Evidence from patients with amnesia suggests that recognition memory span tasks engage both long-term memory (i.e., secondary memory) processes mediated by the diencephalic-medial temporal lobe memory system and working memory processes mediated by fronto-striatal systems. Thus, the recognition memory span task may be particularly effective for detecting memory deficits in disorders that disrupt both memory systems. The presence of unique pathology in fronto-striatal circuits in Dementia with Lewy Bodies (DLB) compared to AD suggests that performance on the recognition memory span task might be differentially affected in the two disorders even though they have quantitatively similar deficits in secondary memory. In the present study, patients with autopsy-confirmed DLB or AD, and Normal Control (NC) participants, were tested on separate recognition memory span tasks that required them to retain increasing amounts of verbal, spatial, or visual object (i.e., faces) information across trials. Results showed that recognition memory spans for verbal and spatial stimuli, but not face stimuli, were lower in patients with DLB than in those with AD, and more impaired relative to NC performance. This was despite similar deficits in the two patient groups on independent measures of secondary memory such as the total number of words recalled from long-term storage on the Buschke Selective Reminding Test. The disproportionate vulnerability of recognition memory span task performance in DLB compared to AD may be due to greater fronto-striatal involvement in DLB and a corresponding decrement in cooperative interaction between working memory and secondary memory processes. Assessment of recognition memory span may contribute to the ability to distinguish between DLB and AD relatively early in the course of disease.

Visual search is an aspect of visual cognition that may be more impaired in Dementia with Lewy Bodies (DLB) than Alzheimer's disease (AD). To assess this possibility, the present study compared patients with DLB (n = 17), AD (n = 30), or Parkinson's disease with dementia (PDD; n = 10) to non-demented patients with PD (n = 18) and normal control (NC) participants (n = 13) on single-feature and feature-conjunction visual search tasks. In the single-feature task participants had to determine if a target stimulus (i.e., a black dot) was present among 3, 6, or 12 distractor stimuli (i.e., white dots) that differed in one salient feature. In the feature-conjunction task participants had to determine if a target stimulus (i.e., a black circle) was present among 3, 6, or 12 distractor stimuli (i.e., white dots and black squares) that shared either of the target's salient features. Results showed that target detection time in the single-feature task was not influenced by the number of distractors (i.e., "pop-out" effect) for any of the groups. In contrast, target detection time increased as the number of distractors increased in the feature-conjunction task for all groups, but more so for patients with AD or DLB than for any of the other groups. These results suggest that the single-feature search "pop-out" effect is preserved in DLB and AD patients, whereas ability to perform the feature-conjunction search is impaired. This pattern of preserved single-feature search with impaired feature-conjunction search is consistent with a deficit in feature binding that may be mediated by abnormalities in networks involving the dorsal occipito-parietal cortex.

Objective

Visual processing abilities of patients with dementia with Lewy bodies (DLB) or Alzheimer disease (AD) dementia were assessed psychophysically using a simple horizontal motion discrimination task that engages the dorsal visual processing stream.

Methods

Participants included patients with mild dementia with DLB, AD dementia or Parkinson disease (PD) with dementia (PDD), without dementia with PD, and normal controls. Participants indicated the left or right direction of coherently moving dots that were embedded within dynamic visual noise provided by randomly moving dots. The proportion of coherently moving dots was increased or decreased across trials to determine a threshold at which participants could correctly indicate their direction with greater than 80% accuracy.

Results

Motion discrimination thresholds of patients with DLB and PDD were comparable and significantly higher (i.e., worse) than those of patients with AD dementia. The thresholds of patients with AD dementia and patients with PD were normal. These results were confirmed in subgroups of patients with DLB/PDD and AD dementia with autopsy-confirmed disease. A motion discrimination threshold greater than 0.23 distinguished between DLB/PDD and AD dementia with 67% sensitivity and 85% specificity.

Conclusions

Differential deficits in detecting direction of simple horizontal motion suggest that dorsal processing stream dysfunction is greater in DLB and PDD than in AD dementia. Therefore, impaired performance on simple visual motion discrimination tasks that specifically engage occipitoparietal brain regions suggests the presence of Lewy body pathology.