| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Neurology (L.T.H., N.K., V.K.), Internal Medicine (A.V., P.N.), and Radiology (R.P.), University of Turku, FIN-20521 Turku, Finland; and Turku PET Centre (L.T.H., A.V., N.K., J.O.R., P.N., V.K.), FIN-20521 Turku, Finland
Address all correspondence and requests for reprints to: Dr. Lauri T. Haltia, Turku PET Centre, University of Turku, P.O. Box 52, FIN-20521 Turku, Finland. E-mail: toivo.haltia{at}tyks.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Methods: In part I of the study, we scanned with magnetic resonance imaging 16 lean (mean body mass index, 22 kg/m2) and 30 obese (mean body mass index, 33 kg/m2) healthy subjects. In part II, 16 obese subjects continued with a very low-calorie diet for 6 wk, after which they were scanned again. Regional brain white and gray matter volumes were calculated using voxel-based morphometry.
Results: White matter volumes were greater in obese subjects, compared with lean subjects in several basal brain regions, and obese individuals showed a positive correlation between white matter volume in basal brain structures and waist to hip ratio. The detected white matter expansion was partially reversed by dieting. Regional gray matter volumes did not differ significantly in obese and lean subjects, and dieting did not affect gray matter.
Conclusions: The precise mechanism for the discovered white matter changes remains unclear, but the present study demonstrates that obesity and dieting are associated with opposite changes in brain structure. It is not excluded that white matter expansion in obesity has a role in the neuropathogenesis of degenerative brain diseases.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Hence, studies concerning degenerative brain diseases support the idea that obesity has a negative impact on brain function, and there are human studies indicating functional differences in the brain between healthy obese and lean individuals. Imaging studies with positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) have found that obesity is associated with alterations in brain blood flow and neurochemistry. A PET study with [11C]raclopride has indicated that the availability of brain dopamine D2 receptors of very obese individuals is decreased in proportion to their body mass index (BMI) (6). Studies using PET and measures of regional cerebral blood flow have shown differential brain responses to satiation in obese and lean individuals (7, 8), and an fMRI study demonstrated that oral glucose ingestion induces an inhibition of the fMRI signal in the parts of hypothalamus and that this central inhibitory response is markedly attenuated in obese subjects (9). Another fMRI study revealed the greater number of brain activation areas in the obese binge eaters (compared with lean binge eaters and lean and obese nonbinge eaters) in response to visual and auditory binge food stimuli (10). In addition, one previous study with single photon emission tomography showed that visual exposure to food is associated with increases in the regional cerebral blood flow of right temporal and parietal cortices in obese women but not normal-weight women (11). A recent structural study with the magnetic resonance imaging (MRI) and voxel-based morphometry (VBM) indicated that obese individuals have significantly lower brain gray matter volume in the postcentral gyrus, frontal operculum, putamen, and middle frontal gyrus, compared with the group of lean subjects, and that BMI in obese (but not lean) subjects is negatively associated with gray matter volume of the left postcentral gyrus (12). Also, a difference in white matter volume was detected in the vicinity of the striatum, in which obese subjects had greater volume than lean subjects.
Most obesity brain imaging studies are static group comparisons. Often the groups have been separated according to BMI, and a chosen central nervous system variable, e.g. regional blood flow, dopamine receptors or gray matter volume, is studied in a cross-sectional manner. To our knowledge, there are no longitudinal analyses of brain function in obesity. In the present study, we were interested in the effects of weight gain and loss on human brain gray and white matter structure. Phospholipids are major components of neuronal and glial membranes, and participate in membrane remodeling and synthesis and signal transduction (13). The metabolism of the brain phospholipids is a dynamic process, which is affected by, for instance, the plasma concentration of free fatty acids. About 5% of nonesterified fatty acids are extracted from blood as it passes through rat brain, and extraction is independent of cerebral blood flow (13). Obesity is accompanied by the excess of free fatty acids in the plasma yielding to fat accumulation into adipocytes as well as into several organs. Therefore, we hypothesized that obese individuals might have differences in the brain fat metabolism and increased fat accumulation in the white matter, and this could have a reflection to the white matter volume.
The study was designed to consist of two parts: 1) a conventional cross-sectional brain comparison of obese and lean individuals and correlation analyses, and 2) a longitudinal follow-up of individual brains after a large rapid body weight loss. In part I, we investigated differences in brain regional gray and white matter volumes between lean and obese individuals. In part II, a subpopulation of obese individuals (n = 16) from the first part started a controlled very low-calorie diet (VLCD) for 6 wk, and the second brain scan followed after they had successfully reduced their weight on average by 12%. Dieting is known to have a beneficial effect on, for example, insulin sensitivity and plasma lipids in obese individuals (14), and weight reduction is also associated with a reduction in plasma leptin level (15). Brain, as a lipid-rich tissue, could also be affected by weight loss. We tested whether weight reduction could decrease brain volume in obese subjects in line with the reduction of fat in the whole body.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Part I.
The study included 30 obese (12 men and 18 women) and 16 lean (eight men and eight women) subjects. Lean individuals were defined as those with BMI less than 26 kg/m2 and obese individuals those with BMI greater than 27 kg/m2. Patients with eating disorder, metabolic diseases, cardiovascular disease, previous or present abnormal hepatic or renal function, anemia, or oral corticosteroid treatment were excluded. The main physical and metabolic characteristics of the subjects are presented in Table 1
. Obese individuals had significantly higher fasting plasma concentrations of glucose, insulin, leptin, and free fatty acids (Table 1). Written informed consent was obtained after explaining the purpose and potential risks of the study to the subjects. The study protocol was approved by the ethical committee of Southwest Finland Healthcare District and conducted according to the principles of the declaration of Helsinki.
|
). All daily meals were replaced by VLCD products for a period of 6 wk (Nutrifast; Leiras Finland, Helsinki, Finland) (2.3 MJ, 4.5 g fat, 59 g protein, and 72 g carbohydrate per day). Added to Nutrifast, subjects drank daily at least 2 liters of water or sugar-free soft drinks. No changes in physical activity were allowed. The diet was regularly controlled by a study nurse with nutritional expertise. After the diet there was a 1-wk recovery period with normocaloric diet to avoid catabolic state. MRI, anthropometric measurements, and laboratory assessments were repeated after the recovery period. Adipose tissue masses in abdominal area were assessed at the level of L2/L3 intervertebral disk before and after dieting using a standardized MRI-based method (16).
|
MRIs were obtained with Philips Gyroscan Intera 1.5 T CV Nova Dual scanner (Philips, Best, The Netherlands). Whole-brain T1-weighted three-dimensional fast field echo (FFE) data set was acquired in transverse plane (time repetition = 25 msec, time echo = 5 msec, flip angle = 30°, number of excitations (NEX) = 1, and field of view = 256 x 256 mm2), yielding at least 160 contiguous slices through the head. Images were transferred to a personal computer and converted to Analyze format using MRIconvert (http://lcni.uoregon.edu/
jolinda/MRIConvert/) and analyzed using SPM2 (Wellcome Department of Cognitive Neurology, London, UK; http//www.fil.ion.ucl.ac.uk/spm) and Matlab 6.5 (The MathWorks, Natick, MA). The optimized VBM protocol was applied to the images (17). Before VBM analysis, a clinical visual evaluation of the MR images was performed by an experienced neuroradiologist (R.P.). One elderly lean subject had a small lacunar infarct near the left insular cortex; no other clinically significant findings were observed in any of the subjects.
Templates
Customized templates were created to facilitate optimal normalization and segmentation of the MRI scans of obese and lean subjects. Template generation was performed using a toolbox extension to the segmentation algorithm of SPM2 (Christian Gaser, University of Jena, Jena, Germany; http://dbm.neuro.uni-jena.de/vbm/). Templates were constructed because the contrast of the present MRI scans might differ from the existing template, the demographics of the present subject population might differ from those used to generate the existing template, and each scanner introduces specific nonuniformities and inhomogeneities. Templates were therefore constructed in an attempt to reduce the potential for bias toward one group during spatial normalization (18).
Optimized VBM
After the creation of study-specific templates, the optimized protocol was applied to the original data (17). The optimized VBM protocol improves spatial normalization by the use of gray matter images and a gray matter template rather than anatomical T1 images. Optimized protocol also involves cleaning up of partitions by applying morphological operations and the optional modulation of partitions to preserve the total amount of signal. Because we were mainly interested in the volumetric differences in obesity rather than differences in concentrations, we chose to use additional modulation in our VBM protocol. Cut-off of spatial normalization was 25 mm, medium nonlinear regularization was used, and the protocol involved 16 nonlinear iterations. The modulated images were smoothed with a 12-mm full width at half maximum (FWHM) isotropic Gaussian kernel. In previous studies, optimized VBM has been properly validated, and the tissue classification technique used in VBM has yielded highly reproducible results (17).
Biochemical analyses
Plasma glucose concentration was determined in duplicate by the glucose oxidase method (Analox GM9 analyzer; Analox Instruments, London, UK). Glycosylated hemoglobin was measured by fast protein liquid chromatography (MonoS; Pharmacia, Uppsala, Sweden). Plasma insulin concentration was measured by double-antibody fluoroimmunoassays (Autodelfia; Wallac, Turku, Finland). Serum total cholesterol and high-density lipoprotein cholesterol were measured using standard enzymatic methods (Roche Molecular Biochemicals, Mannheim, Germany) with a fully automated analyzer (Hitachi 704; Hitachi, Tokyo, Japan). Serum low-density lipoprotein cholesterol was calculated according to Friedewald equation (19). Serum free fatty acids were determined by an enzymatic method (acyl-CoA synthase-acyl-CoA oxidase peroxidase method; Wako Chemicals, Neuss, Germany). Plasma leptin was analyzed with RIA (Linco, St. Charles, MO). In part I, data from blood tests, waist circumference, and waist to hip ratio were not available from four lean subjects and leptin data were missing from one obese subject.
Statistical analysis
The smoothed, modulated data were analyzed using statistical parametric mapping (SPM2) using the general linear model. Volumetric changes were tested by analysis of the modulated data. Because during modulation we incorporated the correction for volume change induced by spatial normalization, it was appropriate to include total intracranial volume (TIV) as a covariate to remove any variance due to differences in head size. TIV was calculated using the get_globals function of SPM2. The number of voxels in each of the tissue compartments was calculated and summed.
For the statistical analysis, voxels with a gray or white matter value less than 0.1 were excluded to avoid possible edge effects around the border between gray and white matter. The differences between obese and lean subjects were tested with analysis of covariance using sex and TIV as confounding covariates. Correlation analyses between physical/metabolic measures and brain white/gray matter volumes were performed with multiple regression analysis using sex and TIV as confounding covariates. The effects of dieting on white and gray matter were tested with paired t tests within SPM2. Correlation analyses for part II were performed with simple regression by calculating delta images (scan 1 – scan 2) and delta values for physical and metabolic measures. Height threshold in SPM analyses was set at P = 0.01 and extent threshold 50 voxels. The MNI space utility (Sergey Pakhomov, Russian Academy of Sciences, St. Petersburg, Russia) extension of SPM was used to interpret SPMs and determine appropriate anatomical labels. The level of statistical significance was set at voxel level-corrected P < 0.01 [corrected for multiple comparisons using false discovery rate (FDR)]. Data are presented as means (SD), unless indicated otherwise.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Greater relative brain white matter volumes were observed in obese subjects, compared with lean subjects, in several regions: superior, middle, and inferior temporal gyri; fusiform gyrus; parahippocampal gyrus; brain stem; and cerebellum (all findings bilaterally) (Fig. 1, A and B). In SPM brain map, the contiguous voxels with significant group difference in the relative white matter volume fused forming two clusters [35,901 voxels, peak voxel (at 6 mm, –23 mm, –29 mm), FDR corrected P = 0.006; 16,228 voxels, peak voxel (at –52 mm, –18 mm, –28 mm), FDR corrected P = 0.006] (Table 3). Lean subjects did not have greater white matter volumes, compared with obese subjects, in any brain region. The mean (SD) global white matter volume was 0.486 liters (0.063) in obese subjects and 0.458 liters (0.044) in lean subjects (TIV corrected P = 0.14).
|
|
There were no statistically significant differences in gray matter volumes between obese and lean subjects, although lean subjects had trend-level greater gray matter volumes in certain brain regions such as the cingulate gyri, superior and medial frontal gyri, brain stem, and the cerebellum (FDR corrected P = 0.025). The mean (SD) global gray matter volume was 0.752 liters (0.070) in obese subjects and 0.734 liters (0.074) in lean subjects (TIV corrected P = 0.79).
Effect of dieting (part II)
Six weeks of VLC dieting induced a highly significant weight reduction in all obese subjects [11 (3.4) kg, range 6.6–19 kg] and a decrease of sc and visceral fat masses in the abdominal area (Table 2
). The weight loss was associated with decreases in blood pressure, cholesterol, leptin, and glycosylated hemoglobin (Table 2), but no significant changes were seen in fasting plasma glucose and insulin concentrations.
Dieting reduced global white matter volume: 0.498 liters (0.051) before and 0.488 liters (0.048) after dieting (P = 0.002). Regional white matter volumes decreased in the left temporal lobe (fusiform gyrus; parahippocampal gyrus; and inferior, medial, and superior temporal gyri) [12,026 contiguous voxels, peak voxel (at –46, –6, and –31 mm), FDR corrected P = 0.009] (Fig. 2, A and B, and Table 3). In addition, the white matter reduction reached trend-level significance in several other clusters (FDR corrected P value between 0.03 and 0.07). None of the brain structures showed increases in the white matter volumes after dieting. Changes in global or regional gray matter were nonsignificant (P > 0.28).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Increased white matter volume has recently been detected in the vicinity of striatum of severely obese subjects (BMI 39.4) (12). In that study, gray matter volume was lower in obese subjects in several brain regions and an inverse association was seen between BMI and gray matter volume in the left postcentral gyrus in obese but not lean subjects. We did not detect significant gray matter differences between lean and obese subjects, although there were several brain areas in which obese subjects showed trend-level lower gray matter volumes than lean individuals (P = 0.025). Because subjects in the present study were less obese, compared with those in the earlier study (12), it is possible that more severe chronic obesity influences gray matter together with white matter.
In the present study, greater white matter volumes in the obese group were seen in basal bilateral regions, and the white matter expansion was associated with increased waist to hip ratio (gender corrected) but not BMI. Numerous studies have demonstrated that the distribution, rather than the amount, of body fat is related to metabolic changes (20, 21, 22). Waist to hip ratio seems to be better than BMI in assessing the risk for cardiovascular diseases and metabolic abnormalities in pre- and postmenopausal women (23). In addition, there is evidence by a recent large study (n = 27,007), that waist to hip ratio adds prognostic information on the cardiovascular risk in women at all levels of BMI and men with normal weight (24). In the present study, we saw a strong sex-corrected positive correlation between waist to hip ratio and white matter volume. This suggests that cerebral white matter may be more related to the accumulation of abdominal fat rather than body fat per se. Within the brain, however, the large total size of the clusters suggests that the relationship could be more general and less region specific. One interpretation could be that the increase in visceral fat is associated with the accumulation of fat in central myelin throughout the brain.
The current study also demonstrated a positive association between serum free fatty acid concentration and brain white matter volume in left temporal and occipital lobes in obese subjects, and obese subjects showed significantly higher concentrations of serum free fatty acids. Therefore, an explanation for the white matter differences in obesity could be abnormal lipid metabolism and accumulation in the brain. Previous studies with rodents have shown that the hypothalamic metabolism of fatty acids can modify feeding behavior and that the hypothalamic levels of long-chain fatty acyltransferase-coenzyme A can be increased by enhanced esterification of circulating or central lipids and/or by the local inhibition of lipid oxidation (25). The findings of the present study together with the results of animal studies suggest that fatty acid excess in obesity could result in pathological lipid metabolism in the brain, and this might have an influence on both brain white matter volume and brain function in the regulation of food intake. On the other hand, although the detected volume differences are in line with the study hypothesis, they do not directly prove that obesity is accompanied by fat accumulation in the brain. To confirm the hypothesis, future studies should provide more evidence that brain fatty acid metabolism is changed in obesity in humans.
It should be noted that although VBM can accurately detect regional volume changes, it does not provide any clues about the causative agent. White matter volume expansion in obesity is therefore not necessarily related to adipose tissue or myelin. In theory, individual hydration status could influence white matter volume because the lack of fluid intake for 16 h has been reported to decrease brain volume by 0.55% (26). However, lean and obese subjects followed identical fasting instructions prior the MRI scan and had normal (and similar) blood hematocrit values (mean 41%in the lean group, 42% in the obese group). Second, the findings were regionally selective and located predominantly in basal brain regions. In the intervention part, obese subjects underwent 1-wk normocaloric diet before the second MRI scan, which presumably normalized the fluid equilibrium. They also had normal blood hematocrit values before and after dieting (39 vs. 37%, respectively), suggesting that there were no significant changes in hydration status.
Dieting is known to improve insulin sensitivity and plasma lipid profile (14), hence preventing comorbidities associated with obesity. Effects of dieting on brain structure have not been studied previously. The white matter decrease induced by dieting in part II of the present study combined with the results of part I suggests that both chronic weight gain and rapid weight loss are linked to brain white matter. It was beyond the scope of this study to investigate the clinical significance of white matter volume changes in obesity. We cannot answer whether the reported brain structural changes are primary or secondary. However, on the basis of the localization of the findings in the myelin rich white matter (with the preservation of gray matter), we speculate that the demonstrated changes are secondary, reflecting fat accumulation. We failed to correlate white matter changes in dieting to changes in physical or metabolic measures, although a trend-level relationship between white matter reduction and abdominal visceral fat loss (in relation to sc fat) was seen. The results do not, however, suggest that central white matter change is an isolated event in weight gain and weight loss but rather that the studied subpopulations of 30 obese subjects (part I) and 16 obese subjects (part II) might have been too small for correlation analyses with the large variation. Finally, due to possible registration errors and smoothing, it is conceivable that although the great majority of the observed differences reflect white matter changes, it cannot be excluded that some gray matter signal is included in the total signal.
To conclude, we have presented data that indicate that obesity is associated with brain white matter volume expansion. The most significant relationship was seen between waist to hip ratio and white matter. In the longitudinal analysis, the results demonstrated brain white matter shrinkage after short-term dieting. Although epidemiological studies have shown that the risk of degenerative brain diseases is increased in obese individuals, the clinical significance of the here presented white matter changes in obesity and dieting remains unclear. Future studies could be designed to investigate the role of central fat accumulation and white matter abnormalities in the neuropathogenesis of degeneration.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Information: L.T.H., A.V., R.P., N.K., J.O.R., P.N., and V.K. have nothing to declare.
First Published Online May 29, 2007
Abbreviations: BMI, Body mass index; FDR, false discovery rate; fMRI, functional MRI; MRI, magnetic resonance imaging; PET, positron emission tomography; TIV, total intracranial volume; VBM, voxel-based morphometry; VLCD, very low-calorie diet.
Received November 13, 2006.
Accepted May 23, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I. Soreca, C. Rosano, J. R. Jennings, L. K. Sheu, L. H. Kuller, K. A. Matthews, H. J. Aizenstein, and P. J. Gianaros Gain in Adiposity Across 15 Years is Associated With Reduced Gray Matter Volume in Healthy Women Psychosom Med, June 1, 2009; 71(5): 485 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. H. Coker, P. E. Hogan, N. R. Bryan, L. H. Kuller, K. L. Margolis, K. Bettermann, R. B. Wallace, Z. Lao, R. Freeman, M. L. Stefanick, et al. Postmenopausal hormone therapy and subclinical cerebrovascular disease: The WHIMS-MRI Study Neurology, January 13, 2009; 72(2): 125 - 134. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
