Cerebral disease with more prominent left sided cerebral involvement may be more closely associated with psychotic phenomena; a comprehensive review of cerebral laterality in relation to psychosis has suggested that a special, although unclear, pathophysiology may be at work.1

A review of patients with treated Parkinson’s disease found an overall incidence of psychiatric side effects of 20% (range 10–50) in 908 patients treated in major studies.2 Psychoses—that is, hallucinations and delusions—occurred with a frequency of 4% and were more likely to occur with concurrent dementia, increasing age and use of higher dose of levodopa. It is assumed that the main precipitant of psychotic phenomena in Parkinson’s disease is dopaminergic excess secondary to treatment. Cognitive impairment, which has been reported to occur in as many as 29% of patients with Parkinson’s disease, associated with increased age and older age at onset, may also increase vulnerability to psychosis.

Asymmetry in Parkinson’s disease is said to remain unchanged over time,3 and patients with unilateral onset of Parkinson’s disease manifestations have greater degeneration of the contralateral substantia nigra at postmortem examination.4 If predominantly left sided pathology increases the vulnerability to psychotic phenomena then initial right sided predominance of parkinsonian symptoms and signs might be a predictor of increased vulnerability. If initial right sided symptom predominance did indeed predict the subsequent development of psychotic phenomena (independent of cognitive decline, age, and medication) this might be clinically useful in identifying a patient subgroup in whom particular care is required in titrating medication.

A retrospective review of all case notes of patients with Parkinson’s disease, identified by their presence on the specialist Parkinson’s disease nurse register in a district general hospital, was carried out (a) to evaluate the presence of psychotic symptoms using a checklist, (b) to record asymmetry of parkinsonism both currently and at onset, and (c) to record handedness. Psychotic symptoms (delusions and hallucinations) were only noted when they occurred outside an acute confusional state. All patients had been diagnosed by one of two consultant neurologists. The level of medication, both current and while having psychotic symptoms, cognitive assessment, number of years of illness, and demographic variables were also recorded. Patients who had had a psychotic episode were compared with the remainder with respect to asymmetry of symptoms at onset of illness, age, duration of illness, medication levels and demographic variables. This was repeated with a subgroup of patients with no cognitive deficit. Logistic regression analysis (forward stepwise) was carried out with psychosis as the dependent variable.

The case notes of 100 patients were reviewed. There were 51 men and 49 women in the sample. Fifty one patients were right handed, four were left handed, and in 45 handedness was unknown. Dementia was noted in 30 patients, with “memory difficulties” recorded in a further eight. Hallucinations were recorded in 28 patients and delusions in six. The patients with hallucinations or delusions were classified as the psychosis group (n=30), and those with dementia and memory difficulties as the cognitively impaired group (n=38), for analysis. Details of illness, for the whole sample and separated into psychotic and non-psychotic subgroup, side of onset, and the demographic data, are shown in the table.

Comparison of side of onset with presence of psychosis disclosed more patients with right sided onset of parkinsonian symptoms having psychotic symptoms, although this was not significant (χ²=3.0, df=1 p<0.09; table). Presence of psychosis was significantly associated with presence of cognitive decline. Eighteen out of 38 in the cognitively impaired group were psychotic by comparison with 12 out of 62 in the cognitively intact group (χ²=13.5, df=1, p<0.003). Presence of psychosis was also associated with duration of illness (t test=2.69, df=36, p<0.02). There were no significant differences between the psychotic symptom group and the other patient groups in age, age at onset of symptoms, and dosage of current medication. Comparative dosage of levodopa and selegiline are shown in the table. Benzhexol, bromocriptine, pergolide, and orphenadrine were taken by 10, 14, 13, and seven patients respectively, and showed no differences between the groups. Furthermore there were no significant differences in the dose of current medication and the dose noted when undergoing a psychotic episode.

Cognitive decline was associated with increasing age but was not related to age of onset or dosage of medication. When patients with cognitive decline (n=38) were removed from the analysis, right sided onset of symptoms was significantly related to the presence of psychosis (χ²=5.0, df=1, p<0.03) in the remainder. In this subsample there were no significant differences between left and right side onset for age, duration of illness, or dosage of different medication.

Logistic regression analysis of the total sample, with psychosis as the dependent variable, confirmed the association between psychosis and cognitive decline (t=3.89, p<0.003) and increased duration of illness (t=2.64, p<0.02). There were no other significant contributing variables, although side of onset was the strongest associated variable remaining (t=1.69, p<0.09). However, when the logistic regression was repeated with the subsample without cognitive decline, right sided onset was the only variable significantly associated with psychotic symptoms (t=2.30, p<0.03).

Our results show only a trend, in the sample as a whole, linking right sided symptoms at onset and the subsequent development of psychosis; perhaps unsurprising in view of the confounding effect of cognitive impairment. In the cognitively intact subsample there was a significant association with right sided onset of symptoms of Parkinson’s disease. This suggests that damage to left hemispheric structures involved in Parkinson’s disease is associated with a predisposition to psychosis. Our results do not support an iatrogenic dopaminergic excess as a cause of psychosis; this may be because of subsequent dosage adjustment, or the explanation may lie in asymmetric upregulation of dopaminergic receptors and supersensitivity to dopaminergic therapy at equivalent doses of medication.

The limitations of using retrospective data are well recognised and the possibility of mild cognitive dysfunction not being detected in the cognitively intact group needs to be recognised. However, this preliminary study provides support for our a priori hypothesis that right sided predominance of neurological deficit at the onset of Parkinson’s disease predicts the subsequent development of psychosis.