Abstract
In his impressive paper, Professor McGorry focused on interventions based on current knowledge which were likely to immediately reduce the likelihood of young people developing schizophrenia [1]. He reported the use of sequential screening in identifying the people at risk for developing schizophrenia and, hence, who would be subjects for interventions designed to prevent this development. He pointed out these interventions would be maximally effective for the prevention of schizophrenia if they were directed at those factors associated with a high risk of the development of the illness. At the same time, interventions directed at factors associated with a lower risk, such as dysfunctional family environment, substance abuse, stress and perinatal brain damage, which increase the risk of other psychiatric disorders in addition to schizophrenia, would have broader preventative effects.
While the interventions discussed by Professor McGorry, which have immediate results, clearly should receive the highest priority, it would seem advisable that some attention, at least at a research level, also be given to risk factors that, while being associated with a low risk of developing the illness, are more specific for schizophrenia. It is possible that if the genes associated with specific low-risk factors could be identified and the range of their effects understood, interventions could be designed which would minimise the vulnerability in the subjects carrying the genes so that less specific interventions may not be required.
Study of specific vulnerability to schizophrenia and, in particular, identification of a psychological marker for this vulnerability, has been a major interest in my research. This interest was stimulated by a clinical observation made when treating patients with insulin coma treatment at Royal Park Hospital (Melbourne, Victoria, Australia) in 1953. Over the required day-to-day contact with the patients over several weeks, I noticed that when they reported that they no longer experienced delusions or hallucinations, they continued to show what I at first thought was a disturbance of abstract thinking. My initial conclusion was that they had not completely recovered. Their thinking appeared to be vague and imprecise, lacking the clarity of meaning I then believed should characterise normal thinking. However, when I discussed with the patients' relatives my belief that the patients had not yet recovered as they did not express themselves clearly, the relatives would assure me that the patients had returned to their normal state, and they considered their thinking to be as it had always been. In these discussions, I started to notice the same lack of precision in the thinking of many of the relatives, and concluded they must also be showing this as a symptom of schizophrenia. However, I then noticed the same thinking in some of my friends when they discussed abstract matters, and concluded it must be a different form of abstract thinking that occurred in many mentally well people, but which was associated with a predisposition to schizophrenia. To emphasise its normality and its possible positive quality, I suggested it be termed allusive thinking.
When Emeritus Professor Sid Lovibond, then a psychology student, came to Royal Park to carry out a project, I interested him in my belief and he derived a scoring method from that used by Rapaport [2] to measure looseness of abstract thinking on an Object Sorting Test. Patients whom I classified as showing allusive thinking compared to those not showing it obtained significantly higher scores on the test, which was administered independently by Sid Lovibond [3]. I then used this test to show that parents of patients with schizophrenia obtained significantly higher scores on this test compared to parents of controls [4]. To my knowledge this was the first research evidence of a familially transmitted marker for the development of schizophrenia. In a subsequent study with Margaret Clancy [5] we showed that parents of university students who obtained higher scores on this test, themselves obtained higher scores than did parents of students with lower scores.
From the nature of allusive thinking I believed that, rather than being learned, it was the result of a difference in physiological brain functioning, and then somewhat grandiosely titled the article reporting the finding, ‘The use of an object sorting test in elucidating the hereditary factor in schizophrenia’ [4]. I suggested it was due to a weakness of cortical inhibition so that subjects with allusive thinking would have a less focused attention process. This would result in their associative processes in abstract thinking activating a broader halo of associations than would be the case in subjects without allusive thinking. Hence, they would choose more remote associational path-ways in their thinking. Michael Armstrong carried out a study supporting this broader ‘word halo’ concept of allusive thinking, demonstrating that university students with higher Object Sorting Test scores showed more remote assocations than did those with low scores [6]. Although suggestive, this did not establish that the form of transmission was genetic rather than environmental. Only after working with Stan Catts, Philip Ward and other colleagues was evidence obtained to support this belief, namely the presence of a reduction of the cortical evoked potential P300 in students with higher as compared to those with lower scores on the Object Sorting Test [7]. There is evidence that genetic factors determine P300 amplitude in healthy subjects. Although statistically significant, the relationship between Object Sorting Test scores and P300 was much weaker than the relationship between reduced P300 and schizophrenia itself.
Allusive thinking has not been used as a marker in intervention studies. This could be due to the fact that it does not represent a strong predisposition to schizo-phrenia, which is consistent with the fact that not all the replications supported at a significant level the finding of my original study [4]; that is, that parents of patients with schizophrenia obtained higher scores on the Object Sorting Test than did control parents. Only when Stan Catts and coworkers carried out a meta-analysis of all the studies was the finding convincingly supported [8]. That meta-analysis was necessary to establish the relationship of course means that it is weak and only a minority of people with allusive thinking will develop schizophrenia. This is, of course, not unexpected in view of the finding that a significant percentage of university students and their parents obtain relatively high Object Sorting Test scores. However markers that are associated with a weak risk of developing schizophrenia are, at times, suggested for identification of subjects for interventions. Gosden cited evidence that nearly half of a general population of adolescents had two or more of the prodromal symptoms [9] supplied by the DSM-III-R, one such suggested marker. Another problem in the use of allusive thinking as a possible marker in intervention studies is that the administration of the Object Sorting Test is time-consuming, taking on average about 1 hour.
If the specific vulnerability to schizophrenia is to be investigated, account will need to be taken of the evidence that there are a number of cognitive factors that are distributed dimensionally in the population, with high-scoring patients for these factors demonstrating only a weak predisposition to schizophrenia. Ward et al. investigated word associations, Object Sorting Test scores, psychoticism, physical anhedonia and perceptual aberration in 93 university students to determine whether a general factor of psychological vulnerability could be demonstrated [10]. There were no substantial relationships between these measures of psychosis proneness, indicating that this proneness is multidimensional. Evidence that differences in brain processing are involved in some measures of this proneness was found in an unpublished study by myself and colleagues again investigating university students. Schizotypy was found not only to be independent of allusive thinking but to correlate weakly with an increase in P300, whereas, as stated earlier, allusive thinking correlates weakly with a reduction in P300. This reduction correlates strongly with the presence of schizophrenia. Although psychoticism was found to correlate highly with schizotypy so that both may measure the same entity, the Thought Disorder Index correlated only weakly with both schizotypy and allusive thinking.
Gosden, in criticising interventions designed to prevent the development of schizophrenia, expressed concern that they could threaten human diversity [9]. However, even if genetic markers are identified for schizophrenia only a small percentage of subjects carrying such markers are likely to develop schizophrenia. Attempts to discourage these subjects from having children would clearly be impractical if they were ever considered. A further reason why this would never be attempted is the recognition of the value of aspects of behaviour associated with the presence of such markers. Having recognised the normality of allusive thinking, I noted its link with certain forms of creativity, in particular the ability to see connections between ideas or impressions which to non-allusive thinkers could seem only remotely related or even illogical. Yet these connections proved to have artistic or scientific validity. In an early article to support the argument that the lack of total relevance of the verbal expressions of subjects with allusive thinking was not due to their lack of verbal skills, I quoted a passage from Patrick White's Voss concerning the heroine, Laura [11]: ‘How her defection (from religion) had come about was problematic, unless it was by some obscure action of antennae… Already as a little girl she had been softly sceptical, perhaps out of boredom; she was suffocated by the fuzz of faith. She did believe, however, most palpably, in wood, with the reflections in it… her mind seemed to be complete. There was in consequence no necessity to duplicate her own image, unless in glass, as now, in the blurry mirror of the big darkish room.’
Apart from the idiosyncratic expressions, in particular ‘some obscure action of antennae’ White's ability to use imagery at both an abstract and a concrete level, most obvious in the expression ‘duplicate her own image’ – referring both to her soul and her reflection – is evidence of his masterly use of allusiveness. In science the concepts of space being bent, and of the ability to see round corners when at speeds approaching the velocity of light, again requires an ability to transcend the rational. To obtain empirical support for the link between allusive thinking and creativity, Sharon Rothwell administered the Object Sorting Test to a group of visual artists and a group of administrators [12]. The artists obtained significantly higher scores in the same range as those of patients with schizophrenia.
The aim in identifying markers for the development of schizophrenia is not to attempt to eliminate the presence of people carrying such markers in the population. Rather it is to lead to an understanding of how they predispose to schizophrenia, in order to prevent this development while maintaining their associated positive contribution to human diversity.