ÐÓ°ÉÔ­´´

Defining ‘madness’, whatever form it takes, has throughout history been a difficult task; it seems that the vagaries of the mind are more elusive than those of the body. Partly as a result of this, the treatments prescribed for people suffering from different forms of mental illness have veered from the ineffective to the downright harmful. But in the past decade a new definition of schizophrenia has evolved which takes into account both environmental and genetic causes – and with it, a new way of helping sufferers.

People labelled ‘psychotic’ suffer the most severe forms of mental illness, in which they lose touch with reality. They experience delusions (false beliefs that are out of step with those of the patient’s social group and cannot be corrected by evidence) and hallucinations (sensory experiences without a valid external source). A person suffering from some of the more common delusions may believe that he or she is somebody of great importance, such as the prime minister or Christ, or is being persecuted by an organisation such as the FBI or the Mafia. Hallucinations can involve any of the senses, but hearing the voices of people who are not present is the most common form.

These severe mental illnesses have been described over the past 2500 years. Examples include the story of Nebuchadnezzar in the Old Testament, clinical descriptions in the Vedic texts of ancient Indian medicine dating from more than 2000 years ago, the story of Christ healing the man possessed by devils as told by Mark and Luke, and Edgar’s simulated madness in King Lear (‘Tom’s a cold.’ ‘Beware the foul fiend’).

The popular concept of ‘madness’ as an homogeneous entity matched professional ideas about severe mental illness until the end of the 19th century, when Emil Kraepelin divided it into ‘manic-depressive insanity’ and schizophrenia. He did not coin the latter term himself – Eugen Bleuler did, some 20 years later – but called the condition ‘dementia praecox’. This phrase suggests a progressive deterioration in mental functioning, but Kraepelin was aware that some patients made a good recovery. Several long-term studies, including one by Bleuler’s son Manfred, revealed that a surprisingly high proportion of patients – about a third – make a good recovery from schizophrenia and achieve a reasonable adjustment to life. At the other end of the spectrum lies a small proportion of patients, less than 10 per cent, who fail to respond to any of the available treatments and experience persistent delusions and hallucinations for the rest of their lives. Even in this unfortunate group of people, however, there does not appear to be a deterioration in their mental function over time, though their social adjustment usually suffers.

Although there is no progressive dementia in patients with schizophrenia, it has long been suspected that there is an underlying abnormality in the structure of the brain and/or in its function. Within the past decade researchers have discovered brain abnormalities in people with schizophrenia. Some, but by no means all, patients have abnormally large ventricles – the fluid-filled spaces in the brain – suggesting that brain substance is smaller than normal. The frontal parts of the brains of people with schizophrenia also seem to use less oxygen and glucose than normal. But these studies do not yet provide a precise picture of the brain abnormality that gives rise to the distorted experiences of people with schizophrenia, nor do the findings point clearly to the cause of schizophrenia.

The most obvious feature of schizophrenia is that it runs in families. This observation led to two opposed schools of thought early in the 20th century. The geneticists, such as Seymour Kety and Einar Kringlen, conducted a series of family studies using increasingly sophisticated epidemiological techniques, including studies of twins who were either monozygotic (from the same egg) or dizygotic (from two different eggs), and children born to schizophrenic mothers but adopted shortly after birth. These studies have demonstrated the unequivocal – but not paramount – role of inheritance in the development of schizophrenia. There remains enough latitude for environmental theories to burgeon, and there were theorists aplenty in the 1950s. Coming from psychoanalytic backgrounds, Theodore Lidz, Gregory Bateson, Lyman Wynne and Margaret Singer in the US, and RD Laing in Britain, each evolved a different theory linking the origin of schizophrenia with some form of abnormal parenting. A weakness of all these theories was their failure to account adequately for healthy siblings growing up in the same family atmosphere as the individual with schizophrenia. The theories had to postulate that the parents behaved very differently to the child who would later develop schizophrenia than they did to his or her siblings. What is the scientific evidence to support these theories?

In 1975 Steven Hirsch and I reviewed more than 200 studies. Many of these failed to find significant differences between the families of schizophrenic or neurotic patients and those with healthy offspring. Where differences were found with respect to some quality, such as distorted communication, families with a schizophrenic member exhibited more of it, but it was not absent from the other families. So these factors could not account for the nature of the schizophrenic illness, although they might conceivably contribute to its origin. In general, there has been a retreat from the extreme environmentalist position. But the idea that parents cause schizophrenia was remarkably pervasive among psychiatrists in the US and Europe, and regrettably led to harsh and dismissive behaviour towards relatives.

In the US, the pendulum has swung towards the extreme biological position, partly impelled by the excitement of the new brain imaging techniques, such as magnetic resonance imaging, and the rapid pace of developments and discoveries in molecular genetics. But it is still possible to take a more moderate position. In this view, schizophrenia is seen as a manifestation of an underlying brain fault which renders the individual very sensitive to environmental stress. The build-up of stress triggers the onset or recurrence of the illness, the form of which is determined by the brain abnormality, which may be genetic in origin or due to some other biological defect. Progress has been made over the past 30 years in identifying the particular stresses which throw a person with schizophrenia off balance.

Stress is a very general notion, which must be carefully defined if it is to be accurately measured. George Brown of Bedford College, London, developed an interviewing technique for detecting the occurrence of life events. These are sudden, discrete happenings in a person’s life which would require them to adapt psychologically; examples are the death of a close relative, moving house or the loss of a job. Brown realised that it is important to distinguish between events that are brought about by the person’s behaviour, such as dismissal because of persistent lateness, which might be an early sign of the appearance of schizophrenia, and those which were truly independent, such as dismissal because the employer’s business was going into liquidation. When this refinement was included, a number of studies found that independent events clustered in the two to three weeks before the onset or relapse of schizophrenia. One of these studies was an international collaborative project conducted in nine different countries by the WHO. The finding was replicated in six of these countries, despite wide variations in the patients’ ways of life.

Other studies used the life events questionnaire and found clustering of events some time before the onset of a number of disparate developments – depression, heart attack and accidental injury to the subject’s children. The crucial period varied according to the condition, indicating that the psychological processes set in train by an event take different lengths of time to act. But the general conclusion must be that life events are nonspecific stress factors, determining the timing of an episode of illness but not its form.

I was recently asked to give my opinion in a court case involving a young man whose first episode of schizophrenia came a few weeks after his family’s home was destroyed by fire. Could this severe event have caused the illness – in which case the family would be entitled to financial compensation from those responsible for the fire? I had to ask myself, if the fire had not happened, would this young man have lived out the rest of his life without developing schizophrenia? There can be no certain answer to this question, but the likelihood is that some other severe event would have come along sooner or later – none of us is protected from life events – and precipitated a schizophrenic illness. This reasoning assumes that a susceptibility to schizophrenia existed in this young man and was revealed by the psychological impact of the fire. This assumption is currently untestable, as we lack any index of susceptibility to schizophrenia, but there is a reasonable expectation that molecular genetics might provide one. On the basis of this reasoning, I argued that the timing of the onset of the schizophrenic illness was of great significance because it interfered with the young man’s education and psychosocial development. He was knocked off course in a way that probably would not have happened had the illness begun, say, 10 years later. The family was awarded compensation.

The other form of stress that has been quantified stems from the emotional atmosphere in the home. The measurement technique, introduced by Brown and Michael Rutter at the Institute of Psychiatry in London and further developed by Christine Vaughn and myself, consists of an interview with a relative of the person who has schizophrenia at the time of first onset or a relapse of the illness. The relative, usually a parent or spouse but occasionally a sibling, is asked about behaviour and symptoms shown by the patient in the preceding three months. We audiotape the interview and make ratings from the tape afterwards. The rating scales address four important aspects of the relative’s expression of emotion when talking about the patients: critical comments, hostility, overinvolvement and warmth. Raters listen not only to the content of speech but also to the way in which emotion is conveyed in the voice. In everyday conversation we use rate of speech, volume and tone to express emotions, and raters rely heavily on these to make the ratings. But individuals have their own style of expression, in addition to regional and cultural variations. So, no absolute judgements can be made; instead, raters listen for fluctuations outside the usual range of each particular individual. These judgements may appear subjective, yet after a two-week training course people agree on how to rate a particular interview.

In exploratory studies we found that a patient was significantly more likely to suffer a relapse within nine months of leaving hospital if relatives expressed a high level of criticism or overinvolvement or any degree of hostility towards the patient at the time of a schizophrenic episode. On the other hand, if the relative rated high on warmth and low on negative emotions, the patient was more likely to remain well. So relatives could apparently affect the short-term course of schizophrenia both for good and for ill. We combined the negative emotions into an index of ‘expressed emotion’ (EE), which has now been used in more than 20 studies of schizophrenia in a wide variety of cultures. Most studies have replicated the original findings – that about half the patients in high-EE homes suffer a relapse of schizophrenia over nine months compared with less than a quarter in low-EE homes. But the proportion of households rated as high EE varies dramatically with culture (see Table).

Stress in the city

The highest levels of expressed emotion are found in cities in the West. Intermediate levels occur in rural areas in the West, such as Galicia, and in groups living in cities but retaining traditional family structures, such as Mexican-Americans in Los Angeles and Indians in Chandigarh, North India – although high-EE attitudes are virtually nonexistent among Indians maintaining a traditional way of life in the countryside. The cultural determinants of high-EE attitudes have not been explored in detail, but these findings suggest a number of possible influences. Urbanisation and industrialisation lead to the dissolution of the extended family. Rural households in agrarian societies are often extended both vertically, through three generations, and horizontally, by bringing in cousins and in-laws, whereas the nuclear family, with parents and children only, has become the norm in cities in the West. Moreover, even the nuclear family is shrinking as a consequence of high divorce and separation rates, and today we often find households made up of one middle-aged schizophrenic patient being cared for by a single elderly parent, usually the mother. Negative emotional responses seem much more likely to develop when the whole burden of care falls on a single individual than when it is shared by 10 or more household members.

Another consequence of industrialisation is an increase in the demands made on the workforce. Punctuality and productivity are essential requisites for keeping a job. These habits of discipline are particularly difficult for people with schizophrenia to sustain, leaving few of them able to compete in the open market in the West. By contrast, agrarian economies offer numerous tasks that people can do even if their speech and concentration are impaired by their schizophrenia. Patients benefit from a feeling of achievement and relatives are satisfied with the patient’s contribution.

Another feature of traditional societies that is fast disappearing from the West is the sense of duty to look after sick and elderly relatives. The care of a family member with schizophrenia is accepted as a matter of course, and even when acutely ill, few patients are admitted to hospital. Traditional views of illness, whether psychiatric or other, often embody fatalistic notions such as the belief that the occurrence of illness, along with other misfortunes, is preordained and so beyond the power of the individual to influence. By contrast, in the West there is a strong tendency to hold the person responsible for the bad things that happen to them, including psychiatric illnesses – an attitude summed up by the phrase ‘Pull yourself together’. This could readily give rise to the critical attitudes that are a major component of the high-EE index.

Whether these factors adequately explain the cultural variations in EE requires further study. Because EE is linked to the outcome of schizophrenia one might expect that people with schizophrenia would fare better in developing countries. A number of studies in such countries have indicated that this is the case, despite such adverse conditions as a paucity of medical staff and facilities and difficulties in arranging long-term outpatient care. In a recent study in Chandigarh, people with schizophrenia making a first contact with the psychiatric services were followed up for one year. The relapse rate was found to be half that of a comparable sample in London. Furthermore, there were half as many high-EE relatives in the Indian sample as in the British sample, and this difference entirely accounted for the better outcome of the patients in Chandigarh.

The relatively poor outcome of patients in the West occurs despite the fact that a high proportion are kept on antipsychotic drugs indefinitely. It has become common practice to give the drugs as a long-acting injection lasting up to four weeks. Even with this regime, which ensures that the patients are receiving the drugs, about 40 per cent experience relapse over a year. For patients on such drugs who are living in a high-EE household, the relapse rate is 50 per cent over nine months. By contrast, the relapse rate for patients on drugs in low-EE homes is 15 per cent over the same period. These findings suggest that the drugs give the patient some protection against relapse, but that the cumulative stress of a high-EE home breaks through this pharmacological shield. If this interpretation is correct, it suggests that further protection might be gained through changing the emotional atmosphere in the home. But there is another way of viewing the link between a high-EE home and relapse of schizophrenia, and this is centred on the patient’s behaviour.

We contrasted the two opposing interpretations by using simple models. Model A depicts a direct impact of the high-EE relative’s behaviour on the patient, leading to relapse. Model B indicated that some aspect of the patient’s behaviour (Factor X) both provokes high-EE attitudes and leads to a relapse. According to model B, therefore, high-EE attitudes are merely a by-product of the patient’s intrinsic susceptibility to schizophrenia.

Of course it is a gross oversimplification to reduce the complexities of human interaction to linear models. A model which depicts the family as a system would be much more appropriate, but there are serious problems in devising experiments to test such models. Consequently my colleagues and I designed a study to determine which of our simple models best described the relationship between relatives’ EE and relapse of schizophrenia. There is one other factor we need to take into account: the amount of social contact between patient and relatives. Patients who are able to keep a social distance from high-EE relatives have a lower relapse rate than those in high social contact. In fact, low contact appears to exert a protective effect of a similar magnitude to that of drugs.

We focused our study on patients in high social contact with high-EE relatives. We knew from previous research that nearly all relapse over nine months if given no medication. It was therefore unethical to keep them off drugs, so all patients entering the study were maintained on antipsychotic drugs. They were then randomly assigned to either a control group where they would receive routine outpatient care, with very little likelihood that the relatives would be offered any help, or a special experimental programme designed to lower relatives’ EE and/or the social contact between patients and relatives. We predicted that if model A is right, the experimental patients would be less likely to suffer relapse than the control patients. But if high-EE attitudes among relatives are a by-product of the patient’s behaviour, lowering the relatives’ expressed emotion would have no impact on the relapse rate, and model B would be a better fit.

In the experimental programme we began by telling relatives about the causes, symptoms, prognosis, and treatment and management of schizophrenia, in two sessions held in their homes. This venue was chosen on the grounds that they were more likely to feel relaxed and receptive to information on their own territory. The education sessions were followed by family sessions held every two weeks in the home and including the patient. These were focused on the everyday problems that arise in such households, such as difficulty in getting the patient out of bed in the morning and neglect of hygiene. We had found that these problems were often the target of relatives’ critical comments. The family was taught how to break each problem down into manageable components; they were then encouraged to come up with a variety of solutions, choose one to attempt, and cooperate in carrying it out. At the end of each session, the family was left with an agreed task; the therapists checked on their progress the next time they visited.

Attention was also paid to the quality of communication between family members. If they talk over each other, ignore one member, or constantly digress from the point, the therapists lay down ground rules for clear communication and tactfully but firmly encourage the family to observe them. The expression of emotion was naturally one of the key areas on which the therapists’ efforts were focused. They intervened to block angry exchanges, attempt to alleviate relatives’ sense of guilt, and generally reframe negative emotions in a positive way.

To reduce social contact, patients who were unemployed were found places in sheltered workshops or day activity centres. If it was a household where parents (rather than a spouse) cared for the patient, we encouraged the parents and the patient to spend their leisure time engaging in different activities. Attempts were made to link patients socially with people of their own age, though this was difficult if the patient was deficient in social skills.

In parallel with the family sessions, we set up a group for relatives, but not patients, to attend. Both high- and low-EE relatives were invited to the group. The aim was that the low-EE relatives would teach their effective styles of coping with everyday problems to the high-EE relatives. Furthermore, it gave the therapists the opportunity to learn about the strategies employed by low-EE relatives. It emerged that these relatives had indeed developed a range of coping strategies which seemed to solve the problems that arose from life with a schizophrenic patient. These were often creative solutions which avoided conflict and the dangers of overinvolvement. The group offered social support to relatives who were often socially isolated; it provided a forum for relatives to express their grief and sense of loss; and the factor of peer pressure did seem to help to induce change in otherwise ‘stuck’ situations.

After nine months in the programme, family members and patients were reassessed for EE, social contact and the course of the schizophrenic illness. We found that the experimental programme reduced EE and/or contact in three-quarters of the families. The relapse rate of 50 per cent in the control families was exactly as predicted from the earlier naturalistic studies. The patients in the experimental programme had a significantly lower rate of relapse (only 8 per cent relapsed), while among experimental families who changed in the desired direction, no patient relapsed. These results strongly favour model A over model B. Furthermore they indicate, rather surprisingly, that social intervention employed in tandem with drug treatment can make a big impact on the short-term outcome of schizophrenia, even though it has a biological basis.

Our own study is not alone in demonstrating an effect of family intervention. Since the late 1970s Michael Goldstein, Ian Falloon, Gerry Hogarty and others have conducted 6- to 12-month studies which obtained comparable results. And Falloon’s findings, for example, showed a relapse rate after two years of 12 per cent for experimental families, against 83 per cent for the control group.

It is clear, then, that there is a remarkably consistent benefit to patients from family interventions both at nine months and two years after discharge from hospital. The build-up of this body of evidence over the past decade holds the promise that a similar approach applied to other stress-related conditions, both psychiatric and nonpsychiatric, could lead to the development of a range of effective psychosocial interventions.

Julian Leff is director of the Medical Research Council’s Social and Community Psychiatry Unit at the Institute of Psychiatry, London.