A voxel-based morphometric magnetic resonance imaging study of the brain detects age-related gray matter volume changes in healthy subjects of 21–45 years old

Abstract

Previous and more recent work of analyzing structural changes in the brain suggest that certain brain regions such as the frontal lobe are among the brain regions profoundly affected by the aging process across males and females. Also, a unified model of structural changes in a normally aging brain is still lacking. The present study investigated age-related structural brain changes in gray matter from young to early middle-age adulthood for males and females. Magnetic resonance images of 215 normal and healthy participants between the ages of 21–45 years were acquired. Changes in gray matter were assessed using voxel-based morphometry and gray matter volumetric analysis. The results showed significant decrease in gray matter volume between the youngest and oldest groups in the following brain regions: frontal, temporal, and parietal lobes. Grey matter loss in the frontal lobe was among the most widespread of all brain regions across the comparison groups that showed significant age-related changes in grey matter for both males and females. This work provides a unique pattern of age-related decline of normal and healthy adult males and females that can aid in the future development of a unified model of normal brain aging.

Introduction

Gray matter (GM) volume of the adult human brain has been shown to diminish gradually with age. However, it is not clear if age-related changes occur uniformly across all cortical regions. Neurodegenerative diseases can reduce volume of GM regionally and globally long before they become symptomatic.^1,2 Therefore, there exists an increased and crucial need for a full understanding of how the brain ages normally. Early detection of GM volume loss can help in distinguishing effects of certain neuropathology on the brain from normal aging. It appears that the pattern of GM volume reduction with age can be relatively easy to detect in elderly subjects when compared to young healthy subjects. However, for individuals who have already lost a substantial volume of GM, the implementation of any intervention, which can slow the accelerated volume loss of the brain, is obviously a late undertaking. Moreover, normal baseline values of GM for both males and females will certainly help better understand and distinguish the effects of certain neuropathology and neurodegenerative diseases on brain structure, and increased knowledge on brain shrinkage and associated patterns in normal aging can lead to more superior understanding of its causes, and may lead to interventions that can lessen the decline of brain functions associated with aging.³

There is plausible and convincing evidence from both post mortem and in vivo studies that the brain shrinks with age, but it remains the case that there is no clear quantification of age-related atrophy patterns. It still remains unclear whether there exist predictable patterns of age-related changes in the brain or whether such age-related changes are idiosyncratic.⁴ Moreover, specificity of brain tissue class loss is crucial towards understanding global brain shrinkage patterns. This issue can be addressed by segmenting magnetic resonance imaging (MRI) brain data into GM, white matter (WM), and cerebrospinal fluid (CSF): then using the voxel-based morphometry (VBM) technique to study and report regional changes in any tissue class, as well as using a volumetric analysis approach to see global volume changes with respect to tissue class.

Previous studies using VBM have reported global decrease in GM volume but with quantitative differences.^5–8 More specifically, previous studies^9,10 showed a significant decline in cortical GM volume, while Liu et al.¹¹ found that global brain volume loss was around 0.385% per year in normal subjects over age 54 years in an MRI-based study. Another study¹² provided similar results for whole brain in normal spouses of Alzheimer's Disease (AD) patients (∼0.33% per year). Furthermore, Coffey et al.¹³ found whole hemisphere loss of 2.79 cm³ per year, and Tisserand et al.¹⁴ also observed similar results.

Diffuse reductions of GM volume correlated with age in the frontal, parietal and temporal lobes, cerebellum and basal ganglia according to Smith et al.³ Also, accelerated brain atrophy was seen in the age range 35–54 years and cross-sectional analysis (i.e. VBM) revealed a significant association between age and reduction in brain volumes.¹¹ Moreover, a voxel-based morphometry for ages between 18–79 years⁴ revealed a global age-related reduction in GM and reductions in the superior parietal gyri, insula, and cingulate sulci.

In a longitudinal study of brain volume changes in normal brain, Scahill et al. concluded that there existed a significant age-related decrease in global and regional brain volume, which may be related to nonlinear acceleration in atrophy rate in aging, and that better understanding of such a process may help differ structural changes in the brain due to normal aging from changes due to neurodegenerative diseases.¹²

Detection and quantification of subtle global and regional GM volume loss in earlier stages of adult life would allow identification of persons at high risk of more substantial accelerated volume loss with age and offer benefits of preventive interventions. Although we are able to measure regional brain volumes quite accurately using voxel-based morphometric MRI, it remains unclear whether there exist any pattern of age-related changes in the brain in adult subjects aged 21–45 years. Studies that suggest that there is no age-related volume loss in healthy subjects in that age span are supported by studies which have shown that blood flow parameters in basal cerebral arteries are fairly stable up to the age of 40–42 years, but the flow velocity starts to decrease and the flow impedance starts to increase definitely from this point in age.^15,16

A lot, if not most, of previous work focused on associated volumetric-brain decline from early adulthood/middle adulthood to elderly age. In this study we examine the effect of normal aging on global brain volumes across an age range from early adulthood to early middle age (i.e. 21–45 years) in both males and females respectively. We performed a cross-sectional study using voxel-based morphometry on 215 normal T1-weighted MRI brain images, separately for males and females, in order to examine any reductions in GM with aging in this age range. We investigated normal brain images across ages 21–45 years and provide a pattern of regional global GM aging from age 21–45 years for males and females in order to help future work towards a unified model of normal brain aging. Furthermore we performed a volumetric analysis of GM volume to investigate global reductions in GM volume, if any, in females and males respectively. We hypothesized that greater regional and global effects of GM decline would be most evident between age groups 21–25 and 41–45 years for both males and females.

Methods

Subjects

All subjects were participants in a project lead by the Biomedical Image Analysis Group at Imperial College London.¹⁷ In that project more than 500 MRI images from normal and healthy subjects were acquired. The cohort included T1-weighted MRI images of normal and healthy subjects. Inclusion criteria were: minimum age of 21 and maximum age of 45 years, neurological and cognitive normality at time of MRI scan. A total of 215 T1-weighted MRI brain images of normal and healthy subjects were acquired from the brain development IXI dataset.¹⁷ The acquired brain images included randomly selected T1-weighted brain images of both males and females to create equal subgroups of data across age groups as follows: 21–25, 26–30, 31–35, 36–40, 41–45 years (see Table 1).

Table 1.

Mean and median age, number of females and males for each age group.

Age group, years	Mean age (n = 43)	Median age (n = 43)	Number of males	Number of females
21–25	24	24	20	23
26–30	28	28	21	22
31–35	33	34	22	21
36–40	38	38	24	19
41–45	42	42	22	21

MRI protocol

The T1-weighted images from brain-development.org¹⁷ were acquired at three different hospitals in London. Hammersmith Hospital used a Philips Medical Systems Intera 3T MRI system (TR = 9.6, TE = 4.6, number of phase encoding steps = 208, echo train length = 208, reconstruction diameter = 240, acquisition matrix = 208 × 208, flip angle = 8). The Institute of Psychiatry used a 1.5T GE MRI system, and Guy’s Hospital used a Philips Medical Systems Gyroscan Intera 1.5T MRI system (TR = 9.8, TE = 4.6, number of phase encoding steps = 192, echo train length = 0, reconstruction diameter = 240, flip angle = 8).

VBM and brain volume

VBM was performed using SPM8 (Wellcome Trust Centre for Neuroimaging, University College London, London, UK) and VBM8 toolbox (C. Gaser, Department of Psychiatry, University of Jena, Germany; http://dbm.neuro.uni-jena.de/vbm8/). All 215 T1-weighted MRI brain images were reoriented so that each respective image was centralized at the anterior commissure to optimize normalization. The images were segmented into GM, WM, and CSF, and normalized to standard space based on the Montreal Neurological Institute (MNI) template. A quality check was performed manually on all images to make sure that segmented images reflected segmentation category with proper and standard orientation. Then for the purposes of this study, the segmented GM data was smoothed using an 8 mm FWHM isotropic Gaussian kernel for later statistical data analysis.

For the VBM statistical analysis a second level analysis was performed in SPM8 (software from the Wellcome Department of Imaging Neuroscience, London, UK) for males and females respectively. The statistical analysis was performed on the smoothed GM data. Multiple regression analysis was performed with age and gender as covariates. Then age-group-specific statistical parametric maps of regional correlated changes in GM volume for age and gender were determined using a one-way analysis of variance (ANOVA) performed on the age groups provided in Table 1. The threshold for statistical significance in all VBM analysis was set to p < 0.05 with family wise error (FWE) correction, and voxel level and extent threshold of 10 voxels. Furthermore, the volume (ml) of segmented GM images was then calculated using Get_Totals (www0.cs.ucl.ac.uk/staff/g.ridgway/vbm/get_totals.m) SPM8 script. Then SPSS (IBM SPSS Statistics for Windows Version 22.0) was used to detect and exclude any outlier GM volumes for both genders (none found in this study), followed by a multiple linear regression where GM volume (ml) was the dependent variable and gender along with age as the independent variables.

Results

There was no significant reduction in GM between the following age groups in both males and females: 21–25 > 31–35, 21–25 > 26–30, 26–30 > 36–40, 26–30 > 31–35, 31–35 > 41–45, 31–35 > 36–40, and 36–40 > 41–45 years. The most significant reduction in GM was found in the following age-group contrasts for males and females respectively: 21–25 > 41–45, 21–25 > 36–40, and 26–30 > 41–45 years.

Comparison of the 21–25 and 41–45 age groups showed the most widespread age-related reduction in GM for both male and female groups. These reductions included the left frontal lobe, left temporal lobe, right frontal lobe, right temporal lobe, and right parietal lobe for males (Figure 1, Table 2) and for females (Figure 2, Table 3). Age-related reductions in GM between age groups 21–25 and 36–40 years were found in the right frontal lobe for both males (only region of GM loss: inferior frontal gyrus) and females while females showed more areas of regional GM loss (Table 4). Also, significant reductions in GM volume due to normal aging were exhibited between 26–30 and 41–45 years age groups for both males and females. In this comparison both males and females showed significant GM loss in the left frontal lobe while females showed more regional GM loss (Tables 5 and 6). Tables 7 and 8 provide the results of multiple linear regression with age and gender as covariates. More specifically Table 7 provides the brain regions that showed decrease in GM with increasing age (negative correlation), while Table 8 provides the brain regions that showed larger GM volume for females compared to males (females > males).

Figure 1.

Gray matter (GM) regional changes rendered onto a single subject brain for the comparison between 21–25 and 41–45 years age groups (21–25 > 41–45) for males. Four section configurations are shown.

Table 2.

Overall gray matter loss for males for the comparison between 21–25 and 41–45 years age groups (21–25 > 41–45).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Middle temporal gyrus	Temporal	Right	63	−28	0	5.93	5.72	0.001
Cingulate gyrus	Limbic	Left	−3	21	34	5.79	5.6	0.001
Inferior frontal gyrus	Frontal	Left	−56	18	4	5.74	5.55	0.002
Superior frontal gyrus	Frontal	Left	−12	51	33	5.7	5.51	0.002
Inferior frontal gyrus	Frontal	Right	50	5	27	5.67	5.49	0.002
Middle frontal gyrus	Frontal	Left	−48	11	31	5.67	5.48	0.002
Medial frontal gyrus	Frontal	Left	−9	51	22	5.63	5.45	0.003
Medial frontal gyrus	Frontal	Left	−3	21	52	5.63	5.45	0.003
Cingulate gyrus (Brodmann area 32)	Frontal	Left	−3	11	40	5.6	5.43	0.003
Supramarginal gyrus (Brodmann area 40)	Parietal	Right	58	−57	27	5.57	5.39	0.004
Superior frontal gyrus (Brodmann area 9)	Frontal	Right	4	53	27	5.54	5.37	0.004
Anterior cingulate	Limbic	Right	3	38	24	5.46	5.29	0.006
Superior temporal gyrus	Temporal	Left	−58	−55	12	5.42	5.26	0.007
Medial frontal gyrus	Frontal	Right	2	56	7	5.37	5.21	0.009
Middle temporal gyrus	Temporal	Right	54	−67	1	5.33	5.18	0.01
Superior frontal gyrus	Frontal	Right	26	6	58	5.23	5.08	0.016
Middle frontal gyrus	Frontal	Left	−27	15	58	5.2	5.06	0.018
Sub-gyral (Brodmann area 6)	Frontal	Left	−26	−4	58	5.1	4.97	0.027

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Figure 2.

Gray matter (GM) regional changes rendered onto a single subject brain for the comparison between 21–25 and 36–40 years age groups (21–25 > 36–40) for females. Four section configurations are shown.

Table 3.

Overall gray matter loss for females for the comparison between 21–25 and 41–45 years age groups (21–25 > 41–45).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Precentral gyrus	Frontal	Left	−56	12	9	7.63	7.21	0
Inferior frontal gyrus (Brodmann area 44)	Frontal	Right	54	9	21	6.52	6.25	0
Inferior parietal (Brodmann area 40)	Parietal	Right	60	−36	46	6.25	6.01	0
Insula (Brodmann area 13)	Insula	Right	38	18	6	5.8	5.6	0.001
Caudate	Sub-lobar	Right	14	8	16	5.75	5.56	0.002
Insula	Insula	Right	46	−9	6	5.65	5.47	0.002
Postcentral gyrus (Brodmann area 2)	Parietal	Right	54	−28	52	5.58	5.4	0.003
Medial frontal gyrus (Brodmann area 9)	Frontal	Left	−10	39	18	5.54	5.37	0.004
Inferior frontal gyrus (Brodmann area 47)	Frontal	Left	−15	18	−21	5.46	5.29	0.006
Precuneus (Brodmann area 7)	Parietal	Right	3	−63	52	5.41	5.25	0.007
Superior temporal gyrus (Brodmann area 22)	Temporal	Right	66	−49	15	5.4	5.24	0.008
Thalamus	Sub-lobar	Right	16	−28	1	5.28	5.13	0.013
Fusiform gyrus (Brodmann area 37)	Temporal	Left	−48	−45	−21	5.25	5.1	0.015
Culmen	Cerebellum	Right	39	−45	−33	5.24	5.09	0.016
Middle temporal gyrus (Brodmann area 22)	Temporal	Left	−64	−39	1	5.18	5.04	0.02
Superior temporal gyrus (Brodmann area 21)	Temporal	Right	64	−25	−3	5.13	4.99	0.025
Superior frontal gyrus (Brodmann area 9)	Frontal	Left	−26	45	37	5.12	4.98	0.026
Middle temporal gyrus (Brodmann area 21)	Temporal	Right	66	−18	−6	5.1	4.96	0.028
Middle frontal gyrus (Brodmann area 6)	Frontal	Left	−44	−1	45	5.09	4.95	0.03

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Table 4.

Overall gray matter loss for females for the comparison between 21–25 and 36–40 years age groups (21–25 > 36–40).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Middle temporal gyrus (Brodmann area 21)	Temporal	Left	−63	−19	−8	6.17	5.94	0
Superior frontal gyrus (Brodmann area 9)	Frontal	Right	10	50	30	6.17	5.93	0
Middle frontal gyrus (Brodmann area 6)	Frontal	Right	−32	5	54	5.76	5.57	0.002
Middle frontal gyrus (Brodmann area 8)	Frontal	Left	−51	5	43	5.67	5.48	0.002
Superior temporal gyrus (Brodmann area 22)	Temporal	Right	60	0	4	5.66	5.48	0.002
Inferior parietal lobe (Brodmann area 40)	Parietal	Right	60	−36	46	5.64	5.46	0.003
Middle temporal gyrus	Temporal	Right	62	−33	0	5.58	5.41	0.003
Superior temporal gyrus	Temporal	Right	66	−27	−3	5.55	5.38	0.004
Middle frontal gyrus	Frontal	Left	−48	9	33	5.54	5.37	0.004
Medial frontal gyrus	Frontal	Left	−4	56	−3	5.52	5.35	0.005
Precentral gyrus	Frontal	Right	57	14	7	5.44	5.28	0.006
Insula (Brodmann area 13)	Insula	Right	45	−9	−6	5.33	5.18	0.011
Insula	Insula	Right	46	−6	4	5.28	5.13	0.013
Anterior cingulate	Limbic	Right	16	36	27	5.21	5.07	0.018
Precentral	Frontal	Left	−57	12	7	5.14	5	0.024
Middle frontal gyrus	Frontal	Left	−48	21	28	5.13	5	0.024

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Table 5.

Overall gray matter loss for males for the comparison between 26–30 and 41–45 years age groups (26–30 > 41–45).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Inferior frontal gyrus	Frontal	Left	−57	18	4	5.56	5.38	0.004
Inferior frontal gyrus	Frontal	Left	−56	14	12	5.31	5.15	0.012
Superior temporal gyrus	Temporal	Left	−58	−55	12	5.2	5.05	0.019

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Table 6.

Overall gray matter loss for females for the comparison between 26–30 and 41–45 years age groups (26–30 > 41–45).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Superior frontal gyrus	Frontal	Left	0	23	57	7.33	6.96	0
Middle frontal gyrus (Brodmann area 11)	Frontal	Left	−40	36	−15	6.57	6.29	0
Subcallosal gyrus (Brodmann area 25)	Frontal	Left	−8	21	−17	6.24	6	0
Superior frontal gyrus (Brodmann area 8)	Frontal	Right	28	26	54	6.22	5.98	0
Superior temporal gyrus (Brodmann area 39)	Temporal	Right	56	−64	19	6.14	5.91	0
Superior frontal gyrus (Brodmann area 9)	Frontal	Right	22	39	40	6.11	5.88	0
Middle temporal gyrus (Brodmann area 21)	Temporal	Right	66	−16	−8	6.1	5.88	0.001
Inferior frontal gyrus	Frontal	Left	−56	15	0	5.75	5.56	0.002
Superior frontal gyrus (Brodmann area 9)	Frontal	Left	−24	42	37	5.67	5.49	0.002
Middle frontal gyrus (Brodmann area 6)	Frontal	Right	27	9	58	5.53	5.36	0.004
Middle frontal gyrus (Brodmann area 9)	Frontal	Right	54	15	31	5.53	5.36	0.004
Postcentral gyrus (Brodmann area 2)	Parietal	Right	54	−30	51	5.51	5.34	0.005
Thalamus	Sub-lobar	Right	18	−28	4	5.45	5.29	0.006
Superior frontal gyrus	Frontal	Right	33	48	25	5.41	5.25	0.007
Medial frontal gyrus (Brodmann area 6)	Frontal	Left	−16	26	40	5.33	5.17	0.011
Middle temporal gyrus (Brodmann area 21)	Temporal	Right	62	−27	−6	5.25	5.1	0.015
Orbital gyrus (Brodmann area 11)	Frontal	Right	6	54	−23	5.22	5.07	0.017
Fusiform gyrus (Brodmann area 19)	Occipital	Right	27	−55	−15	5.22	5.07	0.017
Superior frontal gyrus (Brodmann area 10)	Frontal	Left	−26	66	−6	5.16	5.02	0.022
Insula (Brodmann area 13)	Insula	Right	44	−9	4	5.15	5.01	0.023
Cingulate gyrus	Limbic	Left	0	20	34	5.13	4.99	0.025
Precuneus	Parietal	Left	0	−51	37	5.09	4.96	0.029
Medial frontal gyrus (Brodmann area 10)	Frontal	Left	−10	56	−6	5.07	4.94	0.032

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Table 7.

Multiple linear regression results showing brain regions that have negative gray matter volume correlation with age.

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Superior frontal gyrus	Frontal	Left	−8	47	33	9.89	Inf	0
Superior frontal gyrus	Frontal	Left	−26	18	55	9.76	Inf	0
Precentral gyrus	Frontal	Left	−56	14	7	9.32	Inf	0
Thalamus	Sub-lobar	Right	16	−28	3	6.45	6.2	0
Inferior temporal gyrus	Temporal	Left	−48	−73	−6	6.13	5.91	0
Culmen	Cerebellum	Left	−34	−52	−33	6.09	5.87	0
Cingulate gyrus	Limbic	Right	12	−57	25	6.07	5.85	0
Cuneus	Occipital	Left	−27	−84	31	5.93	5.73	0.001
Precentral gyrus	Frontal	Left	−40	−21	58	5.9	5.7	0.001
Precuneus	Parietal	Right	8	−72	39	5.79	5.6	0.001
Culmen	Cerebellum	Right	38	−48	−32	5.77	5.59	0.001
Culmen	Cerebellum	Right	4	−67	−11	5.55	5.39	0.004
Fusiform	Temporal	Right	27	−40	−18	5.54	5.37	0.004
Lingual gyrus	Occipital	Left	−21	−93	−11	5.46	5.3	0.006
Superior temporal gyrus	Temporal	Left	−45	−37	6	5.35	5.2	0.009
Precentral gyrus	Frontal	Left	−54	−7	37	5.19	5.05	0.019
Thalamus	Sub-lobar	Right	6	−7	9	5.17	5.04	0.02
Thalamus	Sub-lobar	Left	−15	−30	3	5.16	5.02	0.021
Inferior parietal	Parietal	Left	−40	−61	42	5.16	5.02	0.021
Precuneus	Parietal	Left	−2	−66	52	5.14	5	0.023

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Table 8.

Multiple linear regression results showing brain regions having gray matter volume correlated with gender (female > male).

Region	Lobe	Hemisphere	MNI coordinates (mm)			t-value	z score	p (FWE)
			x	y	z
Paracentral lobule	Frontal	Left	−3	−27	45	8.56	Inf	0
Cingulate gyrus	Limbic	Left	−2	−6	42	7.28	6.92	0
Caudate	Sub-lobar	Right	8	4	7	7.06	6.73	0
Cuneus	Occipital	Right	9	−87	4	6.87	6.57	0
Cuneus	Occipital	Left	−6	−87	6	6.82	6.52	0
Declive of Vermis	Cerebellum	Left	−2	−75	−30	6.8	6.51	0
Lingual gyrus	Occipital	Left	−6	−84	−2	6.72	6.44	0
Postcentral gyrus	Parietal	Right	38	−25	42	6.65	6.37	0
Precuneus	Occipital	Left	−24	−82	24	6.53	6.26	0
Superior temporal gyrus	Temporal	Left	−52	−46	9	6.49	6.23	0
Inferior frontal gyrus	Frontal	Left	−12	36	−23	6.36	6.11	0
Parahippocampa gyrus	Limbic	Left	−21	−31	−15	6.21	5.98	0
Inferior frontal gyrus	Frontal	Right	18	28	−20	6.2	5.97	0
Precuneus	Parietal	Left	−8	−72	43	6.16	5.94	0
Inferior frontal gyrus	Frontal	Left	−54	15	3	6.08	5.87	0
Medial frontal gyrus	Frontal	Left	−2	48	13	6.05	5.84	0
Medial frontal gyrus	Frontal	Right	2	2	52	5.95	5.74	0.001
Inferior parietal	Parietal	Right	64	−27	24	5.86	5.67	0.001
Middle occipital gyrus	Occipital	Right	33	−82	12	5.86	5.66	0.001
Superior temporal gyrus	Temporal	Right	48	−33	7	5.78	5.6	0.001
Insula	Insula	Right	46	9	−2	5.72	5.54	0.002
Precuneus	Parietal	Left	−8	−64	58	5.67	5.49	0.002
Precuneus	Parietal	Right	8	−67	31	5.47	5.31	0.005
Postcentral gyrus	Parietal	Left	−40	−25	40	5.28	5.13	0.013
Middle temporal gyrus	Temporal	Right	56	−31	−3	5.25	5.1	0.014
Cerebellum posterior lobe	Cerebellum	Left	−39	−64	−42	5.19	5.05	0.018
Cuneus	Occipital	Left	−26	−84	10	5.18	5.04	0.019
Precentral lobule	Frontal	Right	10	−49	55	5.17	5.03	0.02
Superior frontal gyrus	Frontal	Left	−20	68	10	5.15	5.02	0.021
Medial frontal gyrus	Frontal	Left	−8	63	12	5.15	5.01	0.022

FWE: family wise error; MNI: Montreal Neurological Institute; Region: name of specific brain region with significant gray matter loss.

Results of the multiple linear regression, where GM volume (ml) was the dependent variable and gender along with age as the independent variables, showed standardized beta (B) for age as −0.629 (p < 0.001) at 95% confidence level, while B for gender was shown as 0.187 (p < 0.001) at 95% confidence level. Moreover scatter plots with respective coefficient of determination R² values are shown in Figure 3 for males, Figure 4 for females, and Figure 5 for both males and females.

Figure 3.

Scatter plot of gray matter (GM) volume (ml) across age (21–45 years old) for males.

Figure 4.

Scatter plot of gray matter (GM) volume (ml) across age (21–45 years old) for females.

Figure 5.

Scatter plot of gray matter (GM) volume (ml) across age (21–45 years old) for both genders.

Discussion and conclusions

This study reports on a pattern of GM loss in a large series of normal adults using VBM analysis. GM volumes from normal adult males and adult females aged between 21–45 years were respectively compared across five different age subgroups (21–25, 26–30, 31–35, 36–40, 41–45 years). There was no significant reduction in GM volume between the following comparisons for both males and females: 21–25 > 31–35, 21–25 > 26–30, 26–30 > 36–40, 26–30 > 31–35, 31–35 > 41–45, 31–35 > 36–40, and 36–40 > 41–45 years. As hypothesized, distinct and widespread GM volume loss was seen in the 21–25 > 41–45 years comparison for both male and female subgroups. The reduction in GM for 21–25 > 41–45 years comparison was predominantly in the frontal lobe for both males and females, while such GM loss was also evident in the temporal and parietal lobes for both genders. The 21–25 > 36–40 years comparison showed significant GM loss in the frontal regions of the brain for both males and females, while 26–30 > 41–45 years comparison showed significant GM loss in both frontal and temporal regions for both genders.

In this study we also found that regional GM loss was more widespread in females than males (Figures 1 and 2, and Tables 2–6). We also found that the coefficient of determination (R²) value differed between males and females (females > males) with regards to age-related GM loss (Figures 3 and 4). Overall, according to the multiple linear regression of GM volume, we found that age as a much stronger predictor of GM loss (B = −0.629) than gender (B = 0.187) with 95% confidence level across early adulthood and early middle ages (i.e. 21–45 years of age).

In relating our findings to previous studies, Tisserand et al.¹⁸ found that advancing in age was strongly associated with volumetric decreases in the whole of frontal cortex with strongest age-related volume decrease found in the lateral and orbital frontal GM. In a later study Tisserand et al.¹⁹ found that the prefrontal cortex and temporal lobe were among the brain regions that showed the largest age effects, and concluded that such findings suggest that the prefrontal and temporal cortical brain regions are of particular relevance in aging as well as age-related cognitive decline in healthy elderly persons. Another study²⁰ found that frontal and parietal, compared to temporal and occipital, brain regions showed greater tissue loss. These studies along this one and others^21,22 seem to support the longstanding theory of frontal selectivity in cognitive aging.

The frontal lobe was among the brain regions that had the largest effect for GM volume loss across all the comparisons of this study that showed significant GM loss for both genders (21–25 > 41–45, 21–25 > 36–40, 26–30 > 41–45 years). In the 21–25 > 41–45 years comparison the temporal lobe along with the parietal lobe also showed significant loss in GM for males and females. For the 21–25 > 36–40 years comparison GM loss was only present in the left inferior frontal gyrus, while females showed more widespread regional GM loss that included the insula along with frontal, temporal, limbic, as well as parietal regions. As for the 26–30 > 41–45 years comparison males only showed GM loss in frontal and temporal regions while females showed GM loss in both temporal and frontal regions along with the thalamus, occipital lobe, cingulate gyrus, and insula.

With regards to the largest age gap comparison (21–25 > 41–45 years), along with the frontal lobe, the temporal lobe also showed significant decline in GM. It is well known and documented that the certain regions within the temporal lobe (i.e. entorhinal and perirhinal cortices) are devastated by neurofibrillary pathology and cell loss in very early course of Alzheimer’s disease.^23,24 However it remains less clear whether and how such temporal cortical regions are affected by normal aging due to severely limited investigations.^25–27

The multiple regression results for GM volume along with the scatter plots (Figures 3 and 4) and statistical parametric mapping results (Figures 1 and 2, and Tables 2–6) show that gender as a contributor to regional and global GM differences which is in line with previous work.²⁸ Although gender may be a more minor contributor to GM loss (B = 0.187), age on the other hand is a major contributor to GM loss as indicated by Figures 3, 4, and 5 (R² = 0.30/0.54/0.41 respectively), and B = −0.629.

In this work we provide a pattern of age-related GM loss in normal and healthy adults. Some of the brain regions that showed significant GM loss across different age groups are consistent with previous work. Also, this work is consistent with the frontal selectivity in cognitive aging theory. Moreover, this work provides a unique pattern for age-related GM decline associated with comparisons of young adults to early middle-aged males and females. Therefore, this work will aid in the development of a unified model of normal brain aging for both males and females which in turn may provide brain biomarkers of normal aging for both genders, and this may help ultimately in differentiating normal versus abnormal structural brain changes. On a final note; future work can include an investigation in inter-hemispheric differences, if any, in GM volume across age and gender.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

The author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.