Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arch Gen Psychiatry. Author manuscript; available in PMC 2011 March 28.
Published in final edited form as:
PMCID: PMC3065347

Prediction of Psychosis in Youth at High Clinical Risk

A Multisite Longitudinal Study in North America



Early detection and prospective evaluation of individuals who will develop schizophrenia or other psychotic disorders are critical to efforts to isolate mechanisms underlying psychosis onset and to the testing of preventive interventions, but existing risk prediction approaches have achieved only modest predictive accuracy.


To determine the risk of conversion to psychosis and to evaluate a set of prediction algorithms maximizing positive predictive power in a clinical high-risk sample.

Design, Setting, and Participants

Longitudinal study with a 2½-year follow-up of 291 prospectively identified treatment-seeking patients meeting Structured Interview for Prodromal Syndromes criteria. The patients were recruited and underwent evaluation across 8 clinical research centers as part of the North American Prodrome Longitudinal Study.

Main Outcome Measure

Time to conversion to a fully psychotic form of mental illness.


The risk of conversion to psychosis was 35%, with a decelerating rate of transition during the 2½-year follow-up. Five features assessed at baseline contributed uniquely to the prediction of psychosis: a genetic risk for schizophrenia with recent deterioration in functioning, higher levels of unusual thought content, higher levels of suspicion/paranoia, greater social impairment, and a history of substance abuse. Prediction algorithms combining 2 or 3 of these variables resulted in dramatic increases in positive predictive power (ie, 68%–80%) compared with the prodromal criteria alone.


These findings demonstrate that prospective ascertainment of individuals at risk for psychosis is feasible, with a level of predictive accuracy comparable to that in other areas of preventive medicine. They provide a benchmark for the rate and shape of the psychosis risk function against which standardized preventive intervention programs can be compared.

Can prevention models now common to medicine be applied to psychotic disorders? Advances in early detection and intervention in cardiovascular disease,1 diabetes mellitus,2 and cancer3,4 have led to substantial reductions in morbidity and mortality and improved quality of life among individuals with these conditions. Efforts to extend such a prevention approach to schizophrenia have focused on developing and validating criteria for ascertaining individuals at risk for imminent onset of psychosis (ie, clinical high-risk or prodromal patients) and following them over time.57 The aims are to improve understanding of the mechanisms of disease onset and progression and to facilitate application of interventions before the illness takes hold, thereby reducing or preventing the devastating effects of schizophrenia. An advantage of this approach over high-risk methods based on a family history of schizophrenia is that assessments can be timed much more efficiently in relation to the onset of disorder.8

Using empirically defined criteria for a high-risk clinical state that emphasize recent onset or worsening of subsyndromal psychotic symptoms,9,10 previous studies have reported conversion rates of 9% to 76% in sample sizes of 13 to 110 subjects across 1- to 9-year follow-up intervals.1119 Much larger numbers of cases are required to provide statistically reliable modeling of the survival curve and estimates of the positive predictive power (PPP) of existing prodromal criteria. In addition, there is a great deal of variability in the assessment methods, sample characteristics, and length and frequency of follow-up across these studies.11

The North American Prodrome Longitudinal Study20 is a consortium of 8 research centers, each organized around the goal of improving the accuracy of prospective prediction of initial psychosis by ascertaining individuals who are at high clinical risk and following them at regular intervals for up to 2½ years. Although the research centers originally developed independent studies, they used similar ascertainment and longitudinal assessment methods, making it possible to form a standardized protocol for mapping acquired data into a new scheme representing the common components across the sites.20 This method yielded the largest database of prodromal cases followed up longitudinally worldwide (291 cases).

The primary aims of this study were to determine the rate of conversion to psychosis, to ascertain the shape of the survival function across 2½ years of follow-up, and to develop a multivariate risk prediction algorithm to guide the selection of cases in future studies. Included in the list of potential predictors were variables found to be associated with risk of conversion to psychosis in previous studies of smaller samples, including genetic risk for schizophrenia,8,21,22 severity of prodromal symptoms,11 severity of nonspecific symptoms,19 social and role functioning,19,23,24 and substance abuse.2529 It was hypothesized that a subset of these variables would contribute uniquely to the prediction of psychosis and combine into a multivariate algorithm with higher PPP compared with prodromal syndrome criteria alone.



The study protocols and informed consent documents, including procedures for data pooling, were reviewed and approved by the institutional review boards of the 8 participating study sites (Emory University; Harvard Medical School; University of California, Los Angeles; University of California, San Diego; University of North Carolina, Chapel Hill; University of Toronto; Yale University; and Zucker Hillside Hospital). Each site recruited potential subjects through clinical referrals as stimulated by talks to school counselors and mental health professionals in community settings. At each site, from 30% to 50% of the referred case patients met Structured Interview for Prodromal Syndromes (SIPS)9,13 criteria for study entry. Training workshops, conducted by Miller et al9,13 for the interviewers at each site, included lectures, group rating exercises, and detailed discussion of operational criteria used for diagnostic classification. Post-training agreement with the Yale University expert raters on the distinction between prodromal and psychotic levels of intensity on the positive symptom items (ie, the critical threshold for determining initial eligibility and subsequent conversion status) was excellent overall (κ, 0.90) and at each of the sites (κ range, 0.80–1.00).20 At each site, the raters were mental health specialists with academic credentials consisting of doctorates of medicine, master’s degrees, or doctorates of philosophy. New raters added during the course of the study had to achieve agreement standards with the training set before conducting assessments.

The SIPS criteria9 for a prodromal syndrome emphasize onset or worsening in the past 12 months of attenuated positive symptoms in 1 or more of 5 possible categories: unusual thought content, suspicion/paranoia, perceptual anomalies, grandiosity, and disorganized communication. A 7-point severity scale is used for each symptom, reflecting its frequency, duration, impact on functioning, and degree of loss of insight. Levels of 0 to 2 (none, questionable, or mild) indicate normal to sub-prodromal functioning; levels of 3 to 5 (moderate, moderately severe, or severe), a prodromal state; and a level of 6, a fully psychotic state. For example, a prodromal level of unusual thought content corresponds to an idea of reference or an odd belief that is worrisome or becomes meaningful because it will not go away and may be accompanied by an emerging sense that the event is caused by an external source, but doubt in this notion can be induced by contrary evidence. By contrast, a psychotic level of unusual thought content is an idea of reference or odd belief that is accompanied by full conviction for a specified period or that is acutely disruptive or disabling. The instrument is used to rate the severity of symptoms and to derive a categorical determination of prodromal status.

A subject may also qualify for a prodromal syndrome on the SIPS because of onset in the past 3 months of brief intermittent psychotic symptoms, which are positive symptoms of psychotic intensity but below the threshold required for a DSM-IV Axis I psychotic disorder diagnosis, or by having a genetic risk (defined as having a first-degree relative with a psychotic disorder or as having a diagnosis of schizotypal personality disorder) for psychosis and deterioration of 30% or greater on the General Assessment of Functioning Scale in the past 12 months. A genetic relationship between schizophrenia and schizotypal personality disorder has been detected in samples of families, twins, and adoptees.3036


All sites collected information on demographics, prodromal symptom severity,13 family history of mental illness,37 schizotypal personality disorder diagnosis,9 social and role functioning,38 comorbid psychiatric diagnoses (as assessed by the Structured Clinical Interview for DSM-IV39 or the Schedule for Affective Disorders and Schizophrenia for School-Age Children),40 and substance abuse (as assessed by the Structured Clinical Interview for DSM-IV or the Schedule for Affective Disorders and Schizophrenia for School-Age Children). For all of these variables except social and role functioning, the same instruments were used across sites, permitting straightforward data integration. For the functioning measures, the original data were recoded at each site using new scales developed specifically for this study and shown to have favorable psychometric properties in a reliability/ validity study.38 Further details of the construction of the federated database are published elsewhere.20


The SIPS was readministered at 6-month intervals to a maximum of 30 months. If case managers observed clinical deterioration in the patients under their care, a reassessment was conducted between regularly scheduled assessments. The primary outcome variable for this study was time from baseline evaluation to conversion to psychosis according to SIPS criteria. A SIPS diagnosis of a psychotic syndrome refers to psychotic symptoms of particular intensity (eg, delusional conviction) and frequency or duration (≥1 h/d for ≥4 d/wk during the past month) or of particular impact (seriously disorganizing or dangerous), designed to operationalize the threshold for a DSM-IV41 Axis I psychotic disorder diagnosis. Psychosis is the primary defining feature of schizophrenia but may occur in a number of other DSM-IV categories, including bipolar disorder and major depression. The Structured Clinical Interview for DSM-IV was not applied across all of the sites at follow-up to enable examination of the particular DSM-IV diagnoses attained at the point of conversion.


Of the 370 patients in the study, 83 (22.4%) were enrolled as participants in a randomized comparison of olanzapine vs placebo42 or in other small prospective treatment studies, and the remaining 287 (77.6%) were enrolled as participants in a longitudinal follow-up study with treatment of diagnosable symptoms provided on- or off-site when indicated in the view of the treating physician according to his or her interpretation of the standards for usual and customary care. Because treatment was not standardized across patients or sites, information on the dosing and duration of antipsychotic treatments was not available for most of the cases, but sites were able to indicate whether each case patient received antipsychotic drug treatment during the follow-up.


We used Kaplan-Meier survival analysis to ascertain the shape of the survival function during the 2½-year follow-up interval, the cumulative rate of conversion, and the incidence rates of conversion within successive 6-month epochs. We also sought to derive a multivariate algorithm that optimizes prediction of conversion to psychosis using the Cox proportional hazards model. In this form of analysis, predictors are modeled in relation to the time since baseline to conversion to psychosis, and subjects who do not experience conversion contribute to the prediction until they are no longer available for observation, at which point they are considered censored. A large number of potential predictor variables were available from the baseline assessment, and many of them would be expected to overlap with each other in relation to outcome. Separate multivariate Cox regressions were used to screen sets of potential predictor variables within the following 10 domains: sociodemographic characteristics, genetic risk for schizophrenia, positive and negative symptoms, disorganization, general symptoms, comorbid psychiatric diagnoses, social and role functioning, substance abuse, and antipsychotic drug treatment during the follow-up interval. These analyses used a backward selection approach to ascertain variables that have unique predictive associations with conversion at an initially liberal threshold of P<.10. We then performed an omnibus regression in which variables found to contribute uniquely to conversion in the initial series were considered together. Variables that remained significant at P<.05 in the omnibus analysis were then evaluated for multiplicative (interaction) effects in relation to conversion using the Lifetest procedure in SAS statistical software.43 In addition to the hazards ratio, we monitored the PPP and the sensitivity of each predictor or each combination of predictors.



Of the 370 subjects enrolled in the study, 291 (78.6%) completed at least 1 subsequent clinical evaluation, and 79 (21.4%) were lost to follow-up. As shown in Table 1, the patients with follow-up information did not differ significantly from those lost to follow-up in terms of age; parental education; severity of positive or negative symptoms; social, role, and global functioning; SIPS subdiagnosis; race; ethnicity; year of study entry; schizotypal personality disorder diagnosis; presence of a first-degree relative with psychosis; or presence of a first- or second-degree relative with psychosis. Sex was the only significant effect related to attrition, with a higher percentage of male patients among those lost to follow-up compared with the percentage of male patients among those successfully followed up (ie, 74.7% vs 58.4%).

Table 1
Sociodemographic Characteristics of the Samples With and Without Follow-up Information Available


Eighty-two of the 291 patients experienced conversion to psychosis (hereinafter referred to as converted cases), with a mean±SD time to conversion of 275.5±243.7 days since the baseline evaluation. Seventy-nine of the 82 converted cases met initial eligibility based on attenuated positive symptoms and 3 met initial eligibility based on brief intermittent psychotic symptoms (the corresponding numbers for the nonconverted cases were 203 and 6, respectively). Although only 2 patients were ascertained as prodromal exclusively in the genetic risk and deterioration category, 16 of the converted cases (and 18 of the nonconverted case) had a comorbid attenuated positive symptoms–genetic risk and deterioration prodromal diagnosis. The 209 nonconverted cases were followed up for a mean±SD of 575.4±258.4 days since the baseline assessment. Antipsychotic medications were prescribed for 35.1% of the patients during the follow-up interval. The Figure plots the Kaplan-Meier survival curve reflecting the percentage of subjects who did not experience conversion to psychosis (here in after referred to as nonconverted cases) during the 2½-year follow-up. The cumulative prevalence rate±SE of conversion to psychosis was 12.7%±1.9% at 6 months, 21.7%±2.5% at 12 months, 26.8%±2.8% at 18 months, 32.6%±3.3% at 24 months, and 35.3%±3.7% at 30 months. Thus, the SIPS criteria alone are associated with a PPP of 35% during 2½ years of follow-up. The incidence rate of conversion shows an overall decelerating trend during the follow-up period; this rate is 13% in the first 6 months, slows modestly to 9% from 7 to 12 months, slows to 5% per each 6-month epoch at 13 to 24 months, and then slows again to 2.7% from 25 to 30 months. For comparison, there were no conversions during this period among 134 normal control subjects who were matched demographically (by age, sex, and ethnicity) with the prodromal patients.

Cumulative survival distribution function modeling time to conversion to psychosis in 291 clinical high-risk (prodromal) patients and 134 demographically comparable normal control subjects (dashed line).


Of the 77 potential predictor variables examined (Table 2), 37 were associated with conversion to psychosis in univariate analyses. As shown in Table 3, when multivariate analysis was applied to sets of predictors from each assessment domain, which effectively removes redundancy among related measures, the number of predictors meeting the cutoff for inclusion fell to 16. Treatment with antipsychotic drugs during the follow-up interval was associated with a significant increase in risk of conversion (hazard ratio, 1.55).

Table 2
Potential Predictor Variables by Domain of Assessment
Table 3
Multivariate Proportional Hazards Regression Results Within Domains of Predictor Variables

When the 16 predictors that survived domain-wise multivariate screening were examined in an omnibus (cross-domain) multivariate analysis, conversion to psychosis continued to be related significantly and uniquely to genetic risk for schizophrenia with recent functional deterioration (χ2=10.45; P=.001), unusual thought content (χ2=6.36; P=.01), suspicion/paranoia (χ2=9.24; P=.002), social impairment (χ2=14.98; P<.001), and history of any drug abuse (χ2=6.82; P=.009). With these terms in the model, none of the other predictors that had survived the domain-wise screening procedure (Table 3) were related to conversion risk, indicating that their predictive associations were redundant with the other model terms. In particular, treatment with antipsychotic drugs during the follow-up interval was not significantly associated with conversion in the cross-domain multivariate analysis (χ2=0.59; P=.44).


Prediction statistics for each of the 5 uniquely predictive variables and their 26 possible combinations are given in Table 4. These represent all of the combinatorial algorithms tested. At the univariate level, these factors have approximately equivalent PPP (ie, 43%–52%), and each is superior in this regard to the SIPS criteria alone (35%). Nevertheless, the adjunctive use of these predictors in determining risk status results in a reduction in sensitivity. Sensitivity is excellent for suspicion/paranoia and impaired social functioning (79% and 80%, respectively), moderate for genetic risk for schizophrenia with recent functional decline and unusual thought content (66% and 56%, respectively), and poor for history of substance abuse (29%). Among the algorithms requiring co-occurrence of 2 risk factors, the models including genetic risk for schizophrenia with recent functional decline and unusual thought content or impaired social functioning have the highest PPP (69% and 61%, respectively), both substantially higher than that of the 1-factor models, although sensitivity is again relatively modest (ie, 38% and 55%, respectively). Two of the 3-factor models, involving genetic risk for schizophrenia with recent functional decline, unusual thought content, and either suspicion/paranoia or impaired social functioning, result in even higher PPP (74% and 81%, respectively) compared with the 2-factor models, with only marginal additional loss in sensitivity, and there is no further gain in prediction among any of the 4-factor models or the 5-factor model. Multivariate algorithms not requiring co-occurrence of risk factors have lower PPP (ie, 40%–45%) but substantially higher sensitivity (ie, 70%–95%) compared with those in Table 4. In addition, the algorithm reflecting the sum of the 5 independent risk factors is no better in terms of PPP (ie, 77%–79% among those with 4 or 5 factors) than the best-performing 3–coincident factor model.

Table 4
Prediction Statistics for 5 Uniquely Associated Predictors and Their Combinations

Controlling for antipsychotic drugs during the follow-up interval did not modify the significance or the magnitude of the results for the 5 uniquely predictive variables and their 26 possible combinations shown in Table 4.


The operationally defined criteria for prodromal schizophrenia show substantial predictive validity. Thirty-five percent of individuals identified on the basis of recent onset or worsening of subsyndromal psychotic symptoms experienced conversion to psychosis after 2½ years of follow-up. To our knowledge, the current sample size of 291 is nearly 3 times larger than that of any previous study, providing greater statistical confidence in the survival estimates. This 2½-year conversion rate of 35.3% represents a relative risk of 405 compared with the incident rate of all forms of psychosis in the general population during a comparable period (ie, 0.087%, or 0.034% per annum).45

The survival curve has a decelerating trend, such that progressively fewer cases convert to psychosis with increasing length of follow-up. This finding indicates that the prodromal criteria are sensitive to risk for imminent onset and provide an empirical basis on which to time the application of preventive interventions. After 2½ years, the risk of onset of psychosis is 2.7%, still higher than the annual incidence rate of schizophrenia in the general population but significantly below the rate observed in the first year of follow-up (ie, 20%).

In the 2 largest previous studies of prodromal psychosis,15,16 a conversion rate of 35% was observed among 104 clinical high-risk subjects identified using criteria comparable to the SIPS,16 and a conversion rate of 49% was observed (after 9.6 years of follow-up) among 110 cases identified using the Bonn Scale of Basic Symptoms.15 The Bonn Scale of Basic Symptoms emphasizes changes in social, emotional, and motivational factors and is thought to ascertain individuals in a much earlier stage of developing psychotic illness.46

Prediction algorithms incorporating combinations of 3 baseline variables (genetic risk for schizophrenia with recent functional decline, higher levels of unusual beliefs or suspiciousness, and greater social impairment) resulted in dramatic increases in PPP (74%–81%) compared with SIPS criteria alone (35%). These prediction algorithms were derived empirically, rather than confirmed through hypothesis testing. A relatively conservative empirical approach was used, such that we first screened the potential predictor variables for association with conversion in multivariate models within each assessment domain and retained only those variables that contributed uniquely to prediction in an overall (cross-domain) multivariate model for consideration in combinatorial algorithms maximizing PPP. Nevertheless, because the algorithms were derived empirically, they should be confirmed in an independent study with comparable sample size and selection, assessment, and follow-up criteria, as might be possible in future collaborations of the North American Prodrome Longitudinal Study group, as well as in similar collaborative efforts in Europe.47

Genetic risk for schizophrenia with recent functional deterioration was strongly and uniquely predictive of conversion to psychosis in this sample. Although the SIPS criteria include a prodromal syndrome involving genetic risk with a decline of 30% or more on the General Assessment of Functioning Scale in the past 12 months, patients who meet these criteria exclusively, without evidence of attenuated psychoticlike symptoms, are quite rare. Nevertheless, the risk construct implied by this category appears promising given that schizophrenia spectrum disorders are specifically elevated among first-degree relatives of patients with schizophrenia.34,48,49 Thus, functional decline, although otherwise nonspecific, should be highly predictive of psychosis in those with a genetic background for the disorder. To model this possibility, we created a new genetic risk and functional deterioration metric in which the genetic component is defined as in the SIPS, but the functional deterioration requirement was relaxed to a criterion of decline of 10% or greater in social, role, or psychological functioning in the year before ascertainment, using scales developed specifically for use in adolescent and preonset samples.38 This metric proved to be a more sensitive predictor of conversion to psychosis than a family history of psychosis or schizotypal personality disorder, whose contributions to psychosis risk were not significant once the genetic risk with functional deterioration term was modeled.

Social deficits and prodromal symptom severity at baseline are also key predictors of psychosis. The present findings indicate that the poorer the social functioning and the more severe the subsyndromal symptoms at ascertainment, particularly in the domains of unusual thought content and suspiciousness, the closer the subject is to the onset of psychosis. Deficits in social functioning are among the most robust behavioral correlates of genetic risk for schizophrenia and are present in many at-risk individuals from childhood.17,5053 Given that social deficits and prodromal symptom severity combine with a genetic risk for schizophrenia and recent functional decline in achieving maximal prediction, the onset of psychosis appears to be marked by a changing course of thinking and functioning against a backdrop of preexisting inherited vulnerability traits.17,22 In a previous study of 104 clinical high-risk patients from the Personal Assessment and Crisis Evaluation Clinic in Melbourne, Australia, the coincident requirement of meeting attenuated positive symptoms and genetic risk and deterioration criteria was associated with a PPP of 69% and a sensitivity of 31%.19 The increased sample size in the present study enabled the evaluation of specific symptom predictors and varying thresholds for functional deterioration. The predictive validity of other positive symptoms, such as perceptual abnormalities and grandiosity, is limited by their relatively low base rates in this sample.

A history of substance abuse also predicted conversion, although in multivariate analyses no specific substance class of the 7 tested (ie, alcohol, hypnotics, cannabis, amphetamines, opiates, cocaine, and hallucinogens) was significantly associated with risk. It is possible that larger studies will be needed to determine whether specific substances are associated with psychosis in prodromal cases. Although the low base rate of substance abuse severely limits sensitivity, its association with conversion risk is theoretically important because a drug-related mechanism may be capable of producing psychosis-promoting changes in brain function in some high-risk patients. Furthermore, this association, if confirmed, suggests that abstinence from drugs may help to lower the risk of psychotic illness in this population.

Although rates of conversion were higher among cases ascertained in earlier years of the study than more recently, after controlling for other predictors, this effect was not significant. Given that most prodromal research programs have increasingly engaged in community outreach and education efforts to increase awareness of early warning signs, decrease stigma, and stimulate referral, a higher proportion of more recently recruited patients may be ascertained in an earlier phase of risk, when symptoms are less severe.54 A decreasing transition rate could also reflect the increasing and/or more effective application of pharmacologic and psychosocial interventions in prodromal clinics and the community. Three preliminary studies support the notion that early intervention with either or both approaches is associated with prodromal symptom reduction and possibly with reduced or delayed risk for onset of psychosis.42,55,56 Although most investigators in the prodromal field advocate a highly conservative approach to drug treatment of clinical high-risk individuals, whereby antipsychotic drug therapy is initiated only after symptoms have reached a fully psychotic level of intensity, community-based physicians may sometimes be less conservative. In addition, it is not unusual for clinical high-risk patients to receive psychosocial interventions in the community or in the prodromal research programs themselves because these interventions are generally indicated to address presenting complaints (eg, low motivation, social withdrawal, and school failure) and have a lower risk of adverse events than drug treatment.

In this study, antipsychotic drug treatment was found to be associated with a significant increase in risk of conversion to psychosis at the univariate level, most likely reflecting the fact that most of the patients were treated in naturalistic circumstances in which physicians prescribe antipsychotics in the presence of greater severity of positive symptoms. This effect disappeared in the cross-domain multivariate analyses controlling for symptom severity, and accounting for this treatment variable did not modify the predictive relationships between the other study variables and conversion risk. Thus, the predictive relationships between other risk indicators and conversion, the decelerating survival function, and the 35% conversion rate observed in this study appear to be statistically independent of the application of such treatments. A more rigorous basis for dissociating the effects of treatment from natural factors influencing the risk of conversion to psychosis may be possible in a formalized treatment study with random assignment of patients to an active treatment vs placebo. However, patients who consent to participation in randomized, placebo-controlled studies of antipsychotic drugs may differ in substantial ways from those who are willing to be followed up longitudinally while retaining choice over interventions received. More restrictive exclusion criteria (eg, owing to diagnostic comorbidities or the need for conjoint treatments) and attrition owing to the adverse effects of drug treatments further limit generalizability of prediction findings from samples drawn from randomized treatment studies.

In general, the multivariate algorithms, while achieving a considerably higher PPP than any of the univariate models, were associated with much lower sensitivity. This pattern reflects the lower base rates of coincident occurrences of risk factors. Allowing for noncoincident combinations of risk factors resolves this problem, yielding excellent sensitivity but at the sacrifice of PPP, which falls to the level of the univariate models. Sensitivity may be increased in multivariate algorithms integrating quantitative measures that may have more favorable distributional properties than clinical ratings, such as indicators of brain anatomy or physiology or neurocognitive performance.5762

Attrition was unrelated to the primary variables that predicted conversion to psychosis in this sample. Although more male than female patients were lost to follow-up, conversion to psychosis did not vary according to sex, suggesting that this asymmetry is neutral with respect to the prediction results.

The present results apply to a treatment-seeking population that is recruited and screened for psychosis risk indicators. The results are not expected to be useful in general population screening. Moreover, the present criteria for a prodromal state reflect emerging clinical symptoms and signs that are thought to be on a continuum with fully psychotic states. Thus, the prediction results apply to a population that is already to some extent ill, rather than to a completely clinically unaffected population, and thus it is more appropriate to view prediction in this context in relation to risk of progression and increasing severity of illness than to the risk of illness per se. It is hoped that the knowledge gained from using this approach to monitor neurobiological changes over time will lead eventually to risk ascertainment criteria that can identify at-risk cases before emergence of subpsychotic clinical features.

The shape of the survival function suggests that the initial 2½ years after ascertainment represents a critical window of opportunity for evaluating changes in brain functioning that may underlie the development of psychosis and for the application of interventions that could attenuate or prevent the emergence of psychotic symptoms and functional disability. The present results thus provide a benchmark for the shape and rate of conversion risk against which to compare in future studies assessing comparable populations provided with a standardized intervention program. The use of prediction algorithms with 80% PPP will enable more selective recruitment into prevention programs (minimizing exposure of false-positive cases to potential adverse events) and facilitate studies attempting to elucidate neural and other changes proximal to the onset of psychosis.22,63,64


Funding/Support: This study was supported by investigator-initiated grants from the NIMH and a gift from the Staglin Music Festival for Mental Health.


Author Contributions: Drs Cannon and Cadenhead had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosures: Dr Cannon reports receiving investigator-initiated research funding support from multiple not-for-profit entities, including the National Institute for Mental Health (NIMH), the National Alliance for Research on Schizophrenia and Depression (NARSAD), and the Staglin Music Festival for Mental Health; and has served as a consultant to Janssen Pharmaceuticals and Eli Lilly and Co. Dr Cadenhead reports receiving investigator-initiated research funding support from the NIMH. Dr Cornblatt reports receiving investigator-initiated research funding support from not-for-profit entities, including the NIMH and the Stanley Medical Research Institute; has served as a consultant to Eli Lilly and Co, Bristol-Meyers Squibb, and Janssen Pharmaceuticals; and reports receiving unrestricted educational grants from Janssen Pharmaceuticals. Dr Woods reports receiving investigator-initiated research funding support from multiple not-for-profit entities, including the NIMH and the Donaghue, Stanley, and NARSAD foundation; and reports receiving investigator-initiated research funding support from multiple for-profit entities, including Eli Lilly and Co, Janssen Pharmaceuticals, UCB Pharma, and Bristol-Myers Squibb. Dr Addington reports receiving investigator-initiated research funding support from multiple not-for-profit entities, including the NIMH and Ontario Mental Health Foundation; and has served as a consultant to Pfizer Inc, Astra-Zenca, and Janssen Pharmaceuticals. Dr Walker reports receiving investigator-initiated research funding support from not-for-profit entities, including the NIMH and NARSAD. Dr Seidman reports receiving investigator-initiated research funding support from multiple not-for-profit entities, including the NIMH, the March of Dimes Foundation, the Mental Illness Neuroscience Discovery Institute, the Commonwealth of Massachusetts Department of Mental Health, and the National Association for Research in Schizophrenia and Depression; has served as a consultant to Shire, Eli Lilly and Co, and Janssen Pharmaceuticals; and reports receiving an unrestricted educational grant from Janssen Pharmaceuticals. Dr Perkins reports receiving research funding from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, Otsuka Pharmaceutical Co Ltd, Eli Lilly and Co, Janssen Pharmaceutica Products, and Pfizer Inc, and consulting and educational fees from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, Eli Lilly and Co, Janssen Pharmaceuticals, GlaxoSmithKline, Forest Labs, Pfizer Inc, and Shire. Dr McGlashan reports receiving investigator-initiated research funding support from the NIMH, the Personality Disorder Research Foundation, and Eli Lilly and Co; and has served as a consultant to Eli Lilly and Co, Pfizer Inc, Solvay/Wyeth, and Roche Pharmaceuticals.


1. Berenson GS. Obesity: a critical issue in preventive cardiology: the Bogalusa Heart Study. Prev Cardiol. 2005;8(4):234–243. [PubMed]
2. Peters AL, Davidson MB, Schriger DL, Hasselblad V. Meta-analysis Research Group on the Diagnosis of Diabetes Using Glycated Hemoglobin Levels. A clinical approach for the diagnosis of diabetes mellitus: an analysis using glycosylated hemoglobin levels. JAMA. 1996;276(15):1246–1252. [PubMed]
3. Adams EK, Breen N, Joski PJ. Impact of the National Breast and Cervical Cancer Early Detection Program on mammography and pap test utilization among white, Hispanic, and African American women: 1996–2000. Cancer. 2007;109(2 suppl):348–358. [PubMed]
4. Parekh DJ, Ankerst DP, Higgins BA, Hernandez J, Canby-Hagino E, Brand T, Troyer DA, Leach RJ, Thompson IM. External validation of the Prostate Cancer Prevention Trial risk calculator in a screened population. Urology. 2006;68(6):1152–1155. [PubMed]
5. McGlashan TH, Johannessen JO. Early detection and intervention with schizophrenia: rationale. Schizophr Bull. 1996;22(2):201–222. [PubMed]
6. McGorry PD, Yung AR, Phillips LJ. The “close-in” or ultra high-risk model: a safe and effective strategy for research and clinical intervention in prepsychotic mental disorder. Schizophr Bull. 2003;29(4):771–790. [PMC free article] [PubMed]
7. Yung AR, McGorry PD. The prodromal phase of first-episode psychosis: past and current conceptualizations. Schizophr Bull. 1996;22(2):353–370. [PubMed]
8. Cannon TD. Clinical and genetic high-risk strategies in understanding vulnerability to psychosis. Schizophr Res. 2005;79(1):35–44. [PubMed]
9. Miller TJ, McGlashan TH, Rosen JL, Cadenhead K, Cannon T, Ventura J, McFarlane W, Perkins DO, Pearlson GD, Woods SW. Prodromal assessment with the structured interview for prodromal syndromes and the scale of prodromal symptoms: predictive validity, interrater reliability, and training to reliability. Schizophr Bull. 2003;29(4):703–715. [PubMed]
10. Yung A, Phillips L, McGorry P, Ward J, Donovan K, Thompson K. Comprehensive Assessment of At Risk Mental States (CAARMS) Melbourne, Australia: PACE Clinic, University of Melbourne, Dept of Psychiatry; 2002.
11. Haroun N, Dunn L, Haroun A, Cadenhead KS. Risk and protection in prodromal schizophrenia: ethical implications for clinical practice and future research. Schizophr Bull. 2006;32(1):166–178. [PMC free article] [PubMed]
12. Yung AR, Phillips LJ, McGorry PD, McFarlane CA, Francey S, Harrigan S, Patton GC, Jackson HJ. Prediction of psychosis: a step towards indicated prevention of schizophrenia. Br J Psychiatry Suppl. 1998;172(33):14–20. [PubMed]
13. Miller TJ, McGlashan TH, Rosen JL, Somjee L, Markovich PJ, Stein K, Woods SW. Prospective diagnosis of the initial prodrome for schizophrenia based on the Structured Interview for Prodromal Syndromes: preliminary evidence of interrater reliability and predictive validity. Am J Psychiatry. 2002;159(5):863–865. [PubMed]
14. Yung AR, Phillips LJ, Yuen HP, Francey SM, McFarlane CA, Hallgren M, McGorry PD. Psychosis prediction: 12-month follow up of a high-risk (“prodromal”) group. Schizophr Res. 2003;60(1):21–32. [PubMed]
15. Klosterkötter J, Hellmich M, Steinmeyer EM, Schultze-Lutter F. Diagnosing schizophrenia in the initial prodromal phase. Arch Gen Psychiatry. 2001;58(2):158–164. [PubMed]
16. Yung AR, Yuen HP, McGorry PD, Phillips LJ, Kelly D, Dell’Olio M, Francey SM, Cosgrave EM, Killackey E, Stanford C, Godfrey K, Buckby J. Mapping the onset of psychosis: the comprehensive assessment of at-risk mental states. Aust N Z J Psychiatry. 2005;39(11–12):964–971. [PubMed]
17. Cornblatt BA, Lencz T, Smith CW, Correll CU, Auther AM, Nakayama E. The schizophrenia prodrome revisited: a neurodevelopmental perspective. Schizophr Bull. 2003;29(4):633–651. [PubMed]
18. Lencz T, Smith CW, Auther AM, Correll CU, Cornblatt BA. The assessment of “prodromal schizophrenia”: unresolved issues and future directions. Schizophr Bull. 2003;29(4):717–728. [PubMed]
19. Yung AR, Phillips LJ, Yuen HP, McGorry PD. Risk factors for psychosis in an ultra high-risk group: psychopathology and clinical features. Schizophr Res. 2004;67(2–3):131–142. [PubMed]
20. Addington J, Cadenhead KS, Cannon TD, Cornblatt B, McGlashan TH, Perkins DO, Seidman LJ, Tsuang M, Walker EF, Woods SW, Heinssen R. North American Prodrome Longitudinal Study. North American Prodrome Longitudinal Study (NAPLS): a collaborative multi-site approach to prodromal schizophrenia research. Schizophr Bull. 2007;33(3):665–672. [PMC free article] [PubMed]
21. Cannon TD, Kaprio J, Lonnqvist J, Huttunen M, Koskenvuo M. The genetic epidemiology of schizophrenia in a Finnish twin cohort: a population-based modeling study. Arch Gen Psychiatry. 1998;55(1):67–74. [PubMed]
22. Cannon TD, van Erp TG, Bearden CE, Loewy R, Thompson P, Toga AW, Huttunen MO, Keshavan MS, Seidman LJ, Tsuang MT. Early and late neurodevelopmental influences in the prodrome to schizophrenia: contributions of genes, environment, and their interactions. Schizophr Bull. 2003;29(4):653–669. [PubMed]
23. Owens DG, Miller P, Lawrie SM, Johnstone EC. Pathogenesis of schizophrenia: a psychopathological perspective. Br J Psychiatry. 2005;186:386–393. [PubMed]
24. Yung AR, Stanford C, Cosgrave E, Killackey E, Phillips L, Nelson B, McGorry PD. Testing the ultra high risk (prodromal) criteria for the prediction of psychosis in a clinical sample of young people. Schizophr Res. 2006;84(1):57–66. [PubMed]
25. Arseneault L, Cannon M, Witton J, Murray RM. Causal association between cannabis and psychosis: examination of the evidence. Br J Psychiatry. 2004;184:110–117. [PubMed]
26. Weiser M, Reichenberg A, Grotto I, Yasvitzky R, Rabinowitz J, Lubin G, Nahon D, Knobler HY, Davidson M. Higher rates of cigarette smoking in male adolescents before the onset of schizophrenia: a historical-prospective cohort study. Am J Psychiatry. 2004;161(7):1219–1223. [PubMed]
27. Semple DM, McIntosh AM, Lawrie SM. Cannabis as a risk factor for psychosis: systematic review. J Psychopharmacol. 2005;19(2):187–194. [PubMed]
28. Rosen JL, Miller TJ, D’Andrea JT, McGlashan TH, Woods SW. Comorbid diagnoses in patients meeting criteria for the schizophrenia prodrome. Schizophr Res. 2006;85(1–3):124–131. [PubMed]
29. Kristensen K, Cadenhead KS. Cannabis abuse and risk for psychosis in a prodromal sample. Psychiatry Res. 2007;151(1–2):151–154. [PMC free article] [PubMed]
30. Cannon TD, van Erp TG, Glahn DC. Elucidating continuities and discontinuities between schizotypy and schizophrenia in the nervous system. Schizophr Res. 2002;54(1–2):151–156. [PubMed]
31. Jang KL, Woodward TS, Lang D, Honer WG, Livesley WJ. The genetic and environmental basis of the relationship between schizotypy and personality: a twin study. J Nerv Ment Dis. 2005;193(3):153–159. [PubMed]
32. Kendler KS, Gruenberg AM, Strauss JS. An independent analysis of the Copenhagen sample of the Danish adoption study of schizophrenia, II: the relationship between schizotypal personality disorder and schizophrenia. Arch Gen Psychiatry. 1981;38(9):982–984. [PubMed]
33. Levinson DF, Mowry BJ, Sharpe L, Endicott J. Penetrance of schizophrenia-related disorders in multiplex families after correction for ascertainment. Genet Epidemiol. 1996;13(1):11–21. [PubMed]
34. Onstad S, Skre I, Edvardsen J, Torgersen S, Kringlen E. Mental disorders in first-degree relatives of schizophrenics. Acta Psychiatr Scand. 1991;83(6):463–467. [PubMed]
35. Squires-Wheeler E, Skodol AE, Bassett A, Erlenmeyer-Kimling L. DSM-III-R schizotypal personality traits in offspring of schizophrenic disorder, affective disorder, and normal control parents. J Psychiatr Res. 1989;23(3–4):229–239. [PMC free article] [PubMed]
36. Tsuang MT, Faraone SV. The genetic epidemiology of schizophrenia. Compr Ther. 1994;20(2):130–135. [PubMed]
37. Andreasen NC, Endicott J, Spitzer RL, Winokur G. The family history method using diagnostic criteria: reliability and validity. Arch Gen Psychiatry. 1977;34(10):1229–1235. [PubMed]
38. Cornblatt BA, Auther AM, Neidham T, Smith CW, Zinberg J, Bearden CE, Cannon TD. Preliminary findings for two new measures of social and role functioning in the prodromal phase of schizophrenia. Schizophr Bull. 2007;33(3):688–702. [PMC free article] [PubMed]
39. Spitzer RL, Williams JB, Gibbon M. Instruction Manual for the Structured Clinical Interview for DSM-IV. New York: Biometrics Research Dept, New York State Psychiatric Institute; 1994.
40. Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, Williamson D, Ryan N. Schedule for Affective Disorders and Schizophrenia for School-Age Children–Present and Lifetime Version (K-SADS-PL): initial reliability and validity data. J Am Acad Child Adolesc Psychiatry. 1997;36(7):980–988. [PubMed]
41. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4. Washington, DC: American Psychiatric Association; 1994.
42. McGlashan TH, Zipursky RB, Perkins D, Addington J, Miller T, Woods SW, Hawkins KA, Hoffman RE, Preda A, Epstein I, Addington D, Lindborg S, Trzaskoma Q, Tohen M, Breier A. Randomized, double-blind trial of olanzapine versus placebo in patients prodromally symptomatic for psychosis. Am J Psychiatry. 2006;163(5):790–799. [PubMed]
43. SAS Institute Inc. SAS/STAT 9.1 User’s Guide. Cary, NC: SAS Institute Inc; 2004.
44. Hall RC. Global assessment of functioning: a modified scale. Psychosomatics. 1995;36(3):267–275. [PubMed]
45. Kirkbride JB, Fearon P, Morgan C, Dazzan P, Morgan K, Tarrant J, Lloyd T, Holloway J, Hutchinson G, Leff JP, Mallett RM, Harrison GL, Murray RM, Jones PB. Heterogeneity in incidence rates of schizophrenia and other psychotic syndromes: findings from the 3-center AeSOP Study. Arch Gen Psychiatry. 2006;63(3):250–258. [PubMed]
46. Gross G, Huber G. Psychopathology of basic stages of schizophrenia in view of formal thought disturbances. Psychopathology. 1985;18(2–3):115–125. [PubMed]
47. Klosterkötter J, Ruhrmann S, Schultze-Lutter F, Salokangas RK, Linszen D, Birchwood M, Juckel G, Morrison A, Vázquèz-Barquero JL, Hambrecht M, von Reventlow H. The European Prediction of Psychosis Study (EPOS): integrating early recognition and intervention in Europe. World Psychiatry. 2005;4(3):161–167. [PubMed]
48. Onstad S, Skre I, Torgersen S, Kringlen E. Twin concordance for DSM-III-R schizophrenia. Acta Psychiatr Scand. 1991;83(5):395–401. [PubMed]
49. Parnas J, Cannon TD, Jacobsen B, Schulsinger H, Schulsinger F, Mednick SA. Lifetime DSM-III-R diagnostic outcomes in the offspring of schizophrenic mothers: results from the Copenhagen High-Risk Study. Arch Gen Psychiatry. 1993;50(9):707–714. [PubMed]
50. Calkins ME, Curtis CE, Grove WM, Iacono WG. Multiple dimensions of schizotypy in first degree biological relatives of schizophrenia patients. Schizophr Bull. 2004;30(2):317–325. [PubMed]
51. Dworkin RH, Lewis JA, Cornblatt BA, Erlenmeyer-Kimling L. Social competence deficits in adolescents at risk for schizophrenia. J Nerv Ment Dis. 1994;182 (2):103–108. [PubMed]
52. Hans SL, Auerbach JG, Asarnow JR, Styr B, Marcus J. Social adjustment of adolescents at risk for schizophrenia: the Jerusalem Infant Development Study. J Am Acad Child Adolesc Psychiatry. 2000;39(11):1406–1414. [PubMed]
53. Tyrka AR, Cannon TD, Haslam N, Mednick SA, Schulsinger F, Schulsinger H, Parnas J. The latent structure of schizotypy, I: premorbid indicators of a taxon of individuals at risk for schizophrenia-spectrum disorders. J Abnorm Psychol. 1995;104(1):173–183. [PubMed]
54. Yung AR, Yuen HP, Berger G, Francey S, Hung TC, Nelson B, Phillips L, McGorry P. Declining transition rate in ultra high risk (prodromal) services: dilution or reduction of risk? Schizophr Bull. 2007;33(3):673–681. [PMC free article] [PubMed]
55. McGorry PD, Yung AR, Phillips LJ, Yuen HP, Francey S, Cosgrave EM, Germano D, Bravin J, McDonald T, Blair A, Adlard S, Jackson H. Randomized controlled trial of interventions designed to reduce the risk of progression to first-episode psychosis in a clinical sample with subthreshold symptoms. Arch Gen Psychiatry. 2002;59(10):921–928. [PubMed]
56. Morrison AP, French P, Walford L, Lewis SW, Kilcommons A, Green J, Parker S, Bentall RP. Cognitive therapy for the prevention of psychosis in people at ultra-high risk: randomised controlled trial. Br J Psychiatry. 2004;185:291–297. [PubMed]
57. Byrne M, Hodges A, Grant E, Owens DC, Johnstone EC. Neuropsychological assessment of young people at high genetic risk for developing schizophrenia compared with controls: preliminary findings of the Edinburgh High Risk Study (EHRS) Psychol Med. 1999;29(5):1161–1173. [PubMed]
58. Cannon TD, Bearden CE, Hollister JM, Rosso IM, Sanchez LE, Hadley T. Childhood cognitive functioning in schizophrenia patients and their unaffected siblings: a prospective cohort study. Schizophr Bull. 2000;26(2):379–393. [PubMed]
59. Cosway R, Byrne M, Clafferty R, Hodges A, Grant E, Abukmeil SS, Lawrie SM, Miller P, Johnstone EC. Neuropsychological change in young people at high risk for schizophrenia: results from the first two neuropsychological assessments of the Edinburgh High Risk Study. Psychol Med. 2000;30(5):1111–1121. [PubMed]
60. Job DE, Whalley HC, Johnstone EC, Lawrie SM. Grey matter changes over time in high risk subjects developing schizophrenia. Neuroimage. 2005;25(4):1023–1030. [PubMed]
61. Pantelis C, Velakoulis D, McGorry PD, Wood SJ, Suckling J, Phillips LJ, Yung AR, Bullmore ET, Brewer W, Soulsby B, Desmond P, McGuire PK. Neuroanatomical abnormalities before and after onset of psychosis: a cross-sectional and longitudinal MRI comparison. Lancet. 2003;361(9354):281–288. [PubMed]
62. Lawrie SM, Whalley HC, Abukmeil SS, Kestelman JN, Miller P, Best JJ, Owens DG, Johnstone EC. Temporal lobe volume changes in people at high risk of schizophrenia with psychotic symptoms. Br J Psychiatry. 2002;181:138–143. [PubMed]
63. McGlashan TH, Hoffman RE. Schizophrenia as a disorder of developmentally reduced synaptic connectivity. Arch Gen Psychiatry. 2000;57(7):637–648. [PubMed]
64. Walker EF, Neumann CC, Baum K, Davis DM. The developmental pathways to schizophrenia: potential moderating effects of stress. Dev Psychopathol. 1996;8:647–665.