Accepted Manuscript Aging in Deep Gray Matter and White Matter Revealed by Diffusional Kurtosis Imaging Nan-Jie Gong, Chun-Sing Wong, Chun-Chung Chan, Lam-Ming Leung, Yiu-Ching Chu PII:

S0197-4580(14)00265-6

DOI:

10.1016/j.neurobiolaging.2014.03.011

Reference:

NBA 8813

To appear in:

Neurobiology of Aging

Received Date: 7 August 2013 Revised Date:

8 March 2014

Accepted Date: 13 March 2014

Please cite this article as: Gong, N.-J., Wong, C.-S., Chan, C.-C., Leung, L.-M., Chu, Y.-C., Aging in Deep Gray Matter and White Matter Revealed by Diffusional Kurtosis Imaging, Neurobiology of Aging (2014), doi: 10.1016/j.neurobiolaging.2014.03.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Aging in Deep Gray Matter and White Matter Revealed by Diffusional Kurtosis Imaging

SC

RI PT

(Surnames are underlined) First author: Nan-Jie Gong Department of Diagnostic Radiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong. Address: Room 406, Block K, Queen Mary Hospital, Hong Kong. Tel: 852-28170373, Fax: 852-28170373, Email: [email protected]

M AN U

Co-authors: Chun-Sing Wong (Corresponding author) Department of Diagnostic Radiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong. Address: Room 406, Block K, Queen Mary Hospital, Hong Kong. Tel: 852-22553307, Fax: 852-28170373, Email: [email protected]

Chun-Chung Chan Department of Geriatrics & Medicine, United Christian Hospital, Hong Kong

TE D

Lam-Ming Leung Department of Psychiatry, United Christian Hospital, Hong Kong Yiu-Ching Chu Department of Radiology, Kwong Wah Hospital, Hong Kong

AC C

Word count: 6203

EP

Running title: DKI in normal aging brain

This is an original work that has not been published nor submitted to another journal The authors have no conflict of interest to disclose Keywords: diffusional kurtosis imaging; diffusion tensor imaging; gray matter; white matter; age; Alzheimer’s disease

ACCEPTED MANUSCRIPT

Abstract: Diffusion tensor imaging (DTI) has already been extensively used to probe microstructural

RI PT

alterations in white matter tracts, and scarcely, in deep gray matter. However, results in literature regarding age-related degenerative mechanisms in white matter tracts and parametric changes in the putamen are inconsistent. Diffusional kurtosis imaging (DKI) is a mathematical extension of

SC

DTI, which could more comprehensively mirror microstructure, particularly in isotropic tissues such as gray matter.

M AN U

In this study, we utilised the DKI method and a white-matter-model that provided metrics of explicit neurobiological interpretations in healthy participants (58 in total, age from 25 to 84 years).

Tract-based whole-brain analyses and regions-of-interest (anterior and posterior limbs

of the internal capsule, cerebral peduncle, fornix, genu and splenium of corpus callosum, globus

TE D

pallidus, substantia nigra, red nucleus, putamen, caudate nucleus and thalamus) analyses were performed to examine parametric differences across regions and correlations with age. In white matter tracts, evidence was found supportive for anterior-posterior gradient and

EP

not completely supportive for retrogenesis theory. Age-related degenerations appeared to be broadly driven by axonal loss. Demyelination may also be a major driving mechanism, although

AC C

confined to the anterior brain. In terms of deep gray matter, higher mean kurtosis and fractional anisotropy in the globus pallidus, substantia nigra and red nucleus reflected higher microstructural complexity and directionality compared to the putamen, caudate nucleus and thalamus. In particular, the unique age-related positive correlations for FA, MK and KR in the putamen opposite to those in other regions call for further investigation of exact underlying mechanisms.

ACCEPTED MANUSCRIPT

In summary, the results suggested that diffusional kurtosis can provide measurements in a new dimension that were complementary to diffusivity metrics. Kurtosis together with diffusivity can more comprehensively characterize microstructural compositions and age-related changes

RI PT

than diffusivity alone. Combined with proper model, it may also assist in providing neurobiological interpretations of the identified alterations.

SC

Keywords: diffusional kurtosis imaging; diffusion tensor imaging; gray matter; white matter; age;

AC C

EP

TE D

M AN U

Alzheimer’s disease

ACCEPTED MANUSCRIPT

RI PT

Aging in Deep Gray Matter and White Matter Revealed by Diffusional Kurtosis Imaging

Abstract:

Diffusion tensor imaging (DTI) has already been extensively used to probe microstructural

SC

alterations in white matter tracts, and scarcely, in deep gray matter. However, results in literature regarding age-related degenerative mechanisms in white matter tracts and parametric changes in

M AN U

the putamen are inconsistent. Diffusional kurtosis imaging (DKI) is a mathematical extension of DTI, which could more comprehensively mirror microstructure, particularly in isotropic tissues such as gray matter.

In this study, we utilised the DKI method and a white-matter-model that provided metrics

TE D

of explicit neurobiological interpretations in healthy participants (58 in total, age from 25 to 84 years). Tract-based whole-brain analyses and regions-of-interest (anterior and posterior limbs of the internal capsule, cerebral peduncle, fornix, genu and splenium of corpus callosum, globus

EP

pallidus, substantia nigra, red nucleus, putamen, caudate nucleus and thalamus) analyses were

AC C

performed to examine parametric differences across regions and correlations with age. In white matter tracts, evidence was found supportive for anterior-posterior gradient and not completely supportive for retrogenesis theory. Age-related degenerations appeared to be broadly driven by axonal loss. Demyelination may also be a major driving mechanism, although confined to the anterior brain. In terms of deep gray matter, higher mean kurtosis and fractional anisotropy in the globus pallidus, substantia nigra and red nucleus reflected higher microstructural complexity and directionality compared to the putamen, caudate nucleus and

ACCEPTED MANUSCRIPT

thalamus. In particular, the unique age-related positive correlations for FA, MK and KR in the putamen opposite to those in other regions call for further investigation of exact underlying

RI PT

mechanisms. In summary, the results suggested that diffusional kurtosis can provide measurements in a new dimension that were complementary to diffusivity metrics. Kurtosis together with diffusivity

SC

can more comprehensively characterize microstructural compositions and age-related changes than diffusivity alone. Combined with proper model, it may also assist in providing

M AN U

neurobiological interpretations of the identified alterations.

Keywords: diffusional kurtosis imaging; diffusion tensor imaging; gray matter; white matter; age;

AC C

EP

TE D

Alzheimer’s disease

ACCEPTED MANUSCRIPT

Introduction: It is well recognised that cerebral alterations at macrostructural level, such as gray matter

RI PT

atrophy, progress with advancing age in adults without apparent pathological changes (Walhovd, et al., 2005). Preceding abnormalities detectable on macrostructural level, microstructural deteriorations such as axonal disintegration, neuron cell loss and iron accumulations have taken place much earlier. Diffusion tensor imaging (DTI) along with its two major metrics, namely

SC

fractional anisotropy (FA) and mean diffusivity (MD) has already been extensively used to probe

M AN U

microstructural alterations in white matter tracts, and much less commonly in deep gray matter (Moseley, 2002,Moseley, et al., 2002). Better characterisations of microstructural changes in both white and deep gray matter are clinically important for improving diagnosis of disorders especially those related to the hippocampus and basal ganglia, including Alzheimer’s disease, Parkinson’s disease and Huntington’s disease (Bartzokis, et al., 1999,Bartzokis, et al.,

TE D

2007,Damoiseaux, et al., 2009). Therefore, an investigation of disease-free normal adults that establishes the baseline against which patients’ data should be compared is urgently needed.

EP

In white matter, one general finding across animal model and human studies was, a pattern of decreased FA with increased radial (perpendicular to fibre projection) diffusivity indicating

AC C

demyelination (Schmierer, et al., 2007,Schmierer, et al., 2008,Song, et al., 2002). A pattern of increased MD, radial and axial (along the fibre direction) diffusivity together with decreased FA was suggested to characterize loss of axons and myelin content. The anterior-posterior gradient, which describes a decline of fibre integrity that is more pronounced in the anterior part of the brain than the posterior, has been consistently observed in published DTI studies (Sasson, et al., 2012,Schmierer, et al., 2008,Smith, et al., 2007). However, controversial results have been reported in identifying the driving mechanism leading to degeneration. Some findings support the

ACCEPTED MANUSCRIPT

demyelination and retrogenesis theory, which states that degeneration is more pronounced in late-myelinated tracts (such as splenium of the corpus callosum and fornix) than the earlymyelinated (such as posterior limb of the internal capsule and cerebral peduncles) (Brickman, et

RI PT

al., 2012,Stricker, et al., 2009). Other findings support axonal loss and Wallerian degeneration, which holds that axonal fibre degeneration distal to the point of transection or injury is the major driving mechanism (Damoiseaux, et al., 2009,Davis, et al., 2009). The coexistence of both

SC

mechanisms has also been proposed (Bennett, et al., 2010,Burzynska, et al., 2010). Compared to white matter, deep gray matter has received much less attention from DTI studies and its

M AN U

corresponding degenerative mechanism remain poorly understood. Possible cell loss with aging, which manifested as decreased FA and increased radial diffusivity, was reported in the dorsal substantia nigra (Vaillancourt, et al., 2012). One in vivo DTI study also reported an iron deposition induced increase of FA in the putamen (Pfefferbaum, et al., 2010). To date, DTI studies of deep

TE D

gray matter are scarce and the results are inconsistent (Bhagat and Beaulieu, 2004,Pal, et al., 2011,Pfefferbaum, et al., 2010,Wang, et al., 2010). For example, Wang’s report of decreased MD with aging disagrees with Pfefferbaum’s report of increased MD in the putamen (Pfefferbaum, et

EP

al., 2010,Wang, et al., 2010).

AC C

The DTI model itself is also prohibited by pre-assumption of Gaussian-distributed proton diffusion from fully characterization of the actual non-Gaussian diffusion caused by obstacles such as cell membranes and organelles. The more recent method called diffusional kurtosis imaging (DKI) is an extension of DTI to the fourth order (Jensen, et al., 2005,Liu, et al., 2004). By quantifying the deviation from Gaussian distribution, new metrics such as mean kurtosis (MK) are able to capture the “complexity” of tissue microstructure. In addition to improved accuracy (Veraart, et al., 2011), DKI may outperform DTI with possibly higher sensitivity, particularly in

ACCEPTED MANUSCRIPT

isotropic tissues such as gray matter (Gong, et al., 2013,Jensen, et al., 2005). To our knowledge, there has been only one study that investigated DKI parametric alterations in normal aging adults. However, the relatively small cohort, limited age range and unreported directional metrics

RI PT

impaired its suitability for serving as a normal reference (Latt, et al., 2013).

In a large cohort with a broad age range, here we measure age-related microstructural

SC

changes using the DKI method, advanced tract-based spatial statistics (TBSS) analyses (Smith, et al., 2006) and an improved image registration method called Diffeomorphic Anatomical

M AN U

Registration Through Exponentiated Lie Algebra (DARTEL) (Ashburner, 2007). We hypothesised that brain microstructure could be comprehensively mirrored by DKI with additional kurtosis information, especially in deep gray matter. The primary objective of this study was to provide new insight into the current inconsistencies in the literature using exhaustive analyses of

TE D

diffusivity and kurtosis metrics. In accordance with a DKI-based white-matter-model, we further calculated two metrics, namely intra−axonal space axonal water fraction (AWF) and extra−axonal space tortuosity (Fieremans, et al., 2011). It has been shown that AWF is a measurement of

EP

fraction of intra−axonal water volume and is sensitive to axonal loss. Tortuosity is an indirect measurement of myelinated axonal fraction and is sensitive to demyelination. Based on

AC C

parametric correlations with age and differences between regions of interest (ROI), we also presented plausible interpretations for underlying neurobiological mechanisms in deep gray and white matter. Lastly, kurtosis metrics were compared against diffusivities in sensitivity for detecting microstructural differences across regions and changes with aging.

ACCEPTED MANUSCRIPT

Methods and materials:

RI PT

a. Participants and cognitive assessment This study was approved by an institutional review board. All participants were recruited from a local tertiary referral centre and examined by a research geriatricians or psychiatrist. They were either relatives of patients or staff members. The assessment battery included an interview with physical

and

neurological

SC

participants and collaterals,

examinations,

M AN U

semistructured mental status examination (MSE), global deteriorating scale (GDS) and a global cognitive test using the mini–mental state examination (MMSE). Normal cognition was confirmed through MSE and GDS of stage 1. Participants with a neurological illness, a psychiatric disorder or cognitive impairment were excluded. Fifty-eight healthy participants underwent MRI scans: 18 young adults (25 ~ 40 years), 20 middle-aged adults (42 ~ 56 years) and 20 older adults (65 ~ 84

TE D

years). Informed consent was obtained from each subject. Their demographical and cognitive details are summarised in Table 1.

EP

b. Magnetic resonance images acquisition

AC C

All scans were performed on a Philips 3T MRI Achieva scanner (Philips Healthcare, Best, The Netherlands). A body coil for excitation and an eight-channel head coil for reception were used. DKI images were acquired with 3 b values (0, 1000 and 2000 s/mm2) across 32 diffusion gradient directions. Other imaging parameters were as follows: TR/TE = 2000/69 ms; nominal resolution = 2.5×2.5×3 mm3; reconstruction resolution = 2×2×3 mm3; matrix size = 128 × 128; 33 axial slices with no interslice gap to cover the brain; SENSE-reduction factor = 2; and partial Fourier encoding = 3/4. The acquisition time was 15 min. For anatomical reference, 3D T1-weighted fast-field-echo

ACCEPTED MANUSCRIPT

images were acquired with the following parameters: TR/TE = 7.0/3.2 ms; TI = 800 ms; nominal/reconstruction resolution = 1×1×1 mm3; and 167 slices. The acquisition time was 10 min. Fluid attenuated inversion recovery, T2-weighted turbo-spin-echo and susceptibility-weighted

RI PT

images were also acquired for screening subjects with vascular damage or lesions. No participants were excluded.

SC

c. Derivation of summary, directional and white-matter-model metrics Diffusion weighted images were first corrected for eddy-current distortion and head motion

M AN U

using 12 degrees of freedom FLIRT (FMRIB Software Library) in reference to b0 images (Jenkinson, et al., 2002). Spatial Gaussian smoothing using a full-width-half-maximum of 2.5 mm was then performed. Apparent diffusion and kurtosis coefficients were calculated to obtain diffusion tensor and kurtosis tensor. On basis of these tensors, summary metrics of FA, MD and

TE D

MK, and directional metrics of axial diffusivity (DA), radial diffusivity (DR), axial kurtosis (KA) and radial kurtosis (KR) were derived (Jensen and Helpern, 2010,Tabesh, et al., 2011). White-mattermodel metrics of AWF and tortuosity were further derived according to a DKI-based two-

EP

compartment white-matter-model (Fieremans, et al., 2011). All metric calculations were performed using in-house MATLAB (MathWorks, Natick, MA, USA) programs. Fig. 1 shows

AC C

example parametric maps from one middle-aged subject (age 47 years, male).

d. Whole brain spatial statistics and regional analyses of white matter tracts

For unbiased analyses of whole brain white matter tracts using TBSS, the FA maps of all the participants were first registered into a native space representative target using non-linear

ACCEPTED MANUSCRIPT

transform. They were then affine-aligned into MNI152 space. The aligned FA maps were further averaged to obtain the mean FA map. This mean FA map was thinned (thresholded at a FA value of 0.2) to generate an FA skeleton which represents major white matter tracts common to all

RI PT

participants. Next, parametric maps were projected onto the template. Controlled for sex as covariate, effect of age on the parametric values was modelled using a general linear model and ‘randomise tool’ (FMRIB Software Library) with 5000 permutations. Significance was tested at

SC

cluster level using threshold-free cluster enhancement (TFCE) and family-wise error (FWE) corrections for multiple comparisons at p < 0.05 (Smith and Nichols, 2009). JHU DTI WM atlas was

M AN U

used to identify white matter tracts that exhibited significant correlations. For tract-based ROI analyses, the JHU DTI WM atlas was used to extract masks for anterior and posterior limbs of the internal capsule, cerebral peduncle, fornix and splenium and genu of corpus callosum. Regional

TE D

values of all metrics were later calculated using these masks for every subject.

e. Regional analyses of deep gray matter To draw ROIs for deep gray matter in a time efficient manner and to control for observers’ bias,

EP

we utilised the improved image registration algorithm of DARTEL (Ashburner, 2007) contained in SPM8 (Wellcome Trust Centre for Neuroimaging, London, UK). First, 3D T1 weighted images were

AC C

resliced and coregistered to b0 images for each subject. The resliced native space T1 weighted images were segmented into white matter, gray matter and CSF using unified segmentation approach with default settings (Ashburner and Friston, 2005). Segmented tissue maps were further utilised as three tissue masks. By simultaneously aligning gray and white matter masks across participants through a progressively refining process, a T1 weighted custom template common to all participants was generated. For each subject, flow field that described the

ACCEPTED MANUSCRIPT

transformation from native space T1 weighted images into the custom template space was recorded. Custom b0 template was also generated by transforming native b0 images of one subject into the template space using the recorded flow field. Second, ROIs of the globus pallidus, putamen,

RI PT

caudate nucleus and thalamus were then placed bilaterally on T1 weighted template by one experienced radiologist using ImageJ (v. 1.14, National Institutes of Health, Bethesda, MD, USA). The ROIs covered the whole volume of each tissue compartments (Fig. 2). On the b0 template,

SC

ROIs of the substantia nigra and red nucleus were placed by the same radiologist in the same manner. All of the ROIs were double checked by another radiologist. ROIs were converted into

M AN U

binary masks and eroded by 2 pixels to avoid partial volume effect or contamination from adjacent structures. Third, the ROI masks in the template space were back projected into native space of individual subject using flow fields obtained via DARTEL registration processing (first step). Regional DKI parametric values were then calculated using the back projected native space ROI

TE D

masks for each subject.

f. Statistical analyses for regional parametric values

EP

To investigate differences in microstructural compositions, degrees of complexity and directionality, parametric mean values were compared across six white matter tracts (for FA, MD,

AC C

MK, AWF and tortuosity) and across six deep gray matter regions (for FA, MD and MK). Comparisons were performed using Analysis of Covariance (ANCOVA) with the general linear model, in which region was a within-subjects factor, and age and sex were covariates. Adjustment was not performed for education, which was theoretically not a confounding factor. Multiple comparisons were corrected using Bonferroni method for family wise error.

ACCEPTED MANUSCRIPT

Metrics were also compared in a similar way between averaged anterior (anterior limb of the internal capsule + genu of corpus callosum) and posterior tracts (posterior limb of the internal capsule + splenium of corpus callosum), between averaged late-myelinated (the fornix + splenium

RI PT

of corpus callosum) and early-myelinated tracts (the cerebral peduncle + posterior limb of internal capsule), and between averaged all white and deep gray matters, using the general linear

SC

model covarying for age and sex.

Associations between age and regional parametric values were investigated using multivariate

M AN U

multiple regressions with sex controlled as a covariate. F statistic and p value were recorded from corrected model. Standardized β values were also recorded as rates of age-related parametric changes. P value < 0.05 (two-tailed) was considered as statistically significant. All statistical

TE D

analyses were performed using SPSS (v. 19.0.0, SPSS, Chicago, IL, USA).

EP

Results:

AC C

a. Whole brain spatial statistics of white matter tracts FA exhibited significant age-related decreases in widespread tracts including genu, body and splenium of the corpus callosum, anterior and posterior limbs of internal capsule, cerebral peduncle, uncinate fasciculus, anterior corona radiate and fornix, sagittal stratum, posterior thalamic radiation and hippocampal cingulum (Fig. 3). MK exhibited negative correlations that were distributed similarly to those of FA, with additional correlations observed in the superior

ACCEPTED MANUSCRIPT

longitudinal fasciculus and superior and posterior corona radiate (Fig. 3). MD exhibited negative correlations with age in more tracts than MK, including the cingulated gyrus and external capsule

RI PT

(Fig. 3). Interestingly, age-related decreases of AWF were widely observed in tracts such as genu, body and splenium of the corpus callosum, fornix, anterior and posterior limbs of internal capsule,

SC

cerebral peduncle, anterior, superior and posterior corona radiate, posterior thalamic radiation, sagittal stratum, external capsule, superior longitudinal fasciculus and uncinate fasciculus. These

M AN U

age-related decreases suggested a widespread decrease of axonal density with aging (Fig. 3). Noteworthy results were observed in tortuosity, which exhibited negative correlations with age only in the anterior corona radiate and fornix (Fig. 3).

b. Regional analyses of white matter

TE D

In parametric comparisons across the white matter tracts, except for the cerebral peduncle vs. genu of corpus callosum in FA, anterior vs. posterior limbs of internal capsule, and cerebral

EP

peduncle vs. genu of corpus callosum in MD, anterior limb of internal capsule vs. genu of corpus callosum in AWF, anterior limb of internal capsule vs. cerebral peduncle, and posterior limb of

AC C

internal capsule vs. genu of corpus callosum in tortuosity, all other pair-wise comparisons were significant. Notably, pair-wise comparisons in MK were all significant (Fig. 4). Regarding parametric associations with age, correlations were significant for FA and MD in all tracts and significant for MK and AWF in tracts other than splenium of the corpus callosum and cerebral peduncle. Correlation was only significant in the fornix for tortuosity (Table 2).

ACCEPTED MANUSCRIPT

All parametric comparisons between anterior and posterior tracts and between late- and early-myelinated tracts were significant except for tortuosity in late- vs. early-myelinated tracts (Fig. 5). Regarding parametric association with age, anterior tracts exhibited significant

RI PT

correlations for all metrics. Correlations of the posterior and early-myelinated tracts were significant except for tortuosity. The least number of correlations with age was found in latemyelinated tracts (only for MD and MK). In terms of rates of age-related changes, those of anterior

SC

tracts were all greater than corresponding correlations of posterior tracts. Contrary to expectation, the effect sizes of age on metrics were larger in early-myelinated tracts than in the late-myelinated

M AN U

(Table 3).

c. Regional analyses of deep gray matter

All other pair-wise comparisons across deep gray matter structures were significant except for:

TE D

the globus pallidus vs. substantia nigra in FA, substantia nigra vs. red nucleus vs. putamen for MD, globus pallidus vs. red nucleus, and caudate nucleus vs. thalamus for MK. Observed parametric ranking orders, especially from MK and FA, revealed distinct degrees of diffusional directionality

EP

and heterogeneity, in which the globus pallidus, substantia nigra and red nucleus were much

AC C

higher than putamen, caudate nucleus and thalamus (Fig. 6). Strikingly, the putamen exhibited unique pattern distinctly different from the other regions in parametric associations with age. FA correlated negatively with age in the globus pallidus, while positively in putamen. MK correlated negatively with age in the caudate nucleus, thalamus and globus pallidus, while positively in putamen. Similarly negative regression with age in the putamen opposite to the other regions was also observed for KR. For MD, DA and DR, all significant parametric correlations with age were positive (Table 4, Fig 7).

ACCEPTED MANUSCRIPT

d. Comparisons between averaged deep gray and white matters FA, MD and MK were significantly higher in averaged white matter tracts as compared with

RI PT

averaged deep gray matter. Other parametric regressions with age were significant except for FA

SC

and MK in averaged deep gray matter (Table 5, Fig. 8).

M AN U

Discussion:

For the first time in a large cohort of adults with a broad age range, we utilised the DKI method and a white-matter-model to comprehensively characterise the diffusional features and to provide explicit neurobiological interpretations. Using the improved post-processing methods of TBSS and DARTEL registration, two controversial issues regarding age-related degenerative

TE D

mechanisms in white matter tracts and parametric changes in the putamen were investigated. The merits of kurtosis metrics were also discussed in comparison to diffusivities, especially in deep

EP

gray matters that were relatively isotropic.

AC C

a. Microstructural changes in white matter tracts Some TBSS based studies reported positive correlations between FA and age in the inferior longitudinal fasciculus, anterior thalamic radiation and most consistently in superior longitudinal fasciculus during adolescence (Peters, et al., 2012,Qiu, et al., 2008). These findings highlighted the development of white matter integrity that commonly peaks at approximately 20 years of age (Sullivan and Pfefferbaum, 2006) and subsequently degenerates. Post-mortem, animal and DTI investigations have suggested that axonal loss and demyelination with aging causes the

ACCEPTED MANUSCRIPT

degeneration reflected by increased MD and decreased FA. Consistent with previous reports (Madden, et al., 2012,Peters, et al., 2012), age-related increases in MD and decreases in FA were observed in nearly all white matter tracts. In most of these tracts, loss of diffusional barriers was

RI PT

also captured by the decrease in MK in current study. Declined microstructural complexity reduced the diffusional heterogeneity, and thus MK. To explain such decreases in white matter integrity, several neurodegenerative theories or patterns, such as anterior-posterior gradient,

SC

retrogenesis theory and Wallerian degeneration have been previously proposed.

M AN U

An anterior-posterior gradient of degeneration, which describes a decline of integrity that is more pronounced in the anterior part of the brain than the posterior, has been consistently observed in published DTI studies (Sasson, et al., 2012,Schmierer, et al., 2008,Smith, et al., 2007). Our study provided novel supportive evidence. Axons in genu of the corpus callosum are thinner

TE D

and less heavily myelinated than those in splenium (Aboitiz, et al., 1996). Previous DTI studies reported lower FA in genu than splenium of the corpus callosum (Sullivan, et al., 2001). Lower FA in anterior than posterior limb of the internal capsule was also reported (Kawaguchi, et al., 2010).

EP

In current study, adding to lower FA and higher MD, much lower MK in genu of the corpus callosum and anterior limb of internal capsule than their posterior counterparts were also

AC C

observed, indicating lower microstructural integrity and complexity. White-matter-model metrics offered more explicit interpretations: AWF and tortuosity were much lower in the anterior parts than the posterior, reflecting lower axonal density and degree of myelination. Such anteriorposterior differences were also confirmed in comparisons between averaged anterior and posterior tracts. In spatial statistical analyses, although anterior and posterior tracts exhibited similarly spread age-related decreases in FA, MK and AWF and increases in MD, significant decrease in tortuosity was only observed in the anterior corona radiate and the fornix, indicating a

ACCEPTED MANUSCRIPT

possible age-related demyelination process that is only evident in the anterior brain. Regional parametric correlations with age also exhibited an anterior-posterior gradient. The age effects on FA, MD, MK and AWF in genu of the corpus callosum and anterior limb of internal capsule were

RI PT

much larger than their posterior counterparts, which agreed with previously reported faster agerelated decrease of FA in the genu than splenium (Ota, et al., 2006), and in anterior than posterior limb of the internal capsule (Kawaguchi, et al., 2010). Further analyses of averaged anterior tracts

SC

confirmed their higher vulnerability to aging than averaged posterior tracts.

M AN U

According to retrogenesis theory, demyelination is the major driving mechanism of degeneration and is more severe in late-myelinated tracts (the fornix and splenium of corpus callosum), which may be more vulnerable to aging compared to early-myelinated tracts (the posterior limb of internal capsule and cerebral peduncles) (Brickman, et al., 2012,Kinney, et al.,

TE D

1988,Stricker, et al., 2009). Previous DTI studies that examined the retrogenesis theory have reported controversial results. Although a study by Stricker indicated that late-myelinated tracts were more susceptible to Alzheimer’s disease than early-myelinated tracts (Stricker, et al., 2009),

EP

conflicting results were reported by Brickman in normal aging population (Brickman, et al., 2012). In the latter study, the greatest age effects on decrease in FA and increase in DR were both found

AC C

in the cerebral peduncle (early-myelinated). Furthermore, the superior longitudinal fasciculus (late-myelinated) showed less robust age-related changes than early-myelinated tracts. Similar to the result reported by Brickman, our age-related decrease of MK in splenium of the corpus callosum (late-myelinated) failed to reach significant while that of MK in posterior limb of the internal capsule (early-myelianted) was significant. Averaged late-myelianted tracts also showed slower age-related changes in FA, MD, MK and AWF than the early-myelinated. In addition, splenium of the corpus callosum (late-myelinated) showed the highest FA, AWF and tortuosity

ACCEPTED MANUSCRIPT

values among the four tracts. MD value of averaged late-myelinated tracts was even lower than that of the early-myelinated. Therefore, our results did not completely support higher overall integrity and complexity of early-myelinated tracts or greater aging effects in late-myelinated

RI PT

tracts, which were expected from retrogenesis theory.

Wallerian (secondary) degeneration theory suggests that axonal fibre degeneration distal to

SC

the point of transection or injury is the leading cause of degeneration (Damoiseaux, et al., 2009,Davis, et al., 2009). Our results were in line with this hypothesis. Previous DTI studies

M AN U

reported controversial results in terms of identifying the major driving mechanism that caused the decrease of integrity with aging. Some findings indicated that the decrease in FA was the result of an increased DR that was more strongly associated with demyelination (Bhagat and Beaulieu, 2004,Inano, et al., 2011). Other findings indicated that the decrease in FA and the increases in DR

TE D

with aging were due to decreased axon density. Bennett and Burzynska proposed several degenerative patterns reflecting either axonal loss or demyelination or coexistence of both mechanisms (Bennett, et al., 2010,Burzynska, et al., 2010). In agreement with these studies, age-

EP

related decrease of FA, MK and increase of MD were observed for nearly all tracts in whole brain and regional analyses. Although they suggested lower microstructural integrity, complexity and

AC C

viscosity, their neurobiological meanings are uncertain. Notably, white-matter-model metrics were helpful in providing explicit interpretations specific to underlying mechanisms (Fieremans, et al., 2011). In spatial statistical analyses, our findings of widespread decreases in AWF and smaller scale decreases in tortuosity mainly confined to the anterior corona radiate and fornix suggested the coexistence of both axonal loss and demyelination in the anterior brain and a universal axonal loss. This was further confirmed in regional correlation analyses: age-related correlations were observed in four tracts for AWF, while only in fornix for tortuosity.

ACCEPTED MANUSCRIPT

Briefly, current results suggested that age-related white matter degenerations may be broadly driven by axonal loss. In anterior brain, demyelination was also a probable major mechanism contributing to disintegration. Such coexistence of both mechanisms was in line with the

RI PT

Wallerian degeneration theory. In addition, comparisons of parametric values and age-related changes across regions were more supportive of the anterior-posterior gradient degeneration than the retrogenesis theory. Although AWF and tortuosity provided extended view on in-vivo

M AN U

vivo comparisons are critical for further conclusions.

SC

white matter assessments, additional investigations on proper imaging parameters as well as ex-

b. Microstructural changes in deep gray matter

Correlations with age were observed in all six deep gray matter structures from at least two metrics each. It should be noted that the parametric trends of decrease or increase were not

TE D

uniform among the investigated regions. For example, MK exhibited age-related increases in the putamen, but decreases in globus pallidus, caudate nucleus, and thalamus. Extending on previous DTI studies, we speculated that such distinctly different pattern may result from mixed effects of

EP

several mechanisms such as axonal disintegration, cell loss and iron accumulation. On one hand, the original microstructural complexity, anisotropy and diffusivity varied between the structures.

AC C

On the other, the rates of above degeneration process with aging were different in these structures. Among all six deep gray matter structures, the globus pallidus, substantia nigra and red nucleus have the most complicated microstructural compositions. The globus pallidus is constructed by large disc-shaped neurons that are perpendicular to the myelinated axons (Percheron, et al., 1984). The substantia nigra has neurons similar to globus pallidus with thinly myelinated axons (Yelnik, et al., 1987). The red nucleus consists of densely packed cells and small

ACCEPTED MANUSCRIPT

myelinated axons (Onodera and Hicks, 2009). These three compartments are also among the tissues that have the highest iron concentration in the brain (Haacke, et al., 2005). Iron atoms are stored in ferritin that has large spherical protein coat. Myelinated axons, large cells and ferritin

RI PT

protein constitute a complex microstructure that imposed extensive hindrance to free diffusion of water molecules. The quantified diffusivity will therefore manifest as low values. Typically, the myelinated axons are spatially oriented, which constrain molecule diffusion along the fibre

SC

direction, and subsequently manifest as high anisotropy. All these obstacles also cause proton diffusion to deviate from Gaussian distribution, which leads to an increase in kurtosis. In the

M AN U

present study, MK and FA of the globus pallidus, substantia nigra and red nucleus were much higher, and MD was lower compared to putamen, caudate nucleus and thalamus, confirmed the higher complexities of underlying microstructures.

TE D

One previous study reported reduced FA and increased DR in the dorsal substantia nigra of older adults compared to young adults. The authors concluded that such observations were an accurate reflection of age-related cell loss (Vaillancourt, et al., 2012). Other DTI studies reported

EP

increased diffusivity and decreased FA in the globus pallidus of older adults (Bhagat and Beaulieu, 2004,Pfefferbaum, et al., 2010). Our observation of age-related increase of diffusivity in the

AC C

substantia nigra, red nucleus and globus pallidus, and decrease of FA in globus pallidus corroborated their findings. Our negative correlations between MK, KR and age in the globus pallidus were also supported by a recent DKI study (Latt, et al., 2013). One probable interpretation is that axonal disintegration and cell loss exerted dominating effects on the microstructural alterations in these regions. The decreased diffusional obstructions imposed less restriction on the free diffusion of protons, which manifested as increased diffusivity metrics. The directional

ACCEPTED MANUSCRIPT

propensity of water diffusion was also reduced due to axonal disintegration, which manifested as decreased FA.

RI PT

The microstructural compositions of the caudate nucleus and thalamus were relatively simple. They were predominantly constituted by neurons and glia. Such simple and isotropic microstructures were characterised by much lower MK and FA and higher diffusivity, compared to

SC

those of the globus pallidus, substantia nigra and red nucleus. Iron concentration levels in these structures were also among the lowest (Haacke, et al., 2005). Hallgren and Sourander even

M AN U

reported an age-related decrease in iron content in the thalamus after 30 years of age in a postmortem study (Hallgren and Sourander, 1958). Corresponding ferritin contents that were either decreasing or at consistently low level may contribute substantially less to microstructural complexity than those of the globus pallidus, substantia nigra and red nucleus. Axonal

TE D

disintegration and cell loss were therefore expected to dominate age-related microstructural changes. In the caudate nucleus, two previous DTI studies reported conflicting results. Wang reported an age-related decrease of MD, DA and DR (Wang, et al., 2010), while Pfefferbaum

EP

reported increased DA in older adults (Pfefferbaum, et al., 2010). Our observations of age-related increase in MD and DA were in agreement with Pfefferbaum. In the thalamus, our findings of age-

AC C

related increasing in MD and DA were supported by a previous study from Bhagat and Beaulieu (Bhagat and Beaulieu, 2004), and decreasing in MK was supported by a recent DKI study (Latt, et al., 2013). In addition to increasing diffusivity, the decreasing MK, which suggested a loss of microstructural complexity, confirmed our speculation that axonal disintegration and cell loss may dominate the age-related microstructural alterations in the thalamus and caudate nucleus.

ACCEPTED MANUSCRIPT

Overall, the most striking results were observed in the putamen. Although the microstructural composition of the putamen was relatively simple and similar to caudate nucleus (Kemp and Powell, 1971), as reflected by the similarly low FA and MK in the current study, its iron

RI PT

concentration was among the highest in the brain (Hallgren and Sourander, 1958). The iron deposition rate of the putamen was also higher than that of caudate nucleus (Hallgren and Sourander, 1958). The comparatively simpler composition of neurons and higher level of iron

SC

content may therefore contribute similarly to diffusional obstructions. In the putamen, age-related MD increase (Abe, et al., 2002,Bhagat and Beaulieu, 2004,Pfefferbaum, et al., 2010) and FA

M AN U

increase (Bhagat and Beaulieu, 2004,Camara, et al., 2007,Hasan and Frye, 2011,Pfefferbaum, et al., 2010,Wang, et al., 2010) were widely reported in previous DTI studies. Although the previous DKI study reported contradictory results of decreasing diffusivity (Latt, et al., 2013), our findings were in line with these DTI studies. In contrast to decreasing FA in the caudate nucleus (marginally

TE D

significant), the age-related increasing FA in the putamen may be attributed in part to its high level and fast iron deposition. A previous DTI study that quantified iron content in vivo also

EP

suggested an iron deposition induced increase in FA in the putamen (Pfefferbaum, et al., 2010). More specifically, in contrast to the caudate nucleus and thalamus, a positive rather than

AC C

negative correlation with age was observed for MK in the putamen, indicating increase in diffusional heterogeneity and higher structural complexity. DR in the putamen increased with aging at a much slower rate compared to the thalamus (βstand = 0.338 vs. 0.666). This may suggest that accumulated ferritin protein presented additional obstacles mainly along the ‘radial’ direction. Increasing KR corroborated the presence of more and more densely packed obstructions along the radial direction. The comparatively high rate of increase in DA (βstand = 0.665) and slower increase in DR (βstand = 0.338) in the putamen resulted in higher directionality and further increased FA.

ACCEPTED MANUSCRIPT

Compared with the thalamus, the hindered increases in DA and DR were both slower in the putamen, which manifested as a slower increase of MD (βstand = 0.522 vs. 0.680). Although our finding was coherent with previous study by Pfefferbaum and provided further insight, we cannot

RI PT

assert that iron deposition is the only underlying mechanism. Another recent DTI study suggested that FA increase might be artifactual and was result of decrease in signal to noise ratio (Rulseh, et al., 2013). Further histological and imaging studies for unveiling the exact underlying mechanism

SC

are needed.

M AN U

c. Deep gray matter vs. white matter, kurtosis vs. diffusivity As a measurement of diffusional anisotropy, FA was reflecting microstructural “directionality”. Unlike a sharp decrease with aging observed for FA in anisotropic white matter (averaged), which suggested disintegration of fibre bundles, the sustained low FA in near isotropic

TE D

deep gray matter (averaged) was expected. Sensitive to changes in diffusion magnitude and reflecting microstructural “viscosity”, MD was increased with age similarly in both averaged white and deep gray matters, which indicated a loss of diffusional barriers. Diffusional heterogeneity

EP

decreased with axonal loss and demyelination, manifested as decreasing MK with aging in averaged white matter. Decreasing of MK in averaged deep gray matter was marginally significant,

AC C

which was probably counterbalanced by the increasing MK in the putamen. Taking all regional comparisons into consideration, kurtosis metrics could serve as microstructural biomarkers in a new dimension complementary to FA and MD. MD was sensitive to any alteration of hindrance or restriction in water diffusion, but not specific. It was the most robust metric in detecting microstructural changes by correlating with age in all six white matter and six deep gray matter regions. However, because of low specificity, MD was unable to detect

ACCEPTED MANUSCRIPT

subtle differences in microstructural compositions and exhibited the least number of significant findings in parametric comparisons across regions. FA was responding specifically to change in diffusional directionality. Therefore it was more sensitive in white matter than in gray matter,

RI PT

correlating significantly with age in all six white matter but only two deep gray matter regions. Quantifying diffusional heterogeneity, MK was sensitive to alterations in microstructural “complexity” even in regions without apparent diffusion directionality. It exhibited more

SC

significant correlations with age than FA in deep gray matters. It also has larger age-related effect size than FA and MD in three regions (the globus pallidus, genu of corpus callosum and posterior

M AN U

limb of internal capsule). Most importantly, the unique diffusional feature with aging in the putamen was only captured and manifested by kurtosis (MK, KR). In addition, all pair-wise comparisons of MK across white matter tracts were significant. MK also revealed distinct differences in microstructural compositions of deep gray matter regions and clearly separated

TE D

them into two groups that agreed well with the histological report. Notably, combined with proper biophysical model, diffusional kurtosis information can help derive metrics that bear explicit neurobiological interpretations, which may potentially provide new insights into underlying

AC C

EP

degenerative mechanisms.

Limitations of the present study should be recognized. First, our study of age-related microstructural changes was cross-sectional that needed to be replicated and confirmed in larger longitudinal studies. Second, although current results provided evidence supporting probable neurological interpretations of underlying mechanisms, we cannot definitively determine the extent to which individual mechanisms altered the microstructure without the use of non-invasive

ACCEPTED MANUSCRIPT

biomarkers sensitive to these individual mechanisms. The observed changes in the diffusivity and kurtosis metrics were combinations of different effects that should be interpreted with caution. In white matter, aside from axonal loss and demyelination, post-mortem studies have demonstrated

RI PT

other possible age-related microstructural alterations, including a decline in the length of myelinated fibres (Marner and Pakkenberg, 2003), trapping of fluid between thin or lysed sheaths, or bulbous swelling of oligodendrocytes (Peters and Sethares, 2002). In deep gray matter, other

SC

mechanisms such as loss of dendrite connections, tissue compaction and gliosis have also been proposed to account for changes in DTI metrics (Hasan, et al., 2008,Wang, et al., 2010).

M AN U

Nonetheless, the normative basis established by the current study will be critical for detecting abnormalities in future studies. Given that microstructural degenerations usually occur before apparent disease symptoms, such as those seen in Alzheimer’s disease and Parkinson’s disease,

neurodegenerative diseases.

TE D

our findings will have great clinical importance for improving the early detection of

EP

In summary, we utilised the DKI method in both deep gray and white matters of normal aging adults. The results suggested that diffusional kurtosis can provide measurements in a new

AC C

dimension complementary to those of diffusivity metrics. Kurtosis together with diffusivity can more comprehensively characterize microstructural compositions and age-related changes than diffusivity alone. Combined with a proper biophysical model, it may also assist in providing neurobiological interpretations of the identified alterations. In white matter tracts, evidence supportive for anterior-posterior gradient and not completely supportive for retrogenesis theory was found. Age-related degenerations appeared to be broadly driven by axonal loss.

ACCEPTED MANUSCRIPT

Demyelination may also be a major driving mechanism, although confined to the anterior brain. In terms of deep gray matter, higher MK and FA in the globus pallidus, substantia nigra and red nucleus mirrored the higher microstructural complexity and directionality compared to putamen,

RI PT

caudate nucleus and thalamus. In particular, the unique age-related positive correlations for FA, MK and KR in the putamen opposite to those in other regions call for further investigation of exact

M AN U

SC

underlying mechanisms.

Acknowledgements: The authors would like to thank Dr. Yee Tak Fong for valuable suggestions

AC C

EP

TE D

and Dr. Samantha Holdsworth for proofreading.

ACCEPTED MANUSCRIPT

References

AC C

EP

TE D

M AN U

SC

RI PT

Abe, O., Aoki, S., Hayashi, N., Yamada, H., Kunimatsu, A., Mori, H., Yoshikawa, T., Okubo, T., Ohtomo, K. 2002. Normal aging in the central nervous system: quantitative MR diffusion-tensor analysis. Neurobiol Aging 23(3), 433-41. Aboitiz, F., Rodriguez, E., Olivares, R., Zaidel, E. 1996. Age-related changes in fibre composition of the human corpus callosum: sex differences. Neuroreport 7(11), 1761-4. Ashburner, J. 2007. A fast diffeomorphic image registration algorithm. Neuroimage 38(1), 95-113. Ashburner, J., Friston, K.J. 2005. Unified segmentation. Neuroimage 26(3), 839-51. Bartzokis, G., Cummings, J.L., Markham, C.H., Marmarelis, P.Z., Treciokas, L.J., Tishler, T.A., Marder, S.R., Mintz, J. 1999. MRI evaluation of brain iron in earlier- and later-onset Parkinson's disease and normal subjects. Magn Reson Imaging 17(2), 213-22. Bartzokis, G., Lu, P.H., Tishler, T.A., Fong, S.M., Oluwadara, B., Finn, J.P., Huang, D., Bordelon, Y., Mintz, J., Perlman, S. 2007. Myelin breakdown and iron changes in Huntington's disease: pathogenesis and treatment implications. Neurochem Res 32(10), 1655-64. Bennett, I.J., Madden, D.J., Vaidya, C.J., Howard, D.V., Howard, J.H., Jr. 2010. Age-related differences in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy aging. Hum Brain Mapp 31(3), 378-90. Bhagat, Y.A., Beaulieu, C. 2004. Diffusion anisotropy in subcortical white matter and cortical gray matter: changes with aging and the role of CSF-suppression. J Magn Reson Imaging 20(2), 216-27. Brickman, A.M., Meier, I.B., Korgaonkar, M.S., Provenzano, F.A., Grieve, S.M., Siedlecki, K.L., Wasserman, B.T., Williams, L.M., Zimmerman, M.E. 2012. Testing the white matter retrogenesis hypothesis of cognitive aging. Neurobiol Aging 33(8), 1699-715. Burzynska, A.Z., Preuschhof, C., Backman, L., Nyberg, L., Li, S.C., Lindenberger, U., Heekeren, H.R. 2010. Agerelated differences in white matter microstructure: region-specific patterns of diffusivity. Neuroimage 49(3), 2104-12. Camara, E., Bodammer, N., Rodriguez-Fornells, A., Tempelmann, C. 2007. Age-related water diffusion changes in human brain: a voxel-based approach. Neuroimage 34(4), 1588-99. Damoiseaux, J.S., Smith, S.M., Witter, M.P., Sanz-Arigita, E.J., Barkhof, F., Scheltens, P., Stam, C.J., Zarei, M., Rombouts, S.A. 2009. White matter tract integrity in aging and Alzheimer's disease. Hum Brain Mapp 30(4), 1051-9. Davis, S.W., Dennis, N.A., Buchler, N.G., White, L.E., Madden, D.J., Cabeza, R. 2009. Assessing the effects of age on long white matter tracts using diffusion tensor tractography. Neuroimage 46(2), 530-41. Fieremans, E., Jensen, J.H., Helpern, J.A. 2011. White matter characterization with diffusional kurtosis imaging. Neuroimage 58(1), 177-88. Gong, N.J., Wong, C.S., Chan, C.C., Leung, L.M., Chu, Y.C. 2013. Correlations between microstructural alterations and severity of cognitive deficiency in Alzheimer's disease and mild cognitive impairment: a diffusional kurtosis imaging study. Magn Reson Imaging 31(5), 688-94. Haacke, E.M., Cheng, N.Y., House, M.J., Liu, Q., Neelavalli, J., Ogg, R.J., Khan, A., Ayaz, M., Kirsch, W., Obenaus, A. 2005. Imaging iron stores in the brain using magnetic resonance imaging. Magn Reson Imaging 23(1), 125. Hallgren, B., Sourander, P. 1958. The effect of age on the non-haemin iron in the human brain. J Neurochem 3(1), 41-51. Hasan, K.M., Frye, R.E. 2011. Diffusion Tensor-Based Regional Gray Matter Tissue Segmentation Using the International Consortium for Brain Mapping Atlases. Human Brain Mapping 32(1), 107-17.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Hasan, K.M., Halphen, C., Boska, M.D., Narayana, P.A. 2008. Diffusion tensor metrics, T2 relaxation, and volumetry of the naturally aging human caudate nuclei in healthy young and middle-aged adults: possible implications for the neurobiology of human brain aging and disease. Magn Reson Med 59(1), 713. Inano, S., Takao, H., Hayashi, N., Abe, O., Ohtomo, K. 2011. Effects of age and gender on white matter integrity. AJNR Am J Neuroradiol 32(11), 2103-9. Jenkinson, M., Bannister, P., Brady, M., Smith, S. 2002. Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 17(2), 825-41. Jensen, J.H., Helpern, J.A. 2010. MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed 23(7), 698-710. Jensen, J.H., Helpern, J.A., Ramani, A., Lu, H., Kaczynski, K. 2005. Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 53(6), 1432-40. Kawaguchi, H., Obata, T., Ota, M., Akine, Y., Ito, H., Ikehira, H., Kanno, I., Suhara, T. 2010. Regional heterogeneity and age-related change in sub-regions of internal capsule evaluated by diffusion tensor imaging. Brain Res 1354, 30-9. Kemp, J.M., Powell, T.P. 1971. The structure of the caudate nucleus of the cat: light and electron microscopy. Philos Trans R Soc Lond B Biol Sci 262(845), 383-401. Kinney, H.C., Brody, B.A., Kloman, A.S., Gilles, F.H. 1988. Sequence of central nervous system myelination in human infancy. II. Patterns of myelination in autopsied infants. J Neuropathol Exp Neurol 47(3), 217-34. Latt, J., Nilsson, M., Wirestam, R., Stahlberg, F., Karlsson, N., Johansson, M., Sundgren, P.C., van Westen, D. 2013. Regional values of diffusional kurtosis estimates in the healthy brain. J Magn Reson Imaging 37(3), 610-8. Liu, C., Bammer, R., Acar, B., Moseley, M.E. 2004. Characterizing non-Gaussian diffusion by using generalized diffusion tensors. Magn Reson Med 51(5), 924-37. Madden, D.J., Bennett, I.J., Burzynska, A., Potter, G.G., Chen, N.K., Song, A.W. 2012. Diffusion tensor imaging of cerebral white matter integrity in cognitive aging. Biochim Biophys Acta 1822(3), 386-400. Marner, L., Pakkenberg, B. 2003. Total length of nerve fibers in prefrontal and global white matter of chronic schizophrenics. J Psychiatr Res 37(6), 539-47. Moseley, M. 2002. Diffusion tensor imaging and aging - a review. NMR Biomed 15(7-8), 553-60. Moseley, M., Bammer, R., Illes, J. 2002. Diffusion-tensor imaging of cognitive performance. Brain Cogn 50(3), 396-413. Onodera, S., Hicks, T.P. 2009. A comparative neuroanatomical study of the red nucleus of the cat, macaque and human. PLoS One 4(8), e6623. Ota, M., Obata, T., Akine, Y., Ito, H., Ikehira, H., Asada, T., Suhara, T. 2006. Age-related degeneration of corpus callosum measured with diffusion tensor imaging. Neuroimage 31(4), 1445-52. Pal, D., Trivedi, R., Saksena, S., Yadav, A., Kumar, M., Pandey, C.M., Rathore, R.K., Gupta, R.K. 2011. Quantification of age- and gender-related changes in diffusion tensor imaging indices in deep grey matter of the normal human brain. J Clin Neurosci 18(2), 193-6. Percheron, G., Yelnik, J., Francois, C. 1984. A Golgi analysis of the primate globus pallidus. III. Spatial organization of the striato-pallidal complex. J Comp Neurol 227(2), 214-27. Peters, A., Sethares, C. 2002. Aging and the myelinated fibers in prefrontal cortex and corpus callosum of the monkey. J Comp Neurol 442(3), 277-91. Peters, B.D., Szeszko, P.R., Radua, J., Ikuta, T., Gruner, P., DeRosse, P., Zhang, J.P., Giorgio, A., Qiu, D., Tapert, S.F., Brauer, J., Asato, M.R., Khong, P.L., James, A.C., Gallego, J.A., Malhotra, A.K. 2012. White matter development in adolescence: diffusion tensor imaging and meta-analytic results. Schizophr Bull 38(6), 1308-17. Pfefferbaum, A., Adalsteinsson, E., Rohlfing, T., Sullivan, E.V. 2010. Diffusion tensor imaging of deep gray matter brain structures: effects of age and iron concentration. Neurobiol Aging 31(3), 482-93.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Qiu, D., Tan, L.H., Zhou, K., Khong, P.L. 2008. Diffusion tensor imaging of normal white matter maturation from late childhood to young adulthood: voxel-wise evaluation of mean diffusivity, fractional anisotropy, radial and axial diffusivities, and correlation with reading development. Neuroimage 41(2), 223-32. Rulseh, A.M., Keller, J., Tintera, J., Kozisek, M., Vymazal, J. 2013. Chasing shadows: What determines DTI metrics in gray matter regions? An in vitro and in vivo study. J Magn Reson Imaging. Sasson, E., Doniger, G.M., Pasternak, O., Tarrasch, R., Assaf, Y. 2012. Structural correlates of cognitive domains in normal aging with diffusion tensor imaging. Brain Struct Funct 217(2), 503-15. Schmierer, K., Tozer, D.J., Scaravilli, F., Altmann, D.R., Barker, G.J., Tofts, P.S., Miller, D.H. 2007. Quantitative magnetization transfer imaging in postmortem multiple sclerosis brain. Journal of Magnetic Resonance Imaging 26(1), 41-51. Schmierer, K., Wheeler-Kingshott, C.A.M., Tozer, D.J., Boulby, P.A., Parkes, H.G., Yousry, T.A., Scaravilli, F., Barker, G.J., Tofts, P.S., Miller, D.H. 2008. Quantitative magnetic resonance of postmortem multiple sclerosis brain before and after fixation. Magnetic Resonance in Medicine 59(2), 268-77. Smith, S.M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T.E., Mackay, C.E., Watkins, K.E., Ciccarelli, O., Cader, M.Z., Matthews, P.M., Behrens, T.E. 2006. Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 31(4), 1487-505. Smith, S.M., Johansen-Berg, H., Jenkinson, M., Rueckert, D., Nichols, T.E., Miller, K.L., Robson, M.D., Jones, D.K., Klein, J.C., Bartsch, A.J., Behrens, T.E. 2007. Acquisition and voxelwise analysis of multi-subject diffusion data with tract-based spatial statistics. Nat Protoc 2(3), 499-503. Smith, S.M., Nichols, T.E. 2009. Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage 44(1), 83-98. Song, S.K., Sun, S.W., Ramsbottom, M.J., Chang, C., Russell, J., Cross, A.H. 2002. Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water. Neuroimage 17(3), 1429-36. Stricker, N.H., Schweinsburg, B.C., Delano-Wood, L., Wierenga, C.E., Bangen, K.J., Haaland, K.Y., Frank, L.R., Salmon, D.P., Bondi, M.W. 2009. Decreased white matter integrity in late-myelinating fiber pathways in Alzheimer's disease supports retrogenesis. Neuroimage 45(1), 10-6. Sullivan, E.V., Adalsteinsson, E., Hedehus, M., Ju, C., Moseley, M., Lim, K.O., Pfefferbaum, A. 2001. Equivalent disruption of regional white matter microstructure in ageing healthy men and women. Neuroreport 12(1), 99-104. Sullivan, E.V., Pfefferbaum, A. 2006. Diffusion tensor imaging and aging. Neurosci Biobehav Rev 30(6), 749-61. Tabesh, A., Jensen, J.H., Ardekani, B.A., Helpern, J.A. 2011. Estimation of tensors and tensor-derived measures in diffusional kurtosis imaging. Magn Reson Med 65(3), 823-36. Vaillancourt, D.E., Spraker, M.B., Prodoehl, J., Zhou, X.J., Little, D.M. 2012. Effects of aging on the ventral and dorsal substantia nigra using diffusion tensor imaging. Neurobiol Aging 33(1), 35-42. Veraart, J., Poot, D.H., Van Hecke, W., Blockx, I., Van der Linden, A., Verhoye, M., Sijbers, J. 2011. More accurate estimation of diffusion tensor parameters using diffusion Kurtosis imaging. Magn Reson Med 65(1), 13845. Walhovd, K.B., Fjell, A.M., Reinvang, I., Lundervold, A., Dale, A.M., Eilertsen, D.E., Quinn, B.T., Salat, D., Makris, N., Fischl, B. 2005. Effects of age on volumes of cortex, white matter and subcortical structures. Neurobiol Aging 26(9), 1261-70; discussion 75-8. Wang, Q., Xu, X., Zhang, M. 2010. Normal aging in the basal ganglia evaluated by eigenvalues of diffusion tensor imaging. AJNR Am J Neuroradiol 31(3), 516-20. Yelnik, J., Francois, C., Percheron, G., Heyner, S. 1987. Golgi study of the primate substantia nigra. I. Quantitative morphology and typology of nigral neurons. J Comp Neurol 265(4), 455-72.

ACCEPTED MANUSCRIPT

Table 1. Demographical and cognitive features of participants (n = 58)

Young (n=18) Middle-aged (n=20) Older adults (n=20) 34.94 ± 3.90 46.75 ± 3.63 72.95 ± 5.27 range 25 ~ 40 range 42 ~ 56 range 65 ~ 84 Sex (male: female) 9:9 11 : 9 8 : 12 MMSE ( /30) 30 ± 0 29.8 ± 0.52 27.4 ± 2.23 30 range 28 ~ 30 range 22 ~ 30

AC C

EP

TE D

M AN U

SC

RI PT

Age (years)

ACCEPTED MANUSCRIPT

Table 2 Associations between regional parametric values and age covarying for sex in white matter tracts

anterior limb of internal capsule βstand p value

Posterior Early-myelinated Late-myelinated posterior limb of splenium of fornix internal capsule corpus callosum βstand p value βstand p value βstand p value

genu of corpus callosum βstand p value

RI PT

Anterior

FA F=9.383***, p