Abstract
BACKGROUND: Physical violence and aggression (PVA), defined as behaviors with the potential to cause bodily injury, are unfortunate risks in the management of all-cause neurodegenerative dementias. While dementia in Parkinson's disease (PD) may not be evident for many years after clinical onset, neuropsychiatric disturbances occur at all stages of the disease. At issue is whether PVA in PD is associated with clinical factors that can be targets for prevention and management in the absence of a prevailing dementia syndrome. OBJECTIVE: This systematic review examined the extent to which PVA in PD without dementia is a clinically significant concern and whether it is associated with factors that could warrant proactive management. METHODS: A systematic search of 9 electronic databases used MeSH headings and equivalent terms for PD, aggression, and violence. Eligible manuscripts were original articles that were published in peer-reviewed journals and reported on adults with PD in the awake state with PVA as possible outcomes. Extracted data included study design, PD ascertainment methods and characteristics, PVA assessment methods, subject demographics, psychiatric and medical comorbidities, and pertinent results. Inciting and confounding factors were extracted from case reports. Quality assessment tools were applied in accordance with the study design (e.g., observational, qualitative, or case report). RESULTS: The search identified 10 manuscripts: 2 observational quantitative studies (total n with PD = 545), 1 qualitative study (n with PD = 20), and 7 case reports (n = 7). The observational studies suggested that PVA is less common than other neuropsychiatric disturbances, but heterogeneous methods and quality concerns prevented further conclusions. In the case reports, all patients were male, and most were early onset. In 6 of the reports, PVA occurred in the context of bilateral subthalamic nucleus deep brain stimulation. CONCLUSIONS: PVA, while relatively rare in PD, can be a significant management issue that is associated with select premorbid characteristics and antiparkinsonian motor treatments. As PVA may be under-reported, further understanding of its frequency, causes, risk factors, and outcomes would benefit from its systematic assessment, ideally using self-report and informant-based questionnaires.