American Journal of Respiratory and Critical Care Medicine

Obstructive sleep apnea is associated with abnormalities in neuropsychologic function, and defects in respiratory control may contribute to pathogenesis. Abnormalities may be reflected in structural brain changes. Twenty-seven male untreated patients with severe sleep apnea without comorbidities, and 24 age-matched control subjects, had T1-weighted brain imaging in a high-resolution magnetic resonance scanner. Twenty-three patients with sleep apnea had repeat imaging after 6 months of continuous positive airways pressure treatment. No areas of gray matter volume change were found in patients using an optimized voxel-based morphometry technique, at p < 0.05 adjusted for multiple comparisons (despite the method being sensitive to changes in gray matter fraction of 0.17 or less in all voxels). Furthermore, no differences were seen in bilateral hippocampal, temporal lobe, or whole brain volumes, assessed by manual tracing of anatomical borders. No longitudinal changes were seen in gray matter density or regional volumes after treatment, but whole brain volume decreased slightly. We have found no gray matter volume deficits nor focal structural changes in severe obstructive sleep apnea. Whole brain volume decreases without focal changes after 6 months of continuous positive airways pressure treatment.

Obstructive sleep apnea (OSA) is a disorder characterized by repetitive collapse of the upper airway during sleep, with resulting intermittent hypoxemia, hypercapnia, surges in sympathetic nervous system (SNS) activity with arousals, and cumulative sleep fragmentation. OSA has been associated with impairment in a range of domains of neuropsychologic function (1, 2). These decrements lead to reduced quality of life, impaired work performance, and increased risk of vehicular and industrial accidents (3, 4).

Changes in the regulation of autonomic function have also been described in OSA (5, 6), as have changes in neuroendocrine function (7, 8). It is unclear to what extent these changes are sequelae to sleep disruption, hypoxemia, and repetitive SNS surges, or primary alterations in neural function.

The pathogenesis of OSA remains incompletely defined. Although upper airway anatomy and collapsibility are important, it is clear that central control of upper airway musculature and of ventilation play a role (913).

There has been little exploration of the pathologic basis for these differences in central respiratory, neuroendocrine, or autonomic control nor for the neuropsychologic impairment seen in OSA. Few magnetic resonance imaging studies have characterized brain structure and metabolism in OSA. There is evidence that OSA is associated with alterations in metabolite concentrations in frontal white matter, whose severity correlates with apnea–hypopnea index (AHI) (14). It has also been reported that temporal lobe volume is decreased in healthy subjects who frequently traverse several time zones in long-haul flights (15). Furthermore, rats exposed to intermittent hypoxia during sleep demonstrate impaired spatial learning, and increased apoptosis in the cortex and CA1 region of the hippocampus (16). A recent study reported widespread deficits in gray matter volume including the temporal lobe and right hippocampus in 21 patients with OSA of varying severity compared with 21 age-matched control subjects (17). These authors used a voxel-based morphometry (VBM) technique—without, however, using the optimized form of VBM, and only reporting results at an uncorrected threshold. They did not exclude patients with comorbidities, which might also have influenced brain structure, and did not assess whether these changes were reversible on treatment with continuous positive airways pressure (CPAP).

We hypothesized that (1) structural differences in the brain would exist between patients with severe OSA and age-matched control subjects, whether primary or as a result of sleep fragmentation or hypoxia; and (2) structural differences would exist in a group of patients with severe OSA between scans taken before and after 6 months of therapy with CPAP. We reasoned that if any changes were reversible with CPAP, this would suggest they were secondary to sleep deprivation, repetitive nocturnal hypoxemia, or disturbances in autonomic activity.

Further methodologic details are in the online supplement.


Twenty-seven male patients with severe OSA and 24 age-matched control subjects were recruited. Patients were recruited from the clinic population of the Department of Respiratory and Sleep Medicine, Austin Health. Inclusion criteria: (1) AHI greater than 30 and (2) at least 15% of total sleep time spent at SaO2 less than 90%. Exclusion criteria for both patients and controls were: (1) respiratory disease (historically or abnormal spirometry) and (2) history of cerebrovascular or ischemic heart disease, diabetes mellitus, central nervous system disorders (neurodegenerative diseases, epilepsy, head injury, psychosis, hypothyroidism, current depression), alcohol or illicit drug abuse, or current intake of psychoactive medications.

Controls were recruited by newspaper advertisement. In addition to the exclusion criteria above and age matching, no control subject had a history of snoring or other sleep complaint. All had full screening polysomnography (PSG) to exclude OSA (defined as AHI greater than 5 if airflow measured by thermistor, AHI greater than 15 by nasal cannula) (18, 19).

All subjects gave informed written consent. The study was approved by the institutional Research and Ethics Committee.

Protocol and Data Analysis

T1-weighted images were acquired in the morning using a 3-Tesla scanner (GE Horizon LX, Milwaukee, WI). A fast spoiled, gradient-recalled echo at steady state sequence (time to repeat/time to echo 8.9/1.9 milliseconds, flip angle 20°, matrix size 256 × 256, and a field of view of 25 ×18.75 cm) with contiguous coronal slices of 1.5-mm thickness was used. Patients with OSA had a baseline scan (S1) after diagnosis but before commencement of treatment. In 23 patients, scans were repeated (S2) after 6 months of CPAP treatment (Autoset T, ResMed, Australia).

Region-of-interest analysis.

In the region-of-interest (ROI) approach, manual tracing of anatomical borders of defined structures was performed, followed by automated volume calculation (20). ROI assessment included whole brain, bilateral temporal lobe, and bilateral hippocampal volume measurements.


VBM is a fully automated, whole-brain approach that performs voxel-wise comparisons of regional tissue gray matter volumes between groups (21, 22). VBM has previously been used to assess brain morphology in various neurologic diseases (2325). Briefly, VBM involves warping (“spatially normalizing”) the images into the same stereotactic space, extracting the gray matter from the normalized images, smoothing, and finally performing statistical analysis to localize group differences. The analysis output is a statistical parametric map (SPM) displaying voxels in which the gray matter volume shows a significant difference. VBM has been undergone a number of refinements in recent years to increase its utility in the detection of volume changes (22). In this study, we have used this optimized VBM approach. The most recent version of the software package (SPM2; Wellcome Department of Cognitive Neurology, London, UK) was used for the analysis.

Statistical Analysis

Nonparametric tests were used in ROI analysis. Mann-Whitney U tests were used to compare mean volumes in the control and patient groups. Wilcoxon signed rank tests were used to assess changes over time in patients (S1 to S2). Level of significance was set at p < 0.05.

VBM analysis.

Baseline scans of patients and controls were compared using analysis of covariance, with age as a covariate. For comparisons between S1 and S2 in patients, repeated measures t tests were used. Inferences about regionally specific gray matter volume differences were made using a significance threshold level of p < 0.001, uncorrected for multiple comparisons, as well as a more stringent threshold of p < 0.05 with a multiple-comparison adjustment for false delivery rate (26). A sensitivity map was also constructed that indicated the minimum change in fractional gray matter volume required in each voxel for that voxel to reach significance in our analysis.

Table 1

TABLE 1. Baseline characteristics

OSA (n = 27)

Control Subjects (n = 24)
Age, yr45.7 ± 10.143.3 ± 9.4
BMI, kg/m233.2 ± 4.725.3 ± 2.8
FEV1, % pred102.1 ± 16.1104.3 ± 12.6
FVC, % pred100.5 ± 15.1106.2 ± 13.0
% TST < 90%39.8 ± 20.40.02 ± 0.06
AHI71.7 ± 17.05.9 ± 4.7
13.1 ± 3.9
5.5 ± 3.2

Definition of abbreviations: AHI = apnea–hypopnea index; BMI = body mass index; ESS = Epworth sleepiness score (a measure of subjective daytime sleepiness); OSA = obstructive sleep apnea; % TST < 90% = percentage of total sleep time at diagnostic polysomnography spent at oxygen saturations less than 90%.

All values are expressed as mean ± SD.

contains details of subject demographics, anthropometrics, and PSG results. Twenty-three patients with OSA had repeat scans after 6 months of CPAP treatment. Mean ± SD objectively recorded “time at pressure” CPAP compliance was 5.8 ± 1.7 hours per night.

ROI Analysis

Three patients with OSA were excluded from this analysis for technical reasons (image artifacts in the hippocampal regions). The absolute values of the hippocampal and temporal lobe volumes in the control subjects and in patients in S1 and S2 are shown in Table 2

TABLE 2. Brain volumes determined by region-of-interest analysis


OSA (S1)
 (n = 24)

 (n = 24)

p Value*

OSA (S1)
 (n = 21)

OSA (S2)
 (n = 21)

p Value
Hippocampus, mm32,810 ± 2792,910 ± 3300.32,798 ± 2882,762 ± 3750.6
Temporal lobe, cm371 ± 772 ± 60.771 ± 870 ± 80.6
Whole brain, cm3
1,331 ± 122
1,334 ± 91
1,336 ± 124
1,276 ± 113

*Comparing patients with OSA with control subjects at baseline (Mann-Whitney U tests).

Comparing S1 with S2 in the 21 patients with OSA analyzed using region of interest who had repeat scans (Wilcoxon signed-rank tests).

Definition of abbreviations; OSA = obstructive sleep apnea; S1 = baseline scan, before commencement of treatment with continuous positive airways pressure (CPAP); S2 = scan after 6 months' CPAP treatment.

All values are expressed as mean ± SD.

. There was no difference between volume measurements in patients and control subjects, either in absolute volumes or when normalized as a percentage of whole brain volumes. There was no change either in absolute or normalized temporal lobe or hippocampus volumes between S1 and S2 in patients. There was a slight decrease in whole brain volumes between S1 and S2 (S1: 1,336 ± 124 cm3; S2: 1,276 ± 113 cm3; p = 0.01). This is a volume decrement of approximately 4% over this period (Table 2).


Two patients and one control subject were excluded from this analysis because of technical factors (failures of the normalization step in SPM analyses), leaving 25 patients and 23 control subjects. Assessed at significance threshold of p < 0.001, uncorrected for multiple comparisons, scattered areas of gray matter deficit were observed in patients as compared with control subjects, including in the posterior and mesial temporal lobe bilaterally and in the left insular region (Figure 1)

. None of these regional differences remained significant when adjusted for multiple comparisons using the false discovery rate. The opposite contrast, assessing areas of gray matter volume increase in patients with OSA as compared with control subjects, suggested changes in the right basal ganglia and less prominently in scattered frontal lobe and parietal lobe areas (Figure 2). None of these differences remained significant when adjusted for false discovery rate.

No gray matter volume deficits were detected between S1 and S2, even when assessed at a threshold of p < 0.001, uncorrected for multiple comparisons. The opposite contrast, describing an increase between S1 and S2, suggested an increase in gray matter volume in the area of the perirhinal gyrus on the left side. Again, this difference did not survive adjustment for false discovery rate.

Post hoc assessment of the data was performed using SPM99—as used in (17), but with the optimized protocol (22)—to assess whether the lack of difference between patient and control groups could be explained by a methodologic effect related to the version of SPM. However, this assessment also showed no significant gray matter deficits or increases in our patient group and no change after treatment with CPAP.

Finally, a sensitivity map was constructed to quantify the power of our study to detect a difference between patients and controls (Figure 3)

. The map demonstrates that, with our methods and subjects, changes in a voxel's gray matter fraction of 0.17 or less, depending on brain region, would have been detectable.

The present study, investigating volumetric changes using two different measurement methods in patients with severe OSA found surprisingly little indication of volume deficits in patients as compared with controls. Using VBM, no areas of gray matter volume deficit or increase were found in patients with OSA compared with control subjects when adjustment was made for multiple comparisons. Using ROI analysis, no differences were found in whole brain, temporal lobe, or hippocampal volumes. Treatment with CPAP over 6 months did not result in focal changes in gray matter concentration or in hippocampal or temporal lobe volume changes, although there was a small but significant decrease in whole brain volume.

We selected a group of subjects with very severe OSA with a specific inclusion criterion being oxygen desaturation to less than 90% for greater than 15% of total sleep time on diagnostic PSG. All patients were imaged before commencing treatment. We were careful to exclude any comorbidity in both our patients with OSA and control subjects that might be associated with altered brain structure. All control subjects were screened by full-attended PSG to exclude OSA. In addition, we used a scanner with high magnetic field strength (3 Tesla), which is typically associated with increased signal-to-noise ratio compared with that at lower field strength. All of these design elements should maximize the sensitivity of finding a true difference between patients with OSA and control subjects. Despite this, gray matter volume deficits in OSA were only detected in a few small areas using VBM, and these differences did not survive a threshold of statistical significance adjusted for false discovery rate.

We performed region of interest analysis in two focal brain areas: the hippocampus and the temporal lobes. Atrophy in both areas has been previously reported in conditions causing chronic sleep disruption (15, 17) and in rats exposed to intermittent hypoxia (16). The ROI analysis was confined to a small number of areas in which a volume change was hypothesized, reducing the number of comparisons performed. Despite this increased statistical sensitivity, no volume deficit in severe OSA was detected in the hippocampus or temporal lobes.

The VBM analysis investigated changes in gray matter distribution in the whole brain. This method was fully automated, but its quality relied on the accurate segmentation of the different tissue types and on the normalization paradigms used to bring the individual brains into a standard format allowing group comparisons. We used an optimized form of the VBM method (22), which should have increased sensitivity to detect true areas of volume change, and the most recent version of the SPM package (SPM2), which has improved utilities to facilitate the segmentation and normalization steps compared with SPM99. These advantages, used in this study, augmented the ability of this technique to detect true effects between groups of patients, particularly in subcortical areas, where tissue segmentation has traditionally been difficult. The selection of a reasonable statistical threshold is also important in VBM. Several publications have detailed findings based on a lenient threshold of p < 0.001, uncorrected for multiple comparisons (27, 28). These reports can only validly be interpreted as trends. There are various methods whereby a correction can be made for the multiple comparisons made in the analysis. The SPM99 correction for multiple comparisons was based on a correction for family-wise error (i.e., the probability of a single false positive anywhere) using the theory of Gaussian random fields (29). Recently, it has been recognized that adjustment for false discovery rate (i.e., the estimated fraction of false positive voxels) provides adequate control for multiple comparisons, at typically a more lenient threshold than family-wise error correction (26). Both methods are available in SPM2. We chose to use a threshold of p < 0.05 with a multiple comparison adjustment for false discovery rate; however, none of our comparisons showed a significant effect at this level. We therefore suggest that, using current optimized techniques, no reproducible gray matter differences between patients with OSA (without cardiovascular comorbidities) and control subjects are detectable.

Because previous studies (17) have demonstrated changes in patients with OSA using an earlier version of the SPM package (SPM99), we performed a post hoc analysis of our data using SPM99 to exclude an effect of methodology related to the software version used. This analysis also did not demonstrate a significant difference between patients and control subjects. However, in contrast to the previous study by Macey and colleagues (17), we performed all our analyses with an optimized VBM protocol (22), which may at least partly explain the different results obtained. We believe that the methods used in this study have increased sensitivity and specificity to true disease-related effects on brain volumes.

Because we are essentially reporting a negative finding, it is important to know the power of our study to detect an effect. Our sensitivity map (Figure 3) demonstrates that in all voxels we had sufficient power to detect changes in gray matter fraction of 0.17 or less. To put this into perspective, in areas of any gray matter loss, one would expect large changes in fractional gray matter in the voxels at the edge of the region, because voxels that initially contained mostly gray matter (i.e., values between 0.5 and 1) changed to contain little gray matter (i.e., values approaching 0). Therefore, our sensitivity to changes in gray matter concentration should be more than adequate to detect subtle changes.

Finally, we acquired scans in 23 patients with OSA after 6 months of treatment with CPAP with objectively monitored mean compliance 5.8 ± 1.7 hours per night. No focal differences were found in VBM or ROI analysis between S1 and S2. We did find a small (4%) decrease in whole brain volume using ROI between S1 and S2. This may reflect an effect of OSA on brain structure, or perhaps an effect on cerebral blood volume or whole brain water content from reversal of OSA-induced changes in autoregulation when on CPAP (30). Further studies are required assessing longitudinal changes in brain volume in patients with OSA both before and after treatment with CPAP to confirm this finding.

Our findings are in contrast to the recent report by Macey and colleagues (17), which showed widespread gray matter volume deficits in patients with OSA using VBM. These deficits included the following regions: bilateral anterosuperior frontal gyri, left ventrolateral frontal cortex, bilateral lateral prefrontal cortices, bilateral inferior temporal gyri, bilateral parahippocampal gyri, left anterior cingulate, right hippocampus, an area surrounding the caudal extent of the lateral sulcus, and bilateral quadrangular lobule of cerebellum.

There were several design differences between the current study and that of Macey and colleagues (17). Macey and colleagues recruited 21 patients with OSA who had a range of severity of OSA (mean AHI 38 ± 24) with no oxygen desaturation inclusion criteria. Screening polysomnography was not performed on their 21 control subjects, and 12 patients with OSA had already received treatment for varying periods—11 with CPAP and with a mandibular splint. These population differences from the present study would all be expected to result in a greater likelihood of finding true differences between the two groups in the current study. A further difference, however, is that Macey and coworkers did not rigorously exclude subjects with comorbidities likely to have an impact on brain volumes. Their subjects with OSA suffered from a variety of medical conditions, including neurologic and psychiatric disorders, and four were taking neurologic drugs. It is possible, therefore, that the findings of these investigators were not specific to OSA and reflected comorbidities in their OSA group.

On the other hand, it is conceivable that by meticulous exclusion of subjects with these comorbidities associated with OSA, we excluded the patients who were likely to show changes in brain structure. However, we were interested in the changes attributable to OSA per se, rather than those from the cardiovascular consequences of the condition. Similarly, it could be postulated that we selected a “super-sample” of patients with OSA resistant to the neurologic and cardiovascular complications of OSA (and obesity). We know of no evidence for the existence of such a group of patients, but we cannot entirely discount this possibility. Finally, it is possible that the deleterious effects of OSA on cerebral structure may only occur after decades of untreated disease. The group studied by Macey and colleagues (17) had a mean age (49 ± 11 years) similar to the current study (46 ± 10 years). We would therefore have expected to be able to replicate their results in our much more severe, untreated group. However, we cannot state with certainty that, if our patients had remained untreated for another one or two decades, we would not then have been able to find differences from a group of age-matched controls.

The study by Macey and colleagues (17) reported findings at a threshold of p < 0.001, uncorrected for multiple comparisons. At a similar threshold, the present study also found some areas of gray matter volume deficit in patients with OSA, though they were not as widespread as in Macey's study. In our view, adjustment for the multiple comparisons performed in a whole brain VBM study is essential to allow generalizable conclusions about the assessed disease. Such correction measures were recommended by the group that originally described VBM (21), and were performed in two of the three studies cited by Macey and colleagues in support of their statistical approach (31, 32).

Recently, another group has published a small study at 1.5 Tesla demonstrating a gray matter volume deficit in the left hippocampus in seven newly diagnosed patients with OSA (33). These differences persisted in an expanded group reported in abstract form (34). The authors had an a priori hypothesis of hippocampal damage; therefore, they only corrected for multiple comparisons made within a 15-mm radius of the center of this ROI. No changes were found elsewhere in the brain when correcting for multiple comparisons across the whole brain. They did not report whether the hippocampal changes persisted when the more rigorous statistical conditions were applied.

Davies and colleagues (35) examined 45 patients with OSA and 45 age-matched controls and failed to find an increased incidence of subclinical cerebrovascular disease in the subjects with OSA. We believe the results of Morrell and colleagues (33, 34) and Davies and colleagues (35) are largely in agreement with those of the current study.

It is important to state that the results of the current study do not indicate that there are no changes in neural or psychologic function in OSA patients. Rather, we suggest that the pathologic basis of these deficits is not gray matter loss as measurable with the techniques used or change in absolute volume of whole brain or our ROIs. Further investigation of reported changes in frontal white matter metabolite concentrations (14) would appear warranted.

In conclusion, the present study documents no significant gray matter volume deficits in a group of patients with severe OSA who were carefully screened for comorbidities and scanned before initiation of treatment with CPAP. Furthermore, there were no changes in gray matter distribution after 6 months of CPAP therapy. This indicates that, even in severe OSA, there is no evidence of marked and permanent structural brain damage.

1. Reimer MA, Flemons WW. Quality of life in sleep disorders. Sleep Med Rev 2003;7:335–349.
2. Beebe DW, Gozal D. Obstructive sleep apnea and the prefrontal cortex: towards a comprehensive model linking nocturnal upper airway obstruction to daytime cognitive and behavioral deficits. J Sleep Res 2002;11:1–16.
3. Teran-Santos J, Jimenez-Gomez A, Cordero-Guevara J. The association between sleep apnea and the risk of traffic accidents: Cooperative Group Burgos-Santander. N Engl J Med 1999;340:847–851.
4. Lindberg E, Carter N, Gislason T, Janson C. Role of snoring and daytime sleepiness in occupational accidents. Am J Respir Crit Care Med 2001;164:2031–2035.
5. Harper RM, Macey PM, Henderson LA, Woo MA, Macey KE, Frysinger RC, Alger JR, Nguyen KP, Yan-Go FL. fMRI responses to cold pressor challenges in control and obstructive sleep apnea subjects. J Appl Physiol 2003;94:1583–1595.
6. Hilton MF, Chappell MJ, Bartlett WA, Malhotra A, Beattie JM, Cayton RM. The sleep apnoea/hypopnoea syndrome depresses waking vagal tone independent of sympathetic activation. Eur Respir J 2001;17:1258–1266.
7. Grunstein RR, Handelsman DJ, Lawrence SJ, Blackwell C, Caterson ID, Sullivan CE. Neuroendocrine dysfunction in sleep apnea: reversal by continuous positive airways pressure therapy. J Clin Endocrinol Metab 1989;68:352–358.
8. Luboshitzky R, Aviv A, Hefetz A, Herer P, Shen-Orr Z, Lavie L, Lavie P. Decreased pituitary-gonadal secretion in men with obstructive sleep apnea. J Clin Endocrinol Metab 2002;87:3394–3398.
9. Strohl KP. Pro-Con debate: con: sleep apnea is not an anatomic disorder. Am J Respir Crit Care Med 2003;168: 271–2; discussion 272–3.
10. Suratt PM, McTier R, Wilhoit SC. Alae nasi electromyographic activity and timing in obstructive sleep apnea. J Appl Physiol 1985;58:1252–1256.
11. Younes M, Ostrowski M, Thompson W, Leslie C, Shewchuk W. Chemical control stability in patients with obstructive sleep apnea. Am J Respir Crit Care Med 2001;163:1181–1190.
12. Insalaco G, Kuna ST, Catania G, Marrone O, Costanza BM, Bellia V, Bonsignore G. Thyroarytenoid muscle activity in sleep apneas. J Appl Physiol 1993;74:704–709.
13. Georgopoulus D, Giannouli E, Tsara V, Argiropoulou P, Patakas D, Anthonisen NR. Respiratory short-term poststimulus potentiation (after-discharge) in patients with obstructive sleep apnea. Am Rev Respir Dis 1992;146:1250–1255.
14. Kamba M, Inoue Y, Higami S, Suto Y, Ogawa T, Chen W. Cerebral metabolic impairment in patients with obstructive sleep apnoea: an independent association of obstructive sleep apnoea with white matter change. J Neurol Neurosurg Psychiatry 2001;71:334–339.
15. Cho K. Chronic ‘jet lag’ produces temporal lobe atrophy and spatial cognitive deficits. Nat Neurosci 2001;4:567–568.
16. Gozal D, Daniel JM, Dohanich GP. Behavioral and anatomical correlates of chronic episodic hypoxia during sleep in the rat. J Neurosci 2001;21:2442–2450.
17. Macey PM, Henderson LA, Macey KE, Alger JR, Frysinger RC, Woo MA, Harper RK, Yan-Go FL, Harper RM. Brain morphology associated with obstructive sleep apnea. Am J Respir Crit Care Med 2002;166:1382–1387.
18. Series F, Marc I. Nasal pressure recording in the diagnosis of sleep apnoea hypopnoea syndrome. Thorax 1999;54:506–510.
19. Norman RG, Ahmed MM, Walsleben JA, Rapoport DM. Detection of respiratory events during NPSG: nasal cannula/pressure sensor versus thermistor. Sleep 1997;20:1175–1184.
20. Briellmann RS, Berkovic SF, Jackson GD. Men may be more vulnerable to seizure-associated brain damage. Neurology 2000;55:1479–1485.
21. Ashburner J, Friston KJ. Voxel-based morphometry—the methods. Neuroimage 2000;11:805–821.
22. Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 2001;14:21–36.
23. Burton EJ, Karas G, Paling SM, Barber R, Williams ED, Ballard CG, McKeith IG, Scheltens P, Barkhof F, O'Brien JT. Patterns of cerebral atrophy in dementia with Lewy bodies using voxel-based morphometry. Neuroimage 2002;17:618–630.
24. Merschhemke M, Mitchell TN, Free SL, Hammers A, Kinton L, Siddiqui A, Stevens J, Kendall B, Meencke HJ, Duncan JS. Quantitative MRI detects abnormalities in relatives of patients with epilepsy and malformations of cortical development. Neuroimage 2003;18:642–649.
25. Kassubek J, Juengling FD, Kioschies T, Henkel K, Karitzky J, Kramer B, Ecker D, Andrich J, Saft C, Kraus P, et al. Topography of cerebral atrophy in early Huntington's disease: a voxel based morphometric MRI study. J Neurol Neurosurg Psychiatry 2004;75:213–220.
26. Genovese CR, Lazar NA, Nichols T. Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage 2002;15:870–878.
27. Busatto GF, Garrido GE, Almeida OP, Castro CC, Camargo CH, Cid CG, Buchpiguel CA, Furuie S, Bottino CM. A voxel-based morphometry study of temporal lobe gray matter reductions in Alzheimer's disease. Neurobiol Aging 2003;24:221–231.
28. Kassubek J, Juengling FD, Hellwig B, Spreer J, Lucking CH. Thalamic gray matter changes in unilateral Parkinsonian resting tremor: a voxel-based morphometric analysis of 3-dimensional magnetic resonance imaging. Neurosci Lett 2002;323:29–32.
29. Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ, Evans AC. A unified statistical approach for determining significant voxels in images of cerebral activation. Hum Brain Mapp 1996;4:58–73.
30. Franklin KA. Cerebral haemodynamics in obstructive sleep apnoea and Cheyne-Stokes respiration. Sleep Med Rev 2002;6:429–441.
31. Richardson MP, Friston KJ, Sisodiya SM, Koepp MJ, Ashburner J, Free SL, Brooks DJ, Duncan JS. Cortical grey matter and benzodiazepine receptors in malformations of cortical development: a voxel-based comparison of structural and functional imaging data. Brain 1997;120:1961–1973.
32. Salmond CH, Ashburner J, Vargha-Khadem F, Gadian DG, Friston KJ. Detecting bilateral abnormalities with voxel-based morphometry. Hum Brain Mapp 2000;11:223–232.
33. Morrell MJ, McRobbie DW, Quest RA, Cummin AR, Ghiassi R, Corfield DR. Changes in brain morphology associated with obstructive sleep apnea. Sleep Med 2003;4:451–454.
34. Morrell MJ, Giassie R, Simonds AK, Murphy K, McRobbie DW, Corfield DR. Obstructive sleep apnea is associated with changes in brain morphology in the hippocampus and parahippocampus [abstract]. Eur Respir J 2004;24:446s.
35. Davies CW, Crosby JH, Mullins RL, Traill ZC, Anslow P, Davies RJ, Stradling JR. Case control study of cerebrovascular damage defined by magnetic resonance imaging in patients with OSA and normal matched control subjects. Sleep 2001;24:715–720.
Correspondence and requests for reprints should be addressed to Dr. Fergal J. O'Donoghue, M.B., B.Ch., Ph.D., Institute for Breathing and Sleep, Austin Health, Heidelberg 3081, Australia. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record