ABC Schizophrenia study: an overview of results since 1996.

PURPOSE: The ABC Schizophrenia study, led by a single research team, investigated a schizophrenia sample systematically over quarter of a century. This paper summarises results from 1996 onwards. The initial goals were to explain the considerably higher age at first admission in women, and to obtain precise information on the onset and early course of schizophrenia as a prerequisite for early intervention. METHOD: The study was hypothesis-driven. People with schizophrenia were compared in the prodrome and at first admission to those with unipolar depression and to healthy controls. We analysed the medium-term (5-year) and the long-term (12-year) course of schizophrenia, its symptom dimensions, social parameters and predictors. SAMPLES: (1) 276 population-based first admissions (232 first episodes) of schizophrenia (age range 12-59 years); (2) a subsample of 130 first admissions for schizophrenia; (3) 130 first admissions for unipolar depression; (4) 130 healthy population controls and (5) 1,109 consecutive first admissions for schizophrenia spectrum disorder without an age limit. RESULTS: The prodromal stages of schizophrenia and depression were very similar until positive symptoms appeared. The most frequent symptom in schizophrenia was depressed mood. The course of psychosis from prodrome to 12 years following first admission was very variable. From 5 to 12 years after first admission the course was characterised by irregular exacerbations of the main symptom dimensions, with no overall deterioration or improvement. CONCLUSIONS: Schizophrenic psychosis and severe affective disorder, rather than representing discrete illnesses, probably mark different stages in the manifestation of psychopathology produced by various degrees of brain dysfunction.

[Schizophrenia - a disorder in its own right?: results from 25 years of the ABC study]. / Schizophrenie - eine einheitliche Krankheit?: Ergebnisse aus 25 Jahren ABC-Studie.

BACKGROUND: The ABC schizophrenia study conducted by the same team over 25 years initially aimed at illuminating the onset, prodromal stage and sex differences in age at first hospitalization in schizophrenia. New hypotheses were systematically generated from the results achieved. METHODS: A population-based sample of 276 first admission cases (232 first episodes, age 12-59 years), including a subsample of 130 first admissions (115 first episodes), were assessed to study prodromal stage, first illness episode, medium and long-term course and symptom dimensions in schizophrenia. The samples were compared with age and sex-matched healthy controls and with patients first admitted for unipolar depression. A total of 1,109 consecutive first admissions for schizophrenia spectrum disorders independent from the other study samples were assessed to study changes in symptomatology across the age range. RESULTS: Before the onset of psychotic symptoms the prodromal stages of schizophrenia and severe and moderately severe depression are difficult to distinguish. The most frequent symptom in the course of schizophrenia, depressed mood, also represents the most frequent initial symptom in both disorders. Prodromal depression is a predictor of more depressive and positive symptoms in the first episode but not in the further course of the illness. Psychosis incidence for men, diagnosed according to ICD 9 (295, 297, 298.3/4), shows a pronounced peak at age 15-24 years, for women a lower peak at age 15-29 years and a second, still lower peak at the menopausal age of 45-49 years. The explanation, confirmed in animal experiments, lies in a protective effect of estrogen due to reduced D2 receptor sensitivity. The effect is antagonized by an elevated genetic risk. Functional and social impairment emerge even at the prodromal stage and the severity depends on sex and social status. Young men with schizophrenia show a less favorable social course because of the earlier age of onset and socially adverse illness behavior. Late onset is associated with a milder, primarily paranoid symptomatology and less severe social impairment. Schizophrenia is a disorder of all ages showing roughly equal life time incidence rates for men and women but considerable difference in certain periods of age. The symptom dimensions show a plateau-like course 2-5 years after the first episode. Hidden behind this picture are irregular symptom exacerbations which vary in duration. Schizophrenia conveys the picture of recurrent vulnerability to crisis and not of a stable residual state of disordered brain development or of a progressive neurodegenerative process.

[Schizophrenia and depression]. / Schizophrenie und Depression.

BACKGROUND: Comparison of symptom-related and social role development between patients with schizophrenia, major depression and healthy controls provided insights into the untreated early course of the two disorders. SAMPLES AND METHODS: Symptoms, functional impairment and social disability were assessed and compared using the IRAOS and several other cross-sectional instruments in three samples. RESULTS: At the early illness stages there was considerable overlap in the symptom patterns and impairments presented by persons with schizophrenia and severe depression. The two disorders did not diverge until later in the early illness course with the onset of psychotic symptoms. Depressive symptoms showed a maximum in the first psychotic episode and relapse episodes and decreased with the remitting episode. Due to differences in cognitive appraisal depressed patients reported more functional impairment and social disability than patients with schizophrenia did. CONCLUSION: The early course of symptoms and social impairment in schizophrenia and depression seems to offer an opportunity to distinguish these disorders from variants of normal development fairly early. However, early diagnostic distinction and prediction of psychosis versus depression risk at the pre-psychotic prodromal stage do not seem promising due to the broad overlap in symptoms and impairment.

Recovery from psychotic illness: a 15- and 25-year international follow-up study.

BACKGROUND: Poorly defined cohorts and weak study designs have hampered cross-cultural comparisons of course and outcome in schizophrenia. AIMS: To describe long-term outcome in 18 diverse treated incidence and prevalence cohorts. To compare mortality, 15- and 25-year illness trajectory and the predictive strength of selected baseline and short-term course variables. METHODS: Historic prospective study. Standardised assessments of course and outcome. RESULTS: About 75% traced. About 50% of surviving cases had favourable outcomes, but there was marked heterogeneity across geographic centres. In regression models, early (2-year) course patterns were the strongest predictor of 15-year outcome, but recovery varied by location; 16% of early unremitting cases achieved late-phase recovery. CONCLUSIONS: A significant proportion of treated incident cases of schizophrenia achieve favourable long-term outcome. Sociocultural conditions appear to modify long-term course. Early intervention programmes focused on social as well as pharmacological treatments may realise longer-term gains.

Main risk factors for schizophrenia: increased familial loading and pre- and peri-natal complications antagonize the protective effect of oestrogen in women.

Women fall ill with schizophrenia 3 to 4 years later than men. The neurobiological mechanism, explaining the delay of onset in women until menopause, is presumably due to a sensitivity reducing effect of oestrogen on central d(2) receptors, as we have previously shown in animal experiments and in a controlled clinical study. The gender difference in age at onset seems to disappear in familial cases with schizophrenia, but it increases to highly significant values of 5 years or more in isolated cases according to a recent study by Albus and Maier (Schizophrenia Research 18:51-57, 1995). We tried to replicate these findings and to test the hypothesis of a functional antagonism between genetic predisposition to illness and the protective effect of oestrogen in a population-based sample of 232 first illness episodes of schizophrenia. In women with at least one first-degree relative suffering from schizophrenia, age at onset defined by first psychotic symptom was significantly reduced by several years and the difference with men disappeared. In sporadic female cases (no mental disorder in first-degree relatives) the age at onset was slightly increased compared with the total sample, which was in accordance with our hypothesis. In men with familial schizophrenia, but without a protective agent like oestrogen, the age at onset was only slightly and non-significantly reduced compared with the total group and with sporadic cases. This was in line with Albus and Maier and with our hypothesis that only the protective effect of oestrogen could be antagonized by a strong genetic disposition. The second main risk factor for schizophrenia is pre- and peri-natal complications. We compared men and women from our sample of first illness episodes with a history of pre- and peri-natal complications with those without a history of obstetric complications. In women the age at first psychotic symptom was markedly reduced, but due to small case numbers not significantly, compared with women without the risk factor and with the total group. Again, schizophrenic men with a history of pre- and peri-natal complications showed only a small, non-significant reduction of age at onset compared with the total and the group without the risk factor. Therefore, we concluded that the degree of genetically determined vulnerability and, presumably to a slightly lesser extent, the degree of pre- and peri-natal brain injury antagonizes the onset delaying effect of oestrogen in schizophrenia.

[Methodological problems of longitudinal studies on schizophrenia]. / Methodische Probleme der Verlaufsforschung an der Schizophrenie.

Longitudinal studies are a key to understanding schizophrenia. They are the more informative, the longer the periods covered. Hence, good studies into the course of schizophrenia almost exclusively involve a lot of effort and cost. In practice, however, time-consuming methods and design variables must be avoided. The pitfalls this constraint produces are instructive of the difficulties longitudinal studies are faced with in striving for valid results. For reasons of research economy, requirements must be adjusted to study objectives. Studies into the short term course are less time-consuming, but because of the rapid changes in the illness course study intervals should be defined clearly and observed strictly. In long-term studies, too, one source of error lies in the highly varying lengths of illness of the patients studied. Even some of the classic long-term studies are marred by this error. The beginning of the follow-up period should be comparable across the study cohort and as close to illness onset as possible. To obtain generally valid results the probands must be representative of all the illness cases in the general population not only at the outset, but also all the later stages of the study. Besides the efforts to avoid attrition in the study cohort, ways must be found for correcting and estimating data for an acceptable proportion of drop-outs. In the analysis of course and outcome the indicators chosen must be apt to the traditional subtypes as well as to a theoretical symptom patterns and empirical symptom structures. In the context of typical design variables of longitudinal studies the assets and weaknesses of two retrospective and one prospective design will be discussed. Concerning the social course, importance of disease-independent factors, such as age, sex and level of social development at illness onset, as well as of control groups will be demonstrated. Predictor models will be discussed with reference to the direct and indirect influences involved. Examples of such analyses will be given.

The epidemiology of onset and course of schizophrenia.

Traditionally the heterogeneity of schizophrenia was dealt with by subdividing the syndrome into different subtypes. However, due to lacking standards, the result was an immense variety of subtypes partly based on cross-sectional assessments, partly taken the whole course between onset, resp. first admission and outcome after many years into account. Some solutions were based on symptomatology only, others also relied on social characteristics as the ability to fulfil different roles in family and the world of employment. So it is not surprising that the number of subtypes ranges from two up to more than 70. As one possible solution Carpenter and Kirkpatrick (1988) suggest that attempts to subdivide the schizophrenic syndrome should concentrate on few significant parts of the course thought of to represent specific disease processes. Based on two epidemiological studies finding about the onset, middle course and late course of schizophrenia are presented. In three quarter of the cases the onset of the first psychotic episode in schizophrenia is preceded by a prodromal phase with a mean length of about five years. The earliest signs of the disorder are depressive and negative symptoms. Early depressive symptoms predict higher overall symptom scores in the first illness episode and lower scores for affective flattening in the medium-term course. There is no decrease in the number of patients with acute symptomatology over fifteen years after first hospital admission, rather there is a tendency of an increase. With respect to social abilities we found a significant increase of disability over time. But the change already takes place during the first five years. Approx. 60% of those falling ill with schizophrenia become chronic and approx. 25% will recover during the first five to six years.

Social disability in schizophrenia: its development and prediction over 15 years in incidence cohorts in six European centres.

BACKGROUND: This paper focuses on the long-term course of social disability in schizophrenia assessed at first onset, and after 1, 2 and 15 years in incidence cohorts in six European centres in Bulgaria, Germany, Ireland, The Netherlands, the Czech Republic and the United Kingdom. The study population comprises 349 patients comprising 75% of the original cohorts. METHODS: Social disability was assessed in a standardized way with the WHO Disability Assessment Schedule. RESULTS: Social disability in schizophrenia appears to be a persistent phenomenon. Its severity decreased overall in the period of follow-up, but this was not so in a small group traced to hospital or sheltered accommodation. Only 17% of subjects had no disability and 24% still suffered from severe disability. The great majority lived with their family, a partner, or alone. A deteriorating course was more frequent than late improvement. Gender, age, onset, duration of untreated psychosis or type of remission during the first 2 years did not predict the long-term outcome of disability. Severity of disability at the first three assessments of the illness contributed significantly to the explanation of its variance at 15 years. CONCLUSION: Disability generally ameliorates, but less than expected or hoped. It needs continuing attention and care in this era of de-institutionalization.

Depression, negative symptoms, social stagnation and social decline in the early course of schizophrenia.

OBJECTIVE: The aim of this study was to investigate when social consequences in schizophrenia emerge, and what conditions give rise to the social disadvantage evident in people suffering from schizophrenia. METHOD: Early course in schizophrenia was studied in a population-based sample of 232 first illness-episode cases retrospectively from onset to first admission, and in a representative subsample of 115 patients prospectively at six cross-sections over a period of 5 years. Data on non-specific and negative symptomatology and social development was compared with data from an age- and sex-matched control group drawn from the normal population. RESULTS: In total, 73% of the patients showed a prodromal phase of several years. First signs were depressive and negative symptoms. In 57% of cases social disability emerged 2 to 4 years before first admission. Social consequences depended on the level of social development at onset. An early onset involved social stagnation, and a late onset was associated with social decline. Men's poorer social outcome was determined by their lower level of social development at onset and socially adverse illness behaviour. The 5-year symptom-related course showed no gender difference. At 81% the lifetime prevalence of depressive mood until first admission was several times higher in schizophrenics than in healthy controls. Early depression predicted a lower subsequent score for affective flattening. Suicide indicators were predicted by lack of self-confidence and feelings of guilt early in the illness. CONCLUSION: Taking into account a prodromal phase of several years on average before first hospital admission, early detection, case identification and intervention are urgently needed. The intervention must be targeted at syndromes such as early depression, negative symptoms and certain forms of cognitive and social impairment.

The course of schizophrenia in the light of modern follow-up studies: the ABC and WHO studies.

In schizophrenia most of the social consequences emerge in the prodromal phase of the illness and before treatment is initiated. Further course is determined by the level of social development at illness onset and by age- and sex-related illness behavior. Despite the sex difference in age at onset the disease process seems to be the same in both sexes, since social course in men and women converges in the long run. Although great variation in outcome between the patients is to be observed at each cross-section, the medium and long-term symptom-related course of schizophrenia shows a high degree of stability at the individual level.

The ABC Schizophrenia Study: a preliminary overview of the results.

The ABC Schizophrenia Study, a large-scale epidemiological and neurobiological research project commenced in 1987, initially pursued two aims: (1) to elucidate the possible causes of the sex difference in age at first admission for schizophrenia and (2) to analyse the early course of the disorder from onset until first contact and its implications for further course and outcome. First, transnational case-register data (for Denmark and Germany) were compared, second, a population-based sample of first-episode cases of schizophrenia (n = 232) were selected and third, the results obtained were compared with data from the WHO Determinants of Outcome Study by using a systematic methodology. A consistent result was a 3-4 years higher age of onset for women by any definition of onset, which was not explainable by social variables, such as differences in the male-female societal roles. A sensitivity-reducing effect of oestrogen on central D2 receptors was identified as the underlying neurobiological mechanism in animal experiments. Applicability to humans with schizophrenia was established in a controlled clinical study. A comparison of familial and sporadic cases showed that in cases with a high genetic load, the sex difference in age of onset disappeared due to a clearly reduced age of onset in women, whereas in sporadic cases it increased. To analyse early course retrospectively, a semistructured interview, IRAOS, was developed. The early stages of the disorder were reconstructed in comparison with age- and sex-matched controls from the same population of origin. The initial signs consisted mainly of negative and affective symptoms, which accumulated exponentially until the first episode, as did the later emerging positive symptoms. Social disability appeared 2-4 years before first admission on average. In early-onset cases, social course and outcome, studied prospectively over 5 years, was determined by the level of social development at onset through social stagnation. In late-onset cases, decline from initially high social statuses occurred. Socially negative illness behaviour contributed to the poor social outcome of young men. Symptomatology and other proxy variables of the disorder showed stable courses and no sex differences. Further aspects tested were the sequence of onset and the influence of substance abuse on the course of schizophrenia, primary and secondary negative symptoms, structural models and symptom clusters from onset until 5 years after first admission.

Causes and consequences of the gender difference in age at onset of schizophrenia.

The ABC (age, beginning, course) schizophrenia study was commenced in 1987 to generate and test hypotheses about pathogenic aspects of schizophrenia. One of the main branches of the study focused on how gender influences the age distribution of onset, symptomatology, illness behavior, and early course in schizophrenia. Proceeding from one of the rare, strikingly deviating, consistent findings--the gender difference in age at first admission--we launched a systematic search for explanations by generating and testing hypotheses in a series of substudies. We moved from the epidemiological to the neurobiological and finally to the clinical level. The present article is an attempt to provide a brief overview of the individual stages of the ABC study and the different levels of investigation involved in formulating and testing the estrogen hypothesis in animal experiments and in demonstrating its applicability to human schizophrenia. From these results, three hypotheses were formulated and tested on data from an ABC study sample of 232 first-episode cases of schizophrenia. The analyses described here represent the latest stages of the ABC study.

Epidemiology of schizophrenia.

OBJECTIVE: To characterize the epidemiology of schizophrenia. METHOD: Narrative literature review. RESULTS: Each year 1 in 10,000 adults (12 to 60 years of age) develops schizophrenia. Based on a restrictive and precise definition of the diagnosis and using standardized assessment methods and large, representative populations, the incidence rates appear stable across countries and cultures and over time, at least for the last 50 years. Schizophrenic patients are not born into ecological and social disadvantage. The uneven distribution of prevalence rates is a result of social selection: an early onset leads to social stagnation, a late onset to descent from a higher social status. The main age range of risk for schizophrenia is 20 to 35 years. It is still unclear whether schizophrenia-like late-onset psychoses (for example, late paraphrenia) after age 60 should be classified as schizophrenia either psychopathologically or etiologically. In 75% of cases, first admission is preceded by a prodromal phase with a mean length of 5 years and a psychotic prephase of one year's duration. On average, women fall ill 3 to 4 years later than men and show a second peak of onset around menopause. Consequently, late-onset schizophrenias are more frequent and more severe in women than in men. The sex difference in age of onset is smaller in cases with a high genetic load and greater in cases with a low genetic load. Type of onset and core symptoms do not differ between the sexes. The most pronounced sex difference is the socially negative illness behaviour of young men. CONCLUSIONS: Among the factors determining social course and outcome are level of social development at onset, the disorder itself (for example, genetic liability, severity of symptoms, and functional deficits), general biological factors (for example, estrogen), and sex- and age-specific illness behaviour.

[Can outpatient measures prevent inpatient admission?]. / Können durch ausserstationäre Massnahmen stationäre Aufenthalte verhindert werden?

During the last decades the focus of psychiatric care has shifted from hospital to the "community". As the philosophy of deinstitutionalisation implies that community mental health care is preferable to hospital care and that treatment functions may be performed equally well or even better outside a hospital, one of the main objectives of community-based care consisted in preventing hospital readmission. Despite many objections readmission data and length of stay become most popular as outcome criteria in the evaluation of treatment measures. However, until today evaluative research has failed to demonstrate the overall effectiveness in preventing inpatient treatment. This is mainly due to the fact that the complexity of the research topic could not be adequately modelled and controlled in observational studies as the main source of information.

[The Mannheim long-term study of schizophrenia. Initial results of follow-up of the illness over 14 years after initial inpatient treatment]. / Mannheimer Langzeitstudie der Schizophrenie. Erste Ergebnisse zum Verlauf der Erkrankung über 14 Jahre nach stationärer Erstbehandlung.

A 14-year follow-up was accomplished on 56 out of a total of 70 patients first admitted to hospital with Schneiderian first-rank symptoms. Data collected during 5 years after index admission were also at our disposal. Three quarters of the probands are living alone, and barely one third are in regular employment. At the time of the 14-year follow-up about one third showed delusions or hallucinations. Almost the same number had psychological impairments and 64% were socially disabled. Comparison with the data collected at 1 year and 5 years reveals no difference in the rates of impairments and symptomatology, but a significant increase of social disability.

When and how does schizophrenia produce social deficits?

The present study is an empirical contribution to the controversy over whether the poor social performance and lower social class of schizophrenic patients are consequences of the illness, consequences of changes in the individuals predisposed to develop schizophrenia or are due to the adverse social conditions that lead to schizophrenia. The study focuses on the socioeconomic status at onset, on the performance of social roles in the early course of schizophrenia by taking age, gender and the individual level of social development into account. In a representative sample of 232 first episodes of schizophrenia age and type of onset.type and accumulation of symptoms and social functioning in the prodromal and the psychotic prephase and at first admission were assessed and analysed for their predictive power concerning social disability 2 years after first admission. In a case-control study expected and observed social functioning from onset until first admission were compared. The subsequent course was followed up prospectively in five cross sections until 2 years after first admission. In women the age at onset was significantly higher than in men, whereas symptomatology and type of onset showed no gender differences. In 73% of the sample the prodromal phase covered 5 years on average, and the psychotic prephase (until the maximum of positive symptoms) 1.1 years. Deficits in social functioning occurred predominantly during the prodromal and the psychotic prephase. The course over 14 years showed stable group trends in social and symptom measures. By the end of the prodromal phase it was possible to predict social disability 2 years after first admission with a correct classification of 81%. The main factor determining social outcome appeared to be the acquired social status during the prodromal phase of the disorder. The unfavourable early course in men was due mainly to their significantly lower age at onset. These results raise questions concerning an earlier therapeutic and rehabilitative intervention.

New perspectives in the epidemiology of schizophrenia.

It was with great pleasure that I accepted the invitation to contribute to the present symposium in honour of my friend Jules Angst. It gives me the opportunity to pay tribute to his outstanding scientific work. The topic of my presentation has been inspired by the genius loci. In his foreword to a proceedings volume entitled 'Die Entstehung der Schizophrenie' (The Origin of Schizophrenia), which he edited on the occasion of the 100th anniversary of the Psychiatric University Hospital Zurich in 1970, Jules Angst pointed out that two former heads of the 'Burgholzli', Eugen and Manfred Bleuler, who had held that office longest, had dedicated their life's work to schizophrenia, which had received its name from Eugen Bleuler. Jules Angst's most successful research field has been psychiatric epidemiology. I therefore thought it appropriate to talk about new perspectives in the epidemiology of schizophrenia.

The epidemiology of early schizophrenia. Influence of age and gender on onset and early course.

For the investigation of the early course of schizophrenia starting from onset, the standardised Interview for the Retrospective Assessment of the Onset of Schizophrenia was developed and validated. In a representative sample of 267 first-admitted German schizophrenics of a broad diagnosis from a population of 1.5 million, the age at which different diagnostic and onset definitions were satisfied, the symptoms at the time of the interview, and the accumulation of positive and negative symptoms until first admission were assessed. Comparison between the two sexes and three age groups yielded hardly any differences in the accumulation of symptoms and their course until first admission, except for a slightly shorter period of negative symptoms in young males and a slightly longer one in older women--which contradicts prevailing opinion. At the time of the interview, no significant sex differences were found with respect to the core symptoms of schizophrenia (negative and first-rank symptoms), but clear and substantial differences emerged in disease behaviour. The significantly higher age at first onset in women is explained, on the basis of animal experiments and a clinical study, by the neuromodulatory effect of oestrogen on D2 receptors and by a higher vulnerability threshold in women.

[A chapter in systematic schizophrenia research--the search for causal explanations for sex differences in age of onset]. / Ein Kapitel systematischer Schizophrenieforschung-Die Suche nach kausalen Erklärungen für den Geschlechtsunterschied im Ersterkrankungsalter.

With the aim of detecting causal processes contributing to the onset of schizophrenic symptoms a systematic search strategy was worked out. One of the few epidemiological findings on schizophrenia consistently diverging from expected values, the sex difference in age at first admission, was taken as a basis and replicated on data from the Danish and the Mannheim case registers by controlling for selection and diagnostic artefacts. Danish psychiatrists turned out to have underdiagnosed schizophrenia to a considerable extent at least in 1976, the year from which the analysed case-register data dated. After the exclusion of alternative explanations, the time when symptoms appeared for the first time and the first acute episode occurred was determined for a representative sample of 267 first-admitted cases with a diagnosis of non-affective functional disorder by using the IRAOS interview designed for this purpose. At any of the definitions of first onset applied the mean age of females was significantly higher than that of males, the difference ranging from 3.2 to 4.1 years. The distribution of onsets across the female life cycle showed a clearly delayed increase at young age and a second, lower peak of onsets at the age of 45-54, whereas the cumulative incidence up to the age of 60 years was equal for males and females. On assessing the plausibility of psychosocial versus biological explanations it was hypothesized that due to the effect of estrogens the vulnerability threshold for schizophrenia is raised in females until the menopause. Animal experiments and postmortem analysis showed that chronic estrogen applications significantly shortened dopamine-induced behaviour and reduced D2 receptor sensitivity in the brain. The applicability of this pathophysiological mechanism on human schizophrenia was tested on acutely schizophrenic females with normal menstrual cycles. A significant negative correlation was found between measures of symptomatology and plasma estrogen levels. Apparently, the manifestation of schizophrenic symptoms is influenced by a sufficiently sensitive D2 receptor system in the brain, blocked by neuroleptics and modulated by estrogens.

Generating and testing a causal explanation of the gender difference in age at first onset of schizophrenia.

Motivated by the lack of knowledge of the pathophysiological processes underlying the manifestation of symptoms in schizophrenia, we have worked out a systematic search strategy. Since epidemiological distribution patterns consistently deviating from expected values provide valuable indications of causal relationships, we chose the higher age of females at first admission for schizophrenia, first reported by Kraepelin and since then confirmed in over 50 studies, as the basis for our study. This unexplained epidemiological finding was replicated on Danish and Mannheim case-register data by systematically controlling for selection and diagnostic artefacts and by testing alternative explanations at the individual stage of the study. To check whether the difference in age at first admission was determined by a difference in age at onset, a representative sample of 267 first-admitted patients with non-affective functional psychosis was examined by using an interview for the retrospective assessment of the onset of schizophrenia (IRAOS) designed for this purpose. Any of the definitions of first-ever onset applied--first sign of mental disorder, first psychotic symptom, first acute episode--led to a significant age difference of 3.2 to 4.1 years between the sexes. The distribution of onsets across the life cycle showed a later increase and a second, lower peak between the ages of 45 and 54 years among females compared with males. The lifetime risk for schizophrenia was equal for males and females. After testing the plausibility of psychosocial versus biological explanations we hypothesized that due to the effect of oestrogens the vulnerability threshold for schizophrenia is elevated in females until the menopause. Animal experiments and post mortem analyses showed that chronic oestrogen applications significantly shortened dopamine-induced behaviour and reduced D2 receptor sensitivity in the brain. The applicability of this pathophysiological mechanism to human schizophrenia was tested on acutely schizophrenic females with normal menstrual cycles. A significant negative correlation was found between measures of symptomatology and plasma oestrogen levels. The manifestation of symptoms in schizophrenia appears to be influenced by a sufficiently sensitive D2 receptor system in the brain, blocked by neuroleptics and modulated by oestrogens.