Cognitive and visual processing performance in Parkinson's disease patients with vs without visual hallucinations: A meta-analysis.

Importance: Cognitive and visual impairments in Parkinson's Disease Psychosis (PDP) raise the question of whether a specific profile of impaired cognition and visual function is linked to vulnerability to visual hallucinations (VHs). Previous studies have limited sample sizes and only included a sub-sample of tests. This is the first meta-analysis quantifying visuo-cognitive impairments in PDP patients across a spectrum of tests and taking into account potential confounding factors such as levodopa medication, illness duration and general cognitive ability.

Objective: Compare visual processing and cognitive performance between PD patients with and without VHs (PDVH and PDnoVH).

Methods: Four databases (PubMed, PsychINFO, Scopus, WebOfScience) were searched for studies on visual and/or cognitive performance of PDnoVH and PDVH published up to 02/2020. For each task, means and SDs were extracted and standardized-mean-differences (SMDs) between-groups calculated. Effect-sizes (Hedges' g) were calculated for all comparisons and synthesized in random-effects meta-analyses with robust-variance-estimation (accounting for multiple correlated measures within each study per cognitive/visual domain). Publication bias was assessed with funnel plots and Egger intercept.

Results: N = 99 studies including 2508 PDVH patients (mean age 68.4 years) and 5318 PDnoVH (mean age 66.4 years) were included in the seven meta-analyses. PDVH patients performed worse than PDnoVH across all measures of cognition and visual processing, with the greatest between-group effect-sizes in executive functions, attention, episodic memory and visual processing. Study characteristics were not significantly associated with between-group differences in the domains investigated. Age-differences were significantly associated with performance differences in general cognition, working memory and executive functions.

Conclusion: Models of PDVH need to incorporate a wider range of cognitive and processing domains than currently included. There is a need for studies disentangling the temporal relationship between cognitive/visual deficits and VHs as early identification of risk before the onset of VHs could mitigate later outcomes such as progression to dementia.