Exploring the relationship between cognitive performance and magnetic resonance imaging diffusion parameters in elderly individuals

ABSTRACT

Objective  This study evaluated the association between cognitive function and quantitative magnetic resonance imaging analyses, including brain volume, white matter hyperintensity volume, and diffusivity metrics.

Methods  This retrospective analysis included 504 older adults from São Paulo, Brazil, who underwent 3T magnetic resonance imaging scans. Image analysis was performed using the FMRIB Software Library (FSL), with peak width of mean diffusivity assessed via a public script. FLAIR signal changes were quantified using the Lesion Segmentation Tool and Fazekas scale. Cognitive performance was assessed using MMSE and 3MS tests. Multiple linear regression, adjusting for control variables, was used to evaluate the relationships between magnetic resonance imaging measurements and cognitive scores, validated against a UK Biobank sample.

Results  Magnetic resonance imaging demonstrated strong correlations with UK Biobank dataset. Fractional anisotropy, mean diffusivity, and peak width of the mean diffusivity values were significantly associated with white matter hyperintensities (Spearman’s rho: -0.630, 0.750, and 0.747, p<0.001). Specific brain regions demonstrated strong links between fractional anisotropy and mean diffusivity values and cognitive performance. Fractional anisotropy findings correlated positively with neuropsychological scores (r=0.315 for 3MS and r=0.285 for MMSE, p<0.001).

Conclusion  Diffusivity metrics, including fractional anisotropy, mean diffusivity, and peak width of the mean diffusivity significantly correlated with brain volume, white matter hyperintensities, and cognitive scores. These findings may serve as potential imaging markers for monitoring cognitive decline and dementia.

Keywords
Diffusion tensor imaging; Cognition; Cognitive dysfunction; Dementia; Neuroimaging; Neuropsychological tests; Aged; Magnetic resonance imaging

Highlights

■First study to investigate associations between cognitive performance and brain magnetic resonance imaging parameters in an older Brazilian population.

■Uses advanced magnetic resonance imaging diffusion tensor imaging and automated analysis to quantify brain changes.

■Validates findings through comparison with data from a large UK-based cohort.

■Focuses on region-specific assessments, such as the superior longitudinal fasciculus and other key brain areas related to cognition.

■Demonstrates significant correlation between 3MS and MMSE cognitive scores and diffusion metrics, highlighting the role of white matter integrity in cognitive aging.

In Brief

This study investigates the relationship between cognitive performance and brain magnetic resonance imaging diffusion parameters in 504 older adults from São Paulo, Brazil. Using advanced magnetic resonance imaging techniques and cognitive assessments (MMSE and 3MS), the study found significant correlations between diffusivity metrics (fractional anisotropy, mean diffusivity, and peak width of skeletonized mean diffusivity) and brain volume, white matter hyperintensities, and cognitive scores. Findings were validated by comparing them with a UK Biobank cohort. The results suggest that magnetic resonance imaging metrics may be useful imaging markers for monitoring cognitive decline and dementia.

INTRODUCTION

Despite medical advances and numerous efforts to mitigate cognitive decline, aging remains a universal process affecting all individuals.(¹)Cognitive decline complexity in the older population results from the interplay of various aging-related factors, including cellular senescence, epigenetic modifications, and metabolic dysfunction. Brain aging phenotypes, such as vascular dysfunction, inflammation, and lipid dysregulation, further interact with central nervous system processes, leading to morphological, cognitive, and neuropathological changes, ultimately contributing to neurodegenerative diseases.(²)

Dementia, characterized by severe cognitive decline that interferes with daily activities, is most commonly associated with Alzheimer’s disease as the leading cause,(³) followed by cerebrovascular disease as the second most common cause.(⁴)

Particularly, neuroimaging studies on dementia are scarce in low and middle-income countries. This is relevant considering countries such as Brazil have the highest dementia burden and potential for reduced prevalence through addressing seven modifiable risk factors (low education, physical inactivity, midlife hypertension, midlife obesity, depression, smoking, and diabetes mellitus).(⁵) Despite rising prevalence owing to increased life expectancy, public health investment services remain insufficient.(⁶)

Cognitive tests, particularly the widely validated Mini Mental State Examination (MMSE), are essential for assessing cognitive function, particularly in the older population, both internationally and in Brazil.(⁷)

Magnetic resonance imaging (MRI) serves as a valuable in vivo marker for nervous system characterization. High-resolution MRI allows for comparative intervention analyses, risk factor control, and central nervous system impact evaluation.(⁸) However, access to more advanced MRI modalities is challenging in low-middle income countries. When available, advanced MR techniques and analysis methods enables assessment of tissue microstructure organization in these populations. Among these techniques, diffusion tensor imaging (DTI) analyzes water molecule mobility along white matter tracts, revealing white matter molecular properties under physiologic and altered conditions.(⁹) One of its parameters, fractional anisotropy (FA) is sensitive to microstructural changes, with experimental tests demonstrating a strong correlation with axon number and constitution.(¹⁰)

Fractional anisotropy and mean diffusivity (MD) are DTI parameters that demonstrate good intra-scanner reproducibility, and our sample presents ideal data on which all tests were performed with the same equipment and protocol.(¹¹) Peak width of skeletonized mean diffusivity (PSMD) is a newer automated DTI marker quantifying damage related to small vessel disease; however, possible inconsistent results may be found and special care must be taken with quality assessment for motion and distortion artifacts.(¹²,¹³)

OBJECTIVE

This study aims to explore the association between cognition and quantitative magnetic resonance imaging analysis, including brain volume, white matter hyperintensity volume, and diffusivity metrics. To ensure external validation and reproducibility, differences among diffusion metrics were compared in two different diffusion magnetic resonance imaging acquisitions in a distinct population, considering the extraordinary sample of older individuals in large population-based studies (UK Biobank and OCTAGENE).

METHODS

Our data is part of the SABE multicenter study coordinated by the Pan American Health Organization developed in urban centers to profile living and health conditions of older adults. In Brazil, the study was conducted in São Paulo with a probabilistic cohort of individuals >60 years old. A comprehensive questionnaire, including health status, housing conditions, work history, and income were reported, alongside functional tests and blood collections for biochemical, immunological, and genetic variables. These were reported by Naslavsky et. al., and detailed in previous methodologic, genetic, and brain matter volumetric evaluation studies.(¹⁴-¹⁶)

As part of SABE, we analyzed the OCTAGENE study, including a retrospective sample of this population who underwent brain magnetic resonance imaging in 3T equipment (TIM TRIO; Siemens Healthcare, Erlangen, Germany). The imaging followed the same study protocol: T1-weighted volumetric magnetization-prepared rapid-acquisition gradient echo imaging with 1 mm isotropic voxels, 2500ms repetition time, 3.45ms echo time, 1100ms inversion time, and a 7° flip angle. FLAIR isotropic and DTI with 30 directions were also required. Four trained neuroradiologists analyzed images to exclude abnormalities and movement artifacts.(¹⁵)

UK Biobank MRI acquisition was performed using similar study protocols,including diffusion MRI and volumetric isotropic T1 MPRAGE and T2 Flair, as detailed in table 1. A standardized six-week training program and routine phantom measurements using UK Biobank MRI scanners ensured harmonized imaging data. Acquisition parameters are displayed table 1.

Table 1
MRI acquisition parameters of T1, T2, and diffusion images according to duration, resolution matrix, and other parameters in OCTAGENE and the UK Biobank protocol

Modality	T1 MPRAGE	T2 FLAIR	Diffusion MRI
Duration, UK Biobank (mins)	4:54	5:52	7:08
Duration, OCTAGENE (mins)	4:23	4:40	3:22
Resolution, UK Biobank	1.0 × 1.0 × 1.0	1.0 × 1.0 × 1.05	2.0 × 2.0 × 2.0
Resolution, OCTAGENE	1.0 × 1.0 × 1.0	0.5 × 0.5 × 1.0	2.2 × 2.2 × 2.2
Matrix, UK Biobank	256 × 256 × 208	256 × 256 × 192	104 × 104 ×72
Matrix- OCTAGENE	256 × 256 × 138	256 × 256 × 160	280 × 280 × 128
Other parameters, UK Biobank	TE/TR=2.01/2000ms, TI=880ms	TE/TR=395/5000ms, TI=1800ms	TR=3600ms, 50 directions/shell, b=0,1000, 2000 s/mm2, 51 deg, MB=3
Other parameters, OCTAGENE	TE/TR=3.48/2500ms, TI=1000ms	TE/TR=280/5000ms, TI=1800ms	TR=5900, 30 directions/shell, b=0,1000 s/mm2, 0 deg., MB=2

• MRI: magnetic resonance imaging.:

Images were evaluated by a certified radiologist with specific training in neuroradiology. Substantial structural abnormalities were considered and qualitative evaluation was performed using the periventricular white matter Fazekas scale(¹⁷) (score 0, I, II or III). Qualitative data were supported using automated quantification processing tools, including the lesion segmentation tool (LST) toolbox (LGA, Lesion Growth Algorithm)(¹⁸) and FreeSurfer software (version 5.3.0; Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, USA). Processing was performed on a high-performance computer to build models of the cortical surface Results were verified via quality control assessment.(¹⁹)

White matter hyperintensities were categorized via periventricular region more related to vascular disease. Classification levels were 0, I, II, and III (0 for single punctate white matter hyperintensities, I for multiple punctate white matter, II for some lesion confluency, and III for large confluent lesions).(¹⁷)

Structural image evaluation was performed and statistically controlled in correlation with demographic data (age, years of education, and sex), diffusion metrics (FA, MD, and PSMD), and cognitive tests (MMSE and 3MS).(²⁰-²²)

The images were initially evaluated for the technical acquisition quality, excluding cases with incomplete images, unavailable DTI images, and artifacts that could impede analysis.

Diffusivity parameters were considered individually and group analysis was performed with the FMRIB Software library, including validated protocols. The Brain Extraction Tool was used for extracting brain volumes. The “eddy” function from FDT (Diffusion Toolbox) corrected distortions, movement artifacts, and Eddy currents. The DTIFIT command provided FA and MD values. Tract based spatial statistics (TBSS) was used for group assessment.(²²) Tract based spatial statistics pipeline was applied to skeletonized data as reviewed by Baykara et al.(²³)

Images were processed and analyzed with the FreeSurfer software on a high-performance computer to construct models of the cortical surface.

Statistical analyses

Analyses were performed by a statistician experienced in health data using the IBM SPSS Statistics software package for Windows, Version 20.0, Armonk, NY: IMB Corp., 2011. Bonferroni correction mitigated multiple variable corrections.

Spearman’s rank correlation quantified relationships, where coefficients (rho) of 0.1, 0.3, and 0.5 indicate small, medium, and large effects, respectively.(²⁴) Spearman’s correlation was applied to assess associations between white matter lesion volume and count, FA, MD, cognitive scores (3MS and MMSE), Fazekas scale scores, and age.

Multiple linear correlation regarded MMSE and 3MS as dependent variables in the model, considering sex, age, and years of education in each of the regions of interest (white matter tracts).

Cluster (groups of contiguous voxels that exhibit similar properties) comparisons of white matter hyperintensity and volume observed in FLAIR images between sexes in relation to FA, MD, and PSMD were performed using the Mann-Whitney test.

A comparison of diffusion metrics (FA, MD, and PSMD) was performed based on outcomes, categorizing scores into MMSE ≤17, severe cognitive impairment(²⁵) and MMSE ≥23, and mild cognitive impairment.(²⁶) Receiver Operating Characteristic (ROC) curves and areas under the curve (AUC) with confidence intervals >95% were employed for this assessment.

A significance threshold of 5% was adhered to for all statistical inferences and tests.

RESULTS

In total, 529 patients underwent MRI, and 504 were selected for analysis. Twenty-five patients (4%) were excluded owing to DTI sequence unavailability or poor image quality.

The studied population had a mean age of 74±9 years, with 326 females (64.6%), an MMSE average score of 25.4, and a standard deviation of ±4.5, out of a total of 30 points (Table 2).

Table 2
OCTAGENE demographic data

	n (%)
Age, (Mean)	504 74±9 years
Sex
Female	326 (64.6)
Male	178 (35.3)
Cognitive test, (Mean)
MMSE	25±4.5
3MS	81.7±14.4
Brain volume	956±109 mL
Intracranial volume	1342±226 mL

The 3MS had a mean of 81.7 and standard deviation of 14.4, out of a total of 100 points.

The mean brain volume was 956 mL, with a median of 956 mL, and a standard deviation of 109 mL. The mean intracranial volume in the studied population was 1,342 mL, with standard deviation of 226 mL. The ratio between brain volume and intracranial volume had a mean of 0.72, and a standard deviation of 0.11.

Strong correlations were observed when evaluating lesion volumes and diffusion measures, highlighting volumetric measures and DTI sensitivity (DTI metrics in characterizing white matter abnormalities). The observed correlation coefficients ranged 0.514-0.750 when evaluating lesion volumes alongside diffusion metrics, indicating a robust relationship between these factors (Figure 1).

Figure 1
Relationship between the volume of white matter hyperintensities on FLAIR and the overall values of fractional anisotropy, mean diffusivity, and peak width of skeletonized mean diffusivity

The global fractional anisotropy values for study participants had a mean of 0.42, with a median of 0.43 and standard deviation of 0.021.

Mean diffusivity analysis among participants had an average value of 0.916, with a median of 0.907, and standard deviation of 0.068.

The global PSMD variable had a median value of 0.30 × 10-⁴ mm²/s among study participants, with a standard deviation of 0.068.

An exploratory analysis was conducted on all tracts from the JHU White-Matter Tractography Atlas, considering region of interest (ROI) analysis standardized using the FLS software. A multivariate analysis of these regions of interest was performed, considering sex, age, and years of education. The superior longitudinal fasciculus, inferior fronto-occipital fasciculus, and the cingulum gyrus showed the strongest associations with MMSE and 3MS scores (Tables 3 and 4).

A comparative analysis of diffusion metrics (FA, MD, and PSMD) were performed for MMSE ≤17 (MMSE17) and MMSE ≤23 (MMSE23). ROC curvers showed an AUC >0.769 for MMSE ≤17 and AUC >0.625 for the MMSE23 cutoff (Figures 2 and 3).

In a complementary analysis, data from individuals in the UK Biobank were subsampled to evaluate possible variations related to age group differences. The original sample included individuals with a minimum age of 45 years and a maximum of 75. To adjust the samples, the minimum age of 60 years was used as the cutoff (older population), and the maximum age was equal to that of the UK Biobank study (75 years) (Table 5).

Table 5
Distribution of mean ages after adjustment for the minimum age of 60 and maximum of 75 in the OCTAGENE and UK Biobank studies

Group	Total	Mean age	N [60-75 years old]	Mean age 60-75 years old sample	p value
OCTAGENE	504	74.7	282	67.6	<0.001
UK Biobank	40.521	62.9	21.932	65.7	<0.001
Total	41.025		22.214	63.3

• Created with the BioRender tool.. Available from: https://biorender.com
• Student t test.
• Demographic data: OCTAGENE and UKBiobank.
• Created with the BioRender tool.. Available from: https://biorender.com
• FA: fractional anisotropy.

Mean FA and MD curves across tracts showed better alignment between OCTAGENE and UK Biobank in the 60-75 years subsample (Figure 4).

Figure 4
Distribution of mean fractional anisotropy and mean diffusivity values across different white matter tracts in the OCTAGENE and UK Biobank studies for complete sample and subsample individuals aged 60-75 years

The proportional difference formula: $(OCT-UK)/\left[\right(OCT+UK\left)/2\right]\times 100$ was used to compare mean FA values. The average difference was 0.00±0.11, the maximum (-29.46%) was in the left cingulate gyrus, and the smallest (-0.01%) was in the right hippocampus (Table 6).

Table 6
Mean FA values according to white matter tracts in OCTAGENE and UK Biobank samples and proportional difference formula $(OCT-UK)/\left[\right(OCT+UK\left)/2\right]\times 100$

DISCUSSION

The traditional Fazekas scale revealed weaker correlations, with Spearman’s rho coefficients ranging 0.417-0.414. This suggests that visual assessment methods may have inherent limitations compared to quantitative techniques, supporting the findings of van Straaaten et al.(²⁷)who reported that, in a review of 618 independently living older adults, volumetric measures of white matter hyperintensities demonstrated greater sensitivity in detecting memory-related symptoms. This highlights the added value of objective metrics for evaluating cognitive health, particularly in aging populations.

Additionally, our results indicated a negative correlation between global FA and lesion burden, based on total lesion count and volume. This aligns with Andersen et al.’s findings that decreased FA values correlate with higher white matter lesion burden in patients with multiple sclerosis. Such associations imply that lower FA may reflect structural damage and serve as an important biomarker for ongoing neurodegeneration, contributing to cognitive decline.(²⁸)

Furthermore, significant correlation was observed between the 3MS and MMSE cognitive assessments and diffusion metrics, with the FA showing positive correlations (Spearman’s rho coefficients of 0.315 for 3MS and 0.285 for MMSE, p<0.001) and MD values displaying a negative correlation (Spearman’s rho coefficients of -0.237 for 3MS and -0.258 for MMSE). These findings align with those of Li et al.(²⁹) who reported similar associations in a community-based cohort conducted on a rural population in China. Li et al. also observed global associations between FA and MD values and cognitive performance in the verbal fluency test. The consistent relationship between FA, MD, and various cognitive measures supports the role of white matter integrity in age-related cognitive decline.

This study showed that higher global FA scores correlate positively with higher 3MS and MMSE scores. This result aligns with previous research, including the findings of Xing et al., who, in an analysis of 77 participants with Fazekas 2 or 3, stratified by age, sex, education, and the presence of apolipoprotein E gene 4, identified mean global FA as the strongest risk marker for cognitive decline that exhibited statistical correlation with MMSE scores.(³⁰)

Technological advances in the field now allow quicker image acquisition for DTI analysis with fewer artefacts, making clinical use of such techniques more feasible. Although some MRI metrics still require offline image processing steps, standardized strategies can help mitigate practical limitations.

Notably, this study focuses on region-specific assessment, such as the superior longitudinal fasciculus - the brain’s largest associative fiber bundle.(³¹) Similarly, Koshiyama et al.(³²) found that FA values in the superior longitudinal fasciculus strongly correlated with memory and visuospatial ability in 583 healthy volunteers aged 18-68 (mean age of 30.8) years.

Finally, our findings on associations between cognitive test alterations and specific regions, including the inferior fronto-occipital fasciculus, cingulate gyrus, posterior thalamic radiation, and fornix, align with Deng et al.(³³) Their study found that elevated heart rates were linked to diminished performance in fluid intelligence tasks, suggesting phycological factors may interact with the structural integrity of white matter, further complicating our understanding of cognitive decline.

CONCLUSION

A substantial correlation was observed between white matter hyperintensities, cognitive performance, and diffusion metrics (fractional anisotropy, mean diffusivity, and peak width of skeletonized mean diffusivity). Although Magnetic resonance imaging metric assessment requires offline image processing, standardized strategies can help mitigate practical limitations. This comprehensive neuroimaging study provides valuable insights into the complex interplay among brain structure, cognitive decline, and aging. The findings enhance neuroradiology by highlighting potential biomarkers of cognitive aging and emphasizing the importance of region-specific analyses and quantitative imaging techniques.

Figure 2
ROC curve for the outcome of MMSE ≤17

Figure 3
ROC Curve for the outcome of MMSE ≤23

Table 3
Interaction intensity values from the multivariate analysis of fractional anisotropy in white matter tracts corrected for sex, age, and education, distributed by regions of interest in the OCTAGENE study

Table 4
Interaction intensity values from the multivariate analysis of mean diffusivity in white matter tracts corrected for sex, age, and education, distributed by regions of interest in the OCTAGENE study

REFERENCES

• ¹ Morley JE. An Overview of Cognitive Impairment. Clin Geriatr Med. 2018;34(4):505-13.
• ² Gonzales MM, Garbarino VR, Pollet E, Palavicini JP, Kellogg DL Jr, Kraig E, et al. Biological aging processes underlying cognitive decline and neurodegenerative disease. J Clin Invest. 2022;132(10):e158453.
• ³ Kumar A, Sidhu J, Goyal A, Tsao JW. Alzheimer Disease. StatPearls. Treasure Island (FL): StatPearls Publishing Copyright © 2022.
• ⁴ GBD 2016 Causes of Death Collaborators. Global, regional, and national age-sex specific mortality for 264 causes of death, 1980-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet. 2017;390(10100):1151-210. Erratum in: Lancet. 2017;390(10106):e38.
• ⁵ Oliveira D, Jun Otuyama L, Mabunda D, Mandlate F, Gonçalves-Pereira M, Xavier m, et al. reducing the Number of People with Dementia Through Primary Prevention in Mozambique, Brazil, and Portugal: an analysis of population-based data. J Alzheimers Dis. 2019;70(s1):S283-S291.
• ⁶ Nitrini R, Ferri CP. Burden of dementia in Brazil. Arq Neuropsiquiatr. 2020;78(12):755-6.
• ⁷ Gallegos M, Morgan ML, Cervigni M, Martino P, Murray J, Calandra M, et al. 45 Years of the mini-mental state examination (MMSE): A perspective from ibero-america. Dement Neuropsychol. 2022;16(4):384-7.
• ⁸ Terribilli D, Schaufelberger MS, Duran FL, Zanetti MV, Curiati PK, Menezes PR, et al. Age-related gray matter volume changes in the brain during non-elderly adulthood. Neurobiol Aging. 2011;32(2):354-68.
• ⁹ Maj E, Jamrozy M, Bielecki M, Bartoszek M, Golebiowski M, Wojtaszek M, et al. Role of DTI-MRI parameters in diagnosis of ALS: useful biomarkers for daily practice? Tertiary centre experience and literature review. Neurol Neurochir Pol. 2022;56(6):490-8. Review.
• ¹⁰ Ranzenberger LR, Snyder T. Diffusion Tensor Imaging. StatPearls. Treasure Island (FL): StatPearls Publishing. Copyright © 2022, StatPearls Publishing LLC; 2022.
• ¹¹ Grech-Sollars M, Hales PW, Miyazaki K, Raschke F, Rodriguez D, Wilson M, et al. Multi-centre reproducibility of diffusion MRI parameters for clinical sequences in the brain. NMR Biomed. 2015;28(4):468-85.
• ¹² Zanon Zotin MC, Yilmaz P, Sveikata L, Schoemaker D, van Veluw SJ, Etherton MR, et al. Peak width of skeletonized mean diffusivity: a neuroimaging marker for white matter injury. Radiology. 2023;306(3):e212780.
• ¹³ Campaña Perilla LA, Mahecha Carvajal ME, Cardona Ortegón JD, Barragán Corrales C, Barrera Patiño AM. Diffusion tensor imaging protocol: the need for standardization of measures. Radiology. 2023;308(2):e230470.
• ¹⁴ Naslavsky MS, Scliar MO, Yamamoto GL, Wang JY, Zverinova S, Karp T, et al. Whole-genome sequencing of 1,171 elderly admixed individuals from São Paulo, Brazil. Nat Commun. 2022;13(1):1004.
• ¹⁵ Lebrão ML, Duarte YA, Santos JL, Silva NN. 10 Years of SABE Study: background, methodology and organization of the study. Rev Bras Epidemiol. 2019;21(Suppl 2):e180002.
• ¹⁶ Rodrigues MA, Rodrigues TP, Zatz M, Lebrão ML, Duarte YA, Naslavsky MS, et al. Quantitative evaluation of brain volume among elderly individuals in São Paulo, Brazil: a population-based study. Radiol Bras. 2019;52(5):293-8.
• ¹⁷ Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer's dementia and normal aging. AJR Am J Roentgenol. 1987;149(2):351-6.
• ¹⁸ Ribaldi F, Altomare D, Jovicich J, Ferrari C, Picco A, Pizzini FB, et al. Accuracy and reproducibility of automated white matter hyperintensities segmentation with lesion segmentation tool: A European multi-site 3T study. Magn Reson Imaging. 2021;76:108-15.
• ¹⁹ Fischl B. FreeSurfer. Neuroimage. 2012;62(2):774-81.
• ²⁰ Arevalo-Rodriguez I, Smailagic N, Roqué-Figuls M, Ciapponi A, Sanchez-Perez E, Giannakou A, et al. Mini-Mental State Examination (MMSE) for the early detection of dementia in people with mild cognitive impairment (MCI). Cochrane Database Syst Rev. 2021;7(7):CD010783.
• ²¹ Tombaugh TN. Test-retest reliable coefficients and 5-year change scores for the MMSE and 3MS. Arch Clin Neuropsychol. 2005;20(4):485-503.
• ²² Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage. 2012;62(2):782-90. Review.
• ²³ Baykara E, Gesierich B, Adam R, Tuladhar AM, Biesbroek JM, Koek HL, Ropele S, Jouvent E; Alzheimer's Disease Neuroimaging Initiative; Chabriat H, Ertl-Wagner B, Ewers M, Schmidt R, de Leeuw FE, Biessels GJ, Dichgans M, Duering M. A novel imaging marker for small vessel disease based on skeletonization of white matter tracts and diffusion histograms. Ann Neurol. 2016;80(4):581-92.
• ²⁴ Cohen J. A power primer. Psychol Bull. 1992;112(1):155-9.
• ²⁵ Kochhann R, Varela JS, Lisboa CS, Chaves ML. The Mini Mental State Examination: review of cutoff points adjusted for schooling in a large Southern Brazilian sample. Dement Neuropsychol. 2010;4(1):35-41.
• ²⁶ Monroe T, Carter M. Using the Folstein Mini Mental State Exam (MMSE) to explore methodological issues in cognitive aging research [editorial]. Eur J Ageing. 2012;9(3):265-74.
• ²⁷ van Straaten EC, Fazekas F, Rostrup E, Scheltens P, Schmidt R, Pantoni L, Inzitari D, Waldemar G, Erkinjuntti T, Mäntylä R, Wahlund LO, Barkhof F; LADIS Group. Impact of white matter hyperintensities scoring method on correlations with clinical data: the LADIS study. Stroke. 2006;37(3):836-40.
• ²⁸ Andersen KW, Lasic S, Lundell H, Nilsson M, Topgaard D, Sellebjerg F, et al. Disentangling white-matter damage from physiological fibre orientation dispersion in multiple sclerosis. Brain Commun. 2020;2(2):fcaa077.
• ²⁹ Li WX, Yuan J, Han F, Zhou LX, Ni J, Yao M, et al. White matter and gray matter changes related to cognition in community populations. Front Aging Neurosci. 2023;15:1065245.
• ³⁰ Xing Y, Yang J, Zhou A, Wang F, Wei C, Tang Y, et al. White matter fractional anisotropy is a superior predictor for cognitive impairment than brain volumes in older adults with confluent white matter hyperintensities. Front Psychiatry. 2021;12:633811.
• ³¹ Janelle F, Iorio-Morin C, D'amour S, Fortin D. Superior longitudinal fasciculus: a review of the anatomical descriptions with functional correlates. Front Neurol. 2022;13:794618.
• ³² Koshiyama D, Fukunaga M, Okada N, Morita K, Nemoto K, Yamashita F, et al. Association between the superior longitudinal fasciculus and perceptual organization and working memory: A diffusion tensor imaging study. Neurosci Lett. 2020;738:135349.
• ³³ Deng YT, Kuo K, Wu BS, Ou YN, Yang L, Zhang YR, et al. Associations of resting heart rate with incident dementia, cognition, and brain structure: a prospective cohort study of UKbiobank. Alzheimers Res Ther. 2022;14(1):147.

Edited by

• Associate Editor:
  Maysa Seabra Cendoroglo Universidade Federal de São Paulo, São Paulo, SP, Brazil ORCID: https://orcid.org/0000-0003-2548-2619

Publication Dates

• Publication in this collection
  17 Oct 2025
• Date of issue
  2025

History

• Received
  17 Oct 2024
• Accepted
  13 May 2025