*Tsukasa Sasaki, MD, PhD, Professor, Office for Mental Health Support and Graduate School of Education, University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8655, Japan. Email: firstname.lastname@example.org
Aims: Prospective follow-up studies of high-risk children may help clarify the etiological factors in schizophrenia. While studies from North America, Europe and Israel have estimated the risk of schizophrenia at 7–16% in the offspring of an affected parent, no data have been reported for Asian populations.
Method: We started a follow up of the offspring of Japanese schizophrenia patients in 1978. We investigated the estimated risk of schizophrenia in 51 high-risk offspring at the 24-year follow up. The effects of the parents' status, including history of psychiatric hospitalization and social functioning, on the risk in the offspring were also investigated.
Results: The cumulative incidence of schizophrenia was 13.7 % and the lifetime prevalence was estimated to be 13.5 ± 4.8%. The association between the psychiatric hospitalization in the probands and the risk of schizophrenia in the offspring was not significant, and the Global Assessment of Functioning score was significantly lower in the probands with a history of psychiatric hospitalization than in those without such a history.
Conclusions: The risk of developing schizophrenia in Japanese high-risk offspring might be comparable with the Western results. The present study suggests that the severity of the disease or the level of social functioning may not significantly affect the risk in Japanese offspring.
WHILE THE LIFETIME risk of developing schizophrenia is approximately 1% in the general population, family studies have shown that the risk may be more than ten times higher in the offspring of a parent suffering from schizophrenia (high-risk offspring).1 These family studies of schizophrenia were mainly from Europe and North America, and few studies involved an Asian population.2,3 The elevated risk of schizophrenia in high-risk offspring is considered to be basically a genetic mechanism.4 However, the specific factors at work remain to be clarified. Environmental factors, which could interact with the genetic factors, might also play a role in the elevation of the risk. Such environmental factors might include the child's upbringing environment and the parents' mental status5 as well as several biological factors present in the environment during the early developmental stages.6
High-risk studies of schizophrenia prospectively follow the offspring born to parents with the disease (high-risk offspring) from early developmental stages to adulthood. This strategy helps estimate the risk of psychosis in the high-risk offspring. It also helps search for the etiological factors of schizophrenia and for clues to its prevention.7 Thus far, several groups from Europe, North America and Israel have conducted high-risk studies of schizophrenia (reviewed by Cannon et al.).8 The cumulative incidence of schizophrenia was estimated as follows in those studies: 8.3% in the USA,9 16.2% in Denmark,10 8.0% in Israel,11 13.1% in the USA,12 3.6% in Sweden13 and 6.7% in Finland.14 With the exception of an incidence of 3.6% in Sweden,12 the incidence ranged from 6.7% (in Finland)14 to 16.2% (in Denmark).10 While this relatively wide range may be due to chance or to differences in the estimating methods,7 it could be partly due to differences in the clinical features of the affected parents or in the epidemiological backgrounds among the studies.8
Here we report the 24-year outcome of our high-risk study of schizophrenia in Japan ('The Tokyo Schizophrenia High-Risk Study'). The follow up was started in 1978, and the last survey of the subjects was completed at the end of 2003. A major aim of the present study is to examine whether the risk of developing schizophrenia in Japanese high-risk offspring is comparable with the risk in European, North American or Israeli high-risk offspring. This is the first high-risk study of schizophrenia in Asia. No other Japanese or Asian study has thus far investigated the long-term outcome for high-risk offspring with a prospective follow up.
Another aim of the present study is to examine whether attributes of the affected parents, including social functioning and the presence/absence of psychiatric hospitalization, modify the risk of schizophrenia in high-risk offspring. A previous study observed that the risk of schizophrenia might differ between offspring who grew up in nursing institutions and those who were reared at home.15 The upbringing environment may have an effect on the future precipitation of the disease.5,11,15 The severity of the parents' illness may also biologically or psychologically affect the risk in the offspring.16 In the present subjects, half of the affected parents (the probands) had no history of psychiatric hospitalization at the start of the follow up. The probands were recruited at the authors' outpatient clinic and were under the care of the authors for most of the follow-up period. This may be a unique feature of the present study in contrast with the previous studies, which mainly followed up hospitalized patients and their offspring.9–14 Thus, in the present study, the high-risk offspring were all being reared at home by their biological parents at the start of the follow up. Whether these features of our study significantly affect the risk of schizophrenia in the offspring is of interest.
Recruitment of the probands
The probands were recruited at the outpatient clinic of the principal investigator (Y. O.), at the Department of Psychiatry, University of Tokyo Hospital, during the period 1978–1985. Those who met the following criteria were invited to participate in the study: (i) diagnosed with schizophrenia according to the Ninth Revision of the International Classification of Disease criteria;17 (ii) ethnically Japanese (i.e. with parents in whom no ethnicity other than Japanese was present); (iii) without a comorbidity of alcoholism, any other substance-related disorder, or organic mental disorders; (iv) having one or more healthy child aged ≦18 years at the time of recruitment; and (v) being married with a spouse who was the biological parent of the child and was living with the spouse and the children in the Tokyo area. Among 36 patients who met the criteria, agreement was obtained from 35 patients (six male and 29 female) and their spouses to participate in the study.
The procedure of the diagnostic assessment can be briefly described as follows. At the recruitment, the proband was interviewed and clinical records were reviewed separately by two of the authors including Y. O., and a consensus diagnosis was made. The diagnosis was reconfirmed 5–10 years after the start of the follow up considering the clinical course, and a final consensus diagnosis was made according to the Diagnostic and Statistical Manual of mental Disorders (Third Edition, Revised) (DSM-III-R) criteria18 in 1992 by three authors (T. S., S. H. and Y. O.). Among the initial 35 probands, four were finally diagnosed with schizoaffective disorder and one with delusional disorder. These five subjects were excluded from the follow up at this point, and we continued to follow up the 30 probands diagnosed with DSM-III-R schizophrenia (four male and 26 female, age (mean ± SD) = 47.8 ± 5.2 years as of 1992). At the time of recruitment, 15 (two male and 13 female) of the 30 probands had no history of psychiatric hospitalization. Spouses of the probands were also interviewed, and their mental status was evaluated by Y. O. from1986–1992. All the spouses were confirmed not to be suffering from schizophrenia or other major psychoses.
Recruitment of the high-risk offspring
Follow up of 65 offspring of the initial 35 probands was started between 1978 and 1985. Out of the total of 65, ten offspring born to the five probands whose final diagnosis was not schizophrenia were excluded. Therefore, the follow up of 55 offspring (27 male and 28 female) was continued after the final diagnosis of the probands in 1992. The ages of the 55 offspring at the start of the follow up were as follows: 16 newborns, 15 aged 1–5 years, 16 aged 6–10 years, six aged 11–15 years and two aged 16–18 years (mean ± SD = 5.5 ± 5.6 years).
Follow up of the high-risk offspring and probands
All the probands were regularly cared for by one of the authors at psychiatric outpatient clinics in Tokyo until the mid-1990s. When the care at the authors' clinics was discontinued, yearly contact by telephone or mail was begun. Also, the probands and their families were welcome to contact the authors when they wanted advice on or help with their treatment or with other concerns including the care of their offspring, especially when risky situations arose (e.g. school refusal and withdrawal). In addition to these frequent contacts, detailed investigations of the subjects' outcomes were twice conducted during the follow up. The first investigation was performed during the period 1992–1994. The second investigation was conducted approximately ten years after the first investigation (2002–2003, (mean ± S.D.) follow-up period = 23.7 ± 3.5 years). In these investigations, one or two authors visited the homes of the subjects and interviewed the probands, offspring and, in some cases, the spouses. When the subjects were not amenable to a visit, a telephone interview of the proband and the family was conducted. In the first investigation, the authors were able to visit all of the probands and families. In the second investigation, however, five probands with ten offspring were telephone-interviewed. When the offspring or probands were under the care of other mental health professionals, the authors asked the caregivers to collect the information, with the subjects' agreement. According to the information gathered through the regular contacts and the investigations, three authors including Y. O. evaluated whether the high-risk offspring developed psychotic disorders, including schizophrenia, schizoaffective disorder, mania and major depression with psychotic features, according to the DSM-III-R criteria. The mental condition and social functioning of the probands were assessed using the Global Assessment of Functioning (GAF) Scales19 in the second investigation.
Informed consent and ethical review
The objective of the study was explained, and informed consent was obtained from the probands and spouses at the initial recruitment and in the outcome surveys during the follow up. Consent was also obtained from the offspring when they were ten years old or older. The study was approved by the ethical boards of the faculties of medicine at the University of Tokyo and Mie University.
Association between the proband's features and the development of psychosis in the offspring was analyzed using the χ2 or Fisher's exact test. The cumulative incidence of psychosis in the offspring was calculated as of the end of 2003. Lifetime morbidity risk and survival probability were estimated with age correction using the Kaplan–Meier method. These analyses were conducted using SPSS, 11.0.
Outcome of the probands
The 24-year follow up was completed in 29 probands (four male and 25 female) out of the 30. A female proband with one female offspring dropped out after 10 years of follow up when they moved to a distant area. When the follow up was stopped in this case, the offspring was 21 years old and had not developed any of the psychotic disorders.
Out of the 29 probands, two (one male and one female) died due to suicide. Six others (one male and five female) experienced psychiatric hospitalization due to the exacerbation of schizophrenia during the follow up. Two (both female) of these six had no history of hospitalization before the start of the follow-up. Thus, 13 probands (two male and 11 female) out of the 29 had no experience of psychiatric hospitalization when the follow-up was completed. Four probands (all female) were divorced or separated from their spouses, and three of the four experienced no hospitalization during the follow up.
Risk of the psychoses in the offspring
Out of the 55 offspring, the follow up was stopped in four offspring (two male and two female) before any psychotic disorders had developed. Among the four, two male offspring died of physical illnesses (leukemia and heart failure) at the ages of 17 and 30, respectively. Two female offspring dropped out of the follow up at the ages of 21 and 22, when one moved to a distant area with her family and the proband of the other died.
Thus, the follow up was completed in 51 high-risk offspring (25 male and 26 female), who were born to 28 probands (four male and 24 female). The age of the offspring ranged from 18 to 43 years (mean ± SD = 29.2 ± 6.4) at the end of the follow up. Out of the 51 offspring, seven (three male and four female, two from affected fathers and five from affected mothers, [Table 1]) developed a psychotic disorder by the end of 2003. All of them were diagnosed with schizophrenia according to the DSM-III-R criteria. The cumulative incidence of schizophrenia was therefore 13.7% (95% CI = 6.8 – 25.7%). The age at onset was 15 years in two, 18 years in one, 19 years in three and 22 years in one offspring. The age-corrected lifetime morbidity risk using the Kaplan–Meier method was estimated to be 13.5 ± 4.8 %. The accumulated hazard curve using the Kaplan–Meier method is shown in Figure 1. An analysis according to the sex of the proband showed that among the offspring of the affected mothers (n = 39, cumulative incidence = 12.8 %: [95% CI = 5.6 – 26.7 %]), five developed schizophrenia, while out of the 12 offspring of the affected fathers two (17%) developed the disease.
Table 1. The demographic characteristics of the proband with or without affected offspring
Proband with affected offspring
Proband without affected offspring
(n = 7)
(n = 21)
Fisher's exact test.
Independent t-test. GAF, Global Assessment of Functioning; ns, not significant.
The risk of schizophrenia in the offspring according to the proband's history of psychiatric hospitalization is summarized in Table 2. Among the 51 offspring, 27 were from probands with a history of hospitalization (n = 15) and 24 were from probands without such a history (n = 13). Out of the 27 offspring of probands with a hospitalization history, four developed schizophrenia (14.8%), while out of the 24 offspring of probands without such a history, three (12.5%) developed the disease. No significant difference was observed in the risk of schizophrenia between the offspring of probands with a hospitalization history and the offspring of those without (14.8% vs 12.5%, P = 1, Fisher's exact test). Age was not significantly different between the 27 offspring of probands with a history and the 24 offspring of those without (29.3 ± 5.9 years vs 29.1 ± 7.0 years [mean ± SD] at the end of the follow up). GAF scores were significantly different between the probands with and without a hospitalization history (48.6 ± 19.0 and 63.1 ± 13.3 [mean ± SD], respectively) at the second investigation (t = 2.2, d.f. = 25, P = 0.03). When the probands with and without affected offspring (n = 6 and 21, respectively) were compared, the GAF scores were not significantly different (54.5 ± 13.8 vs 55.9 ± 19.1 (mean ± SD), respectively, t = 0.16, d.f. = 25, P = 0.87) (Table 1).
Table 2. Comparison of the GAF scores and the rates of the affected offspring between proband with and without history of psychiatric hospitalization at the end of the 24-year follow-up
† Independent t-test. ‡ Fisher's exact test. GAF, Global Assessment of Functioning; ns, not significant.
The present study investigated the risk of schizophrenia in the offspring of affected parents in the Japanese population. This is the first Asian high-risk study of schizophrenia, to our knowledge. The risk of developing schizophrenia was approximately 14% in the Japanese high-risk offspring. The cumulative incidence was 13.7% (95% CI = 6.8 – 25.7%), with an estimated lifetime risk of 13.5 ± 4.8%. In Western studies, the incidence has been reported to range from 6.7% to 16.2%.14 The risk of developing schizophrenia in Japanese high-risk offspring might therefore be comparable with the average of the reported rates or rather close to the two highest rates (13.1% in the study by Erlemeyer-Kimling et al., 1997,12 16.3% in that by Parnas et al., 199310 among the Western studies).
Approximately half of the probands (13 out of 28) never experienced psychiatric hospitalization. The disease, therefore, might not, on average, be severe in the present probands. This lack of severity did not appear to lower the risk in Japanese offspring, considering the relatively high incidence or lifetime-risk. In addition, the association between the presence/absence of psychiatric hospitalization in the probands and the risk of schizophrenia in the offspring (12.5% vs 14.8%, respectively) was not significant, while the GAF score was significantly lower in the probands with a history of psychiatric hospitalization than in those without such a history. These findings suggest that the severity of the disease or the level of social functioning may not significantly affect the risk in Japanese offspring. The present offspring were all raised at home by their biological parents. Therefore, a poor upbringing environment due to hospitalization of the parents or to the parent's lower social functioning might not significantly change the risk of schizophrenia in the offspring in the present sample.
We maintained close contact with the probands and their families. The probands were recruited at the authors' clinic and were clinically cared for for several years, as in the study by Fish.9 Regular contacts were continued if the probands moved to other clinics. The probands and spouses were welcome to ask the authors for help and advice on clinical and other mental health issues. This enabled us to continue close observation of the probands and families for more than 20 years. We were thus able to determine the onset of psychosis in the offspring using direct information from the families. This unique feature of the study might have helped us to achieve a high follow-up rate (93.0% or 51 out of 55 offspring). We are not able to evaluate the effects of this close care on the outcome of the proband and the rate of the development of psychoses in the offspring, because control data without the close care is not available in our sample.
A limitation of the present study may be the relatively small sample size, which resulted in the large 95% confidence interval of the cumulative incidence (6.8 – 25.7%) by reducing the statistical power. No significant association was observed in the present study between the risk in the offspring and the psychiatric hospitalization or social functioning of the proband. This finding could also be related to the lack of statistical power.
Another limitation is that we did not employ a structured interview. Information was mainly obtained for the psychiatric assessment from clinical interviews/records and from regular contacts with the probands and families. This is why we focused on the risk of psychotic disorders including schizophrenia and did not study other mental disorders in the offspring. Finally, the age of the offspring ranged from 18 to 43 years as of the end of the follow up. This range may be larger than those in previous studies. Considering these limitations, the present results should be interpreted with caution.
This study was supported by a Grant-in-Aid for Scientific Research (Exploratory Research) from the Ministry of Education, Science, Sports, and Culture of Japan (Dr. Yuji Okazaki). We are grateful to the clinicians for their corporation with this study (Shoji Nagakubo, MD, Michio Onai, MD, Masami Onai, MD, Reinin Haba, MD, and Shinya Hasegawa, MD).