In vivo Age-related Changes in Cortical, Subcortical Nuclei, and Subventricular Zone: A Diffusion Tensor Imaging Study

Abstract

Gradual decline in adult stem cells over time at Subventricular zone (SVZ) may contribute to age related changes and neurodegenerative diseases. Study was aimed to evaluate in vivo age-related neuroimaging changes in cortical, subcortical, and SVZ. Sixty four healthy volunteers were recruited from various ongoing studies and subjects were grouped in to pediatric with age less than 18yrs (n=13, females=6) aged 11.8 ± 2.9 yrs, middle aged between 19 to 59 yrs (n=38, female=15) aged 40.4 ± 12.2yrs and elderly above 60yrs (n=13, females=6) aged 65.3 ± 6.0yrs, subgroups. Subjects underwent MRI scanning on a 3T MR scanner and Diffusion Tensor Imaging data with 3d T1TFE data was acquired. DTI was processed using region of interest (ROI) analysis method and the results were observed at p < 0.05 corrected for multiple corrections. Cortical, WM and subcortical GM volumes were extracted using a fully automated method. The cortical volumes (grey, white & whole brain) were least in elderly and highest in pediatric group. Among subgroup analysis following subcortical nuclei significantly differed on; fractional anisotropy (FA): bilateral hippocampus, right pallidum and left amygdale & caudate; Mean diffusivity (MD): bilateral thalamus, right pallidum, left caudate & accumbens; radial diffusivity (RD): bilateral pallidum, left caudate, left thalamus and left accumbens; axial diffusivity (AD): bilateral caudate, bilateral thalamus and left accumbens. The MD, RD and AD values of at SVZ around caudate were also significantly different between subgroups. Study observes patterns of volumetric and DTI changes across normal aging. Alterations in DTI parameters in subcortical and SVZ may indicate changes in neurogenic region in aging process; however longitudinal studies are required for further validation.

Brain is a complex structure which shows regular, nonlinear changes throughout aging process [1]. During aging several changes occur from molecular to morphological level involving physical, chemical, and biological changes. Different tissues may be more or less susceptible to age-induced changes [2, 3]. Also as age progress there is decline in the regenerative capacity resulting in functional deterioration and poor repair. Recently, it is observed that even in adults the neurogenesis is evident, with endogenous stem cell activity. The Subventricular zone (SVZ) is one of the neuro-genic regions in brain. SVZ contains multi potent neural stem cells, lies immediately beneath the ependymal layer of the lateral ventricles, and is separated from caudate nucleus by a layer of myelin. These stem cells gradually decrease in number over aging and may contribute to age related physiological changes [4, 5, 6]. Thus study of SVZ could provide important information about fundamental changes associated with the brain aging.

Diffusion Tensor Imaging (DTI) is a relatively new MRI technique where Magnetic resonance Imaging (MRI) contrast is created by relative motility of protons due to Brownian motion. This permits the study of the brain’s microstructure and allows characterization of white matter abnormalities [7]. Integrity of cerebral white matter (WM) and also about the brain tissue integrity can be assessed using various DTI parameters like fractional anisotropy (FA), Mean Diffusivity (MD), radial diffusivity (RD) and axial diffusivity (AD). The changes in DTI parameters values during lifespan are presumed to be due to changes in cerebral myelin levels and myelin packing [8, 9]. Currently the literature examining the age related DTI parameter changes in subcortical structures and also evaluating the tissue integrity at the SVZ zone to understand the age related changes associated with the stem-cell load/density is very limited. The present study was undertaken to evaluate the global and regional brain volume, structural and diffusion changes in subcortical structures with age using the DTI and MRI volumetry, specifically looking at age related alterations in the SVZ.

MATERIALS AND METHODS

Study site and subjects

The subjects were healthy volunteers with normal activities of daily living, and normal global cognitive function. The subjects were excluded if they had any of these manifestations: 1) major medical illnesses, e.g., diabetes (not stabilized), obstructive pulmonary disease, or asthma; hematologic/oncologic disorders; endocrine, or cardiovascular system disease; newly treated hypothyroidism; 2) co morbidity of primary psychiatric or neurological disorders (e.g., schizophrenia, major depression, stroke, Parkinson disease, seizure disorder, head injury with loss of consciousness) and any other significant mental or neurological disorder; 3) known or suspected history of alcoholism or drug dependence and abuse during lifetime; 4) MRI evidence of focal parenchymal abnormalities and 5) MRI evidence of tumors/neoplasm. The study was conducted at National Institute of Mental Health And Neurosciences (NIMHANS) Bangalore, India. During the study period, 64 healthy subjects were selected on basis of inclusion and exclusion criteria as stated above.

Data Collection and Management

Healthy volunteers recruited from various ongoing projects during the period of 2009 to 2011. The selected volunteers MRI data was acquired on Phillips Achieva 3 Tesla MRI system with conventional T1 weighted, T2 weighted and Fluid attenuation and inversion recovery (FLAIR) images acquired along with the T1-weighted magnetization prepared rapid gradient echo (MP-RAGE) sequence covering the whole brain (Orientation-sagittal; FOV-256 x 256 x 155 mm; voxel size- 1mm x 1mm x 1mm; slice thickness-1 mm; acquisition matrix-256 x 256; TR=8.1 ms; TE=3.7 ms; flip angle=8 degree; sense factor=3.5; total scan time= 7 min 41 sec.) and DTI (Number of directions-32; orientation-transverse; FOV-224mm x 224 mm x 144 mm; voxel size-2mm x 2mm x 2mm; slice thickness-2 mm; acquisition matrix-112 x 109; TR=5661 ms; TE=59 ms; flip angle=90 degree; EPI factor=47; b-value - 0, 1000; sense factor-3.5; total scan time=13 min 56 sec).

All image processing was performed using FMRIB Statistical Laboratory (FSL) 4.1 software [10] using methods similar to those described in previous studies [11]. Briefly Image distortions induced by eddy currents and head motion in the DTI data were corrected by applying a full affine (mutual information cost function) alignment of each image to the mean no diffusion-weighted image. After distortion corrections, DTI data were averaged and concatenated. A diffusion single tensor model was fit at each voxel, generating FA, MD, RD, AD maps. The FA maps created were then registered to brain-extracted whole-brain volumes from T1-weighted images using a full affine (correlation ratio cost function) alignment with nearest-neighbor resampling. The calculated transformation matrix was then applied to the MD maps with identical resampling options. For each subject, the co registered FA map was superimposed to the original T1-weighted volume, and the resulting images were visually assessed accuracy of registration.

We were interested in the regions separating the lateral ventricles from the subcortical structures bordering the ventricles themselves. Consequently, we identified in each subject the following structures: lateral ventricles, caudate, thalamus, using the FSL Integrated Registration and Segmentation Tool (FIRST) 1.2[12]. Subsequently, we further identified the intersection regions between lateral ventricles and two neighboring structures: caudate, thalamus. For the purpose of the present study, we identified the region of intersection between the caudate, thalamus and the lateral ventricles as the neuro-genic SVZ. The intersection regions included the voxels that were comprised in a laminar volume surrounding the ventricle with a thickness of 2 mm. This volume was defined by the surface of separation between the ventricle itself and the corresponding neighboring structure, and by a second surface, parallel but external to the first surface.

For each subject and each hemisphere, we calculated the volumes of the two boundary regions, defined as above. Before statistical comparison, individual volume values were multiplied by a normalization factor obtained with the FSL Structural brain change analysis, for estimating brain atrophy (SIENAX) tool [13] from the corresponding T1-weighted image. These segmented areas defined the region of interest (ROI) from where mean and standard deviation of DTI parameters were calculated for each individual and for each hemisphere. For each subject, the results of ROI segmentation and the co registered FA map were all superimposed to the original T1-weighted volume, and the resulting images were visually assessed and confirmed for accuracy by a trained neuro radiologist to exclude misregistration or erroneous ROI identification.

Statistics

The statistics of the data were achieved by SPSS statistical software version 15. For better understanding age related changes, it was grouped into pediatric (less than ≤18yrs), middle aged (19–59yrs), & elderly (≥60yrs) aged. Data expressed using descriptive statistics such as mean, standard deviation for continuous variables and frequency, percentage for categorical variables. Among these groups’ brain volumes and DTI changes in right and left side of cortical, subcortical, and SVZ regions were analyzed. Multivariate Analysis of Covariance (MANCOVA) test was used to compare between the subgroups for all the volume data with correction for sex and intracranial volume as covariates in the statistical design. Similarly MANCOVA test was used to compare between the subgroups, for all the DTI data with correction for sex and intracranial volume as covariates in the statistical design. Pearson’s correlation test was used to correlate between the age and the DTI parameters. Statistical significance level of p<0.05 was considered as significant.

RESULTS

Of total 64 healthy voluntaries, the pediatric age accounts for 13 subjects with mean age of 11.8 ± 2.9 yrs (6 females), the middle age subjects were 38 with mean age of 40.4 ± 12.2yrs (females 15), and elderly included 13 subjects with age of and 65.3 ± 6.0yrs (females).

Volumetry

On MANCOVA analysis the design was statistically significant with wilk’s lambda =0.023 with observed power of 0.985. These results of cortical, subcortical and SVZ volumes were tabulated in Table-1. The GM (p<0.001) & WM volumes (p<0.001) showed significant differences between the three subgroups; lowest volumes were observed in elderly group while greatest volumes in the pediatric group. Among the subcortical volumes bilateral hippocampus (left p=0.01, right p=0.024), left thalamus (p=0.17), left caudate (0.013) and brain stem (p=0.01) differed significantly between the subgroups, with highest volumes seen in middle age group and least in pediatric group.

Table 1.

The socio-demographic and volumetric matrices of whole brain and subcortical nuclei

Volumes		Pediatric (≤18yrs)	Middle age (19–59yrs)	Elderly age (≥60yrs)	Tests of Between-Subjects Effects
Age (Mean)		11.76 ± 2.9	40.44 ± 12.2	65.30 ± 6.0
Sex (M/F)		7/6	23/15	7/6
Grey matter vol. (mm3)		1029.1 ± 71.8	835.4 ± 60.3	789.8 ± 29.5	<0.001
White matter vol.(mm3)		767.8 ± 11.9	758.3 ± 42.0	719.0 ± 30.0	0.001
Total brain vol. (mm3)		1796.9 ± 71.0	1593.7 ± 89.6	1508.8 ± 53.6	<0.001
Hippocampus (mm3)	Left	34.31 ±.4.71	38.13 ± 5.39	34.69 ± 4.59	0.011
	Right	34.45 ± 3.94	38.39 ± 4.96	34.39 ± 6.21	0.014
Amygdale (mm3)	Left	10.81 ± 1.75	12.03 ± 2.13	10.88 ± 2.51	0.137
	Right	9.49 ± 2.49	11.96 ± 2.82	10.49 ± 3.12	0.063
Caudate (mm3)	Left	35.78 ±3.37	30.58 ± 6.71	27.19 ± 2.95	0.013
	Right	36.44 ± 4.59	32.96 ± 5.71	29.56 ± 3.26	0.321
Putamen (mm3)	Left	48.32 ±4.19	45.73 ± 7.46	41.09 ± 5.32	0.101
	Right	47.17 ±5.75	46.23 ± 7.84	41.23 ± 4.95	0.147
Thalamus (mm3)	Left	71.71 ±4.14	72.76 ± 9.26	65.22 ± 5.48	0.017
	Right	68.50 ±4.84	69.69 ± 8.96	63.05 ± 6.06	0.056
Accumbens (mm3)	Left	5.03 ±0.72	4.89 ±1.03	4.62 ± 1.22	0.845
	Right	3.42 ±.1.59	3.77 ±1.08	3.12 ± 0.78	0.234
Sub-ventricular zone (mm3)	Left	6.92 ± 0.93	6.73 ± 1.17	6.52 ± 1.1	0.491
	Right	6.85 ± 0.69	6.55 ± 1.26	6.14 ± 0.94	0.175
Brain Stem (mm3)		174.70 ± 12.03	212.24 ± 30.85	193.30 ± 18.37	0.001

The statistical comparisons for sub-cortical volumes were performed after correcting for the Intra-cranial volume.

Volumes of other structures namely bilateral putamen, bilateral amygdale, bilateral accumbens, bilateral pallidum, right caudate, right thalamus, and SVZ showed no significant difference among various subgroups. Although statistically not significant (p>0.05), we observed trends of higher volumes of caudate, putamen, thalamus and accumbens structures except hippocampus and amygdale in pediatric group with lower volumes in the elderly group.

On correlation analysis grey matter volume (r=−0.376, p=<0.001), white matter volume (r=−0.341, p=0.006) and Intracranial volume (r=−0.711, p=<0.001) showed negative correlation with aging. Bilateral caudate (Right r=−0.558, p<0.001, Left r=−0.475, p<0.001), bilateral putamen (Right side r=−0.438, p<0.001, Left r=−0.369, p=0.003), bilateral thalamus (Right side r=−0.343, p=0.006, Left r=−0.369, p=0.003) and left accumbens (r=−0.257, p=0.04) showed negative correlation with aging. Right side SVZ volume (r=−0.272, p=0.030) showed negative correlation with aging.

DTI parameter

On MANCOVA analysis of individual DTI parameters the design was not statistically significant with wilk’s lambda =0.11 with observed power of 0.384. These results of DTI parameters are tabulated in Table-2.

Table 2.

The DTI parameters in the different subgroups of SVZ and at the sub-cortical nuclei

DTI parameters		Pediatric (≤18yrs)	Middle age (19–59yrs)	Elderly age (≥60yrs)	P-value	F-value
Sub ventricular zone
FA	Left	0.1385 ±.0218	0.1423 ±.0334	0.1214 ±.0099	0.045	2.9
	Right	0.1442 ±.0301	0.1498 ±.0462	0.1206 ±.0163	0.062	2.7
MD	Left	0.0015 ±.0003	0.0016 ±.0003	0.0018 ±.0002	0.037	5.1
	Right	0.0018 ±.0003	0.0018 ±.0003	0.0020 ±.0002	0.511	1.4
RD	Left	0.0014 ±.0003	0.0015 ±.0003	0.0018 ±.0002	0.024	5.6
	Right	0.0016 ±.0003	0.0017 ±.0003	0.0019 ±.0002	0.371	1.7
AD	Left	0.0017 ±.0003	0.0018 ±.0003	0.0020 ±.0002	0.089	4.4
	Right	0.0020 ±.0003	0.0021 ±.0004	0.0022 ±.0002	0.814	0.8
Subcortical nuclei
Pallidum
FA	Left	0.296206 ±.024	0.306834 ±.032	0.319468±.0455	.226	1.5
	Right	0.298077 ±.03592	0.33495 ±.0349	0.342714±.0389	.004	6.1
MD	Left	0.000772 ±.00002	0.000734 ±.00002	0.00075±.00009	.061	2.9
	Right	0.000759 ±.00003	0.0007199 ±.00003	0.00071±.00006	.016	4.4
RD	Left	0.000651 ±.00003	0.0006132 ±.00003	0.00062±.00007	.048	3.2
	Right	0.0006353±.00003	0.0005831 ±.00003	0.00057±.00006	.001	7.9
AD	Left	0.0010014±.00003	0.0009926 ±.00003	0.000976±.0001	.162	1.8
	Right	0.0009990±.00005	0.0009829 ±.00005	0.000982±.0001	.720	0.3
Hippocampus
FA	Left	0.1479679 ±.0052	0.1508446 ±.0104	0.1408349 ±.011	.015	4.5
	Right	0.1600817±.00913	0.1530685 ±.01353	0.141758±.0081	.001	8.0
MD	Left	0.0010655±.00005	0.0010646 ±.00007	0.00106 ±.00007	.978	0.02
	Right	0.0010610±.00005	0.0010635±.00007	0.00106 ±.00007	.927	0.9
RD	Left	0.0009808±.00004	0.0009765 ±.00007	0.00097±.00006	.883	0.1
	Right	0.000972 ±.00005	0.0009687 ±.00007	0.000973 ±.0009	.977	0.02
AD	Left	0.0012243±.00006	0.0012502 ±.00008	0.00126±.00009	.602	0.5
	Right	0.0012407±.00007	0.0012347 ±.00010	0.00124 ±.00010	.971	0.02
Caudate
FA	Left	0.18099 ±.016	0.20802±.0363	0.19540±.0165	.020	4.1
	Right	0.16436 ±.01686	0.19023 ±.0469	0.19441 ±.0236	.068	2.8
MD	Left	0.000823 ±.00005	0.0008564 ±.00009	0.00093 ±.0001	.006	5.5
	Right	0.0008843±.00007	0.0009639 ±.00015	.001024 ±.00021	.083	2.5
RD	Left	0.0007481±.00004	0.0007675 ±.00009	0.00084±.00011	.008	5.2
	Right	0.0008116±.00006	0.0008717 ±.00014	0.000920 ±.0002	.172	1.8
AD	Left	0.0009704±.00004	0.0010406 ±.0010	0.001098 ±.0001	.020	4.2
	Right	0.0010165±.00007	0.0011458 ±.00017	0.001206±.0002	.025	3.9
Amygdale
FA	Left	0.18 00322 ±.0120	0.1700132 ±.013	0.17 0034±.0140	.035	3.5
	Right	0.1758232±.01221	0.1804410 ±.01852	0.179858 ±.0154	.798	0.2
MD	Left	0.0008638±.00002	0.0008490 ±.00005	0.00086 ±.00004	.441	0.8
	Right	0.0008671±.00003	0.0008550 ±.00005	0.00086±.00006	.512	0.6
RD	Left	0.0007824±.00001	0.0007709 ±.00004	0.00077±.00004	.703	0.3
	Right	0.0007718±.00003	0.0007761 ±.00005	0.00077±.00005	.556	0.5
AD	Left	0.0010266±.00003	0.0010174 ±.00006	0.00105±.00005	.247	1.4
	Right	0.0010170±.00003	0.0010097±.00004	0.00104±.00006	.184	1.7
Thalamus
FA	Left	0.29947±.0258	0.30723 ±.0229	0.30233±.01517	.453	0.8
	Right	0.2960± 0.027	0.30592 ±.0266	0.29348±.01242	.157	1.8
MD	Left	0.00085±.00004	0.0008804 ±.00004	0.00092 ±.0001	.014	4.6
	Right	0.0008±.00005	0.00089±.00006	0.00094±.00013	.038	3.4
RD	Left	0.0007 ±.00004	0.0007386±.00005	0.00077±.00009	.047	3.2
	Right	0.0007±.00005	0.0007526 ±.00006	0.000799±.0001	.052	3.1
AD	Left	0.0011 ±.00003	0.0011634 ±.00005	0.001212±.0001	.004	6.1
	Right	0.0011 ±.00004	0.0012±.00006	0.00121 ±.00015	.035	3.5
Accumbens
FA	Left	0.17787±.0141	0.19539 ±.0258	0.19215±.0337	.115	2.2
	Right	0.176216 ±.022	0.19135 ±.0270	0.1811634±.03235	.112	2.2
MD	Left	0.0008±.00003	0.00073±.00002	0.0007497±.00002	.000	12
	Right	0.0007±.00003	0.00080±.00007	0.0007652 ±.00006	.317	1.1
RD	Left	0.0007±.00002	0.00067±.00002	0.0006594 ±.00003	.000	13
	Right	0.0007±.00003	0.00072±.00007	0.0006901 ±.00004	.276	1.3
AD	Left	0.00092±00003	0.00085±00006	0.0009116 ±.00007	.003	6.2
	Right	0.0009±.00004	0.00095±.00008	0.0009126 ±.00008	.292	1.2

FA: Fractional Anisotropy, MD: Mean Diffusivity, RD: Radial Diffusivity, AD: Axial Diffusivity

Correlation analysis of subcortical structures DTI parameters with the age showed a positive correlation between FA and, left pallidum (r=0.267, p=0.036), putamen (r=0.716, p<0.001) & right caudate nucleus (r=0.327, p=0.009) and negative correlation was observed between bilateral hippocampi [right (r=−0.599, p=0.001), left (r=−0.307, p=0.015)], right thalamus (r=−0.27, p=0.028). MD with aging showed positive correlation with bilateral thalamus [(Right, r=0.27, p=0.034), left (r=0.316, p=0.012)] & bilateral caudate [(Right r=0.313, p=0.013), Left (r=0.313, p=0.013)] and negative correlation was seen in left putamen (r=−0.486, p=0.001) and right pallidum (r=−0.42, p=0.001).

RD showed positive correlation with age in bilateral thalamus [Right r=0.255, p=0.45) Left (r=0.28, p=0.028)], left caudate (r=0.298, p=0.018) and accumbens(r=0.468, p<0.001) and negatively correlation in the bilateral pallidum (Right r=−0.485, p<0.001) Left (r=−0.0265, p=0.037). AD in bilateral thalamus [Right (r=0.295, p=0.020), Left (r=0.334, p=0.008)] and caudate (Right (r=0.372, p=0.003) Left (r=0.301 p=0.017)) showed positive correlation with age. Bilateral SVZ MD (r=0.275, p=0.029), RD (r=0.291, p=0.020) and SVZ AD (r=0.258, p=0.041) also revealed positive correlation with age.

DISCUSSION

In our study we evaluated age related changes in various brain structures using the MRI techniques of DTI and volumetry. The volumes of cortical gray and white matter, sub cortical bilateral hippocampus, left caudate, left thalamus and brain stem were significantly different between the various subgroups (p<0.05) with higher volumes observed in the pediatric group and lower volumes in middle and elderly age groups. We observed significant negative correlation between the age and volumes of WM & GM, bilateral caudate, putamen, thalamus and right sided SVZ (p<0.05). DTI measures also revealed significant correlation with age. The FA in bilateral hippocampus and right thalamus showed significant negative correlation with age while diffusivity indices showed significant positive correlations with age in the bilateral thalamus (in MD, RD and AD), bilateral caudate (in MD, AD), left caudate (in RD), left accumbens (in RD) and left SVZ (in MD, RD and AD) (p<0.05).

Normal aging is associated with gradual loss of brain tissue and it has been studied previously in various MRI investigations using the techniques of VBM, volumetry and DTI. GM reduction starts in early adulthood and continues linearly there after throughout life [1, 14–16]. WM volume change shows non linear relationship and it initially increases and then starts declining after 5^th decade of life [1, 16–18]. Cerebral cortex has been reported to be more affected by the age related GM loss compared to sub cortical structures [19]. However results of previous studies on age related changes in the sub cortical nuclei have not been consistent. For example thalamic volume loss with age has been reported in many studies [20, 21] however on the contrary few studies have found no such relationship between age and thalamus [19, 22]. Similar observations were also noted for the other sub cortical structures. Various reasons for these discrepancies in the results of various past studies include different study populations, different age groups used for study, use of different MRI scanners, variable imaging protocols and differences in analysis methods. In our study we noted continuous linear decrease in GM volume and whole brain volume from pediatric to old age. Volumes of multiple subcortical structures showed significant negative correlation with age (p<0.05). Cortical volumes were negatively correlating with age (p<0.05). Similar results were also noted in previous studies. Studies pertaining to cortical volume changes with aging had similar findings except for whole brain, where until the 20 years volume increases and later decline has been reported [23, 24].

Cherubini et al studied physiological aging changes of deep grey matter nuclei by simultaneously measuring volumes, iron deposition, microstructural damage using DTI in various sub cortical structures [25]. Authors reported significant correlation of volumetry with age in thalamus, putamen and caudate nucleus but did not observe similar correlation in the hippocampus, amygdale, pallidum, or accumbens [25]. In thalamus authors [25] found that volume and MD showed significant correlation with age while they did not show any correlation with each other which was suggested that volume loss and DTI metrics are measures of different microscopic degenerative processes associated with normal aging. In putamen and caudate nucleus also similar results were noted however DTI metrics and volumetry results showed negative correlation with each other. This difference between the thalamus and striatum might be related to the greater iron deposition in striatum compared to the thalamus. Similar results were noted in our study and bilateral caudate nucleus, putamen and thalamus volumes revealed negative correlation with the age. However, in addition we also noted negative correlation of age with the left accumbens volume.

We observed positive correlation between age and FA in the left pallidum, putamen and right caudate nucleus and negative correlation in the bilateral hippocampi and right thalamus. MD with aging showed positive correlation in the bilateral thalamus and caudate and negative correlation in the left putamen and right pallidum. RD showed positive correlation with age in bilateral thalamus, left caudate and accumbens and negatively correlation in the bilateral pallidum. AD in bilateral thalamus and caudate nucleus showed positive correlation with age.

Interpretation of DTI metrics in deep gray matter structures like basal ganglia is different and more complex when compared to WM. Previously Cherubini et al reported positive correlation between age and MD of thalamus and striatum [25]. No correlation between age and FA was noted. No correlation of age was observed with the DTI metrics of other sub cortical structures. Wang et al studied DTI measurements in the basal ganglia structures and found significant reduction of MD with aging the head of caudate nucleus and putamen [26]. Increase in FA with aging was found in the putamen [26]. RD showed a significant age-related decrease in caudate and putamen but no significant changes of RD were found in globuspallidus [26]. Pal et al used quadratic regression model to study the age related changes in the deep gray matter nuclei and found that FA showed increase with age while MD decreased with age in the deep sub cortical nuclei [27]. Pfefferbaum et al also reported increase in FA value in putamen and caudate nucleus with age while thalamus did not show any significant change with age. MD was noted to increase with age in globuspallidus and remaining structures revealed no significant differences [28]. Study of AD showed higher values in all the sub cortical structures in older age group subjects [29]. Increase in MD values in these nuclei suggest neurodegenerative processes like neuronal loss, widening of extracellular space and shrinkage of neuronal soma changes with aging. It could also be related to increased iron deposition as supported by increased R2* values with age and its correlation with diffusivity measurement [25]. Metal deposition may cause tissue damage by generation of free radicals as well by affecting the blood flow.

In our study we observed negative correlation between total WM and age. We also observed that WM volume was not significantly different between the paediatric and middle age groups.Courchesne et al reported steep increase in the WM volume followed in children followed by slower increase till the 4^th decade of life when it reaches the plateau [30]. This phase is followed by decrease in WM volume among older individuals. Pfefferbaum et al in there study reported no significant change in WM volume up to 5^th decade of life [17]. Giorgio used a VBM-style analysis and found linear negative association between WM volume and age especially in external capsule, internal capsule, anterior thalamic radiations, cerebral peduncle and cerebellum [31]. They also found areas with significant nonlinear (quadratic) relationships between age and WM volume of superior corona radiata and at superior longitudinal fascicle.

Although statistically not significant (p>0.05), we observed trends of higher volumes of the SVZ on the left side and the both side volumes decreases with aging. These findings are interesting, and probably suggest a decrease in proportion of larger volume components (multi-potent neurons and myelinated axons) with aging. Our data of DTI indices’ on SVZ suggests that the anisotropy value decreases and diffusivity value increases with aging. MD, RD and AD of SVZ revealed positive correlation with age. Right SVZ volume also showed negative correlation with the age. A previous DTI study on SVZ in Alzheimer’s disease had reported significantly higher diffusivity values among subjects with Alzheimer’s disease and mild cognitive impaired patients as compared to healthy controls. These findings were suggestive of neurodegenerative process in form of micro-structural damage in the neuro-genic areas of brain [25]. The decrease in volume and increased diffusivity with decreased FA in SVZ can be partly explained by loss of the multi-potent neurons and other micro-structural changes in this vital area of brain. There are no substantial studies on SVZ and evolution of DTI and volumetric changes with normal aging. Such information may be useful for study of neurodegenerative changes and explore the role of neuro-genic areas of brain in there pathogenesis. Limitations of study were collection of MRI data of healthy controls that were participated in various ongoing projects and not prospectively collected for the current study purpose with no data about subclinical neuropsychiatric conditions. Another limitation is modest sample size used in this study. Ideally longitudinal study is required for evaluation of age related changes as cross sectional design has its limitations. However the data was acquired with the higher MRI magnet of 3T with better tissue contrast and advanced data processing methods were applied with careful manual selection of the SVZ ROI.

Conclusion

The volume and DTI changes in normal age pattern illustrate the process of normal brain maturation and aging. Profiles of DTI indices and volume changes in SVZ in our study may be useful in interpreting deviations from this physiological profile as risk for neuro-degenerative diseases. Further study observations of age related alteration of DTI parameters in sub cortical structures and at SVZ may indicate the changes in stem cell distribution across the aging process, however longitudinal studies are required for further validation.

Figure 1.

Global, Sub-cortical and SVZ volume changes with aging. A. The various whole brain volumes and Intra-cranial volume corrected brain stem volume and bilateral amygdale volumes on Y axis plotted against the age in X-axis with corresponding marginal box-plots on the left side of each graphs. B. The Intra-cranial volume corrected volumes of bilateral hippocampi, bilateral thalami and bilateral SVZ regions on Y axis plotted against the age in X-axis with corresponding marginal box-plots on the left side of each graphs.

Figure 2.

Subcortical and SVZ regional DTI changes with aging. A. The DTI Fractional Anisotropy values of bilateral hippocampi, bilateral pallidum, left SVZ and DTI- Mean Diffusivity of right SVZ on the Y-axis plotted against age in years on the X-axis with marginal box-plots. B. The DTI Mean Diffusivity values of Mean diffusivity of bilateral caudate, bilateral pallidum and bilateral thalami on the Y-axis plotted against age in years on the X-axis with marginal box-plots.