COLOR VISION IMPAIRMENT IN DEMENTIA WITH LEWY BODIES: A NOVEL AND HIGHLY SPECIFIC DISTINGUISHING FEATURE FROM ALZHEIMER DEMENTIA

Abstract

Dementia with Lewy Bodies (DLB) is frequently misdiagnosed for Alzheimer dementia (AD), especially in its earlier stages. Identification of clinical features that lend to accurate and earlier diagnosis remains paramount. We characterized the prevalence of color vision impairment (CVI) in DLB vs. AD patients and investigated how CVI is associated with other features of DLB. We retrospectively reviewed charts of patients with AD, DLB, and patients with mild cognitive impairment suspected to be in the prodromal phase of DLB (pro-DLB) or AD (pro-AD). All patients underwent an online 15-hue color vision arrangement test as part of the clinical assessment. The median age of the 47 included patients was 76 years and 47% were female. No significant differences in age or gender existed between patients with AD (n=14 patients) and pro-AD (n=6) versus pro-DLB (n=7) and DLB patients (n=20). Three patients (2 AD, 1 DLB) had CVI history from young age. New-onset CVI prevalence was: pro-AD patients (0%), AD patients (8%), pro-DLB patients (29%), and DLB patients (79%). CVI type was as follows: AD (1 protanopia), pro-DLB (2 deutanopia), DLB (10 protanopia, 3 deutanopia, and 2 tritanopia). The clinical signs and symptoms of patients with DLB were analyzed to evaluate factors associated with color vision impairment. In the 20 DLB patients, the symptom prevalence was: visual hallucinations (90%), tremor (80%), REM-behavior disorder (75%), postural instability (65%), bradykinesia (55%), rigidity (55%), and autonomic instability (50%). Normal mixtures iterative clustering analysis identified two DLB patient groups. While Cluster1 (n=10) and Cluster2 (n=8) had similar total number of Parkinsonian symptoms (2.4 vs. 2.0) and autonomic instability (50% vs. 38%), compared to Cluster2, Cluster1 patients were more likely to have CVI (100% vs. 50%), lower MOCA scores (14 vs. 19), lower visuospatial scores (1.4 vs. 2.5), and REM behavior disturbances (100% vs. 38%). Color vision impairment in DLB patients showed prevalence similar to the core features of DLB (∼80%) and was highly specific for DLB versus AD. Further larger scale investigation will determine if incorporation of color vision tests improve the accuracy of earlier diagnosis of DLB.