Clinical Staging, Early Intervention, and Youth Mental Health: An Interview with Patrick McGorry

How do we intelligently connect multifactorial, pluripotential syndromes to treatment decisions?

Patrick D. McGorry, MD, PhD, is an Irish-born Australian psychiatrist and Professor of Youth Mental Health at the University of Melbourne, where he also serves as Executive Director of Orygen, Australia National Centre of Excellence for Youth Mental Health. After his family immigrated to Australia, McGorry earned bachelor’s degrees in medicine and surgery from the University of Sydney (1977) and doctorates in psychiatry from Monash University (1991) and the University of Melbourne. In 1992, he founded and became director of the Early Psychosis Prevention and Intervention Centre (EPPIC), the first Australian organization to focus specifically on young people rather than adolescents or adults, and its model has since influenced mental health services across Asia, Europe, and North America. In 2001 he established Orygen as a translational medical research institute for youth mental health and it is now the largest mental health research organization in Australia. His advocacy led to the establishment by the Australian Government in 2005 of the National Youth Mental Health Foundation, which was branded as Headspace in 2006 and this model of care now operates in 174 Australian communities. His contributions have earned him numerous honors, including the Australian Government Centenary Medal (2003), the 2010 Australian of the Year award, and the Founders’ Medal of the Australian Society for Psychiatric Research (2001). He has done pioneering scientific work in the multidisciplinary treatment of first-episode psychosis, the development of clinical staging models in psychiatry, and youth mental health.

From “Dementia Praecox” to “First Episode Psychosis”

Aftab: I’d like to start by going back to the beginning of your career. You trained in medicine in Sydney in the 1970s, completed your psychiatric training in Melbourne, and by 1992 you had founded EPPIC, the Early Psychosis Prevention and Intervention Centre. I’m struck by how you’ve been at the forefront of rethinking how we conceptualize and respond to emerging psychopathology in young people. So, I’m curious: what was your big-picture understanding of mental disorders in your early days as a trainee, and how has it changed over time?

McGorry: When I was young, I was caught up in a lot of revolutionary thinking. As a medical student in the 1970s, that spirit was still very much alive in Europe. Antipsychiatry currents were strong in places like Heidelberg, where the Socialist Patients’ Collective had emerged. It was a radical time, with broad dissatisfaction with institutions, and within psychiatry that dissatisfaction had a particularly intense focus. The old asylums were oppressive places, and many had clearly lost their purpose.

I was drawn to psychiatry partly because I’d always been as interested in the humanities as the sciences. Psychiatry seemed to offer a unique blend, from philosophy through to medicine, an integrative, diverse field that still sat inside medicine. But I was also deeply wary of psychiatry because of the antipsychiatry literature I’d consumed. Then a suitable opportunity emerged: I’d spent a few years working in internal medicine, and a very humane professor of psychiatry, someone I respected a lot, the late Professor Beverley Raphael, was appointed where I was working in Newcastle. She gave me hope that a different psychiatry was possible. With her as a role model, I thought I could give psychiatry a chance. When I started training, it felt like stepping back decades compared to the general hospitals, in terms of the quality of care, the custodial atmosphere of the old institutions. But I persisted.

Near the end of my training, I moved to Melbourne to work with a newly appointed professor, Bruce Singh, who gave me the opportunity to establish a research unit in one of the old asylums there, a small facility near the city center. I decided to focus on first-episode psychosis, and this decision was shaped by an earlier experience in medicine. I’d worked in an education and stabilization center, created by the late Dr. Paul Moffitt, for newly diagnosed diabetics, many of them adolescents having to come to terms with a life-changing diagnosis. The psychological impact of being diagnosed with diabetes was real and often profound, and diabetes is relatively straightforward compared to what it means to be told you have schizophrenia or psychosis. That contrast stayed with me.

I also observed unconscionably long delays before young people with psychosis accessed treatment, and I witnessed how harmful many treatment experiences could be, both the neglect and the coercive aspects. It became clear to me that you first had to remove the iatrogenesis of the system itself before you could build something genuinely hopeful, recovery-oriented, and humane. That’s what we set out to do.

Aftab: How was your understanding of psychopathology developing at this early point in your life?

McGorry: My PhD was fundamentally about mapping, in minute detail, the psychopathology of a first episode of psychosis. I developed a multidiagnostic instrument that would allow me to capture the full phenomenological landscape of these experiences. I drew extensively on historical concepts from classical descriptive psychiatry, and where those concepts hadn’t been properly operationalized, I worked to operationalize them myself. The result was a comprehensive set of definitions covering psychotic and severe mood phenomena within a single instrument, administered both at the beginning and end of the episode, supplemented with detailed information from family members and other informants.

The aim was to produce a rich, textured portrait of what actually happens during a first episode of psychosis, not just a symptom checklist or a diagnostic label. I wanted to understand the phenomenological complexity, how these experiences unfold, what their constituent elements are, how they evolve over time and how people react to these typically profound experiences. That process immersed me deeply in psychopathology in a way that fundamentally shaped my subsequent thinking. It taught me to pay close attention to what young people were actually experiencing.

Aftab: How did you go from a more rigid schizophrenia concept, from Kraepelinian thinking, to the more fluid “first-episode psychosis”?

McGorry: In a way, moving to “first-episode psychosis” as a conceptual frame was strategic, because I’d come to believe that the Kraepelinian dichotomy created a terrible problem for the field—a fatal flaw. I’d read Kraepelin’s original text closely and realized that the basis for the concept of dementia praecox was essentially a deteriorating outcome. It was an attempt to unify diverse clinical presentations under a common end state, a syndrome defined by progressive deterioration.

I think Kraepelin was inspired by neurological diseases like Alzheimer’s disease, where you genuinely do find a common underlying biology producing a syndrome that culminates in a specific, severe outcome. He tried to apply that same logic as a heuristic for functional psychosis: gather together cases with poor outcomes and assume they share a common underlying basis.

But that approach wouldn’t work in other areas of medicine. If you grouped together all cases of chronic renal failure and assumed they were a single entity with a common cause, you’d be committing a fundamental conceptual error. You’d be confusing a prognostic pathway with etiological unity. And that mistake has had severe consequences in psychiatry which persist to this day.

It leads directly into the twin problems of heterogeneity and pleiotropy, which are two sides of the same coin. Even something with a relatively simple biological cause can present with diverse syndromal expressions; conversely, a syndrome like psychosis or depression can arise from multiple distinct causes and pathways. The Kraepelinian framework conflates these issues by reifying diagnostic categories as if they were disease entities.

This framework became the cornerstone of clinical attitudes that I found deeply problematic. In case conferences, I watched teams spend enormous energy debating: is this schizophrenia, which implicitly meant hopelessness and deterioration, or is it bipolar disorder, which felt like “winning the lottery” because the prognosis was assumed to be so much better? But in real patients, outcomes overlap far more than that stark dichotomy suggests. Many people diagnosed with schizophrenia recover well, and many with bipolar disorder struggle chronically.

More troubling, communicating low expectations for recovery, whether explicitly or implicitly, was profoundly iatrogenic. It contributed to demoralization, hopelessness, and suicide risk. The prognosis narrative was harmful as well as empirically incorrect. Even older longitudinal work, including the WHO cross-cultural studies, showed substantial rates of recovery in schizophrenia. The idea that you should tell someone experiencing their first episode that their outlook is inevitably terrible violates basic principles of medical practice. You wouldn’t approach someone with a serious cancer diagnosis that way, telling them at the outset there’s no hope. Why would we do that in psychiatry?

McGorry: The idea that you should tell someone experiencing their first episode that their outlook is inevitably terrible violates basic principles of medical practice.

Aftab: At that point in your career, what was your general understanding of what mental disorders are? Conceptually, how were you thinking about psychiatric constructs?

McGorry: I started with a strong pull toward environmental, sociological, and psychological explanations, partly shaped by antipsychiatry influences, but also because psychoanalysis was still prominent in Australian psychiatry at the time. I had substantial psychodynamic influence in my training, which made me receptive to ideas of multiple causation and psychological meaning.

I was quite resistant to biological explanations initially, partly because Kraepelinian determinism felt inseparable from biological reductionism and therapeutic nihilism. But the reality of immersion in frontline care and the success of careful biological treatments for most patients rapidly helped me to see that biology was undeniably part of the picture. Diabetes became a useful analogy for me: you can have a clear biological basis for an illness while recognizing that environmental and psychological factors profoundly influence its course, the degree of disability, and the prospects for recovery. So I tried to move toward a genuinely integrative biopsychosocial framework, grounding it in what was actually happening on wards and in clinical services.

One experience that profoundly shaped my thinking in those early years was witnessing what was being done with antipsychotic medications, particularly in acute settings. There was a prevailing fashion for “rapid neuroleptization.” The philosophy being that if some medication is good, more must be better, so very high doses were given very quickly with the belief that people would get well faster. There was no evidence supporting this approach, and it caused a lot of harm.

In the first year I worked at the hospital, Royal Park, where we developed our first episode psychosis unit, eight patients died of torsades de pointes, fatal cardiac arrhythmias directly caused by massive doses of antipsychotics. That was catastrophic, and it clarified something fundamental for me at the time: the first principle for young people experiencing their first episode had to be protecting them from the iatrogenic effects of high-dose medication practices and also mixing newly diagnosed young people with much older chronically ill patients.

If you look back historically, quite small doses were sufficient when neuroleptics were first introduced in the 1950s. The problem developed when medication increasingly became used for behavioral control rather than careful treatment of core psychotic symptoms. We returned to very low doses, something like 2 mg of haloperidol as a starting point, and we used benzodiazepines to reduce anxiety, distress, and behavioral disturbance rather than simply escalating antipsychotic doses. Patients’ experiences were dramatically better. They weren’t zombified, they could think more clearly, they maintained more of their sense of self.

Then we set about developing comprehensive psychosocial treatments. At that time, there weren’t many effective psychosocial approaches available in routine psychiatric care. There were pockets of good practice, some social skills work, and Ian Falloon had written compellingly about family interventions, but nothing like the integrated recovery-oriented program we envisioned.

So we built what we called a recovery program: psychosocially oriented, actively involving families, infused with hope and realistic optimism. It was pragmatic and evidence-informed, drawing on whatever worked while remaining grounded in patients’ actual experiences and needs. And we began to conduct research, especially in new psychosocial interventions, and integrate these with much more careful medication strategies.

Aftab: Within the psychiatric community at that time, was there resistance to the idea that psychotic disorders could be effectively managed through psychosocial interventions?

McGorry: Psychiatry was deeply split along ideological lines and sadly this still lingers. On one side, you had psychoanalysts, who believed psychoanalysis could essentially treat anyone, including people with severe psychotic disorders. On the other side, you had biologically oriented psychiatrists, especially those working in the asylums and large state hospitals, who more or less believed psychological treatments were useless, that psychosis was fundamentally a brain disease requiring medication, and that talk therapy was at best irrelevant, at worst a distraction. Sadly, that latter attitude remains common in many acute psychiatric settings even today. The former attitude of psychological reductionism also dies hard.

I did engage seriously with some of the more thoughtful psychoanalytic writers—Silvano Arieti comes to mind—who at least attempted to understand the phenomenology and meaning of psychotic experience in sophisticated ways. But much of the mainstream psychoanalytic culture involved people writing quite confidently about psychosis without actually treating actively psychotic patients in any sustained way. There was often a disconnect between the theory and the clinical reality.

We were trying to chart a different course. Our approach became: use medication, but carefully and at low doses, always combined with robust psychosocial support, family involvement, and attention to the social and psychological dimensions of recovery. We weren’t interested in ideological purity. As clinical scientists we wanted to know what actually helped young people recover and get their lives back.

McGorry: I was quite resistant to biological explanations initially, partly because Kraepelinian determinism felt inseparable from biological reductionism and therapeutic nihilism. But the reality of immersion in frontline care and the success of careful biological treatments for most patients rapidly helped me to see that biology was undeniably part of the picture.

George Bellows, Forty-two Kids (1907)

Antipsychotics versus psychosocial care in first-episode psychosis

Aftab: You’ve been involved in trials in first-episode psychosis comparing antipsychotics with cognitive-behavioral therapy–based approaches, along with intensive case management. These kinds of trials are extraordinarily difficult to design, fund, and implement, and often controversial. I’m thinking here of the Francey et al, 2020 paper. What drove that work?

McGorry: Part of the historical background is that Philip May conducted related work back in the 1960s, but the more immediate motivation came from the intense debate about duration of untreated psychosis (DUP).

Tom McGlashan came to Melbourne in the 1990s after we had established our early psychosis program and prodromal clinic. He engaged deeply with what we were doing and subsequently took many of these ideas to Norway and the United States. At the invitation of the pioneering Dr Jan Olav Johannessen in Stavanger he became scientifically involved in the TIPS Study in Scandinavia, which used a quasi-experimental design: in one sector, massive efforts were made to shorten DUP through public education campaigns and mobile early detection teams, while another sector continued with standard care as a naturalistic comparison. They found substantially better outcomes in the early detection sector, sustained even at 10-year follow-up.

At the time, there were prominent skeptics who argued that the correlation between longer DUP and worse outcomes was spurious, that it simply reflected a confound. Their claim was that poor-prognosis patients present later because they have a more insidious onset, greater negative symptoms, and less social support, so DUP doesn’t cause worse outcomes, it merely reflects underlying illness severity. McGlashan believed it was unethical to randomize people to early versus deliberately delayed treatment, which is why he pursued the quasi-experimental approach. But that design ultimately couldn’t definitively settle the causal question. I asked myself: under what conditions could you ethically conduct a randomized study that gets closer to answering the causal question about early intervention?

In our setting, we were seeing many young people with remarkably short DUP, often just weeks, not the months or years typical of chronic first-episode cohorts. That demographic reality led me to a different hypothesis: perhaps some first-episode patients with very recent onset, where psychosis was still in its earliest stages, could recover with intensive psychosocial treatment alone, at least initially. Some of these individuals might be clinically closer to the subthreshold or clinical high-risk population. Potentially more responsive to psychosocial interventions before psychotic processes became deeply entrenched.

At the same time, I was thinking carefully about the relationship between symptom severity and medication necessity. The severity of overt psychosis isn’t necessarily the only guide to whether someone needs medication. Some people with subthreshold presentations might genuinely require antipsychotic medication; conversely, some people who meet full threshold criteria for psychosis might not need it immediately, or might recover adequately without it under the right conditions. The clinical question became: especially with much shorter DUPs, can we identify a subgroup that can do well without antipsychotics, at least as a first-line approach?

That question became the basis for what’s often called the STAGES study. We had to design the trial meticulously and take it through ethics committees with great care. We excluded anyone at high risk: people who were severely disturbed, experiencing intense suffering, actively suicidal, at risk of harm to themselves or others, or otherwise unsafe to delay medication. We focused exclusively on individuals who were relatively early in their psychotic presentation, not in acute crisis, and who could provide genuinely informed consent to postponing antipsychotic treatment.

We drew heavily on the capacity-to-consent framework developed by Paul Appelbaum and colleagues. Some of our FEP patients were clearly able to meet that standard. They could understand the risks and benefits, appreciate their situation, reason about the choice, and communicate a decision. Importantly, some patients were ambivalent about or resistant to medication anyway, so we could approach them honestly: we had an open mind about whether this particular individual needed antipsychotic medication immediately or not, and we were conducting rigorous research to find out.

We randomized approximately 90 patients over about five years. Since our service was seeing many more first-episode cases during that period, the recruited cohort wasn’t at all representative of the entire spectrum. They were deliberately a carefully selected subset who met our safety and consent criteria. But the trial demonstrated clearly that some people who meet diagnostic criteria for first-episode psychosis can improve substantially without antipsychotics, provided they receive intensive, structured psychosocial care.

Now, that finding doesn’t mean medication-free treatment is appropriate for the majority of young people with first-episode psychosis. What it does suggest is that for a small but meaningful minority, less than 10 percent, it should be presented as a legitimate option, with careful monitoring and a plan to introduce medication promptly if needed. The key is informed choice, safety, and very close clinical attention, which currently and sadly very few services are resourced to provide.

Aftab: Were you surprised by the results? And what can be said about their generalizability?

McGorry: I wasn’t surprised that a carefully selected subgroup could do well without antipsychotics under very specific conditions. That was precisely our hypothesis. But generalizability is always the critical question with findings like these, and I think we need to be quite cautious and precise about the implications.

I also think it’s important to be transparent about investigator perspectives and motivations in this area. Without naming names, there are some researchers studying these questions with a rather dogmatic agenda. They want to demonstrate that antipsychotics are mostly harmful and should be avoided whenever possible. That’s not my position at all, and I want to be clear about that. Psychosis is a very serious condition. Medication is needed by and works for most but not all people, and it remains a central pillar of treatment. New and more diverse medications are also urgently needed.

One of the reasons I moved away from antipsychiatry thinking early in my career is that I witnessed firsthand how beneficial antipsychotics can be, especially when used very precisely, at very low doses, well below the threshold where you’re producing severe extrapyramidal side effects or inducing a “neuroleptic” state that robs people of their vitality. This was known in the 1950s as the “neuroleptic threshold” and this “sweet spot” was validated by brilliant imaging studies in the 1990s in Canada and Sweden. I don’t have any ideological bias against these medications. What I do object to is careless, high-dose, and often coercive prescribing practices that prioritize behavioral control over therapeutic benefit. A much gentler and rational way of managing acute distress, agitation and behavioral disturbance is with short-term use of benzodiazepines, which tend to make the patients feel subjectively better, not worse, when in the grip of acute psychosis.

At the same time, we know from longitudinal data that some people experience a single episode and recover completely; others may need medication for a limited period but not indefinitely. I’ve debated this extensively with John Kane, whose work I deeply respect. His position, roughly speaking, is that because the majority of people with first-episode psychosis will require longer-term treatment to prevent relapse, you should treat everyone uniformly according to that statistical reality. It’s a population-level risk management approach.

My view is different. I believe it’s worth making the effort to identify those individuals who don’t need medication at all, or who don’t need it for very long, even if they represent a minority. That’s what genuine personalization means in medicine, tailoring treatment to the individual rather than applying a one-size-fits-all approach based on group averages. The price of not personalizing is unnecessary exposure to metabolic side effects, neurological risks, and the stigma associated with long-term psychiatric medication for a subset of people who could potentially do well with a different therapeutic pathway. We owe it to patients to try to identify who they are. Currently we struggle to identify them prospectively as a subgroup and hence the one size fits all approach which is overall safer but not optimal.

McGorry: I believe it’s worth making the effort to identify those individuals who don’t need medication at all, or who don’t need it for very long, even if they represent a minority.

Aftab: My reading of the DUP literature is that causal inference remains genuinely contested in parts of the academic community. What’s your current assessment? And as psychotherapy and psychosocial interventions are increasingly considered as potential first-line treatments for certain patients, do you think the DUP concept should expand beyond simply “time to first antipsychotic treatment”?

McGorry: Within the early psychosis clinical community, the people actually running services and treating these patients, and I think even in the schizophrenia field generally now I’d say the matter is essentially settled: shortening DUP improves outcomes, at least in the short to medium term, and there’s evidence for longer-term benefits as well.

The researchers who continue to dispute this tend to rely on a relatively narrow set of arguments and studies that, in my view, have significant methodological limitations or don’t adequately account for the full picture. But here’s what I think is crucial: even if you completely set aside the “brain toxicity” hypothesis, the idea that untreated psychosis causes progressive biological injury or neurotoxicity, you don’t need that mechanism to understand why leaving a young person actively psychotic for months or years is profoundly damaging.

Common sense and clinical reality are important here. Who would really like to argue that it is a good idea to advocate for treatment delay once psychosis is established and sustained? That is what we call in Australia “the pub test”—what would the average reasonable person think? If someone is experiencing florid psychosis throughout late adolescence or early adulthood, those critical developmental years, they lose friends, drop out of school or university, withdraw from social networks, fail to achieve key developmental milestones like completing education, forming intimate relationships, or establishing vocational identity. That accumulating social and psychological fallout makes recovery substantially harder, regardless of eventual medication response. Even if you eventually achieve good symptomatic control, the task of rebuilding a life, reestablishing social connections, and catching up developmentally becomes exponentially more difficult the longer psychosis has been allowed to run unchecked.

Research examining the duration of active psychosis after the first episode, looking at relapse duration or time spent symptomatic in longitudinal cohorts, is also broadly consistent with the idea that prolonged time in an actively psychotic state is not benign. It has cumulative costs.

So I genuinely don’t think anyone can be arguing for delaying treatment once psychosis is clearly present and causing distress or functional impairment. Psychiatry has too many fruitless ideological debates, and this strikes me as one of them. The preponderance of evidence, both clinical and empirical, points in one direction.

Now, you raise a genuinely interesting conceptual question: should effective psychosocial treatments count as “treatment” within a DUP framework, rather than defining DUP narrowly as time to first antipsychotic? In principle, yes, absolutely, if those interventions demonstrably reduce distress, diminish symptom intensity, preserve or restore functioning, and prevent the cascading social and developmental losses we’ve been discussing.

McGorry: I genuinely don’t think anyone can be arguing for delaying treatment once psychosis is clearly present and causing distress or functional impairment. Psychiatry has too many fruitless ideological debates, and this strikes me as one of them.

Critical Psychiatry

Aftab: You’ve described being influenced by the antipsychiatry movement in the 1960s and 1970s. Over the last two decades, there’s been a resurgence of critical perspectives on psychiatry and a revival of arguments originally made by figures like Thomas Szasz and R.D. Laing. How do you relate intellectually to the earlier antipsychiatry wave versus the contemporary one? What frustrates you about current critiques of psychiatry?

McGorry: I think part of what’s happening reflects broader polarization in society, not just in psychiatry but across many domains. You get increasingly entrenched thesis and antithesis positions, and hopefully something more nuanced and workable eventually emerges from the dialectic. I’ve appreciated the work you’ve done in moving the debates towards a productive synthesis.

I genuinely understand what many critics are reacting to. The concerns that originally motivated me to practice psychiatry from a human rights and humanitarian position were entirely real: patients getting a raw deal, experiencing iatrogenic harm, systematically dehumanizing care in many settings. Those realities resonated very strongly with me then, and they still do. When critics point to coercion, to over-medication, to the dismissal of patients’ subjective experiences, to conflicts of interest with the pharmaceutical industry, those are legitimate concerns that deserve serious engagement.

But I also think that some prominent critics of psychiatry today have become deeply ideological and dogmatic. They will reject this characterization as unfair, but if you look carefully at what they are doing, they frequently start with a predetermined conclusion and then selectively marshal confirming evidence while dismissing or ignoring contrary data. That’s not genuine knowledge-seeking or scientific inquiry. There’s a lack of epistemic humility, a reluctance to genuinely engage with complexity or acknowledge trade-offs.

What these critics often deny or minimize is another set of empirical realities: when antipsychotic medications are used carefully and judiciously, they help people enormously. They prevent suffering, enable recovery, save lives. Without these medications, we would be in a vastly worse situation. I’ve seen that with my own eyes over decades of clinical work. Now, are there valid criticisms of pharmaceutical industry influence? Absolutely. Are there serious problems with careless prescribing practices, with polypharmacy, with using medication primarily for behavioral control? Without question. But acknowledging those real problems doesn’t justify a blunt dismissal of antipsychotics itself.

I don’t have a black-and-white stance on any of this. Frankly, I don’t have a black-and-white stance toward biologically reductionistic psychiatrists either, some of them are far too one-dimensional in the opposite direction. Psychiatry has lurched historically from psychoanalytic reductionism to biological reductionism, and what we desperately need is to find our way back to something genuinely integrative and biopsychosocial, while still demanding real scientific and therapeutic progress.

We need substantially better psychological treatments. CBT for psychosis has shown benefits, but it’s honestly quite limited in what it can do. We need more innovation, perhaps through immersive technologies like virtual reality, through newer forms of psychotherapy, through better integration of peer support and social interventions. But we also desperately need advances in biological treatment. There’s a subset of first-episode patients, roughly 20 to 30 percent, who don’t respond adequately to low-dose dopamine blockade. What do you do for them? We don’t have good answers. One answer we do have is clozapine but the mental health systems around the world are very poor at delivering early access to this life changing and extending medication once treatment non-responsiveness to standard dopamine antagonists is clearly established.

It’s scientifically plausible that glutamatergic agents could help this population, but drug development in this area has been painfully sluggish and frustrating. When you look at what’s happened in oncology over the same time period, the explosion of targeted therapies, immunotherapies, the translation of basic science into clinical benefit, the contrast is striking and frankly dispiriting.

There simply isn’t enough momentum or investment behind the discovery of novel psychiatric agents. We’re also hampered by an excessively rigid regulatory and scientific culture at places like the NIMH in US that insists you must definitively demonstrate the mechanism of action before you can explore clinical efficacy. That stifles innovation. We need more space for intelligently guided empiricism and serendipitous discovery alongside mechanistic neuroscience, the way psychiatric pharmacology actually progressed in its most productive eras.

McGorry: Some prominent critics of psychiatry today have become deeply ideological and dogmatic. There’s a lack of epistemic humility, a reluctance to genuinely engage with complexity or acknowledge trade-offs.

Clinical staging, transdiagnostic thinking, and “new diagnosis”

Aftab: You’ve worked on clinical staging concepts for decades. How has your thinking about staging developed over that time?

McGorry: We arrived at staging through prospective longitudinal work attempting to define and characterize a subthreshold psychotic state, what eventually became known as the clinical high-risk or ultra-high-risk for psychosis concept. In our early cohort studies in the mid-1990s, transition rates to full-threshold psychosis were quite high, 30 to 40 percent within relatively short follow-up periods.

The fundamental principle underlying staging was this: if you’re going to intervene early, before clear diagnostic thresholds are crossed, you must prioritize safety above all else. “Primum non nocere.” That’s precisely why we argued that psychosocial interventions should be prominent at early stages, while pharmacological approaches should be more conservative and reserved for those who truly need them or who progress despite initial psychosocial treatment.

Over time, as we accumulated longer follow-up data and larger cohorts, the picture became more complex and interesting. The ultra-high-risk group didn’t just transition to psychosis or remain stable, many developed other comorbid or single syndromal outcomes entirely. Some developed severe mood disorders, some anxiety disorders, some substance use problems. That empirical reality pushed us strongly toward a more transdiagnostic conceptual framework. We were clearly dealing with evolving, fluid syndromal pictures in young people, not with discrete disease entities that simply declare themselves over time.

We’ve worked hard to create research infrastructure for studying these developmental pathways, the ebbs and flows of psychopathology as it emerges and transforms in young people. But I have to be honest: I don’t think we’ve pinned down the staging model well enough yet. We need many more high-quality longitudinal studies that carefully track how symptoms cluster, how comorbidity emerges and changes, how functional trajectories diverge… the kind of intensive, long-term cohort work that’s expensive and difficult but absolutely essential. It is already a very good framework for the conduct of clinical trials however.

This all loops back to your fundamental question about ontology: what exactly are we dealing with in psychopathology? If these are syndromes with complex, shifting, multifactorial causal mixtures rather than natural kinds, how do we intelligently connect our classification systems to treatment decisions and biological, psychological, and social interventions?

Aftab: Do you see potential synergy between staging approaches and dimensional classification systems like HiTOP? And what about complex systems approaches, network analysis and related methods?

McGorry: We’ve certainly tried to bring these different theoretical worlds into productive conversation. Just before the pandemic, we organized an international workshop in Amsterdam on what we ambitiously called “new diagnosis.” We invited researchers working on HiTOP, people involved with RDoC, those developing staging models, and others thinking innovatively about psychiatric classification. The goal was to see whether we could build some kind of shared conceptual framework or at least identify areas of convergence. That work was eventually written up and published in Psychological Medicine recently.

To be candid about my own personal view, I think staging has the best chance of achieving real clinical utility in routine practice. What staging does is add an orthogonal temporal dimension, a developmental or longitudinal axis, over the dimensional syndromal space. That combination potentially gives you something useful for clinicians and patients.

With HiTOP, the dimensional structure often makes excellent conceptual sense and fits the psychometric data beautifully. But the critical clinical questions are: how do these dimensions evolve over time in individual patients? How do they guide specific treatment decisions? Network analysis and related complex systems approaches may ultimately capture this kind of dynamic evolution more effectively… if we can generate sufficiently high-quality longitudinal data to feed into these models. Technically, this work is above my pay grade, but conceptually I think that’s where the field needs to move.

There’s also an important practical reality we sometimes overlook: treatment is often inherently transdiagnostic anyway, especially psychosocial interventions. CBT, family work, supported employment, substance use interventions. These don’t neatly map onto diagnostic categories. Even pharmacological treatments are trans-syndromal in their effects but are forced by the FDA and pharma into a procrustean bed of artificially discrete entities.

McGorry: To be candid about my own personal view, I think staging has the best chance of achieving real clinical utility in routine practice. What staging does is add an orthogonal temporal dimension, a developmental or longitudinal axis, over the dimensional syndromal space.

DSM and attenuated psychosis syndrome

Aftab: When DSM-5 was being developed, there was intense controversy around including attenuated psychosis syndrome or clinical high risk for psychosis in the manual. Allen Frances was particularly vocal in his opposition. How do you reflect on that debate now, more than a decade later?

McGorry: [Laughs] It’s quite tricky, actually. I find myself agreeing with Allen Frances on many things these days. But at the time, I think he misunderstood what we were doing and approached the debate in an unnecessarily heated manner.

His argument was that attenuated psychosis syndrome represented boundary expansion of the kind that had legitimately troubled people with ADHD, autism or pediatric bipolar disorder, that we were pathologizing normal adolescent experiences and creating a new diagnostic category to label people who didn’t genuinely need psychiatric intervention.

But that characterization was not aligned with empirical realities. The data already clearly showed that these young people were actively help-seeking. They weren’t being identified through population screening or dragged into services by worried parents. They had significant functional impairment, elevated rates of distress, and concerning levels of suicidality. They clearly warranted clinical attention and support. The genuine question, the one we were actually trying to answer through rigorous research, was what constituted the right care for this population.

We consistently used staged, sequential or adaptive designs with safety as the paramount concern. The default approach was simpler supportive needs-based psychosocial interventions first, with intensive case management, CBT, family work, with medications held in reserve unless and until someone’s clinical state deteriorated or they developed full-threshold psychosis. Now, we did also explore low-dose antipsychotic treatment as one option in some studies, always with informed consent, because that can be a legitimate clinical choice depending on symptom severity, level of distress, suicide risk, and patient preference. But the entire enterprise got caught up in a politicized battleground over DSM-5 revision politics and broader cultural anxieties about diagnostic inflation and pharmaceutical company influence.

Aftab: Do you think something like attenuated psychosis syndrome is ready to be formally included in diagnostic classification systems now?

McGorry: I think some form of diagnostic recognition is necessary to make evidence-based treatment accessible and fundable, but we have to be careful about what this syndrome actually represents. It’s pluripotential. It’s a phenotype that clearly warrants clinical care, but the syndromal and developmental outcomes are genuinely heterogenous and extend beyond psychosis alone, even though there is a strong valence for subsequent psychotic illness.

Current data suggest that transition to full-threshold psychosis occurs in roughly 20 percent of these individuals, possibly somewhat higher with extended follow-up, though rates vary considerably depending on how rigorously your sample is ascertained and where you’re recruiting from. But transition to psychosis is only one outcome trajectory among several.

Attenuated psychosis or clinical high risk syndrome is also associated with progression to mood disorders, to substance use disorders, and to persistent subthreshold attenuated symptoms that still cause considerable distress and impairment. In longitudinal studies, only a small minority, perhaps less than 10%, failed to meet criteria for any DSM syndrome over extended periods of time.

So I think it’s most accurately conceptualized as a kind of gateway syndrome, one important pathway into serious mental illness, but not the only one, and not deterministically leading to any single outcome. It identifies a population of young people at elevated risk who need much more than minimal “watchful waiting” but who shouldn’t automatically receive specific antipsychotic pharmacological treatment, especially as first line.

The question of nosological placement is an important one. Placing it exclusively within the psychotic disorders section would be misleading, because it’s emphatically not simply “early schizophrenia” or “prodromal psychosis” in a straightforward sense. But it is a recognizable, clinically meaningful state with characteristic features and predictable risk profiles. It is only in retrospect in the minority who do develop psychotic illness that the term “prodrome” can be applied.

McGorry: Clinical High Risk for Psychosis syndrome is most accurately conceptualized as a kind of gateway syndrome, one important pathway into serious mental illness, but not the only one, and not deterministically leading to any single outcome.

Youth Mental Health

Aftab: You’ve also been a champion of youth mental health system reform and policy initiatives, including the Lancet Commission on youth mental health. Say more about that.

McGorry: After about 10 to 15 years focused intensively on early psychosis, it became increasingly obvious that the issue for mental health reform was much broader: the vast majority of serious mental disorders have their onset in adolescence and young adulthood. We needed to shift from an early psychosis paradigm to a youth mental health paradigm.

The traditional split in mental health services, child and adolescent psychiatry ending at 18, then adult psychiatry beginning at 18, is very problematic developmentally and clinically. What we actually need is a robust, coherent system of care that spans from puberty through the mid-20s, capturing people during the peak incidence period for most mental disorders while being developmentally attuned to the distinctive challenges of the transition to adult roles and responsibilities.

Historically, this has been the weakest, most neglected part of the entire health system. Child and adolescent services were never adequately resourced or conceptually equipped to handle the serious disorders that emerge in late adolescence. Adult services, meanwhile, often only became seriously engaged once chronicity had already developed, essentially waiting for people to fail before offering intensive support. Young people fell through the gap.

So we developed Orygen as a research institute specifically focused on youth mental health, and we advocated persistently with federal and state governments to build a national system of care. In Australia, that eventually resulted in Headspace, a network of youth-friendly primary mental health centers, along with more specialized early intervention services as a “second tier” for those who need more intensive treatment. Other countries have developed variations on this model in different ways, and we tried to synthesize the broader vision and evidence base in the Lancet Commission work.

During the same decades we’ve been building this infrastructure and advocating for reform, the underlying problem appears to have been getting significantly worse. The Lancet Commission devoted considerable attention to assessing whether there’s genuinely a youth mental health crisis unfolding globally, or whether this perception reflects reporting artifacts, increased awareness, or diagnostic boundary expansion.

The data we examined looked disturbingly robust in certain settings, particularly in Australia and Denmark, where high-quality population-level surveillance has been maintained over time. The picture may not be uniform everywhere, and in low- and middle-income countries we often lack adequate epidemiological infrastructure to know with confidence. But where we have reliable data, the trends are deeply concerning.

There’s certainly a “prevalence inflation” phenomenon we have to reckon with, broadening diagnostic criteria over successive DSM editions, self-diagnosis movements facilitated by social media, increased awareness leading to more help-seeking. Some proportion of people who now believe they have a mental disorder genuinely may not meet rigorous diagnostic criteria or have a need for mental health care. That complicates interpretation considerably.

But if you look at hard data from well-designed community epidemiological surveys using consistent methodology, Australia appears to have experienced something on the order of a 50 percent increase in the prevalence of mental disorders among young people over roughly two decades, with the increase particularly concentrated in common mental disorders like anxiety and depression. That’s not an artifact, something real is happening. If prevalence genuinely is rising, you’re forced to think seriously about prevention and about the socioeconomic determinants and upstream causes of mental disorder. It becomes a public health and social policy problem.

Some governments have been eager to pin the youth mental health crisis primarily on social media and smartphone use. There are prominent public intellectuals like Jonathan Haidt pushing that narrative very forcefully. And look, I think it’s plausible that these technologies play some role. But even if that’s a contributing factor, I think broader socioeconomic and cultural megatrends are probably more powerful and more fundamental. Weakening social cohesion and community ties, worsening economic prospects for young people relative to previous generations, housing affordability crises in many developed countries, precarious employment, educational pressure and competition, climate anxiety and existential uncertainty about the future. They create chronic stress, undermine the conditions for healthy development, and erode protective factors.

Mental health is inseparable from social and economic policy. Decisions about education, employment, housing, inequality, and social infrastructure have mental health consequences that dwarf what we can accomplish through downstream clinical intervention alone. I think psychiatry has a genuine responsibility to understand and speak to these broader societal changes, not just to respond clinically after the psychological damage has already been done. We need to be engaging in prevention and advocacy upstream, not just mopping up downstream.

Aftab: Thank you!

McGorry: Thanks Awais, and I want to acknowledge that much of the research I have talked about has been the result of a team effort, with great colleagues at Orygen as well as internationally. It would not have been possible without them and they share the credit.

---

This Q&A is part of a series featuring interviews and discussions intended to foster a re-examination of philosophical and scientific debates in the psy-sciences. See prior interviews here.

---

If you are a regular reader of this newsletter, and it deepens your intellectual life, consider a paid subscription to support this effort.