Abstract
Objective:
To study age-related metabolic changes in different brain regions.
Methods:
Point-resolved spectroscopy (repetition time/echo time = 2000 ms/30 ms) was performed in the left and right hippocampus, the left thalamus and the left centrum semiovale of 80 healthy subjects (37 females and 43 males aged 7–64 years). Analysis of covariance and linear regression were used for statistical analysis. Both metabolite concentration ratios with respect to total creatine (tCr) and absolute metabolite concentrations were included for analysis.
Results:
Ins (myo-inositol)/tCr (p < 0.001) and absolute Ins concentration (p = 0.031) were significantly increased with age after adolescence. NAA (N-acetylaspartic acid)/tCr (p < 0.001) and absolute NAA concentration (p = 0.010) significantly declined with age after adolescence.
Conclusion:
Age-related increase of Ins and decline of NAA are found in all three regions, especially at the hippocampus, indicating possible gliosis in the ageing brain.
Advances in knowledge:
We could use NAA/tCr and Ins/tCr as an indicator to estimate the neurons-to-glial cells ratio at the thalamus. This may be an index to distinguish normal tissues from gliosis.
INTRODUCTION
Development and ageing are two issues of common concern that scientists have studied for many years. As summarized in Table 1, these two issues have been investigated in several studies using proton MR spectroscopy (1H-MRS).
Table 1.
MR spectroscopy (MRS) findings of brain development and ageing in the present study and published data
| Study | Number of subjects (male/female ratio) | Age range (years) | Regions of interest | Field strength (T) | MRS (I) | Quantification analysis | Trends seen with increasing age |
|---|---|---|---|---|---|---|---|
| Chang et al1 | 36 | 19–78 | Frontal WM | 1.5 | MRS | A + R | tCr↑, Cho↑, Ins↑, NAA/tCr↓ |
| Lim et al2 | 10 (10/0) | 20–28, 65–75 | Centrum semiovale | 1.5 | MRSI | R | NAA (GM)/NAA (WM)↓ |
| Fukuzako et al3 | 36 (17/19) | 24–79 | Left, medial, temporal | 2 | MRS | R | NAA/tCr↓ |
| Pfefferbaum et al4 | 34 (24/10) | 25.3 ± 2.9, 73.3 ± 4.1 | Centrum semiovale | 1.5 | MRSI | A | tCr↑, Cho↑ |
| Saunders et al5 | 30 (17/13) | 24–89 | Parietal, WM, occipital, GM | 1.5 | MRS | A | tCr↑ |
| Lundbom et al6 | 12 (0/12) | 35 ± 6, 74 ± 7 | Frontal, parietal, temporal, WM, GM | 1.5 | MRSI | R | NAA/Cho↓, NAA/tCr↓ |
| Leary et al7 | 44 (22/22) | 22–62 | Parietal, WM | 1.5 | MRS | A | tCr↑, Cho↑ |
| Angelie et al8 | 32 (14/18) | 21–61 | Cortical GM, centrum semiovale, temporal | 1.5 | MRSI | R | NAA/tCr↓, NAA/Cho↓ |
| Brooks et al9 | 50 (50/0) | 20–70 | Frontal | 1.5 | MRS | A + R | NAA↓, NAA/tCr↓ |
| Grachev and Apkarian10 | 35 (23/12) | 19–31, 40–52 | Cingulate, thalamus, insula, DPFC, orbit frontal cortex, sensorimotor cortex | 1.5 | MRS | R | NAA/tCr↓, glutamate/tCr↓, Gln/tCr↓, Glc/tCr↓, Lac/tCr↓, GABA/tCr↓ |
| Harada et al2 | 21 | 20–70 | Frontal, lentiform nucleus | 1.5 | MRS | A | NAA↓ |
| Kadota et al11 | 90 (45/45) | 4–88 | Centrum semiovale | 1.5 | MRSI | R | NAA/Cho↓ |
| Pouwels et al12 | 97 | 1–39 | Cortical parietal GM, parieto-occipital WM, cerebellum, thalamus, basal ganglia | 2 | MRS | A | NAA↑, NAAG↑, Tau↓ |
| Driscoll et al13 | 32 (16/16) | 20–39, 60–85 | Hippocampus, frontal WM | 1.5 | MRSI | R | NAA/tCr↓ |
| Kreis et al14 | 21 | 1–3 | Centrum semiovale, thalamus, occipital grey matter | 1.5 | MRS | A | NAA↑ |
| Horská et al15 | 15 | 3–19 | Putamen, thalamus, posterior pre-frontal cortex, deep pre-motor/motor/parietal WM, dorsal parietal cortex | 1.5 | MRSI | A + R | NAA/Cho↑ |
| Szentkuti et al.16 | 35 (16/19) | 22–27, 60–75 | Hippocampus, extra hippocampus | 1.5 | MRS | R | NAA/(tCr + Cho)↓ |
| Sailasuta et al17 | 50 | 21–71 | Frontal GM/WM, parietal, GM, basal ganglia | 3 | MRS | NAA↓, glutamate↓ | |
| Gruber et al18 | 63 (24/39) | 18–65 | DPFC, frontal, parietal, thalamus, caudate nucleus, lentiform nucleus | 3 | MRSI | A + R | NAA↓, tCr↑, Cho↑, Ins↑, NAA/tCr↓, Cho/tCr↑, Ins/tCr↑ |
| Raininko and Mattson19 | 57 (32/25) | 13–72 | Supraventricular WM | 1.5 | MRS | A | NAA↓, Ins↑ |
| Reyngoudt et al20 | 90 (48/42) | 18–76 | Posterior cingulate, hippocampus | 3 | MRS | A + R | Ins/tCr↑, Ins↑, tCr↑ |
| Present study | 80 (43/37) | 7–64 | Centrum semiovale, hippocampus, thalamus | 3 | MRS | A + R | NAA↓, Ins↑, NAA/tCr↓, Ins/tCr↑, NAA/Ins↓, glutamate plus glutamine/tCr↓ |
A, absolute; Cho, choline; DPFC, dorsolateral pre-frontal cortex; GM, grey matter; Ins, myo-inositol; MRSI, MRS imaging; NAA, N-acetylaspartic acid; R, relative; tCr, total creatine; WM, white matter.
Previous studies of brain development have found that the increase in the number of neuronal cells reaches its peak by 6 months after birth, and then the centrum semiovale attains maturity with a significant increase in the number and size of axons and dendrites.21 Although major events in the maturation of the cerebrum take place during the first few years of life,22 the myelination, which is most closely related to the full functional capacity of neuronal interconnections, is not completed before adolescence.21 In previous 1H-MRS studies, the age range of the studied subjects rarely included childhood. Girard et al23 have studied the growth of the brain in 22- to 39-week foetuses and reported reductions in Ins (myo-inositol) and Cho (choline) and increases in NAA (N-acetylaspartic acid) in the foetal brains. Kadota et al11 found that NAA/Cho reduced at the centrum semiovale with progressive age from 4 to 88 years. Kreis et al14 scanned 21 infants from 23 to 43 weeks and found that absolute NAA concentration significantly increases with age in the white matter and grey matter. Horská et al15 have employed MRS imaging (MRSI) and studied 15 subjects from 3 to 19 years old and found that NAA/Cho increased to a maximum at 10 years old in grey matter regions, while NAA/Cho increased linearly with age in the white matter.
The normal ageing of the brain has been associated with loss of both neurons and brain volume,24 and some age-related diseases such as mild cognitive impairment and Alzheimer's disease also have these symptoms. 1H-MRS has the potential to observe these changes in the brain by the non-invasive measurement of brain metabolites. A large variability in observations was found in previous studies, partly due to the different inclusion criteria (size, age range), regions of interest, data acquisition [field strength, single-voxel vs multiple-voxel studies, spectroscopy sequence, repetition time (TR)/echo time (TE)] and data processing (relative vs absolute quantification). Some infant studies found the increase of NAA from 22 weeks to 10 years in most grey and white matters.14,15,23 On the contrary, a decreased amount of NAA with increasing age was observed in frontal,6,7,11,17,18 parietal17,18,25 and temporal cortices.3,6,8 And some of these studies did not find age-related NAA changes either in the frontal cortex4 or in the parietal cortex.5,7 The results of NAA in these studies seem unclear and contradictory sometimes. In several studies, an age-related increase in total creatine (tCr) was found in the frontal white matter,1,4,5,7,18 the parietal white matter,5,7 the temporal lobe8 and the posterior cingulate cortex.20 The age-related increase in Cho was also found in the frontal white matter1,4,7,18 and the parietal white matter.7,18 In a small number of studies, an increased amount of Ins was found in the frontal white matter,1,18,19 the parietal white matter,18,19 the hippocampus20 and the posterior cingulate cortex.20
LCModel (Stephen Provencher, Oakville, ON, Canada) is widely used and accepted because it performs spectral fitting based on prior knowledge. Our study used LCModel software (http://s-provencher.com/pages/lcmodel.shtml) to analyse the MRS data from three different brain regions (centrum semiovale, hippocampus and thalamus) of volunteers between 7 and 64 years old. We chose a relatively short TE (30 ms) to reduce the influences of T2 relaxation time to absolute metabolite signals.20,26 We chose the centrum semiovale (partly represents the white matter) and the thalamus (partly represents the grey matter) to observe the development and ageing processes in different regions of the human brain. The hippocampus is important for development because it contains neural progenitor cells27 for neuron meiosis, and it has also been reported to be connected with memory functions and degenerated by ageing and other diseases.13,16,28–30 Hence, the hippocampus was also chosen as a region of interest in this study. Our study uses absolute metabolite concentration of these three brain regions to analyse the trends in chemical changes in these three different brain regions from childhood to middle age.
METHODS AND MATERIALS
Subjects
We studied 80 healthy volunteers (43 males and 37 females) with a mean age of 31 years (range, 7–64 years). None had any history of disorders affecting the central nervous system or demonstrated any abnormal signs on MRI. All subjects were right-handed, non-smokers and non- or light drinkers. This study was approved by the Ethics Committee of Wuhan University, Wuhan, China. Written informed consent was obtained from all the subjects or (for subjects under 18 years) their guardians.
MR examination
MR examination was performed on a Siemens MAGNETOM® Trio 3.0-T whole-body MR system (Siemens Healthcare, Erlangen, Germany) with a 32-channel head coil. First, high-resolution T2 weighted images in three orthogonal planes (sagittal, coronal and transversal with continuous slicing) were performed to permit localization of the voxels of interest (VOIs), including the left and right hippocampus, the left thalamus and the left centrum semiovale (Figure 1). Then single-voxel 1H-MRS was performed using point-resolved spectroscopy (PRESS) with TR/TE = 2000 ms/30 ms, spectral bandwidth = 1200 Hz, 1024 data points, 128 signals averaged and sampling interval = 0.83 ms. The normal size of the VOI was 2000 mm3 (normal of regions sizes are: hippocampus, 8 × 8 × 30 mm3; centrum semiovale, 10 × 10 × 20 mm3; thalamus, 10 × 10 × 20 mm3), and the normal sizes of the voxels in all three regions are nearly the same, while in some subjects the VOI was adapted to fit the hippocampal region in order to avoid cerebrospinal fluid (CSF) contamination. A similar PRESS sequence without water suppression was scanned at the same voxels to acquire a water signal as the internal reference for absolute quantification. Field shimming and water suppression were automatically accomplished by the spectral software package. Pre-saturation bands for outer volume suppression were placed around the VOI to avoid contamination from neighbouring CSF or scalp fat. For the volunteers' comfort, no more than three spectra were acquired from each subject, and only two spectra from the seven subjects under 18 or over 55 years old so that total scanning time did not exceed 30 min.
Figure 1.
An example of the volume of interest (VOI). The VOIs (indicated by the white rectangles) are placed at the hippocampus (a), the centrum semiovale (b) and the thalamus (c).
Post-processing of MR spectroscopy data
All spectra were taken for analysis with LCModel v. 6.2 using the water-scaling option. Raw data were selected in two subsequent steps: first, spectra were rejected if full width at half maximum (FWHM) of NAA was >0.1 ppm (13 Hz) or if the signal-to-noise ratio (SNR) of the LCModel output file was <6 for the NAA peak; second, metabolites could be accepted for further statistical analyses only if the estimated standard deviation (SD) provided by LCModel was <20%. LCModel fits a linear combination of the full spectral pattern for each metabolite included in the basis set. In the LCModel fitting result, the metabolite concentration ratio (/tCr) and absolute concentration by water scaling (mmol kg−1 of wet matter, mmol kgww−1) of NAA, tCr, Cho, Ins, glutamate plus glutamine (Glx), macromolecules at 0.9 ppm (MM09), macromolecules at 1.2 ppm (MM12) and macromolecules at 1.4 ppm (MM14), referenced at 2.01, 3.03, 3.19, 3.52, 2.12–2.46, 0.9, 1.2 and 1.4, respectively (Figure 2), were chosen for statistical analysis. At the hippocampus, since CSF is unavoidable, a correction for CSF was needed. To correct the concentrations for the amount of CSF contained in the VOIs of the hippocampus, segmentation of the high-revolution transversal T2 weighted images was performed as we described previously.31 The volume of CSF in each slice was calculated by multiplying their areas by the slice thickness, and their summation gave the total volume of CSF parts within the voxel. We assumed that CSF parts contributed to tissue water rather than to metabolite, and the corrected metabolite concentration was calculated as follows: corrected metabolite concentration at the hippocampus = metabolite concentration × VOI volume/(VOI volume − CSF volume).
Figure 2.
An example of an LCModel result of spectroscopy. N-acetylaspartic acid (NAA), total creatine (tCr), choline (Cho), myo-inositol (Ins), glutamate plus glutamine (Glx), macromolecules at 0.9 ppm (MM09), macromolecules at 1.2 ppm (MM12) and macromolecules at 1.4 ppm (MM14) were chosen for statistical analysis.
Statistical analysis
Statistical analysis was performed using the SPSS 16.0 software package (SPSS® 16.0 for Windows; SPSS Inc., Chicago, IL). Subjects younger than 19 years were classified as “Children”; from 19 to 39 years as “Adults”; and older than 40 years as “Middle-aged”. An unpaired Student's t-test was applied to compare metabolite concentration ratios and absolute metabolite concentrations between males and females and between the left and right hippocampus. Analysis of covariance (ANCOVA) was used to test the effects of region and age on metabolite concentration ratios (/tCr) and absolute metabolite concentrations. Subsequent multiple comparisons (using the Bonferroni correction method) of the three age-dependent subgroups in the same regions (the centrum semiovale in Children, Adults and Middle-aged groups; the hippocampus in Children, Adults and Middle-aged groups; the thalamus in Children, Adults and Middle-aged groups) were taken to exclude interactions of regions and to carry out the region-dependent analysis. When there were significant differences in metabolite concentration ratios and absolute metabolite concentrations at different age levels, a Pearson correlation analysis was used to analyse the correlation between age and metabolite concentration ratios or absolute metabolite concentrations. p < 0.05 was taken as statistically significant.
RESULTS
A total of 196 spectral data were acquired from 80 volunteers. Data with poor spectral quality [low SNR <5, large FWHM >0.1 ppm (13 Hz)] were excluded, and the remaining 153 data (details in Table 2) were included for further analysis. The SNR of the data from the centrum semiovale, the hippocampus and the thalamus was 19.1, 7.74 and 9.89, respectively. The FWHM (mean ± SD) of NAA was 0.057 ± 0.017 ppm (7.41 ± 0.22 Hz).
Table 2.
The distribution of MR spectroscopy data acquired and analysed
| Group | Acquired | Analysed |
|---|---|---|
| Female/male | 90/106 | 71/82 |
| Children/Adults/Middle-aged | 40/90/66 | 25/76/52 |
| Centrum semiovale/hippocampus/thalamus | 67/74/55 | 59/53/41 |
The total acquired spectral data was 196, and the total analysed spectral data was 153.
Table 3 and Table 4 show the results of ANCOVA of the metabolite concentration ratios and the absolute metabolite concentrations, respectively, with p-values for age, region and the region-by-age interaction effect. Table 5 shows the quantitative analysis result of all data in different age groups and region groups.
Table 3.
p-values of analysis of covariance of metabolite concentration ratios [/tCr (total creatine)] vs age and region
| /tCr | Age (at the same region) | Region (at the same age level) |
|---|---|---|
| Macromolecules at 0.9 ppm | 0.606 | 0.001a |
| Macromolecules at 1.2 ppm | 0.845 | 0.010a |
| Macromolecules at 1.4 ppm | 0.721 | 0.012a |
| N-acetylaspartic acid | 0.001a | 0.008a |
| Glutamate plus glutamine | 0.011a | 0.001a |
| Choline | 0.428 | 0.001a |
| Myo-inositol | 0.001a | 0.001a |
Significant difference (p < 0.05).
Table 4.
p-values of analysis of covariance of absolute concentrations vs age and region
| Water scaling | Age (at the same region) | Region (at the same age level) |
|---|---|---|
| Macromolecules at 0.9 ppm | 0.544 | 0.002a |
| Macromolecules at 1.2 ppm | 0.606 | 0.027a |
| Macromolecules at 1.4 ppm | 0.786 | 0.004a |
| N-acetylaspartic acid | 0.010a | 0.001a |
| Myo-inositol | 0.031a | 0.001a |
| Choline | 0.389 | 0.269 |
| Glutamate plus glutamine | 0.251 | 0.001a |
| Total creatine | 0.522 | 0.001a |
Significant difference (p < 0.05).
Table 5.
The metabolite concentration ratios [/tCr (total creatine)] and absolute concentrations (mmol kgww−1, expressed as mean ± standard deviation) in different ages and regions (the first-line data are metabolite concentration ratios and the second-line data are absolute concentrations)
| Metabolites | Age groups | Centrum semiovale | Hippocampus | Thalamus |
|---|---|---|---|---|
| N-acetylaspartic acida | Children | 1.21 ± 0.39 6.62 ± 0.37 |
1.14 ± 0.20 6.18 ± 0.73 |
1.14 ± 0.18 7.71 ± 1.13 |
| Adults | 1.57 ± 0.44 7.50 ± 0.67 |
1.12 ± 0.21 6.19 ± 1.01 |
1.12 ± 0.22 7.73 ± 0.91 |
|
| Middle-aged | 1.05 ± 0.26 6.98 ± 0.92 |
1.01 ± 0.33 5.58 ± 0.65 |
1.03 ± 0.20 7.08 ± 1.02 |
|
| Choline | Children | 0.35 ± 0.08 2.05 ± 1.45 |
0.30 ± 0.03 2.06 ± 0.29 |
0.28 ± 0.04 1.72 ± 0.23 |
| Adults | 0.34 ± 0.06 1.78 ± 0.25 |
0.29 ± 0.04 1.87 ± 0.22 |
0.25 ± 0.03 1.61 ± 0.15 |
|
| Middle-aged | 0.36 ± 0.04 1.79 ± 0.20 |
0.29 ± 0.27 1.79 ± 0.31 |
0.26 ± 0.05 1.58 ± 0.30 |
|
| Myo-inositola | Children | 0.71 ± 0.15 4.00 ± 0.57 |
0.30 ± 0.03 2.05 ± 0.43 |
0.28 ± 0.04 1.72 ± 0.33 |
| Adults | 0.51 ± 0.10 3.60 ± 0.57 |
0.40 ± 0.09 4.86 ± 0.97 |
0.60 ± 0.15 4.32 ± 0.82 |
|
| Middle-aged | 0.82 ± 0.17 4.12 ± 0.76 |
0.99 ± 0.22 6.86 ± 1.26 |
0.71 ± 0.16 4.75 ± 0.96 |
|
| Glutamate plus glutamine | Children | 1.24 ± 0.15 5.40 ± 1.05 |
1.45 ± 0.13 9.90 ± 1.82 |
1.54 ± 0.16 9.55 ± 1.77 |
| Adults | 1.16 ± 0.16 5.31 ± 0.89 |
1.30 ± 0.25 9.07 ± 1.76 |
1.33 ± 0.30 8.24 ± 1.65 |
|
| Middle-aged | 1.08 ± 0.27 5.04 ± 1.18 |
1.41 ± 0.24 8.77 ± 1.52 |
1.14 ± 0.30 6.83 ± 1.21 |
|
| Macromolecules at 0.9 ppm | Children | 1.21 ± 0.15 5.32 ± 1.06 |
0.81 ± 0.12 5.57 ± 1.00 |
1.09 ± 0.09 6.75 ± 0.83 |
| Adults | 1.05 ± 0.15 5.21 ± 0.65 |
0.98 ± 0.19 6.31 ± 1.02 |
0.98 ± 0.16 6.12 ± 0.79 |
|
| Middle-aged | 1.02 ± 0.15 4.98 ± 1.19 |
0.91 ± 0.39 5.34 ± 2.13 |
1.03 ± 0.20 6.43 ± 0.99 |
|
| Macromolecules at 1.2 ppm | Children | 0.34 ± 0.05 1.52 ± 0.33 |
0.24 ± 0.06 1.65 ± 0.40 |
0.37 ± 0.03 2.31 ± 0.47 |
| Adults | 0.36 ± 0.06 1.79 ± 0.54 |
0.31 ± 0.05 1.85 ± 0.56 |
0.32 ± 0.04 2.08 ± 0.45 |
|
| Middle-aged | 0.32 ± 0.07 1.51 ± 0.24 |
0.30 ± 0.06 1.79 ± 0.36 |
0.33 ± 0.09 2.07 ± 0.42 |
|
| Macromolecules at 1.4 ppm | Children | 1.13 ± 0.48 4.51 ± 1.43 |
0.65 ± 0.17 4.50 ± 1.21 |
0.99 ± 0.20 6.17 ± 1.14 |
| Adults | 0.94 ± 0.23 4.39 ± 1.10 |
0.85 ± 0.29 5.83 ± 1.33 |
0.85 ± 0.21 5.27 ± 1.28 |
|
| Middle-aged | 0.86 ± 0.12 3.96 ± 0.90 |
0.88 ± 0.13 5.38 ± 0.91 |
0.89 ± 0.11 5.76 ± 0.98 |
|
| tCr | Children | 4.53 ± 0.45 | 6.89 ± 0.93 | 6.21 ± 0.36 |
| Adults | 4.57 ± 0.85 | 6.88 ± 0.97 | 6.22 ± 0.57 | |
| Middle-aged | 4.91 ± 0.37 | 6.10 ± 0.94 | 6.25 ± 0.49 |
Significant difference (p < 0.05).
In the first-round analysis, there was no significant difference between males and females (p = 0.09) and between the left and right hippocampus (p = 0.70); therefore, the male/female and the hippocampus data were combined for further analysis.
Metabolite concentration ratios
Age difference
ANCOVA demonstrated that NAA/tCr in the Middle-aged group was significantly lower than in the Children group and Adults group (Tables 3 and 5). The Ins/tCr in the Middle-aged group was significantly higher than in the Children group and Adults group (Tables 3 and 5). The Glx/tCr in the Children group was significantly higher than in the Middle-aged group and Adults group (Tables 3 and 5).
Region difference
ANCOVA demonstrated that NAA/tCr, MM09/tCr, MM12/tCr and MM14/tCr in the centrum semiovale and thalamus were higher than in the hippocampus (Tables 3 and 5). The Cho/tCr in the centrum semiovale was the highest of the three regions (Tables 3 and 5). The Glx/tCr in the centrum semiovale was the lowest in the three regions (Tables 3 and 5).
Post hoc analysis
Myo-inositol/total creatine
The post hoc multiple comparisons (using the Bonferroni correction method) of Ins/tCr shows (Table 5): in the centrum semiovale, the Adults group was significantly lower than the Children (pAdults vs Children < 0.001) and Middle-aged groups (pAdults vs Middle-aged < 0.001); in the hippocampus, the Middle-aged group was significantly higher than the Children (pChildren vs Middle-aged < 0.001) and Adults groups (pAdults vs Middle-aged < 0.001); in the thalamus, the Children group was significantly lower than the Adults (pAdults vs Children < 0.001) and Middle-aged groups (pChildren vs Middle-aged < 0.001).
N-acetylaspartic acid/total creatine
The post hoc multiple comparisons (using the Bonferroni correction method) of NAA/tCr shows (Table 5): in the centrum semiovale, the Adults group was significantly higher than the Middle-aged groups (pAdults vs Middle-aged = 0.002); in the hippocampus, the Middle-aged group was significantly lower than the Adults group (pAdults vs Middle-aged = 0.005); in the thalamus, the Middle-aged group was significantly lower than the Children (pChildren vs Middle-aged < 0.001) and Adults groups (pAdults vs Middle-aged = 0.002).
Glutamate plus glutamine/total creatine
The post hoc multiple comparisons (using the Bonferroni correction method) of Glx/tCr found: in the centrum semiovale and hippocampus, the Glx/tCr have no significant difference among the three age levels, whereas in the thalamus the Middle-aged group was significantly lower than the Children (pChildren vs Middle-aged = 0.003) and Adults groups (pAdults vs Middle-aged = 0.005).
Absolute metabolite concentrations
Age difference
ANCOVA demonstrated that the absolute NAA concentration in the Adults group was significantly higher than in the Children group and Middle-aged group (Tables 4 and 5). Absolute Ins concentration in the Middle-aged group was significantly higher than in the Children group and Adults group (Tables 4 and 5).
Region difference
ANCOVA demonstrated that the absolute NAA, MM09 and MM12 concentrations in the thalamus were the highest of the three regions (Tables 4 and 5). Absolute Glx concentration in the hippocampus was the highest of the three regions and in the thalamus was higher than in the centrum semiovale (Tables 4 and 5). Absolute Ins concentration in the centrum semiovale was higher than in the hippocampus and in the thalamus in the Children group. In the Adults and Middle-aged groups, absolute Ins concentration in the hippocampus was higher than in the centrum semiovale and in the thalamus.
Post hoc analysis
Myo-inositol
Table 5 shows: in the centrum semiovale, the Ins absolute concentration in the Adults group was significantly lower than in the Children (pAdults vs Children < 0.001) and Middle-aged groups (pAdults vs Middle-aged < 0.001); in the hippocampus, the Ins absolute concentration in the Children group was significantly lower than in the Adults (pAdults vs Children < 0.001), and the Adults group was significantly lower than the Middle-aged groups (pAdults vs Middle-aged < 0.001); in the thalamus, the Ins absolute concentration in the Children group was also significantly lower than in the Adults (pAdults vs Children < 0.001) and Middle-aged groups (pChildren vs Middle-aged < 0.001).
N-acetylaspartic acid
The post hoc multiple comparisons (using the Bonferroni correction method) of NAA suggests that (Table 5): in the centrum semiovale, the absolute concentration of NAA in the Adults group was significantly higher than in the Children (pAdults vs Children = 0.005) and Middle-aged groups (pAdults vs Middle-aged < 0.001); in the hippocampus, the absolute concentration of NAA in the Middle-aged group was significantly lower than in the Children (pChildren vs Middle-aged = 0.004) and Adults groups (pAdults vs Middle-aged = 0.005); in the thalamus, the absolute concentration of NAA in the Middle-aged group was also significantly lower than in the Children (pChildren vs Middle-aged < 0.001) and Adults groups (pAdults vs Middle-aged < 0.001).
Correlations
Linear regression analysis confirmed that there is a negative correlation between NAA/tCr and age in the thalamus (Figure 3; rthalamus = −0.451; p = 0.003) and a positive correlation between Ins/tCr and age in the thalamus (Figure 4; rthalamus = 0.522; p < 0.001).
Figure 3.
Negative correlations between NAA (N-acetylaspartic acid)/tCr (total creatine) and age in the thalamus.
Figure 4.
Positive correlations between Ins (myo-inositol)/tCr (total creatine) and age in the thalamus.
DISCUSSION
We found that the concentrations of NAA and MM09, MM12 in the thalamus are higher than in the hippocampus and in the centrum semiovale. This may be proof of a previous study21 that showed the grey matter may have a higher density in neurons than in the white matter.
NAA, a prominent peak in the proton spectrum, is an amino acid thought to be present in neurons and their dendritic and axonal extensions. It is localized primarily in the central and peripheral nervous system. Brain NAA is commonly believed to provide a marker of neuronal density. Sometimes, the decreased NAA levels may reflect neuronal dysfunction.32 In our study, the absolute concentration of NAA is significantly lower in the Middle-aged group than in the Children or Adults groups in the hippocampus and thalamus, and the absolute concentration of NAA in the Adults group was significantly higher than in the Children or Middle-aged groups in the centrum semiovale. These trends were nearly the same with the NAA/tCr ratio in all three regions. This may show a possibility that, in the white matter, the neural cells mature in adults, but in the grey matter, they might mature earlier than in the white matter. Furthermore, neural cells in both white matter and grey matter will decrease in middle age. Horská et al15 studied 15 subjects from 3 to 19 years old and found NAA/Cho increased to a maximum at 10 years old in the grey matter regions, but NAA/Cho increased linearly with age in the white matter. Both studies show the maximum of NAA/tCr or NAA/Cho came much earlier in the grey matter (at about 6–15 years old) than in the white matter (after 19 years old). These findings may correspond with myelination processes: the full development period of neurons was always before 7 years old, and the dendritic and axonal extensions finish and neural networks mature usually by 19 years old, which has been proved by previous anatomy and diffusion tensor imaging studies.22,33–35 And another observation in our research is that the increased concentration value of NAA from child to adult in the centrum semiovale (Table 5; 7.50 − 6.62 = 0.88 mmol kgww−1) is higher than in the hippocampus (Table 5; 6.19 − 6.18 = 0.01 mmol kgww−1) and in the thalamus (Table 5; 7.73 − 7.71 = 0.02 mmol kgww−1). The maturation of the white matter may lead to a higher increase in NAA concentration in the centrum semiovale. Our finding may correspond with axonal development processes in the white matter elucidated in an anatomic study22 that the white matter contains more axons. The declining trend in NAA concentration from the Adults group to the Middle-aged group is also the same as the declining trend of cognitive function. This decline in NAA concentration may be a consequence of neuronal dysfunction.32
Ins is believed to be involved in messaging and to act as an osmolyte in glial cells. Its altered levels have been associated with Alzheimer's disease and ageing.32 In our study, we found that the absolute Ins concentration in the white matter (centrum semiovale), the grey matter (thalamus) and the hippocampus increases from the Adults group to the Middle-aged group. Our finding of the high level of Ins concentration in the Middle-aged group corresponds to a high density of glial cells in the ageing brain tissues.24,32
We also found that there is a negative correlation between NAA/tCr and age. With increasing age and brain ageing, the neurons may become dysfunctional and gradually turn into glial cells. Our findings indicated a reduced neuronal density and an increased glial cell density with age. The ageing process has an overall effect on the whole brain, which may be the reason that the neuron density declined in different brain tissues. Especially in the thalamus, NAA/tCr has an obvious negative correlation with age (Figure 3; r = −0.451; p = 0.003), and neuronal cell density declines the most in the grey matter. The Ins/tCr has a positive correlation with age in the thalamus (Figure 4), demonstrating that the glial cell density increases with age, indicating that gliosis is possibly related to brain ageing. This finding agrees with several previous 1H-MRS studies.19,20 We could use NAA/tCr and Ins/tCr as an indicator to estimate the neurons-to-glial cells ratio in the thalamus. This may be an index to distinguish the normal tissues from gliosis, and it may also be an important marker for differentiating normal brain from the age-related pathologies such as mild cognitive impairment and Alzheimer's disease.
In our study, we found that the absolute concentrations of Cho, Glx and tCr do not change with age. Although we found that Glx/tCr significantly declines with increasing age we could not find the same declining trend in absolute concentration of Glx.
Our study has several limitations. First, although we studied a wide range of subjects from 7 to 64 years old, we did not include subjects below 6 years old. The first six years of life is a critical period of development in the brain.21,22 We did not include these subjects because it is too hard to obtain reliable and stable spectra from them without anaesthesia. Second, MRSI has been widely used to analyse the metabolic changes in the brain with age and region.2,4,6,8,11,13,18 In order to obtain a high reliability of absolute quantitative metabolite levels, we stopped using MRSI to acquire the data but chose single-voxel PRESS to guarantee the high quality of spectrum. We are going to use MRSI in our further studies of the brain to obtain a map of the whole brain showing changing metabolite levels clearly. Third, the water scaling method compares the unsuppressed water signal intensity in the basis set and the acquired water signal intensity with an assumed tissue water concentration.37 This water scaling method in LCModel has its limitations. This method may involve potential uncertainty since the water relaxation time may change owing to different tissue compositions.
Acknowledgments
ACKNOWLEDGMENTS
The authors are grateful to the Huaxi MR Research Center and Elekta (Shanghai) Instruments Ltd for their assistance with the project.
Contributor Information
Z-Y Yang, Email: evonneovertake@163.com.
Q Yue, Email: qiangmoon@126.com.
H-Y Xing, Email: kevinxhy@163.com.
Q-Y Tan, Email: 15513037138@163.com.
H-Q Sun, Email: sunhuaiqiang@gmail.com.
Q-Y Gong, Email: qiyonggong@hmrrc.org.cn.
Z-J Tan, Email: zjtan@whu.edu.cn.
H Quan, Email: csp6606@sina.com.
FUNDING
This study was supported by the Natural Science Foundation of China (Grant No. 10875092) and the Natural Science Foundation of Hubei province of China (Grant Nos. 2012FKB04449).
REFERENCES
- 1.Chang L, Ernst T, Poland RE, Jenden DJ. In vivo proton magnetic resonance spectroscopy of the normal aging human brain. Life Sci 1996; 58: 2049–56. doi: 10.1016/0024-3205(96)00197-X [DOI] [PubMed] [Google Scholar]
- 2.Lim KO, Spielman DM. Estimating NAA in cortical gray matter with applications for measuring changes due to aging. Magn Reson Med 1997; 37: 372–7. doi: 10.1002/mrm.1910370313 [DOI] [PubMed] [Google Scholar]
- 3.Fukuzako H, Hashiguchi T, Sakamoto Y, Okamura H, Doi W, Takenouchi K, et al. Metabolite changes with age measured by proton magnetic resonance spectroscopy in normal subjects. Psychiatry Clin Neurosci 1997; 51: 261–3. doi: 10.1111/j.1440-1819.1997.tb02595.x [DOI] [PubMed] [Google Scholar]
- 4.Pfefferbaum A, Adalsteinsson E, Spielman D, Sullivan EV, Lim KO. In vivo spectroscopic quantification of the N-acetyl moiety, creatine and choline from large volumes of brain gray and white matter: effects of normal aging. Magn Reson Med 1999; 41: 276–84. doi: [DOI] [PubMed] [Google Scholar]
- 5.Saunders DE, Howe FA, van den Boogaart A, Griffiths JR, Brown MM. Aging of the adult human brain: in vivo quantification of metabolite content with proton magnetic resonance spectroscopy. J Magn Reson Imaging 1999; 9: 711–16. doi: [DOI] [PubMed] [Google Scholar]
- 6.Lundbom N, Barnett A, Bonavita S, Patronas N, Rajapakse J, Tedeschi G, et al. MR imaging segmentation and tissue metabolite contrast in 1H spectroscopic imaging of normal and aging brain. Magn Reson Med 1999; 41: 841–5. doi: [DOI] [PubMed] [Google Scholar]
- 7.Leary SM, Brex PA, MacManus DG, Parker GJ, Barker GJ, Miller DH, et al. A 1H magnetic resonance spectroscopy study of aging in parietal white matter: implication for trials in multiple sclerosis. Magn Reson Imaging 2000; 18: 455–9. doi: 10.1016/S0730-725X(00)00131-4 [DOI] [PubMed] [Google Scholar]
- 8.Angelie E, Bonmartin A, Boudraa A, Gonnaud PM, Mallet JJ, Sappey-Marinier D. Regional differences and metabolic changes in normal aging of the human brain: proton MR spectroscopic imaging study. AJNR Am J Neuroradiol 2001; 22: 119–27. [PMC free article] [PubMed] [Google Scholar]
- 9.Brooks JC, Roberts N, Kemp GJ, Gosney MA, Lye M, Whitehouse GH. A proton magnetic resonance spectroscopy study of age-related changes in frontal lobe metabolite concentrations. Cereb Cortex 2001; 11: 598–605. doi: 10.1093/cercor/11.7.598 [DOI] [PubMed] [Google Scholar]
- 10.Grachev ID, Apkarian AV. Aging alters regional multichemical profile of the human brain: an in vivo 1H-MRS study of young versus middle-aged subjects. J Neurochem 2001; 76: 582–93. doi: 10.1046/j.1471-4159.2001.00026.x [DOI] [PubMed] [Google Scholar]
- 11.Harada M, Miyoshi H, Otsuka H, Nishitani H, Uno M. Multivariate analysis of regional metabolic differences in normal ageing on localised quantitative proton MR spectroscopy. Neuroradiology 2001; 43: 448–52. doi: 10.1007/s002340000513 [DOI] [PubMed] [Google Scholar]
- 12.Kadota T, Horinouchi T, Kuroda C. Development and aging of the cerebrum: assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 2001; 22: 128–35. [PMC free article] [PubMed] [Google Scholar]
- 13.Pouwels PJ, Brockmann K, Kruse B, Wilken B, Wick M, Hanefeld F, et al. Regional age dependence of human brain metabolites from infancy to adulthood as detected by quantitative localized proton MRS. Pediatr Res 1999; 46: 474–85. doi: 10.1203/00006450-199910000-00019 [DOI] [PubMed] [Google Scholar]
- 14.Driscoll I, Hamilton DA, Petropoulos H, Yeo RA, Brooks WM, Baumgartner RN, et al. The aging hippocampus: cognitive, biochemical and structural findings. Cereb Cortex 2003; 13: 1344–51. doi: 10.1093/cercor/bhg081 [DOI] [PubMed] [Google Scholar]
- 15.Kreis R, Hofmann L, Kuhlmann B, Boesch C, Bossi E, Hüppi PS. Brain metabolite composition during early human brain development as measured by quantitative in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 2002; 48: 949–58. doi: 10.1002/mrm.10304 [DOI] [PubMed] [Google Scholar]
- 16.Horská A, Kaufmann WE, Brant LJ, Naidu S, Harris JC, Barker PB. In vivo quantitative proton MRSI study of brain development from childhood to adolescence. J Magn Reson Imaging 2002; 15: 137–43. [DOI] [PubMed] [Google Scholar]
- 17.Szentkuti A, Guderian S, Schiltz K, Kaufmann J, Münte TF, Heinze HJ, et al. Quantitative MR analyses of the hippocampus: unspecific metabolic changes in aging. J Neurol 2004; 251: 1345–53. doi: 10.1007/s00415-004-0540-y [DOI] [PubMed] [Google Scholar]
- 18.Sailasuta N, Ernst T, Chang L. Regional variations and the effects of age and gender on glutamate concentrations in the human brain. Magn Reson Imaging 2008; 26: 667–775. doi: 10.1016/j.mri.2007.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gruber S, Pinker K, Riederer F, Chmelík M, Stadlbauer A, Bittsanský M, et al. Metabolic changes in the normal ageing brain: consistent finding from short and long echo time proton spectroscopy. Eur J Radiol 2008; 68: 320–7. doi: 10.1016/j.ejrad.2007.08.038 [DOI] [PubMed] [Google Scholar]
- 20.Raininko R, Mattson P. Metabolite concentrations in supraventricular white matter from teenage to early old age: a short echo time 1H magnetic resonance spectroscopy (MRS) study. Acta Radiol 2010; 51: 309–15. doi: 10.3109/02841850903476564 [DOI] [PubMed] [Google Scholar]
- 21.Reyngoudt H, Claeys T, Vlerick L, Verleden S, Acou M, Deblaere K, et al. Age-related differences in metabolites in the posterior cingulated cortex and hippocampus of normal ageing brain: a 1H-MRS study. Eur J Radiol 2012; 81: e223–31. doi: 10.1016/j.ejrad.2011.01.106 [DOI] [PubMed] [Google Scholar]
- 22.Yakovlev PI, Lecours AR. The myelogenic cycles of regional maturation of the brain. In: Minkowski A, ed. Regional development of the brain in early life. Oxford, UK: Blackwell Scientific Publications; 1967. [Google Scholar]
- 23.Dobbing J, Sands J. Quantitative growth and development of human brain. Arch Dis Child 1973; 48: 757–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Girard N, Fogliarini C, Viola A, Confort-Gouny S, Fur YL, Viout P, et al. MRS of normal and impaired fetal brain development. Eur J Radiol 2006; 57: 217–25. doi: 10.1016/j.ejrad.2005.11.021 [DOI] [PubMed] [Google Scholar]
- 25.Mrak RE, Griffin ST, Graham DI. Aging-associated changes in human brain. J Neuropathol Exp Neurol 1997; 56: 1269–75. doi: 10.1097/00005072-199712000-00001 [DOI] [PubMed] [Google Scholar]
- 26.Chang L, Jiang CS, Ernst T. Effects of age and sex on brain glutamate and other metabolites. Magn Reson Imaging 2009; 27: 142–5. doi: 10.1016/j.mri.2008.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Träber F, Block W, Lamerichs R, Gieseke J, Schild HH. 1H metabolite relaxation times at 3.0 tesla: measurements of T1 and T2 values in normal brain and determination of regional differences in transverse relaxation. J Magn Reson Imaging 2004; 19: 537–45. [DOI] [PubMed] [Google Scholar]
- 28.Manganas LN, Zhang X, Li Y, Hazel RD, Smith SD, Wagshul ME, et al. Magnetic resonance spectroscopy identifies neural progenitor cells in the live human brain. Science 2007; 318: 980–5. doi: 10.1126/science.1147851 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Trenerry MR, Jack CR, Jr, Ivnik RJ, Sharbrough FW, Cascino GD, Hirschorn KA, et al. MRI hippocampal volumes and memory function before and after temporal lobectomy. Neurology 1993; 43: 1800–5. doi: 10.1212/WNL.43.9.1800 [DOI] [PubMed] [Google Scholar]
- 30.Fernández G, Weyerts H, Schrader-Bölsche M, Tendolkar I, Smid HG, Tempelmann C, et al. Successful verbal encoding into episodic memory engages the posterior hippocampus: a parametrically analyzed functional magnetic resonance imaging study. J Neurosci 1998; 18: 1841–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Schuff N, Amend DL, Knowlton R, Norman D, Fein G, Weiner MW. Age-related metabolite changes and volume loss in the hippocampus by magnetic resonance spectroscopy and imaging. Neurobiol Aging 1999; 20: 279–85. doi: 10.1016/S0197-4580(99)00022-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Yue Q, Shibata Y, Isobe T, Anno I, Kawamura H, Gong QY, et al. Absolute choline concentration measured by quantitative proton MR spectroscopy correlates with cell density in meningioma. Neuroradiology 2009; 51: 61–7. doi: 10.1007/s00234-008-0461-z [DOI] [PubMed] [Google Scholar]
- 33.Govindaraju V, Young K, Maudsley AA. Proton NMR chemical shifts and coupling constants for brain metabolites. NMR Biomed 2000; 13: 129–53. doi: [DOI] [PubMed] [Google Scholar]
- 34.Hasan KM, Sankar A, Halphen C, Kramer LA, Brandt ME, Juranek J, et al. Development and organization of the human brain tissue compartments across the lifespan using diffusion tensor imaging. Neuroreport 2007; 18: 1735–9. doi: 10.1097/WNR.0b013e3282f0d40c [DOI] [PubMed] [Google Scholar]
- 35.Klingberg T, Vaidya CJ, Gabrieli JD, Moseley ME, Hedehus M. Myelination and organization of the frontal white matter in children: a diffusion tensor MRI study. Neuroreport 1999; 10: 2817–21. doi: 10.1097/00001756-199909090-00022 [DOI] [PubMed] [Google Scholar]
- 36.Voineskos AN, Rajji TK, Lobaugh NJ, Miranda D, Shenton ME, Kennedy JL, et al. Age-related decline in white matter tract integrity and cognitive performance: a DTI tractography and structural equation modeling study. Neurobiol Aging 2012; 33: 21–34. doi: 10.1016/j.neurobiolaging.2010.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wellard RM, Briellmann RS, Jennings C, Jackson GD. Physiologic variability of single-voxel proton MR spectroscopic measurements at 3T. AJNR Am J Neuroradiol 2005; 26: 585–90. [PMC free article] [PubMed] [Google Scholar]