Frontal White Matter Hyperintensities and Executive Functioning Performance in Older Adults

Abstract

Frontal lobe structures decline faster than most other brain regions in older adults. Age-related change in the frontal lobe is associated with poorer executive function (e.g., working memory, switching/set-shifting, and inhibitory control). The effects and presence of frontal lobe white matter hyperintensities (WMH) on executive function in normal aging is relatively unknown. The current study assessed relationships between region-specific frontal WMH load and cognitive performance in healthy older adults using three executive function tasks from the NIH Toolbox (NIHTB) Cognition Battery. A cohort of 279 healthy older adults ages 65-88 completed NIHTB and 3T T1-weighted and FLAIR MRI. Lesion Segmentation Toolbox quantified WMH volume and generated lesion probability maps. Individual lesion maps were registered to the Desikan-Killiany atlas in FreeSurfer 6.0 to define regions of interest (ROI). Independent linear regressions assessed relationships between executive function performance and region-specific WMH in frontal lobe ROIs. All models included age, sex, education, estimated total intracranial volume, multi-site scanner differences, and cardiovascular disease risk using Framingham criteria as covariates. Poorer set-shifting performance was associated with greater WMH load in three frontal ROIs including bilateral superior frontal (left β = -0.18, FDR-p = 0.02; right β = -0.20, FDR-p = 0.01) and right medial orbitofrontal (β = -0.17, FDR-p = 0.02). Poorer inhibitory performance associated with higher WMH load in one frontal ROI, the right superior frontal (right β = -0.21, FDR-p = 0.01). There were no significant associations between working memory and WMH in frontal ROIs. Our study demonstrates that location and pattern of frontal WMH may be important to assess when examining age-related differences in cognitive functions involving switching/set-shifting and inhibition. On the other hand, working memory performance was not related to presence of frontal WMH in this sample. These data suggest that WMH may contribute selectively to age-related declines in executive function. Findings emerged beyond predictors known to be associated with WMH presence, including age and cardiovascular disease risk. The spread of WMH within the frontal lobes may play a key role in the neuropsychological profile of cognitive aging. Further research should explore whether early intervention on modifiable vascular factors or cognitive interventions targeted for executive abilities may help mitigate the effect of frontal WMH on executive function.

Keywords: FLAIR MRI; NIH toolbox; cognitive aging; executive function; frontal lobes; region-specific hyperintensities; white matter hyperintensities.

Figure 1

Summary conceptual model of MRI processing. Participant's MPRAGE and FLAIR were collected and processed using FreeSurfer and Lesion Segmentation Toolbox. Resultant images were transformed to FreeSurfer space, the Desikan-Killiany atlas, via FSL FLIRT. Regional-specific lesion volumes were calculated for final analysis.

Figure 2

Region-specific WMH Desikan-Killiany ROIs included in final analysis (n = 16). Regions included right and left hemisphere Superior Frontal (red), Rostral (blue), and Caudal (light blue) Middle Frontal, Medial (green) and Lateral Orbitofrontal (yellow), Pars Opercularis (orange), Pars Triangularis (cyan), and Periventricular/Deep Frontal (beige).

Figure 3

Region-specific WMH ROIs significant for (A) dimensional change card sort test and (B) Flanker inhibitory control and attention test. Significant Desikan-Killiany regions rendered as 3D isosurface mesh files and presented in surface. R, right; L, left; A, anterior.

Figure 4

Region-Specific WMH ROIs significant for (A–C) dimensional change card sort test and (D) Flanker inhibitory control and attention test. Partial regression scatter plots for relationships between NIHTB DCCS and Flanker performance and significant Desikan-Killiany frontal regions WMH load for (A) left hemisphere superior frontal, (B,D) right hemisphere superior frontal, (C) right hemisphere medial orbitofrontal.