Exploring the Contribution of Myelin Content in Normal Appearing White Matter to Cognitive Outcomes in Cerebral Small Vessel Disease

Abstract

Background:

Myelin damage is a salient feature in cerebral small vessel disease (cSVD). Of note, myelin damage extends into the normal appearing white matter (NAWM). Currently, the specific role of myelin content in cognition is poorly understood.

Objective:

The objective of this exploratory study was to investigate the association between NAWM myelin and cognitive function in older adults with cSVD.

Methods:

This exploratory study included 55 participants with cSVD. NAWM myelin was measured using myelin water imaging and was quantified as myelin water fraction (MWF). Assessment of cognitive function included processing speed (Trail Making Test Part A), set shifting (Trail Making Test Part B minus A), working memory (Verbal Digit Span Backwards Test), and inhibition (Stroop Test). Multiple linear regression analyses assessed the contribution of NAWM MWF on cognitive outcomes controlling for age, education, and total white matter hyperintensity volume. The overall alpha was set at ≤0.05.

Results:

After accounting for age, education, and total white matter hyperintensity volume, lower NAWM MWF was significantly associated with slower processing speed (β = –0.29, p = 0.037) and poorer working memory (β= 0.30, p = 0.048). NAWM MWF was not significantly associated with set shifting or inhibitory control (p > 0.132).

Conclusion:

Myelin loss in NAWM may play a role in the evolution of impaired processing speed and working memory in people with cSVD. Future studies, with a longitudinal design and larger sample sizes, are needed to fully elucidate the role of myelin as a potential biomarker for cognitive function.

Keywords

Cerebral small vessel disease myelin myelin water imaging executive functions vascular cognitive impairment white matter hyperintensities

INTRODUCTION

Cerebral small vessel disease (cSVD) is the second most common cause of cognitive impairment and dementia [1]. cSVD predominantly manifests as white matter hyperintensities (WMHs) and though WMHs are consistently and robustly associated with cognitive decline, the observed effect is small (r = –0.10 [95% confidence interval = –0.13 to –0.08]) [2]. The small association between WMHs and cognitive function may be due to the microstructural pathologic heterogeneity of cSVD. Postmortem pathology studies suggest that small vessel ischemic white matter microstructural damage can involve an array of abnormalities including discontinuities of the ependymal lining, gliosis, and loosening of white matter fibers. The majority of studies also report myelin loss as a salient feature [3].

Myelin is responsible for saltatory conduction of nerve impulses that allow for high-speed and high-fidelity signal transmission. Optimal nerve conduction is crucial for integrating information across dispersed neural networks that underlie cognitive abilities [4]. Myelin loss can severely compromise cerebral communication; yet, the specific role of myelin on cognitive outcomes is poorly understood. This is, in part, due to the technical limitations in quantifying myelin in vivo. Conventional T₁-weighted and T₂-weighted imaging can only capture macrostructural white matter changes such as WMHs or white matter volume [5]. Other imaging techniques, such as diffusion tensor imaging and magnetization transfer imaging, are influenced by myelin changes but are not specific to myelin content [6]. Diffusion tensor imaging metrics such as mean diffusivity or fractional anisotropy measure the movement of water within cerebral tissue, but are unable to specify the white matter substrate producing the observed signal. Diffusion tensor imaging metrics non-specifically reflect microstructural complexity, membrane permeability, axonal density, and myelination [7]. Standard magnetization transfer imaging techniques measure magnetization transfer ratio, which is strongly influenced by inflammation and edema and is not specific to myelin content [8].

Notwithstanding the limitations of diffusion tensor and magnetization transfer imaging in specifically assessing myelin content, these techniques have demonstrated the importance of white matter microstructure for cognitive function. Of particular relevance are changes to normal appearing white matter (NAWM), which may represent early microstructural damage before the formation of WMHs [3]. Disrupted NAWM integrity, characterized by increased mean diffusivity and reduced fractional anisotropy, is associated with impaired executive functions and processing speed in older adults with cSVD [9 –11]. In addition, lower NAWM magnetization transfer ratio is associated with disrupted executive functions and memory in older adults who are healthy or at increased risk for vascular disease [12, 13]. Furthermore, diffusion tensor and magnetization transfer imaging data report greater NAWM microstructural damage with increasing WMH severity [14]. Together, these studies indicate the relevance of NAWM microstructure in assessing cognitive function in people with cSVD.

Despite the growing recognition that subtle damage in NAWM microstructure impacts cognitive outcomes, the contribution of myelin content remains unclear. To better understand the role of myelin, studies need to use advanced neuroimaging techniques that allow myelin specific quantification in vivo. Myelin can be measured with myelin water imaging using a multi-component T₂-relaxation technique. Different water environments such as cerebrospinal fluid, intra- and extracellular water, and “myelin water” trapped tightly between the myelin bilayers will exhibit different T₂ components [6]. Myelin can be quantified as the ratio of myelin water to total water (the sum of all T₂ components), termed myelin water fraction (MWF). Histopathological studies report strong correlations between MWF and myelin staining in animal models and human tissue [15 –19] and MWF is a reliable marker of myelin in people with stroke [20 –22]. To date, MWF correlates with myelin content better than other magnetic resonance imaging (MRI) measures associated with myelin, such as radial diffusivity, fractional anisotropy, or magnetization transfer ratio [5, 23].

Both pathological and MRI studies demonstrate widespread microstructural white matter damage in cSVD, that includes the NAWM [3 , 25]. Though myelin is a critical component of white matter, studies specific to myelin are currently lacking. Thus, the objective of this exploratory study was to investigate the association between NAWM myelin content and cognitive function, specifically processing speed and executive functions, in older adults with cSVD.

METHODS

Study design and participants

Participants from two studies were included. Thirty-three participants were included from a randomized controlled trial assessing the effect of resistance training on cognitive function in older adults with cSVD (NCT02669394). Another 22 participants were included from a prospective study assessing changes in cognitive function between fallers and non-fallers with mild cognitive impairment. We used baseline data from both studies and only included participants who exhibited WMHs on MRI—cognitive assessments and MRI scans were completed within two weeks of each other. Participants were recruited from either the University of British Columbia Hospital Clinic for Alzheimer’s Disease and Related Disorders or from advertisements placed in the community (newspaper advertisements, flyers, and brochures).

cSVD was defined as the presence of WMHs on MRI [26] with a Fazekas score of ≥1 [27]. Participants were also cognitively impaired defined as a Montreal Cognitive Assessment score (MoCA) < 26/30 at baseline [28]. To screen for mild cognitive impairment, we used the MoCA instead of the Mini-Mental State Examination (MMSE) because the MoCA has greater sensitivity for identifying mild cognitive impairment (90%) than the MMSE (18%) [28]. In addition to the presence of cSVD and mild cognitive impairment, individuals were eligible if they: 1) were ≥50 years old; 2) had an MMSE score ≥20/30 to exclude participants with possible dementia [29]; 3) were community-dwelling (i.e., not living in a long-term care facility) and living independently and; 4) provided informed consent. Individuals were excluded if they: 1) were diagnosed with dementia or a neurodegenerative/neurological condition (e.g., Alzheimer’s disease, multiple sclerosis, Parkinson’s disease, etc.); 2) were taking medications that may negatively affect cognitive function, such as anticholinergics, including agents with pronounced anticholinergic properties (e.g., amitriptyline), major tranquilizers (typical and atypical antipsychotics), and anticonvulsants (e.g., gabapentin, valproic acid, etc.); 3) were concurrently participating in a clinical drug trial or; 4) had contraindications to MRI scanning.

Ethical approval was obtained from the University of British Columbia’s Clinical Research Ethics Board (H15-00972 and H07-01160) and the Vancouver Coastal Health Research Institute (V15-00972 and V07-0172). All subjects gave written informed consent in accordance with the Declaration of Helsinki.

Descriptive variables

Information on age, sex, body mass index (weight in kilograms/square of height in meters), education, and global cognition (MoCA and MMSE) were collected. We also acquired information on co-morbidities using the Functional Comorbidity Index [30], where a higher score indicates a greater number of comorbidities.

Measures of cognitive function

We focused on processing speed and executive functions because these cognitive processes are strongly associated with cSVD [1]. Cognitive tests were administered by trained lab personnel (L.T.B, R.C, C.K.B., S.H.D).

Trail making test part A

This test measures processing speed, specifically psychomotor speed and visual scanning speed [31]. In Part A, participants were asked to draw lines connecting encircled numbers sequentially.

Trail making test part B minus A

This test measures set shifting [31]. Participants were asked to draw lines connecting encircled numbers sequentially (Part A) and to alternate between numbers and letters (Part B). The difference in time to complete Part B and Part A was calculated; smaller difference indicates better performance.

Verbal digit span backwards test

This test measures working memory [32]. Participants are verbally presented with number sequences and were asked to repeat the number sequences in the reversed order. A lower score indicates worse working memory performance.

Stroop test

Participants completed three conditions (80 trials each): 1) reading out color words printed in black ink; 2) reading out the display color of colored-Xs; and 3) shown a page with color-words printed in incongruent colored inks, participants were asked to name the ink color in which the words were printed. To measure inhibition [31], the time difference between the third condition and second condition was calculated; smaller difference indicates better response inhibition.

MRI acquisition

Magnetic resonance data was acquired at the University of British Columbia MRI Research Centre on a 3T Philips Achieva MRI scanner (Philips Medical Systems, Best, The Netherlands) using an eight-channel sensitivity encoding head coil and parallel imaging. The following scans were collected: 1) 3D T₁-weighted MPRAGE (TR = 7.7 ms, TE =3.5 ms, TI = 800 ms, Shot TR = 1900 ms, flip angle θ= 8°, FOV = 256×200 mm, transverse orientation, 170 slices, acquired and reconstructed voxel =1.00×1.00×1.00 mm); 2) 3D T₂-weighted VISTA (TR = 2500 ms, TE = 363 ms, flip angle θ= 90°, FOV = 256×256 mm, sagittal orientation, 200 slices, acquired voxel = 1.00×1.00×1.60 mm, re-constructed voxel = 0.80×0.80×0.80 mm); 3) Pro-ton density (PD)-weighted (TR = 3000 ms, TE =30 ms, flip angle θ= 90°, FOV = 250×250 mm, sagittal orientation, 170 slices, acquired voxel =0.99×1.00×1.00 mm, reconstructed voxel = 0.98×0.98×1.00 mm); 4) whole-brain 48-echo 3D gradient and spin echo (GRASE) for T₂ measurement (TR = 1073 ms, TE = 8, 16, 24 ... 384 ms, flip angle θ= 90°, FOV = 230×190 mm, slice oversampling factor = 1.3, SENSE = 2, transverse orientation, acq-uired in-plane voxel = 0.99×2.04 mm, reconstructed in-plane voxel = 0.96×0.95 mm, 20 slices acquired at 5 mm slice thickness, 40 slices reconstructed at 2.5 mm slice thickness) [33].

WMH volume quantification

WMHs were quantified using T₂- and PD- weighted MR images. Full details on WMH segmentation procedures are described previously [34]. Briefly, WMHs were first identified and digitally marked by a radiologist. Next, WMHs were segmented by a method that automatically computed the extent of each marked lesion. Specifically, the seed points were processed by a customized Parzen windows classifier [35] to estimate the intensity distribution of the lesions. The algorithm included heuristics to optimize the accuracy of the estimated distributions by dynamically adjusting the position and the number of seed points used for the Parzen window computation, as well as a spatial method that approximated visual shape partitioning to identify areas that were likely to be false positives [34]. The WMH lesion masks were then used to quantify WMH volumes in cubic millimeters (mm³). This segmentation method has been validated in large data sets with a wide range of lesion loads, has high agreement with a gold standard manual method and is robust to variations in seed point placements [34]. In addition, all lesion masks were visually inspected to ensure accuracy.

NAWM MWF quantification

T₂ relaxation analysis used a modified non-negative least squares approach with no a priori assumptions about the number of T₂ components [36] to reconstruct MWF maps from the acquired T₂ decay curves. This algorithm included correction for stimulated echo contamination using an extended phase graph algorithm as well as a non-local spatial regulariser to make the fit more robust against noise in the time domain. Also, this approach improved border delineation of structures by using weighted averaging to avoid combining T₂ distributions that are very different [36]. The T₂ signal arising from myelin water was defined as 15–40 ms [6]. Voxel-wise MWF was quantified as the amplitude of the short T₂ component (myelin water) divided by the amplitude across the entire distribution (total water, the sum of all T₂ components) using in-house software code developed at the University of British Columbia [33, 36].

A NAWM mask was created for each participant using the first echo of the GRASE image. First, white matter masks were generated from high-resolution 3DT₁ images using the automated brain segmentation algorithm in FSL-FAST [37], followed by in-plane thresholding of voxels <1. These white matter masks were then registered to the GRASE image using FLIRT (transformation = 6 DOF rigid body; interpolation = nearest neighbor). To remove WMHs from the white matter mask, the T₂ image was registered to the GRASE image using FLIRT (transformation = 6 DOF rigid body; interpolation = spline) to obtain a transformation matrix. Using the applywarp command in FSL, the WMH mask was resampled to the GRASE image using the transformation matrix and nearest neighbor interpolation. The resampled WMH mask was then subtracted from the white matter mask to generate a NAWM mask for each participant. Each participant’s NAWM mask was multiplied by their MWF map to generate a NAWM MWF map in native space (Fig. 1). We used linear registration because we co-registered images from the same scan session of the same participant and found that this method worked well. All NAWM masks were visually inspected and edited as needed.

Fig. 1

Sample WMH and NAWM MWF segmentations. A) PD-weighted scan for WMH segmentation; B) T₂-weighted scan for WMH segmentation; C) WMH mask; D) MWF map; E) NAWM MWF mask; F) NAWM MWF map.

To determine average NAWM MWF for each participant, histograms were calculated by summing the number of voxels with MWF values between 0 and 30% into 100 uniform bins. To account for differences in brain size and tissue volume, each participant’s histogram was scaled by the total number of data points used in the histogram calculation. The histogram mean reflects the average value of the NAWM MWF map.

Statistical analysis

All statistical analyses were performed using Statistical Package for the Social Sciences 22.0. A bivariate correlation analysis assessed the relationship between NAWM MWF and total WMH volume. Multiple linear regression analyses were conducted to obtain estimates for the independent contribution of NAWM MWF on cognitive function. Four separate models were constructed to assess the effect of NAWM MWF on processing speed (Trail Making Test Part A), set shifting (Trail Making Test Part B minus A), working memory (Verbal Digit Span Backwards Test), and inhibition (Stroop Test). Each statistical model controlled for age, education, and total WMH volume. NAWM MWF was entered last to determine its unique and additional contribution to cognitive function. We did not correct for multiple comparisons as Type II error is a consideration in exploratory analyses [38]; therefore, statistical significance was based on an alpha of ≤0.05 and we report 95% confidence intervals [39, 40].

For each regression model, we computed collinearity statistics (tolerance and variance inflation factor), histograms of the residuals, and scatterplots of the predicted versus residual values to ensure that the assumptions of linear regression were met. In all models, multicollinearity was not an issue among predictor variables, and the residuals were normally distributed and homoscedastic.

RESULTS

Participants

Fifty-five participants (34 females) were included in this cross-sectional study. The mean age was 75.7±5.8 years. The mean MoCA score was 21.4 and MMSE score was 27.4. Table 1 reports all descriptive characteristics.

Table 1

Descriptive Characteristics

Variable (N = 55)	Mean or No.	SD or %	Range
Age (y)	75.7	5.8	57.0 –90.0
Female sex, No. (%)	34	61.8	-
MoCA (max. score 30)	21.4	3.5	11.0 –26.0
MMSE (max. score 30)	27.4	1.8	22.0 –30.0
Body mass index (kg/m²)	26.6	4.7	17.5 –43.6
Education, No. (%)
Less than grade 9	2	3.6	-
Grades 9–13, without certificate	2	3.6	-
Highschool certificate	8	14.5	-
Trades or professional certificate or diploma	3	5.5	-
Some university training	10	18.2	-
University degree	30	54.5	-
Functional Comorbidity Index	3.2	1.8	0.0 –6.0
Neuroimaging biomarkers
NAWM MWF	0.10	0.01	0.07 –0.13
WMH volume (mm³), median (interquartile range)	4435.2	10590.9	51.0 –39796.7
Cognitive function tests
Trail Making Test Part A (s)	41.0	17.0	23.0 –114.5
Trail Making Test Part B minus A (s)	88.5	106.2	18.4 –636.3
Verbal Digit Span Backward Test (max. score of 12)	5.2	1.9	2.0 –11.0
Stroop Test (s)	56.5	34.4	9.3 –229.8

MoCA, Montreal Cognitive Assessment; MMSE, Mini-Mental State Examination; NAWM, normal appearing white matter; MWF, myelin water fraction; WMH, white matter hyperintensity.

Bivariate correlation

NAWM MWF was significantly correlated with total WMH volume (r = –0.29, p = 0.031, Fig. 2).

Fig. 2

Bivariate correlation analysis.

Associations between NAWM MWF and cognitive function

For processing speed, lower NAWM MWF was significantly associated with slow processing speed (Trail Making Test Part A; β= –0.29; p = 0.037); the total variance accounted for by the model was 24.5%. For executive functions, NAWM MWF was not significantly associated with set shifting (Trail Making Test Part B minus A; β= –0.20; p = 0.132). Lower NAWM MWF was significantly associated with worse working memory (Verbal Digit Span Backwards Test, β= 0.30; p = 0.048); the total variance accounted by the model was 8.6%. NAWM MWF was not significantly associated with inhibitory control (Stroop Test; β= –0.02; p = 0.895). All analyses accounted for age, education, and total WMH volume. Please refer to Table 2 for detailed regression results.

Table 2

Multiple Linear Regression Results

Dependent variables	R²	Adjusted R²	R² change	Standardized β for NAWM MWF	Unstandardized B (standard error) for NAWM MWF	Confidence Interval for B	p
Trail Making Test Part A
Step 1	0.08	0.04	0.08	–	–	–	0.126
Step 2	0.18	0.13	0.10	–	–	–	0.017
Step 3	0.25	0.19	0.07^*	–0.29	–385.17 (179.65)	–746.00 ––24.33	0.037
Trail Making Test Part B-A
Step 1	0.25	0.22	0.25	–	–	–	0.001
Step 2	0.27	0.23	0.02	–	–	–	0.241
Step 3	0.30	0.25	0.03	–0.20	–1647.25 (1076.69)	–3809.56 –515.35	0.132
Verbal Digit Span Backwards
Step 1	0.01	–0.03	0.01	–	–	–	0.767
Step 2	0.01	–0.05	0.00	–	–	–	0.833
Step 3	0.09	0.01	0.08^*	0.30	45.25 (22.34)	0.39 –90.11	0.048
Stroop Test
Step 1	0.05	0.01	0.05	–	–	–	0.303
Step 2	0.09	0.04	0.04	–	–	–	0.106
Step 3	0.09	0.02	0.00	–0.02	–52.57 (396.87)	–849.70 –744.60	0.895

Independent Variables in Step 1 = age and education; Independent Variables in Step 2 = age, education, and white matter hyperintensity volume; Independent Variables in Step 3 = age, education, white matter hyperintensity volume, and normal appearing white matter (NAWM) myelin water fraction (MWF). ^*Significant at p≤0.05.

DISCUSSION

In this study, we found that greater WMH volume is associated with reduced myelin content in the NAWM. This further affirms reports from histology and MRI studies indicating deterioration in the NAWM in cSVD [3 , 41]. Critically, we found that lower NAWM myelin content may be associated with slow processing speed and worse working memory performance after accounting for the effects of age, education, and total WMH volume. Our results provide preliminary evidence that myelin damage in what appears to be healthy cerebral white matter may have negative cognitive consequences. To our knowledge, this is the first study to use a myelin specific imaging technique to study the association between myelin content and cognitive performance in older adults with cSVD. Future studies are needed to replicate our current preliminary results.

Using quantitative MRI to assess the role of myelin in relation to cognitive function is a growing area of research interest. In healthy middle-aged and older adults, myelin-probing approaches include using a T₂ time measure [42] and a T_1w/T_2w ratio [43]. A study by Lu et al. [42] reported an association between increased T₂ time in the prefrontal white matter and the genu of the corpus callosum with slow processing speed in healthy older adults. However, there is no quantitative evidence specifically linking T₂ time to myelin content [44]. In another study examining healthy middle-aged and older adults, Chopra et al. [43] reported an association between increased T_1w/T_2w ratio in the anterior limb of the internal capsule and the left splenium of the corpus callosum and slow processing speed. In addtion, the same study reported no significant association between whole brain white matter T_1w/T_2w ratio and processing speed [43]. These results are in contrast to our current study which demonstrates whole brain NAWM MWF is associated with processing speed performance. These divergent findings are likely due to differences in MRI-based myelin measurements, where T_1w/T_2w ratio has limited validity as an index of myelin content in white matter [45, 46].

Our results also suggest an association between myelin content and working memory. Working memory involves widespread cerebral networks, including bilateral frontal-parietal, frontal-temporal, and parietal-temporal tracts [47, 48]. Consequently, working memory performance may be influenced by the strength of the anatomical connectivity between these cerebral networks. Notably, myelin is crucial for sharing and integrating information across these distant brain regions. Our results suggest that a widespread reduction in myelin content may negatively impact working memory.

We did not find a significant association between myelin content and inhibition as measured with the Stroop test or set shifting as measured with the Trail Making Test Part B minus A. These results are in contrast to diffusion tensor and magnetization transfer imaging studies [49 –52]. Decreased white matter integrity, as measured by fractional anisotropy, in frontoparietal and occipital regions have been associated with greater Stroop interference [52] and task switching costs [49, 50]. Moreover, a cross-sectional and longitudinal analysis of the Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) [12] found that reduced whole-brain magnetization transfer ratio, interpreted by the authors as a marker of macromolecular degeneration, was associated with greater Stroop interference. Age related alterations in magnetization transfer ratio also contributes to set shifting costs in older adults [51]. Though diffusion tensor and magnetization transfer imaging data lack tissue specificity, these studies do suggest that damage to white matter microstructure may contribute to reduced set shifting and inhibitory control. Our data suggests that this relationship might not be associated with myelin specific changes and may instead be associated with other microstructural properties of white matter (e.g., axonal loss or alterations in membrane permeability).

Our findings are consistent with studies in other clinical populations assessing the association between myelin, as quantified with MWF, and cognitive function. In people with multiple sclerosis, reduced NAWM MWF was associated with slower processing speed, poorer working memory, and reduced attention [53]. Greater MWF heterogeneity was also associated with slow processing speed [54]. In people with mild cognitive impairment, lower regional MWF was associated with impaired verbal episodic memory [55]. In the same study, lower MWF was also associated with impaired semantic fluency (a component of executive functions) in a combined sample of mild cognitive impairment and nondemented older adults [55]. However, this study did not include participants with signs of cSVD (i.e., WMHs, lacunes, enlarged perivascular spaces, cerebral microbleeds, or cortical superficial siderosis) [55]. Here we provide the first evidence suggesting that myelin deterioration is also detrimental for cognitive function in people with cSVD, and that the effects of myelin are independent of WMHs.

Our preliminary results should be interpreted within the study limitations. First, there are limitations inherent to myelin water imaging such as low signal-to-noise ratio, partial-volume contamination, potential water exchange effects, and the possibility of global increases in water content confounding the measurement. However, as a strength, myelin water imaging is currently considered the most specific measure of white matter myelin [6 , 56]. Second, this is a cross-sectional study and no conclusions can be made about causality. Third, the overall explained variance of NAWM MWF was quite modest, 24.5% and 8.6% for processing speed and working memory, respectively. Though we did control for the effects of WMHs, we note that we did not account for other cSVD pathologies (i.e., lacunes, cerebral microbleeds, perivascular spaces, and atrophy) and it remains unclear whether the effects of myelin would be additive, synergistic, or attenuated when other lesions are included. In addition, different regions of the brain may exhibit different rates of myelin damage and our global measure of MWF may overlook regional contributions of cerebral demyelination on cognitive function. Specifically, myelin breakdown in the brain may begin in anterior regions before progressing to posterior regions [57 –59]. Our results align with the notion of an anterior-to-posterior gradient in myelin damage, as processing speed and working memory rely predominantly on anterior brain regions [50, 60]. Within the current analyses, we chose to focus on global NAWM, instead of a region-specific analyses, because our sample size would not have sufficient power for Type I error adjustments. We also did not perform a sex stratified analysis though sexual dimorphism in myelin content has been observed in several brain regions [57, 58]; however, there were no significant sex differences in myelin content in our cohort (t(53) = –0.42, p = 0.677). Lastly, we note that our results would not survive multiple testing correction and should be interpreted with caution as Type I error is a concern. However, we note that these are exploratory analyses and we did not correct for multiple comparisons because Type II error is a consideration in exploratory studies [38]. Thus, we report all p values and confidence intervals so readers can interpret this study’s results within the context of the number of analyses conducted [39, 40]. Studies are needed to replicate and confirm our reported findings. Future studies with a healthy control group and a longitudinal design will further elucidate the unique pattern of myelin loss in cSVD and how these changes may relate to changes in cognitive function.

Footnotes

ACKNOWLEDGMENTS

Thank you to the study participants and the MRI technologists at the UBC MRI Research Centre. We would like to also thank Kevin Lam for his assistance with the WMH segmentations. Funding for this research was provided by the Heart and Stroke Foundation of Canada (G-15-0009019) and the Alzheimer Society Research Program (15–18). T.L.A. is a Canada Research Chair (Tier 2) in Physical Activity, Mobility, and Cognitive Neuroscience. E.D. is the recipient of the Canadian Institutes of Health Research Doctoral Award.

Authors’ disclosures available online ().

References

Sachdev

, Kalaria

, O’Brien

, Skoog

, Alladi

, Black

, Blacker

, Blazer

, Chen

, Chui

, Ganguli

, Jellinger

, Jeste

, Pasquier

, Paulsen

, Prins

, Rockwood

, Roman

, Scheltens

, International Society for Vascular Behavioral and Cognitive Disorders (2014) Diagnostic criteria for vascular cognitive disorders: A VASCOG statement. Alzheimer Dis Assoc Disord 28, 206–218.

Kloppenborg

, Nederkoorn

, Geerlings

, van den Berg

(2014) Presence and progression of white matter hyperintensities and cognition: A meta-analysis. Neurology 82, 2127–2138.

Gouw

, Seewann

, van der Flier

, Barkhof

, Rozemuller

, Scheltens

, Geurts

(2011) Heterogeneity of small vessel disease: A systematic review of MRI and histopathology correlations. J Neurol Neurosurg Psychiatry 82, 126–135.

Bartzokis

(2011) Alzheimer’s disease as homeostatic responses to age-related myelin breakdown. Neurobiol Aging 32, 1341–1371.

Laule

, Vavasour

, Kolind

, Li

, Traboulsee

, Moore

, MacKay

(2007) Magnetic resonance imaging of myelin. Neurotherapeutics 4, 460–484.

MacKay

, Laule

(2016) Magnetic resonance of myelin water: An in vivo marker for myelin. Brain Plast 2, 71–91.

Beaulieu

(2002) The basis of anisotropic water diffusion in the nervous system - a technical review. NMR Biomed 15, 435–455.

Vavasour

, Laule

, Li

, Traboulsee

, MacKay

(2011) Is the magnetization transfer ratio a marker for myelin in multiple sclerosis? J Magn Reson Imaging 33, 713–718.

van Norden

, de Laat

, van Dijk

, van Uden

, van Oudheusden

, Gons

, Norris

, Zwiers

, de Leeuw

(2012) Diffusion tensor imaging and cognition in cerebral small vessel disease: The RUN DMC study. Biochim Biophys Acta 1822, 401–407.

10.

O’Sullivan

, Morris

, Huckstep

, Jones

, Williams

, Markus

(2004) Diffusion tensor MRI correlates with executive dysfunction in patients with ischaemic leukoaraiosis. J Neurol Neurosurg Psychiatry 75, 441–447.

11.

Shenkin

, Bastin

, Macgillivray

, Deary

, Starr

, Rivers

, Wardlaw

(2005) Cognitive correlates of cerebral white matter lesions and water diffusion tensor parameters in community-dwelling older people. Cerebrovasc Dis 20, 310–318.

12.

Sala

, de Roos

, Blauw

, Middelkoop

, Jukema

, Mooijaart

, van Buchem

, de Craen

, van der Grond

(2015) Association between changes in brain microstructure and cognition in older subjects at increased risk for vascular disease. BMC Neurol 15, 133.

13.

Seiler

, Pirpamer

, Hofer

, Duering

, Jouvent

, Fazekas

, Mangin

, Chabriat

, Dichgans

, Ropele

, Schmidt

(2014) Magnetization transfer ratio relates to cognitive impairment in normal elderly. Front Aging Neurosci 6, 263.

14.

Maniega

, Valdes Hernandez

, Clayden

, Royle

, Murray

, Morris

, Aribisala

, Gow

, Starr

, Bastin

, Deary

, Wardlaw

(2015) White matter hyperintensities and normal-appearing white matter integrity in the aging brain. Neurobiol Aging 36, 909–918.

15.

Laule

, Kozlowski

, Leung

, Li

, Mackay

, Moore

(2008) Myelin water imaging of multiple sclerosis at 7 T: Correlations with histopathology. Neuroimage 40, 1575–1580.

16.

Laule

, Leung

, Lis

, Traboulsee

, Paty

, MacKay

, Moore

(2006) Myelin water imaging in multiple sclerosis: Quantitative correlations with histopathology. Mult Scler 12, 747–753.

17.

Kozlowski

, Raj

, Liu

, Lam

, Yung

, Tetzlaff

(2008) Characterizing white matter damage in rat spinal cord with quantitative MRI and histology. J Neurotrauma 25, 653–676.

18.

Laule

, Yung

, Pavolva

, Bohnet

, Kozlowski

, Hashimoto

, Yip

, Li

, Moore

(2016) High-resolution myelin water imaging in post-mortem multiple sclerosis spinal cord: A case report. Mult Scler 22, 1485–1489.

19.

McCreary

, Bjarnason

, Skihar

, Mitchell

, Yong

, Dunn

(2009) Multiexponential T2 and magnetization transfer MRI of demyelination and remyelination in murine spinal cord. Neuroimage 45, 1173–1182.

20.

Borich

, Mackay

, Vavasour

, Rauscher

, Boyd

(2013) Evaluation of white matter myelin water fraction in chronic stroke. Neuroimage Clin 2, 569–580.

21.

Lakhani

, Borich

, Jackson

, Wadden

, Peters

, Villamayor

, MacKay

, Vavasour

, Rauscher

, Boyd

(2016) Motor skill acquisition promotes human brain myelin plasticity. Neural Plast 2016, 7526135.

22.

Lakhani

, Hayward

, Boyd

(2017) Hemispheric asymmetry in myelin after stroke is related to motor impairment and function. Neuroimage Clin 14, 344–353.

23.

Alonso-Ortiz

, Levesque

, Pike

(2015) MRI-based myelin water imaging: A technical review. Magn Reson Med 73, 70–81.

24.

Schmidt

, Schmidt

, Haybaeck

, Loitfelder

, Weis

, Cavalieri

, Seiler

, Enzinger

, Ropele

, Erkinjuntti

(2011) Heterogeneity in age-related white matter changes. Acta Neuropathol 122, 171–185.

25.

Wardlaw

, Valdes Hernandez

, Munoz-Maniega

(2015) What are white matter hyperintensities made of? Relevance to vascular cognitive impairment. J Am Heart Assoc 4, 001140.

26.

Wardlaw

, Smith

, Biessels

, Cordonnier

, Fazekas

, Frayne

, Lindley

, O’Brien

, Barkhof

, Benavente

, Black

, Brayne

, Breteler

, Chabriat

, Decarli

, de Leeuw

, Doubal

, Duering

, Fox

, Greenberg

, Hachinski

, Kilimann

, Mok

, Oostenbrugge

, Pantoni

, Speck

, Stephan

, Teipel

, Viswanathan

, Werring

, Chen

, Smith

, van Buchem

, Norrving

, Gorelick

, Dichgans

(2013) Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration.. Lancet Neurol 12, 822–838.

27.

Fazekas

, Chawluk

, Alavi

, Hurtig

, Zimmerman

(1987) MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging.. AJR Am J Roentgenol 149, 351–356.

28.

Nasreddine

, Phillips

, Bedirian

, Charbonneau

, Whitehead

, Collin

, Cummings

, Chertkow

(2005) The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J Am Geriatr Soc 53, 695–699.

29.

Folstein

, Folstein

, McHugh

(1975) Mini-Mental State: A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

30.

Groll

, To

, Bombardier

, Wright

(2005) The development of a comorbidity index with physical function as the outcome. J Clin Epidemiol 58, 595–602.

31.

Strauss

, Sherman

, Spreen

(2006) A compendium of neuropsychological tests: Administration, norms, and commentary, American Chemical Society.

32.

Wechsler

(2008) Wechsler adult intelligence scale–Fourth Edition (WAIS–IV). NCS Pearson, San Antonio, TX.

33.

Prasloski

, Rauscher

, MacKay

, Hodgson

, Vavasour

, Laule

, Madler

(2012) Rapid whole cerebrum myelin water imaging using a 3D GRASE sequence. Neuroimage 63, 533–539.

34.

McAusland

, Tam

, Wong

, Riddehough

, Li

(2010) Optimizing the use of radiologist seed points for improved multiple sclerosis lesion segmentation. IEEE Trans Biomed Eng 57, doi: 10.1109/TBME.2010.2055865.

35.

Parzen

(1962) On estimation of a probability density function and mode. Ann Math Statist 33, 1065–1076.

36.

Yoo

, Tam

(2013) Non-local spatial regularization of MRI T2 relaxation images for myelin water quantification. Med Image Comput Comput Assist Interv 16, 614–621.

37.

Zhang

, Brady

, Smith

(2001) Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging 20, 45–57.

38.

Schoenfeld

(1980) Statistical considerations for pilot studies. Int J Radiat Oncol Biol Phys 6, 371–374.

39.

Rothman

(1990) No adjustments are needed for multiple comparisons. Epidemiology 1, 43–46.

40.

Saville

(1990) Multiple comparison procedures: The practical solution. Am Stat 44, 174–180.

41.

Zhang

, Wong

, van de Haar

, Staals

, Jansen

, Jeukens

, Hofman

, van Oostenbrugge

, Backes

(2017) Blood-brain barrier leakage is more widespread in patients with cerebral small vessel disease. Neurology 88, 426–432.

42.

, Lee

, Raven

, Tingus

, Khoo

, Thompson

, Bartzokis

(2011) Age-related slowing in cognitive processing speed is associated with myelin integrity in a very healthy elderly sample. J Clin Exp Neuropsychol 33, 1059–1068.

43.

Chopra

, Shaw

, Sachdev

, Anstey

, Cherbuin

(2018) More highly myelinated white matter tracts are associated with faster processing speed in healthy adults. Neuroimage 171, 332–340.

44.

Whittall

, MacKay

, Graeb

, Nugent

, Li

, Paty

(1997) In vivo measurement of T2 distributions and water contents in normal human brain. Magn Reson Med 37, 34–43.

45.

Arshad

, Stanley

, Raz

(2017) Test-retest reliability and concurrent validity of in vivo myelin content indices: Myelin water fraction and calibrated T1 w/T2 w image ratio. Hum Brain Mapp 38, 1780–1790.

46.

Uddin

, Figley

, Solar

, Shatil

, Figley

(2019) Comparisons between multi-component myelin water fraction, T1w/T2w ratio, and diffusion tensor imaging measures in healthy human brain structures. Sci Rep 9, 2500.

47.

Charlton

, Barrick

, Lawes

, Markus

, Morris

(2010) White matter pathways associated with working memory in normal aging. Cortex 46, 474–489.

48.

Takeuchi

, Taki

, Sassa

, Hashizume

, Sekiguchi

, Fukushima

, Kawashima

(2011) Verbal working memory performance correlates with regional white matter structures in the frontoparietal regions. Neuropsychologia 49, 3466–3473.

49.

Gold

, Powell

, Xuan

, Jicha

, Smith

(2010) Age-related slowing of task switching is associated with decreased integrity of frontoparietal white matter. Neurobiol Aging 31, 512–522.

50.

Kennedy

, Raz

(2009) Aging white matter and cognition: Differential effects of regional variations in diffusion properties on memory, executive functions, and speed. Neuropsychologia 47, 916–927.

51.

Serbruyns

, Leunissen

, van Ruitenbeek

, Pauwels

, Caeyenberghs

, Solesio-Jofre

, Geurts

, Cuypers

, Meesen

, Sunaert

(2016) Alterations in brain white matter contributing to age-related slowing of task switching performance: The role of radial diffusivity and magnetization transfer ratio. Hum Brain Mapp 37, 4084–4098.

52.

Wolf

, Zschutschke

, Scheurich

, Schmitz

, Lieb

, Tuscher

, Fellgiebel

(2014) Age-related increases in Stroop interference: Delineation of general slowing based on behavioral and white matter analyses. Hum Brain Mapp 35, 2448–2458.

53.

Baumeister

, Lin

, Vavasour

, Kolind

, Kosaka

, Li

DKB

, Traboulsee

, MacKay

, McKeown

(2019) Data fusion detects consistent relations between non-lesional white matter myelin, executive function, and clinical characteristics in multiple sclerosis. Neuroimage Clin 24, 101926.

54.

Abel

, Vavasour

, Lee

, Johnson

, Ackermans

, Chan

, Dvorak

, Schabas

, Wiggermann

, Tam

, Kuan

, Morrow

, Wilken

, Laule

, Rauscher

, Bhan

, Sayao

, Devonshire

, Li

, Carruthers

, Traboulsee

, Kolind

(2020) Myelin damage in normal appearing white matter contributes to impaired cognitive processing speed in multiple sclerosis. J Neuroimaging 30, 205–211.

55.

Kavroulakis

, Simos

, Kalaitzakis

, Maris

, Karageorgou

, Zaganas

, Panagiotakis

, Basta

, Vgontzas

, Papadaki

(2018) Myelin content changes in probable Alzheimer’s disease and mild cognitive impairment: Associations with age and severity of neuropsychiatric impairment. J Magn Reson Imaging 47, 1359–1372.

56.

Lee

, Hyun

, Lee

, Choi

, Shin

, Min

, Nam

, Kim

, Oh

(2021) So you want to image myelin using MRI: An overview and practical guide for myelin water imaging. J Magn Reson Imaging 53, 360–373.

57.

Arshad

, Stanley

, Raz

(2016) Adult age differences in subcortical myelin content are consistent with protracted myelination and unrelated to diffusion tensor imaging indices. Neuroimage 143, 26–39.

58.

Bouhrara

, Rejimon

, Cortina

, Khattar

, Bergeron

, Ferrucci

, Resnick

, Spencer

(2020) Adult brain aging investigated using BMC-mcDESPOT-based myelin water fraction imaging. Neurobiol Aging 85, 131–139.

59.

Papadaki

, Kavroulakis

, Kalaitzakis

, Karageorgou

, Makrakis

, Maris

, Simos

(2019) Age-related deep white matter changes in myelin and water content: A T2 relaxometry study. J Magn Reson Imaging 50, 1393–1404.

60.

Collette

, Van der Linden

, Laureys

, Delfiore

, Degueldre

, Luxen

, Salmon

(2005) Exploring the unity and diversity of the neural substrates of executive functioning. Hum Brain Mapp 25, 409–423.