J.H.K.C. Psych. (1991) 1, 23-32

REVIEW ARTICLE

pdf Full Paper in PDF

INTRODUCTION

Despite extensive research, the neuropathogeneses of schizophrenia remains elusive. Various theories previously put forward had been discarded, modified, or found to be deficient. For example, the dopamine theory was unable to explain the negative symptoms of schizophrenia, and the virus theory was not substantiated by the finding of any specific virus. For a hypothesis to be plausible, it must consider and account for the major features of schizophrenia. Explaining one aspect of the disease while ignoring the whole is just like 'the blind men touching the elephant'. In particular, it is important to realize the dynamic and interlocking condition of the brain functions. In this article I am going to outline the common clinical and laboratory findings in schizophrenic research, discuss the prevalent theories and attempt to give a hypothesis that can accommodate the findings basing on our current knowledge.

CLINICAL FINDINGS

The main clinical findings are summarized as follows:
(1) Hereditary factors in the transmission of schi- zophrenia. (2) Excess of winter births in schizophrenic (3) Onset in late adolescence or early adulthood.
(4) Relapsing nature of the illneses.
(5) Role of stress in onset and relapse.
(6) Therapeutic efficacy of neuroleptic drugs.
(7) Effect of neuroleptic drugs on the course of illness

a. The natural history of (untreated) schizophrenia is less severe nowadays than it used to be (Hare, 1983). This suggestion has been offered to argue for a viral aetiology for schizophrenia. But an alternative possibility is that modem treatment, especially drug treatment, not only limits the acute psychotic phase of the illness, but also protects to some extent against the otherwise inevitable decline. This protective effect is supported by studies comparing the outcome in patients admitted before and after the introduction of neuroleptic drugs in the 1950's (Pritchard, 1967) and by long term controlled comparisons of neuroleptic and placebo treated patients (Englehardt et al, 1967).

b. Early treatment is able to achieve a higher rate of complete remission and a lower rate of defect

c. Since negative symptoms correlate with ventricular enlargement, it follows that prolonged and repeated episodes of active untreated psychosis can lead to progressive cell loss (Andreasen et al, 1982).

d. With neuroleptic treatment, schizophrenia is still a progressive disorder. Impairment acquired earlier cannot be dissipated by later neuroleptic treatment. The greater the number of admissions, the less the likelihood of a successful therapeutic outcome. The more episodes a patient had, the more insidiously developing and longer lasting these have been, the more likely it is that some residual damage will remain.

(8) Positive symptoms of schizophrenia usually precede negative symptoms (Pfohl & Winokur, 1982).

(9) Only about 35% of the patients remain in the defect state (Gross & Huber, 1986).

(10) Rate of progression of illness

a. In schizophreniform disorder there is remission in between episodes with insignificant or mild residual negative

b. More commonly, positive symptoms progress to negative symptoms and defect state within weeks or months but the progression usually becomes static after a few eposodes. Sometimes positive and negative symptoms exist together.

c. Positive symptoms tend to disappear in chronic cases (although acute on chronic exacerbation do occur). Negative symptoms also lessen in old age.

d. Individual patients could show either deterioration, improvement or, occasionally recovery, many years after

LABORATORY FINDINGS

BRAIN IMAGING STUDIES (CT SCAN AND MRI)

Johnstone et al (1976) were the first to report ventricular dilatation in schizophrenic brains by the use of CT scan. Subsequent studies gave rise to different groups of results which are worth mentioning.

a) A number of studies found no abnormality: e.g. in relatively young schizophrenics or chronic cases having rather benign illness (Benes et al, 1982).

b) Some found abnormality correlating with duration of illness, hospital care or number of episodes (Kemali et al, 1985). The correlational studies do provide some support for the idea that ventricular enlargement progresses during the course of the manifest illness. Some patients who had been ill for many years, so that all progression that might occur, would already have occurred. That might be the reason why Nasrallah et al (1986) had failed to find progressive enlargement over a 3 year period.

c) Weinberger et al (1981) claimed that there is a tendency for ventricular size in normal siblings to show correlaHowever, in sibships where one member is schizophrenic, the affected member always had larger ventricles. Reveley et al (1982) compared the ventricular size in twins discordant for schizophrenia and found that only the affected twin had an increased VBR (ventricular brain ratio). Therefore ventricular size is partly under genetic control, but also that enlargement is related to the manifestation of schizophrenia.

d) Monozygotic twins who were discordant for schizophrena were studied by Most affected twins had smaller hippocampus, larger lateral ventricle and larger 3rd ventricle. No difference were found in MZ twin pairs without schizophrenia (Suddath et al, 1990). The fact that an affected MZ can be differentiated from an unaffected twin on the basis of structural difference in the CNS probably means that the cause of the underlying neuropathological process is at least partly not genetic.

e) Ventricular enlargement can result from such early trauma in normal subjects, without relation to schizophreHowever, those scanning abnormalities which are most characteristic of schizophrenia (i.e. loss of tissue in the temporal lobes) seem not to be related to such early insults. When ventricular enlargement is found in schizophrenia in association with early head rrauma or birth complications, it probably results from associations that apply as much to normal subjects as to schizophrenic patients. Reveley et al (1984) found amongst MZ twins without schizophrenia, ventricular enlargement was greater in those with evidence of birth complications. But in MZ twin with schizophrenia ventricles were smallr in those with birth complications than those without. Thus enlarged ventricles can occur as a result of birth trauma, without leading to schizophrenia and in schizophrenia it could be acquired for reasons other than birth trauma. Since twins have a much higher incidence of obstetric complications it is possible that the original finding in twin reflects an increase in ventricular size resulting from a complicated birth such as is found in normal subjects irrespective of their development of schizophrenia. There are other ways in which ventricular enlargement can be acquired in schizophrenia.

PET STUDIES

Farkas et al (1984) found relative hypofrontality in chronic schizophrenia. Other studies (Gur et al, 1987), using similar PET procedures, have failed to find hypofrontality in schizophrenia. Some researchers (Szechtman et al, 1988) even found hyperfrontality in acute cases. These findings served to confirm earlier studies of regional cerebral blood flow and cerebral metabolic rate in acute and chronic schizophrenia. Generally speaking, in acute schizophrenia with productive (positive) symptoms, the postcentral region and possibly the frontal cortex are overactive while in chronic schizophrenia with non-productive (negative)symptoms, the frontal region is underactive.

NEUROANATOMICAL AND NEUROPATHOLOGICAL STUDIES

a) A significant number of schizophrenic brains as compared with controls had some degree of localized cerebral pathology (Bruton et al, 1990).

b) Cytoarchitectural changes.

Benes & Bird (1987) found unusual cellular organization in the superficial laminae of the cingulate cortex in the brains from elderly schizophrenic brains. There were significantly more vertical axonal connections in laminae II and III of the cingulate cortex and these excess connections were tentatively identified as connections from other areas of association cortex. Altered orientation of the hippocampal pyramidal cells has also been reported (Kovelman & Scheibe! 1984). These changes were considered suggestive of a defective pattern of neuronal migration during gestation. However it is known that in normal brain activity, dendrites are in a state of continual change, some degenerating, some continue to grow, even in advanced age. An alternative interpretation would be that excitotoxic influences in the adult could lead to degeneration of some dendrites which then stimulates the affected neurones to grow more dendrites, which lack their normal orientation (Miller, 1989).

c) Medial temporal lobe abnormality (Roberts, 1990)

Medial temporal lobe structures are believed to have a crucial role in the integration and processing of the input from association cortex. Direct connections link the medial temporal cortex to the frontal cortex, striatum and amygdala. Temporal cortex, hippocampus and amygdala are thought to be central in the integration of emotional responses and intellect. It may be that frontal dysfunction results, at least in part, from structural abnormalities in regions such as the temporal lobe.

d) Left sided

Several observations indicate laterality in motor functioning of schizophrenic patients. Early et al (1987) reported higher relative blood flow in the left globus pallidus of neuroleptic naive patients. Laterality in the distribution of 02 dopamine receptor densities in the postmortem brain (Reynolds et al, 1987) and living putamen (Pakkenberg, 1987) have been reported.

c) Absence of gliosis (Roberts et al, 1987).

In recent years the question of cerebral gliosis has figured prominently in many discussions of schizophrenia. Stevens (1982) found 75% of schizophrenic brains showed an increased fibrous gliosis affecting principally the periventricular structures. However later studies generally found absence of gliosis after excluding cases which showed histological evidence of Alzheimer-type change, cerebra-vascular disease and any from of focal pathology (Bruton et al, 1990).

The presence or absence of reactions in the glial cells is held to be an abnormal criterion with which to ascertain whether the loss of neurones occurring in the brains of some schizophrenic patients is a result of degeneration or of an earlier abnormality. The rational is that degeneration of neurones in the adult is said to be invariably associated with a reaction of glial cells. On the other hand, loss of neurones occurring in immature neuronal system, whether as part of normal development, or as a result of early injury to the nervous tissue, is not accompanied by such a reaction.

The majority view of the recent studies is that there is no glial cell reaction in the brains of schizophrenic patients; at least, not one that endure the many years until the brains become available for post-mortem studies. However, the extent and durability of glial cell reaction appears to vary considerably according to type and severity of damage. It is plausible to think that very gradual loss of neurones in the adult could occur without leaving a permanent trace of a glial cell reaction. The association between a glial cell proliferation amd neuronal damage may not be inevitable. It can be said that neuronal loss without glial cell abnormalities does not necessarily mean that the neuronal loss is of early origin. it is likely that the mechanism of cell loss in schizophrenia is a less severe pathology than in some other neuordegenerative disorders (e.g. Huntington).

BIOCHEMICAL STUDIES

Besides an upregulation of D2 receptors and increase in dopamine turnover, other neurotransmitter systems are found to be altered. Of particular interest is the study on glutamate receptors. Kim et al (1980) found low glutamate level in the CSF of schizophrenic patients. Toru et al (1988) reported increased frbntal cortical 3H-kainate (a glutamate receptor ligand) bindings, a finding shared by Deakin et al (1989). There is a negative correlation between 3H-kainate binding in the medial frontal cortex and glutamate content in various brain areas. Deakin suggested there is excessive glutamatergic innervation of orbital frontal cortex in schizophrenia.

CURRENT THEORIES

ABERRATION OF EARLY DEVELOPMENT

Current thinking seems to favour aberration of early development to be the underlying cause of schizophrenia. These workers first try to give reasons that oppose the idea that schizophrenia may be due to neurodegeneration, an idea that was revived since the discovery of dilated ventricles in schizophrenia. They argue that:

a) Most studies failed to show the correlation between ventricular size and duration of illness (Williams et al, 1985). But there are studies that had shown such a correlation (Kemali et al, 1985; Nasrallah et al, 1986).

b) Several studies found increased ventricular size predate the onset of psychotic symptoms (Vita et al, 1988), but the number is comparatively small to be of significant value

c) Increase ventricular size remained static during the course of the illness (Vita et al, 1988).

d) Absence of gliosis in those with reported decreased tissue volume and abnormal cytoarchitecture. See discussion above.

e) Weinberger (1987) pointed out that, if schizophrenia is analogous to a metabolic or a neurodegenerative disorder, then the pathology should parallel the disease process and become more extensive over time as the process conTo answer this consideration should be given to: (i) Excitotoxic effect depends on the excitation, which becomes quiescent as the stress is over; (ii) clinical experience which shows that reparative mechanisms are going on trying to repair the deficit if it is not too extensive. Hence we find schizophreniform disorder remits without deficit in terms of months. Long term follow up of over tens of years show ameliation of symptoms, both positive and negative, when the age advances.

WEINBERGER'S NEURODEVELOPMENTAL HYPOTHESIS

Weinberger (1987) has recently put forward a hypothesis of schizophrenia (fig. 1). The main argument is that in schizophrenia a fixed lesion of the brain exists early in life, which expresses itself as symptoms of acute psychosis in late adolescence or early adulthood or chronic schizophrenia much later on. The tendency for schizophrenia to start at early adulthood may reflect a combination of an innate overconnectivity and some chemical state (perhaps hormonal) which mature at this stage of life.

It is found that neurological disorders such as tumours and trauma are associated with schizophreniform symptomatology (positive) more frequently when they involve the limbic system and diencephalon. Negative symptoms (flat affect, social withdrawal, lack of initiative) resemble symptoms seen in patients with frontal lobe lesions, especially the dorsolateral prefrontal cortex. Weinberger argues that the weight of evidence suggests that the lesion in schizophrenia renders mesocortical dopamine function underactive. In the rat, it is shown that after selectively destroying DA afferents within the prefrontal cortex, chronic subcortical dopamine hyperactivity develops. The prefrontal cortex projects to DA cell bodies in the midbrain and to mesocortical terminal fields in the amygdala, nucleus accumbens, hypothalamus and hippocampus. It is known that dopamine metabolism in the brain increases when an animal is under physical stress such as pain or shock. While all dopamine systems respond to visceral stress, the prefrontal afferentation system is uniquely sensitive to experiential stress. Stress (where judgement and abstract concepts must be called into action) places a heavy demand on dorsolateral prefrontal function. If a lesion that disrupts the dopaminergic afferents to the prefrontal cortex, the individual would be incapable of making appropriate cognitive responses to stressful situations. Early adulthood is the time of maximum dopaminergic activity in the brain. Faced with stress that demand maximum prefrontal cognitive function and a lesion impairing one of the mechanisms that augment this function, the individual cannot make the physiological and cognitive adaptations and manifests inappropriate behaviour, confused thinking, delusions, social withdrawal. And when faced with mesolimbic dopamine activity that is peaking for normal developmental reasons, the individual cannot control this activity and become agitated, fearful and hallucinative. The lesion may be caused, according to Weinberger, by hereditary encephalopathy, infection, perinatal trauma, early exposure to toxins and so on.

Weinberger suggests that the 'hypofrontality' is present early in life, but not expressed until early adulthood. He points out that the dorsolateral prefrontal cortex (DLPFC) is the last brain area to begin myelination. Because it reaches maturity relatively late brain in life, a lesion of the DLPFC would have little impact on behaviour during childhood. That is how a fixed congenital lesion could remain relatively unapparent until early adulthood. However he fails to consider that this is more a characteristic of chronic than acute schizophrenia and 'hyperfrontality' and excessive neuronal activity are occurring in the early stage of schizophrenia. Weinberger thinks frontal lobe lesion is primary while temporal and limbic lesion are secondary. However negative symptoms, hypofrontality and defect state only occur in about 35% of the schizophrenic patients. Other common features of schizophrenia that a nurodevelopmental hypothesis cannot explain are the relapsing nature of the illness, positive preceding negative symptoms, the progressive deterioration with relapses, hyperfrontality in acute stage and hypofrontality in chronic stage., etc.

GENETIC CAUSE OF BRAIN ANOMALY

Roberts (1991) suggests that schizophrenia is a neurodevelopmental anomaly of the temporal lobe, probably genetic in most cases and affects brain development during the third trimester. He points out that birth difficulties can give rise to intraventricular haemorrhages and infarcts, which lead to enlarged ventricles and small brains. Yet such pathologies are accompanied by gliosis, yst formation and leukomalacia. Since the volume of the white matter is not reduced in schizophrenic brain, and gliosis and inflammatory changes are absent, there is insufficient ground to support it. There is little evidence for traumatic injury either. The evidence indicates that the cause of proposed developmental anomaly responsible for structural changes in the brains is probably genetic in origin in most cases. Environmental factors account for a minority (5- 10%) of cases. The asymmetrical pattern of development in the normal brain can explain why such a defect affecting both hemispheres simultaneously will disproportionately affect the left dominant hemisphere.

Again this argument cannot explain the findings of relapsing nature of the illness, positive preceding negative symptoms, protective effect of neuroleptics, and hyperfrontality in acute schizophrenia.

DISCUSSION

HOMOGENEOUS OR HETEROGENEOUS?

Some workers regard schizophrenia as a homogeneous condition while others treat it as heterogeneous. The present article tends to favour the former. We do see schizopreniform disorders that remit rapidly without residual defect and schizophrenia that deteriorate gradually. During relapses, however, the clinical presentations do not differ from each other and the positive symptoms respond to neuroleptic treatment in both conditions. We have to agree that they share similar abnormalities during these states, i.e. a hyperactivity of the dopamine system. However, relation between positive symptoms and brain structural changes cannot be established. Changes in brain structure have been shown in patients with or without positive symptoms, negative symptoms and varying degrees of assumed genetic loading. Thus there is little convincing evidence for a rational subgrouping of schizophrenia on the basis of the presence or absence of structural change or other etiological factors. It seems prudent to say that all schizophrenics have a degree of brain abnormality and that schizophrenia is best viewed as a homogeneous, but graded condition. Why some cases progress to the chronic state while others remit without residual defect is the point to be considered.

FRONTAL OR TEMPORAL LOBE, CORTICAL OR SUBCORTICAL LESIONS?

Generally speaking, in acute schizophrenia the abnormality lies in the temporal-limbic region whereas in chronic schizophrenia the abnormality lies in the frontal lobe Hyperactivity of the dopamine system is found in acute schizophrenia and neuroanatomical lesions are found in chronic schizophrenia. Yet two conditions are but different stages of the same disease. Can we give a unified theory to account for the various findings?

Positive symptoms are produced as a reaction to physical or psychosocial stress. In fact symptomatic psychoses occurring in cases of infection, toxic confusional state etc. are schizophrenia-like and amenable to treatment by dopamine antagonists. Other symptoms such as the thought disorders and speech abnormalities are most likely cortical dysfunctions. Psychosocial stress involves the cognitive process which is again a higher cortical function. Together with hallucinations and agitation, supposedly temporolimbic dysfunctions, acute schizophrenia represent an involvement of both cortical and subcortical structures. On the other hand, negative symptoms represent a more fundamental defect. It can either be the underlying cause of the positive symptoms or the result of the overactive neuronal discharges.

GENETIC, DEVELOPMENTAL OR ACQUIRED?

If the anomaly is solely genetic or developmental in origin, one would expect a constant or deteriorating clinical picture once the illness sets in. Yet the usual schizophrenic process is an initial acute phase (positive symptoms) followed by the development of chronic state (negative symptoms). The chronic symptoms become static after some time without fruther deterioration and there may even be improvements. The disease process is punctuated by episodes of positive psychotic symptoms when the patient faces psychosocial stresses beyond his capacity to cope. The course of the illness has been altered by neuroleptic treatment and is much milder nowadays than 50 years ago. These are some of the evidences that can be put forward to oppose an entirely genetic and developmental origin of the disease. It would be reasonable to assume that the genetic and developmental effect on the patient only work by making the patient more vulnerable to stress instead of directly producing ventricular dilatation and cytoarchitectural changes. There is as yet no direct evidence that ventricular dilatation or cytoarchitectural changes precede the onset of schizophrenia in the same way as schizoid personality predisposes schizophrenia.

Loss of neurones is conjectured by many researchers to be of early developmental origin. Critical scrutiny of the evidence shows that the arguments supporting this conjecture are flawed. Absence of glial veil reactions is not a sure way of excluding degenerative loss of neurones in the adult if that loss is slow and mild. The workers who gave evidence for structural changes early in life probably had considered those cases that become defective later on, and these constitute but 35-50% of the cases. It is more likely that the structural changes, except for a small number of cases, occur after the onset of illness.

THE HYPOTHESIS

The aim is to propose a hypothesis that can explain most facets of schizophrenia. An attempt is made to look at the schizophrenic priocess in the following ways: [fig. 2]

Contributory Factors

a) The hereditary factor in the pathogenesis of schizophrenia is indisputable. It may act in various ways, for example, on the predisposition to an excessive neurotransmitter (dopamine and glutamate) activity, calcium influx and excitotoxic effect in reaction to stress; or on an intense reaction to hormonal influence on the nerve connection.

b) Occurring in the 2nd or 3rd trimester, there may be an innate misconnection of neurons (in the medial temporal lobe and hippocampus). When demand is put on them in early adulthood, coupled with hormonal influence, there may be an aberrant reaction in the affected brain areas (basal ganglia and medial temporal region). Brain injury and viral infection account only for a small percentage of the

c) The early adult onset of schizophrenia suggests that hormonal effects are present in the pathogenesis of the illIn fact steroid hormones may have wider effects at various stages of development. The evidence for hormonal influence on brain structure can be seen in the following studies.

(i). It is reported that perinatal steroid hormone secretion can influence the developing central nervous system by altering the pattern of nerve connection (DeVoogd, 1987), which has influence on the future make up of the animal brain. The central nervous system remains plastic in adult animals, capable of substantial structural reorganisation in response to altered levels of circulating steroids and seasonal changes, There are thus two actions of steroids on the dendritic structure; one is lasting and occurs in a restricted time window (critical period) early in development; the other is rapid and reversible and may occur throughout the adult life according to seasonal and hormonal changes. Hence the excessive winter birth of schizophrenics, besides a possible increased susceptibility to virus infections, could be the result of changes in steroid hormones with season and day length which in turn affect nerve connections.

(ii). In rats, an environmental manipulation occurring early in life resulted in changes in the adrenocortical axis that persisted throughout the entire life of the animal. Rats handled during infancy had a permanent increase in concentrations of receptors for glucocorticoids in the hippocampus. Rats that were not handled secreted more glucocorticoids in response to stress than did the handled ones. Consequently, increased exposure to adrenal glucocorticoids can accelerate hippocampal neuronal loss in aging. (Meanly et al, 1988).

These studies show that the perinatal environment and manipulations early in life and in adulthood can modify the reaction of the brain to hormonal influences. But then how neurotransmitters (e.g. glutamate and dopamine) and hormonal effects act together to exert influence on the brain structure and function is an area that need to be further explored.

d) It is well known that the onset and relapse of schizophrenia can be precipitated by psychosocial As suggested by Weinberger (1987), "stress" calls into action judgement and abstract concepts and places a heavy demand on prefrontal function, which is mediated by neurotransmitters.

The principal neurotransmitters of the cortex are the excitatory amino acids, primarily L-glutamate and L-aspartate. In fact they are the most abundantly found neurotransmitters in the mammalian brain (Cotman & Iversen, 1987). The excitatory amino acid pathways have been delineated and include the cortico-cortical, cortico-fugal and sensory systems (Cotman et al, 1987). Messages from the cerebral cortex to the subcortical structures are mediated by glutamate or a closely related compound. There are generally speaking two types of receptors, the NMDA and the nonNMDA (kainate and quisqualate) receptors. The NMDA receptors are the most widely distributed type and have a wide range of properties. [fig. 3] (Cotman et al, 1987)

When stimulated by the excitatory amino acids, the nonNMDA receptors generate fast depolarising responses until a certain level of membrane potential is reached, and then the NMDA receptors are activated. The NMDA induced current is longer in duration and Ca + + ions enter along with Na + + through the ion channels. In the event of prolonged and excessive stimulation, excitotoxicity comes into play when there is too much glutamate and too much calcium. It results in nerve cell damage and death. When the neurons are briefly exposed to glutamate, the cells swell transiently and then go on to recover, but when influx of calcium also occurs, the cells slowly degenerate (Rothman

& Olney, 1987). Available evidence suggests that there is in fact a continuum of action whereby a given excitatory neurotransmitter can at low levels stimulate sprouting, at only slightly higher levels halt neurite outgrowth, and at still higher concentrations, cause neuronal death. The neurotransmitter exert these effects largely by modulating the levels of specific second messengers such as intracellualr free calcium (Lipton & Kater, 1989).

NEUROPATHOLOGICAL CHANGES

a) It is proposed that during acute psychotic stage, excessive neuronal activity occurring in association with an unchecked psychotic state may lead to cell loss as a result of the excitotoxic effects of natural excitatory transmitters. Loss of brain substance occurs as schizophrenic illness progresses. The unusual organisation of neurones and their processes suggest an overconnection in certain pathways in the hemisphere. It is not yet known to what extent these abnormalities are of developmental or of degenerative origin, but it is probable that both processes are involved. Unusual overconnection (or misconnection) of early origin may be an initiator of the early stages of schizophrenia, because it could lead to excess signal amplification in certain parts of the brain.

It is known that brain damage can be followed by extensive "rewiring" in the form of axonal sprouting, axonal rerouting and modified axonal weeding. Focal misconnections between specific neural systems could result in a variety of bizarre symptoms (psychotic phenomenon) and personality traits, whereas diffuse misconnections may be more likely to produce cognitive and behavioural impairment due to poor signal: noise ratio. If focal and diffuse misconnections occur together, the result may be a mixture of bizarre symptoms and cognitive impairments. Schizophrenia may be a case in point. Perhaps a genetic tendency to misconnections accounts for the high rate of attentional problems in the children of schizophrenia (Goodman, 1989).

b) Dopamine possibly enhances the effectiveness of other glutamatergic pathways, and they together intensify the excitotoxic effect (Globus et al, 1987). Other factors, such as hormonal influences, may also be

As noted from above, the corticofugal pathways run from the cortex to the basal ganglia (caudate/putamen and nucleus accumbens), the hippocampus and the dorsal lateral septum, the parts of the brain where highest level of NMDA receptors are found. The striatal NMDA receptors are located on the heads of the dendritic spines of the striatonigral projection neurons while dopamine receptors are found on the neck of the dendritic spines of the same neurons (Smith & Bolam, 1990). Synaptic transmission in these pathways is blocked by glutamate antagonists.

It is conceivable that during severe psychosocial stress the excitatory amino acids are overstimulated, which leads to an increase of glutamate transmission and calcium influx. This results in excitotoxic effect on the striatal neurons where NMDA receptors are located. This in turn activates the dopaminergic pathways which give off additional nerve terminals and synaptic connections. The over-reaction and subsequent neuronal degeneration produce the schizophreniform reaction and progressive schizophrenic process as mentioned earlier, depending on the severity of the damage done and the vulnerability of the dopamine system. In support of this some recent laboratory studies reported an increase in presynaptic glutamate uptake sites in schizophrenic temporal lobe and orbital cortex (Deakin et al, 1989), and increased specific binding if 3H-kainic acid in the prefrontal cortex in schizophrenia (Toru et al, 1988).

c) The premorbid personality and other subtle changes present before the onset of illness are considered results of hereditary effect and neurodevelopmental changes, where hormonal influences and EM activity may also have some parts to

d) Another point to note is the lateralisation effect in schhizophrenic brain. Cerebral lateralisation studies have established the asymmetric distribution of some neurotransmitters in the human brain. Dopaminergic systems are left hemisphere biased (Glick et al, 1982). It can be deduced that lesions of the dopaminergic systems are reflected as left hemisphere dysfunction and such are reported in the CT scan, regional blood flow and PET studies, as well as a specific increase if dopamine in the left hemisphere. It is interesting to note that glutamate transmission is symmetrically distributed in the

CONCLUSION

The present article tries to show that although the neurodevelopmental hypothesis on the pathogenesis of schizophrenia is in current favour, it is not without its shortcomings, By looking at the various contributory factors and the course of the illness as a whole, a hypothesis basing on the interaction between the glutamate and dopamine systems is advanced. It does not reject the neurodevelopmental influence, but serves to give a more complete view of the dynamic process. The psychosocial factor in the initiation and recurrence of the clinical course is emphasized, and is proposed to involve the excitatory amino acids, although it has not been experimentally substantiated. Some other postulations are derived from the studies of other leading workers in the field. They are incorporated into the present article to form a unified view of the schizophrenic process. The hypothesis also has implication in the therapeutic aspect of the disease. Control of the glutamatergic excitotoxic effect and calcium influx may be promising areas in future research. Preliminary trials in this direction have been done in Alzheimer's disease but not in schizophrenia. Further development is awaited.

REFERENCES

Andreasen, N.C., Olsen, S., Dement, J.W., and Smith, M.R. (1982) Ventricular enlargement in schizophrenia, relationship to positive and negative symptoms. Am. J Psychiatry 139, 297-302.

Benes, F.M., Sunderland, P., Jones, B.D., LeMay, M., Cohen,B.M. and Lipinski, J.F. (1982) Normal ventricles in young schizophrenics. Brit. J. Psychiatry 141, 90-93.

Benes, F.M. amd Bird, E.D. (1987) An analysis of the arrangement of neurons in the cingulate cortex of schizophrenic patients. Arch. Gen. Psychiatry 44, 608-616.

Bruton, C.J., Crow, T.J., Frith, C.D., Johnstone, E.C., Owens,

D.G.C. and Roberts, G.W. (1990) Schizophrenia and the brain: a prospective clinico-neuropathological study. Psychological Medicine 20, 285-304.

Cotman, C.W. and Iversen, L.L. (1987) Excitatory amino acids in the brain-focus on NMDA receptors. TINS 10, 263-265.

Cotman, C.W. Monaghan, D.T., Ottersen, O.P. and StormMathisen, K. (1987) Anatomical organisation of excitatory amino acids receptors and their pathways. TINS 10, 273-280.

Deakin, J.F.W., Slater, P., Simpson, M.D.C., et al (1989) Frontal cortical and left temporal glutamate dysfunction in schizophrenia. J. Neurochemistry 52, 1781-1786.

DeVoogd, T.J. (1987) Androgens can affect the morphology of mammalian CNS neurons in adulthood. TINS 10, 341-342.

Early, T.S., Reiman, E.M., Raiche, M.E. and Spitznagel, E.L. (1987) Left globus pallidus abnormality in never medicated patients with schizophrenia. Proc. Natl. Acad. Sci. USA. 84, 561- 563.

Englehardt, D.M., Rosen, B., Freedman, N. and Margolis, R. (1967) Phenothiazines in prevention of psychiatric hospitalisation-a reevaluation. Arch. Gen. Psychiatry 16, 98-101.

Farde, L., Wiesel, F.A., Stone-Elander, S., Halldin, C. et al (1990) D2 Dopamine receptors in neuroleptic naive schizophrenic patients. Arch. Gen, Psychiatry 47, 213-219.

Farkas, T., Wolf, A.P., Jaeger, J., Brodie, J.D., Christman, D.R. and Fowler, J.S. (1984) Regional brain glucose metabolism in chronic schizophrenia. Arch. Gen. Psychiatry 41, 293-300.

Glick, 6.D., Ross, A.R. and Hough, L.B. (1982) Lateral asymmetry of neurotransmitters in human brain. Brain Research 234, 53- 63.

Globus, M.Y.T., Ginsberg, M.D., Harik, S.I., Busto. R., and Dietrich, W.D. (1987) Role of dopamine in ischemic striatal injury: metabolic evidence. Neurology 37, 1712-1719.

Goodman, R. (1989) Neuronal misconnections and psychiatric disorder. Is there a link? Brit. J. Psychiatry 154, 292-299.

Gross, G. and Huber, G, (1986) Classification and prognosis of schizophrenic disorders in light of the Bonn follow up studies. Psychopathology 19, 50-59.

Gur, R.E., Resnick, S.M., Alavi, A., et al (1987) Regional brain function in schizophrenia. 1. A PET study. Arch. Gen. Psychiatry 44, 119-125.

Hare, E. (1983) Age incidence and schizophrenia: Part II. Beyond age incidence. Schizophr. Bull. 15, 75-80.

Johnstone, E.C., Crow, T.J., Frith, C.D., Husband, J. and Kreel,

• (1976) Cerebral ventricular size and cognitive impairment in chronic schizophrenia. Lancet 11, 924-926. Kemali, D., Maj. M., Galderisi, S., Ariano, M.G., Cesarelli, M., Milici, N., Salvati, A., Valente, A. and Volpe, M. (1985) Clinical and neuropsychological correlates of cerebral ventricular enlargement in schizophrenia. J. Psychiat. Res. 19, 587-596.

Kim. J.S., Komhuber, H.H., Schmid-Burgk, W., and Holzmuller,

• (1980) Low cerebrospinal fluid glutamate in schizophrenic patients and a new hypothisis on schizophrenia. Neurosci. Lett. 20, 379-382.

Kovelman, J.A. and Scheibe!, A.B. (1984) A neurohistological correlate of schizophrenia. Biol. Psychiatry 19, 1601-1621.

Lipton, S.A. and Kater, S.S. (1989) Neurotransmitter regulation of neuronal outgrowth, plasticity and survival. TINS 12, 265-270. Meanly, M.J., Aitken, D.H., van Berke!, C., Bhatnagar, S., and

Sapolsky, R.M. (1988) Effect of neonatal handling on agerelated impairments associated with the hippocampus. Science 239, 766-768.

Miller, R. (1989) Schizophrenia as a progressive disorder: Relation to EEG, CT, Neuropathological and other evidence. Progress in Neurobiology 33, 17-44.

Nasrallah, H.A. Olson, S.C., McCalley-Whitters, M., Chapman, S. and Jacoby, C.G. (1986) Cerebral ventricular enlargement in schizophrenia: A preliminary follow-up study. Arch. Gen. Psychiatry 43, 157-159.

Pakkenberg, B. (1987) Post-mortem study of chronic schizophrenic brains. Brit. J. Psychiatry 151, 744-752.

Pfohl, B. and Winokur, G. (1982)

The evolution of symptoms in institutionalised Hebephrenic/Catatonic schizophrenics. Brit. J. Psychiatry 141, 567-572.

Pritchard, M. (1967) Prognosis of schizophrenia before and after pharmacotherapy. 1. Short term outcome. 2. Three years follow up. Brit. J. Psychiatry 113, 1345-1359.

Reveley, A.M., Reveley, M.A., Clifford, C.A., and Murray, R.M. (1982) Cerebral ventricular size in twins discordant for schizophrenia. Lancet 1, 540-542.

Reveley, A.M., Reveley, M.A., and Murray, R.M. (1984) Cerebral ventricular enlargement in non-genetic schizophrenia: A controlled twin study. Brit. J. Psychiatry 144, 89-93.

Reynolds, G.P. Czudek, C., Bzowej. N and Seeman, P. (1987) Dopamine receptor asymmetry in schizophrenia. Lancet 1, 979.

Roberts, G.W., Colter, N., Lofthouse, R., Johnston, E.C. and Crow, T.J. (1987) Is there gliosis in schizophrenia? Investigation of the temporal lobe. Biol. Psychiatry 22, 1459-1468.

Roberts, G.W. (1990) Schizophrenia: the cellular biology of a functional psychosis. TiNS 13, 207-211.

Roberts, G.W. (1991) Schizophrenia: a neuropathological perspective. Brit. J. Psychiatry 158, 8-17.

Rothman, S.M. and Olney, J.W. (1987) Excitatotoxicity and the NMDA receptors. TINS 10, 299-302.

Smith, A.O., Bolam, J.P. (1990) The neural network of the vasal ganglia as revealed by the study of synaptic connections of identified neurones, TINS 13, 259-265.

Stevens J.R. (1982) Neuropathology of schizophrenia. Arch. Gen. Psychiatry 39, 1131-1139.

Suddath, R.L., Christison, G.W., Torrey, F., Casanova, M.F. and Weinberger, D.R. (1990) Anatomical abnormalities in the brains of MZ twins discordant for schizophrenia. New. Engl. J. Medicine 322, 12: 789-794.

Szechtman, H., Nahmias, C., Garnett, S. et al, (1988) Effect of neuroleptics on altered cerebral glucose metabolism in schizophrenia. Arch. Gen. Psychiatry 45, 523-532.

Toru, M., Watanabe, S., Shibuya, H. et al, (1988) Neurotransmitters, receptors and neuropeptides in post-mortem brains of chronic schizophrenic patients. Acta. Psychiat. Scand. 78, 121-137.

Vita, A., Sacchetti, E., Valvassori, G., Cazzullo, C.L. (1988) Brain morphology in schizophrenia, a 2-5 year CT scan follow up study. Acta. Psychiatr. Scand 78, 618-621.

Waddington, J.L. (1990) Sight and insight: regional cerebral metabolic activity in schizophrenia visualised by PET, and competing neurodevelopmental perspectives. Brit. J. Psychiatry 156, 615- 619.

Weinberger, D.R., Delisi, LE., Neophytides, N. and Wyatt, R.J. (1981) Familial aspects of CT scan abnormalities in chronic schizophrenic patients. Psychiat. Research 4, 65-71.

Weinberger, D.R. (1987) Implications of normal brain development for the pathogenesis of schizophrenia. Arch. Gen. Psychiartry 44, 660-669.

Williams, A.O., Reveley, M.A., Kolalowska, T, Ardem, M., and Madelbrote, B.M. (1985) Schizophrenia with good and poor outcome. II. Cerebral ventricular size and its clinical significance. Brit. J. Psychiatry 146, 239-246.

Wong, D.F., Wagner, H.N. Jr., Tune, LE., Dannals, R.F. et al (1986) Positron emission tomography reveals elevated D2 receptors in drug-naive schizophrenics. Science 234, 1558-1563.

C.W. Lo MBBS, MRCPsych. Consultant, Kowloon Hospital Psychiatric Unit, 147A, Argyle Street, Kowloon, Hong Kong.