Article Text

Download PDFPDF

Ultra-high-risk paradigm: lessons learnt and new directions
  1. Patrick D McGorry1,2,
  2. Cristina Mei1,2
  1. 1 Orygen, The National Centre of Excellence in Youth Mental Health, Parkville, Victoria, Australia
  2. 2 Centre for Youth Mental Health, University of Melbourne, Parkville, Victoria, Australia
  1. Correspondence to Professor Patrick D McGorry, Orygen, The National Centre of Excellence in Youth Mental Health, Parkville, VIC 3052, Australia; pat.mcgorry{at}


Within the embryonic early psychosis field in the early 1990s, the conceptualisation and definition of an at-risk or ultra-high-risk (UHR) mental state for psychosis was a breakthrough which transformed the clinical and research landscape in psychiatry. Twenty-five years later, we have a new evidence base that has illuminated the neurobiology of the onset phase of psychotic disorder, delivered Cochrane level 1 evidence showing that the onset of full-threshold sustained psychotic disorder can be at least delayed, and is paving the way to a new generation of transdiagnostic research. Here, we document the contribution of the UHR approach to understanding the underlying mechanisms of psychosis onset as well as the long-term outcomes. Particularly, we highlight that psychosis onset can be delayed in those meeting UHR criteria and that these criteria have a higher valence for subsequent psychotic disorders and some valence for persistent non-psychotic syndromes. Critiques have helped to identify some of the limitations of this paradigm, which are acknowledged. These include evidence that psychotic disorders can emerge more acutely and from other, as yet undefined, precursor states. Rather than defending, or alternatively questioning the value of, the UHR approach, we propose a broader, transdiagnostic staging model that is consistent with the pluripotent and variably comorbid trajectories for mental disorders. This approach moves beyond psychosis to capture a wider range of subthreshold symptoms and full-threshold disorders, thus enhancing prediction for the emergence and progression of a range of mental disorders, as well as providing new avenues for early intervention and prevention.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The concept of an at-risk or ultra-high-risk (UHR) for psychosis was developed as a prospective version of the older concept of prodrome. It had been understood for decades that psychotic disorders such as schizophrenia did not emerge fully fledged, but typically emerged on a background of milder symptoms. Diagnostic threshold was achieved only once a certain level of intensity, severity and/or pathognomonic symptom status was reached. The concept of prodrome was a retrospective one and once the diagnosis was made it seemed to have arisen inevitably from the prodromal stage. This fitted well with the deterministic mindset that spawned the schizophrenia concept, namely that onset and deterioration are inevitable. The at-risk mental state concept was a major advance because it meant that progression to full diagnostic clarity and threshold, while much more likely, was not a given. Similarly, ‘first episode psychosis’ (FEP) is also a heterogeneous way station along the course of presumed increasing diagnostic clarity and stability. This mindset offered genuine hope for remission and recovery and accorded well with the evidence.

The heuristic value of this breakthrough cannot be underestimated. First, it opened the door to a dynamic clinical staging model, which now can be applied transdiagnostically, in which clarity, stability and specificity increases with time in the subset of cases in which symptoms persist or recur. Second, it allows for remission of symptoms at an early stage; indeed the clinical staging model allows for this at all stages. Third, it has been revealed that, while the valence for schizophrenia and other psychotic disorders is high, it is not exclusive. This challenges conventional siloed approaches to psychiatric diagnosis (Diagnostic and Statistical Manual of Mental Disorders, International Classification of Diseases), and provides a way forward unlike some other proposed alternatives and critiques.1 2

The at-risk mental state or UHR paradigm has catalysed a new wave of research and service reform over the last two decades, leading to improved knowledge of the mechanisms underlying the emergence of psychotic disorders and subsequent outcomes. Despite this progress, it has attracted controversy. This has arisen from diverse sources.3 Such critiques have helped the paradigm to evolve, diversify and potentially to strengthen.

Challenges and confusions: UHR and need for care

Early critics were quick to assert that psychotic symptoms were not necessarily always associated with a need for care. This assertion was based on population-based studies which seemed to reveal that the psychosis phenotype was more widely distributed in the general population.4 5 Conversely, all samples of UHR patients with even subthreshold psychotic symptoms were functionally impaired,6 distressed and seeking help. They clearly were in need of care even prior to reaching full diagnostic threshold. Moreover, other population-based studies showed that the presence of psychotic-like experiences and subthreshold psychotic symptoms predicted a poorer outcome and other morbidity, notably suicidal behaviour.7 The conclusion from this research appears to be that psychotic symptoms, even subthreshold ones, if they occur together with distress and other symptoms of anxiety and depression, and if they persist,8 are a predictor of poorer outcomes. Follow-up studies of UHR cohorts confirm this, namely that in addition to the transition to psychotic disorders, persistence of the subthreshold symptoms, of other syndromes such as depression and anxiety, and poor functional outcome, are all more likely.

While the at-risk mental state concept was originally developed within the schizophrenia and early psychosis research fields, it clearly has a wider significance, and raises the notion what really is a psychotic disorder. Is it possible to have psychotic symptoms which are distressing and disabling without being regarded as psychotic? It is perfectly understandable that antipathy to the pessimism and stigma of the schizophrenia concept has driven some of the critique of the at-risk/UHR concept, including that of van Os and Guloksuz.3 Ironically, however, the at-risk/UHR concept has shown the nexus is finite yet fluid. Subthreshold or early symptoms of psychosis are often non-specific and do not necessarily lead to schizophrenia, with nearly half of patients with FEP failing to meet diagnostic criteria for schizophrenia or schizophreniform disorder.9 The potential to progress to a full range of psychotic disorders, including psychotic mood disorder, has long been recognised by the UHR approach, with its target being psychosis and not merely schizophrenia.

Another set of critiques deliberately exploited the presence of the word ‘risk’ in its title to misrepresent the UHR paradigm as overtreating asymptomatic people who might be harmed as a result. Now in general medicine, even asymptomatic people with hypertension, prediabetes and those at risk for stroke or ischaemic heart disease are offered treatment, including medication, based on risk–benefit considerations. However, the UHR phenotype is a highly symptomatic and disabling stage of illness not asymptomatic or presymptomatic, as such critics well knew, but chose to ignore and misrepresent.10 Sequentially, staged treatment does benefit UHR patients greatly, and this means medications being offered only as second or third line. So while overtreatment (which can occur in an off-label and non-evidence-based manner in some clinical settings) can be inhibited with a staged approach, we know all too well that untreated, they do poorly as a group.

Transition and its limitations

Transitioning to a full-threshold psychotic disorder is not seen as inevitable, with just over a third receiving such a diagnosis after 3 years.11 Although this could be interpreted as a somewhat low rate, those meeting criteria for UHR are at a greatly elevated risk of psychosis compared with the general population. Transition can be defined objectively, and while there are limitations, it is a clinically meaningful concept, marking the need for a key change in treatment (eg, the point at which the introduction of antipsychotic medication is agreed to be appropriate). The statistically significant differences on clinical and neurobiological markers between those who do and do not transition provides partial validation for the concept of transition.12 13 While these markers have not been able to define a precise point of transition or add to clinical utility, it is increasingly possible to predict those UHR patients who are at an increased risk of transition.14

While the high transition rates in the first generation of studies validated the paradigm, rates of transitioning to psychosis have been declining.15 This is likely due to sampling less enriched populations of referrals with a lower risk of transition to sustained psychosis. Another concern is that even when UHR clinics exist in parallel with FEP programmes, a low percentage of FEP cases (4.1%–12.5%) enter via this channel.16 17 While this may indicate that more effective treatment approaches delivered earlier by UHR services could have averted a proportion of FEP, it seems probable that the entry channel is too narrow or that alternative and unmapped phenotypic pathways leading to FEP coexist. This means that specialised UHR services that focus exclusively on early psychotic symptoms may not capture all cases who transition to psychosis as alternative pathways leading to a psychotic disorder have emerged. For instance, they have been shown to emerge following a period of non-psychotic at-risk states,18 with approximately one-third of patients with FEP not experiencing subthreshold psychotic symptoms.19 This suggests that the wider enhanced primary care youth mental health model of headspace is likely to capture a fuller range of pathways and be more inclusive.20 Headspace, as described elsewhere,21 is an enhanced primary care service that offers a broad entry point to care for young people aged 12–25 years that is backed by specialist services (including UHR) for those with more severe or complex presentations who typically require more intensive and specialised care. The service is not triaged based and is available to any young person requiring mental healthcare. This is a key strength of the model as it is able to effectively manage a high volume of cases through its broad-based entry, which is not possible through narrow approaches such as UHR clinics and traditional child and adolescent mental health services.22 The number of headspace clients with subthreshold psychotic symptoms ranges from 23% (based on preliminary findings from a large trial of UHR cases)23 to 38%.24 It is likely that not all of these cases would meet the full UHR criteria, however, access to care is clearly enhanced for earlier stage cases. This indicates that a substantial number (though still an unknown proportion) of UHR cases are now able to gain access to care. We will study whether that leads to a higher proportion of subsequent FEPs being identified prior to transition and even whether transition rates are affected by such access and care. While the UHR criteria do indeed have a higher valence for psychotic disorders,25 they also have, to a lesser extent, some valence for non-psychotic syndromes. Most UHR cases do not progress to FEP and may either manifest at baseline or progress to (comorbid) non-psychotic disorders (eg, mood, anxiety and/or substance use disorders).26

Nevertheless, UHR services are of clear value. They have demonstrable cost-effectiveness27 and even when an individual does progress to an FEP, prior engagement with UHR services can delay transition,28 reduce trauma and acute risk, and improve clinical outcomes (eg, due to shorter duration of untreated psychosis and reduced risk of hospital admission).29 These outcomes can be achieved through interventions that pose fewer risks as demonstrated by trials comparing low-dose risperidone and cognitive–behavioural therapy.30 Although these trials have attracted criticism from commentators who have blurred the distinction between clinical practice and ethically approved research, they have informed the development of evidence-based guidelines that have recommended the withholding of antipsychotic medication as a first-line treatment for UHR cases. Yet transition, while important, is increasingly regarded as not the only or even the most crucial target for intervention, and that improved functional outcomes, which are certainly attainable during the UHR state, and even post-transition if this occurs,31 should be a key focus23 across the various trajectories.32

Moving towards a transdiagnostic at-risk mental state

In light of the modest transition rates to psychosis and the heterogeneous outcomes associated with the UHR criteria, a more definitive and ambitious strategy for predicting psychosis and other mental disorders is needed.33 However, rather than abandoning the at-risk/UHR concept, which has created a highly successful research paradigm, the lessons learnt from this concept can be used to develop a wider at-risk mental state approach (ie, one that targets a broader range of mental disorders and outcomes, not solely psychosis). The advantage of this technique is that it could provide the statistical power needed to accurately predict low incidence mental disorders (eg, schizophrenia, bipolar disorder),34 potentially overcoming the prevention paradox that has been linked to the UHR strategy.3 A broader focus that goes beyond the confines of psychosis risk identification and prevention aligns with the proposition that pathways to mental disorders are pluripotent and transdiagnostic.

Such an approach has been developed by our team, known as the clinical high-risk mental state (CHARMS).35 CHARMS has been adapted from the UHR concept and is guided by the clinical staging model, which classifies mental disorders along a continuum of stage 0 (asymptomatic) to stage 4 (treatment resistant) and links each stage to intervention strategies and biomarkers that denote specific stages and risk of progression.36 Subthreshold states (stage 1b) considered by CHARMS include attenuated psychotic symptoms as well as subthreshold bipolar and borderline personality symptoms, and mild-moderate depression. Even the use of the terms ‘risk’ and ‘subthreshold’ might turn out to be unhelpful since this first stage justifies a need for care and risk for progression exists in every stage. The target outcome has similarly been broadened to any exit syndrome (eg, psychosis, bipolar disorder, depressive disorders, borderline personality disorder). Trait vulnerability criteria of the traditional UHR have also been expanded to include family history of serious mental disorder in addition to functional decline. An advantage of the CHARMS approach is that research into specific at-risk states and single disorders (eg, UHR and psychosis) can still occur within a broad, transdiagnostic approach. Thus, the latter does not need to occur at the expense of a disorder-specific approach, with both strategies complementing each other. This means we can continue to sharpen prediction for psychosis and schizophrenia and other classical syndromes using neurobiological markers, machine learning and other prediction techniques.


The at-risk/UHR approach has provided greater insight into the early stages of psychosis, including the overlapping subthreshold symptoms occurring across multiple mental disorders. The lessons learnt from this approach have been used to guide the development of a broad set of transdiagnostic at-risk criteria that capture the pluripotent trajectories of mental disorders. It is anticipated that this wider focus will enhance the ability to predict the onset of mental disorders, strengthen the clinical staging model and broaden preventive opportunities. It is also synergistic with and supported by major redesign and reform in early intervention and youth mental health services.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.


  • Contributors PDM and CM drafted and revised the manuscript.

  • Funding PDM is supported by an NHMRC Senior Principal Research Fellowship (1060996).

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.