Please cite this article as follows: Myles-Worsley M, Blailes F, Ord LM, Weaver S, Dever G, Faraone SV. 2007. The Palau Early Psychosis Study: Distribution of Cases by Level of Genetic Risk. Am J Med Genet Part B 144B:5–9.
The Palau early psychosis study: Distribution of cases by level of genetic risk†
Article first published online: 10 OCT 2006
Copyright © 2006 Wiley-Liss, Inc.
American Journal of Medical Genetics Part B: Neuropsychiatric Genetics
Volume 144B, Issue 1, pages 5–9, 5 January 2007
How to Cite
Myles-Worsley, M., Blailes, F., Ord, L. M., Weaver, S., Dever, G. and Faraone, S. V. (2007), The Palau early psychosis study: Distribution of cases by level of genetic risk. Am. J. Med. Genet., 144B: 5–9. doi: 10.1002/ajmg.b.30362
- Issue published online: 19 DEC 2006
- Article first published online: 10 OCT 2006
- Manuscript Accepted: 27 APR 2006
- Manuscript Received: 25 JAN 2006
- US Public Health Service. Grant Number: MH54186
- Grable Foundation NARSAD award
- genetic high risk;
- prodromal symptoms;
- clinical high risk;
The Palau Early Psychosis Study (PEPS) was designed to examine the pathogenesis of early psychosis in a high-risk population isolate. This paper describes the characteristics of our community-based, non-help seeking sample of 404 Palauan adolescents and quantifies the presence of early psychosis by level of genetic risk. The sample included 53 offspring of a schizophrenic parent designated as “Genetically Highest Risk” (GHR+) and 68 nieces/nephews of sib-pairs/trios, designated as “Genetically High Risk” (GHR). The remaining subjects were recruited through a high school survey that identified 62 “Genetically Moderate Risk” (GMR) adolescents with an affected second or third degree relative and 221 “Genetically Low Risk” (GLR) subjects with no close affected relatives. The GLR adolescents included 117 symptomatic or “Clinically High Risk” (CHR) adolescents and 104 asymptomatic normal controls. Based on a modified K-SADS-PL assessment, we identified 221 adolescents with early psychosis, 62 or 28% of whom had already transitioned to a psychotic disorder. Together, the two highest risk groups contributed 31% of the adolescent-onset psychosis cases and 27% of the prodromals. More than half of the early psychosis cases (53%) were GLR adolescents. The mean age of onset for DSM-IV psychosis was 12.9 years, and males transitioned at an earlier age than females. Our results indicate that Palauan adolescents, even GLR adolescents with no close affected relatives, have elevated rates of early psychosis. These young subjects can contribute valuable information about the familial transmission of schizophrenia, the developmental course of the illness, and rates of transition to frank psychosis. © 2006 Wiley-Liss, Inc.
Schizophrenia is a complex neuropsychiatric disorder that typically appears during late adolescence or early adulthood. A number of prospective studies of high risk (HR) subjects are being conducted in an effort to identify the risk factors that predict future onset of schizophrenia. Prospective studies can identify HR subjects either genetically, through family history [e.g., the New York High Risk Project, Erlenmeyer-Kimling et al., 1997], or clinically, through referrals of young people with pre-psychotic or prodromal symptomatology [e.g., the Early Psychosis Prevention and Intervention Center in Melbourne Australia, McGorry et al., 1996].
The Palau Early Psychosis Study (PEPS) was designed to examine the pathogenesis of early psychosis in a HR population isolate by combining the genetic (GHR) and clinical high risk (CHR) strategies. We are studying the clinical, psychosocial, and neurocognitive functioning of adolescents who are either at high genetic risk because of their relationship to affected family members or at high clinical risk because they have symptoms of early psychosis but no close affected relatives. The advantage of combining the GHR and CHR strategies is that we can evaluate the independent effects and also the joint effects of genetic liability and emerging psychopathology on adolescent functioning, as recommended by Cannon 2005].
The isolated island nation of Palau in Western Micronesia is well-suited for a combination of these two strategies. Palau has extended, high-density schizophrenia families that contain a large number of GHR adolescents who were identified during our previous genetic epidemiological study when we completely ascertained all cases and their families [Myles-Worsley et al., 1999]. Furthermore, this small island nation of 20,000 people has only about 1,200 high school students who are regularly surveyed by the Palau government as part of the country's public health program. We were able to utilize these government-sponsored surveys to screen for “potentially prodromal” adolescents and an age-matched sample of normal control adolescents [Ord et al., 2004]. Thus, our adolescent subjects represent a relatively unprecedented community-based sample that was recruited prior to any referral for psychiatric treatment.
The purpose of the present report is to describe the characteristics of the Palau adolescent sample and the results of the baseline clinical assessment. These findings were used to test the hypothesis that early psychosis would be most prevalent in adolescents with the highest levels of genetic risk.
A community-based sample of 404 Palauan adolescents 14–19 years of age (mean age = 16.5 years), who were not help-seeking, participated in the study. Table I presents sample sizes by level of genetic risk and gender.
|Level of genetic risk||Definition||Total No.||Female No.||Male No.||% Male|
|Highest: GHR+||Offspring of affected parent||53||28||25||47.2|
|High: GHR||≥2 affected aunts/uncles (second degree)||68||34||34||50.0|
|Moderate: GMR||≥1 affected second or third degree relative||62||37||25||40.3|
|Low: GLR||No first, second, third degree affected relatives|
|Clinical HR: CHR||Symptomatic||117||68||49||41.9|
We assessed 53 offspring of a parent with a psychotic illness, predominantly schizophrenia, designated as “Genetically Highest Risk” or GHR+, and 68 offspring of an unaffected parent with two or more affected siblings, designated “Genetically High Risk” or GHR. All of these adolescents were identified during the original ascertainment of schizophrenia cases and their families. As part of the clinical interview with schizophrenia patients and their participating siblings, we documented all offspring and asked the parent for the name, gender, and age of each child. Only children in the study age range were contacted. In the GHR+ group, two adolescents refused and seven adolescents, who were no longer living in Palau with the affected parent, were inaccessible. In the GHR group, 3 adolescents refused and 16 adolescents could not be contacted. The overall recruitment rate for these two groups was 81.2%.
The remaining 283 adolescent subjects were identified by surveying high school students in all four Palauan high schools as described in Ord et al. 2004]. Briefly, we developed a self-report questionnaire, the Y-PARQ (Youth Psychosis At Risk Questionnaire) based on the Comprehensive Assessment of at Risk Mental States [CAARMS, Yung et al., 2005] which presents 92 questions covering positive, affective, and negative symptoms. The total number of endorsements on the 24 most discriminating positive symptom items was used as a positive symptom score. We identified “potentially prodromal” and “probably normal” adolescents by identifying students in the upper and lower tails of the distribution of positive symptom scores. The high school survey was conducted in year one at the largest high school and in year two at the other four private high schools in Palau. A total of 898 (75%) of the approximately 1,200 high school students in Palau completed the Y-PARQ. The 25% who did not participate were either absent from the classroom on the day of the survey or declined/failed to complete the questionnaire.
The 283 high school students recruited for the study included 149 “potentially prodromal” adolescents and 134 “probably normal” adolescents. These students were disproportionately female (56.2%) as males more frequently cancelled appointments. For all participating subjects, the maternal and paternal pedigrees were constructed in order to identify any affected relatives out to the third degree. We found 62 adolescents with at least one affected second or third degree relative who were designated as “genetically moderate risk” or GMR. The remaining 221 subjects with no first, second, or third degree relatives with a psychotic disorder were designated as “Genetically Low Risk” or GLR. As previously reported, the Y-PARQ positive symptom score had a positive predictive value of 82.4% for identifying early psychosis [Ord et al., 2004]. The clinical assessment identified 117 GLR adolescents as symptomatic or CHR and 104 GLR adolescents as normal controls.
Study protocols and consent agreements for the PEPS were approved by both the University of Utah and the Palau Institutional Review Boards. All consent agreements were translated into Palauan. All subjects plus one parent or legal guardian gave written informed consent before being admitted into the study. After potential subjects were identified, the Palauan Principal Investigator, Francisca Blailes, sought consent from each adolescent's parents, and, if consent was granted, she then met with the adolescent to explain the study in the Palauan language and obtain his/her consent.
Clinical Assessment and Diagnosis
All adolescents were assessed prior to any psychiatric treatment over a 4-year period from 2000 to 2004. The clinical assessment instrument was a modified Kiddie-Schedule of Affective Disorders and Schizophrenia, Present and Lifetime Version [SADS-PL, Kaufman et al., 1997]. Modifications were made to incorporate prodromal and early psychotic symptomatology as evaluated by the Comprehensive Assessment of At Risk Mental States [CAARMS, McGorry et al., 2003; Yung et al., 2005] and to reflect cultural norms in Palau. One experienced Palauan clinician, Francisca Blailes, BSN, conducted all adolescent interviews, and the subject's responses to the questions were recorded verbatim by the interviewer. Interviews were conducted in the Palauan language unless the subject preferred to be interviewed in English.
The interview data were summarized by placing all positive, affective, and negative symptoms reported by the adolescent on a timeline representing the adolescent's lifespan. Each symptom was assigned a severity rating (1–6 scale ranging from 1 = “questionable” through 6 = severe) and a frequency rating according to the scoring system developed for the CAARMS. The interview data and the summary timeline data with symptom ratings were reviewed by a two-member diagnostic panel in order to reach a diagnosis. Each diagnostician made best-estimate diagnoses using DSM-IV criteria for a psychotic disorder (other than Brief Psychotic Disorder) and the PACE criteria [McGorry et al., 2003] for subsyndromal psychosis, often referred to as prodromal schizophrenia.
Briefly, DSM-IV criteria require at least two positive symptoms (only one if the symptom is a command voice, conversing voices, or bizarre delusions) that are present for a “significant portion of time during a 1-month period,” “continuous signs of the disturbance persist for at least 6 months (less for Schizophreniform Disorder),” and positive symptoms are accompanied by some social and/or occupational dysfunction. The PACE criteria for the prodrome require that positive symptoms (perceptual abnormalities/hallucinations and/or disorders of thought content/delusions) have been present over the past year and occur with at least moderate severity (a severity rating ≥3) and a frequency greater than once a month. In the case of prodromal subjects, the diagnostic panel distinguished between “severe” prodromal, a category reserved for adolescents with psychosis-like symptoms that failed to meet DSM-IV criteria, and “moderate” prodromal, a category covering a broad spectrum of symptom levels ranging from a relatively low level that just met PACE criteria to a high level that was not yet “severe.”
Table II presents the clinical diagnosis by level of genetic risk. The affection rates in the GHR+ and GHR adolescents were equivalently high (χ2 = 2.14, df = 1, ns). Almost half (48.8%) of these adolescent offspring were diagnosed with early psychosis, including 15.7% who had already transitioned to a DSM-IV psychotic disorder at the time of the baseline assessment.
|Number (% of group)|
|Transitioned||Severe prodromal||Moderate prodromal||Normal||Total|
|GHR+: Genetic highest risk||9 (17.0%)||4 (7.5%)||15 (28.3%)||25 (47.2%)||53|
|GHR: Genetic high risk||10 (14.7%)||6 (8.8%)||18 (26.4%)||34 (50.0%)||68|
|GMR: Genetic moderate risk||10 (16.1%)||10 (16.1%)||22 (35.5%)||20 (32.3%)||62|
|GLR: Genetic low risk|
|Clinical HR||33 (28.2%)||21 (17.9%)||63 (53.9%)||0 (0.0%)||117|
|Normal controls||0 (0.0%)||0 (0.0%)||0 (0.0%)||104 (100%)||104|
Table III presents the distribution of early psychosis cases by level of genetic risk. We identified a total of 221 Palauan adolescents with early psychosis, 62 or 28.1% of whom had already transitioned to a DSM-IV psychotic disorder, predominantly schizophrenia. Together, the two highest risk groups, the GHR+ and GHR adolescents, contributed 30.6% of the adolescent-onset DSM-IV psychosis cases and 27.1% of the prodromals. However, more than half of the early psychosis cases (52.9%) had no affected first, second, or third degree relatives.
|Psychosis||Prodromal||Total early psychosis|
The gender ratio of the early psychosis cases is presented in Table IV. We found a higher percentage of females (57.5%) than males (42.5%) among the symptomatic adolescents. This gender imbalance is partially explained by noting that participants recruited via the high school survey were disproportionately female (56.2%). However, two-thirds (66.1%) of the adolescents who had already transitioned to psychosis at the time of the baseline assessment were female, indicating that adolescent-onset psychosis was more prevalent in Palauan females 14–19 years of age.
|Transitioned to psychosis||Prodromal||Total early psychosis|
|No.||% Male||No.||% Male||No.||% Male|
Figure 1 shows the distribution of transition age by gender for the 62 adolescents with frank psychosis at the time of the baseline assessment. The mean transition age, defined as the age when symptoms first met DSM-IV criteria for a psychotic disorder, was 12.9 years (SD = 2.53). Males transitioned earlier than females (11.9 vs. 13.5 years, t = 2.37, P < 0.05).
Figure 2, which shows one of the high density schizophrenia pedigrees used to identify the GHR+ and GHR adolescents, illustrates the emergence of early psychosis by degree of relationship to affected family members. In this family with eight cases of broadly defined schizophrenia, we found 7 adolescents with early psychosis among the 10 adolescents who have completed the clinical assessment, and 5 of these are offspring of an affected parent, the GHR+ group.
We exhaustively assessed all consenting Palauan adolescents who are offspring of a parent with a psychotic disorder (GHR+) or the nieces/nephews of affected sib-pairs/trios (GHR). In addition, by conducting a survey of high school students, we were able to identify most of the “potentially prodromal” adolescents in Palau, all of whom were invited to participate in the research. Therefore, the 221 cases of early psychosis identified in Palau represent a relatively unprecedented community-based sample that was not help-seeking.
The results indicate that Palauan adolescents, even GLR adolescents with no close affected relatives, have an elevated rate of early psychosis. Our hypothesis that level of genetic risk would predict symptoms of early psychosis was not confirmed. Over half of the 221 cases of early psychosis were found among GLR adolescents with no close affected relatives, indicating that community-wide screening of adolescents for prodromal or “at risk mental states” is an important adjunct to genetic high risk strategies for identifying early psychosis.
Our young Palauan subjects will be able to provide new data about the familial transmission of schizophrenia. The present study indicates that level of genetic liability, as determined by the closest degree of relationship to an affected family member, is an important predictor of early psychosis. In our GHR+ group, 17.0% of the 53 adolescent offspring of an affected parent had already transitioned to a DSM-IV psychotic disorder at the time of the baseline assessment. However, in the past two generations of the Palauan schizophrenia families, recurrence risk to offspring of an affected parent was 23.4% [Myles-Worsley et al., 2006]. Therefore, this affection rate in GHR+ subjects is expected to rise over the next decade as adolescent offspring with prodromal symptoms progress into adulthood. In the case of GLR adolescents with no close affected relatives, the high school survey identified 33 out of approximately 1,200 Palauan students or 2.75% of the adolescent sample with a DSM-IV psychotic disorder, an affection rate that is close to the 2.67% prevalence for broadly-defined schizophrenia in the general Palauan population [Myles-Worsley et al., 1999]. This finding suggests that in a high risk population isolate like Palau, where ancestors can carry unexpressed susceptibility genes for schizophrenia, adolescents who are more distantly related to affected family members also carry an elevated risk for developing a psychotic disorder.
The 62 adolescent-onset cases represent a 2:1 female-to-male ratio, a reversal of the 2:1 male-to-female ratio found in the original Palau epidemiological study [Myles-Worsley et al., 1999]. We hypothesize that this shift in the gender distribution of cases in the youngest generation reflects the higher probability of referral for males versus females with schizophrenia in Palau. Drug abuse, crime, and unemployment are more prevalent in Palauan males compared to females, and these problems often precipitate referral for treatment.
Studies of early psychosis in Australia, North America, and Europe have found no significant gender differences in young people with prodromal symptoms, nor in the subset of patients that transition to psychosis. However, the Palau adolescent sample is not directly comparable to samples studied in other countries because the Palau adolescents were not help-seeking. Our data suggest that adolescent females with psychotic symptoms are less likely to seek help or to be referred for treatment. However, longer-term follow-up data is needed to determine if rates of transition to psychosis in prodromal adolescents reduce the gender imbalance that we have identified in untreated cases of psychosis in Palau.
Our results show a mean age of onset of DSM-IV psychosis of 12.9 years, a significantly earlier age of onset than previously reported. Research has shown that prodromal symptoms typically appear many years before the individual seeks treatment [Yung and McGorry, 1996a,b]. The CHR approach to prodromal research has primarily relied on referrals, and age of onset has generally been defined as the age when the referred patient meets criteria for full-blown psychosis and begins antipsychotic medication. Our study is relatively unique in assessing adolescents prior to referral and at an earlier age than previous prodromal studies. Furthermore, unlike previous prodromal research studies, we did not exclude young people who already met criteria for frank psychosis unless they had already begun treatment. Our clinical data suggest that symptoms may reach threshold for DSM-IV psychosis long before referral for treatment, the period known as “duration of untreated psychosis” (DUP). Among the 62 Palauan adolescents who had transitioned by the time of the baseline assessment, only three males have been treated with the conventional antipsychotics available in Palau. The remaining adolescents have declined to be referred to the Behavioral Health Division of Belau National Hospital for intervention and have stated that they prefer to manage their symptoms without medication.
This unique sample of adolescents can contribute valuable information about the developmental course of early psychosis in the 14- to 19-year-old age group and transition rates from prodromal to frank psychosis. Yung et al. 2004] have reported a 35% transition rate over a 12-month period in young subjects referred to the PACE Clinic in Melbourne Australia for an “At Risk Mental State.” Because our sample was younger and not help seeking, the adolescents we assessed may have been at an earlier stage of the prodrome than the PACE Clinic subjects. Therefore, transition to frank psychosis is expected to occur over a longer time period, and the transition rate may be lower than 35% now that regular clinical assessment and symptom monitoring has begun. Furthermore, some of the adolescents who had already transitioned at the time of the baseline assessment will improve in terms of the severity and frequency of their psychotic symptoms as they progress into adulthood, and they may remain unmedicated indefinitely.
In addition to examining transition rates and the developmental course of early psychosis, our follow-up reassessments will evaluate the role of environmental risk factors in triggering transition and the role of protective factors in delaying onset and preventing further deterioration. Our long-term goal is to accelerate progress toward preventive intervention strategies for young people with early psychosis.
The authors thank all members of the Palauan community who assisted in this research.
- 2005. Clinical and genetic high-risk strategies in understanding vulnerability to psychosis. Schizophr Res 79: 35–44. .
- 1997. The New York High Risk Project: Prevalence and comorbidity of axis I disorders in the offspring of schizophrenic parents at 25-year follow-up. Arch Gen Psychiatry 54: 1096–1102. , , , , , , , , , .
- 1997. 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 36: 980–988. , , , , , , , .
- 1996. EPPIC: An evolving system of early detection and optimal management. Schizophr Bull 22: 305–326. , , , , .
- 2003. The “close-in” or ultra high-risk model: A safe and effective strategy for research and clinical intervention in prepsychotic mental disorder. Schizophr Bull 29: 771–790. , , .
- 1999. A genetic epidemiological study of schizophrenia in Palau, Micronesia: Prevalence and familiality. Am J Med Genet: Neuropsychiatr Genet 88: 4–10. , , , , , , , , .
- 2006. Recurrence risk to offspring in extended multiplex schizophrenia pedigrees from a Pacific Island isolate. Am J Med Genet: Neuropsychiatr Genet (in press). , , , , .
- 2004. Screening for prodromal adolescents in an isolated high risk population. Schizophr Res 71: 507–508. , , , .
- 1996a. The initial prodrome in psychosis: Descriptive and qualitative aspects. Aust N Z J Psychiatry 30: 587–599. , .
- 1996b. The prodromal phase of first-episode psychosis: Past and current conceptualizations. Schizophr Bull 22: 353–370. , .
- 2004. Risk factors for psychosis in an ultra high-risk group: Psychopathology and clinical features. Schizophr Res 67: 131–142. , , , .
- 2005. Mapping the onset of psychosis: The comprehensive assessment of At-risk mental states. Aust N Z J Psychiatry 39: 964–971. , , , , , , , , , , , .