Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 24, 2003, 10.1148/radiol.2263011937
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2263011937v1
226/3/889    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sijens, P. E.
Right arrow Articles by Oudkerk, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sijens, P. E.
Right arrow Articles by Oudkerk, M.
(Radiology 2003;226:889-896.)
© RSNA, 2003

Neuroradiology

Brain Changes with Aging: MR Spectroscopy at Supraventricular Plane Shows Differences between Women and Men1

Paul E. Sijens, PhD, Tom den Heijer, MSc, Daniela Origgi, PhD, Sarah E. Vermeer, MD, Monique M. B. Breteler, MD, PhD, Albert Hofman, MD, PhD and Matthijs Oudkerk, MD, PhD

1 From the Dept of Radiology, Univ Hosp Groningen, Hanzeplein 1, 9713 GZ Groningen, the Netherlands (P.E.S., M.O.); Dept of Epidemiology and Biostatistics, Erasmus Medical Center, Rotterdam, the Netherlands (T.d.H., S.E.V., M.M.B.B., A.H.); and Dept of Health Physics, Istituto Europeo di Oncologia, Milan, Italy (D.O.). Received Nov 27, 2001; revision requested Feb 1, 2002; final revision received Jun 10; accepted Jul 16. Rotterdam Scan Study supported by Netherlands Organization for Scientific Research (NWO) and Health Research and Development Council (ZON). M.M.B.B. supported as a fellow of Royal Netherlands Academy of Arts and Sciences. Address correspondence to P.E.S. (e-mail: p.e.sijens@rad.azg.nl).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To assess the effect of aging on the proportions of choline (Cho), creatine, and N-acetylaspartate (NAA) in the brains of elderly women and men.

MATERIALS AND METHODS: A transverse plane above the ventricle of the brain was mapped with magnetic resonance spectroscopy. Examinations were performed in 1995–1996 with 271 healthy subjects (age range, 60–90 years; mean age, 73 years) and were repeated 4 years later (1999–2000). Student t tests were used for statistical analysis.

RESULTS: Difference analysis of the changes in 4 years (paired data) reproduced the decrease in Cho in women only (2.9% per year, P < .001) that had been indicated with intersubject correlation analyses. Decreases in NAA, though significant in both men and women according to age correlation analyses (P < .01 for both), did not reach significance. The resulting sex difference in the Cho/NAA ratio at a mean age of 77 years, while not yet significant at a mean age of 73 years, was especially manifest in the posterior half of the plane analyzed.

CONCLUSION: Increasing sex differences in Cho/NAA ratios in a supraventricular plane indicate that brain metabolite levels differ between women and men at advanced age.

© RSNA, 2003

Index terms: Aging • Brain, metabolism, 18.12145 • Magnetic resonance (MR), spectroscopy, 18.12145


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Women and men have different brain morphology at magnetic resonance (MR) imaging (1–4). In MR spectroscopic studies of brain metabolism, however, sex differences were absent (5–8) or were limited to the spatial distributions of brain metabolites rather than the brain as a whole (9,10). An exception is our recent demonstration that the prevalence of detectable brain lactate and lipid signals at MR spectroscopy differs between healthy elderly women and men (11).

Decreased brain water content (12) and increased cerebrospinal fluid (12,13) with aging have been reported. Despite the numerous experimental studies with results that show that cellular metabolism declines with aging, MR spectroscopic studies on the effects of aging on the brain metabolite signals of adults have included limited sample sizes, and results are often discrepant. In the frontal lobe, for example, observations vary from increased creatine (Cr) and choline (Cho) levels (12) to decreased N-acetylaspartate (NAA) levels (13) and lack of any significant change (14). Results of some MR spectroscopic studies show aging differences between white matter and gray matter, including decreased Cr in parietal white matter only (15) versus increased Cho and Cr, that were more significant in gray matter than in white matter (16). Since compounds that contain Cho are involved with lipid metabolism, Cr with oxidative metabolism, and NAA presumably with neurotransmission (17), sex differences in these metabolites could have potential clinical implications.

The purpose of this MR spectroscopic study was to assess the effect of aging on the proportions of Cho, Cr, and NAA in the brains of elderly women and men.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Study Population
Nondemented elderly subjects aged 60–90 years were sampled in strata of age and sex from the cohort of the Rotterdam Scan Study of the causes and consequences of cerebral abnormalities seen at MR imaging (18–20). Five hundred eighteen subjects underwent successful MR spectroscopic examinations in 1995–1996. Four years later, all surviving participants were invited for a second examination (1999–2000), and 271 subjects were examined successfully for a second time. At the time of the first MR examinations, the mean ages were 73 years ± 8 (SD) in women and 73 years ± 8 in men. The design of the study, with a sampling scheme that led to similar numbers of men and women in each age cohort, enabled us to investigate the effect of aging on the MR spectra of the human brain. Institutional review board approval from the University Hospital Rotterdam and informed consent from the patients were obtained to select a sample from the Rotterdam Scan Study as the basis of the present study. The use and evaluation of information from healthy volunteers (134 men and 137 women; age range, 60–90 years; mean age, 73 years) were part of the validation of this study, with institutional review board approval and informed consent from the volunteers.

Of the original 518 subjects, 31 died during the 4 years between both examinations, 190 refused to participate (for various reasons, including deteriorated health), and 26 either did not undergo MR spectroscopy or had images of poor quality. For the Cho, Cr, and NAA ratios and for the age at the first MR spectroscopic examination (1995–1996), the subjects who underwent the second MR spectroscopic examination (1999–2000) did not have any significant differences from those who did not undergo the second examination.

MR Imaging and MR Spectroscopy
Brain MR imaging and MR spectroscopy were performed within 30 minutes at 1.5 T with the standard head coil (Vision; Siemens Medical Systems, Erlangen, Germany). MR imaging was performed with the following sequences: spin echo T1 weighted (700/14 [repetition time msec/echo time msec]), proton density weighted (2,500/6), and T2 weighted (2,500/90). An automated hybrid two-dimensional chemical shift imaging sequence (point-resolved spectroscopy [1,500/135]) was used for MR spectroscopy (21). A transverse plane above the ventricle was measured approximately parallel to the canthomeatal line (generally rotated between -10° and 0°) to obtain maps of 8 x 8 spectra within an excited volume of 8 x 8 x 2 cm, or 128 cm3.

Region-of-interest selection was the only nonautomated step in data acquisition and was performed by MR technologists. This large volume was generally fitted tightly between the ventricle and the skull, with little room for variations in inter- and intrasubject positioning. The spectral maps were collected with 2.56-msec sinc Hanning–shaped radio-frequency pulses preceded by 25.6-msec Gaussian-shaped radio-frequency pulses for chemical shift selective excitation and subsequent spoiling of the resultant water signal. The use of one acquisition per phase-encode step, with four prescans and repetition time of 1,500 msec, resulted in acquisition time of 6 minutes 31 seconds at MR spectroscopy.

Frequency-Domain Curve Fitting
Time-domain data were multiplied by a Gaussian function (center, 0 msec; half width, 256 msec), two-dimensional Fourier transformed, and phase and baseline adjusted. The software program (Numaris-3; Siemens Medical Systems) provided with the MR system was used for fully automated quantification by means of frequency-domain curve fitting. In the curve fitting, the number of peaks and their chemical shift ranges were fixed (Cho, 3.3–3.1 ppm; Cr, 2.9–3.1 ppm; NAA, 1.9–2.1 ppm), whereas line widths and peak intensities were unrestricted (21). On the supraventricular maps of 8 x 8 spectra, or 64 spectra, the central 6 x 6 spectra, or 36 spectra, were extracted for assessment of the regional and intersubject variation in the metabolite peak areas and ratios. Time restraints did not allow absolute quantification, which would have required additional phantom measurements or non– water-suppressed baseline measurements. Although differences in coil loading with ensuing variations in sensitivity were not accounted for, monthly phantom evaluations ensured that the system was stable over the course of the study. The goodness of frequency-domain curve fitting with the software was tested with the data sets for six subjects (the content of one of the disks [selected at random] on which the MR data were stored) by means of comparison with the result of a method of time-domain fitting developed in Milan, Italy (D.O.).

Time-Domain Curve Fitting
Packets of a Daubechies wavelet were used for automated time-domain decomposition of the spin-echo signal through successive low-pass then high-pass filtering and subsequent subsampling following a classical decomposition tree structure of wavelet packets (22). Thus, sets of orthonormal subband signals were obtained. Among the eight distinct equal-shaped subbands of the third decomposition level, bands that contained potential peaks at hydrogen 1 MR spectroscopy were the following: band3,2, 3.68–2.69 ppm (inositol, Cho, Cr); band3,3, 2.69–1.70 ppm (NAA); band3,4, 1.70–0.71 ppm (lactate and lipids). Residual water peak to metabolite peak contribution ends up in band3,1, thus automatically isolating water contribution and avoiding separate preprocessing of the filter. Each subsignal was then processed to extract the metabolite peak parameters of amplitude, chemical shift, damping factor, and phase. The spin echo was modeled as the sum of damped complex sinusoids, and the parameters were quantified by means of the identification method of linear prediction singular value decomposition (22).

Automated identification of peaks was based on the frequency information for each spectral component. Among components, the one that corresponded to a metabolite of interest was selected automatically as that closest to the theoretic position. Only the peaks that had a damping factor greater than -0.04 arbitrary units were included in the selection to reduce possible misdetections due to the superimposed noise. In addition, effects of spatial-dependent frequency shifts due to the field inhomogeneity were corrected by setting the water peak at 4.67 ppm. The amplitude parameter was used for the relative quantification of metabolic ratios (23,24).

This method was tested first on a homemade cylindric two-chamber phantom filled with a 100 mol/L lithium lactate solution (inner chamber) and a 100 mol/L sodium lactate solution (outer chamber). In subsequent chemical shift imaging measurements in volunteers, the Cho/Cr, NAA/Cr, and NAA/Cho ratios were reproducible (SD, 2.8%).

Statistical Analysis
The paired Student t test was used to assess differences between the peak areas and ratios of Cho, Cr, and NAA in the first MR spectroscopic examination and those in the same subjects 4 years later. Results obtained in women and men were compared with the t test (independent samples). A P value of less than .05 was considered to indicate a statistically significant difference.

We used a multivariate linear regression model to assess the cross-sectional analyses of the relation between age and peak areas and whether this relation was different for men versus women. The model was peak area = A + age x B1 + sex x B2 + age x sex x B3, where A, B1, B2, and B3 are constants. From this model, the variance and P value of the age-sex interaction term were determined. To investigate whether differences in the relation between age and metabolites were a result of hormone replacement therapy in women, we also performed the analyses without hormone replacement users.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
The volume of interest above the ventricle, a representative spectral map, and a single spectrum are shown in Figure 1, which illustrates the good homogeneity over the whole plane obtained with automated measurement and standard postprocessing. The spectrum is the sixth spectrum from the fourth row of the spectral map. The transverse T1- and T2-weighted MR images were acquired at an angle of 0° to -10° relative to the spectral maps, which prohibited display of these MR images with the spectra. The goodness of the frequency-domain curve fitting with the software was assessed by correlating the results with the numbers obtained with time-domain fitting on a portion from the total data set (sampled at random). Significant correlations were obtained for Cho (r = 0.78, P < .001), Cr (r = 0.77, P < .001), and NAA (r = 0.91, P < .001). Results for NAA were best, as might be expected. Because of the closeness in frequency (peak overlap) and comparative smallness of Cho and Cr peaks, more errors were introduced as a result of correcting phases and baselines (frequency-domain curve fitting).



View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. (a) Sagittal and (b) coronal T1-weighted MR images (15/6) depict location of chemical shift imaging voxels in a plane above the ventricles (volume of interest, 8 x 8 x 2 cm) in a 67-year-old woman. (c) Spectral map (A = anterior, R = right side, W 346 = window, C 198 = contrast) is shown with (d) one of the spectra with peak intensity (arbitrary units) plotted as a function of frequency in parts per million. Acquisition time was 6 minutes 31 seconds (point-resolved spectroscopy, 135/1,500).

 


View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. (a) Sagittal and (b) coronal T1-weighted MR images (15/6) depict location of chemical shift imaging voxels in a plane above the ventricles (volume of interest, 8 x 8 x 2 cm) in a 67-year-old woman. (c) Spectral map (A = anterior, R = right side, W 346 = window, C 198 = contrast) is shown with (d) one of the spectra with peak intensity (arbitrary units) plotted as a function of frequency in parts per million. Acquisition time was 6 minutes 31 seconds (point-resolved spectroscopy, 135/1,500).

 


View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c. (a) Sagittal and (b) coronal T1-weighted MR images (15/6) depict location of chemical shift imaging voxels in a plane above the ventricles (volume of interest, 8 x 8 x 2 cm) in a 67-year-old woman. (c) Spectral map (A = anterior, R = right side, W 346 = window, C 198 = contrast) is shown with (d) one of the spectra with peak intensity (arbitrary units) plotted as a function of frequency in parts per million. Acquisition time was 6 minutes 31 seconds (point-resolved spectroscopy, 135/1,500).

 


View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d. (a) Sagittal and (b) coronal T1-weighted MR images (15/6) depict location of chemical shift imaging voxels in a plane above the ventricles (volume of interest, 8 x 8 x 2 cm) in a 67-year-old woman. (c) Spectral map (A = anterior, R = right side, W 346 = window, C 198 = contrast) is shown with (d) one of the spectra with peak intensity (arbitrary units) plotted as a function of frequency in parts per million. Acquisition time was 6 minutes 31 seconds (point-resolved spectroscopy, 135/1,500).

 
Metabolite Changes in Healthy Elderly Women and Men
In 1995–1996, at a mean age of 73 years, the peak areas and ratios of Cho, Cr, and NAA for the quantified 6 x 6 x 2-cm plane above the ventricle as a whole did not differ significantly between women and men. However, intersubject comparisons revealed significant correlations of Cho and NAA peak areas with age (10). Table 1 shows the percentage annual changes in brain metabolite peak areas (derived from differences between pairs, 1999–2001 vs 1995–1996) in the plane above the ventricle, together with the changes predicted by means of the correlations. In both sexes, the actual decreases in NAA were similar to but less significant than those inferred from the correlations. In contrast, the decrease in Cho in women was both larger and more significant than the predicted change. In the correlation plots, we observed that the scatter of metabolite signal versus age was considerable. This phenomenon is illustrated in Figure 2 for Cho, the metabolite that showed changes in signal strength with aging that differed between the sexes. The decrease in Cho during advancing age is steeper in women than that in men and is more significant than that in men because of the lesser degree of scatter. According to a model that included baseline age and sex, the variance of Cho and the interaction term between age and sex was 40%. This interaction term had borderline significance (P = .07), which suggests that the effect of age on Cho is different for men than that for women. For Cr and NAA, there were no statistically significant interaction terms between age and sex.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Annual Percentage Changes in Brain Metabolite Peak Areas

 


View larger version (36K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Scatterplots show brain Cho levels as a function of age for the 1995-1996 ({bullet}) and 1999-2000 ({circ}) data sets for (left) women and (right) men.

 
Table 2 shows that in 4 years, the ratios of Cho/Cr and NAA/Cr decreased in both men and women (P < .001), which amounted to mean annual decreases of up to 3.8%. The Cho/NAA decrease was observed in women only (P < .001). At a mean age of 77 years, men had higher Cho/Cr and Cho/NAA ratios than those in women (P < .02 for both comparisons). At reexamination in 1999–2000, all mean metabolite ratios had higher SDs than those in 1995–1996. This effect probably reflected a decrease in spectral quality with aging (decreases in the signal) and increasing spatial heterogeneity as a result of an increasing prevalence of lactate and lipid signals (11). Nevertheless, the quality of spectra did not correlate significantly with age, and elimination of the results of subjects with lactate or lipid signals had little effect on the intersubject variations in Cho, Cr, and NAA. We adjusted the values for both sexes by excluding all subjects in whom the larger value of any of the three metabolite ratios exceeded the smaller value by more than 50%; results showed reduced SDs in the 1999–2000 results (Table 3). This procedure filtered out less accurate results (comparatively noisy spectra) and abrupt changes due to development of (hidden) disease.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Brain Metabolite Ratios in Plane above Ventricle and Annual Decrease

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Brain Metabolite Distributions in Plane above Ventricle

 
Spatial Distribution of Metabolite Changes with Aging
The correlations, changes, and sex differences reported earlier were averages over the whole plane of 36 voxels (6 x 6 x 2 cm = 72 cm3). For NAA/Cho, the ratio with the changes most divergent between the sexes, sex differences and changes in the brain metabolite levels over time were evenly distributed over the plane. Figure 3, a color representation of the quantified 6 x 6 voxels, with the right posterior part at bottom left and left anterior part at top right, shows that in all but 1 voxel the sex difference in NAA/Cho was increased at the second examination (1999–2000 compared with 1995–1996). Initially, Cho/NAA in men relative to that in women varied between 0.99 and 1.05 compared with up to 1.19 at 4 years later. The Cho/NAA sex difference increased in all but 3 voxels, especially in the posterior part of the volume of interest (increases varied between 7% and 18%). Spatial distributions of Cho, Cr, and NAA (ie, the metabolite peak area in each voxel divided by the total area of the corresponding metabolite in all 36 voxels) in men relative to those in women are shown in Table 3. Metabolic maps in 1999–2000 were similar to those in 1995–1996 (also when both sexes were considered separately [data not shown]). In a block of eight central-frontal voxels, the proportions of Cho and Cr were reduced in men compared with those in women at both 1995–1996 and 4 years later. Note that unlike the way data are presented in Table 2 and Figure 3, Table 3 is intrinsically independent of the existence of any difference in Cho/Cr, NAA/Cr, and NAA/Cho between 1995–1996 and 1999–2000 data. When hormone replacement therapy users (n = 7) were excluded, results were similar.



View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Color representations of spatial distributions of Cho/NAA in men divided by that in women (a) at first examination (1995-1996) (mean age, 73 years) and (b) at second examination (1999-2000) (mean age, 77 years), as well as (c) changes in 4 years.

 


View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Color representations of spatial distributions of Cho/NAA in men divided by that in women (a) at first examination (1995-1996) (mean age, 73 years) and (b) at second examination (1999-2000) (mean age, 77 years), as well as (c) changes in 4 years.

 


View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c. Color representations of spatial distributions of Cho/NAA in men divided by that in women (a) at first examination (1995-1996) (mean age, 73 years) and (b) at second examination (1999-2000) (mean age, 77 years), as well as (c) changes in 4 years.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
The proportions of Cho, Cr, and NAA in elderly subjects changed substantially with aging, whereas the distribution of these metabolites over the quantified 6 x 6 x 2-cm plane above the ventricle remained similar. The mean annual decreases observed in 271 subjects (approximately 3%) for Cho/Cr and NAA/Cr were one order of magnitude larger than those previously indicated by correlating the age of subjects with their metabolite distributions (10); this difference should not be interpreted as discrepancy. On the basis of this finding, the annual changes in the three metabolites relative to each other are larger than the changes in the distribution of each metabolite when considered separately. The significance of the correlations of peak areas of NAA and Cho (women only) with age (P < .01) appear to indicate that the peak area changes relate to changes in brain metabolite concentrations. The implication is that the concentrations of Cho (women only) and NAA decrease with aging in a large brain region that contains white matter and gray matter (plus minor contributions from ventricular space in the center).

By means of cross-sectional correlation and follow-up, we demonstrated metabolic changes with aging that are significant and involve the whole plane above the ventricles. Previously, we analyzed brain mass relationships and showed in the 1995–1996 data that the differences in brain metabolites between women and men remained after adjustment for brain mass (10,25). Although the 1999–2000 data have not yet been analyzed, it is reasonable to assume that the results of the present study do not suffer from bias due to differences or changes in brain size.

Reductions of Cho/Cr and NAA/Cr probably reflect decreasing concentrations of Cho (women only) and NAA (trends in both women and men) in the aging brain. It is widely acknowledged that NAA/Cr and Cho/Cr are variable in certain regions of the brain, particularly those with deep gray matter structures and temporal lobes. The peri-Rolandic mantle, the brainstem, and the lenticular nuclei contain key elements of earlier neuronal maturation and have variable neuronal densities and are therefore anticipated to exhibit variable metabolite profiles with age. At this time, therefore, we cannot be certain that the indicated decreases of Cho (women only) and NAA levels with aging involve the entire brain. Another question is how to explain the losses of Cho and NAA.

The 1995–1996 MR imaging data in the present study show that white matter lesions are more severe and atrophy is less severe in women compared with those in men (25). Periventricular and subcortical white matter lesions were observed in 92% (126 of 137) and 80% (107 of 134) of subjects, respectively; these proportions increased similarly with age for women and men. Subcortical white matter lesions tended to be larger in women than those in men (total volume, 1.45 vs 1.29 mL, respectively; P = .33) as a result of differences in the volume of frontal white matter lesions (0.89 vs 0.70 mL, respectively; P = .08). Women also had more periventricular white matter lesions than those in men (P = .07). Cerebral atrophy, on the other hand, occurred more frequently in men than in women for both cortical and subcortical atrophy (P < .001 for both) (25).

At this time, we conclude the following: Our observations of decreased NAA associated with the loss of neurons that is known to occur with aging in men and women fit with recently demonstrated correlations of decreasing NAA with progression of cortical atrophy (26,27). Nevertheless, the observation that for all subjects the distributions of Cho, Cr, and NAA remained similar indicates that our results were not influenced by any progression of atrophy between examinations. The quantified 6 x 6 x 2-cm plane above the ventricles contained mainly white matter, with some gray matter in the periphery of the volume of interest and negligible amounts of ventricular space (well below 10% of the volume of any voxel).

In 1995–1996 MR images, subcortical white matter lesions occurred more frequently in women than in men as a result of differences in the frontal white matter lesion volume (25). This phenomenon appears to coincide with reduced Cho and Cr levels in men relative to those in women (Table 3, top rows of columns C3 and C4). It is possible that white matter lesions, more frequent in women, have increased contents of Cho and Cr. In this region of the brain, we previously showed that the respective contents of Cho (in particular) and Cr (to a lesser degree) are comparatively high, up to 114% and 108% of the average in the quantified plane in men and up to 121% and 111% in that in women (10). The implication that there are increased Cho and (to a lesser degree) Cr levels in the presence of white matter lesions fits with previous associations of increased Cho/NAA and Cho/Cr with white matter lesions in two small studies of healthy elderly subjects (28,29).

In women only, decreases in Cho observed in the plane examined as a whole could reflect the loss of myelin lipids, which is reported to be particularly severe in female brains after the age of 70 years (30). The relevance of this is enhanced by the fact that our study was performed with 271 healthy subjects aged 60–90 years in 1995–1996 who were reexamined 4 years later. In any case, the findings in our study are probably confined to this particular age group instead of being applicable to younger adults. Although our observations involved the entire plane examined, the changes in metabolite levels with aging may not involve the entire brain. Knowledge of the changes at MR imaging and the vascular factors in these subjects will increase our understanding of the metabolic changes in women and men with aging.


   ACKNOWLEDGMENTS
 
The authors thank Bart Schraa, RT, and Deni J. A. Kraus, RT, for the 1995–1996 MR acquisitions, and Bart Schraa, RT, and Peter Kappert, RT, for the 1999–2000 MR acquisitions.


   FOOTNOTES
 
Abbreviations: Cho = choline, Cr = creatine, NAA = N-acetylaspartate

Author contributions: Guarantors of integrity of entire study, P.E.S., M.O.; study concepts, T.d.H., D.O., P.E.S., M.M.B.B., A.H.; study design, A.H., M.M.B.B., P.E.S., T.d.H., S.E.V.; literature research, P.E.S.; clinical studies, P.E.S., M.O.; experimental studies, D.O.; data acquisition, T.d.H., D.O., P.E.S.; data analysis/interpretation, S.E.V., P.E.S., T.d.H., D.O.; statistical analysis, P.E.S., S.E.V., T.d.H.; manuscript preparation, P.E.S., T.d.H., D.O., S.E.V.; manuscript definition of intellectual content, A.H., M.M.B.B., M.O.; manuscript editing, P.E.S.; manuscript revision/review, S.E.V., T.d.H., M.O., M.M.B.B., A.H.; manuscript final version approval, A.H., M.O.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Grant R, Gordon B, Lawrence A, et al. Human cranial CSF volumes measured by MRI: sex and age influences. Magn Reson Imaging 1987; 5:465-468.[CrossRef][Medline]
  2. De Lacoste M, Cadesanya T, Woodward DJ. Measurement of gender differences in the human brain and their relationship to brain weight. Biol Psychiatry 1990; 28:931-942.[CrossRef][Medline]
  3. Gur RC, Mozley PD, Resnick SM, et al. Gender differences in age effect on brain atrophy measured by magnetic resonance imaging. Proc Natl Acad Sci U S A 1991; 88:2845-2849.[Abstract/Free Full Text]
  4. Agartz I, Saaf J, Wahlund LO, Wetterberg L. Quantitative estimations of cerebrospinal fluid spaces and brain regions in healthy controls using computer-assisted tissue classification of magnetic resonance images: relation to age and sex. Magn Reson Imaging 1992; 10:217-226.[CrossRef][Medline]
  5. Charles HC, Lazeyras F, Krishnan KKR, et al. Proton spectroscopy of human brain: effect of age and sex. Biol Psychiat 1994; 18:995-1004.
  6. Bernard D, Walker PM, Baudouin-Poisson N, et al. Asymmetrical metabolic profile in mesial temporal lobes: localized H-1 MR spectroscopy in healthy right-handed and non-right-handed subjects. Radiology 1996; 199:381-389.[Abstract/Free Full Text]
  7. Pouwels PJW, Frahm J. Regional metabolite concentrations in human brain as determined by quantitative localized proton MRS. Magn Reson Med 1998; 39:53-60.[Medline]
  8. Komoroski RA, Heimberg C, Cardwell D, Karson CN. Effects of gender and region on proton MRS of normal human brain. Magn Reson Imaging 1999; 17:427-433.[CrossRef][Medline]
  9. Wilkinson ID, Paley MNJ, Miszkiel KA, et al. Cerebral volumes and spectroscopic proton metabolites on MR: is sex important? Magn Reson Imaging 1997; 15:243-248.[CrossRef][Medline]
  10. Sijens PE, Oudkerk M, de Leeuw FE, et al. 1H chemical shift imaging of the human brain at age 60–90 reveals metabolic differences between women and men. Magn Reson Med 1999; 42:24-31.[CrossRef][Medline]
  11. Sijens PE, den Heijer T, de Leeuw FE, et al. MR spectroscopy detection of lactate and lipid signals in the brains of healthy elderly people. Eur Radiol 2001; 11:1495-1501.[CrossRef][Medline]
  12. Chang L, Ernst T, Poland RE, Jenden DJ. In vivo proton magnetic resonance spectroscopy of the normal aging brain. Life Sci 1996; 58:2049-2056.[CrossRef][Medline]
  13. 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.[Abstract/Free Full Text]
  14. Harada M, Miyoshi H, Otsuka H, Nishitani H, Uno M. Multivariate analysis of regional metabolic differences in normal aging on localized quantitative proton MR spectroscopy. Neuroradiology 2001; 43:448-452.[CrossRef][Medline]
  15. Saunders DE, Howe FA, van den Boogaart A, Griffiths JR, Brown MM. Aging of the adult human brain: in vivo quantitation of metabolite content with proton magnetic resonance spectroscopy. J Magn Reson Imaging 1999; 9:711-716.[CrossRef][Medline]
  16. 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; 41:276-284.[CrossRef][Medline]
  17. Sijens PE, Oudkerk M. Magnetic resonance spectroscopy. In: Andra W, Nowak H, eds. Magnetism in medicine. Berlin, Germany: Wiley-VCH, 1998; 373-394.
  18. de Leeuw FE, de Groot JC, Oudkerk M, et al. A follow-up study of blood pressure and cerebral white matter lesions. Ann Neurol 1999; 46:827-833.[CrossRef][Medline]
  19. de Leeuw FE, de Groot JC, Oudkerk M, et al. Aortic atherosclerosis at middle age predicts cerebral white matter lesions in the elderly. Stroke 2000; 31:425-429.[Abstract/Free Full Text]
  20. de Groot JC, de Leeuw FE, Oudkerk M, et al. Cerebral white matter lesions and cognitive function: the Rotterdam Scan Study. Ann Neurol 2000; 47:145-151.[CrossRef][Medline]
  21. Sijens PE, van den Bent MJ, Nowak PJCM, van Dijk P, Oudkerk M. 1H chemical shift imaging reveals loss of brain tumor choline signal after administration of Gd-contrast. Magn Reson Med 1997; 37:222-225.[Medline]
  22. Barkhuijsen H, de Beer R, Bovee WMMJ, Creijghton JH, van Ormondt D. Retrieval of frequency amplitudes, damping factors and phases from time-domain signals using a linear least-squares procedure. J Magn Reson 1985; 61:465-481.
  23. Mainardi LT, Origgi D, Lucia P, Mainardi LT, Scotti G, Cerutti S. A wavelet packets decomposition algorithm for quantification of in vivo 1H-MRS parameters. Med Eng Phys 2002; 24:201-208.[CrossRef][Medline]
  24. Origgi D, Mainardi LT, Ghezzi G, et al. Automatic 1H-magnetic resonance spectra quantification and metabolic map extraction with wavelet packets algorithm. Book of abstracts: 15th Annual Meeting of the ESMRMB European Society for Magnetic Resonance in Medicine and Biology 1998.
  25. Sijens PE, den Heijer T, de Leeuw FE, et al. Human brain chemical shift imaging at age 60–90: analysis of the causes of the observed sex differences in brain metabolites. Invest Radiol 2001; 36:597-603.[CrossRef][Medline]
  26. Lundbom N, Barnett A, Bonavita S, et al. MR image segmentation and tissue metabolite contrast in 1H spectroscopic imaging of normal and aging brain. Magn Reson Med 1999; 41:841-845.[CrossRef][Medline]
  27. Ende G, Braus DF, Walter S, et al. Effects of age, medication, and illness duration on the N-acetyl aspartate signal of the anterior cingulate region in schizophrenia. Schizophr Res 2000; 41:389-395.[CrossRef][Medline]
  28. Sappey-Marinier D, Calabrese G, Hetherington HP, et al. Proton magnetic resonance spectroscopy of human brain: applications to normal white matter, chronic infarction, and MRI white matter signal intensities. Magn Reson Med 1992; 26:313-327.[Medline]
  29. Constans JM, Meyerhoff DJ, Gerson J, et al. H-1 MR spectroscopic imaging of white matter signal hyperintensities: Alzheimer disease and ischemic vascular dementia. Radiology 1995; 197:517-523.[Abstract/Free Full Text]
  30. Svennerholm L, Bostrom K, Jungbjer B. Changes in weight and compositions of major membrane components of human brain during the span of adult human lives of Swedes. Acta Neuropathol (Berlin) 1997; 94:345-352.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Cereb CortexHome page
N. M. Zahr, D. Mayer, A. Pfefferbaum, and E. V. Sullivan
Low Striatal Glutamate Levels Underlie Cognitive Decline in the Elderly: Evidence from In Vivo Molecular Spectroscopy
Cereb Cortex, October 1, 2008; 18(10): 2241 - 2250.
[Abstract] [Full Text] [PDF]

Home page
 Eur Respir JHome page
M. Alchanatis, N. Deligiorgis, N. Zias, A. Amfilochiou, E. Gotsis, A. Karakatsani, and A. Papadimitriou
Frontal brain lobe impairment in obstructive sleep apnoea: a proton MR spectroscopy study
Eur. Respir. J., December 1, 2004; 24(6): 980 - 986.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2263011937v1
226/3/889    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sijens, P. E.
Right arrow Articles by Oudkerk, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sijens, P. E.
Right arrow Articles by Oudkerk, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE