Relationships among the metabolic patterns that correlate with mnemonic, visuospatial, and mood symptoms in Parkinson's disease

Objective: A multivariate analysis of baseline brain metabolism was used to investigate the relationships among pathophysiological mechanisms responsible for cognitive dysfunction and dysphoria in nondemented patients with Parkinson's disease. Method: Using [F-18]fluorodeoxyglucose positron emission tomography and neuropsychological tests, the authors studied 15 nondemented patients who had Parkinson's disease without major depression (DSM-III-R). Their mean age was 59.2 years (SD=9.2), the mean rating of Parkinson's disease stage (Hoehn and Yahr scale) was 3.3 (SD=0.9), and all had Mini-Mental State Examination scores of 24 or higher. To identify specific regional patterns of brain metabolism associated with abnormal cognitive and mood functioning, the data were analyzed by using brain-behavior partial least squares, This multivariate voxel-based analysis allowed detection of significant topographic patterns of metabolic activity and quantification of the extent to which each topographic pattern correlated with scores on mnemonic, visuospatial, and dysphoric tests. Results: Two significant, independent topographic patterns were identified. Pattern 1 included parieto-occipito-temporal and medial temporal brain regions, and pattern 2 included the lateral frontal and anterior limbic cortex. Patterns 1 and 2 exhibited a double dissociation in their behavioral correlates: pattern 1 correlated with both visuospatial and mnemonic functioning but not with dysphoria; pattern 2 correlated with dysphoria but not with the cognitive measures. Conclusions: The authors used the independence of topographic patterns and the size of correlations between topographic patterns and behavior to infer relationships among the pathophysiological processes responsible for the correlations. The finding that mildly abnormal mnemonic and visuospatial functioning correlated with the same topographic pattern suggests that a common pathophysiology underlies this marker of cognition in Parkinson's disease. By contrast, the independence of the two topographic patterns supports the notion that different mechanisms underlie cognitive and dysphoric symptoms in nondemented patients with Parkinson's disease.