Objective: To investigate the effect of white matter hyperintensity (WMH) severity on cognitive function according to presence of the apolipoprotein E (APOE) ε4 allele.

Method: From participants in a nationwide, multicenter, hospital-based cohort study of dementia by the Clinical Research Center for Dementia of South Korea (November 2005 to December 2011), data for 5,077 elderly subjects (mean [SD] age = 71.37 [8.40] years) who had available data for APOE genotype and WMH severity were studied retrospectively. We used the diagnostic criteria for mild cognitive impairment proposed by Petersen et al; the diagnostic criteria for vascular dementia included in DSM-IV; and, for probable Alzheimer's disease, the criteria issued by the National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association. WMH severity was evaluated using modified criteria of Fazekas et al and Scheltens et al using T2 axial or fluid-attenuated inversion recovery magnetic resonance images, yielding 3 groups for WMH severity level. APOE genotype was determined by analysis of venous blood, and all participants were classified into 2 groups depending on presence or absence of the APOE ε4 allele. The Seoul Neuropsychological Screening Battery-Dementia Version was used for all subjects. Cognitive impairment, classified by 6 cognitive test scores, was the primary outcome measure. Using multiple logistic regression, we investigated which cognitive domains were associated with WMH severity and the APOE ε4 allele, and, using analysis of covariance, we examined the interaction effects of these 2 factors on cognitive test scores.

Results: After multivariable adjustments, logistic regression analyses showed that WMH severity was associated with higher odds of cognitive impairment on frontal/executive function tests in both APOE ε4 carriers (odds ratio [OR] = 2.49; 95% CI, 1.65-3.76) and noncarriers (OR = 2.36; 95% CI, 1.83-3.03). WMH severity was not significantly associated with memory function in APOE ε4 carriers: for verbal memory, ε4 noncarriers had an OR of 1.44 (95% CI, 1.13-1.84), and ε4 carriers had an OR of 1.36 (95% CI, 0.87-2.04); for visuospatial memory, ε4 noncarriers had an OR of 1.86 (95% CI, 1.45-2.37), and ε4 carriers had an OR of 1.35 (95% CI, 0.89-2.04). Moreover, a significant interaction effect between APOE ε4 and WMH severity was confirmed on memory tests by analysis of covariance (verbal memory: F = 3.40, P = .033; visuospatial memory: F = 8.49, P < .001).

Conclusions: Severe WMHs appear to be predominantly associated with frontal/executive dysfunction, irrespective of APOE ε4 allele presence. WMH severity and APOE ε4 had an interactive effect on memory function, with WMH severity affecting memory impairment only in APOE ε4 noncarriers.