Abstract
Background/Objectives: Cerebral small vessel disease (CSVD) is a leading cause of cognitive decline and dementia. The comparative prognostic value of MRI-based neuroimaging markers and genetic risk factors such as the APOE ε4 allele for cognitive outcomes remains uncertain. The objectives of this study were to estimate the pooled prevalence of cognitive impairment in CSVD, evaluate the associations of key neuroimaging markers (white matter hyperintensities [WMHs], cerebral microbleeds [CMBs], lacunes) and APOE ε4 with cognitive outcomes, and assess their diagnostic performance. Methods: This study included a systematic review and meta-analysis in accordance with PRISMA and MOOSE guidelines, searching five databases (2005-2025). Eligible studies included adults with CSVD and MRI-visible markers reporting cognitive outcomes (mild cognitive impairment [MCI], global cognitive impairment [GCI], all-cause dementia [ACD], vascular dementia [VaD], and Alzheimer's disease [AD]). Thirty-nine studies comprising 18,425 participants were included. Pooled prevalence and associations were estimated using random-effects models, and diagnostic accuracy was evaluated. Certainty of evidence was assessed using the GRADE framework. Results: The pooled prevalence of GCI in CSVD was 57% (95% CI: 51-62%), while MCI prevalence was 46% (95% CI: 42-51%). WMHs were strongly associated with VaD (OR 10.35, 95% CI: 7.32-14.64), lacunes with ACD (OR 3.18, 95% CI: 1.24-8.20), and CMBs with AD (OR 1.52, 95% CI: 1.04-2.24). APOE ε4 carriage increased the risk of GCI (OR 1.80, 95% CI: 1.41-2.29). Across markers, diagnostic sensitivity was low, specificity was moderate-to-high, and AUROC values were modest. GRADE certainty ranged from low to moderate, with the highest confidence for WMHs and VaD. Conclusions: CSVD-related MRI markers and APOE ε4 are significantly associated with both early and late cognitive outcomes, supporting the integrated vascular-neurodegenerative continuum. The limited diagnostic sensitivity and variable certainty of evidence highlight the need for harmonized definitions, lesion quantification, and multimodal imaging-genetic approaches to improve early detection and risk stratification of CSVD-related cognitive impairment.