Serial assessments of cognition, mood, and disability were carried out at nine month intervals over a 54 month period on a cohort of 87 patients with Parkinson's disease (PD) and a matched cohort of 50 control subjects. Dementia was diagnosed from data by rigorously applying DSM-III-R criteria. Initially, 6% (5187) PD patients were demented, compared with none of the 50 control subjects. A further 10 PD patients met the dementia criteria during the follow up period; this was equivalent, with survival analysis, to a cumulative incidence of 190%.. With the number of person years of observation as the denominator, the incidence was 47*6I1000 person years of observation.None of the control subjects fulfilled dementia criteria during the follow up period. The Serial assessments of patients can overcome these difficulties. Firstly, poor performance on cognitive testing on a single occasion may be seen to be transient, in which case dementia is unlikely to be the cause. Secondly, decline can be directly observed rather than inferred from the use of unvalidated tests of premorbid intelligence. Thirdly, cross sectional assessment can give only an estimate of prevalence of dementia in PD but serial assessments can enable the incidence of dementia to be estimated and may provide information about the evolution and prognosis of dementia in PD.Mayeux et al recently estimated the incidence of dementia in PD.'0 They rightly emphasised that incidence is a better indicator than prevalence of the occurrence of dementia in PD: if coincident PD and dementia result in a greater mortality than PD uncomplicated by dementia the prevalence of dementia in PD will be lowered. We carried out longitudinal assessments of cognition, mood, and disability at nine month intervals over a 54 month follow up period on a cohort of patients with PD and a matched cohort of control subjects. On the basis of the data we judged whether DSM-III-R dementia was present in each subject and, if present, when it occurred. We present our findings in the form of a survival analysis.

An alternative to the conventional separation of extrapyramidal and catatonic symptoms exists in the 'conflict of paradigms' hypothesis, which proposes that there is a relative rather than absolute distinction between the two. The hypothesis predicts that a clinical association should exist between extrapyramidal and catatonic symptoms in schizophrenia. After rating 75 schizophrenic patients, a highly significant correlation between scores on the two classes of disorder was indeed found. This was composed of separate correlations between tardive dyskinesia and 'positive' catatonic phenomena, and Parkinsonism and 'negative' catatonic phenomena. The associations were not easily attributable to confounding factors and they were supported by factor analysis.

Forty seven patients with idiopathic Parkinson's disease (PD) were assessed to ascertain the presence or absence of emotionalism. Thirty eight per cent of the sample reported being more tearful since the onset of PD than previously. Five patients (10.6%) were found to exhibit emotionalism according to specified criteria. Emotionalism in PD is not solely associated with lowered mood or global cognitive impairment.

No abstract

Matters arising 2 Secondary prevention of vascular disease by prolonged antiplatelet treatment-Antiplatelet trialists' collaboration. BMJ 1988;