Background
Despite research advances, schizophrenia remains one of the most debilitating chronic illnesses in medicine (Hegarty etal. Reference Hegarty, Baldessarini, Tohen, Waternaux and Oepen1994). Although course and severity vary, the illness is generally characterised by recurrent episodes, functional deterioration, residual negative symptoms and enduring cognitive impairment (Larsen etal. Reference Larsen, Johannssem and Opjordsmoen1998; Clarke etal. Reference Clarke, Whitty, Browne, McTigue, Kamali, Gervin, Kinsella, Waddington, Larkin and O'Callaghan2006; Crumlish etal. Reference Crumlish, Whitty, Clarke, Browne, Kamali and Gervin2009; Tandon etal. Reference Tandon, Nasrallah and Keshavan2009). Findings that there may be greater treatment responsiveness in the first episode of psychosis (Bertelsen etal. Reference Bertelsen, Jeppesen, Petersen, Thorup, Øhlenschlaeger, le Quach, Christensen, Krarup, Jørgensen and Nordentoft2008), that intervention early in the first episode may be associated with better prognostic outcomes (Henry etal. Reference Henry, Amminger, Harris, Yuen, Harrigan and Prosser2010), reports of progressive grey matter decline during the early illness phases and evidence that most people who develop a sustained psychotic disorder experience a significant period of sub-threshold symptoms, distress and functional decline long before they become frankly psychotic have impelled the development of early recognition programmes around the world (McGlashan & Johannessen, Reference McGlashan and Johannessen1996; Yung & McGorry, Reference Yung and McGorry1996). This area of research is nascent and not without controversy. In particular, concerns have been raised about the ethics of intervening in a population that may or may not go on to develop mental illness.
Aims
This article aims to review the evidence to date on the ‘ultra-high risk’ (UHR) for psychosis population. This article is not a systematic review; however, it aims to:
• Review the underlying principles of early identification and intervention during the pre-psychotic/prodromal phase.
• Summarise the naturalistic follow-up studies conducted to date in this ‘at-risk’ population.
• Discuss the identified clinical risk factors for transition to psychosis.
• Summarise the interventional studies both psychological and pharmacological completed to date.
• Briefly review the controversy around the proposed inclusion of the psychosis risk syndrome (PRS) in DSM 5.
Methods
Electronic databases EmBase, MedLine and PsychInfo were searched using the key words ultra-high risk/at-risk mental state/risk syndrome/pre-psychotic/prodrome/prodromal and psychosis/schizophrenia. References of all identified studies were searched for further relevant studies.
Results
What is UHR?
Early intervention for psychosis services aims to detect emergent symptoms, reduce the duration of untreated psychosis and improve access to effective treatments. Early intervention for psychosis services differs from standard care in two ways: early detection and phase-specific treatment (phase-specific treatment is a pharmacological, psychological or social intervention adapted or developed for use specifically with people at an early stage of the illness) (Marshall & Rathbone, Reference Marshall and Rathbone2011). Early detection typically involves the early identification of people who are already psychotic, but who have not yet received adequate treatment. However, some early intervention services also endeavour to detect using standardised instruments people who display prodromal symptoms and as such are considered ‘at risk’ for developing psychosis (Schaffner & McGorry, Reference Schaffner and McGorry2001; Wyatt & Henter, Reference Wyatt and Henter2001). These ‘at-risk’ people have not met the criteria for a psychotic disorder and their identification and any intervention offered aims to prevent or ameliorate an emerging psychotic illness.
The ‘UHR’ group, also referred to as the ‘At Risk Mental State’ (ARMS) group or the ‘PRS’, are defined as a group of help-seeking individuals identified using reliable measurement tools as being at high risk of developing a psychotic illness.
Some authors have raised concerns about the validity of the UHR concept and the potential harm associated with intervention during the ‘prodromal phase’. Sub-threshold psychotic symptoms have a reported prevalence in the general population of around 5% (van Os etal. Reference van Os, Linscott, Myin-Germeys, Delespaul and Krabbendam2009), with higher rates of 7.5–23% being reported in children and adolescents (Kelleher etal. Reference Kelleher, Keeley, Corcoran, Lynch, Fitzpatrick, Devlin, Molloy, Roddy, Clarke, Harley, Arseneault, Wasserman, Carli, Sarchiapone, Hoven, Wasserman and Cannon2012a, Reference Kelleher, Connor, Clarke, Devlin, Harley and Cannon2012b). There is emerging evidence that sub-threshold psychotic symptoms may exist on a continuum with psychosis. This continuity is suggested by evidence that known risk factors for the development of schizophrenia such as urbanicity, social disadvantage and cannabis use are also associated with risk of sub-clinical psychotic symptoms in the general population transitioning to clinical psychosis (Binbay etal. Reference Binbay, Drukker, Elbi, Tanık, Özkinay, Onay, Zaǧli, van Os and Alptekin2011; Dominguez etal. Reference Dominguez, Wichers, Lieb, Wittchen and van Os2011). Some authors have expressed concern about the risk of stigmatisation and psychological distress, especially for those false positives identified who were never going to develop a psychotic illness (Raven etal. Reference Raven, Stuart and Jureidini2012). A false-positive diagnosis of ‘UHR’ for psychosis could create unnecessary anxiety and demoralisation about prognosis. An even more serious risk is the potential for unnecessary use in this population of interventions, especially antipsychotics, which have potentially serious side-effects, including weight gain, diabetes, metabolic syndrome and neurological symptoms (Foley & Morley, Reference Foley and Morley2011). Most involved in UHR research acknowledge the potential jeopardy associated with research and treatment of the UHR group and propose that it is only with high-quality research that clarity can be realised on these issues (Yung & Nelson, Reference Yung and Nelson2011). These concerns will be discussed in some more detail in the section on the proposed inclusion of a UHR syndrome in DSM 5.
Prodromal symptoms of psychosis have long been recognised (Meares, Reference Meares1959; Bowers, Reference Bowers1965). The concept of being able to prevent the onset of schizophrenia and other psychotic illness by detecting and intervening at the prodromal phase has long been sought (Sullivan, Reference Sullivan1927). Initial researchers in this area focused with little success on the early identification of florid psychotic symptoms. It was in the 1980s when a series of retrospective studies redirected attention to the fact that often more subtle, less florid signs of psychotic illness are present for days up to years before the onset of the first episode of frank psychotic illness (Reference Hafner, Maurer, Loffler, Bustamante, an der Heiden, Riecher-Rossler and NowotnyHafner etal. 1995). This change in focus and the emerging data in the area facilitated the development of reliable measurement tools for the recognition of young people at ‘UHR’ of developing a psychotic illness.
Current research in this area began in Melbourne, Australia, with the work of Yung, McGorry and colleagues. In 1994, the Personal Assessment and Crisis Evaluation (PACE) clinic was established, aimed at identifying, monitoring and providing care for young help-seeking people described as being at high risk for developing psychosis. Yung and colleagues developed operationalised criteria: the UHR criteria, which aim to overcome the non-specific nature of prodromal symptoms using a ‘close in’ strategy (Philips etal. Reference Philips, Yung and McGorry2000). The features identified comprise a combination of state and trait factors (Box 1).
Box 1 Ultra-High Risk Criteria
PACE Clinic, Melbourne
Assessment tool: Comprehensive Assessment of At-Risk Mental States (CAARMS)
Person is referred for help to a clinical service and meets criteria for one or more of the following groups:
State-based criteria
a. Attenuated Psychotic Symptom syndrome (APS):
• One or more sub-thresholds, attenuated positive psychotic symptoms, e.g., unusual thought content/delusional ideas, suspiciousness/persecutory ideas, grandiosity, perceptual abnormalities/distortions, conceptual disorganisation.
• Held with either sub-threshold frequency or sub-threshold intensity; present for ⩾1 week within the past year and for ⩽5 weeks.
b. Brief Limited Intermittent Psychotic Symptom syndrome (BLIPS)
• Transient fully psychotic symptoms: symptoms in the realm of delusions, hallucinations, disorganisation.
• Duration of the episode < 1 week.
• Spontaneous remission; symptoms occurred within 1 year but for not longer than 5 years.
Trait-based criteria
c. The Genetic Risk and Deterioration Symptom syndrome (GRD)
• First degree relative with a psychotic disorder
OR
• Personally meeting the DSM-IV criteria for schizotypal personality
AND
• Significant drop in functioning as defined by a GAF (global assessment of functioning) drop of 30% or more for at least 1 month over the past year.
PRIME Clinic, Yale
Assessment tool: Structured Interview for Psychotic Symptoms (SIPS)
Person is referred for help to a clinical service and meets criteria for one or more of the following groups:
State-based criteria
a. Attenuated Positive Symptom Prodromal Syndrome (APSS):
• One or more sub-threshold-positive symptoms, not fully psychotic in intensity: Unusual thought content/delusional ideas, suspiciousness/persecutory ideas, grandiosity, perceptual abnormalities/distortions, conceptual disorganisation.
• Currently present at a frequency of at least once per week in past month, onset or worsening in the past year.
b. Brief Intermittent Psychosis Prodromal Syndrome (BIPS):
• One or more fully psychotic symptoms:
• Present intermittently for at least several minutes/day at least once per month, but less than 1 hour/day, 4 days/week over 1 month.
Trait-based criteria
c. Genetic Risk and Deterioration Prodromal Syndrome (GRD):
• Has a first-degree relative with a psychotic disorder
OR
• Personally meeting criteria for schizotypal personality disorder
AND
• Precipitous decline in functioning rated on the General Assessment of Functioning (GAF) scale as a drop of at least 30% compared with 12 months ago.
PACE: Personal Assessment and Crisis Evaluation Clinic, Melbourne, Australia.
PRIME: Psychosis Prodrome Research Clinic, Yale, Connecticut, USA.
Underlined text indicates differences in SIPS criteria when comparedwith CAARMS.
Using these UHR criteria, Yung and colleagues completed a number of studies to assess the validity of their UHR criteria (Yung etal. Reference Yung, Philips, Yuen, Francey, McFarlane, Hallgren and McGorry2000). These studies found rates of transition to psychosis of 20–40% by 12-month follow-up. These rates are several thousand-fold over the expected incidence rate for first-episode psychosis in the general population. This occurred despite the provision of supportive counselling, case management and antidepressant medication if required. The peak time of risk for transition to psychosis was found to be within 4.5 months of entry to the clinic (Yung etal. Reference Yung, Philips, Yuen and McGorry2004).
The PACE UHR criteria have since been adopted and adapted by a number of other settings around the world. UHR population-specific assessment tools have been developed. These tools now include the Comprehensive Assessment of At-Risk Mental States (CAARMS) (Yung etal. Reference Yung, Yuen, McGorry, Phillips, Kelly and Dell-Olio2005) developed by Yung and colleagues and the Structured Interview for Prodromal Symptoms and the Scale of Prodromal Symptoms (SIPS/SOPS) developed by McGlashan and colleagues in Yale (see Box 1) (Miller etal. Reference Miller, McGlashan, Rosen, Somjee, Markovich, Stein and Woods2002).
The Bonn Scale for Assessment of Basic Symptoms (BSABS) has been proposed to identify an earlier phase of UHR [also known as Early Initial Prodromal State (EIPS)] (Gross etal. Reference Gross, Huber, Klosterkötter and Linz1987). These ‘basic symptoms’ emerge from the German psychiatric literature and include disturbances of self-perception, stress tolerance, thought organisation and social and non-verbal interactions that are subjectively observed and not usually noticed by others (Schultze-Lutter, Reference Schultze-Lutter2009). Klosterkotter etal. (Reference Klosterkotter, Hellmich and Steinmyer2001) reported that 70% of 110 participants who had endorsed one or more items on the BSABS had developed schizophrenia at 9.6-year follow-up. However, other authors have questioned the generalisability of this sample, which was made up of people referred to a tertiary clinic on the grounds of possible emerging psychosis (Warner, Reference Warner2002).
The basic symptoms concept has been operationalised by the ‘Schizophrenia Proneness Instrument’ Adult Version (SPI-A) (Schultze-Lutter etal. Reference Schultze-Lutter, Ruhrmann, Picker and Klosterkotter2007), which allows for a frequency-based severity rating of basic symptoms.
The SIPS/SOPS has become the established instrument in North American studies, whereas the CAARMS has a prevailing influence in Australian and European studies. The SPI-A is used in some European centres and usually applied together with the SIPS/SOPS and, to a lesser degree, with the CAARMS in order to allow for the assessment of the PRS according to both approaches.
Both CAARMS and SIPS/SOPS are semi-structured interviews that measure positive, negative, disorganised and common symptoms. All interviews require training typically of postgraduate-level clinicians, although good to excellent inter-rater reliability has not been difficult to attain (Miller etal. Reference Miller, McGlashan, Rosen, Somjee, Markovich, Stein and Woods2002).
Why is identification of the UHR group important?
Research findings from the past 15 years have provided some evidence for identification, follow-up and intervention in the UHR group.
The findings include:
1. Identification of those at UHR of developing psychosis could facilitate prevention or amelioration of the psychosocial deficits that in most cases have their onset during the ‘prodromal phase’ and are treatment resistant in the wake of the first episode (McGorry etal. Reference McGorry, Nelson, Amminger, Bechdolf, Francey, Berger, Riecher-Rössler, Klosterkötter, Ruhrmann, Schultze-Lutter, Nordentoft, Hickie, McGuire, Berk, Chen, Keshavan and Yung2009).
2. A number of studies have shown that brain changes are already present in the UHR group compared with healthy controls, and additional progressive brain changes have been documented in UHR individuals who transition to psychosis compared with those who do not (Pantelis etal. Reference Pantelis, Velakoulis, McGorry, Wood, Suckling and Philips2003; Velakoulis etal. Reference Velakoulis, Wood, Wong, McGorry, Yung, Phillips, Smith, Brewer, Proffitt, Desmond and Pantelis2006; Fornito etal. Reference Fornito, Yung, Wood, Phillips, Nelson, Cotton, Velakoulis, McGorry, Pantelis and Yücel2008).
3. UHR individuals according to current prodromal criteria already suffer from a variety of social, psychological, neuropsychological and neurobiological problems and consequently are in need of psychological and/or psychiatric treatment (Bechdolf etal. Reference Bechdolf, Pukrop, Kohn, Tschinkel, Veith, Schultze-Lutter, Ruhrmann, Geyer, Pohlmann and Klosterkotter2005; Woodberry etal. Reference Woodberry, Seidman, Giuliano, Verdi, Cook and McFarlane2010; Giuliano etal. Reference Giuliano, Li, Mesholam-Gately, Sorenson, Woodberry and Seidman2012).
4. In this early phase of illness, neuroprotective intervention strategies with little if any side effects may be effective, e.g., ω-3 fatty acids (Amminger etal. Reference Amminger, Schafer, Papageorgiou, Klier, Cotton, Harrigan, Mackinnon, McGorry and Berger2010).
5. A preventative approach to psychosis might offer an opportunity to positively change the public perception of the predictability and treatability of psychoses, thereby reducing public stigmatisation and discrimination of those suffering from the disorder.
6. Early detection and intervention might prevent self-stigmatisation by increasing self-empowerment early and by preventing symptoms that might lead to stigmatisation and discrimination by others, e.g., odd and eccentric behaviour or significant psychosocial decline.
Naturalistic follow-up studies
Although no naturalistic follow-up studies of this population have been completed in Ireland, a good number of generally small studies have been completed across the world (Olsen & Rosenbaum, Reference Olsen and Rosenbaum2006). Two large multi-centred studies have also been completed in Europe (Ruhrmann etal. Reference Ruhrmann, Schultze-Lutter, Salokangangas, Heinmaa, Linszen, Dingemans, Birchwood, Patterson, Juckel, Heinz, Morrison, Lewis, von Reventlow and Klosterkötter2010) and North America (Cannon etal. Reference Cannon, Cadenhead, Cornblatt, Woods, Addington, Walker, Seidman, Perkins, Tsuang, McGlashan and Heinssen2008).
A number of methodological issues are associated with UHR studies, which makes them difficult to directly compare. The UHR studies vary in how they define UHR/ARMs/PRS, i.e., the study inclusion criteria. ‘Transition to psychosis’ has been the outcome of interest in UHR studies. However, the point at which an individual crosses the line from high risk or prodromal state to psychosis threshold is ill-defined (Yung etal. Reference Yung, Nelson, Thompson and Wood2010). Furthermore, UHR studies are principally conducted in academic centres and therefore their findings may not be suitable for extrapolation to the general population.
A recent meta-analysis reported a consistent transition risk, independent of the psychometric instruments used, of 18% after 6 months of follow-up, 22% after 1 year, 29% after 2 years and 36% after 3 years of follow-up (Fusar-Poli etal. Reference Fusar-Poli, Bonoldi, Yung, Borgwardt, Kempton, Valmaggia, Barale, Caverzasi and McGuire2012). Significant moderators accounting for heterogeneity across studies and influencing the transition risk included the age of the participants, publication year, treatments received and diagnostic criteria used (Fusar-Poli etal. Reference Fusar-Poli, Bonoldi, Yung, Borgwardt, Kempton, Valmaggia, Barale, Caverzasi and McGuire2012).
A reduction in the transition rates has been reported more recently in the literature, from initially over 40% in 6 months to only 6.6% in 6 months and 12% in 1 year (Haroun etal. Reference Haroun, Dunn, Haroun and Cadenhead2006).
The two largest studies conducted to date the European Prediction of Psychosis Study (EPOS) study in Europe, which included 245 patients, and the North American Prodrome Longitudinal Study (NAPLS) study in North America, which included 291 patients, reported transition rates of 19% and 26.8% at 18 months of follow-up, respectively (Cannon etal. Reference Cannon, Cadenhead, Cornblatt, Woods, Addington, Walker, Seidman, Perkins, Tsuang, McGlashan and Heinssen2008; Ruhrmann etal. Reference Ruhrmann, Schultze-Lutter, Salokangangas, Heinmaa, Linszen, Dingemans, Birchwood, Patterson, Juckel, Heinz, Morrison, Lewis, von Reventlow and Klosterkötter2010). However, at 30 months of follow-up NAPLS found that 40% had transitioned to psychosis (Woods etal. Reference Woods, Addington, Cadenhead, Cannon, Cornblatt and Heinssen2009). Although still considerable, clinically significant and validating the selection criteria for UHR, these rates are lower than those reported in the initial studies.
The reduction in transition rates has caused some concern among the research field and several potential reasons have been put forth for this trend (Yung etal. Reference Yung, Yuen, Berger, Francey, Hung and Nelson2007). These include the provision of treatment to the UHR group, which may reduce transition rate. Treatment of UHR individuals often involves supportive therapy, cognitive therapy (CT) and medications only if indicated, e.g., an antidepressant if there is a comorbid depressive disorder. Antipsychotic prescribing is not routine in UHR centres; however, in NAPLS, in subjects not enrolled in clinical trials, 25% received antipsychotic medication and at least 13% in EPOS (Walker etal., Reference Walker, Cornblatt, Addington, Cadenhead, Cannon, McGlashan, Perkins, Seidman, Tsuang, Woods and Heinssen2009; Ruhrmann etal. Reference Ruhrmann, Schultze-Lutter, Salokangangas, Heinmaa, Linszen, Dingemans, Birchwood, Patterson, Juckel, Heinz, Morrison, Lewis, von Reventlow and Klosterkötter2010). As services become more established and staff experience of treating this cohort increases, the treatment effect could also be increasing. A related effect of UHR clinics becoming more established is that referrers become more vigilant and this could result in a ‘lead-in’ effect, in that patients tend to be identified in an earlier phase of the syndrome (Nelson, 2011, personal correspondence). This earlier recognition could result in enhanced prevention or could simply lead to a longer lag time to transition, i.e., the UHR populations being reported on in the literature could now differ from those of 5–10 years ago in that they are in an ‘earlier’ phase of the UHR syndrome and thus responding differently at least in the shorter term to identification and possibly treatment (Nelson, 2011, personal correspondence). As mentioned, the mean follow-up of the studies conducted to date is only 6–12 months, and thus it may be that longer periods of follow-up are required. Another possible consequence of this earlier ‘diagnosis’ of the UHR individuals is that more false positives are being included in the UHR sample. That is, the apparent UHR phenotype may have a number of different outcome trajectories, and that early detection increases the probability people never truly at risk of developing a psychotic disorder are being identified. Many individuals who at one stage meet UHR criteria diverge from the path leading to psychosis; some may have resolution of symptoms and difficulties, whereas others may develop non-psychotic disorders. Although these outcomes are seen in all UHR cohorts, it may be that these alternative pathways are more common earlier on. A number of medium- to long-term longitudinal follow-up studies of early UHR cohorts, e.g., PACE 400 (Nelson, 2011, personal correspondence), are currently underway and may assist in clarifying the longitudinal course of the UHR syndrome.
What predicts psychosis?
Yung etal. (Reference Yung, Phillips and Yuen2004) found significantly lower global assessment of functioning (GAF) scores (greater impairment), significantly longer delays accessing help and significantly higher levels of depression, measured with the Hamilton Rating Scale for Depression in the group that did transition. There were no significant differences between the groups on measures of anxiety, mania symptoms and negative symptom levels.
Schultze-Lutter etal. (Reference Schultze-Lutter, Ruhrmann, Hoyer, Klosterkotter and Leweke2007) also reported that higher levels of depression at baseline were associated with transition as were higher levels of negative symptoms. The EPOS found in 124 patients that six baseline features predicted psychosis independently: schizotypal personality, SIPS positive scale score >16, bizarre thinking, sleep disturbance, greater social impairment and years of education. The NAPLS found in 291 patients that five baseline features predicted psychosis independently: genetic risk for schizophrenia with recent deterioration in functioning, higher levels of unusual thought content, suspicion/paranoia, greater social impairment and a history of substance abuse (Cannon etal. Reference Cannon, Cadenhead, Cornblatt, Woods, Addington, Walker, Seidman, Perkins, Tsuang, McGlashan and Heinssen2008).
There has been considerable interest in the overlap and relationship between substance use, in particular cannabis use and psychosis, given that many of the symptoms experienced by people in the at-risk group parallel the effects of the use of substances. However, research findings have varied. Higher levels of cannabis use in the UHR population have been reported compared with controls. Cannabis use has been associated with lower levels of functioning within the UHR sample; however, although NAPLS reported substance misuse (not specifically cannabis use) as a clinical factor that predicted transition, a number of other studies have reported that cannabis use, misuse or dependence did not predict transition (Phillips etal. Reference Phillips, Curry, Yung, Yuen, Adlard and McGorry2002; Auther etal. Reference Auther2008).
Treatment options: interventional studies
A variety of interventions have been offered to those meeting criteria for being UHR. These interventions aim to reduce symptoms, as well as possibly delay or, even, fully prevent the onset of psychosis. However, as previously mentioned, there is methodological heterogeneity among these studies, thus making direct comparison difficult.
Psychosocial interventions
Supportive counselling and needs-based interventions are consistently used in UHR services and consist of assessment, education and empathetic but unstructured support (Yung etal. Reference Yung, Philips, Yuen, Francey, McFarlane, Hallgren and McGorry2000). The aim of this approach is reducing stress and enhancing coping skills. Supportive counselling and needs-based intervention have only been evaluated against more elaborate interventions as a control condition in randomised control trials. However, this help-seeking UHR population usually have high levels of anxiety before presenting to UHR programmes (Yung etal. Reference Yung, Philips, Yuen and McGorry2004). They frequently suspect that they have a psychotic illness as a result of the information they have gathered before presenting to services and the provision of supportive counselling with education and assistance with social and role functioning is considered a basic tenet in the provision of care for UHR patients.
Cognitive behavioural therapy (CBT)
Four randomised controlled trials (RCTs) have evaluated the effectiveness of CBT or CT as a treatment option for the UHR population (Morrison etal. Reference Morrison, French, Walford, Lewis, Kilcommons, Green, Parker and Bentall2004, Reference Morrison, French, Parker, Roberts, Stevens, Bentall and Lewis2007, Reference Morrison, French, Stewart, Birchwood, Fowler, Gumley, Jones, Bentall, Lewis, Murray, Patterson, Brunet, Conroy, Parker, Reilly, Byrne, Davies and Dunn2012; Addington etal. Reference Addington, Epstein, Liu, French, Boydell and Zipursky2011; Bechdolf etal. Reference Bechdolf, Wagner, Ruhrmann, Harrigan, Putzfeld, Pukrop, Brockhaus-Dumke, Berning, Janssen, Decker, Bottlender, Maurer, Möller, Gaebel, Häfner, Maier and Klosterkötter2012). Details of the four CT/CBT trials are summarised in Table 1.
Table 1 RCT of psychological interventions
RCT, Randomised controlled trial; SBRCT, single-blind randomised controlled trial; OLRCT, open-label randomised controlled trial; CAARMS, Comprehensive Assessment of At-Risk Mental States; UHR, ultra-high risk; CBT, cognitive behavioural therapy; CT, cognitive therapy.
CBT was considered an acceptable and rational treatment for a number of reasons. CBT works with processes such as meta-cognitions and self-schemas, which are believed to be abnormal in people at risk of psychosis (Birchwood etal. Reference Birchwood, Smith and Macmillan1989). Birchwood etal. (Reference Birchwood, Smith and Macmillan1989) showed that cognitive–behavioural monitoring of prodromal signs in clients with an existing diagnosis of psychosis enabled early intervention to prevent relapse or ameliorate mental state, and it has since been shown to significantly reduce relapse rates and hospital admissions in people at high risk of relapse (Gumley etal. Reference Gumley, O'Grady, McNay, Reilly, Power and Norrie2003). CBT has proven efficacious in the treatment of both acute and chronic psychosis (Birchwood & Trower, Reference Birchwood and Trower2006). Furthermore, most people in an ARMS show significant affective symptoms, for which CBT is an effective treatment. CBT with its use of collaborative problem identification and goal setting may also be a useful intervention for the false-positive group, who are seeking help for distressing symptoms but will not proceed to psychosis.
The CBT interventions used in all four trials used written manuals and were based on general principles of CBT. Key features of CBT for UHR individuals used in these trials included some of the basic tenets of CBT for psychosis; normalisation of experiences, de-catastrophising symptoms, generation and evaluation of alternative more reality-based interpretations, as well as testing them in behavioural experiments. In addition, CBT interventions included stress management, problem solving, coping and psycho-educational features.
Three of the four RCTs have demonstrated that treatment with CBT/CT is not associated with a reduced likelihood of developing psychosis in UHR patients. The studies vary in size (n = 51–288), duration of active treatment (6–24 months) and in the number of psychotherapy sessions (9 sessions in 6 months–60 sessions in 12 months). The largest trial by Morrison etal. (Reference Morrison, French and Parker2007) (n = 288) is a multi-site RCT comparing CT and monitoring of mental state with monitoring of mental state alone found that CT plus monitoring did not significantly reduce transition to psychosis or symptom-related distress, but reduced the severity of psychotic symptoms in young people at high risk. However, the mean number of CT sessions received over the 6 months of active treatment were considerably lower than in previous studies [mean 9.11 (s.d. 6.69; range 0–26)] and the transition rates were considerably lower than expected in both groups, 6.9% for CT and 9.0% for monitoring, which raise concerns about how truly ‘at risk’ the participants in this study were.
Morrison etal.'s (2004) RCT demonstrated that CT significantly reduced the likelihood of transition to psychosis over 12 months; however, this reduction was not maintained at 36 months’ follow-up (Morrison etal. Reference Morrison, French and Parker2007). The participant numbers in this study are small (n = 58) and the follow-up study at 3 years was very vulnerable to attrition with a loss to follow-up of 53% (n = 33). Bechdolf etal. (2012) in a multi-site RCT compared CBT with supportive counselling in 128 patients. CBT was provided for 12 months with an additional 12-month follow-up period. CBT significantly reduced the rate of transition to psychosis in the treatment group at 12 and 24 months. Addington and colleagues (2011) (n = 51) compared CBT with supportive counselling (Addington etal. 2011). Conversions to psychosis only occurred in the group who received supportive therapy, although the difference was not significant. Both groups improved in attenuated positive symptoms, depression and anxiety and neither improved in social functioning and negative symptoms. There were no differences between the two treatment groups.
Psychopharmacologic interventions
To date, five RCTs regarding psychopharmacologic interventions in the UHR state have been conducted (see Table 2). Four studies were double blind and one single blind. Active treatments included low-dose antipsychotic medication: risperidone + CBT (n = 59) (McGorry etal. Reference McGorry, Yung, Philips, Yuen, Francey, Cosgrave, Germano, Bravin, McDonald, Blair, Adlard and Jackson2002), olanzapine (n = 60) (Woods etal. Reference Woods, Breier, Zipursky, Perkins, Addington, Miller, Hawkins, Marquez, Lindborg, Tohen and McGlashan2003), amisulpride (n = 124) (Ruhrmann etal. Reference Ruhrmann, Bechdolf, Kühn, Wagner, Schultze-Lutter, Janssen, Maurer, Häfner, Gaebel, Möller, Maier and Klosterkötter2007), ethyl EPA (ω-3 fatty acids) (n = 81) (Amminger etal. Reference Amminger, Schafer, Papageorgiou, Klier, Cotton, Harrigan, Mackinnon, McGorry and Berger2010) and risperidone plus CBT or CBT in one three-arm study (n = 115) (Yung etal. Reference Yung, Phillips, Nelson, Francey, PanYuen and Simmons2011). Three studies were placebo controlled.
Table 2 RCT of pharmacotherapy
RCT, Randomised controlled trial; DBRCT, double-blind randomised controlled trial; OLRCT, open-label randomised controlled trial; SBRPCT, single-blind placebo controlled randomised controlled trial; DBRPCT, double-blind placebo controlled trial; CAARMS, Comprehensive Assessment of At-Risk Mental States; UHR, ultra-high risk; CBT, cognitive behavioural therapy; RIS, Risperidone; OLZ, Olanzapine.
Of the five trials, two demonstrated significantly lower transition rates associated with psychopharmacological interventions. The omega fatty acids and the risperidone plus CBT studies were associated with significantly lower transition rates in the treatment arm. However, in the risperidone plus CBT group the reduction was not maintained in the at 30–40-month follow-up. Rather, patients who had been on treatment tended to catch up with the control condition once treatment was discontinued, whereas transition rates in patients who originally had been randomised to the control condition did not increase significantly over time. The amisulpride trial curiously does not report on transition rates, but does report on symptomatic and functional improvement in the treatment group. Only the ω-3 fatty acids study showed preservation of the low transition rate over the following 9 months. In this study, the transition rates of 2.6% versus 21.1% (p < 0.05) after the acute 2.5–3.5 months treatment with omega fatty acids versus placebo remained remarkably stable (4.6% v. 27.5%, p < 0.05) over the next 9 months post-treatment in both groups. A finding that the large multi-centre double-blind randomised placebo-controlled trial North America, EURope, Australia PROdrome (NEURAPRO) study is currently attempting to replicate.
The most recent Cochrane review published in June 2011 concluded that it is unclear whether treating UHR patients provides benefit and that further evidence is needed before recommendations on treatment in this cohort can be given (Marshall & Rathbone, Reference Marshall and Rathbone2011). Since this systematic review was conducted, a further four studies – three CBT RCT's (Birchwood etal. Reference Birchwood, Smith and Macmillan1989; Gumley etal. Reference Gumley, O'Grady, McNay, Reilly, Power and Norrie2003; Birchwood & Trower, Reference Birchwood and Trower2006) and one RPCT – have been published (Woods etal. Reference Woods, Walsh, Saksa and McGlashan2010). Of these studies, Birchwood etal. (Reference Birchwood, Smith and Macmillan1989) (n = 128) finds that CBT does reduce transition rates at 12 and 24 months; however, the other three studies fail to demonstrate a significant effect on transition rates of CBT (Gumley etal. Reference Gumley, O'Grady, McNay, Reilly, Power and Norrie2003) (n = 51), CBT + monitoring of mental state (Birchwood & Trower, Reference Birchwood and Trower2006) (n = 288) or Risperidone and CBT (Corcoran etal. Reference Corcoran, First and Cornblatt2010) (n = 115).
Ethical issues and DSM 5
The Psychosis Working Group for DSM 5, chaired by Prof. William Carpenter, was proposing the inclusion of a new diagnostic category ‘Attenuated Psychotic Symptoms Syndrome’ (APS) for DSM 5, which is expected to be released in May 2013. This proposal provoked considerable discussion and debate in the scientific community.
In the work towards a new classification system as part of DSM 5, dimensional and longitudinal aspects of psychiatric disorders are to be given more significance both in the definition and characterisation of psychiatric disorders (DSM Task Force, 2012). It is in this context that UHR research, which has at its core the dimensional measurement of psychopathology and the prospective evaluation of outcomes, was being considered for inclusion as part of the newly introduced risk syndromes section. However, although the move to a more dimensional approach for DSM 5 has been generally welcomed, opinions were divided on the inclusion of APS as a risk syndrome.
Those who favoured inclusion contended that (a) UHR patients are currently ill, (b) they are at high risk of deteriorating, (c) no DSM-IV diagnosis accurately captures their current illness or future risk, (d) the diagnosis has been made with reliability and validity in the research setting and (e) inclusion in DSM 5 would make UHR a more visible and a legitimate subject of research and increase its funding potential (Corcoran etal. Reference Corcoran, First and Cornblatt2010; Woods etal. Reference Woods, Walsh, Saksa and McGlashan2010).
Those who opposed the inclusion of an ‘at-risk’ syndrome agreed that those meeting criteria for UHR are ill; however, the evidence base with regard to intervention in this population remains sparse. The inclusion of an ‘at-risk’ diagnosis in DSM 5 could increase the use of non-evidence-based interventions in this population (Yung etal. Reference Yung, Nelson, Thompson and Wood2010). Those opposed to inclusion generally agreed that ‘at-risk’ patients are at risk for deterioration; however, the rising rate of false positives noted in the naturalistic follow-up studies (well above 50% in nearly all recent studies, as discussed above) is of grave concern and further work to improve accuracy of detection is required. Both sides agree that no DSM-IV diagnosis accurately captures an ‘at-risk’ individual's current illness or future risk (Woods et al. 2010) and also agree that the diagnosis of ‘at risk’ has been made with reasonable reliability and validity in the research setting. However, to date the diagnosis has been studied almost exclusively in academic centres and may not be generalisable to the community settings where the DSM 5 is routinely applied.
In the United States, in particular, large studies of pharmacological interventions are usually funded by pharmaceutical companies. However, these companies are reluctant to fund large studies unless they can use the results to promote their products. Food and Drug Administration (FDA) approval is required for companies to promote medication legally in the United States. Although there is no requirement for a disorder to be recorded in the DSM to receive an FDA indication, it is likely to facilitate the process. However, whether FDA approval or inclusion of an at risk for psychosis diagnosis in DSM 5 would have facilitated the study of relatively benign treatment options including psychotherapeutic options or less ‘commercial’ medications, e.g., omega fatty acids or serine was less certain.
In April 2012, it was announced that due to the nascency of the UHR research and the lack of substantive field trials, a PRS would not be included in the main text of DSM 5. However, ‘Attenuated Psychotic Syndrome’ will be included in Section III for conditions being recommended for further study (DSM Task force, 2012).
Conclusions
The evidence suggests that it is possible to identify individuals who may be at risk of developing psychosis. It may also be possible to reduce or delay the transition to psychosis and improve the severity of non-psychotic symptoms and distress. Results from intervention studies, mostly involving second-generation antipsychotics and CBT, are currently insufficient to make treatment recommendations for this ‘at-risk’ group. The emerging research with regard to possible neuroprotective factors like omega fatty acids is promising, but will require replication in larger cohorts before it can be recommended.
An area of particular concern for the UHR research field is the failing transition rates. The coming years are likely to see the emergence of substantial longitudinal data from some of the original UHR centres, e.g., PACE in Melbourne. Some of these data will hopefully address the ambiguity underlying the significance of falling transition rates and better define both the psychotic and non-psychotic outcomes within this population. In particular, the illness course of those who do transition to psychosis needs to be compared with patients who receive treatment from the onset of psychotic illness only.
Interventional research in the field appears to be moving away from antipsychotic medication and towards neuroprotective and low-risk pharmacologic and non-pharmacologic interventions. Data from the ongoing larger multi-centred interventional studies will hopefully offer clarification around the risks and benefits of UHR treatment strategies where the extent of potential harm needs to be carefully balanced against the risk of transition to psychosis.
