liver failure, congestive cardiac failure, or renal disease like glomerulonephritis. The term dropsy has become obsolete.

Robin suggests the syndrome of psychosis will eventually be replaced by more causative categories. While rare secondary causes of psychosis can occasionally be found, these do not apply to the majority of patients.

One attempt at better classification, based on biomarkers, is the Bipolar-Schizophrenia Network for Intermediate Phenotypes (previously covered by Awais Aftab and Michael Halassa). This initiative aimed to use clinically relevant investigations, like EEG, MRI and cognitive tests, to produce more meaningful categories across the bipolar-schizophrenia psychosis spectrum.

The BNSIP project identified three ‘Biotypes’ :

• Biotype 1: cognitive impairment, ↓ background EEG activity, ↓ sound-evoked EEG activity, ↓ eye-movement responses

• Biotype 2: cognitive impairment, ↑ background EEG activity, ↓ inhibitory control

• Biotype 3: intact cognition, EEG markers overlapping with healthy controls

I am hopeful these Biotypes will produce insights into the neurobiology of psychosis, though they are still a distance away from changing clinical practice.

Truth be told, we are already under-utilising our current diagnostic classification when it comes to these disorders.

This is due to the expansion of a diagnosis, not found in official classification guidelines, first episode psychosis.

Diagnosis: first episode psychosis

Consider this common scenario. A young woman (let’s call her Patient A) presents with acute manic psychosis. She has had a few weeks poor sleep, culminating in not sleeping at all, and going out at all hours. Her speech is fast and uninterruptible. She becomes fixated on religious themes, carrying a Bible and hearing the voice of God, telling her she is the chosen one. She is admitted to hospital due to dangerous and chaotic behaviour, which has put her safety at risk. In hospital, she is treated with olanzapine and makes a good recovery.

In another scenario, a young man (Patient B) has become increasingly isolated over the past year. He drops out of university. He spends more time in his room, refusing to leave the house, as he worries that the secret service are tracking his movements. He suspects these agencies are listening to his thoughts through an advanced AI. He stops eating because he believes trackers are concealed in food. His GP refers him to community mental health services who attempt to engage with him and encourage him to take antipsychotic medication.

In a third example, a thirty-something man (Patient C) presents to the ED with acute paranoia. He believes gang-members, intent on killing him, have been pursuing him across the city. He thinks his flatmates are involved and are communicating behind his back using hand gestures. Of note, he engages in recreational drug use and has been experimenting with methamphetamine (crystal meth) at parties.

In each of these hypothetical (but recognisable) examples, the patient receives the same clinical diagnosis of first episode psychosis. This is despite our current diagnostic systems (imperfect as they are) having better categories.

How did we get here?

Before moving forward, it’s worth taking a step back. How did first episode psychosis become the term for all initial presentations of psychotic illness? After all, you may remember the Kraepelinian dichotomy, dividing the major psychoses into dementia praecox (now called schizophrenia) and manic depressive insanity (now called bipolar disorder). You may also have read about schizoaffective disorder which has features of both.

The conceptualisation of first episode psychosis, can be traced back to a 1991 NIMH workshop - Research on First-Episode Psychosis. A century after Kraepelin, his dichotomy was unified. Much of the motivation was tied to the emerging movement of early intervention in psychosis. This saw first episode psychosis as a heterogenous category, a pluripotent state that could result in a single episode of illness, an episodic course with intermittent recovery, or a deteriorating trajectory where a person never really gets back to their baseline.

Defining patients by their first episode makes sense from a research perspective. Studying patients early in their trajectory, before the confounding effects of chronic mental illness, could provide a cleaner phenotype. Recruiting participants at their first episode might help identify predictors of illness outcome and maybe even treatments to improve outcomes.

First episode psychosis has become the de rigueur diagnosis of early intervention services, the biggest transformation in psychiatric services in recent years (though see my skeptical take). However, in losing the distinction between different psychotic presentations, the baby has been thrown out with the bathwater.

Say what you see

My contention is that our current diagnostic systems, for all their faults, have advantages when used optimally. Some of the qualities we want in a better classifications - biological relevance, consistency over time and raters, and predictive power for both treatment and prognosis - are already present when we diagnose psychotic disorders properly.

From my examples above, Patient A meets criteria for an ICD-11 diagnosis of 6A60.1 Bipolar type I disorder, current episode manic, with psychotic symptoms. ICD allows the diagnosis of bipolar I based on a single manic episode (without the need to wait for further depressive or manic episodes). As I’ve previously discussed, we can say with near certainty that someone experiencing a manic episode will also go on to experience depressive episodes. We know that bipolar disorder, unlike schizophrenia, should be treated with lithium.

Patient B meets ICD 11 criteria for 6A20.0 Schizophrenia, first episode. This requires two symptoms from persistent delusions, persistent hallucinations (most commonly auditory), disorganised thinking, self-disorder (thought alienation, passivity or control) lasting for at least one month. While schizophrenia, like bipolar disorder, can be treated with dopamine antagonists (antipsychotics), if a patient does not respond to at least two adequate trials, their illness would meet criteria for treatment resistance and should be treated with clozapine (licensed solely for treatment resistant schizophrenia).

Compared with bipolar disorder, schizophrenia patients tend to have more severe cognitive impairment and worse functional outcomes. Copy number variants (deletions or duplications of large chunks of the genome) are a feature of schizophrenia but not bipolar disorder - this type of genetic variant is also associated with cognitive impairment.

Patient C meets criteria for 6C46.6 Stimulant-induced psychotic disorder including amphetamines, methamphetamine or methcathinone, what we call drug-induced psychosis. Symptoms should occur shortly after use or withdrawal of the stimulant but should be beyond what we would expect from drug intoxication. It shouldn’t persist after a long period of abstinence (ICD-11 suggests the cutoff of a month). We know that about a quarter of drug-induced psychotic episode will eventually develop into schizophrenia. These patients may well require antipsychotics initially, but a big focus of treatment will be to abstain from drugs, in the hope this will prevent further psychotic episodes.

These three patients could all acquire the same diagnosis of first episode psychosis which then obscures important clinical differences, treatment decisions and prognostic indicators. Grouping these presentations together may be helpful in some research scenarios but for individual patients, we should be using the best diagnostic descriptors available.

Zones of rarity

A regular criticism of psychiatry diagnoses is an absence of dividing lines (zones of rarity) between categories. Diagnoses aren’t distinct, they overlap with fuzzy boundaries. This is true for psychotic illness, with substantial overlap between bipolar disorder and schizophrenia. It manifests clinically in a category of schizoaffective disorder, in which features of both bipolar and schizophrenia exist within a patient. It manifests in substantial overlap in genetic propensity for both illnesses. It manifests in certain medications (antipsychotics) being effective for both conditions.

Diagnoses existing on a spectrum isn’t that unusual in medicine though. Inflammatory bowel disease is traditionally divided into Crohn’s disease and ulcerative colitis. However, both conditions have overlapping pathologies, genetic architecture and treatments. There is even a category, indeterminate colitis, that has features of both diseases.

Closer to the brain, we traditionally think of dementia as being made up of distinct categories; Alzheimer’s disease, vascular dementia, Lewy body disease. However, in reality most patients have a mixture of these pathologies.

It is therefore plausible that the overlap between bipolar and schizophrenia isn’t a deficiency of our diagnoses but a reflection of shared biological mechanisms contributing to these categories.

Diagnostic stickiness

Validity of diagnostic categories also depends on their stability over time and between raters. We don’t want a patient continually changing categories over the course of their illness or two psychiatrists disagreeing as to which diagnosis fits best.

The good news is that both schizophrenia and bipolar type I have among the highest inter-rater agreements for any ICD-11 metal disorder with joint rater agreement verging into ‘near perfect’ (kappa values of 0.87 and 0.84, respectively). This reassures us that these diagnoses are meaningful, that there is a consistency between psychiatrists. Over time, such diagnoses tend to stick too. A meta-analysis reports that 84% of people initially classified as having schizophrenia maintained the diagnosis at three-years follow-up, while a record-linkage study found 74% of those with bipolar type I maintained this at follow-up.

Psychiatrists can be confident that, when diagnosed properly using standardised criteria, categories of schizophrenia and bipolar disorder are reliable and stable.

The art of differential diagnosis

Michael Halassa points out that, during an acute episode, bipolar disorder can be clinically indistinguishable from schizophrenia.

It isn’t easy at first assessment to make a diagnosis, particularly if past psychiatric and medical history is unknown, if substance use is unknown and collateral information not available. However, this uncertainty is nothing to be afraid of - doctors thrive on incomplete information, using this to build up a differential diagnosis.

Differential diagnosis is how doctors think. We draw up a list of possibilities, weighted by likelihood. We keep in mind serious causes not to be missed - could this be encephalitis, a brain tumour, delirium? Then we gather evidence, in an effort to refute or confirm these possibilities - what did the drug screen show? is there a family history of psychiatric illness? is there evidence of cognitive and functional decline? are there neurological symptoms, abrupt onset or other red flags? Eventually we may converge on a single diagnosis or we may keep a few possibilities.

Patient A

Diagnosis: first episode psychosis

Working diagnosis: Bipolar type I disorder, current episode manic, with psychotic symptoms

Differential diagnosis:

• Schizophrenia, first episode

• Schizoaffective disorder, first episode

• Acute and transient psychotic disorder

• Drug induced psychosis (drug screen awaited)

• Autoimmune encephalitis (less likely, no red flags detected)

Patient B

Diagnosis: first episode psychosis

Working diagnosis: Schizophrenia, first episode

Differential diagnosis:

• Single episode depressive disorder, severe, with psychotic symptoms

• Schizoaffective disorder, first episode

Patient C

Diagnosis: first episode psychosis

Working diagnosis: Stimulant-induced psychotic disorder

Differential diagnosis:

• Acute and transient psychotic disorder

• Schizophrenia, first episode

• Schizoaffective disorder, first episode

Unlike other areas of medicine, we don’t typically have a brain scan or blood test to confirm the diagnosis. In some respects, we only ever have working diagnoses. Nonetheless, these provisional categories, based on clinical reasoning have explanatory power. They guide treatment and give an indication of prognosis. Eventually they may even help untangle the underlying biology.

So be as precise as possible. We can do better than first episode psychosis which, let’s be honest, isn’t a diagnosis.

1. Transition time from manic and mixed episodes to depression: a retrospective cohort study

   There is something compelling about the phenotype of bipolar disorder. It was described (as manic depressive insanity) by Emil Kraepelin and much of his characterisation persists today. I’ve previously argued that people who have experienced a manic episode will invariably go on to have a depressive episode but what is the temporal connection between these illness poles?

   In a (*not yet peer-reviewed*) pre-print Dr Chloe Yap and colleagues at Oxford examine this question using the NeuroBlu platform of electronic health records. They found that 1/4 of manic episodes were followed by a depressive episode within a month. Lithium, other mood stabilisers, and sedating anti-histamines were associated with delayed development of depression. While antipsychotics, antidepressants and benzodiazepines were not.

   The study lends support to Athanasios Koukopoulos’s view of bipolar disorder whereby depressive states are secondary to mania.

2. The Lithium Landscape Has Dramatically Changed

   In this American Journal of Psychiatry commentary, Dr Robert Post argues that lithium’s side-effects are manageable, that it is neuroprotective and is disease-modifying for bipolar disorder. He argues it is generally underutilised and should be recommended in first episode mania - I can’t disagree.

3. Time to lithium: A case register study of lithium initiation in bipolar disorder

   A stark demonstration of this underutilisation comes from a study led by Dr Sameer Jauhar, examining the length of time taken to prescribe lithium at South London and Maudsley NHS Trust. Median time between first assessment and bipolar diagnosis was 220 days, while the median time until lithium prescription was 659 days. There was a median 2.5 mood episodes and two antipsychotics prior to lithium.

   We can do better.

4. Acceptability and accuracy of point-of-care monitoring of lithium levels

   One potential solution is lithium point-of-care blood testing (finger-prick sampling, similar to monitoring blood sugar in diabetes). Usual initiation of lithium requires titration and checking of plasma levels until these are within the therapeutic range. Point-of-care testing is more acceptable to patients than venous blood sampling, this study shows the new method was closely correlated with conventional venous measurements.

5. Clozapine After 1 Failed Antipsychotic Drug Trial in First-Episode Psychosis A Randomized Clinical Trial

   Another medication we probably under-use in psychiatry is clozapine. In this RCT it was shown to be superior to other antipsychotics following a single non-response to another medication (current guidance is that patients should be offered clozapine after non-response to two other antipsychotics).

   I’m particularly impressed, and surprised, that clozapine had double the response rate of olanzapine.

6. Cutting through stigma: psychiatry and neurosurgery working together

   Is functional neurosurgery underutilised in psychiatry? A British Journal of Psychiatry editorial argues it is, for refractory forms of OCD and depression. It seems that, like ECT, the history of psychosurgery (frontal lobotomies) has created stigma which inhibits its use as a therapeutic option.

7. Psychiatric and behavioural sequelae following encephalitis: a systematic review and meta-analysis

   A stellar team of authors bookended by Drs Cameron Watson and Tom Pollak provide evidence of psychiatric complications following encephalitis. With so much focus in recent years on autoimmune psychosis, I was surprised that infective encephalitis was associated with more psychiatric symptoms. It also surprised me that, while psychiatric sequelae was common (>25%), it occurred at roughly the same frequency as in other neurological disorders.

8. Diminishing Returns of Novel Autoantibody Discovery in Encephalitis

   Prominent autoimmune neurologists, including Professor Sarosh Irani argue that discovering novel autoantibodies has less benefit than understanding the pathophysiology behind autoimmune encephalitis. They also lament the misdiagnosis of autoimmune encephalitis based on serum autoantibodies, of questionable clinical significance.

9. Ethical issues in the clinical use of diagnostic technologies for rare causes of psychosis

   With autoimmune psychosis in mind, I’d say the biggest ethical issue is the possibility of false positives and unnecessary immunosuppressives.

10. Variability and Magnitude of Choline Levels Across the Psychosis Spectrum: A Meta-analysis

    There is renewed interest in the cholinergic system following the success of xanomeline (Cobenfy) for schizophrenia. The meta-analysis showed elevated choline in each patient group but interestingly the effect size increases from clinical high risk for psychosis to frank psychosis to treatment resistant schizophrenia.

11. Neurometabolites and Antipsychotic Response in Psychosis - A Mega-Analysis

    Like London buses, you wait for one choline meta-analysis and two come along at once. Reassuringly this one also finds elevated choline in antipsychotic non-responders, as well as increased glutamate.

12. Risk of perinatal psychiatric disorder among women with a history of premenstrual disorder: a nationwide register-based study from Sweden

    Women with a history of PMDD had increased risk of perinatal mental disorders, with the strongest association being bipolar disorder, followed by perinatal depression.

Legacy

1. ‘Drug of last resort’ gets new guidance after Times reporting

   Following the death of William ‘Wim’ Northcott in 2021, the Royal College of Psychiatrists has developed a new protocol for clozapine monitoring. Wim had been prescribed the medication for a decade but collapsed and died due to myocarditis, caused by toxic levels of clozapine.

   I’m wary of not prescribing clozapine when appropriate but if this initiative improves safety of the medication while not decreasing its use, it will be a net positive.

2. I had postnatal depression — yes, dads can get it too

   Obviously depressive episodes can be triggered by stressful life events, including a new baby, but I don’t think its helpful to categorise this under the umbrella of postnatal depression. In women, we have a pretty good understanding that postpartum depression is caused, at least in part, by hormonal changes. Allopregnanolone is a hormonal treatment specifically developed for this condition (in women).

3. Young women now have ‘close to zero’ risk of cervical cancer death after HPV jab.

   Love to see it!

Stacks

1. Psychiatry at the Margins - Some Predictions About the Future of “Withdrawal Studies”

   When it comes to controversial topics in psychiatry, Awais usually finds the right balance. Here is a good example where he acknowledges the potential harms of SSRI withdrawal, without readily accepting overly simplistic and alarmist takes on this issue. His predictions are similar to my own views (I’ve also written of the need for balance in this debate) and I hope research is done to prove him right or wrong.

2. Cognitive Wonderland - Blinded by Blindsight

   You may have heard simplistic versions of blindsight discussed on podcasts, particularly in philosophical discussions about consciousness. It is presented as someone who says they are blind (they do not consciously have sight) but their visual system is unconsciously working (for example they are able to navigate past objects even though they don’t ’see these consciously’).

   I’ve always felt a bit skeptical towards this account and in a great post, Tommy Blanchard *a neuroscientist* digs through the original research. As with many scientific stories, it’s a bit more complicated/nuanced than is commonly presented.

3. Asterisk Magazine - The case against randomized controlled trials

   This isn’t really a case against randomised controlled trials but more the case against RCTs being the best form of evidence for every question. I wasn’t totally convinced that emulated target trials are as unbiased as the author suggests. What would convince me? ETTs being shown to predict the results of RCT, before the trials are run.

4. The Clinical Trials Abundance blog - Why were Covid vaccine trials so fast?

   Maybe instead we should do more trials cheaper, better and faster. This is the argument of clinical trials abundance and here Saloni Dattani shows lessons can be learned from the covid vaccine trials.

5. The Munro Report - The truth about the shingles vaccine and dementia

   The shingles vaccines protecting against dementia hypothesis is being tested, to destruction. Here Dr Alasdair Munro presents the evidence against, mainly based on a wonderful talk by Professor George Davey Smith, who failed to replicate the protective effect in a much larger sample size.

   Undoubtedly, the vaccine is good at preventing shingles, which is reason to get it alone.

Pods

1. Triggernometry: The Truth About Depression - Dr Joanna Moncrieff

   The straw man put forward by Professor Moncrieff is that depression isn’t caused by a chemical imbalance. The hosts, Francis and Konstantin (perhaps unintentionally) find that her views extend much further though:

   Francis: When people talk about mental health, they conflate it with mental illness. So we talk about depression and schizophrenia as if they’re under the same umbrella when as anybody knows and I have people in my family have suffered from schizophrenia. It is a very very different condition to depression.

   Prof Moncrieff: I think it’s a very different condition but there’s not very good evidence for any biological basis for schizophrenia either.

   Francis: Really?!

   Tweets

1. Like humans, AI does not choose random numbers randomly

2. How to give a bad talk

3. End organisation-wide mandatory training

4. Is this a problem that a younger version of me was once dreaming about?

5. If the effect size feels too big to be true…

‘Rebound’ on abruptly stopping medication exists across medicine. Stoping anti-hypertensives will shoot your blood pressure up, above its baseline level. Stopping anti-epileptic medication rapidly lowers the seizure threshold. Stopping beta-blockers results in rebound tachycardia .

Rebound is described with certain psychiatric medications too. Rebound insomnia takes places when sleep meds are stopped. Rebound anxiety and dysphoric symptoms are a component of SSRI withdrawal.

The pharmacology of rebound is intuitive. The body adapts to prolonged exposure to a drug, for example by down-regulating receptors which a drug activates, or up-regulating receptors which a drug blocks. When the medication is then withdrawn, the body has overcompensated and doesn’t have time to re-calibrate. Blood pressure goes up, seizure threshold goes down, heart rate goes up.

Whether patients may be put at risk of rebound psychosis after stopping antipsychotic medication is hotly contested. Theoretically, it is something we should expect. The dopamine hypothesis proposes that excessive dopamine activity in the striatum is the final common pathway of psychosis. Antipsychotics exert their action predominately by blocking dopamine receptors. The degree to which these medications occupy the receptor predicts their effectiveness in controlling psychotic symptoms. What happens with long-term antipsychotic treatment? Our simple pharmacological model suggests that dopamine receptors would be up-regulated to compensate for being blocked. When the drug is suddenly stopped, there are so many dopamine receptors that even ‘normal’ levels of dopamine are enough to cause dopamine overactivity. The result? Rebound psychosis. Well, that’s the theory anyway.

Clinically, we do see dysregulation of dopamine when patients receive chronic high-dose dopamine blockade, in the form of the movement disorder tardive dyskinesia. Thankfully, this side-effect is now rare, to the extent that I’ve come across it only a handful of times in my career.

A more speculative concern, voiced by leading figures in psychiatry like Professor Sir Robin Murray, is that treatment with antipsychotics results in up-regulation of dopamine receptors, sometimes called ‘super-sensitivity’. Could our medications actually be making patients worse, with relapses of schizophrenia representing medication withdrawal and subsequent rebound?

Such concerns led to a 2021 paper, A Method for Tapering Antipsychotic Treatment That May Minimize the Risk of Relapse. The author list is interesting. First is Dr Mark Horowitz who, it’s fair to say, is evangelical about the dangers of psychotropic drug withdrawal and has both an NHS and private clinic for tapering antidepressants. The last author is Professor David Taylor, a leading pharmacist, who wrote the Maudsley Deprescribing Guidelines with Mark Horowitz. Robin Murray is a coauthor, as is Dr Sameer Jauhar who, it’s fair to say, is more skeptical of severe withdrawal syndrome.

The paper advocates hyperbolic tapering (I’ve written about this before in the context of antidepressants). This proposes to keeping halving the doses of psychiatric medications over increasingly longer intervals until they reach minuscule, homeopathic amounts. The idea is to give the body time to adapt to each dose change.

The diagram below shows: A. up-regulation of dopamine receptors after chronic antipsychotic use; B. abruptly stopping antipsychotics; C. gradual, hyperbolic reduction in antipsychotic, in this case going down to a tiny (~0.5% of the maximum) dose of haloperidol.

Hyperbolic dose reduction could therefore decrease the rate of relapse following antipsychotic cessation. That’s the theory anyway. So far, there haven’t been randomised trials of hyperbolic tapering or even a clear description of an antipsychotic withdrawal syndrome.

There is one antipsychotic, though, which does have something of a withdrawal syndrome, interestingly, it’s the exceptional one which probably doesn’t have dopamine blockage as its mechanism of action - clozapine. Patients who miss clozapine doses or stop suddenly (often due to blood monitoring regulation), seem to relapse suddenly with a prominent agitated and confusional presentation.

Before going further, let’s think about what we’d expect if rebound psychosis is a consequence of antipsychotic withdrawal. Firstly, we’d expect rapid withdrawal of antipsychotics to result in more relapses than gradual withdrawal (which gives the body more time to regulate receptors in response to lowering doses of the medication). Secondly, we’d expect more relapses around the time medication is stopped and for relapses to gradually even out over time.

Of course, we’d also expect stopping antipsychotic medication to be associated with increased psychotic relapses than continuing the medication but that could be due to the protective effect of antipsychotics against relapse rather than a withdrawal effect.

So let’s test the hypothesis of rebound psychosis.

Stopping, fast and slow

A landmark paper testing the hypothesis was published in Schizophrenia Bulletin two years ago, Does Slow and Steady Win the Race? Rates of Antipsychotic Discontinuation, Antipsychotic Dose, and Risk of Psychotic Relapse. The lead author is Robert McCutcheon, an Associate Professor at the University of Oxford. The coauthors include the familiar names of Sameer Jauhar, Robin Murray and David Taylor.

Rob made use of the Yale Open Data Access project to find antipsychotic discontinuation trials, whereby patients in remission are randomised to either stop or continue medication. Thanks to long-acting injectable antipsychotics (colloquially, ‘depots’) , we have a readily available slow-stopping medication. These type of antipsychotics are typically given at intervals of weeks-months and gradually dissipate from the body. By contrast, oral antipsychotics are typically given once a day. When stopped, they leave the body much faster. Thus we have two readily comparable methods of stopping medication: depots (the tortoise) and oral antipsychotics (the hare). Even better, some depots are dosed monthly while others are three monthly.

Using discontinuation trials of the antipsychotic paliperidone (the active metabolite of risperidone), Rob was able to calculate relapse risks for oral, monthly and three-monthly formulations. You can see the differences in speed of withdrawal in the diagram below (the dotted green line representing discontinuation drops off so quickly it is almost vertical).

Which do you predict would result in rebound psychosis? According to our hypothesis it would surely be stopping oral meds, which leave the body quickly and so give the brain less time to adapt to the change in dopamine blockade.

What they found might surprise you; there was no difference in relapse risk when stopping oral, monthly or three monthly paliperidone. This means that gradually stopping antipsychotics was not associated with reducing risk of relapse, compared with stopping abruptly.

The results are consistent with another pooled analysis of randomised discontinuation trials by Takeuchi and Watabe which showed no difference in psychotic symptoms between gradual and abrupt stopping. It has also been supported by a subsequent prospective study from the team of Professor Iris Sommer which did not find that the speed of stopping antipsychotics was associated with relapse.

A critique of the study by, you guessed it, Mark Horowitz, and a response from the original authors were published in last month’s issue of Schizophrenia Bulletin. I can’t pretend to be unbiased, as most of the authors including Rob are colleagues. Nevertheless, I think they provide a strong and comprehensive rebuttal.

In his article, titled Comparing Apples to Oranges Obscures Tortoises Beating Hares, Mark notes that discontinuation of oral medication wasn’t compared directly with discontinuation of long-acting medication. He calculates that comparing relapses in the arm who discontinued medication in oral trials versus the arm who discontinued in depot trials, shows the risk of relapse was five times higher in those stopping oral meds.

Is it valid to compare one arm of trial with one arm of another trial? No, of course it isn’t. There are various factors which might affect relapse rates in a trial: how the study was conducted, how patients were selected and how outcomes were measured. As McCutcheon et al. point out in their response, that’s the reason we have comparator arms.

They illustrate this point below, plotting relapse rates in a trial of brexpiprazole (trial 1) versus aripiprazole (trial 2). If we only took the active arm of each trial, it would seem that there are much less relapses in those randomised to brexiprazole than aripiprazole (the dotted blue line versus the solid blue line). It is only when we compare differences between active treatment and placebo (dotted blue line versus dotted red line, and solid blue line versus solid red line) that we can see that the risk of relapse is the same across both interventions.

Similarly you can’t just compare relapses in people stopping oral medication versus those stopping depots, instead you need to compare the difference in stopping oral medications versus continuing oral medication against the difference in stopping depots versus continuing depots.

Another consideration is that someone stopping antipsychotics suddenly, say going from 4mg of risperidone to none over two weeks will have less antipsychotic exposure than someone who stops over two months. In the second case, using a linear dose reduction, the patient would still be exposed to an effective dose of risperidone for a month more than someone who stops abruptly.

It turns out that having an effective dose of medication, enough to block ~60% of dopamine receptor is what is needed to prevent relapse, not a slow reduction in the rate of receptors that are blocked. Absolute receptor occupancy, rather than the rate of reducing receptor occupancy is the key.

Receptor occupancy is all you need

The defining feature of antipsychotics is blocking the dopamine receptor. The degree of this blockade is associated with clinical response (>60%) and extra-pyramidal side-effects (>80%). For relapse prevention, it is also receptor occupancy, rather than rate of dose changes that prevents relapse.

A Stefan Leucht meta-analysis showed that, for relapse prevention, maintenance doses above the equivalent of 2.5mg of risperidone were needed. Above 5mg of risperidone, there was not much more benefit but increased side-effects, suggesting the sweet spot of 60%-80% receptor occupancy for control of symptoms while minimising side-effects.

McCutcheon et al. found that absolute receptor occupancy (based on antipsychotic dose) predicted relapse risk in discontinuation trials, while speed of stopping did not. This is more consistent with relapse being related to dopamine receptor occupancy falling below a therapeutic level, rather than dopamine receptor up-regulation in response to antipsychotic treatment.

My clinical experience fits this. I’ve met many patients whose psychotic symptoms worsen in the days before they are due their next depot. It’s not uncommon for an injection to be brought sooner for this reason and for subsequent doses to be increased. This scenario doesn’t fit with relapse due to rebound psychosis from sudden antipsychotic withdrawal. By definition, depots are slowly absorbed and then metabolised by the body. Instead it does fit with the slow decrease of medication eventually falling below a therapeutic threshold of receptor occupancy, which then results in the re-emergence of psychosis.

The importance of receptor occupancy rather than rate of decline means that fiddling with very low antipsychotic doses is probably irrelevant for reducing relapse risk (0.5mg of risperidone is ~30% receptor occupancy, much lower than minimum needed to prevent relapses).

The real cause of rapid relapse

In randomised trials of antipsychotic withdrawal, there is a proportion of participants (roughly a quarter) who experience a rapid relapse, with almost immediate worsening symptoms and mean time to relapse of just 10 days after stopping oral medications. Are these people experiencing rebound psychosis? Actually, no.

Although there is a group who experience this rapid relapse when randomised to stopping medication, a similar proportion experience rapid relapse after being randomised to medication continuation.

Huh?

The answer here doesn’t like in pharmacology but in the practicalities of conducting clinical trials. In order to meet inclusion criteria for discontinuation studies, participants need to be stable in mental state, defined as scoring below a threshold on a measure of psychotic symptoms.

Inevitably, researchers recruiting patients to meet their target sample size will be incentivised to score patients slightly below that threshold (perhaps unconsciously) in order to include them in the study. Following randomisation, symptom ratings do not affect the participant taking part in the study and researchers are free to score symptoms without bias.

Dr Matthew Taylor, a psychiatrist in Oxford1 looked at risk of relapse over time in randomised trials of discontinuation. Consistent with what we’ve seen so far, the risk of relapse (discontinuation vs maintenance) did not vary over time - in other words, there was not a greater risk of relapse in the first month following discontinuation compared with later time points.

The real reason you should stop antipsychotics slowly

The evidence against rebound psychosis is compelling. Let’s not overstate the case though. Most of the studies I’ve discussed have used antipsychotics which are relatively selective for dopamine blocking (like risperidone/paliperidone). As I’ve said, clinically there is rebound after stopping clozapine, to the extent that it would be difficult to ethically justify a discontinuation trial. Perhaps other antipsychotics which are pharmacologically more ‘clozapine-like’ (I’m thinking olanzapine) could also produce rebound psychosis?

Likewise, a lack of evidence for rebound psychosis doesn’t discount long-term dysfunction of the dopamine system due to chronic antipsychotic use, such as tardive dyskinesia. Long-term antipsychotic medication should be prescribed judiciously, at the lowest possible dose. Which brings me on to my last point.

Reducing doses slowly increases the length of time on a therapeutic dose of medication. Perhaps more importantly, by reducing the dose step-by-step it allows the clinician and patient to collaboratively find the optimum dose which effectively controls symptoms with minimum side-effects. With careful monitoring, higher doses can be reinstated in the case of early signs of deterioration, nipping a potential relapse in the bud.

Antipsychotics should still be stopped gradually.

Even if rebound psychosis isn’t a thing.

1. Genetic architecture of postpartum psychosis: from common to rare genetic variation

   Postpartum psychosis is considered a ‘psychiatric emergency’ due to the sudden onset following childbirth, fluctuation in symptoms, and risk to the mother and child. It is widely thought to be heritable but its rarity (1-2 per thousand births) makes statistical power tricky.

   In this Swedish record linkage study, the familial-concordance-based-heritability was 55% (roughly in the same ballpark as bipolar disorder and schizophrenia).

   Using genetic data from All of Us, the number of estimated risk genes (81) was smaller i.e. less polygenic than other psychiatric disorders - which bodes well for future studies that don’t have massive sample sizes.

   Intriguingly, the biggest hit was for a gene HMGCR which is involved in cholesterol synthesis. There was substantial genetic overlap with bipolar and schizophrenia (expected) as well as autoimmune conditions (more interesting).

   Overall, this left me optimistic about the potential for genetics to reveal the causes of postpartum psychosis and indeed other hormone-mediated psychiatric conditions.

2. Poor mood after oral contraceptive use is associated with increased vulnerability to peripartum depression, premenstrual dysphoric disorder, and higher genetic risk for depression

   A subgroup of women have a severe negative reaction to oral contraceptives. It’s not uncommon in my clinical work to hear patients say it made them suicidal. This study links ‘hormone-sensitivity’ to the pill with depression at other times of hormonal flux, postpartum depression and PMDD. More and more it is becoming clear that certain individuals are susceptible to these hormone fluctuations - next, we need to find out what causes this vulnerability.

3. Bidirectional Association Between Premenstrual Disorders and Psychiatric Disorders

   Another Swedish record linkage study, this one confirming that premenstrual disorders are highly co-morbid with other mental disorders, particularly depression and anxiety. Almost half of women with a premenstrual disorder also had another mental health diagnosis, compared to less than a third of unaffected controls. It highlights that conditions like PMDD need input from psychiatry, we can’t leave their management solely to the gynaecologists!

4. Gold-standard evidence and best practice guidance for menstrual cycle-informed clinical care: An overview for clinicians

   Nice update on the psycho-biology of the menstrual cycle from three excellent authors, Ellen Lambert, Louise Nolan and Katja Schmalenberger.

5. Polyendocrine metabolic ovarian syndrome, the new name for polycystic ovary syndrome: a multistep global consensus process

   PCOS is being renamed PMOS! I’m generally skeptical about name changes in medicine (remember dopamine dysregulation disorder for schizophrenia). However, if this can stimulate research and encourage us to look at PMOS beyond cystic ovaries, it should be welcomed.

6. Cognition and the menopause transition: cross-sectional evidence from a large community cohort

   This large study (n=14,000) using the REACT-Long Covid Study found perimenopausal and postmenopausal women had more subjective cognitive complaints compared with a pre-menopausal group but there were minimal differences in objective cognitive measures.

7. The prenatal sex steroid theory of autism after 25 years

   Simon Baron-Cohen looks back at his idea that sex steroids in utero can influence the development of autism.

8. Lithium: challenges of being king

   British Journal of Psychiatry editorial extolling lithium and lamenting its under-use. Predictable response from Professor Moncrieff below:

9. Lithium effects on renal functioning: an expert opinion and management algorithm

   A useful update on managing a genuine concern about long-term lithium, renal impairment. I’ve met more than a couple of patients who developed end-stage renal failure secondary to lithium; this paper is important for any prescriber. The big message is to dose judiciously, monitor levels and keep these on the lower side.

   What caught my eye was the recommendation for the potassium-sparing diuretic amiloride for ‘lithium-induced arginine vasopressin resistance’ (formerly known as nephrogenic diabetes insipidus). It blocks lithium entry into the collecting duct principal cells via the epithelial sodium channel. The recommendation is based on limited evidence, but it hints that one of the biggest adverse effects of this medication could be ameliorated.

10. The pharmacological rationale for amiloride nasal spray: potentially a portable and rapid-acting treatment for panic disorder

    You wait for one psychiatric repurpose of amiloride, then two pop along at once. Preventing panic attacks is even more theoretical, but there’s only one way to find out if it works (a randomised controlled trial).

11. Regional Blood Flow Signatures of Opioidergic Modulation of Ketamine in Major Depressive Disorder: A Randomized Crossover Study

    Another excellent paper from the PhD of Luke Jelen, this one looking at how pre-treatment with naltrexone alters regional cerebral blood flow associated with ketamine treatment.

12. Genotype-Guided Antidepressant Prescribing for Patients With Depression via Marios Georgakis

    Various genetic markers are associated with metabolism and response to antidepressants. Prescribing a medication based on someone’s genetic profile is therefor an appealing idea. Unfortunately, it has yet to be translated to clinical practice and had a negative primary outcome in this randomised trial of 1460 participants.

13. Tell me lies, tell me sweet little lies: power calculations in clinical trials

    Power calculations, as I long suspected, are mostly make-believe.

14. Effectiveness and acceptability of different psychotherapies for obsessive–compulsive disorder: network meta-analysis

    No significant differences between psychotherapy modalities for OCD. Another Dodo bird verdict.

Legacy

1. Doctors discouraged from sectioning black patients

   Prof Sir Robin Murray, one of the world’s leading researchers in psychosis at King’s College London, said there is “political pressure” to reduce the number of black people being sectioned, but added the reality is that physical and mental health issues affect racial groups differently.

   “It’s akin to saying that it’s very unfortunate that so many black patients are having treatment for prostate cancer, so we should decrease the number having operations,” he told The Telegraph. “Black people have an increased risk of sickle cell anaemia and prostate cancer: we’re not saying urologists are racist.”

   There has been a lot of coverage of the enquiry into the Nottingham attacks, including the suggestion that psychiatrists are being encouraged not to detain Black patients under the Mental Health Act (you can see my take here).

2. Vegan student’s vitamin deficiency ‘fuelled delusions’ before death

   Tragic case of a young woman developing psychiatric symptoms linked to vitamin B12 deficiency, secondary to a vegan diet.

3. The Times goes with, How an arthritis drug could treat severe depression. The Telegraph report that it could help beat depression. The Independent runs the headline, Depression treatment breakthrough as arthritis drug could ‘help millions’

   The researchers themselves:

What did the study, a small (n=30) randomised trial of tocilizumab for depression associated with systemic inflammation actually show?

‘no results reached statistical significance’

See for yourself:

Over-hyping such results does everyone a disservice. It creates public mistrust in science, gives patients false hope and loses credibility for a field.

Stacks

1. Welcome to my body

   I’m a big fan of Lucy Foulkes, a psychologist at the University of Oxford, and have previously covered her work on the dangers of ‘mental health awareness’. Her Substack debut does not disappoint. Foulkes discusses her diagnostic odyssey, with remarkable vulnerability, ending in hypermobile Ehlers Danlos Syndrome. Required reading for clinicians.

2. Psychiatry at the Margins

   From the new to the old. Everyone knows I’m a long-time Awais Aftab stan but he has really been on a hot streak in the past month:

   Twilight of the Psychopharmacologists. The Szaszian Heart of MAHA Psychiatry. Make Deprescribing Boring. What I Wish People Understood About Mental Health Problems, The Brain Architecture of Schizophrenia-Bipolar Psychosis Spectrum

3. What really happened to Lisa Turtle?

   An engaging discussion of the diagnostic boundaries between psychotic disorder from another Psychiatry-Substack stalwart Mike Halassa.

4. The Schizophrenogenic Mother

   Musings on genetic and environmental causes of schizophrenia from Eric Turkheimer.

5. Drug development funnel: what I learnt building one from scratch

   It’s kind of amazing that the pharmaceutical industry works to produce new drugs at all but there’s definitely ways to make it work better

   

Pods

1. Manifold: Embryo Selection and Frontier Genomics with Dr. Alex Young

   Discussion of genomics from someone at the cutting edge. Alex also speaks about his ongoing cancer treatment; I was struck by the grace with which he is facing this serious illness.

2. Dwarkesh podcast: David Reich — How one small tribe conquered the world 70,000 years ago

   The Reich lab is leading research into ancient DNA analysis and what it says about human evolution. No-one I’d rather be asking questions than Dwarkesh Patel.

3. Evolving psychiatry podcast: Overdiagnosis and AI Therapists | Allen Frances

   I’ve long enjoyed this podcast by Adam Hunt. Here he talks to Allen Frances, a well-known skeptic of psychiatric expansion (as well as an architect of DSM-IV). He slightly lost me on discussing how good chatbots are as therapists - this is not my impression of LLMs.

Tweets

1. All these risk factors are modifiable.

2. Berkson’s paradox spotted in the wild.

3. So much slop out there. We can all tell!

4. For the first time, one of my Tweets went viral and therefore my biggest contribution to the internet is a snarky line about surgeons.

Non, je ne regrette rien

The point was expanded by sitting Conservative MP Neil O’Brien, Violent crime and mental health - lessons are not being learned, which covers changes to the Mental Health Act intended to increase autonomy and reduce restriction. A major focus of Professor Sir Simon Wessely’s report into modernising the Mental Health Act was the need to address racial disparities.

O’Brien makes the case that disparities are not necessarily indicators of racism and that public protection has been deprioritised:

For too long there has been an acceptance (without any debate) that any disparities in the rates of people sectioned must automatically reflect discrimination and be a problem.

This is odd, because we do not make this assumption in any other field of public health. We accept, and deal with, the fact that some communities are more likely to suffer from certain things, from sickle cell to diabetes. And in the case of potentially dangerous mental health cases, we have to think about not just the patient’s own wellbeing, but also the risk of harm to others.

O’Brien argues that the victims of attacks like those in Nottingham, which took the lives of Barnaby Webber, Grace O'Malley-Kumar and Ian Coates, should be prioritised over the rights of patients who are subject to detention and coercion under the Act.

The competing narrative is that the Act is disproportionally used against people of minority ethnicities, a demonstration of institutional racism. An oft-stated fact is that Black people are four times more likely to be Sectioned than White people, and are seven times more likely to be subject to restrictive Community Treatment Orders (which compels treatment following discharge from hospital). This narrative is encapsulated by a 2020 open letter to the Royal College of Psychiatrists, signed by more than 150 of its members:

Psychiatry, psychology, and psychotherapy are deeply embedded in the historical roots through which social systems and structures created by colonisation, slavery, and economic exploitation have become institutionalised and incorporated into our ways of life and perceptions of each other.

We share a history with psychology and psychotherapy of not just ignoring the effects of discrimination, but painting other cultures as psychologically primitive and casting their approaches to understanding distress as backward superstitions. For the Royal College of Psychiatrists to be able to advocate on behalf of those most marginalised in our society, it must first put its house in order and root out all examples of institutional racism and colonial mentality in its training curricula and various practice guidelines.

The perception that there are systemic biases against Black people with mental health problems isn’t just based on statistics. There are a a number of cases of Black men dying after being restrained by police following a mental health crisis. Men like Seni Lewis, Sean Rigg and Kevin Clarke in South London, or Sheku Bayoh in Scotland.

The question of whether British psychiatrists contribute to racism faced by Black mental health patients is topical but not new, it has been asked by previous generations of psychiatrists.

Are British psychiatrists racist?

The disparity of diagnoses like schizophrenia were noted since the 1960s. Several facts point to this having an environmental (or societal) rather than a genetic cause. Firstly, rates of schizophrenia were higher in Caribbean immigrants to the UK, compared with both the UK baseline rate and their country of origin. Secondly, rates appear to be higher in second-generation than first generation immigrants. Thirdly, a number of replicated risk factors for schizophrenia, such as urbanicity, childhood adversity and social disadvantage, might seem to be more common in Black people. Another established risk factor, cannabis use, could be more common in the Black community, though this is a bit more disputed.

Could racism underlie the disparity in diagnosis? It’s certainly a hypothesis. In the 1990s, two young psychiatrists, now prominent professors at UCL, Glynn Lewis and Tony David, along with Caroline Croft-Jeffreys published Are British Psychiatrists Racist? in which they surveyed 140 psychiatrists on a clinical vignette, randomly changing the sex (male/female) and ethnicity (White/Afro-Caribbean) of the patient. While respondents were more likely to diagnose schizophrenia and to recommend antipsychotic medication in scenarios in which the patient was White, they were more likely to rate the risk of violence as high when the patient was Black.

Going beyond hypothetical scenarios, Professor Sir Robin Murray invited Professor Frederick Hickling to evaluate diagnoses of inpatients at the Maudsley Hospital in the study, A Jamaican psychiatrist evaluates diagnoses at a London psychiatric hospital. Prof Hickling diagnosed schizophrenia at roughly the same proportion as his White British counterparts, though the overall agreement as to which patients had the disorder was low.

One of the most instructive aspects of the ‘Jamaican psychiatrist’ study isn’t included in the paper but tells us something about British society of the day. Prof Hickling, recounts that when visiting the Maudsley, he rented a new Rover. The following day, he was stopped and accosted by three police officers who had suspected the car was stolen. Unfortunately, this was just one of a number of racist experiences Hickling had as a medical student and doctor in the UK. I wish things had improved but have heard similar stories from Black colleagues working in London today.

Prof Murray, the doyen of British psychiatry, clearly does not believe elevated rates of schizophrenia in Black people is down to racist psychiatrists. Instead, he highlights the multiple risk factors that may be experienced by Black people, including racism in wider society.

Even 20 years ago, psychiatrists were wary of the unintended consequences of what we now call anti-racism. Dr Lade Smith, who would go on to become the first Black president of the Royal College of Psychiatrists, spoke about an all White tribunal panel discharging one of her Black patients:

‘He was really at risk getting hurt because of the illness that he had, and the tribunal discharged him’, she says. ‘I don’t know what was going on in their minds other than they were too scared of thinking that they might be being racist towards him.’

‘Racism remains a problem for black people,’ she says, ‘and undoubtedly is one of the reasons why there are more black people involved in psychiatric services in the first place, but psychiatry is not where the problem starts.’

Disproportionate rates of mental illness in this population not being be due to racism was robustly argued by Professor Swaran Singh in a BMJ article, Race and mental health: there is more to race than racism. Singh (once treated as a persona-non-grata for his views but now experiencing a renaissance) argues that diseases are not egalitarian, that we see disparate outcomes in cardiovascular disease, diabetes, and indeed mental illness. He goes further, highlighting that false attributions of racism can further fuel mistrust between ethnic minority groups and psychiatric services in the UK.

However, as Post Liberal Pete pointed out on Twitter, diagnoses like schizophrenia are not comparable to diabetes or cardiovascular disease. People with schizophrenia may be detained in hospital without committing a crime, forcibly treated with antipsychotic medication (possibly by intramuscular injection), and compelled to continue medication in the community (or else be brought back to hospital). Diagnoses like schizophrenia are based on clinical assessment. They have fuzzy boundaries. They are, to an extent, subjective. Of course we should be extremely sensitive and wary if certain ethnic groups are more subjected to these powers.

I worry though, that in cases like the Nottingham attacks, polarising discussions of racism versus anti-racism can obscure other, more pertinent considerations.

Individualism

I hear a lot of criticism about individualism in psychiatry. This is usually along the lines of us locating mental illness within the individual rather than in the context of wider societal factors. They say that for Western psychiatry, there is no such thing as society.

While this point is overblown (of course we take someone’s cultural background into context) it does have a kernel of truth. We do treat psychiatric disorders as originating within a person and we treat that person as an individual, rather than as a member of a ethnic, racial, religious, or socioeconomic group.

Can I shock you? I think individualism in this context is good. We should at least strive to treat each person on their own merits. Not regard them at higher risk of violence because of their ethnicity or conversely be less restrictive in an attempt to reduce inequalities between groups.

Having an individualistic approach to patients is an attempt to reduce the impact of someone’s race, religious background or gender on clinical judgement. They might not say so explicitly but I suspect most psychiatrists are individualistic in their assessments.

Liberty (vs restriction)

There is an uncomfortable balance between liberty and restriction at the heart of psychiatry. Restricting a person’s freedom can be necessary to protect the public or to protect the patient from him/her-self. Where we draw this line is a daily question for psychiatrists.

The decision isn’t just down to psychiatrists, though. We need to decide, as a society how much we weigh the protection of people from severe mental illness against restricting the liberty of these patients.

In the England, we are moving towards less restriction and more patient autonomy. Currently, the Mental Health Act Code of Practice states that treatment should be in the form of the least restrictive option. A major focus of NHS England has been in reducing restrictive practices, such as coercion, restraint, and involuntary medication.

Incoming changes to the Mental Health Act will increase the threshold needed to detain someone, going from ought to be so detained in the interests of the patient’s own health / the patient’s own safety / with a view to the protection of other persons to a higher bar of serious harm may be caused to the health or safety of the patient or of another person unless the patient is so detained.

It will reduce the duration of treatment detentions (Section 3) from six months to three months. It will increase the frequency of Mental Health Tribunals which have the power to discharge patients from detention. It will increase the use of advance choice documents and give patients more control over who is consulted if they are mentally unwell.

These changes to mental health legislation receive much less coverage than racism/anti-racism but will have a more tangible impact on the levels of restriction placed on those who are mentally ill and at risk to themselves or others.

Money (beds)

Perhaps you are persuaded by Neil O’Brien and want to prioritise public safety, at the expense of the freedom and autonomy of psychiatric patients. Perhaps you think the threshold to detain someone in a psychiatric hospital should be lower and the duration of their treatment should be longer. I have bad news for you. We can’t do it. We don’t have the beds. And by beds, I mean money.

Deinstitutionalisation, the transfer of long-term inpatient care to community psychiatric services, began in the mid 20th century. The discovery of effective antipsychotic medication was the catalyst but not the sole reason. As asylums closed, funding was diverted to community mental health care. There is a moral case to be made for treating mentally ill people in society, rather than sequestering them in long-term institutions. However, like much of life, a lot of the incentive boils down to money.

To put it bluntly, the NHS has a relatively fixed amount of funding to cover all mental and physical healthcare. Inpatient beds are expensive. They require more overheads than outpatient services. A major motivation of NHS trusts (my opinion) is to break even. To do this, they are highly incentivised to reduce inpatient admissions and shorten the duration of admissions as much as possible (this is often euphemistically called ‘patient flow’).

A national shortage of inpatient beds means that patients get admitted ‘out of area’. They may wait days in the emergency department for their admission. While patients in the community who are relapsing may face extended delays to be admitted. This benefits neither mental health patients, nor public safety.

Two brothers

Finally, I want to reflect on the racism and dehumanisation of Valdo Calocane, or VC as he is referred to in the Nottingham enquiry. This shouldn’t detract from the horrific, monstrous acts committed by him, nor the tragedy of the three entirely innocent victims who lost their lives and others who received physical or psychological injury at his hands.

VC has been called a monster. On Twitter, much of the discourse refers to his race and his immigrant background. The most common photograph, which I’m not going to reproduce, is a police mugshot. Eyes half-closed, a blank, glazed expression. It is easy to see the face of someone, whom you know has committed terrible acts, whom you know has a long history of psychotic illness, as other. If not less than human then certainly dangerous and apart from civic life. Many people with severe mental illness experience this stigma, even if they haven’t become violent like VC. When I see patients in hospital who have had their lives wrecked by illnesses like schizophrenia, their alienation from wider society is evident.

It can be illuminating to meet their close relatives. Usually, they are people who bear a physical resemblance to the patient but without the chronic effects of a severe mental illness. Often, they are successful, productive members of society. It makes you think of the hopes and dreams of the patient, their potential, what they could have achieved if they hadn’t been afflicted by an illness like schizophrenia.

VC wasn’t from a dysfunctional family. His mother was an ICU nurse. He was brought up as a church-going Christian. His brother, Elias, gave evidence to the Nottingham enquiry. Like Valdo, Elias is Black and from an immigrant background. He had the same family upbringing, a similar early environment. However, whatever genetic or developmental quirks led Valdo to develop schizophrenia, were not present in Elias. He was a straight-A student, graduating in Computer Science at Cambridge. He is demonstrably intelligent, articulate and well-presented.

I believe VC’s life could have taken a completely different course if he hadn’t developed schizophrenia. In psychiatry we sometimes get people back onto the path their life was heading, pre-illness. More commonly we are able to control symptoms and allow them to function in society. At least, we should be able to keep them and the public safe, if their symptoms do not respond to treatment.

Why did we not do so, in the case of VC? We need to wait for the enquiry’s findings, though presumably both individual and institutional failings will be identified. I hope the preoccupation about race will not distract from the real change we need in psychiatry.

1. Women and psychosis: a guide to evidence-based, hormone-informed care. My colleague Bodyl Brand has written a Lancet Psychiatry article (I’m a co-author), arguing that our approach to psychosis cannot be sex-neutral. We need to be aware of hormonal transitions like the menstrual cycle and menopause transition, recognise sex differences in drug pharmacology, and utilise exogenous hormones where appropriate.

   This work builds on her PhD thesis, Towards unequal care, which you should check out too!

   

2. Why Women With Psychosis Need Extra Attention—An Introduction to Special Theme Issue Women’s Mental Health. Together with Iris Sommer, Bodyl has also found time to edit a special issue of Schizophrenia Bulletin, focussing on women and psychosis. Highlights include prescribing hormone replacement therapy, impact of age and sex on antipsychotic levels, and clozapine levels around the age of menopause.

3. Secondary Psychoses. Another of my friends, Graham Blackman, together with Vaughan Bell has edited this special issue of Schizophrenia Research on organic causes of psychosis. It includes articles on psychosis in Huntington’s disease and in autoimmune encephalitis.

   Graham has recently been awarded an NIHR postdoctoral fellowship on the assessment and management of first episode psychosis - congratulations!

4. Clozapine rechallenge after neutropenia: a retrospective cohort study in the UK. Important study from the best pharmacy team I know. 3/4 of people who stopped clozapine because of neutropenia were successfully re-challenged. Only 7% had another episode of neutropenia. Much needed data to inform our approach to a patient group with limited therapeutic options!

5. Mismatch negativity in patients with bipolar affective disorder: a systematic review and meta-analysis. Reduced mismatch negativity (an EEG signal in response to surprising sensory input) is one of the most robust biomarkers of schizophrenia. Turns out it is, to a lesser degree, decreased in bipolar disorder too.

6. Dialectical Behavior Therapy vs Schema Therapy for Patients With Borderline Personality Disorder. Both DBT and schema therapy had substantial, long-term symptom reduction, with almost identical improvements between therapies.

   

   Firstly, this shows that BPD is, contrary to popular belief, highly treatable. Secondly, it is a good example of the Dodo bird verdict (via Crémieux), based on the Alice in Wonderland character declaring that everyone is a winner (or all psychotherapies are equally effective). There certainly seems to be something which causes improvement irrespective of the therapy modality. No, don’t say it’s all placebo.

   

7. Ancient DNA reveals pervasive directional selection across West Eurasia. You may have already come across this Nature paper showing evolutionary selection of traits in the past 10,000 years. I’m most interested in the negative selection of markers for both schizophrenia and bipolar disorder. It has long been held as an evolutionary paradox as to why genetic risk for these mental disorders persists in the population, despite reduced fertility in affected individuals - this new research suggests there is no paradox at all, genetic risk for these mental disorders is decreasing!

8. Menopausal hormone therapy and long term mortality: nationwide, register based cohort study. Reassuringly, HRT did not increase mortality and in younger women who had undergone surgical menopause, it actually decreased mortality.

9. General Anesthesia and Discrete Components of Ketamine Neurophysiology. Another cool ketamine in unconscious patients study from The Boris Lab. This group previously showed blinding by general anaesthesia blocked the effect of ketamine over placebo (but remember it was underpowered). The new study shows that general anaesthesia does not block ketamine’s effect on gamma oscillations but does block lower frequency theta oscillations.

10. An international mega-analysis of psychedelic drug effects on brain circuit function. In a field of small sample sizes, questionable research practices, and hype - this is what we want to see, an international consortium pooling functional neuroimaging studies of psychedelics. The total sample size still isn’t massive (n=267), which is indicative of research to date.

    One of the authors, Manoj Doss, has previously criticised the quality of psychedelic research. Over on Twitter, Manoj highlights some ‘strange takes’ on the paper (some of which come from Robin Carhart-Harris who, checks notes, is a fellow coauthor).

11. Mandatory Training Modules in Health Care—Time to Reassess, via Alan Carson. Please no more.

Legacy

1. How autoimmune conditions can unexpectedly drive mental illness. New Scientist article covering autoimmune psychosis (if it exists) featuring Substacker Tom Pollak and Oxford’s Belinda Lennox.

2. Brain drain risks costing Britain ‘science superpower’ status. Clinical academia in the UK is dwindling.

   Data from the Medical School Council showed that only 3.4 per cent of NHS consultants were now clinical academics, down from 4.7 per cent in 2009. The academic workforce was also ageing due to a lack of training posts, meaning that one third were aged over 55 and nearing retirement. “If we continue this decline,” Strain said, “it’s going to be a disaster.”

3. Mental health patients in crisis put at risk by ‘legal grey areas’, investigation finds. In England, when someone is going to be detained under the Mental Health Act (‘sectioned’), the legal framework isn’t in place until an inpatient bed is allocated. These days, a national shortage of psychiatric beds means more and more patients are stuck in limbo, waiting for days, if not weeks, in emergency departments or assessment units. Until a bed is found, they aren’t technically detained so can’t be stopped from leaving using the Mental Health Act. Surely a gap we need to address. Or could we actually, just, y’know, fund more psychiatric beds!

4. Giving us life-saving drugs shouldn’t be such a trial. Happy to see the misaligned incentives of running clinical trials getting coverage in the Times.

5. The Nottingham Inquiry has fallen out the headlines but is still ongoing. All written evidence and recordings of hearings is available to view. I truly hope this tragedy will lead to improvements to mental healthcare in the UK. Clearly, psychiatric services don’t come out well but there are individuals like Professor Nigel Blackwood and Professor Seena Fazel, two titans of forensic psychiatry, whose evidence is worth watching.

Stacks

1. Schizophrenia beyond psychosis. When you read about schizophrenia in academic papers, it is usually introduced along the lines of…this is a chronic psychotic disorder characterised by hallucinations, delusions and disordered thinking. While this is partly true, it fails to capture the essence of the illness. Often, when I interact with a patient with this diagnosis, I am struck most by non-psychotic symptoms. The inappropriateness of emotional responses. A bizarreness of manner. Negative syndrome (apathy, avolition, anhedonia).

   Mike Halassa has written a nice summary of the non-psychosis side of schizophrenia and the underlying biology. Could xanomeline (KarXT/Cobenfy) be a non-dopaminergic solution? Let’s wait and see.

2. Clinical Trials Were Not Always This Complicated. Make clinical trials simple again, please.

3. Curious cases of financial engineering in biotech. Strategies for de-risking drug development, one of the highest risk investments.

4. Tales from the Wards 1.0. Glad to see Anil Makam migrate to Substack. This post contains a couple of clinical gems.

   I too once misdiagnosed a unilateral red leg not responding to antibiotics as cellulitis, though in my case it turned out to be a DVT!

Pods

1. Dwarkesh podcast: Michael Nielsen – How science actually progresses Patel is fast becoming my favourite podcaster. Here he speculates whether AI would come up with the laws of planetary motion or general relatively, given enough data points? It may me think about whether AI could discover what causes schizophrenia given enough MRI scans or biomarker data. My bet is not any time soon.

2. The Works in Progress Podcast: How to speed up clinical trials. Loved this conversation with Ruxandra Teslo on clinical trial abundance. One thing she mentions is the ability in Australia to do small studies of drugs in humans without the full regulatory requirements of a clinical trial. We have that in the UK too! Any study which does not test safety or efficacy (perhaps instead testing a drug’s effect on neuroimaging metrics, biomarkers, or cognition) is classified as ‘non-CTIMP’ allowing it to be run at a fraction of the time and cost of a standard clinical trial.

   Examples include a trial of whether cannabidiol dampens the psychotomimetic effects of THC and another using naltrexone to block the antidepressant effects of ketamine. Each of these studies was run more-or-less single-handedly by two PhD students at KCL, Ed Chesney and Luke Jelen, respectively.

Tweets

1. Ethics reviews at it again

2. Sad that this is believable

3. 🚀🚀🚀

4. You all know which team I’m on

5. Dr David Berger: On the first day of my medical house job in 1991, the registrar said to me: "You will quickly start to think you're irreplaceable. You're not. Remember that. If you get hit by a bus tomorrow, the consultant will look round and say 'Didn't we used to have a house officer? Get me another one!'"

6. The remains of the day

Enjoy yourself, it’s later than you think!

My wife was sick throughout both her pregnancies. It’s not an exaggeration to say she vomited every day and struggled to eat a meal without retching. She tried various anti-emetics - cyclizine, metoclopramide, ondansetron - with little success. Her nausea didn’t improve after the first trimester (as hoped) but resolved immediately and completely after giving birth.

Nausea and vomiting in pregnancy is estimated to affect 70% of women. At the most severe end of this spectrum is hyperemesis gravidarum, in which women struggle to maintain oral intake, become dehydrated and require intravenous fluids. Catherine, Princess of Wales, famously suffered this and required inpatient treatment.

Something with such a clear biological precipitant and onset/offset of symptoms must surely have a hormonal cause. Right?

You’d think so, but at various points in medical history, hyperemesis gravidarum was viewed, at least partially, as psychosomatic. Psychoanalytic theories proposed that it represented an unconscious desire to orally expel the fetus, or to avoid sex, or to receive parental care. Even today, there is suspicion that sickness in pregnancy, may provide secondary gain for women. During my medical training I remember an older gynaecologist stating that women didn’t suffer from hyperemesis during wartime, a theory given credence in a 1956 Lancet paper:

In recent years, research has not only shown that pregnancy sickness has a biological rather than psychological origin, we have now pinpointed the exact hormone (GDF15) which causes symptoms. The story of GDF151 is one of scientific breakthrough and it provides a reminder that similar conditions may have biological triggers, yet to be discovered.

Growth Differentiation Factor 15

The protein GDF15 (then called macrophage inhibitory cytokine 1) was discovered in 1997 by researchers investigating inflammatory processes. They later found that it was highly expressed in placental tissue and elevated in pregnant women.

Much of the subsequent research focussed on GDF15’s role in metabolism, as a potential target for weight-loss. In cancer, elevated GDF15 drives cachexia, i.e. wasting and loss of appetite. The discovery in 2017 of GDF15’s receptor, expressed in the brainstem’s chemoreceptor trigger zone (or vomiting centre), was another piece of the puzzle. The finding was of keen interest to a Cambridge metabolism professor Sir Stephen O’Rahilly,2 who remembered the reports connecting GDF15 with pregnancy two decades earlier.

On the other side of the Atlantic, geneticist Marlena Fejzo, who herself had suffered from hyperemesis gravidarum, was looking for a genetic signal in women with self-reported nausea and vomiting in pregnancy. The biggest ‘hit’ was GDF15, while a signal near the receptor was associated with symptom severity. She went on to replicate the finding using a bigger, multi-ancestry sample, with GDF15 remaining the strongest signal, shown below.

Fejzo and O’Rahilly collaborated on a blockbuster 2023 Nature paper (#4 on my top studies of that year). They established a number of key findings, which at first glance may seem contradictory:

1. GDF15 is elevated in women experiencing nausea and vomiting in pregnancy

2. GDF15 circulating in the mother during pregnancy is mostly from the fetus

3. Women with low pre-pregnancy levels of GDF15 (due to a rare mutation) have markedly increased rates of nausea and vomiting in pregnancy

4. A condition causing high GDF15 levels pre-pregnancy (beta-thalassaemia) has incredibly low levels of pregnancy-related nausea

So GDF15 is associated with hyperemesis during pregnancy but high levels of GDF15 prior to pregnancy is protective against nausea, and low prior levels increase the risk. What gives? The answer is desensitisation.

Women who have naturally low levels of endogenous GDF15 have a dramatic increase in circulating levels, from their developing fetus. However, in women with high baseline GDF15, they are already habituated or desensitised to its effects - when the fetus starts producing it, there’s a less dramatic increase.

Not only does this give a clear biological explanation as to why some women experience severe hyperemesis, it provides a clear pathway for new treatments. In mice, Fejzo and O’Rahilly showed that pre-treatment with gradually increasing levels of GDF15 could prevent the food aversion associated with the sudden increase in GDF15 (as a model for human pregnancy).

For humans, we already have a medication that increases levels of GDF15, the world’s most prescribed diabetes drug, metformin. Indeed, Marlena Fejzo recently published a study showing that mothers who had been prescribed metformin prior to pregnancy were much less likely to develop nausea and vomiting. It won’t surprise you to hear that Fejzo is also conducting a trial of metformin to prevent hyperemesis gravidarum, though results are not expected until the 2030s.

The harms of disbelief

I suspect the discovery of GDF15, a clear cause of pregnancy-related vomiting, will increase sympathy for women with this condition. Relatedly, it will be difficult for any reasonable and informed person to believe this condition is psychologically mediated. However, it will take time to filter out into the medical consciousness, and it is clear that some women with hyperemesis continue to face dismissive attitudes from health professionals.

Abi Stephenson writes about her experience in the Guardian, Don’t call it morning sickness, facing a lack of concern from health professionals but also her own family:

I asked my mother-in-law if she had gone through anything similar. “Oh no, I never got sick like that,” she said, looking at me sceptically. “Actually, there was one time. I made myself a bacon sandwich and then I just couldn’t face it. I always remembered that, because I love bacon sandwiches.”

“I was so sick, too!” said my mother. “I couldn’t stand the smell of perfume counters. I could only eat beige foods for the first trimester.” I asked how many times a day she threw up, how she survived it. “Oh, I never actually threw up,” she said, “but I did feel just so awful.”

As expected, the consequences of poor oral intake constant vomiting in pregnancy can have impact both the mother and fetus. Quite shockingly 5% of women with hyperemesis gravidarum underwent a termination of a wanted pregnancy, while 1/4 had suicidal ideation.

Hamilton Morrin, a colleague and friend who is an academic psychiatrist at King’s College London, looked at psychiatric outcomes in women diagnosed with vomiting in pregnancy. He compared them to matched pregnant women without this diagnosis, in a large health record study (each group was around half a million). Unsurprisingly, he found higher rates of almost every negative mental health outcome, including both common mental disorders like depression, anxiety, PTSD but also outcomes like psychosis.

More surprisingly, women with severe hyperemesis (those with metabolic dysfunction) actually seemed to have less risk of developing depression. A lived-experience coauthor on the paper speculates that those with the most severe hyperemesis might receive more medical care, buffering some of the negative mental health consequences of feeling dismissed by healthcare providers. Requiring hospital admission for rehydration with intravenous fluids, is a pretty good marker to others that you are unwell.

Understanding the unknown

I can’t help but relate this back to conditions where we have not yet understood underlying biological triggers. Premenstrual dysphoric disorder is similarly a hormone-mediated disorder. The onset of symptoms in the 7-10 days prior to menstruation and offset of symptoms following menstruation is sometimes described by patients as akin to a switch.

Like nausea in pregnancy, there is a spectrum of severity, with the vast majority of women experiencing mild premenstrual symptoms and only a small minority (<5%) meeting strict diagnostic criteria for PMDD, which requires prospective tracking of symptoms and impairment to normal functioning. Of course, this spectrum can lead to the more severely affected being dismissed, it’s only morning sickness! it’s only PMS!

Our understanding of PMDD is not as complete as hyperemesis gravidarum but there is convincing evidence that hormonal fluctuations of oestradiol and progesterone trigger symptoms. PMDD might not have its GDF15 yet, but allopregnanolone seems like a plausible candidate. GWAS is still in its infancy for premenstrual symptoms, the first was released as a pre-print in December 2025.

Like hyperemesis gravidarum, PMDD is found across the world, in developed and developing countries. Despite this, its is still debated as to whether PMDD is a culture-bound syndrome and/or psychosomatic.

I feel hopeful we’ll describe the pathophysiology of PMDD in enough detail to make obsolete arguments that it is a Western, culture-bound syndrome or that it is psychosomatic. However, even for conditions which are truly psychosomatic, kindness and empathy goes a long way.

The elephant in the room

It goes without saying that disorders like hyperemesis gravidarum and PMDD, as well as endometriosis, postpartum psychosis and polycystic ovarian syndrome are specific to females. Is it a coincidence that these conditions are some of the least understood in medicine?

Unfortunately, it reflects a systemic bias against female diseases. Male animal models have historically been the preferences, with female models seen as too complex. Female participants are difficult to include in clinical trials. (What if they get pregnant!) Female reproductive factors like the menstrual cycle, pregnancy and the menopause have largely been ignored by neuroscience, until quite recently.

Similarly, it isn’t a coincidence that some of leading researchers of conditions like hyperemesis gravidarum and PMDD have personal lived experience. I’m sure some of these scientists have been struck by the clearly biological triggers and seen potential to understand underlying hormonal mechanisms.

GDF15 is a great example of what collaborative, curious science can achieve - I’m hopeful it will spark a renewed focus on female specific health problems.

1. Automating thematic review of prevention of future deaths reports: concordance study of a child-suicide analysis using large language models. Very practical use of LLMs from my old registrar, now Professor Dan Joyce.

2. Isolating the Genetic Component of Mania in Bipolar Disorder. Feels like genetics is making genuine progress into the nosology of psychiatric illness. Mania explained 81.5% of the variance in bipolar disorder, with major depressive only explaining 18.5%. ‘Pathway analysis highlighted a significant enrichment in the voltage-gated calcium channel activity pathway’, watch this space.

3. Skin Biopsy Diagnosis of Late-Onset Psychosis as Prodromal Dementia With Lewy Bodies: A Case Series. Keep LBD in mind for patients presenting with psychosis in old age.

4. The prevalence of Lewy body pathology across nine international community-based cohorts. Overall, 27.3% of participants had Lewy Body pathology. Nice work from my colleague Lucy Gibson.

5. GH001 vs Placebo in Patients With Treatment-Resistant Depression A Randomized Clinical Trial. This trial of 5-MeO-DMT had a failure of the placebo arm.

6. The unique efficacy of clozapine is attributable to muscarinic receptor agonism. The effectiveness of Cobenfy has led to a re-evaluation of clozapine’s mechanism of action. Written by not one, not two, not three but four Scottish psychiatrists (3/4 are Glaswegian).

7. Why Is Clozapine Uniquely Effective in Treatment-Resistant Schizophrenia? A Review and Proposal. Another muscarinic-take on clozapine featuring Robin Murray and Sameer Jauhar, alongside various eminent (non-Scottish) coauthors. Quite incredible to see Robin still publishing new ideas into his 80s. He gave me my first break (and still most cited paper) 15 years ago.

8. No Evidence for Cognitive Decline in First-Episode Psychosis: A Meta-Analysis of Long-Term Longitudinal Studies. This study is a strike for the neurodevelopmental hypothesis amid the debate on cognitive trajectories in schizophrenia.

9. Dementia in Severe Schizophrenia. In those with chronic, treatment-resistant schizophrenia, almost half had cognitive impairment which would meet the threshold for severe dementia. The cognitive and genetic profile did not fit with typical dementia syndromes such as Alzheimer’s or LBD. The neurodegenerative hypothesis hits back!

10. Regional Blood Flow Signatures of Opioidergic Modulation of Ketamine in Major Depressive Disorder: A Randomized Crossover Study. Another excellent study from Luke Jelen’s PhD - I’m impressed.

Legacy

1. Briton was diagnosed with rabies after psychiatrist raised fears, inquest told. Unfortunately this one doesn’t have a happy ending, but well done to Dr Alexander Burns for making the diagnosis. Add rabies to your list of rare causes of psychosis and be vigilant with patients presenting for the first time in middle-age, with recent foreign travel.

2. They said I was crazy and put me on antipsychotics - then police found a man in my loft’. ‘Phrogging’ is apparently the act of secretly living in another person's home without their knowledge or permission.

3. Woman found out she had terminal brain cancer after suitcase fell on her head. Glioblastoma initially diagnosed as ADHD and emotional dysregulation.

4. I’m dyeing … man rushed to A&E after turning blue. A builder sparked alarm at a hospital after he woke up with bright blue skin, until a swab revealed the colour had simply rubbed off from new bedding.

5. When mentally ill teenagers ask to be put to death. These cases disturb me, I have big questions for the psychiatrists involved.

6. Autism study is my life’s work. The spectrum has lost all meaning. Uta Frith no longer believes in the autism spectrum. For a counter-point, see Simon Baron-Cohen’s Triggernometry episode.

7. NHS will offer non-hormonal hot flush treatment to menopausal women. Good news as fezolinetant gains NICE approval. However…I covered this medication three years ago, why does did it take so long?!

8. Ten doctors have prescribed half of all cannabis drugs. Private cannabis clinics getting some negative media attention - there will be more to come. Also featuring an excellent quote by Prof-Sir.

Sir Robin Murray, a professor of psychiatric research at King’s College, London, said: “Usually if a person has a medical condition, they see a doctor who specialises in a particular area of medicine, for example, respiratory or kidney disease. After diagnosis the doctor prescribes from a range of treatments.

“One doesn’t go to a clinic called after the name of a drug. There are no penicillin clinics or steroid clinics. But here we have clinics set up to prescribe what the patients say they want, ie cannabis, irrespective of whether there is any evidence that cannabis can help the complaint.”

Stacks

1. Michael Halassa: Did John Nash Really Have Schizophrenia? Wonderful post about the genetic dissection of psychosis, featuring research by Cameron Watson. Years ago, Cam kindly mentioned by Substack at an Institute of Psychiatry journal club. Feels like things have now come full circle.

2. Dark Matter Pragmatism: The Route Problem in Psychiatry. Delivering a drug can be as much of a challenge as its effectiveness.

3. Thinking Psychiatry: Losing faith in the SAINT protocol. I’ve seen a lot of hype about the SAINT protocol - here is a more skeptical take.

4. Asimov Press: AI Won’t Automatically Accelerate Clinical Trials. Completely agree that AI will struggle to get round the bottle-neck of clinical trials to develop new medications.

5. Cognitive Wonderland: Why We're Right-Handed But Not Right-Brained. Lateralisation is surprisingly interesting. See also my post about why people with schizophrenia have higher rates of left-handedness.

Tweets

1. The university you go to matters (in the UK at least). Not seen for medicine because five years post graduation, everyone will still be in NHS training programmes.

2. Manual systematic reviews look to be an early casualty of AI.

3. I cry.

4. Studies are still being eviscerated in tweets, science-twitter is not dead yet!

5. How to deliver bad news.

6. Most side-effects on labels aren’t real.

7. RCPsych elections delivering drama. Oh to be a fly on the wall of UCL faculty meetings.

Pods

1. Science Fictions: Electroconvulsive therapy. Most psychiatrists have seen ECT work clinically, sometimes in dramatic ways, but the RCT evidence base is actually pretty thin (correct me if I’m wrong).

2. The Thinking Mind Podcast: The Nazi & The Psychiatrist. Fascinating insight into the complex character of Dr Douglas Kelley.

3. Radical with Amol Rajan: Over-Diagnosis: Are Too Many People Being Given Medical Labels? (Dr Suzanne O’Sullivan). This neurologist has been taking a lot of heat for suggesting that long-covid is psychosomatic. She also covers the expanding spectrum of neurodiversity, with some challenges from Amol who is a top class interviewer.

4. Dr Nassir Ghaemi Podcast: Emil Kraepelin: The most misunderstood psychiatrist in history - and the most relevant. All roads lead back to Kraepelin.

Marc Andreessen, the venture capitalist and tech mogul, has been widely ridiculed for not introspecting:

You don’t have any levels of introspection?

Yes. Zero. As little as possible.

Why?

Move forward. Go.

I’ve just I found people who dwell in the past get stuck in the past.

He goes on to to say that introspection is a modern concept, developed in ‘1920s Vienna’ and that the great men of history didn’t sit around introspecting. Naturally, it has been pointed out that introspection didn’t begin with Freud and that great men of history, such as Marcus Aurelius, actually did meditate on their inner world.

Within his social circle of tech entrepreneurs, I suspect Andreessen isn’t an outlier. It won’t surprise you that Elon is similarly opposed to introspection.

While America is associated with a therapy culture that has been exported across the world, it also pioneers a ruthless pursuit of progress and profit. This form of ultra-capitalism has been inherited by figures like Elon Musk and Marc Andreessen, but it might be best personified in the protagonist of Mad Men, Don Draper.

Born as Dick Whitman, Don had a traumatic life. His mother (a prostitute) died in childbirth. His father was kicked to death by a horse. Dick Whitman almost died in the Korean war, where he took on the identity of his fallen commanding officer (the real Don Draper) and almost never looked back. Don’s philosophy is to suppress trauma and move forward. When his protégé Peggy Olson finds herself pregnant out of wedlock, Don encourages her to give up the baby and move on with her life.

This strategy doesn’t work for Peggy or indeed Don: they are both forced to confront their pasts. No surprise to psychiatrists, who see how traumatic experiences can shape our relationships and interactions with the world.

Unlike Marc Andreessen, I do have levels of introspection above zero. Psychiatrists need to understand how our own psychological makeup affects how we respond to patients. But you might not know that much of our work involves reducing a patient’s level of introspection. Perhaps not to zero but certainly to a level that does not preoccupy their mind.

I’m going out on a limb here to suggest Andreessen is right, for the benefit of our mental health, we should introspect less.

How flat is your mind?

Andreessen cites Nick Chater’s work claiming that the mind is flat. By that he means we don’t have a deep unconscious. Our desires and drives do not come from hidden depth but are improvisations, rationalisations that our mind generates from moment to moment.

Chater hits many familiar aspects of popular cognitive neuroscience like predictive processing, unreliable memory and inattentional blindness to show that the mind mainly reacts to situations off-the-cuff and the inner-world of unconsciousness is another illusion. To Chater, the mind is only able to process one thing at a time, there is no background or offline activity.

While I have sympathy with some of these claims, it clearly isn’t the full story. Offline replay is crucial for the formation of memory, while spaced repetition is a far better learning technique than cramming. This shows that the background work of the brain is crucial, at least for memory consolidation.

More subjectively, if you (dare I say it) introspect, you may find you indeed have a subconscious. For example, if you’ve ever revisited a childhood place or gone back to your university town, you may experience a flood of lost memories. Sights, sounds and smells, previously unconscious, suddenly fill your mind. It is noticeable how sensory inputs (most famously a madeleine) can bring memories back to the surface.

Psychiatrists see how developmental disruptions, particularly early attachments, can determine your relation to the world. Someone who experiences childhood adversity is much more likely to develop a disorder like borderline personality in adulthood: manifesting in relationship difficulties and an unstable self-identity. Part of the treatment involves understanding how past difficulties affect how the patient’s view of themselves and their relations to others: in a sense, making the unconscious, conscious.

Another complication of trauma, PTSD can develop when a person experiences a exceptionally threatening or catastrophic event. Symptoms usually develop after a latency period of weeks to months and include re-living the event in intrusive memories, flashbacks and nightmares. The person may avoid triggering situations that remind them of the trauma. Again the treatment involves reprocessing the traumatic memories.

For both BPD and PTSD, introspection is crucial in order to process past trauma and stop it interfering (unconsciously) with a person’s life. Too much introspection though, can be harmful. In particular, forcing someone to introspect is increasingly seen as counter-therapeutic.

Stop trying to make introspection happen

As part of the wider expansion of ‘mental health’ we are told to introspect. Go to therapy. Talk about our problems. Prioritise our mental health. Unfortunately, this approach has not brought widespread benefits. Rates of serious mental illness are relatively stable, while rates of anxiety appear to be increasing, particularly in the young.

Schools in the UK have implemented lessons in mental health. Children are taught mindfulness and speak about difficult emotions. It feels like teaching kids to introspect and to reflect about their emotions should be helpful. It should help them deal with current difficulties as well as equipping them with skills for life.

Unfortunately, high quality research shows these interventions have few benefits and actually lead to harms for a subgroup of pupils. This kind of universal intervention isn’t enough for those children who are experiencing mental illness, while for kids without mental health problems see the lessons as irrelevant.

Similarly, mandating that people who have experienced a traumatic evident undergo psychological debriefing - getting them to talk through the event and their emotional reaction - isn’t beneficial. Rates of PTSD are actually higher in people who are made to undergo this intervention.

Forcing people to introspect just isn’t helpful and likely results in more mental health problems down the line. This shouldn’t be surprising because a big part of psychological interventions across various mental disorders involves getting patients to be less introspective and more outward-looking.

Anxiety

Anxiety disorders, defined by pathological worry, are some of the most common mental disorders. We have excellent psychological treatments, which train individuals to focus less on their worries, less on themselves.

Social anxiety is a good example, in which someone’s attentional focus is on themselves. I look awkward. Others can see I’m nervous. They’re going to think I’m stupid. This spotlight on the person’s self‑image, and subsequent attempts to suppress their nervousness, paradoxically makes them come across as more awkward and nervous. Therapy trains them to shift attention outward, onto the environment or the task at hand. Only when they are no longer monitoring themselves can they appear relaxed to others.

In generalised anxiety disorder, worry becomes a coping mechanism, a way of dealing with the uncertainty of the world. The treatment involves tolerating this uncertainty and recognising excessive mental preparation as a maladaptive coping mechanism - shifting attention away from catastrophic thinking.

While therapy is the mainstay of treatment for these disorders, it is focused, structured and brief CBT (Cognitive Behavioural Therapy) which enables the patient to introspect less and to redirect their attention outward.

Depression

A major component in the psychological treatment of depression is behavioural activation. When people are depressed, they tend to withdraw: going out less and becoming more isolated. In these situations it is all too easy to ruminate negatively, on past indiscretions and personal failings, which can lead to a sense of hopelessness.

Behavioural activation breaks this cycle by getting the person out of the doom‑loop and back into the world. Scheduling small, manageable tasks creates a positive feedback loop: doing more increases energy, which makes it easier to socialise, exercise, and reconnect with things that matter. Over time, this re‑engagement lifts mood.

Instead of introspection, a big part of treating depression is helping someone rebuild contact with the world, their friends, and their interests. In doing so, a sense of hope for the future is rebuilt.

Psychosis

In more severe mental illness, such as psychotic disorders, people can become so preoccupied with their inner world and their misinterpretation of reality that everyday functioning becomes difficult. When someone is experiencing a delusion (a fixed false belief) it is futile to try to argue them out of it. Confronting the belief tends to escalate distress, while colluding with it (for example, agreeing that government agencies are pursuing them) reinforces the conviction.

Instead, we focus on maintaining trust, reducing distress, and supporting the person while treatment begins to work. Medication helps make the beliefs less compelling and less intrusive, and psychological support helps the person stay grounded and re‑engage with the external world. Over time, as the intensity of the delusion fades, people become less consumed by their inner world and more able to connect with the reality around them.

What is introspection for?

Lots of smart people, that I admire, introspect. I’m thinking less about Plato and Marcus Aurelius, more Erik Hoel and Sam Harris. Hoel’s Substack, one of my favourites, is called The Intrinsic Perspective - the epitome of introspection. Sam Harris is evangelical about mindfulness and the benefits of turning attention inward.

I know many people who have spent years in open‑ended psychotherapy, which, unlike CBT, has no specific goal beyond exploring the inner landscape. These forms of introspection clearly have intrinsic value for those who pursue them.

But many others (myself included) don’t prioritise that value. In a life with limited capacity, maybe 500,000 waking hours, I’ve chosen not to spend much of it on meditation, psychoanalysis, or other inward‑facing practices.

I won’t be booking a silent retreat anytime soon. In my free time I’d rather invest in the external world: seeing friends, exercising, family time, doing things.

For me, the unexamined life is still worth living.

I passed my DPhil viva.1 Adam Rochussen will tell you this doesn’t mean much, but it meant something to me.

This final milestone of my DPhil journey was surprisingly enjoyable. My internal examiner Beata Godlewska had done each of my assessments over the last three years. The external examiner, Ian Jones, is a professor from Cardiff who has pioneered research into postpartum psychosis. Both had read my thesis, gave interesting comments and were engaging on the general topic of how hormones affect mental health.

My submitted thesis was the culmination of research conducted during my DPhil and the last milestone before the viva. As Adam says, very few students fail a doctorate at that late stage, because of earlier checkpoint assessments. In Oxford these are the Transfer of Status around the end of the first year and Confirmation of Status at the end of the second year. Each offers an assessment of progress, a potential off-ramp if a student is failing, and a milestone on the journey of being awarded a DPhil:

This post isn’t about academic milestones, though. Instead it is about life’s milestones that are otherwise hidden. We all go through such milestones in life. We are born. We go to primary school and later secondary school. Some of us get our first job. Have our first pint. Learn to drive. Unlike academic achievements, these milestones are usually absent from public life.

The first milestone en route to a doctorate is the acquisition of funding. This step, particularly for clinicians, is highly competitive. You usually require a governmental organisation to pay for your student fees, your research costs and your salary as healthcare professional over three years. The value of my fellowship awarded by the MRC was £274,429 - not to be sniffed at. You probably don’t know that the process of obtaining this award was far from straight forward for me.

Looking back, the year 2017 (five years before commencing my fellowship) was pivotal. At the start of that year, I was set on applying for a fellowship to investigate brain activity in psychosis using EEG. I got as far as starting the application and wrote a relatively well-cited review article. Various events led to a course-correction to investigating the role of the menstrual cycle in psychosis. The most personally significant was my father being diagnosed with metastatic prostate cancer.

Embarking on a doctoral fellowship is a huge commitment. With a close family member experiencing a life-limiting illness, I realised that life’s too short and a doctorate too long to be less than fully committed. Two more things happened to me that year. The first was treating a young adolescent girl who experienced strikingly brief but severe episodes of psychosis linked to her menstrual cycle, known as menstrual psychosis. The second was seeing a talk by my future head of department at Oxford, Belinda Lennox, about the potential of sub-typing patients in psychiatry based on biological markers.

A complete change in research direction meant much more time was needed to get back to the point where I was ready to apply for funding. I managed to get (at the second attempt) a ‘preparatory fellowship’ which gives protected time to develop a research idea and work up an application.

During this time I was fortunate to attend the National Female Hormone Clinic with Professor Michael Craig, a psychiatrist who previously trained in gynaecology. Meeting patients with PMDD and other hormone-mediated conditions was incredibly rewarding. In another strange coincidence, one of the treatments we recommend for PMDD, leuprolide, was also keeping my father’s prostate cancer at bay. This medication is a gonadotrophin releasing hormone agonist which works in prostate cancer by suppressing testosterone and in PMDD by suppressing ovarian hormone fluctuations. Attending the clinic convinced me that we could dampen hormonal pathways which worsen mental disorders, much like we hack the hormones which drive prostate cancer.

Around that time I experienced another milestone, my first child was born. Thinking of prognosis in cancer and five year survival rates, I didn’t always appreciate the preciousness of time. These years can be the ones in which a patient meets their grandchildren and when a child gets to know their grandparent.

In 2022, after gaining pilot data, publishing relevant studies, submitting a competitive application and defending my proposed research at a panel interview, I was awarded an MRC fellowship to investigate the effect of the menstrual cycle in women with psychosis. That pales in comparison, though, to the most significant personal milestone of that year, my wedding day.

Starting my DPhil led to a number of professional relationships which I consider important milestones too. It was an honour to meet Emeritus Professor Ian Brockington, the world leading expert in menstrual psychosis whose career was based at Birmingham University. Ian was kind enough to host me at his home and provided me with his collection of menstrual psychosis case histories. Analyses of these reports became a crucial chapter in my thesis.

Moving to Oxford was another milestone that brought opportunities, not least the chance to meet many wonderful colleagues in the department of psychiatry. In particular, I was thrilled to discover the department had their very own endocrinologist, Valeria Frighi. Given the intersection between endocrinology and menstrual disorders, she quickly became a trusted source of advice and eventually a supervisor. She brought incisive critique to much of the hormonal psychiatry literature, mine included. This demand for rigour contrasted with her personal warmth and generosity of spirit. Tragically, Valeria died suddenly in a car accident in March 2024, a loss that was felt by me and colleagues across the department.

Of course, the loss was far greater for Valeria’s family and friends. I got a sense of this when attending a beautiful celebration of her life in 2025. From our conversations, I knew how important family was to Valeria but seeing her as a much loved mother and grandmother, brought insight into part of her life that had been hidden from me. I found out how happy she had been to learn of the birth of her latest grandchild, whom she had been travelling to see when she died. It gave new meaning to the gift she had sent my second child who was born half-way through my DPhil.

One of my proudest achievements (children and marriage aside) during my DPhil has been growing this Substack. I have sent a newsletter every month over the past four years and have written more words to you than in my entire thesis. It has improved my writing and has connected me with people at my university and across the world. Hitting 1000 subscribers back in December 2023 was one of my most satisfying milestones.

My father died in February 2025 and his anniversary (another milestone) coincided with the week of my viva.

Reflecting back, it’s clear that a lot of life happened while I was busy making plans.

I hope you appreciate that my DPhil journey was longer and had more bumps than the records show.

Most of my achievements weren’t academic.

And the most important milestones were, until now, hidden.

You sometimes hear that mental disorders aren’t contagious. Well yes, it is true that they are not infectious (i.e. spread by disease-carrying organisms), but in some cases they can be transmitted from person-to-person. The vector isn’t a virus or bacterium, of course. Instead they spread from mind-to-mind through the communication of ideas.

The term meme was coined by Richard Dawkins, though he simply named what has existed since forever. He explained how ideas or other cultural phenomena (melodies, catchphrases, clothing-fashion) spread through populations under evolutionary pressure. Memes are the units of cultural transmission, somewhat equivalent to genes transmitting biological information.

As the information expressed in a meme can be replicated (and altered), it can change over time. As more transmittable memes are more likely to spread from one person to another, memes undergo a form of natural selection, with the most transmittable memes proliferating over time. The concept of memes really entered our cultural consciousness with the advent of the internet and social networks. Have you heard of something going viral online?

Usually, when people say ‘meme’ they’ll be referring to something like the picture above. But other kinds of meme can spread just as virally across the internet. Think conspiracy theories. Cat videos. Viral dances. These are just different types of memes. Can symptoms, even disorders, be spread across populations by memes? I believe they can; symptoms have always exhibited memetic transmission and this spread has been accelerated by the internet.

When describing disorders being transmitted by memes we should define exactly what we mean. For me, memetic transmission cannot just be a person recognising the same symptoms in another (like the parent of a school child who has just been diagnosed with ADHD seeing the same traits in themself). Similarly, the transmission can’t simply be through shared risk factors (for example a peer network developing a psychotic illness after shared heavy cannabis use).

Instead, memetic transmission has to mean the transfer of symptoms from one person to another. The route of this transmission is by memes, and can involve any form of communication including text, photos and videos. We should be able to track memetic spread of disorders, by clusters of people becoming affected or by rates of a mental disorder increasing in response to cultural events.

The archetypal example of memetic transmission, mass psychogenic illness, has existed long before social media, the internet, or even Richard Dawkins.

Mass psychogenic illness

Previously known as mass hysteria, this is a well described condition that spreads rapidly between members of a community, as if driven by an infective cause but with no organic basis. It isn’t included in the official psychiatric diagnostic systems, but has five defining features according to Sir Simon Wessely:

1. Presence of somatic (physical) symptoms

2. Pre-existing social connection between two or more of the affected people

3. Epidemic spread of symptoms (rapid spread through a large group of people)

4. Attribution of symptoms to a threatening external agent of a physical

   (usually chemical, biologic or radiologic) or spiritual nature

5. Symptoms and signs that are not compatible with the

   environmental exposure

Symptoms are typically non-specific - fatigue, dizziness, headaches, nausea, loss of consciousness. They are benign and transient, with complete recovery. Usually there is a trigger, like false claims of poisoning or chemical exposure. Intense media coverage seems to worsen outbreaks (keep this in the back of your mind).

Epidemics of mass psychogenic illness have been described throughout history, going back to the Middle Ages, where demonic possession was the usual culprit. Post-industrialisation, factories were the settings of outbreaks, triggered by false reports of toxins or noxious gas. These aren’t as rare as you might think, in one study, 7% of chemical incidents were found to have a psychogenic origin. Schools have also been a setting, with girls being particularly vulnerable. This is portrayed well in The Falling, starring Florence Pugh and Maisie Williams where attacks of fainting spreads through pupils.

An illustrative example was published in the New England Journal of Medicine. In 1998, a Tennessee schoolteacher noticed the smell of petrol in her classroom and soon experienced a headache, nausea, dizziness and shortness of breath. Similar symptoms developed in several of her pupils, some of whom were taken to hospital by ambulance with the teacher, in full view of the school.

Over the next few days 170 students and staff attended hospital with symptoms associated with the exposure. No evidence of toxin was found in the blood or urine of those affected and no abnormalities could be identified in the local environment. No explanation for the symptoms was identified. A questionnaire one month later found that people were more likely to have been affected if they saw another person become unwell, knew a classmate who was unwell, or noticed an unusual smell.

How do these symptoms, with no organic basis, spread so rapidly through a community? They spread by memes - in this case, the meme of being exposed to a toxic substance. When a toxic or infectious cause has been ruled out, the meme is left as the vector of transmission. While historically, memes had to be transmitted person-to-person and therefore spread within close communities like schools or workplaces, another route of transmission has emerged in the past few years, social media.

Mass social media induced illness

Tourette’s is a syndrome characterised by involuntary movements or sounds, known as tics. It tends to start in childhood and improve over time. While tics spreading through schools is not new, around the time of the Covid pandemic reports began of tics spreading through social media.

Doctors at a German specialist Tourette’s clinic noticed a rising number of patients presenting with tics which, while superficially similar to Tourette’s, were better characterised as ‘functional’ - they conformed to what the patient’s expectation of what Tourette’s looks like (mainly vocal obscenities or obscene gestures).

The ‘index case’ in this outbreak was a German Youtuber Jan Zimmermann,1 one of the biggest social media personalities in the country, with a large following among teenagers. Zimmermann’s videos of vocal obscenities went viral, at the same time the rise in referrals for suspected Tourette’s.

These patients had a number of features in keeping with a psychogenic origin rather than Tourette’s syndrome:

• Symptoms had an abrupt onset

• Common Tourette’s symptoms such as eye-blinking or throat clearing were absent

• Their ‘tics’ resembled those of Zimmermann

• They gave their symptoms a name, just as Zimmermann called his ‘Gisela’

• Patient were released from obligations such as school or chores, due to the ‘tics’

• The ‘tics’ disappeared when the patients were engaging in enjoyable activities

• In some patients, there was a rapid resolution of symptoms after a diagnosis of Tourette’s was excluded

Similar spread of functional tics have been reported through TikTok. This too followed exposure to a single social media personality. Here, the meme of Tourette’s syndrome was transmitted by videos across social media. It highlights that clinicians should be wary of apparent rapid rises in Tourette’s-related symptoms.

Thankfully, outbreaks of mass psychogenic illness either transmitted person-to-person or through social media tend to be transient and benign. However, the most serious psychiatric outcomes, even suicide, can be transmitted memetically.

Suicide contagion

The tragedy of suicide cannot be overstated; both for the person who has died, as well as their grieving friends and family. We sometimes think of suicide as a final definitive act but it can impulsive. Suicidal ideation may be fleeting and those who survive suicide attempts are often filled with regret, going on to live fulfilling lives.

Can ideas of suicide be spread as memes? They can. We already know this. There are clear guidelines on how to report a suicide in the media. You should not describe the method in detail. You shouldn’t use emotive language, like ‘suicide epidemic’ or mention locations of ‘suicide hotspots’. Suicides shouldn’t be glamorised. Particular caution is advised when the suicide involves a young person.

The reason for careful reporting is the possibility of suicide contagion or copycat suicide. It is sometimes called the Werther effect after the novel by Goethe, which was followed by imitation suicides. These can occur after the death of a celebrity or a prominent figure in a local community. It is thought that the suicide might provide a model for others. Particularly if they over identify with the victim or the suicidal act.

Suicides can occur in clusters. These may be in a local community, such as a school, prison, or psychiatric hospital, (point cluster) or more widespread across geographical areas (mass cluster).

When ideas as grave as suicide can be spread memetically, you know it is something to be taken seriously. Curiously though, it is not something that can be seen in all mental health problems. Severe mental illnesses, are not spread as memes.

Severe mental illness is not memetic

Illnesses like schizophrenia tend to be lifelong with a major impairment of functioning. Their causes are multi-factorial and complex; a mix of genetics with environmental risk factors (such as cannabis, social deprivation, and stressful life events).

Unlike the examples we’ve got to so far, these illnesses do not cluster in space and time. They do not spread rapidly through a population, like an epidemic. Rates of these illnesses are not influenced by media reports. There are no media guidelines saying, ‘don’t speak about schizophrenia, don’t glamorise, it might plant the idea in others to develop symptoms’.

The exception that proves the rule is in rare cases where psychotic beliefs are shared between two people, folie à deux. These types of cases tend to represent a power imbalance, with a dominant person influencing the other. In my clinical career, it is not something I have ever come across.

The reason to mention such illnesses, are as a useful negative control. It simply isn’t the case that all mental disorders can be spread through memes.

Shh! but these probably are memetic

Other conditions previously regarded as biological show, at least in part, memetic spread. While some are more contested and less understood, don’t discount the meme as a vector. For this to be the case, we would expect:

• Clustering of cases (either relating to direct contact or mass media)

• Non-specific symptoms which are not in keeping with a pathological syndrome

• Exclusion of a biological trigger (e.g. noxious substance or infective agent)

There additionally needs to be some idea that is being transmitted through the meme. This might be fear of an exposure or identification with an affected person.

Of course symptoms being influenced by psychological or social factors doesn’t diminish the impact on the individual experiencing them. Being, to some extent, memetic doesn’t make them less real. It does mean we should be cautious in how they are discussed over both social and legacy media.

Remember needle-spiking? Reports of thousands of young women being injected with date-rape drugs in British nightclubs over a few months? No toxicology reports produced evidence of poisoning and no charges were ever brought against any perpetrator. While not seeking to downplay the distress for the women involved, this was almost certainly memetic.

Havana syndrome, in which hundreds of diplomats experienced the onset of fatigue, memory/concentration problems, headaches, dizziness, tinnitus, nausea? It wasn’t caused by a sonic weapon, it was memetic.

Just last week the BBC featured the ‘mystery brain disease’ affecting hundreds of Canadians in New Brunswick. Once feared to have been a form of Creutzfeldt-Jakob Disease, sufferers experienced fatigue, memory/concentration problems, muscle spasms, balance difficulties, visual disturbances, weight loss and limb pain. This outbreak was, yes again, likely to have been memetically driven.

Finally, Emeritus Professor of Psychological Medicine at Oxford, Michael Sharpe was vilified for suggesting that alarmist reports about long Covid could be contributing to its spread (George Monbiot: Apparently just by talking about it, I’m super-spreading long Covid).

In slides available online, Professor Sharpe emphasises that long Covid is unlikely to be a unitary condition and notes biological factors such as lung damage and immune dysfunction. He suggests psychosocial factors play a role too. Here is his example of how the media can unintentionally contribute to the (memetic) spread:

Yes, long Covid has a biological trigger. Yes, in a minority there was demonstrable virus-related damage to organs or the complications of life-threatening illness. However, in keeping with other post-viral fatigue syndromes, most sufferers experienced non-specific symptoms after a mild illness. Symptoms of fatigue, memory/concentration problems, shortness of breath, joint and muscle pain, headaches, dizziness, heart palpitations, sleep disturbance, chest pain, gastrointestinal issues. While these symptoms undoubtedly had a severe impact on a proportion of the population and left a smaller proportion completely debilitated, they may have been amplified by warnings in the media that we should all be living in fear.

More and more, it is acknowledged that psychological factors are important for conditions like long Covid.

Whisper it, Professor Sharpe was probably right.

The original tweet from notable eugenicist Diana Fleischman, criticised the idea (from anthropologist Rebecca Sear) that complex polygenic traits can’t be bred out of a population. Things got heated when Sasha Gusev reasserted that highly polygenic conditions cannot be bred out. This provoked a backlash from people like Damien Morris, Richard Hanania, Geoffrey Miller, and Philippe himself, essentially criticising Sasha for attacking a straw-man. Of course no-one really thinks traits like criminality, alcoholism or low intellect can literally be bred out, but policies can be put in place to reduce their prevalence. These policies (eugenics) may be morally repugnant but that’s different than saying they just don’t work.

Sasha is accused of motte-and-bailey tactics. He argues that complex polygenic traits cannot practically be bred out of a population then retreats when the real argument is put forward - that policies can be used to improve the genetic health of a population.

The opposite side, encapsulated by Richard Dawkins in 2020, is that we should separate the morality of eugenics from the question of its practicality and, as animals, humans can be bred to amplify or suppress certain traits.

I enjoy the genetics discourse on Substack but I rarely offer an opinion given a) I’m not a geneticist and b) there are plenty of smart people who understand the field contributing to the discussion. However, in this instance the history of psychiatry offers valuable lessons. People seem to be unaware that we already had a movement to ‘breed out’ mental illness. So let’s revisit psychiatric eugenics.

Schizophrenia, an archetypal eugenic target

Eugenicists target schizophrenia with zeal - Diana Fleischman continually brought it up during a recent debate with Lyman Stone, arguing that sufferers should be incentivised to produce less offspring, for example by monetary compensation for long-acting contraception.

Schizophrenia is indeed substantially heritable. In most cases it has also a chronic, relapsing, lifelong course. It usually presents in adolescence or early adulthood, meaning that is has huge costs for the person affected, their family and wider society, despite having a relatively low prevalence (<1%) in the general population.

Unlike neurodiverse conditions like autism or ADHD, there are few advantages (perhaps none) associated with schizophrenia. Patients are tormented by voices. They experience delusions of being watched, persecuted, controlled by external agencies. They may believe they are being poisoned, that their mind is being read, that loved ones have been replaced by imposters. Over time, the so-called negative syndrome dominates: anhedonia, amotivation, social isolation. It is highly disabling, impairing the ability to fulfil roles in society or goals in life. If we magically had a cure for schizophrenia, it would be an enormous benefit to humanity.

Like other complex diseases (think of type II diabetes, inflammatory bowel disease, coronary artery disease), the genetic architecture of schizophrenia is highly polygenic, meaning the risk is spread over a number of genes. This includes common variations in the genome each having a tiny effect, as well as rarer variations in larger sections of DNA (copy number variants) with bigger effects. There are no single gene mutations that cause schizophrenia, it is the accumulation of genetic risk that makes an individual susceptible. Then perhaps in combination with environmental factors or by twist of fate, individuals cross a threshold and manifest with the clinical phenotype.

Given the genetic contribution to schizophrenia, it seems plausible (if unethical) that the disorder could be ‘bred out’ of a population. Would this work in practice? We don’t need to speculate, it has already been barbarically attempted, by the Nazis.

Psychiatric genocide

The history of psychiatric eugenics precedes the rise of fascism and taints the reputation of our profession. This wasn’t a fringe movement, it was driven by leaders of the field who are still reverential figures. Eugene Bleuler who coined the name schizophrenia, favoured the sterilisation of people with the diagnosis. Emil Kraepelin, the father of modern psychiatry, promoted ideas of racial hygiene and ‘degeneration theory’ of the Aryan race. Kraepelin died before the Third Reich came to power but his followers played key roles in the Nazi eugenic programme.

Ernst Rüdin, mentored by Kraepelin, proposed that traits such as mental illness, criminality and alcoholism could be bred out of the German population. He help design the Nazi Law for the Prevention of Genetically Diseased Offspring, imposing compulsory sterilisation for genetic conditions such as ‘feeble-mindedness’, schizophrenia, manic-depressive insanity, genetic epilepsy, Huntington’s chorea, genetic blindness or deafness, and alcoholism. Approximately 400,000 Germans were forcibly sterilised between 1934-1939, many because of schizophrenia.

In later years, under Aktion T4, up to 300,000 were killed (sometimes euphemistically called ‘involuntary euthanasia’), for disabilities and mental illnesses. Distinguished psychiatrists of the day such as Carl Schneider, Max de Crinis, Werner Villinger, Friedrich Mauz, Friedrich Panse and Paul Nitsche were involved in the programme. Hans Asperger is accused of referring children to a clinic conducting euthanasia.

Fuller Torrey estimates between 220,000 and 269,500 people with schizophrenia were sterilised or killed, representing 73%-100% of all schizophrenia patients living in Germany between 1939 and 1945. The number of patients in psychiatric hospitals in 1945 was only 14% of what it had been in 1939. An unspeakable horror. Many were slowly starved to death in mental institutions. Others died by lethal injection or in gas chambers. What I find most shocking is that psychiatrists weren’t just the functionaries carrying out the programmes, they weren’t bystanders caught up in the wave of Nazism, or cogs in the machine. Instead, they were the intellect behind the movement and the architects of its implementation.

Thus almost all people with schizophrenia were systematically eliminated from the German population. The psychiatric genocide was ruthless and nearly complete. It was far more brutal, far more comprehensive than proposals from today’s eugenicists, which seem anodyne in comparison:

The proposals I endorsed in the debate were…

Incentives for violent criminals and individuals institutionalized for psychotic disorders to use long-term contraception or undergo sterilization.

Crucially, I was not arguing for coercion. No jailing. No force. No reproductive compulsion.

Diana Fleischman

The eugenics programme instigated by the Nazis will, I hope, never be repeated. The existence of schizophrenia was more or less eradicated from a generation, using methods that even today’s most ardent eugenicist would find reprehensible.

What happened next to rates of schizophrenia in Germany may surprise the hereditarians.

Failure of Nazi eugenics

The prevalence rates of schizophrenia at the end of World War II were predictably low. This wasn’t due to the illness being bred out, of course, it was because virtually everyone with that diagnosis had been murdered. What happens when those with severe mental illness are removed from a population or forcibly sterilised? It turns out that new cases of mental disorders immediately bounce back and over time the prevalence returns to pre-war levels.

By 1965, rates of new cases of schizophrenia were higher in Germany than just about any other comparable countries. While horrifyingly effective in psychiatric genocide, the Nazi’s goal of eradicating schizophrenia from the Aryan bloodline was an abject failure. Schizophrenia, even when applying the most extreme eugenic measures, cannot be bred out of a population.

This should not surprise clinicians who see patients with schizophrenia, the vast majority (~90%) of whom do not have a family history of the illness. It won’t surprise evolutionary psychiatrists who are aware of the paradox that people with schizophrenia already have lower fertility, yet the disorder remains stable across time.

I heard the reason behind this more than 10 years ago at lecture in King’s College London by the Scottish psychiatrist and world-leading schizophrenia geneticist Professor Mick O’Donovan: most genetic risk alleles for carried not by people with schizophrenia but by people without mental disorders.

Complex disorders are complex

The reason why eugenics cannot selectively remove complex disorders like schizophrenia from a population is explained in this excellent Substack post by, you guessed it, Sasha Gusev. He discusses the threshold liability model in which there is an unobserved (latent) liability for developing an illness. This is modelled as a normally distributed bell curve, consisting of common genetic variants of tiny effect, rarer DNA copy number variants with larger effect, as well as environmental risk factors. It is only when the accumulation of risk factors crosses a certain threshold will an individual manifest with the phenotype of that illness, illustrated below.

For disorders with such imprecise boundaries as psychiatric illnesses, it is very plausible two individuals with similar liabilities for schizophrenia might not receive the same diagnosis. Furthermore, there may be individuals that have a relatively high liability to develop schizophrenia, which does not reach the threshold of a florid psychotic episode. Clinically, I think of people presenting with schizotypy who have odd beliefs that don’t quite become delusional or perceptual abnormalities that don’t quite have the quality of a hallucination.

If you think of the normal distribution curve, you would expect a greater number of individuals to have genetic risk for schizophrenia which doesn’t quite meet the threshold of illness than people diagnosed with the disorder. Said another way, most of the genetic risk for schizophrenia is harboured by unaffected individuals.

Sasha models what would happen if we removed all individuals with a 50% heritable polygenic condition with a normally distributed liability from a population, over successive generations. Even after ten successive generations, new cases of the disorder remain relatively common - highly polygenic disorders keep popping up, in contrast to monogenic (single gene) disorders which can be more easily bred out.

I’ve been using schizophrenia as an example favoured by eugenicists because I know it well. However other complex traits, like alcoholism or criminality, are likely to follow a similar threshold liability model. Even extreme measures like we saw in the 1930s would not be enough to substantially suppress new cases in a population.

Breeding out undesirable traits isn’t coming back, because it doesn’t work. However, a more socially acceptable way to reduce the genetic risk of mental illness has recently emerged, polygenic embryo screening. Could this be the new respectable face of psychiatric eugenics?

Maybe we’re all eugenicists

While I think Diana Fleischman is wide of the mark when it comes to breeding, she raises a more astute point in, You’re probably a eugenicist. In many ways modern society does accept what might be regarded as eugenic practices. For decades, antenatal screening has been employed to selectively terminate pregnancies in which certain abnormalities are detected, such as Down’s syndrome (Trisomy 21).

Despite rising maternal age (a strong risk factor for Down’s syndrome) the rates of babies born with this condition has fallen in recent years due to termination of such pregnancies. With some notable exceptions (such as the BBC documentary A World Without Down’s featuring the actor and campaigner Sally Phillips) there is very little opposition to this form of eugenics. It is almost universally accepted.

Pre-natal screening is typically used for malformations, chromosomal abnormalities and severe disabilities. However, now polygenic screening of embryos allows couples undergoing IVF to selectively choose offspring on the basis of traits of like intelligence and height, as well as risk for disorders like schizophrenia.

Typically, couples undergoing IVF will be able to select from a number of embryos to be implanted. Companies, such as Orchid and Herasight now offer embryo reports, calculating polygenic risk scores (based on adding together genetic risk markers across the genome) for a variety of traits and conditions.1

One claim is that this screening can produce a relative risk reduction of 50% for schizophrenia (though with a relatively low baseline prevalence this might not change the absolute risk by much). For most couples, shifting an absolute risk from 0.5% to 0.25% might not matter much, however, if they are concerned perhaps due to family history and already have a higher liability, shifting these risks might seem worthwhile.

This technology is not endorsed by the scientific establishment and I don’t know whether it is ready to be widely deployed but it seems everything is in place to select embryos based on their lifetime risk of developing a mental illness. The time to to discuss whether we, as a society, endorse this new form of psychiatric eugenics is now.

When I first arrived at the Maudsley hospital in 2012, a consultant psychiatrist (now professor) said that in the next few years neurologists would assess each new case of psychosis. Only after they had examined and investigated the patient for reversible causes, would they then pass on management to psychiatry. While the ongoing shortage of UK neurologists precludes this scenario for the foreseeable future, I could see his logic.

The same year, a memoir was released by the journalist Susannah Cahalan about her experience suffering from anti-NMDA receptor autoimmune encephalitis, aptly titled, Brain on Fire: My month of madness. Cahalan initially developed mood swings, anxiety, irritability and paranoia. She struggled with concentration and memory at work. She went on to have auditory hallucinations, persecutory delusions and seizures. First misdiagnosed with a psychiatric disorder or even alcohol withdrawal, testing her spinal fluid revealed the true problem - antibodies against her brain’s NMDA receptors.

Luckily for Cahalan, her astute neurologist Dr Souhel Najjar was familiar with this recently discovered disease and she makes a good recovery with immunotherapy to dampen the inflammation and plasma exchange to remove the harmful antibodies. Cahalan contemplates the alternative, remaining diagnosed with a mental rather neurological illness:

‘…how many people currently are in psychiatric wards and nursing homes denied the relatively simple cure of steroids, plasma exchange, more intense immunotherapy…?’

Another patient, Bex, is featured in BBC Radio 4’s Medical Mysteries. She is a university student who experiences acute anxiety and self-consciousness, along with concentration difficulties and memory lapses. She later develops headaches and is admitted to hospital after having a seizure. In hospital, her condition deteriorates: she loses the ability to speak and hears voices. The medical team initially think she is experiencing an episode of psychosis. Fortunately, working at the hospital was a world-leading expert in autoimmune neurology, Sarosh Irani, who makes the diagnosis of anti-NMDA receptor encephalitis. Bex receives appropriate immunosuppressive treatment and plasma exchange, eventually making a good recovery.

What if patients with this form of encephalitis are not under the care of an eminent neurologist, might they continue to be misdiagnosed with a mental disorder? Sadly, such cases have also been described. Lucy Dawson was detained under the Mental Health Act and treated with ECT, before a diagnosis of autoimmune encephalitis was made. An international rugby player Abi Burton similarly was being treated on a psychiatric ward before a clinician made the connection between her seizures and autoimmune encephalitis.

There are even reports of patients with chronic psychotic disorders like schizophrenia having an autoimmune cause - for example going into spontaneous remission when being treated with immunosuppression, even with no speicific autoantibody identified. Whether there is a type of disorder, autoimmune psychosis, caused by pathological autoantibodies but without neurological symptoms (seizures, cognitive deficits, autonomic dysfunction) is an open question with two possibilities: either we are missing a proportion of reversible psychosis cases or we are talking about a condition that doesn’t actually exist.1

Autoimmune encephalitis

Remarkably, anti-NMDA receptor encephalitis didn’t exist when I started medical school; the syndrome hadn’t been defined and the antibody hadn’t been discovered. The first description of the phenotype was published as a case series of four women in 2005, by a team led by Josep Dalmau (then of Penn, now in Barcelona). Each case was a young woman with a specific type of ovarian tumour, called a teratoma. They all developed acute behavioural change and psychiatric symptoms (often including hallucinations) followed by neurological deterioration (seizures or reduced consciousness). One patient died, the others made a full recovery.

A subsequent 2007 paper from Dalmau of 12 women confirmed the illness course of psychiatric symptoms→memory problems→seizures/movement disorders→reduced consciousness. Half of the patients were first assessed or admitted under psychiatry before the correct diagnosis was made. All bar one developed seizures and all bar two required mechanical ventilation. All patients had tumours and these were usually ovarian teratomas. This time, they found that all patients had antibodies against the NMDA receptor; a new disorder had been discovered. Before this antibody was identified, were cases which fitted the phenotype simply undiagnosed? It certainly seems that way. In 1997 The Lancet published a case report, An acutely confused 15-year-old girl, about a patient with memory impairment and psychosis, who was found to have an ovarian teratoma. She did not respond to antipsychotics but recovered when the tumour was removed. This was almost certainly an early description of the undiscovered disorder.

Over the past 20 years, this new autoimmune disease has been intensively researched and much of its initial characterisation holds up. It has an incidence of 1.5 per million people per year and is around four times more common in women. It is a disorder of young people, with an average age of onset at 21 years. Around a third of patients have an underlying tumour, which is usually an ovarian teratoma, explaining the marked female:male predominance. Most patients respond to immunotherapy and tumour removal, though recovery can be protracted and the mortality is around 7%.

The initial presentation is usually psychiatric which over a period of a couple of weeks progresses to movement disorders, seizures, reduced respiration and autonomic dysfunction (changes in blood pressure, temperature and sweating). A seminal paper led by Adam Al-Diwani synthesised the psychopathology across cases, shown below. The most common individual symptoms were agitation and hallucinations. Approximately 2/3 had documented psychotic symptoms, almost half had mood changes and 1/3 had catatonic symptoms.

Knowing the symptoms and clinical course of this disorder is important. Brain MRI is often normal (another reason not to do these routinely), so clinicians have to go on clinical suspicion to arrange the crucial diagnostic tests of EEG and lumbar puncture.

Most now recognise anti-NMDA receptor encephalitis as a rare but important secondary cause of psychotic symptoms. However, does a forme fruste exist in which antibodies to the NMDA receptor cause psychotic in the absence of typical neurological progression? Before we delve into contentious issue of autoimmune psychosis, it might be worthwhile thinking about the autoimmune pathological process.

Attack of the clones

Autoimmune encephalitis literally means brain inflammation caused by the immune system mistakingly targeting the body’s own proteins, rather than foreign pathogens. How this happens and why it results in psychiatric and neurological symptoms can be thought about within the framework of antigen (NMDA receptor), antibody and phenotype (clinical presentation).

The NMDA receptor sits across the membrane of neurones and is activated by the neurotransmitter glutamate. It is important for memory and learning (i.e. neuroplasticity). The types of drugs that block NMDA receptors (like ketamine and PCP) can induce psychosis-like states and have been used as models for schizophrenia. This led to the glutamate hypothesis which proposes reduced function of the NMDA receptor as a key mechanism, upstream to the most established theory of psychosis dopamine dysregulation. It is not surprising then, that attacking this receptor could produce psychotic symptoms - the next question is why the immune system would target it in error.

The immune system’s job is to protect the body from harmful invaders. A major component is the adaptive immune system, adaptive because it learns from experience. After a first encounter with a pathogen, it builds a memory that allows for a faster, stronger response next time. Antibodies are central to this process: when B cells recognise a specific antigen, they undergo clonal expansion, multiplying into an army of antibody-producing cells. On re-exposure, this force can mount a rapid and highly targeted defence.

A feature of the immune system is its ability to distinguish between self and non-self: this is immune tolerance. When this ability goes awry, the immune system mounts a response against the body, leading to autoimmune disease. Generally, the brain is thought as immune privileged meaning that it is less exposed to the parts of the immune system which recognise antigens and mount a response. When the immune system is presented with neural tissue (which can form part of ovarian teratomas), it may recognise it as ‘foreign’. If the tissue expresses NMDA receptors, we can end up with a clone army of B cells producing anti-NMDA receptor antibodies (not good).2

These circulating antibodies then bind to parts of the receptor, causing it to be internalised into the neuron. Essentially, this results in reduced numbers of the NMDA receptors on cell membranes, sometimes called receptor hypofunction. It disrupts the receptors role in memory and learning, causing amnesia as well as psychiatric symptoms. It also disturbs the brains excitation/inhibition balance, resulting in seizures. In this way, the immune response, mounting an antibody attack against its own brain’s NMDA receptors results in the classic encephalitis phenotype.

Autoimmune psychosis

I hope everything I’ve said so far is uncontroversial. Next we’ll wade into murkier waters. Firstly, can cases of confirmed anti-NMDA receptor encephalitis present with an isolated psychiatric syndrome (i.e. without neurological features)? One study from Dalmau’s group examined 571 patients with a confirmed diagnosis. They found 23 patients had episodes with a purely psychiatric phenotype. However, 18 had previous neurological symptoms in an early episode. There were only 5 (<1%) who presented for the first time in the absence of neurological symptoms. All these patients had antibodies to the NMDA receptor in their cerebrospinal fluid, rather than just in their blood.

One percent of a rare disease is a small number. If we go for the quoted incidence of 1.5 per million people per year, we might expected 100 cases of anti-NMDA receptor encephalitis in the UK each year, with perhaps one of them presenting with an isolated psychiatric syndrome. Given the possibility of misclassifying or overlooking symptoms, this number begins to look like a rounding error.

Of course, a reason that the proportion is so low could be because only patients with seizures and gross cognitive impairment get tested for anti-NMDA antibodies. Perhaps testing everyone with psychosis for these antibodies, might identify a number with antibodies who would never have been thought of as having encephalitis?

A team led by Belinda Lennox in Oxford (COI - my head of department) screened 228 patients with first episode psychosis for anti-NMDA receptor and other neuronal autoantibodies. None of the patients had signs of encephalitis. Seven patients (3%) were positive for antibodies to the NMDA receptor while thirteen were positive for other neuronal autoantibodies. While 3% may seem like a small proportion, approximately 10,000-20,000 patients are diagnosed with a psychotic disorder in the UK each year, meaning up to 500 of these could be positive for such antibodies. However, patients positive for NMDA receptor antibodies did not differ from those who were negative in terms of clinical presentations or outcomes. None of the antibody-positive cases developed neurological symptoms.

Crucially, these antibodies were measured from blood samples rather than cerebrospinal fluid so we don’t know whether they could bind to receptors within the brain. Curiously too, a proportion of healthy controls had neuronal autoantibodies, though to other targets than the NMDA receptor. The proportion was similar (4%) in a study of 387 patients with first episode psychosis from Substack’s own Tom Pollak, though there were no differences in symptomatology, treatment response or outcome. Another study from Dalmau’s team recruiting 105 patients with first episode psychosis (who had undergone lumbar puncture) did not identify NMDA receptor antibody in blood or cerebrospinal fluid, except in patients who additionally had the diagnosis of autoimmune encephalitis.

So where do we stand? We can say that a small proportion of patients with first episode psychosis have antibodies to the NMDA receptor from blood samples, at a higher rate than is found in healthy controls. However, lack of cerebrospinal fluid samples and lack of a clinical phenotype specific to antibody-positive patients, calls into question whether these antibodies are pathological. That the antibodies are found in healthy control blood samples suggests they do not necessarily cause a clinical syndrome.

In my mind there are two ways to resolve whether the autoantibodies found in first episode psychosis patients are indeed pathological. The first is to screen large numbers of patients using lumbar puncture to obtain cerebrospinal fluid. This has been advocated by Belinda and Tom but it would require a change in clinical practice and is a logistical challenge. Most patients I see who are acutely psychotic would lack decision-making capacity for a somewhat invasive procedure with uncertain benefits. The second is to do what we always do to resolve clinical equipoise, a randomised trial.

SINAPPS

The Study of Immunology in Antibody Positive Psychosis (SINAPPS) is a randomised placebo-controlled trial of immunotherapy (intravenous immunoglobulin followed by rituximab) as an add-on treatment for patients with acute psychosis who test positive for anti-neuronal antibodies. Importantly the trial isn’t enrolling patients with suspected encephalitis (seizures are an exclusion criterion) but those with an isolated psychiatric syndrome. The inclusion criteria allows for patients who test positive based on blood or cerebrospinal fluid samples, though from the protocol it seems that blood serum will be the main source.

I say this a lot: clinical trials are hard. They are difficult to setup due to ever increasing bureaucracy, which contributes to them being expensive to run. Recruitment, particularly from people experiencing severe episodes of mental illness, is tough. Even as clinical trials go, SINAPPS is more challenging than most. By focussing on the <5% of psychosis patients who are antibody-positive the study needs to enrol a much greater number (thousands) of patients for antibody screening. Only those who test positive will be offered the 50/50 chance of receiving a treatment which, while standard for encephalitis, has some serious side-effects and requires intravenous infusion.

The study (which I should say is partly run out of my department, though I have no inside knowledge) has so far screened almost 6k people with psychosis from more than 40 health services in the UK. Of the n=80 target sample size, 57 antibody-positive participants have been randomised since it launched in 2017 and its projected end date is 2027.

By now I think you all know that I love randomised controlled trials and this one is a good example of why. It is hard, it is long, it is expensive, it is ambitious. It is attempting to definitively answer whether patients with psychosis associated with autoantibodies but without features of encephalitis respond to immunotherapy. If they do, a new treatment paradigm in psychiatry will be unlocked. If the trial is conclusively negative, then we will have disappointing but ultimately important information about the clinical relevance of these antibodies.

So, does autoimmune psychosis exist? You’ll have to wait and see.

When I started in psychiatry, I was sure that many of my patients suffering from psychosis would actually turn out to have an underlying medical (‘organic’) cause. The consultant psychiatrists running the inpatient wards, so I thought, had drifted from their medical training and forgotten the value of investigations. I bought a copy of The Diagnosis of Psychosis, by Rudolph Cardinal and Ed Bullmore, which lists hundreds of secondary causes of psychosis, from genetic to inflammatory, from rare neurological to infective. I would go through the list of patients on my ward and think: could this one be acute intermittent porphyria? Should I add-on ceruloplasmin to the bloods to exclude Wilson’s disease? No-one wants to miss a brain tumour causing psychotic symptoms. Naturally, I would request an MRI for every patient to exclude cerebral pathology, writing on the request, FEP, routine screening, ?organic cause.

As time went on, I realised that rare secondary causes of psychosis were indeed rare. Most of the scans I ordered came back normal and none showed a clear cause for a patient’s symptoms. The few times where there really was a medical cause for a psychotic episode, there were other unusual features; signs and symptoms that raised suspicions before any investigations were ordered.

I suspect others have similarly shifted their thinking. My friend Matt Butler, published a provokingly titled editorial Psychiatrists should investigate their patients less, arguing that tests should be targeted and based on clinical assessment.

When it comes to MRI scans in first episode psychosis, guidance is conflicting. The Royal Australian and New Zealand College of Psychiatrists recommend MRI but acknowledge that opinions differ on whether this is necessary in all patients. The American Psychiatric Association recommends MRI if indicated on the basis of history or examination. Indian clinical practice guidelines state that neuroimaging ‘may be indicated’ in first episode psychosis. In the UK, NICE does not recommend MRI as a routine investigation.

Having worked in South London and Maudsley for ten years, I know it is common to request MRI scans in first episode psychosis. I also know we have a great tool called CRIS (Clinical Record Interactive Search) which allows researchers to anonymously examine electronic health records of patients. So I asked Matt whether he’d be up for finding out what actually happens after patients have a scan - does it change their clinical management? As we both embarked on PhDs, this became a growing side-project, the results of which have just been published in World Psychiatry.

Making the most of the Maudsley

I’ve written about the importance of institutions like the Maudsley and I don’t think this project could have been done anywhere else, certainly in the UK. In South London we have high rates of psychosis, we have some of the most established early intervention in psychosis services, we have access to MRI scanners, we have had electronic health records since 2007, we have CRIS. Most importantly, despite the ongoing mess of UK medical recruitment, we have a large number of motivated psychiatry resident doctors.

Matt and I were able to extract all cases of first episode psychosis in South London between 2007-2021, we were able to find out which patients were referred for an MRI scan (n=1,693) but we’d each have to review hundreds to find out why the scan was ordered, the results and whether it changed the patient’s care.

So we put out a call to resident doctors and it was answered by Ahmed Abdelsamie, Ishaac Awatli, Sneha Barai, Luke Baxter, Theo Boardman-Pretty, Rebecca Phelps and Anam Waqar. Together, we divided up cases and began the long process of finding out what happens after we do a brain scan in psychosis.

A meta-analysis is published

In 2023, about a year into our project, a meta-analysis was published on MRI abnormalities in first episode psychosis. The lead author was my friend Graham Blackman and the senior author was my doctoral supervisor Philip McGuire. This synthesised results of 1613 participants, across 12 studies. Graham found that 1/4 of patients had an abnormality on their scan, while ~5% had an abnormality that was ‘clinically relevant’, meaning it was reported to have changed management. The most common types of abnormality were white matter changes (often a marker of small vessel cerebrovascular disease) and cysts (usually an incidental finding).

The analysis caused a fierce debate, typified in this Twitter thread by Anil Makam. He points out that ‘clinically relevant’ didn’t necessarily mean ‘reversible cause of psychosis’, instead it could mean clinicians dealing with incidentalomas. He disagreed with the conclusion that routinely screening with MRI should be part of the workup for first episode psychosis.

For us, it was great that Graham’s analysis got attention and sparked debate. It meant that we could harmonise the way we categorised MRI abnormalities and hopefully we could address some of the criticism of the meta-analysis.

Our results

The flow of patients from MRI request to results to change in management is shown below. The most common reason for requesting a scan was ‘routine screening’, in 40% of the sample. Other reasons why patients experiencing psychosis were referred for a scan were atypical features, like cognitive impairment or neurological symptoms.

Our main results replicated the findings of Graham’s meta-analysis: 1/4 of patients had an abnormal scan. We had a higher proportion of scans that led to a change in management, 1 in 10. This might be because all our cases were taken from patients in a health service, rather than those who have agreed to take part in a research study.

Delving into the results a little deeper is illuminating. The reason why the patient was referred for a scan (what you might think of as the pre-test probability) had a big influence on whether the scan was abnormal and whether it changed the patient’s management. For patients who were suspected of having a space occupying lesion, 60% of the scans were abnormal. In cases when a scan was ordered for cognitive impairment, 60% were abnormal and in 25% it changed clinical management. If we stick to ‘routine screening’ MRIs, 15% were abnormal but none changed the patient’s diagnosis or identified a secondary cause of their psychosis.

In the small proportion of scans that led to a change in diagnosis (2.4%) almost all were changed to dementia. This hints at another important factor that was strongly associated with abnormal scans, age. As shown below, in patients under 16, less than 10% had an abnormal scan, whereas all patients over 85 had some sort of abnormality and in 42% the scan changed management. You will see our sample wasn’t restricted to the typical first episode age-range and it was older patients that scans changed management.

Everything in context

You might think, even if the rate of abnormalities is low, some of them are so serious (like a brain tumour) it is still worth scanning everyone with psychosis. However, you’ll pick up rare serious abnormalities in any population if you do enough scanning.

One BMJ meta-analysis looked at MRI brain scans that were done in healthy people, for example in research studies or workplace screening. They found 0.7% of scans had neoplastic findings (brain tumours). In our sample 1% had neoplastic findings, but this reduced to 0.3% when only including scans that were done as part of routine screening.

Another BMJ meta-analysis found ‘potentially serious’ MRI brain abnormalities in 1.3% of asymptomatic people. Using the same criteria, 0.5% of our sample had potentially serious abnormalities and 0.2% of those being scanned routinely.

Overall, it isn’t clear that routinely scanning people with first episode psychosis uncovers serious MRI abnormalities at a greater rate than scanning the general population.

Investigating first episode psychosis

Assessing someone who is experiencing their first psychotic episode comes with a weight of responsibility. It is not unlike seeing someone who has had a stroke - you know this event may change their life-course. It often involves discussing difficult medical concepts with family members, who may be coming to terms with what has happened. On such occasions, it strikes me that I may be seeing patients and their families on the worst day of their lives.

It is important to conduct a thorough assessment, including a good history (including collateral) and examination. In cases in which this suggests atypical features (acute onset of symptoms, neurological features, cognitive impairment, signs of a space occupying lesion) ordering brain imaging is mandatory. If you are referred an older person presenting for the first time with psychosis, get a scan every time. However, if you can’t find features of an ‘organic’ cause, if the patient is young, if you’re ordering the scan as a routine screening, it probably won’t change clinical management.

Whether you think your patient with psychosis needs an MRI scan or you think they don’t, you’re probably right.

You may or may not have heard of the research study Our Future Health, but you will almost certainly know its predecessor UK Biobank. Initiated in 2006, the Biobank is one of British science’s great success stories, described by the Government as a jewel in the crown of UK science, and an envy of the world.

In the four years after it was set up, UK Biobank recruited 500k participants, aged 40-69 across Scotland, England and Wales. Participants gave blood samples for genetic testing, as well as other biomarkers. They did various questionnaires covering health, lifestyle and socio-demographics. Some had scans of their brains, hearts, bones and abdomens. Importantly, this data could be linked with their NHS records.

The data collected by UK Biobank is available to any reputable researcher, whether they are from academia, the charity sector, or commercial industries. This has resulted in more than 16,000 peer-reviewed publications. The latest achievement, published last month in Nature, involved whole-genome sequencing all participants.

The Biobank will continue to produce findings that improve our understanding of genetics, health, disease and aging over the coming years. It has been such a success that various countries are emulating it with their own large, data-rich studies: All of Us Research Program (USA), China Kadoorie Biobank, FinnGen (Finland), German National Cohort, Estonian Biobank, Lifelines Cohort Study.

The most ambitious Biobank-inspired study is based once again in the UK. Larger in scale (with a target of 5 million participants) and aiming to be more ethnically and socially diverse, Our Future Health has already enrolled 2 million volunteers (myself included). With an order of magnitude larger sample size, it has been dubbed the ‘UK Biobank on steroids’ in a highly critical recent BMJ article. In response I’m going to defend and make the case for valuing Our Future Health.

Our Future Health

The researcher behind this charity (its CEO and chief medical officer) is Raghib Ali, a remarkable man. He was born in the UK to parents who had emigrated from India in the 1960s. From the age of five, his family hit hard-times, with his father losing his job and later his eyesight from glaucoma. Raghib was on free-school meal vouchers due his family’s low income. Having attended a poorly performing local school, he went on to win a scholarship at a private school from the age of 11 and subsequently studied medicine at Cambridge. His work on the frontline of the covid pandemic was acknowledged with an OBE. He works as a clinical epidemiologist and has advised the government on ethnic disparities in covid outcomes.

From reading interviews, it is clear that Raghib’s family circumstances and life-story drives him to reduce the differential impact of ethnicity and socioeconomic status on health outcomes. He describes how preventative medicine has been employed successfully in treating glaucoma which caused his father’s blindness.

Our Future Health was awarded funding from various sources, which includes an initial £79 million from the UK Government (rising to £354 million by 2030) and £180 million from the life sciences sector and health research charities. The study partners with a range of organisations including the NHS, pharmaceutical companies, sports organisations and even supermarkets. It is made clear that funding from partnerships and fees to access the data will only be used for research.

Volunteering as a participant is straightforward; the participant information sheet and consent form are available online. After completing a questionnaire, you are invited to attend an appointment - mine was in the car park of my local supermarket - in which various measurements are taken, including height, weight and blood pressure. A blood sample is obtained for genetic testing. Volunteers are now given a £10 shopping voucher as compensation - sadly this was not the case when I enrolled.

The study explains what will and will not be done with participants’ data. Our health records (hospital and primary care) will be linked. We may be recontacted about future research studies. We may even be contacted for personal feedback about our health, which I assume is based on genetic testing like polygenic risk scores. Researchers can apply to use de-identified (essentially anonymous) data, for the purposes of health research in the public interest. Participant data will not be used for advertising or insurance purposes.

The study will build on the success of UK Biobank, disentangling the effects of genetic and environmental factors on the risk of disease. Importantly, by increasing diversity it will be able to investigate conditions that were under-represented in the Biobank, such as psychiatric disorders, which tend to present in early adulthood and are associated with socio-economic disadvantage.

There will eventually be close to a million participants in Our Future Health with depression or anxiety. Could we identify among those individuals who are at highest risk of developing a severe mental illness, like bipolar disorder (~15-20% of those presenting with a depressive episodes will eventually be diagnosed with bipolar). Perhaps those people would benefit from targeted interventions, such as optimising sleep, or they might respond better to treatments like lithium. This is just one example I’ve taken off the top of my head - the range of possibilities from a study this large is difficult to envisage.

Like the Biobank, it will become a platform for health research but with pharmaceutical companies engaged since inception. This could result in better recruitment of participants into clinical trials and identification of novel therapeutic targets. No-one can guarantee this will result in the development of new medications but that’s the hope.

All in all, I think this is a well-intentioned, well-funded, well-run study that could potentially improve the health outcomes of people in the UK and beyond. Not everyone agrees though. Taking a more skeptical slant, Margaret McCartney and Deborah Cohen recently published an exposé of sorts Our Future Health: consent, clinical risk, and industry issues plague the UK’s biggest ever health research programme.

The skeptics

You might be surprised to know that these two investigators are people I have long admired. Margaret McCartney is a Glasgow-based GP and evidence-based medicine stalwart. She previously had a longstanding newspaper column in the Financial Times (think a business-class equivalent of Ben Goldacre’s Bad Science).

Margaret’s route to academia is different but no less inspiring than Raghib Ali’s. She graduated from my own alma mater, the University of Aberdeen, and has worked in the same general practice for more than two decades. Her foray into evidence-based medicine was through public communication. She gained an academic position in 2019 and recently completed a PhD at St Andrew’s University, having first visited the university when her son started as an undergraduate.

Consuming someone’s writing for a number of years gives you a sense of their thought processes and positioning. I would summarise Margaret as being suspicious of unproven interventions, particularly unsubstantiated screening programmes that suck up primary care resources. Here she is on screening for prostate cancer, breast cancer and atrial fibrillation. She is wary of pharmaceutical influence on medical research and of academics’ conflicts of interest. As a duo with Deborah Cohen (ex-medic, current investigative journalist), they have written excellent pieces on the ZOE app, new Alzheimer’s disease drugs, and HRT.

Both researchers are seasoned and highly diligent. So what have they brought to light about Our Future Health?

The hit-piece

The first criticism is that participants don’t understand what they are signing up for; as a consequence of both inadequate information and the use of the NHS-logo on Our Future Health materials.

Participants are required to read a 17-page document explaining the study. If anything, there is too much information on how data will be used. Despite the length of the document, the study has done its best to make the information digestible, it’s not among the worst research documents I’ve read. The Our Future Health site has specific pages on how they are funded, how they are governed, their partnerships (including the NHS), their values and ways to get in touch.

For many, this information will be more than enough to decide if they want to take part. And after all that, if participants do not feel they have enough information to donate their personal data, they are under no obligation.

Professor of sociology Paul Martin (described as an ‘expert in public trust’) warns that it is not clear in the consent process as to how participants’ data will be used:

I’m not sure how much more explicit Our Future Health can be. They have a whole section of their information sheet dedicated to how commercial organisations are involved with a link to the companies they have partnered with. The consent form requires participants to agree that their data can be accessed by ‘academic organisations, charities or companies, in the UK or overseas’.

Helen Wallace, director of GeneWatch UK has concerns that police would be able to access genetic data for the detection, investigation and prevention of crime, based on a recent Act of Parliament. Given this came into being years after Our Future Health had started and Raghib Ali has explicitly said they would not allow access unless ordered by a court, this criticism seems unfair and would presumably apply to every study that obtains genetic information.

The next bone of contention is paying participants with a £10 shopping voucher, with the BMJ re-using quotes from Trustpilot:

This might not be obvious if you don’t do clinical research but paying participants a nominal amount for their time is standard practice. A tenner is less than the national hourly minimum wage, so it’s not exactly a big incentive. Should we try to incentivise participation from under-represented populations? In my mind, we absolutely should.

Other attacks are on the scientific merit of doing this study at all:

Throughout the article, Our Future Health is compared unfavourably with UK Biobank. It seems pessimistic to me that a more diverse, much larger biobank is unlikely to provide novel findings. In recent years, fields as varied as genetics, neuroimaging and machine learning have demonstrated that scaling matters. Increasing a sample size by an order of magnitude can provide novel insights that are not foreseen. More is different. Just because the Biobank is a strong cohort are we going to give-up on creating bigger, better cohorts?

The quotes that most encapsulates the article though, falls to KCL’s Richard Sullivan:

Put aside that the cost of Our Future Health is roughly one tenth of one percent of what the UK government’s benefit spending per annum. If all the study achieves is helping the pharmaceutical industry develop new medications, isn’t that good? Doesn’t that improve health?

Next I’ll linger on one criticism that does seem to hold-up - inaccurate cholesterol results. Participants do a finger-prick test which gives an immediate result. There have been concerns about these tests, namely that participants have had elevated levels which are later much lower on gold-standard venous blood sampling. I experienced this myself when taking part! My immediate cholesterol level was quite raised but when double-checked by my GP was only marginally high. This is exactly the sort of thing that people like Margaret McCartney warn about. Insufficiently evaluated screening tests resulting in increased anxiety for participants and increased workload for the nations over-worked GPs.

In my case, it didn’t do any harm and led to some positive changes in my diet, but it is correct to be cautious about the downsides of medical investigations. In response the test was removed from the study.

An imperfect study

Our Future Health is not a perfect study. As in all research, things can go wrong, like suboptimal cholesterol tests. It is impossible to prevent every negative eventuality. The most we can do is ensure a study has appropriate governance, is adequately managed and that risks are explained to participants. Just because a study isn’t perfect or has risks, that doesn’t mean it should be cancelled. The potential upsides have to be balanced against these downsides.

The BMJ compares it negatively to the UK Biobank but remember that this study faced similar opposition. The original biobank was criticised for inadequate consent procedures, for allowing researchers from insurance companies to access data and for allowing data to be used by race scientists.

Some of the very same critics pop up again:

Can you ever satisfy medConfidential that use of personal data is proportional and that the consent process is adequate? Will an organisation like Genewatch be positive about any study holding the genetic information of millions of Brits? My hunch is that large-scale linking of health records, personal data and genetics is fundamentally against the world-view of these organisations. Of course I believe individuals should be able to opt out of their health records being used for research purposes but I also believe individuals should be free to take part in research studies like Our Future Health and we should minimise barriers to participation.

My biggest criticism of UK Biobank is that it hasn’t (yet, to my knowledge) resulted in any novel treatments. Our Future Health might do better by partnering with pharmaceutical companies from the start. It isn’t flawless; some things have already gone wrong and there may be more controversies to come. However, this kind of large-scale health study is what we do well in this country and leveraging the life-science industry can maximise the gains while reducing costs to the public sector. Our Future Health may not be a perfect study but it is exactly the type of project that the UK should invest in.

Cannabis → psychosis is the closest psychiatry has to smoking → lung cancer. The evidence is summed up in a brilliant 2022 thread by Psych Unseen. Intravenous administration of THC (tetrahydrocannabinol, the component of cannabis thought to be most psychotogenic) causes acute psychotic symptoms in healthy volunteers. Multiple longitudinal studies show that cannabis use in childhood/adolescence has a dose-related increased risk of psychosis later in life. High potency cannabis (high THC-content) is associated with higher incidence of psychosis.

In people who have experienced a psychotic episode triggered by cannabis, 45% are diagnosed with a schizophrenia spectrum disorder within three years. The rate of schizophrenia diagnosis in people with cannabis use disorder (9%) is an order of magnitude higher than the general population (0.6%).

Obviously most people who smoke cannabis don’t develop psychosis, just as 80-90% of cigarettes smokers don’t develop lung cancer. But psychosis is such a serious outcome that I would be similarly worried by a young person smoking high potency cannabis as them smoking cigarettes.

The research evidence chimes with clinical experience. Many patients I meet with chronic psychotic disorders have a history of cannabis use. Psychotic relapse is often associated with cannabis use. Young people presenting with first episode psychosis will often be heavy users.

While THC content has been increasing over time, there are other important components of cannabis. The one which excites clinical researchers is cannabidiol, or CBD for short. Unlike THC, CBD doesn’t make people ‘high’ and in fact is thought to balance out negative effects. As today’s cannabis contains higher concentrations of THC, other cannabinoids such as CBD have been pushed down. Could this be making today’s strong cannabis strains more psychotogenic?

Promising results from small randomised trials in schizophrenia have led to the speculation that CBD could represent a new class of antipsychotic. A £16.5 million multi-centre trial led by my supervisor Philip McGuire aims to definitively answer whether CBD is an effective antipsychotic and whether it can prevent development of psychosis in young people at risk.

So the story is that THC in cannabis causes psychosis, while CBD might prevent it. And the shift towards higher THC and lower CBD might be worsening the psychotic effects of cannabis. What’s more, the harms from cannabis might eventually be outweighed by it providing a new class of antipsychotic!

It’s a nice story. But in science, we like to put nice stories to the test.

Does CBD mitigate the effect of THC?

A new study, Does cannabidiol reduce the adverse effects of cannabis in schizophrenia? A randomised, double-blind, cross-over trial provides an interesting twist on the CBD story. Before going further, I’ll admit I’m not an objective observer. Not only is the last author Philip McGuire but the first author is a friend, Edward Chesney. Perhaps even more significantly, I think the study is really cool. It asks an important clinical question and uses a clever study design to answer it.

Ed wanted to find out whether giving patients CBD prevents the psychosis-inducing effects of THC. To do this, he recruited patients with schizophrenia whose illness was known to be exacerbated by cannabis use. Patients had to be stable for three months and be using cannabis on at least a weekly basis.

He then, to put it erm bluntly, got them high on potent medical-grade inhaled cannabis (20% THC content). In case this dose did not induce acute psychotic symptoms (remember these are regular, long-term cannabis users with high tolerance) he got them back in for another visit and doubled the dose. If this didn’t result in psychotic symptoms, he invited them back to have 3x the original dose (only one patient required this super-dose).

If you think it was brave of Ed to get patients who are known to be susceptible to cannabis induced psychotic relapses and give them high strength cannabis in the lab until they become acutely psychotic, I agree with you! This study was brave. One patient developed persecutory delusions about the study team. They managed to get one patient’s blood pressure high - up to 224/78 mmHg. I’m sure at points during the experiments Ed’s blood pressure was similarly elevated.

This study wasn’t risk free and wouldn’t have been easy. That it was done at all is testament to the research team, local clinical services, regulatory authorities and most of all, the patients themselves - which the paper acknowledges.

Each participant had two randomly ordered experimental visits, in one they were given CBD (at 1g, a treatment dose) and in the other they received placebo. This was a proper double-blind trial, less than half of participants guessed their treatment allocation and researchers correctly guessed the treatment allocation for less than a fifth of patients. Ed wanted to see if CBD prevented THC-induced psychosis; the results were surprising, a clear effect, but not in the predicted direction.

You won’t believe what happened next

In the placebo experiments, THC resulted in PANSS scores (a standard measure of psychotic symptoms) increasing by an average of 3 points. Pre-treatment with CBD resulted in worse psychotic symptoms, an average increase of 5 points. Seven participants experienced a large increase in psychotic symptoms (>9 points) - all of these were after pre-treatment with CBD. Performance on a memory test, Hopkins Verbal Learning Test-Revised, was also worse after pre-treatment with CBD. Interestingly, CBD did not affect self-rated level of intoxication.

When I first saw these results, I assumed the CBD and placebo groups had been mislabelled during the analysis. It was no mistake though - the CBD and THC blood levels were taken at regular intervals so we can be sure which was which. We can also see the pre-treatment with CBD did not effect THC levels, or its active metabolites (shown below) so it wasn’t a simple case of THC concentrations driving the differences.

Why did CBD worsen THC-induced psychotic symptoms, contrary to everyone’s expectations? The authors don’t really give an explanation, apart from saying they don’t believe the effect is mediated through pharmacokinetic interaction with THC. Both THC and CBD have complex effects on the endocannabinoid system, including cannabinoid receptors and enzymes that act on cannabinoids, such as FAAH (fatty-acid amide hydrolase 1).

If there is a simple explanation of how THC and CBD produce psychotic symptoms, I’m not aware of it.

Taking a step back

I believe this study has advanced our understanding of cannabinoids in psychosis. Firstly, it answered the question of whether CBD prevents THC-induced psychotic symptoms in schizophrenia: no. Treating patients who have schizophrenia and co-morbid cannabis misuse with CBD is likely a non-starter. This has practical implications for people desiging larger clinical trials of CBD for schizophrenia. If I were involved in these trials (I’m not) I would not want to enrol people with schizophrenia who are regular cannabis users. Finally, it shows that simplistic explanations of cannabinoids influencing psychosis risk (THC=bad, CBD=good) are too simplistic.

Taking another step back, I’m hugely impressed and slightly envious that Ed completed a significant contribution to clinical research during his PhD. We now know that most neuroimaging or genetic research requires sample sizes that are prohibitively large for stand-alone PhDs but it is still possible to do meaningful experimental studies with sample sizes that can be realistically recruited by a dedicated and industrious PhD student.

Coincidentally, on the same day Ed’s study was published, another psychiatry registrar at South London and Maudsley and former doctoral training fellow at KCL, Luke Jelen, published an n=26 double-blind placebo controlled cross-over trial of naltrexone prior to ketamine infusion for depression in Nature Medicine. Luke showed that naltrexone attenuated ketamine’s antidepressant effect, implicating the opioid system in its mechanism of action.

If you are an ambitious clinician planning a clinical research training fellowship, you’d do worse than using these two studies as inspiration. You don’t need massive sample sizes, a large research team or multi-centre recruitment for impactful clinical research. A well-designed experimental study can provide definitive answers to important questions.

Another reminder that the randomised controlled trial continues to reign supreme.

Antidepressants, so the story goes, work by boosting levels of serotonin. The most widely prescribed class, selective serotonin reuptake inhibitors, stop the removal of this neurotransmitter from the gaps between neurones. While there is endless debate on the precise role of serotonin in depression (too much ink has been already been spilled), this transmitter is clearly involved in regulating mood and emotion.

SSRIs are an effective treatment for depression, but their benefit above placebo is smaller than we’d like. They seem to be better at certain depressive symptoms, particularly anxiety, than others. Many people report characteristic side-effects which can be intolerable. They are associated with initiation anxiety and nausea, and longer-term sexual dysfunction. I’ve written about difficulties faced by patients when trying to stop these medications - withdrawal from SSRI can result in a syndrome of unpleasant symptoms like ‘brain zaps’, dizziness and nausea.

Serotonin’s role in mood-regulation may lead to restriction in the range of emotion, sometimes called blunting or numbing. Some people would have you believe antidepressants’ mechanism of action is through inhibiting emotions. I’ve never been sold on this theory - not all antidepressants act on serotonin or cause these emotional side-effects.

Depression, we all know, is a heterogenous condition. Some patients experience prominent anxiety and might respond well to SSRIs. Some patients at the severe end experience psychosis and are best treated with a combination of antidepressant (serotonergic) and antipsychotic (dopamine receptor blockade). Some patients experience episodes of low mood interspersed by episodes of elevated mood (bipolar depression) and respond better to mood stabilisers like lithium.

There is another group of depressed patients who seem to have marked impairment in motivation, inability to experience joy (anhedonia) and low energy. Theoretically, we might expect these patients to respond less to SSRIs and more to drugs targeting the dopamine system, which is closely linked with reward. We already have an antidepressant with dopaminergic activity called bupropion, though frustratingly in the UK it is licensed only for smoking cessation rather than depression.

Having new classes of antidepressant which target dopamine might be helpful for the subgroup of patients that don’t respond well to conventional antidepressants - sometimes called, a little uncharitably, treatment resistant. Looking beyond serotonin might also increase the range of symptoms which can be effectively treated.

By happy coincidence we already have groups of medications designed for the exact purpose of increasing dopamine activity, drugs for Parkinson’s disease. Dopamine agonists are medications which increase activation of dopamine receptors. One dopamine agonist in particular, pramipexole, has previously shown promise in small trials for both unipolar and bipolar depression.

This week, The Lancet Psychiatry published a largish (n=150) randomised placebo-controlled double-blind trial of pramipexole as augmentation (added-on to usual treatment) for treatment-resistant depression. If you’re a psychiatrist treating patients with depression, I suggest you start learning how to spell and pronounce this medication, because the results are impressive.

PAX-D

The study, called PAX-D, was set up from the University of Oxford1 and launched in 2021. It was a multi-centre trial, recruiting patients from across England. To be eligible, participants had to be diagnosed with depression, with at least moderate scores on the Quick Inventory of Depressive Symptomatology. They had to be treatment resistant meaning a lack of response to at least two antidepressants at therapeutic doses within the current episode. Participants were ineligible if they had ever been diagnosed with bipolar disorder, psychosis (including psychotic depression) or Parkinson’s disease. They couldn’t be prescribed an antipsychotic (this has the opposite effect to dopamine agonists) or have ‘serious suicide or homicide risk’.

After being randomised to either placebo or pramipexole, the main outcome measure was the QIDS at 12-weeks. Importantly (depression trials are often criticised for being short-term) the effectiveness was measured up to 48-weeks. The study wanted to determine whether this intervention was effective at treating depressive symptoms and whether its side-effects were tolerable.

Before we get into the results, let me set the scene. We know that drug development is hard. Most candidates fail. The costs are huge and the rewards uncertain. Most compounds which look promising in small trials are duds. Most effect sizes shrink or disappear completely when tested in an adequately powered sample.

The bureaucratic hurdles in conducting clinical research in the UK are well known, to the extent that even completing a trial of this size is an impressive; an immense effort from many researchers, patients, and clinicians. A well-conducted trial significantly contributes to medical knowledge, even if the results are negative. However, for PAX-D the results weren’t negative, they were better than anyone could reasonably expect.

The primary outcome is plotted below.

The primary efficacy measure shows an effect size (standardised mean difference) of 0.87. As effect sizes go, this falls into the ‘large’ category. For an antidepressant, it is huge. SSRIs have an effect size of ~0.3, which is still interpreted by most clinicians as meaningful. It’s even bigger than the effect sizes of psychedelic trials. Effect sizes approaching 0.9 are rarely seen in serious trials of psychiatric medications. Even more remarkably, it is in a patient population that by definition has failed to respond adequate to conventional antidepressants. On average, participants had been experiencing a depressive episode for a period of 8-10 years, indicating chronic illness.

The effect size is big enough to make me nervous I’ve missed a glaring error in the study methodology. But their analysis plan was straightforward and pre-specified. The protocol was published in advance. Sensitivity and additional analyses are provided. Maybe the biggest issue identified is that around 3/4 of participants correctly guessed their treatment allocation, threatening the ‘double-blind’ design. I do predict that, simply due to regression toward the mean, subsequent trials of pramipexole will have smaller effects.

Of course, in medicine, there is no free lunch. All treatments have side-effects and we should expect those with large effects to have more severe side-effects. Adverse events were similar across groups 99% in pramipexole versus 97% in placebo, as were serious adverse events (5% versus 3%). However, there was a higher rate of dropouts due to intolerance (20% versus 5%). There were two cases of impulse control disorder in the treatment group and none in the controls. These impulsivity side-effects, which can result in compulsive gambling, excessive spending or hypersexuality, are a known complication of dopamine agonists.

Psychiatrists who spend their time prescribing dopamine-blocking meds for bipolar disorder and psychosis will naturally feel nervous about a medication that increases dopamine activity. Perhaps related to this, a sister trial, PAX-BD, which examined pramipexole for bipolar depression, was terminated early due to recruitment problems. PAX-BD aimed to recruited 290 participants but only randomised 39. In this underpowered sample, there were hints at an effect against depressive symptoms (not statistically significantly) while the drug did cause a statistically significant increase in hypomanic symptoms.

Answers and questions

PAX-D gives us a clear answer: pramipexole is an effective add-on to other antidepressants in people with chronic long-term depressive disorders which have not responded to treatment. The most surprising aspect of the trial was the magnitude of this effect.

However, like many research findings, it provokes a flurry of new questions. Could pramipexole work as a monotherapy? Will it work even better in patients with less chronic problems? Where is its place in the treatment algorithm? Almost half of the trial participants were prescribed an SSRI, is it this combination of serotonin plus dopamine action that really makes the difference? Like bupropion, could it be useful for patients who cannot tolerate SSRIs due to emotional numbing or sexual dysfunction?

Most provocatively: could it be more effective than the antidepressants already at our disposal?

I suspect psychiatrists seeing patients with difficult to treat depression will start thinking about pramipexole, particularly for anhedonic symptoms. Its precise role in clinical practice should become clearer with further research in the coming years.

Takeaways

This study has several take-home-messages at different levels of analysis. Firstly it consolidates the dopaminergic system as a therapeutic target in depression. As a corollary to this, depression isn’t just about serotonin. And antidepressant action probably doesn’t rely on restricting normal emotion.

Taking a step back, the trial is another success story for drug repurposing. Pramipexole was licensed for Parkinson’s disease, its side-effect profile was known, it was ‘oven ready’. A mixture of clinical and theoretical reasoning led to it being an antidepressant candidate - and it worked! How many more medications are on the shelf, waiting to be tested for the right indication?

Finally, this trial was funded by NIHR, effectively part of the UK government. For repurposing studies without pharmaceutical financial incentives, we need university research departments to propose plausible candidates and for government agencies to fund the studies that determine effectiveness. As a country we need to back this type of research, because the UK does it well.

Ultimately, I think this new treatment will have tangible improvements for people suffering from chronic depression. Let’s take these victories in psychiatry where we can.

One of the most remarkable phenotypes in psychiatry, maybe in the whole of medicine, is bipolar disorder. It is defined by an episodic course with episodes at either ‘poles’ of the mood spectrum.

At one extreme are manic episodes, consisting of elated, euphoric or irritable mood, accompanied by increased drive and racing thoughts. People experiencing mania have decreased need for sleep and increased motor activity. They may behave unlike their usual selves; taking risks, embarking on business ventures or spending excessively. They may exhibit ‘grandiosity’, elevated self-esteem beyond normal limits. They may experience psychosis which typically fits with their mood state. This could manifest as grandiose delusions, such as believing they are a celebrity, a royal or even God himself.

Invariably, a manic episode will bring a person to the attention of the police or health services. The disinhibition and impulsivity makes them extremely vulnerable which can result in detention under the Mental Health Act for their own safety. I have met many people experiencing mania. I can’t think of any cases in which it looked enjoyable. These aren’t pleasurable ‘highs’. They are the kind of episodes that leave you at risk of financial and reputational ruin. Recovering from a manic episode may require rebuilding relationships, careers, or even self-worth.

At the other extreme are depressive episodes. These eventually predominate the clinical picture and are associated with profound debilitation, as well as increased risk of suicide. Symptomatically, they are similar to episodes of major depressive disorder (sometimes called ‘unipolar depression’). Bipolar depression, though, is more likely to include psychotic features and more somatic symptoms like increased appetite and hypersomnia. It has treatment implications too: we worry about antidepressants triggering mania (‘manic switch’) and generally expect a poorer response. Instead, antipsychotics or mood stabilisers are first-line.

We know that depressive episodes can take place in the absence of manic episodes. What has not been recognised in either DSM (North American) or ICD (rest-of-the-world) diagnostic systems is unipolar mania. Does the phenotype of manic episodes in the absence of depression exist? If so, we might have to reconsider whether mania is really synonymous with bipolar disorder.

Kraepelinian dichotomy

Our conceptualisation of bipolar disorder goes back to Emil Kraepelin. Who else? He divided the major psychoses into ‘dementia praecox’ (literally ‘premature dementia’ but what we now recognise as schizophrenia) and ‘manic depressive insanity’ (which is now called bipolar disorder).

The distinction between these two was the marked progressive decline and cognitive impairment of dementia praecox contrasted with the episodic affective disturbance associated with manic depression. Kraepelin didn’t get everything right, it is still debated whether schizophrenia shows progressive cognitive decline, but for better or worse, his dichotomy persists more than a hundred years later.

The boundary between schizophrenia and bipolar disorder is not clear-cut. There is substantial genetic overlap, with a genetic correlation of around 0.6. While a group of patients exhibit major affective disturbance with more schizophrenia-like features, forming the diagnostic category of schizoaffective disorder.

Nevertheless, the dichotomy does have explanatory power. For example, copy number variants (deletion or duplication of large parts of DNA) are associated with schizophrenia and schizoaffective disorder but not bipolar. It makes sense that these types of genetic changes are implicated in neurodevelopmental problems and cognitive impairment.

While the broad category of bipolar disorder has remained largely intact from Kraepelin’s description, over time and subsequent DSM editions, this has expanded to include what I will call bipolar-spectrum disorders. Notably this spectrum does not include unipolar mania, despite this being first described (yes, by Kraepelin) as periodic mania in 1889.

The bipolar spectrum

Kraepelin’s manic-depression is now known as bipolar type I. A diagnosis requires at least one manic episode but does not require an episode of depression. Bipolar 1 has high inter-rater reliability; as might be expected, after all mania should be a barn-door diagnosis for a competent clinician.

The DSM chapter called Bipolar and related disorders now encompasses much more than the Kraepelinian manic-depression. In DSM-IV, bipolar type II was introduced. Diagnostic criteria requires a depressive episode but accompanied by at least one hypomanic episode (literally meaning below-mania) and no previous episodes of mania.

Hypomanic episodes, by definition, are less severe than manic ones. They might still involve increased drive, energy and risk-taking but without psychosis or an inability to function in society. These episodes probably wouldn’t come to the attention of health services on their own. The concept of bipolar II is controversial and involves recategorising a subgroup of depressive patients as having bipolar disorder, usually by retrospective description of hypomania. It is thus much more subjective, with lower reliability than bipolar I.

Surprisingly to me, depressive episodes massively outnumber hypomanic episodes, at a ratio of 39:1. In some ways, you might think of bipolar II as a specific type of depression. One in which a history of hypomania dictates treatment choices. Another disorder included in this spectrum is cyclothymic disorder. For this diagnosis, there has to be fluctuation in mood but not enough to meet criteria for clear-cut hypomanic or major depressive episodes.

While the diagnosis of bipolar has undergone expansion over time, it has not included the subtype of unipolar mania. Patients experiencing mania will be classified as having bipolar disorder whether or not they have experienced episodes at the depressive pole. Whether this is warranted rests on mania always heralding an episodic illness encompassing both poles of the mood spectrum.

Defining unipolar mania

Simply put, unipolar mania is at least one manic episode occurring in the absence of depressive episodes. However, to be scientific, we need a more precise definition. If we want to show that manic illnesses can take place without depression, we should exclude mania which is secondary to other causes. This might include drugs, either prescription (the textbook example is corticosteroids) or recreational. We should exclude an underlying medical/neurological condition (this should be suspected if the first episode presents late in life).

Furthermore, we should probably require substantial follow-up before declaring that the person doesn’t experience depressive episodes. After all, a proportion of those with a bipolar illness course (up to 40% in some studies) will first present with mania, later followed by depressive episodes. We should stipulate that manic symptoms can’t be as part of a schizophrenic or schizoaffective illness, as these conditions do not require distinct depressive episodes. Episodes which include both manic and depressive symptoms, known as mixed affective states, should be excluded too.

When taking all of the above into account, and bearing in mind that bipolar I is relatively rare (~0.5% global prevalence), finding reliable estimates of mania in the absence of depression is a challenge that will require large sample sizes.

Searching for unipolar mania

There have been various efforts to identify cases of unipolar mania. One is to simply ask patients diagnosed with bipolar I whether they have had previous depressive episodes. The problem with this approach is that retrospectively reporting anything to do with health is unreliable. When cases are defined in this way and then prospectively monitor them, up to 80% go on to develop at least mild depressive episodes over a 20 year follow-up.

Another method is record-linkage, whereby the hospital records containing diagnostic codes and prescriptions are obtained, often for a whole country. This was done in Taiwan and found that 14.9% of patients with bipolar 1 had no depressive episodes and no prescription of antidepressants. However, this kind of study relies on (i) depressive episodes coming to the attention of health services, (ii) depressive episodes being correctly coded on the system, (iii) antidepressant medication being used to treat bipolar depression. I think each of these assumptions is questionable.

Another approach is to search electronic health records of patients with bipolar disorder. This is a slightly enhanced version of a simple record-linkage study. Not only can researchers search for patients who have not had depressive episodes coded on their health record, they can search the text of patients de-identified medical notes for mention of depressive symptoms.

This method was used by a team at the Maudsley, led by Paul Stokes and Allan Young. They identified records of patients who had a diagnosis of mania or bipolar but never a diagnosis of schizophrenia, schizoaffective disorder or depressive episode. They then searched the text of these notes for words associated with depression. To ensure an adequate length of time was available to develop depressive episodes, patients had to have at least 10 years of notes available. The team identified 17 patients who experienced manic episodes, in the absence of depressive episodes. This represented 1.2% of 1458 patients with bipolar 1. Put differently, 98.8% of patients who experience a manic episode also suffer a depressive episode.

While I expect this proportion will be refined by future replications in larger electronic health record samples, this is the most reliable estimate that I’ve come across to date.

Nothing is 100%, but…

It is a truism that nothing in medicine can be said with 100% certainty. We have all heard stories of cancer going into spontaneous remission. Or of patients being cured by unconventional treatments. Psychiatry is particularly bad at prediction when it comes to prognosis. So it’s pretty remarkable that we can say with ~99% certainty that if someone experiences an episode of mania, they will experience an episode of depression at some point as well.

It is astounding that we can tell people who have experienced a manic episode - we know that around 99% of people who have experienced an episode like yours will experience episodes of depression too. This has massive prognostic and treatment implications.

You may quibble with Stokes and Young’s estimation; it is certainly lower than other (though I’d argue less rigorous) studies. However, it’s also possible that their 98.8% would creep toward 100% using a longer timeframe of say 20 or 50 years, or using stricter definitions of mania to exclude any evidence of schizophrenia. It’s possible that some of that 1.2% did experience mild depressive episodes that just weren’t recorded in their health record. Equally, it’s possible that initiatives to improve services for first episode mania might reveal different patterns of illness when patients are systematically and periodically assessed after an index manic episode.

For now though, there is no need for an urgent DSM revision.

Depressive episodes are ubiquitous in patients who experience mania.

The concept of bipolar disorder is robust.