Sex-specific risk factors and clinical dementia outcomes for white matter hyperintensities in a large South Korean cohort

Abstract

Objective: White matter hyperintensities (WMH) on brain MRI images are the most common feature of cerebral small vessel disease (CSVD). Studies have yielded divergent findings on the modifiable risk factors for WMH and WMH's impact on cognitive decline. Mounting evidence suggests sex differences in WMH burden and subsequent effects on cognition. Thus, we aimed to identify sex-specific modifiable risk factors for WMH. We then explored whether there were sex-specific associations of WMH to longitudinal clinical dementia outcomes.

Methods: Participants aged 49-89 years were recruited at memory clinics and underwent a T2-weighted fluid-attenuated inversion recovery (FLAIR) 3T MRI scan to measure WMH volume. Participants were then recruited for two additional follow-up visits, 1-2 years apart, where clinical dementia rating sum of boxes (CDR-SB) scores were measured. We first explored which known modifiable risk factors for WMH were significant when tested for a sex-interaction effect. We additionally tested which risk factors were significant when stratified by sex. We then tested to see whether WMH is longitudinally associated with clinical dementia that is sex-specific.

Results: The study utilized data from 713 participants (241 males, 472 females) with a mean age of 72.3 years and 72.8 years for males and females, respectively. 57.3% and 59.5% of participants were diagnosed with mild cognitive impairment (MCI) for males and females, respectively. 40.7% and 39.4% were diagnosed with dementia for males and females, respectively. Of the 713 participants, 181 participants had CDR-SB scores available for three longitudinal time points. Compared to males, females showed stronger association of age to WMH volume. Type 2 Diabetes was associated with greater WMH burden in females but not males. Finally, baseline WMH burden was associated with worse clinical dementia outcomes longitudinally in females but not in males.

Discussion: Older females have an accelerated increase in cerebrovascular burden as they age, and subsequently are more vulnerable to clinical dementia decline due to CSVD. Additionally, females are more susceptible to the cerebrovascular consequences of diabetes. These findings emphasize the importance of considering sex when examining the consequences of CSVD. Future research should explore the underlying mechanisms driving these sex differences and personalized prevention and treatment strategies.

Clinical trial registration: The BICWALZS is registered in the Korean National Clinical Trial Registry (Clinical Research Information Service; identifier, KCT0003391). Registration Date 2018/12/14.

Fig. 1

Older females appear to have an accelerated increase in cerebrovascular burden as they age and are more susceptible to the cerebrovascular consequences of diabetes. A.) Females are observed to have a higher increase in WMHV as they age compared to males. B.) Diabetic females were associated with larger WMHV compared to non-diabetic females. There were no differences observed for males. For visualization purposes, WMHV is plotted versus age and grouped by sex and type 2 diabetes status. C.) Higher HbA_1c levels were associated with larger WMHV for females but not males, but the interaction effect for HbA_1c*sex did not survive multiple comparisons. WMHV = white matter hyperintensity volume

Fig. 2

Sex-specific risk factors for WMH. Beta coefficient values from linear regression models for individual risk factors are displayed

Fig. 3

Higher WMHV is associated with clinical dementia outcomes over two years among females but not males. A.) A significant interaction effect between WMHV and visit time on CDR sum of boxes was observed for females in a linear mixed effect model. The predicted slopes for a participant with a mean and +- 1 standard deviation of WMHV for females is plotted. B.) The predicted slopes for a participant with a mean and +- 1 standard deviation of WMHV for males is plotted. There was no observed interaction effect between WMHV and visit time on CDR sum of boxes score for males Follow-up visits were acquired approximately one year after the prior visit. WMHV = white matter hyperintensity volume