Alzheimer's Imaging Consortium

The relationship between falls and dementia risk remains unclear. This study aims to investigate this association and assess the impact of genetic risk on this relationship.

We included 351,770 participants from the UK Biobank who were free from dementia at baseline. A Cox model was used to estimate the relationship between falls and dementia risk.

After a median follow-up of 13.6 years, 5,821 developed dementia. Compared to individuals without a history of falls, individuals with single and multiple falls had hazard ratios (HRs) for dementia of 1.24 (95% CI, 1.16-1.33) and 1.79 (95% CI, 1.64-1.94), respectively. This risk was more pronounced in individuals with better cognitive function at baseline. Those experienced single falls demonstrated reduced whole-brain gray matter volume (β = -1526.601, P = 0.036) and parahippocampal gyrus volume (β_left = -22.420, P = 0.004; β_right = -16.862, P = 0.037), whereas those experienced multiple falls, in addition to these findings, demonstrated increased deep white matter hyperintensities (β = 76.602, P = 0.023) and decreased gray matter volume in the hippocampus (β_left = -10.940, P = 0.027; β_right = -13.641, P = 0.008). Furthermore, falls and polygenic risk scores showed a significant interaction (P = 0.001), with higher dementia risk observed in individuals with multiple falls and high genetic risk.

Falls are an early predictor of dementia risk, and genetic susceptibility facilitates this association. Falls indicate potential atrophy in cognition-related brain regions. Falls are a simple and easily observable sign for dementia screening.