From psychoanalysis to epidemiology: autobiographical notes


  • Jules Angst

The psychoanalysis of schizophrenia

From the age of 10 years I wanted to become a doctor. During my last years at school in Zurich, in the early 1940s, I discovered the works of Carl Gustav Jung, Sigmund Freud and Alfred Adler and later also of Leopold Szondi. In class I presented a Jungian interpretation of the protagonists of Hermann Hesse’s Narziss und Goldmund as prototypes of introverted and extraverted personalities. I knew that Hesse had been treated by a Jungian analyst in Lucerne.

It was in these years and under these influences that I decided to become a psychotherapist and a psychiatrist and started training as a medical student in the fall of 1946. In the second year of my medical studies, devoted mainly to anatomy, physiology and biochemistry, I attended in parallel the lectures on psychotherapy given by Kurt Binswanger on C. G. Jung, Medard Boss and C. A. Meier on dreams, Bänziger on symbolism, and Jakob Lutz on child psychiatry. Later I also listened to Herbert Binswanger on Neo-Psychoanalysis (the school of Harald Schultz-Henke in Berlin) and the neurologist and psychoanalyst Rudolf Brun. In 1947 I started a Jungian Lehranalyse with Kurt Binswanger and after a little over a year’s analysis became a guest of the Jungian Psychological Club and the Swiss Society of Practical Psychology (SGPP). All this extra training had to be paid for, which I did by working as a carpenter in my uncle’s business. I also did night stints in the main post office and worked as a swimming bath attendant, which financed my first typewriter and first bicycle as well.

In 1951, one year before qualifying as a doctor, Manfred Bleuler, who had supervised my dissertation on the personality of patients with familial diabetes insipidus (1), allowed me to work in his hospital during the semester break for 2 months. At that time in Switzerland, as elsewhere, psychoanalysis dominated large swathes of psychiatry (in the USA all departments of psychiatry were headed by analysts). At Zurich University, psychoanalysis was mainly represented by Gustav Bally and Medard Boss, both of whom worked in private practice but worked hard to strengthen the status of psychoanalysis inside the hospital. The Burghölzli psychiatric hospital was a European centre for the psychoanalysis of schizophrenia. Under Eugen Bleuler and C. G. Jung it had been the first hospital to introduce psychoanalysis into academic psychiatry. Following in these footsteps, Manfred Bleuler invited the psychologist Marguerite Sechehaye [trained in the USA by John Rosen (2)] to Zurich in order to report on ‘symbolic realization’ as part of the direct analysis of schizophrenia. This tradition was upheld at the Burghölzli from 1947 to 1956 by the impressive activity of Gaetano Benedetti from Catania. We even had Finnish colleagues who came to Zurich for training and worked as psychotherapists in our hospital for years.

And during my stay as a student in the hospital I was allowed – under Gaetano Benedetti’s supervision – to treat psychoanalytically a mentally subnormal schizophrenic male patient with an acute paranoid-hallucinatory psychosis (Pfropfschizophrenia). Until then intensive psychotherapy had mainly been reserved for more intelligent schizophrenic patients. The 1-h sessions took place on the ward 6 days a week and the patient’s acute psychosis remitted slowly.

In that same period while working at the Burghölzli, I met the psychoanalysts Paul Parin and Fritz Morgenthaler, who had both trained in neurology but Morgenthaler worked for 1 year in psychiatry. They both were critical of Bleuler and warned me against spending more than a year in psychiatry, predicting that Manfred Bleuler’s systematic scepticism would ruin me.

In 1952, I finished medical school. One year later, a year in which I started my military service as a doctor in the Swiss Army officers’ training school and worked for a few months as a locum GP, I finally started my training as a psychiatrist under Manfred Bleuler, with a paid post as a junior doctor, something which was very difficult to get at the time.

I continued to treat some schizophrenic patients with intensive psychotherapy under Benedetti. Interestingly, on one occasion C. G. Jung invited us young doctors to his home and encouraged us in this approach, although he also believed that the disorder was associated with an endogenous intoxication of the brain. One day I was asked by the Jung Institute to give a presentation on such a therapy, but Manfred Bleuler would not allow it: he thought I was too young! Over a period of two and a half years my psychotherapeutic treatments (not only of schizophrenics) were controlled in parallel by the three famous schools (Jungian, Freudian and existential) in group sessions of analysts. It gradually became evident to me that all the treatments based on these three very divergent theories led to the same results and that there must be serious flaws in the theories themselves. I was no longer able to identify with any single analytical school, although originally I had felt myself to be a Jungian analyst.

  • image

Over the years I saw no convincing effects of all our efforts with schizophrenic patients. Based on these facts I decided to give up the analytic treatment of schizophrenic patients. At one of the weekly staff conferences I had to present to Manfred Bleuler the previous history and clinical findings of the schizophrenic daughter of a rabbi, whom he knew personally. After Manfred Bleuler had interviewed the patient himself, he asked me whether I would have Zeit und Lust to treat her with psychoanalysis; my response was that I had the time but no wish to do so. I think Bleuler was rather surprised, if not shocked, by such an unexpected refusal. I then treated the patient on my ward in the usual way with drugs until she could be discharged. This marked the end of my career as a psychoanalyst of schizophrenia in the hospital, but I continued throughout my life as a doctor to treat some patients with psychotherapy whenever this was appropriate.

Some years later, in 1961, Christian Müller (3), who had not himself treated any of the patients concerned and was therefore unbiased, published a follow-up study of all those schizophrenics who had received psychoanalysis in the Burghölzli and the out-patient department. Manfred Bleuler concluded that the outcome was identical to conventional treatment.

These were also interesting years due to the initiative of Medard Boss, who each year would invite Martin Heidegger to Zurich for a few weeks for the ‘Heidegger seminars’ which were held in his house overlooking the lake of Zurich. We had all read at least Heidegger’s Sein und Zeit (4). The seminars were later published, with the exception of one, and I like to think that this was because of me. Heidegger had stated that a main difference between human beings and animals was that humans could look into the future, whereas animals only lived in the past and the present. I did not agree and stressed the similarities between animals and humans, arguing that a horse, for instance, which wants to jump over a ditch has to estimate the distance and height in advance because in the air it can no longer change its trajectory. Of course Heidegger never accepted my argument. The published record of the seminars merely noted that this one failed, without giving any reasons.

During this period I was very aware of the fact that many clever people were investigating schizophrenia, whereas the affective disorders were considered to be much less interesting and were a relatively unexplored field, so that was where I decided to put my efforts.

Manfred Bleuler

  • image

Manfred Bleuler was a great teacher and my model of a good clinician and highly critical scientist. He knew practically all the patients in the hospital (at that time up to 530), whom he visited daily, including weekends. He was an excellent psychopathologist and diagnostician. No patient ever received a diagnosis without prior physical examination and a history taken from relatives or significant others. Bleuer stressed the importance of carrying out long-term follow-up studies in order to judge the results of treatments. He had the idea of examining the response of relatives to imipramine (pharmacogenetics) and he encouraged me to do family studies as he had done with alcoholics. He was so sceptical that he never trusted any new treatment, whether psychotherapy, neuroleptics or antidepressants. His belief and value system was based on work, which gave him a high regard for occupational therapy. His staff used to joke that he would have liked the women patients to be issued with their knitting needles already in the admissions room. And he was very thrifty; originally he was hierarchically superior to the administrative director (today it is the reverse). Manfred Bleuer had been an officer in the Swiss army during the Second World War, when soldiers were not given butter for breakfast. In character, he applied this frugal rule to the hospital staff and patients. It was only after the change in hierarchy that butter was introduced.

When Bleuler wrote his important monograph on the course of schizophrenia (5), he was convinced and wanted to show through his data that a schizoid premorbid personality would deteriorate the long-term outcome. He did not know much about statistics and asked me to do the simple chi-squared calculation, which turned out to be non-significant. Over the next few weeks he gave the same task to a professor of bio-statistics in our medical school and finally to a professor of mathematics and statistics of the Swiss Federal Institute of Technology; the finding continued to be negative and he was finally willing to change his mind.

With the introduction of neuroleptics, most hospital wards could now be opened; but Bleuler did not want the walls around the hospital to be dismantled, arguing that the patients could then be seen in the gardens. In the end against Bleuler’s wishes the walls were demolished on the administrative director’s instructions.

Although Bleuler himself had trained in the USA, he effectively dissuaded me from taking up Lawrence C. Kolb’s invitation to join his staff at the Psychiatric Institute of Columbia University in New York, telling me that on my return my post would not be guaranteed. I learnt from that experience to take the opposite approach and to encourage young colleagues to grasp such opportunities.

There is no doubt that I owe my career to Manfred Bleuler’s training and continued backing (in 1967 he gave me a research position as an assistant professor) and later to Konrad Akert and Erwin Ackerknecht, who in 1969 proposed me to the medical school for the post as an independent research director. This was a complete innovation in Switzerland, a position which allowed me, supported by Niklaus Ernst, the clinical director, to devote more than half my time purely to research. Otherwise, I was always cautious not to accept positions I did not feel competent or motivated to hold, like dean of the medical school, head of a planned mega-department of psychiatry (a development which I contributed to thwarting); I was never interested in positions of administrative power.

Depression and the fate of my monograph

In 1951 I experienced psychiatry without any modern drug treatment. Schizophrenic patients were treated with ECT, Insulin shock treatment and strong sedatives (scopolamine combined with morphine and barbiturates). Depressed patients were given ECT, opium drops, amphetamine and barbiturates against insomnia. Life in the hospital was dangerous for both patients and staff. On closed wards, the patients ate with spoons only, knives and forks were considered too dangerous. The walls were high and there was no great therapeutic optimism, especially not among the nurses.

I was still in training as a young psychiatrist when chlorpromazine and reserpine were first used in our hospital in 1953. My first studies in psychopharmacology were devoted to the hematological side effects of iproniazide, used in internal medicine against tuberculosis and multiple sclerosis (6); psychoses induced by disulfiram treatment for psoriasis in dermatological patients (7); and disulfiram psychosis in patients with alcohol dependence (8).

My real psychopharmacological research started in 1959 when I worked full time for 18 months conducting a large open study on 200 depressive in- and out-patients treated with imipramine – a period during which I very nearly became depressed myself. Follow-ups of the hospitalized sub-sample enriched by all admissions for depression or mania over 5 years (1959–1963) were carried out until 1985 and the mortality was studied until 2003 (9–11). In 1961 my data was punched by Geigy in Basle (Dr Peter Weis) and we did most of the counts mechanically by entering the punch cards. Peter Weis was a chemist and statistician and made me familiar with some computational possibilities, especially multiple regression; I quickly became a computer enthusiast and when I took over the chair and hospital from Manfred Bleuler in 1969 I immediately set about computerizing the admissions statistics. We were the first hospital in the University of Zurich to do so.

My studies on imipramine in the late 1950s included the first pharmacogenetic study on familial therapeutic response, confirmed by Michael Pare’s study on MAOIs, and, especially, a systematic study of the heterogeneity of ‘endogenous depression’. Although Kleist (12) and his pupil Leonhard (13) had distinguished between uni- and bipolar disorders, the genetic evidence brought forward in the monograph of Kleist’s pupil Neele (14) had not convinced the scientific community. She had shown that the relatives of bipolar (zweipolige) patients had a higher load of psychoses in general than the relatives of unipolar (einpolige) depressive and unipolar manic patients.

At that time, schizo-affective psychoses were considered to be a subgroup of schizophrenia by ICD and also by Bleuler, and involutional melancholia to be an independent disorder according to Kraepelin and the ICD. A hospitalised endogeneous depressive sample of 331 patients formed the basis for my main genetic study on depression, mixed (schizo-affective) psychosis and involutional melancholia (15). I knew most of the patients personally as well as some of their relatives. I collected data systematically on the relatives: year of birth, sex, psychiatric problems and diagnoses. These data were extensively supplemented by information from other institutions, GP records, etc. Finally, I did a follow-up of all patients, mostly by phone, and on this occasion contacted the relatives again.

Statistics was not then taught at medical school, so I had to teach myself. I punched the data on cards (Sichtlochkarten) and evaluated the data by hand. I bought a basic calculator, a machine which could add and subtract: divisions had to be performed stepwise by subtractions on the basis of digits. Everything was extremely time consuming. Most work was performed over the weekends and in the evenings. From today’s perspective, the statistics applied in my monograph are outdated. However, nevertheless I could compute the morbid risk of relatives in per cent (corrected for age). Manfred Bleuler then gave me a few months’ leave to write the monograph up.

My idea was to look for the heterogeneity of endogenous depression. I systematically tested all sorts of clinical characteristics as subclassifiers of depression, and I used genetic morbid risk of affective disorders among first-degree relatives, sex distribution and course characteristics as validators. The article of Robins and Guze (16) on validators had not yet been published. As potential classifiers for depression, I used age of onset, number of episodes, precipitation by stressors, psychopathological syndromes (psychotic; agitation; retardation; hypochondriasis; hypomania/mania and bodily constitution (leposome, pyknik etc.).

The best classifier turned out to be the presence or absence of hypomania or mania in depressed probands. I called the depressives with hypomania/mania ‘circular’ according to the terminology of Falret (17) (folie circulaire). I found marked differences in two validators: i) sex distribution; the male and female relatives of probands with circular depression were equally affected by mood disorders, whereas, among the relatives of depressives, there was a strong preponderance of females; ii) another main finding was that depressives with hypomania or mania had an increased morbidity risk not only for depression but also for the circular subtype, whereas the pure depressives did not have an elevated risk of the circular subtype but only of depression. In terms of course, the circular subtype of depression showed more and shorter episodes than pure depression.

In contrast to Perris (18) I found the risk not only of bipolar disorder but also of depression to be increased among the relatives of bipolar patients. I assumed that both genetic and environmental factors would play a major role. This hypothesis was derived from Manfred Bleuler’s concept of the multifactorial (genetic and environmental) genesis of mind and especially also of schizophrenia (19). Breaking down subjects by (decades of) age of onset, I found that the genetic load was inversely correlated with the precipitation of depression: somatic and psychological precipitation were rarer in subjects whose depression had an early as opposed to a later onset and, conversely, the genetic load decreased systematically with increasing age of onset of depression. On the basis of these findings, I proposed abandoning the concept of involutional melancholia, which played a role in Kraepelin’s classification and still survived in the ICD.

The sample also included depressed patients who had received the diagnosis of mixed psychosis or who had schizophrenic features. The genetic investigation of these patients showed that there was a genetic continuum of subtypes from depression to schizo-affective in terms of affect-dominant and schizo-dominant subtypes of mixed psychosis (schizo-affective).

My monograph on this question was accepted for publication in 1965, but to my great alarm and distress three great authorities of clinical psychiatry in turn, Manfred Bleuler, Erik Stroemgren and Sir Aubrey Lewis, did not believe in the results. Manfred Bleuler told me in 1965 that he had discussed my findings over the phone with Erik Stroemgren and they had agreed that they must be false, for otherwise they would have been described long ago. This came as a real shock, and for one long year I had repeated visions of the ruin of my scientific career.

In the late 1965, during my 3-month stay at the Psychiatric Institute in London, I presented my conclusions to the director, Sir Aubrey Lewis, who did not believe me at all either; he was a convinced unitarian. Finally, to my intense relief, two geneticists, Eliot Slater and James Shields, who had their offices in a provisional wooden hut in the grounds of the Institute, found my data interesting and thought them to be real and true.

It took another 6 months after my monograph was published for Manfred Bleuler and Erik Stroemgren to change their minds under the influence of Carlo Perris’ monograph (18); Perris in Sweden also demonstrated that there was a qualitative and quantitative difference in the morbid risk of bipolar and depressive patients’ first-degree relatives. Manfred Bleuler was kind enough to invite Perris to Zurich, and together we wrote a paper comparing our results and became friends (20). Our findings received further confirmation when Winokur et al. (21) in the USA published their monograph on the dichotomy of unipolar and bipolar disorders. In this way bipolar disorder, which was first conceptualised as a specific psychiatric disorder by Falret (17) in 1851, was re-born (22).

My 1966 monograph also showed that involutional melancholia (as mentioned earlier, the Kraepelinian diagnostic entity still survived in the ICD of the time) was none other than late-onset depression, and that, from a genetic point of view, schizo-affective psychoses, differentiated into three sub-groups (a core group, and schizo-dominant and affect-dominant groups), took an intermediate position on a spectrum between affective and schizophrenic psychoses.

These patients were followed up throughout their lives and formed the nucleus of several multicentre studies on the natural history of affective disorders with special emphasis on their recurrence over lifetime (23–25).

Nathan S. Kline and psychopharmacology

  • image

Nathan Kline, too, was an important figure in my life, a man of great energy and humour, who was curious about everything that was new. He was one of the pioneers in clinical psychopharmacology and introduced the first MAO inhibitor, iproniazide, into psychiatry. He was also a pioneer in the use of reserpine, clozapine and lithium in the USA. Kline was an idealist and extremely generous with his time and money, for example founding psychiatric facilities in Haiti and Liberia and helping as many as 200 refugee colleagues from Eastern Bloc countries to start a new life in the USA. In the 1960s and 1970s, there were few ways in which scientists could meet: Nathan Kline founded the Denghausen Group – sponsored by a sculptor whose wife he had successfully treated – which met mainly in the Caribbean and provided a forum for researchers into depression (26). Nathan Kline was one of the first to recognize the implications of my monograph and had it translated and published in English.

Kline’s style was at the antipodes of Bleuler’s. Kline organized and chaired the first international symposium on psychopharmacology during the World Psychiatric Congress in Zurich in 1957, which I could not attend because I was doing my periodic service in the Swiss army. After the congress, Manfred Bleuler commented briefly that Kline had been behaving like a clown at the congress banquet, but Nathan Kline himself later described the whole scene to me somewhat differently. Bleuler was worried about the costs of the congress and seemed to want to be very economical in a typically Swiss way, although he was himself a wealthy man. Bleuler and Kline were sitting together at the banquet and at the end of the meal Nathan Kline lit up a cigar with a rolled 10 dollar note deliberately to shock his neighbour, who did not realize that the note was a fake.

Identifying myself with Manfred Bleuler’s scepticism, I have always had problems with what I regard as the myths and misbeliefs about drug action; these are still playing a big role in psychopharmacology, despite modern randomized controlled trials, and have even been partly integrated into important treatment guidelines.

One such myth is ‘drug-induced hypomania’, a hypothesis which is not based on empirical evidence but which is a good example of autistic thinking as defined by Eugen Bleuler (27). Switches to hypomania can only occur in response to treatment, non-responders, who prevail among placebo-treated patients, cannot switch; the statistics have therefore to correct for responders (28); unbiased meta-analyses were negative (29, 30). Another example is the hypothesis of the delayed onset of action of antidepressants, which is not based on proper statistical analysis; in fact, the latter shows an early onset of action within the first days of treatment (31, 32).

More amusing is the story of clozapine. In 1968, the AMP group began a multicentre study in university psychiatric hospitals in Austria (Vienna), Germany (Berlin) and Switzerland (Zurich), which resulted in the detailed description of the antipsychotic effects of clozapine, its lack of extra-pyramidal effects and its comparable efficacy to levomepromazine in the treatment of schizophrenia and mania (33, 34). Wander, the company which developed clozapine was taken over by Sandoz, which proceeded to organize in Berne a big meeting on the drug for its managers worldwide. After we had presented our results, they concluded that if it had no extra-pyramidal effects, the drug could not be effective. This was the first attempt to kill off clozapine, the second was a consequence of agranolocytosis observed in Finland. The drug was withdrawn from the market in most countries but remained available for hospitals in Switzerland: their directors insisted that it would be unethical not to produce the drug for patients who had not responded to conventional neuroleptics. Many patients continued on clozapine after discharge and the drug slowly came back to the market throughout Europe. Clozapine remained my first-line treatment for mania and schizophrenia for many years.

In 1988 John Kane et al. had the merit to demonstrate again the clozapine’s usefulness in treating schizophrenia and together with Herbert Meltzer even showed that it had a suicide preventive effect (35). Later clozapine’s suicide preventive action was also shown in bipolar disorders (11) and we also found some evidence that it has a neuroprotective effect in bipolar patients (36, 37). More research on long-term medication and dementia is urgently needed.

In the 1980s and 1990s, double-blind trials became routine and the methodology rigid without innovation, which made the field increasingly uncreative. In contrast to earlier decades when methodology had been in the hands of the clinicians (e.g. the develoment of the AMP systems and the CPRS), the methodology of trials was taken over by the drug companies and focused on the requirements of the FDA, so reducing the invention of new measures. Moreover, in Switzerland at least there was a very strong political anti-psychiatric movement supported by the mass media, which tried to stop drug treatment even in hospitalized patients who needed it. This opposition spread to doctors and nurses in our hospital and seriously impaired both treatment and research. Patients were encouraged to refuse enrolment in research projects. I was a convinced pharmacotherapist and was dubbed Pillen-Angst by my critics: psychotherapy continued to be considered by many colleagues as the most fundamental and humane treatment of psychiatric patients. In the magazine ‘Red Infusion’ I was attacked for treating private patients in the hospital (two-class medicine) and was incidentally ‘promoted’ from major to colonel in the Swiss army, making me a ‘military chief’, an even more sinister target. Plans to carry out biological research on schizophrenia together with the Brain Institute had to be halted; the necessary CSF had to be collected in Munich. Systematic trials were stopped in this context. This situation was a contributory factor in my moving away from clinical to epidemiological research.

My path to epidemiology

My first personal contact with epidemiologists occurred in 1965 during my stay at the Psychiatric Institute in London under Michael Shepherd and listening there to the lectures of John Wing. Michael Shepherd pointed out to me that Switzerland had no epidemiological studies comparable with those carried out in Camberwell (38, 39), although our small country seemed to be particularly suitable in view of the stability of its population. At that time, I was far too much of a clinician to recognize the importance of Shepherd’s remarks.

In 1967 as an officer in the Swiss army, I was appointed (rather unwillingly because I loved the service in the Alps) as chairman of one of the many small cantonal doctors’ committees which oversaw conscription of Swiss men into the army. I did this job once a year until 1987 and with hindsight I can see the irony of my initial reluctance.

At the age of 19 years all Swiss men had to be screened for military service. In my area, the cantons of Zurich and Schaffhausen, where about one-sixth of the Swiss population lived in cities and pre-alpine areas, this meant examining over 6300 males per year. During my annual 3- to 4-week period of duty, we examined about 50 men per day both physically and psychologically, on the basis of a sports test and a signed personal previous medical history accompanied by all school, medical and criminal records. In doubtful cases, we could postpone the decision on suitability for military service and refer the potential recruit to the University medical services. Apart from physical handicaps, the main reasons for exclusion were low intelligence and psychiatric disorders. In this way, I saw and briefly interviewed about 1000 men every year and was increasingly struck by the great variability in their physical maturation (some looked like 15-year-old adolescents, others like 30-year-old men) and by their enormous psychological variability.

In addition, we had to see all 19-year-old prison inmates, which took a few days; they were brought to us in groups of about 20 under guard. They were not usually suitable for army service but examining them broadened my view of human beings as did my experience as a part-time psychiatrist in a prison for serious offenders.

All this brought home to me how restricted our clinical experiences are, how selective professional medical training is, and that we can never know enough about the variability of human beings, especially in their longitudinal development. Moreover, working mainly in the hospital context, I became aware of the very great changes in admissions over the decades: the patients were becoming increasingly unrepresentative of the disorders, for instance being highly selected by therapy resistance, suicidality and violence. From a research viewpoint, this was a serious limitation and I decided to start working on army conscripts using epidemiological methods.

My ideal of epidemiology was based on the prospective studies carried out in Scandinavia (40). I greatly admired the famous Lundby Study, which had inspired me to start long-term prospective investigations. In a similar way, the Stirling County Study of Alexander H. Leighton (41) and Jane M. Murphy (42) encouraged me to use questionnaires in addition to a structured interview. In September 1981, I managed to visit Ole Hagnell in Lund, who had taken over the Lundby Study from Erik Essen-Moeller. Ole drove me around the country areas where the original population had lived and I was very pleased to get a direct impression.

In Germany Heinz Häfner, who in 1975 founded the Central Institute of Mental Health in Mannheim, became a leading figure in social psychiatry and epidemiology, focusing mainly on longitudinal studies of schizophrenia. He recently reported that depression often antecedes schizophrenia (43). During the 13 years (1973–1985) in which I was on the reviewing committee of one of his major epidemiological projects, sponsored by the German Science Foundation (DFG), I learnt a lot and became very familiar with his excellent methods applied to his Mannheim case register. These records had finally to be destroyed for political reasons. On Heinz Häfner’s initiative, I received an honorary doctorate in Heidelberg in 1988.

Around 1970, the use of cannabis became a growing problem in Switzerland. At conscription I saw many intelligent adolescents who had dropped out of school, and I devoted my first large epidemiological study to this important problem. This study started in 1971 when I got permission from the Swiss army to carry out a research project supported by the Swiss Science Foundation. Over a period of 1 year, we examined the whole birth cohort of 6315 nineteen-year-old men (from cantons Zurich and Schaffhausen) at conscription in separate rooms using questionnaires. The participants received written confirmation signed by me that any information they supplied would be covered by medical secrecy and nothing reported to the army. The questions referred to socio-demographic data, physical and psychological problems, school performance, consumption of legal and illegal substances and delinquency, and we included a 213-item personality test. Half the participants, randomly selected, had to give their identities and the others remained anonymous. The statistical comparisons demonstrated no significant differences between the two groups, except in church attendance, which was more frequently claimed by the non-anonymous group (social desirability). Interestingly, we identified much higher rates of delinquency than those officially recorded. The main focus was on substance use; ultimately the study followed the cohort up to the age of 36 years, with analyses of the course of illicit substance use, delinquency and suicide. We published a 3-year follow-up (44) and a 12-year follow-up of substance use disorders (45), and an analysis of the true premorbid personality measures of patients who later developed unipolar depression, bipolar disorder and schizophrenia or committed suicide (10, 46). Just recently another follow-up of delinquency in the same cohort up to the age of 55 years has appeared (47).

My interest in epidemiological research derived partly from my readings on transcultural psychiatry and my visiting professorships in Australia and New Zealand (1976) and South Africa (1979). I was aiming at a more holistic study of nature. I became convinced that psychiatric disorders could not be investigated without studying somatic symptoms and syndromes. I did not believe in the mind/body split, or in ‘somatisation’; I have believed since then that western psychiatry is biased by a psychologizing intellectual approach which neglects somatic symptoms and the unity of mind and body. Western psychiatrists psychologize and assume that their patients somatize but this is not so: somatic symptoms are part and parcel of psychological disorders. What was called ‘masked depression’ is true depression (48). As a consequence, the later Zurich Study integrated more than 12 somatic syndromes.

David Goldberg shared this view to some extent and has investigated somatizers vs. psychologizers (49, 50). I had always admired his work on mental illness in the community (51) and his development of the General Health Questionnaire. In February 1986, he invited me to Manchester. In the late morning of 14 February 1986, David gave me a tour of his department, during which a patient phoned to warn him that a bomb would explode during my lecture. He reassured me that nothing would happen, as the patient’s doctor would be there and the doors to the event would be secured by the police, who had been alerted but was not inclined to believe it. In the afternoon, we passed through the police controls and I started my lecture; after about 10 minutes, my lecture was interrupted by loud shouting. David took over and the treating doctor talked to the patient, persuaded him to leave and the police took him away. It turned out that this was the same man who had earlier entered the bedroom of the Queen of England on the first floor of Buckingham Palace, an event which received a lot of publicity. He was a Swiss, like me, but living in Manchester and married to a British woman. Fortunately, for the reputation of the police there was no report in the newspapers the next day.

The Zurich study

My main, still ongoing, epidemiological study, which came to be called the Zurich Study, began in 1975 with a series of methodological pre-studies in small samples (200–500) of the conscripts, in which I tested several instruments and confirmed my growing interest and enthusiasm for epidemiological research. David Goldberg had already developed the General Health Questionnaire (GHQ) as a valid screening instrument for mental disorders (52). We used the GHQ as a validator in order to test the SCL-90 R as a new screening instrument in one of our Zurich Study pre-studies (53). The high correlation (r = 0.81) was decisive for the application of the SCL-90 R prospectively in our study; compared with the GHQ the SCL-90 R gave 9.6% false positives and 4.0% false negatives. Finally, I chose the SCL-90 R because it had a broad spectrum with nine subscales of psychopathology.

I was also looking for a suitable interview for the embryonic Zurich Study but could not find any which met my ambition to assess nature as it is, from normality to disorders, and so to include: i) functional somatic syndromes and ii) not just the diagnostic groups but mild and subdiagnostic psychiatric syndromes. As a result we had to develop our own instrument, the SPIKE (54, 55), which at that time comprised 26 syndromes ranging from gastro-intestinal, cardio-vascular and respiratory syndromes to allergies, headache, sleep, exhaustion, menstruation, sexuality and so on, in addition to psychiatric syndromes (anxiety, phobias, depression, suicidality, obsessive–compulsive syndromes and substance use.

In another small prestudy, we checked the reliability of the previous history given by 100 subjects from the community at the age of 20 years comparing it with the documented visits and treatments by school doctors, psychiatrists and psychologists. It turned out that both externalizing and internalizing problems, which led to consultations, were in 25% no longer remembered/reported at the age of 20 years. On account of this ‘forgetting’ lifetime prevalence rates based on one or two assessments have to be considered as minima. Forgetting was a further reason for planning a prospective study with multiple interviews over decades comparable with the recent Christchurch study from birth to age 20 years (56).

Supported by the Swiss Science Foundation, we were able to start the Zurich Study in 1978. This prospective epidemiological investigation of depressive, neurotic and psychosomatic syndromes was based at outset on a sample of 4547 young subjects (male 2201; female 2346) representative of the canton of Zurich in 1978 (population 1.1 million). All subjects were screened with the Symptom Checklist 90-R (SCL-90-R) The screening took place in 1978, when the male participants were aged 19 years (at conscription) and the females 20 years (at their enrolment on the electoral register). In order to enrich the sample with cases at risk of the development of psychiatric syndromes, a stratified sampling procedure was chosen: a subsample of 591 subjects (292 male and 299 female) was selected for interview, with two-thirds consisting of high scorers (i.e. above the 85th percentile on the Global Severity Index of the SCL-90), and one-third being a random sample of those with lower scores (below the 85th percentile). Details of sampling procedures and refusal rates were presented in the first publication (54) and recently again (55).

Due to the stratified sampling, with many more high risk cases of psycho-somatic symptoms than controls, the sample is representative of 2600 persons of the canton of Zurich. The subjects were followed by six interviews up to 1999, when they were 40/41 years old; a seventh interview wave has now been launched with subjects who are 49/50 years old; it is being conducted by Wulf Rössler and his team.

As I have mentioned, the Zurich Study subjects were first investigated at the age of 19/20 years; if I had my life over again, I would want to be a child–adolescent–adult psychiatrist and plan to study development lifelong from birth on, embracing physical and psychological health (limited by my own half-life).

  • image

The Zurich Study has more than any other clinical experience changed my perspective on the normal development of human beings and their psychiatric disorders; it led me to change my view of the separation of pathology from normality to a continuum (spectrum) approach and to relinquish the dichotomy of Kraepelin (manic-depressive insanity vs. dementia praecox) in favour of a diagnostic spectrum, including subgroups of schizo-affective disorders as a bridge between them (57–59). The evidence of the continuum from normal to pathological was so obvious that only a dimensional view and dimensional measures could account for it. Apart from sleep, mood and anxiety disorders, the Zurich Study recently found this phenomenon in the effects of cannabis consumption, which created a remarkable amount of schizophrenia-like symptoms under the threshold of a diagnosis, proportional to the consumption (60). An analogous and amazing finding is that adolescents and young adults in love scored as high as bipolar II patients on the hypomania checklist-32 compared with controls not in love.

Stigmatization thrives on the absence of a dimensional view of nature and on ignorance. Among a certain group of Amazonian Indians severely affected by a parasitic facial skin disease, the healthy, unscarred, minority were considered abnormal, were stigmatized and were less marriageable.

We are all in the same boat, but we focus mainly on inter-individual differences, although we share far more than we notice. Epidemiological studies and biological research should gradually help to eliminate stigmatization.


I thank Professor Konrad Akert and Professor Erwin H. Ackerknecht (1906–1988) for having proposed me in 1969 as director of research and my friend Prof. Dr Niklaus Ernst, who took over most clinical responsibilities from 1970 to 1989 as the clinical director, for his continuous friendship until today, which gave me a wonderful freedom and atmosphere for my work.