Neuroanatomy of comorbid schizophrenia and learning disability: a controlled study

Summary

Background

Reasons for the higher frequency of schizophrenia in learning-disabled populations are uncertain. We investigated the neuroanatomical basis for this phenomenon by structural magnetic resonance imaging (MRI) in patients with learning disability and schizophrenia, learning-disabled patients, and patients with schizophrenia.

Methods

Age-matched and sex-matched patients with learning disability (20 cases), schizophrenia (25), and both disorders (23) underwent MRI scans of the brain. Whole brain areas and specific regions of interest were examined. 29 normal controls were also scanned.

Findings

The scans of the group with both disorders were closely similar to those of the schizophrenic group, in terms of both general structures and the structure of the amygdala-hippocampus. However, the amygdala-hippocampus was significantly smaller on both sides than that of normal controls (left 4.1 vs 4.5 cm³, p=0·011; right 4.2 vs 4.99 cm³, p<0·0001). The brains of learning-disabled patients were generally smaller than those of the other three groups, but the amygdalohippocampal complexes were larger.

Interpretation

In terms of brain structure, patients with comorbid learning disability and schizophrenia resemble patients with schizophrenia and not those with learning disability. We suggest that the higher frequency of schizophrenia in learning-disabled patients is due to a greater tendency of schizophrenic patients to develop cognitive deficits, and that within the learning-disabled population there may be individuals whose deficits result from undiagnosed schizophrenia.

Introduction

Epidemiological studies consistently show that people with mild learning disability have a higher frequency of mental disorder than the general population.¹ The prevalence of mild learning disability in northern Europe is about 0·5%.² The point prevalence of schizophrenia in these individuals is about 3%³–ie, three times that in the general population. The reasons for this higher frequency are uncertain, but possibilities include that the presence of schizophrenia (whether diagnosed or destined to develop) increases the likelihood of an individual being diagnosed as having learning disability, and that the presence of learning disability increases the likelihood of an individual being diagnosed as having schizophrenia. The former possibility is consistent with the view that cognitive dysfunction, and sometimes motor and social impairment, in childhood may be the initial features of severe schizophrenia.⁴ The latter possibility implies that the cognitive deficits of some individuals with learning disability convey an increased susceptibility to develop the impairments associated with schizophrenia.⁵ By way of analogy, people with deafness are more susceptible to the development of paranoid psychoses.⁶ and the suggested mechanism behind this association is the overload on comprehension imparted by partly understood stimuli.

Knowledge of the biological basis of schizophrenia remains limited, although the view that there are underlying structural brain abnormalities in a large proportion of cases of schizophrenia⁷ is widely accepted. Evidence of structural brain abnormalities provided by imaging⁸ and post-mortem studies9, 10 has been assessed alongside data on perinatal events,¹¹ minor physical anomalies,¹² and childhood development in individuals destined to develop schizophrenia in adult life.¹³ These elements have been drawn together to form the basis of the neurodevelopmental hypothesis of schizophrenia.7, 14 If the view that comorbid learning disability and schizophrenia represents schizophrenia in a very severe form is correct, individuals with this disorder would be expected to have particularly profound neurodevelopmental deficits and to show an exaggeration of the findings in individuals with schizophrenia alone.

We investigated this issue in terms of structural imaging, by examining matched groups of patients with comorbid learning disability and schizophrenia, schizophrenia alone, and learning disability alone.⁵ Initial reports of structural imaging in patients with schizophrenia showed enlarged ventricles and reduced brain substance, and these findings have been widely replicated.15, 16 Such general findings are, of course, not diagnosis-specific,¹⁷ but several structures, particularly the amygdala and hippocampus, are affected to a greater extent than would be expected from the overall reductions in brain volume.⁸ The structural brain changes in schizophrenia may thus be classified as general (ie, reduction of brain volume, ventricular enlargement, cortical sulcal dilatation) and focal (eg, reductions in amygdala or hippocampal size). We postulated that if comorbid learning disability and schizophrenia represents a particularly severe form of schizophrenia of neurodevelopmental origin, the focal changes in schizophrenia (ie, the reduction in amygdalo-hippocampal size), will appear in an exaggerated form, and, in general, the brains of the group with both disorders will resemble those of the schizophrenic group. If, on the other hand, the schizophrenic psychosis in the comorbid group is a result of a tendency of people with learning disability to develop psychotic symptoms, probably as a consequence of overload imposed by partial comprehension, the focal changes in the comorbid group will be less than those of the schizophrenic group, and, in general, the brains of the comorbid group will resemble those of the learning-disabled group. Brain structure in all three groups will differ from that of an appropriately matched control group of normal individuals, but the nature of the difference will vary.