White matter hyperintensities alter functional organization of the motor system
Article Outline
- Abstract
- 1. Introduction
- 2. Methods
- 3. Results
- 4. Discussion
- Disclosure statement
- Acknowledgements
- References
- Copyright
Abstract
Severe white matter hyperintensities (WMH) represent cerebral small vessel disease and predict functional decline in the elderly. We used fMRI to test if severe WMH impact on functional brain network organization even before clinical dysfunction. Thirty healthy right-handed/footed subjects (mean age, 67.8 ± 7.5 years) underwent clinical testing, structural MRI and fMRI at 3.0T involving repetitive right ankle and finger movements. Data were compared between individuals with absent or punctuate (n = 17) and early confluent or confluent (n = 13) WMH. Both groups did not differ in mobility or cognition data. On fMRI, subjects with severe WMH demonstrated excess activation in the pre-supplementary motor area (SMA), frontal, and occipital regions. Activation differences were noted with ankle movements only. Pre-SMA activation correlated with frontal WMH load for ankle but not finger movements. With simple ankle movements and no behavioral deficits, elderly subjects with severe WMH demonstrated pre-SMA activation, usually noted with complex tasks, as a function of frontal WMH load. This suggests compensatory activation related to disturbance of frontosubcortical circuits.
Keywords: White matter hyperintensities, Aging, Motor system, fMRI, SMA, Pre-SMA
1. Introduction
Some degree of age-related cerebral white matter hyperintensities (WMH) can be observed almost endemically on brain magnetic resonance imaging (MRI) scans in otherwise healthy elderly subjects. In particular more severe WMH (i.e., early confluent or confluent WMH according to the Fazekas scale (Fazekas et al., 1987) have been associated with small vessel disease. As they also demonstrate significant progression over time (Schmidt et al., 2003), WMH grades 2 and 3 are commonly considered as biologically more malignant (Schmidt et al., 2004).
In line with this notion, severe WMH have been associated with cognitive dysfunction, depression, disturbed micturition (Frisoni et al., 2007, Pantoni et al., 2005) and, most importantly, progression to disability (Inzitari et al., 2009) in the elderly. In the Leukoaraiosis and Disability (LADIS) study, the risk of transition to disability or death was more than 2-fold higher in the presence of severe WMH and 29.5% of individuals with severe WMH reached this end point after a mean follow-up of only 2.4 years. If and to what extent such transition is preceded by compensatory brain changes has not yet been investigated. Locomotor disability appears to represent an ideal focus of research in this context given both the correlation between increasing WMH severity and impaired walking and its impact on successful aging (Baezner et al., 2008).
Functional MRI (fMRI) greatly enhanced our understanding of central human motor control (Picard and Strick, 2001). Using simple motor paradigms, it allows characterizing the distinct functional neuroanatomy of upper and lower limb movements (Enzinger et al., 2008, Luft et al., 2002, Rotte et al., 2002). Ankle movements as a key component of gait (Capaday, 2002) have been used recently to assess the functional effects of gait training using fMRI (Enzinger et al., 2009).
We hypothesized that more severe WMH might impact on functional network organization of the brain in yet clinically intact elderly individuals. We here therefore used such motor paradigms to test for preclinical functional changes in relation to WMH severity.
2. Methods
2.1. Subjects
An a priori defined sample of 30 individuals was randomly selected from the Austrian Stroke Prevention Study (ASPS; Schmidt et al., 2003) by a blinded study nurse on the basis of records on their degree of WMH on a previous structural MRI brain scan. The Austrian Stroke Prevention Study is a single-center prospective follow-up study on the cerebral effects of vascular and genetic risk factors in the normal elderly population of Graz, Austria. A history of neuropsychiatric disease (including stroke) is considered an exclusion criterion from this study. Once patients suffer from a stroke or any other of the exclusion criteria this is considered an endpoint for the study. Thus none of the participants invited for this study had suffered from a stroke. Also, participants were selected in that they did not have a clinically silent infarct or lacunae on MRI of the brain.
Subjects were then invited to undergo structural and functional MRI, as well as gait and cognitive testing on the same day. The local ethics committee approved the study. The study cohort consisted of 21 female and 9 male right-hand and right-foot dominant subjects (according to the Edinburgh Handedness Inventory) with a mean age of 67.8 ± 7.9 years (range, 48–84). They had normal or corrected to normal vision and a mean duration of education of 10.8 ± 2.4 years. On the new fluid attenuated inversion recovery (FLAIR) scans obtained at 3.0 T, the WMH grading according to the modified Fazekas scale (Fazekas et al., 1987, Inzitari et al., 2009) had to be reassigned in 2 individuals from grade 2 to a grade 1. We then dichotomized the group according to their WMH grade, as group A: grades 0 or 1, and group B: grades 2 or 3. Further characteristics are given in Table 1.
Table 1. Sociodemographic data, risk factors, morphological MRI, and motor/mobility-related data for the entire cohort as well as for the 2 WMH severity subgroups
| Variable | Total cohort (n = 30) | WMH 0 and 1 (n = 17) | WMH 2 and 3 (n = 13) | p-value |
|---|---|---|---|---|
A Sociodemographic data | ||||
| Sex, m/f, n (%) | 9/21(30/70) | 4/13(23.5/76.5) | 5/8(38.5/61.5) | χ2 (1) = 0.782, p = 0.376 |
| Age, M ± SD, years | 67.80±7.50 | 63.59±6.26 | 73.31±5.06 | t(28) = −4.565, p < 0.001b |
| Education duration, years, Med (IQR) | 10.0(9.0–13.0) | 10.0(9.5–13.0) | 10.0(9.0–11.5) | U(n1 = 17, n2 = 13) = 104.5, |
| p = 0.788 | ||||
| BClinical profile | ||||
| Arterial hypertension, n (%) | 23(76.7) | 10(58.8) | 13(100) | χ2 (1) = 6.982, p = 0.010c |
| Diabetes mellitus, n (%) | 5(16.7) | 2(11.8) | 3(23.1) | χ2 (1) = 0.679, p = 0.628 |
| Current smoking, n | 0 | 0 | 0 | |
| Hypercholesterolemia, n (%) | 12(40) | 7(41.2) | 5(29.4) | χ2 (1) = 0.023, p = 0.880 |
| Body mass index, M ± SD | 27.68±5.12 | 28.23±5.51 | 26.95±4.69 | t(28) = 0.669, p = 0.509 |
| CMRI data | ||||
| Global WMH load, cm3, Med (IQR) | 1.65(0.15–5.60) | 0.19(0.04–1.35) | 6.25(4.23–19.89) | U(n1 = 17, n2 = 12) = 4.000, |
| p < 0.001 b | ||||
| Regional (frontal) WMH load, cm3, Med (IQR) | 0.95(0.09–3.58) | 0.14(0.03–0.83) | 4.27(2.86–12.31) | U(n1 = 17, n2 = 12) = 12.000, |
| p < 0.001 b | ||||
| Normalized brain volume, mm3, M ± SD | 1,487,347±63,546 | 1,503,646±57,598 | 1,464,257±66,791 | t(27) = 1.698, p = 0.101 |
| DUpper limb function | ||||
| Purdue Pegboard Test, M ± SE | F(4,24) = 0.154, p = 0.959 | |||
| Pins right | 13.25±0.31 | 13.37±0.46 | 13.13±0.55 | |
| Pins left | 12.48±0.29 | 12.16±0.43 | 12.35±0.51 | |
| Pins right and left | 10.56±0.41 | 10.89±0.61 | 10.22±0.73 | |
| Pins assembly | 23.49±0.96 | 23.57±1.44 | 23.41±1.71 | |
| ELower limb function | ||||
| SPPB, M ± SD | 7.40±1.55 | 7.29±0.92 | 7.54±2.15 | U(n1 = 17, n2 = 13) = 106.0, |
| p = 0.869 | ||||
| SPPB score ≤ 10, n (%) | 29(96.7) | 17(100) | 12(92.3) | χ2 (1) = 1.353, p = 0.433 |
| Single leg stance time, Med (IQR) | 20(14–20) | 20(14–20) | 20(14–20) | U(n1 = 17, n2 = 13) = 110.0, |
| p = 1.000 | ||||
| Single leg stance time < 15 sec, n (%) | 8(26.7) | 4(23.5) | 4(30.8) | χ2 (1) = 0.197, p = 0.698 |
| Gait velocity (4 m), Med (IQR) | 0.4(0.330–0.444) | 0.4(0.327–0.489) | 0.4(0.327–0.444) | U(n1 = 17, n2 = 13) = 110.5, |
| p = 1.000 | ||||
| Gait velocity < 1.2 m/sec, n (%) | 29(96.7) | 17(100) | 12(92.3) | χ2 (1) = 1.353, p = 0.433 |
2.2. Magnetic resonance imaging
Imaging was performed on a 3.0 T Tim Trio system (Siemens Medical Systems, Erlangen, Germany) using a 12-element head array coil. Functional data were acquired using a single shot gradient echoplanar imaging (EPI) sequence (repetition time [TR] = 3000 ms, echo time [TE] = 30 ms, spin angle 90°, matrix size 64 × 64, pixel size 3 × 3 × 3 mm), with 210 volumes per functional run (scanning time 10.5 minutes).
Morphological imaging data were acquired using conventional turbo-spin-dual echo, FLAIR (TR = 9000 ms, TE = 70 ms, inversion time [TI] = 2500 ms, in plane resolution = 0.9 × 0.9 mm2, slice thickness = 3 mm) and T1-weighted 3-dimensional magnetization prepared rapid gradient echo (MPRAGE) sequences (TR = 1900 ms, TE = 2.10 ms, TI = 900 ms, flip angle = 9°, resolution = 1 × 1 × 1 mm3).
After rating and identification of WMH by a blinded experienced investigator (CE) on FLAIR-images, a trained technician (PL) segmented WMH. Global and frontal lobe lesion volumes were calculated by local thresholding and region growing using home-developed software, which was not feasible in 1 subject due to technical problems. Our semiautomatic lesion segmentation program limits user interaction to placing a seed point in each lesion. By region growing, i.e., stepwise inclusion of all neighbor pixels that fall within a predefined variation of the seed point, the borders of the lesion are found. To make this approach less sensitive for signal variations caused by noise and coil sensitivity, image preprocessing is needed which includes filtering and signal intensity normalization. Based on high resolution T1 scans, brain tissue volume was estimated using SIENAX as part of FSL (www.fmrib.ox.ac.uk/fsl).
2.3. The fMRI experiment
The paradigm involved unilateral active movements of the fingers II–V of the right hand in a wooden apparatus limiting the maximum degree of extension (Wegner et al., 2008). In a separate run, active right ankle movements to a maximum of 30° were performed in a purpose-built wooden apparatus as described previously (Enzinger et al., 2008, Enzinger et al., 2009). Blocks of active movement (30 seconds), visually paced at a fixed rate (1000 ms for dorsiflexion and plantarflexion, 1 Hz for finger movements), alternated with interspersed periods of absolute rest (21 seconds). The sequence of runs was pseudorandomized. Prior to entering the scanner, subjects practiced the paradigm using the same devices. In an attempt to reduce stimulus-correlated motion, subjects' heads were secured with Velcro straps in a foam-cushioned holder and their knees were flexed to approximately 135° using a soft roll placed beneath the knees.
2.4. Test of distal upper limb function - Purdue Pegboard Test
Visuopractical skills were evaluated using the Purdue's Pegboard Test, providing a score for motor function of the distal upper limb.
2.5. Test of lower limb function - gait and balance tests
Equilibrioception, gait, and motor proficiency were assessed with the Short Physical Performance Battery (SPPB) and 2 additional simple tests measuring the same construct (Baezner et al., 2008).
The SPPB incorporates static balance, walking speed, and repeated chair rise. Balance was evaluated by the ability to maintain side-by-side, semitandem, and tandem position standing with closed feet for 10 seconds each. To determine gait speed, subjects walked twice a 4 m distance at their usual pace. The better performance value was used for further statistical analyses. Repeated chair rise competency was obtained by asking the subjects to stand up from a sitting position without help (having their arms folded across the chest), 5 times in a row. The SPPB yields a composite score ranging from 0 to 12 (with higher values representing better function).
In addition, to obtain information on gait velocity, participants walked 8 m at their usual pace, again for 2 times. The fastest time was then used for calculation of walking speed. Further, for the assessment of single leg stance ability, the longest standing time out of 4 attempts (demanding both legs) was used.
2.6. Neuropsychological testing
Cognitive (in particular frontal executive) function was assessed using the Wisconsin Card Sorting Test, Version B of the timed Trail Making Test, the Digit Span from the Wechsler Adult Intelligence Scale-Revised, the Mini Mental State Examination, and a word fluency task.
2.7. fMRI data analysis
fMRI data were analyzed using FMRI Expert Analysis Tool (FEAT, version 5.92, part of FSL, www.fmrib.ox.ac.uk/fsl). The following prestatistics processing was applied: motion correction using MCFLIRT; nonbrain removal using BET; spatial smoothing using a Gaussian kernel of full-width half maximum (FWHM) 5 mm; mean intensity normalization by a single multiplicative factor; high-pass temporal filtering (Gaussian-weighted least squares straight line fitting, with sigma = 54.0 seconds). Time series statistical analysis was carried out using FILM with local autocorrelation correction. Registration to high resolution structural and/or standard space images was carried out using FLIRT. Z (Gaussianised T/F) statistic images were thresholded using clusters determined by Z > 2.3 and a (corrected) cluster significance threshold of p = 0.05. No subject had to be excluded due to artifacts, slice dropouts, or excessive head motion (defined by > 3 mm in any direction).
In first level analyses, contrasts for active movements versus rest were computed for each individual, including head motion parameters as covariates. Registration results were checked visually. Fixed effects analysis was used to generate second-level contrasts at the group level. In third-level analyses, we tested for differences between subject groups defined by WMH grade, adding age as covariate of no interest to the model. Similar analyses were done using sex as a covariate.
Region of interest (ROI) analyses served to compute mean signal changes within ROIs for each individual for active movement conditions versus rest using FEATQUERY. Functional ROIs of bilateral supplementary motor area (SMA) and pre-SMA activation were defined on the basis of higher level contrasts of subjects with WMH grades 2 and 3 versus subjects with WMH grades 0 and 1 for active movement versus rest. Additionally, higher-level linear correlation analyses were performed to investigate the effect of frontal lesion load on the brain's response to active movements versus rest.
All images are shown in radiological convention in which the left side of the image is the right side of the brain.
2.8. General statistical analyses
SPSS (Statistical Package of Social Sciences, Chicago, IL, USA; version 16.0) was used for general statistical analyses. For normally distributed continuous variables, the significance of any differences in means was tested by uni- and multivariate analyses of (co) variance and Student t tests. Mann-Whitney U tests were used to test for significant differences of nonnormally distributed variables. Any differences in proportions were assessed by χ2 statistics.
3. Results
3.1. Sociodemographic, clinical, behavioral, neuropsychological, and conventional MRI data
As expected, subjects with more severe WMH (grades 2 and 3; group B) were older than subjects with absent or minor WMH (grades 0 and 1; group A). We therefore corrected all subsequent fMRI analyses for age. Arterial hypertension was the only risk factor found more frequently in group B (Table 1, A and B). Group B had a significantly higher volume of global and regional (frontal) WMH compared with group A. There was no significant difference in the global brain volume (Table 1, C). Regarding the comprehensive set of motor- and mobility-related and cognitive tests, there were no significant between-group differences (corrected for age, education, and sex; Table 1, D and E).
3.2. FMRI results
3.2.1. Mean group activation associated with finger or ankle movement versus restContrasts of active finger and ankle movements versus rest showed significant activation in expected somatotopy in both groups, including the primary and secondary sensorimotor cortices (SMC and SII), supplementary and cingulate motor areas (SMA and CMA), and the cerebellum as described previously (Enzinger et al., 2008, Wegner et al., 2008; see also Table 2).
Table 2. fMRI results
| Contrast and anatomical region | Hemisphere | Number of voxels | Maximum Z score | MNI coordinates of maximum Z score | ||
|---|---|---|---|---|---|---|
| X | Y | Z | ||||
| Right ankle subjects WMH grades 0 and 1 | ||||||
| Precuneus | L | 4467 | 21.8 | −10 | −44 | 72 |
| Cerebellum, IV/V | R | 899 | 13.5 | 16 | −36 | −30 |
| Insula | R | 611 | 10.1 | 48 | 2 | 2 |
| Insula | L | 454 | 9.91 | −44 | 2 | 2 |
| Rolandic operculum | R | 341 | 8.98 | 44 | −32 | 20 |
| Right ankle subjects WMH grades 2 and 3 | ||||||
| Paracentral lobule | L | 4959 | 18.6 | −6 | −30 | 74 |
| Cerebellar vermis, IV/V | R | 749 | 11.5 | 2 | −50 | −10 |
| Superior temporal gyrus | R | 650 | 10 | 56 | −22 | 18 |
| Superior temporal gyrus, temporal pole | R | 583 | 9.12 | 56 | 14 | −4 |
| Rolandic operculum | L | 285 | 8.89 | −46 | −28 | 18 |
| Right finger subjects WMH grades 0 and 1 | ||||||
| Precentral gyrus | L | 6099 | 23.3 | −38 | −22 | 58 |
| Cerebellum, VI | R | 1978 | 17.1 | 20 | −52 | −26 |
| Superior temporal gyrus | R | 749 | 9.14 | 60 | −20 | 16 |
| Superior temporal gyrus, temporal pole | R | 420 | 8.53 | 58 | 14 | −4 |
| Postcentral gyrus | R | 414 | 8.82 | 46 | −28 | 46 |
| Right finger subjects WMH grades 2 and 3 | ||||||
| Postcentral gyrus | L | 2970 | 18.7 | −36 | −26 | 56 |
| Cerebellum, VI | R | 1170 | 15 | 20 | −52 | −26 |
| Superior temporal gyrus | R | 813 | 8.81 | 64 | −22 | 16 |
| Supplementary motor area | L | 763 | 11.1 | −2 | −4 | 60 |
| Cerebellum, VI | L | 308 | 10.5 | −28 | −64 | −26 |
| Right ankle contrast subjects WMH 0 and 1 > subjects WMH 2 and 3 | ||||||
| Precuneus | L | 723 | 9.11 | −12 | −44 | 72 |
| Right finger contrast subjects WMH 0 and 1 > subjects WMH 2 and 3 | ||||||
| Precentral gyrus | L | 1297 | 10.1 | −32 | −26 | 60 |
| Inferior frontal gyrus, pars orbitalis | L | 392 | 4.59 | −50 | 40 | −10 |
| Right ankle contrast subjects WMH 2 and 3 > subjects WMH 0 and 1 | ||||||
| Middle temporal gyrus | L | 2612 | 6.38 | −56 | −70 | 4 |
| Inferior frontal gyrus, pars triangularis | L | 2362 | 6.21 | −48 | 44 | 6 |
| Middle frontal gyrus, pars orbitalis | R | 1612 | 5.51 | 26 | 56 | −10 |
| Pre-supplementary motor area | L | 1529 | 6.25 | −6 | 12 | 46 |
| Right finger contrast subjects WMH 2 and 3 > subjects WMH 0 and 1 | ||||||
| Correlation analyses right active ankle movements with frontal WMH | ||||||
| Precentral gyrus | R | 1649 | 4.65 | 36 | −6 | 40 |
| Supramarginal gyrus | R | 749 | 4.37 | 56 | −40 | 30 |
| Correlation analyses right finger ankle movements with frontal WMH | ||||||
| Precuneus | L | 582 | 4 | −4 | −40 | 60 |
Contrasting active finger movements versus rest in group B over group A, no significant differences were noted. The reverse contrast (i.e., group A vs. B) showed activation clusters in the contralateral precentral and postcentral gyri.
In contrast, for active ankle movements versus rest, profound activation differences between subjects with more severe WMH compared with subjects with absent or negligible WMH were found. This contrast demonstrated significantly stronger bilateral pre-SMA recruitment in group B, together with more pronounced bihemispheric activation of the middle and inferior frontal gyri. Further areas of activation difference included temporo-occipital regions, the precentral gyrus, and the cerebellum. The reverse contrast (i.e., group A vs. group B) only showed stronger activation in nonmotor areas (generally ascribed to the default network such as the precuneus) and the contralateral postcentral gyrus (Fig. 1b; Table 2). As there was an imbalance in sex distribution in the entire cohort (though not statistically significant between the subgroups), we reran the analyses with sex as additional covariate. This did not change the findings.
Fig. 1.
Activation maps associated with (A) finger movements versus rest and (B) ankle movements versus rest. Selected slices (z-coordinates) of higher level group contrasts show areas of significant activation difference for the contrasts white matter hyperintensities (WMH) 2 and 3 > WMH 0 and 1 and WMH 0 and 1 > WMH 2 and 3, including age as a covariate. For explanations please see text. All images shown are in radiological convention (see Table 2 for cluster coordinates and activation significance for contrasts).
To closer investigate the magnitude of variation of activation in relation to WMH severity in selected brain areas, we performed ROI analyses using finger and ankle movement conditions versus rest from individual subjects. No activation differences were found for active finger movements. However, with ankle movements, subjects with more severe WMH demonstrated higher signal changes in pre-SMA and SMA proper, compared with individuals with negligible WMH (Fig. 2).
Fig. 2.
Region of interest (ROI) analyses to illustrate quantitatively the magnitude of the activation differences for the 2 subject groups based on white matter hyperintensities (WMH) grade, using functional data from individual subjects. Exclusively stronger pre-supplementary motor area (SMA) and SMA proper activation depending on WMH grade becomes evident with ankle movements only.
To specifically test for a possible interaction between frontal lobe WMH volume and motor network activation elicited by ankle movements, we performed correlation analyses. Whereas a significant positive correlation between finger movements and increasing frontal lesion load was found for the precuneus only, several significant positive correlations were noted for ankle movements. These comprised pre-SMA, the contralateral middle and superior frontal gyri, the precentral and postcentral gyri, the supramarginal gyrus, the cingulate cortex, and the inferior parietal gyrus (Fig. 3; Table 2).
Fig. 3.
Influence of frontal white matter hyperintensities (WMH) load on brain activation elicited by the motor paradigms. Upper panel: brain regions (selected slices, z-coordinates) in which finger movement-associated brain activation shows a significant positive correlation with frontal lesion load (mixed effects analysis; Z > 2.3; corrected cluster significance threshold p = 0.05). Lower panel: analogue results for ankle movements.
4. Discussion
The objective of this study was to test for preclinical differences in functional motor network organization related to the severity of age-related WMH using fMRI. We hypothesized that increasing age-related WMH might lead to compensatory functional changes in brain motor network organization. If these changes were compensatory, one would postulate the functional deviations from normal to increase with increasing WMH severity, in the absence of paralleling abnormalities in behavior. In line with this assumption, as a primary finding, we here observed more widespread activation in the presence of more severe WMH with ankle movements. This excess activation also comprised of nonprimary motor areas like fronto-temporal and occipital regions, as well as pre-SMA. No such differences were found for finger movements.
Our study thus provides first evidence that WMH related functional changes in motor network organization of lower limb movements might precede behaviorally detectable declines in mobility related assessments. As there were no differences in comprehensive mobility and cognitive testing, it is likely that the functional differences observed indeed represent compensatory changes related to WMH. This view is strengthened by the positive correlation between functional deviations and increasing frontal lobe WMH volume. This in turn appears to be consistent with the frontal lobe hypothesis of aging (Raz, 2000, West, 1996), aggravated by an interaction with frontal WMH.
fMRI revealed stronger pre-SMA activation in the presence of more severe WMH. In contrast to the caudal SMA (i.e., the SMA proper) that is connected to the precentral gyrus and attributed to motor execution (i.e., a “lower” function), this rostral part of the SMA is connected to the prefrontal cortex and considered to be responsible for movement preparation (i.e., a “higher” function; Johansen-Berg et al., 2004, Picard, 1996). In humans, SMA proper activation occurs during simple motor tasks, whereas pre-SMA usually is preserved to complex motor tasks (Picard, 1996). Moreover, a reciprocal relationship between SMA proper and pre-SMA exists for motor skill acquisition, with SMA proper being involved in automatic movements and pre-SMA being activated during new unlearned motor tasks (Nachev et al., 2009).
These findings are complemented by stronger frontal and occipital activations in subjects with more severe WMH. The frontal cortex plays a role in advanced motor preparation (Connolly et al., 2007) and higher cognitive functions like cognitive control and executive processes (DuBoisgueheneuc et al., 2006, Koechlin et al., 2003). Consistent with this, positron emission tomography (PET) studies (Jueptner et al., 1997) found frontal activation during new compared with automated (learned) key press sequences, in particular when participants paid attention to the task. “Holding something in mind” therefore coupled this frontal component with occipital activation. Occipital activation has also been discussed in the context of enhanced attention (Deyoe, 2002).
Together, this suggests that even the simple movement paradigm imposed greater effort and cognitive demand on people with severe WMH. This is a surprising finding regarding the absence of clinically detectable differences between groups, very much in line with findings in patients with early forms of multiple sclerosis (MS) in clinical remission without clinical signs of disability (Rocca et al., 2005). Given the activation differences almost exclusively found with ankle movements and their relation to gait (Capaday, 2002, Enzinger et al., 2008, Enzinger et al., 2009), as well as the relationship between WMH and gait impairment (Baezner et al., 2008), it is tempting to speculate on a disruption of frontosubcortical loops by more severe WMH. A positive linear correlation between increasing WMH lesion load and activation of pre-SMA and frontal areas strongly supports this assumption.
Within the parallel organization of functionally segregated circuits linking basal ganglia and frontal cortex, the basal ganglia connection appears to be fundamental for intact movement execution (Cummings, 1993). WMH are predominately located in frontal and periventricular areas (Blahak et al., 2009). Frontal lesions constituted also the majority of the whole lesion volume in our cohort. It is thus conceivable that WMH interfere with these frontosubcortical loops and that the excess activation in subjects with more severe WMH observed with ankle movements might represent compensatory activation, in line with the “disconnection hypothesis” (Galluzzi et al., 2008).
Our findings also remind of a long-standing hypothesis regarding the underlying pathophysiology of gait disorders in subcortical arteriosclerotic encephalopathy (SAE; comprising extensive WMH and lacunar infarcts) (Thompson and Marsden, 1987). Thompson and Marsden noted that patients with subcortical arteriosclerotic encephalopathy faced much more difficulty in using their legs to walk than regarding other movements of the lower limbs when lying or seated, and especially compared with relatively preserved upper limb mobility. They thus proposed damage to the afferent and efferent interconnections of the leg areas of the motor and supplementary motor areas of the cerebral cortex with the cerebellum and basal ganglia as possible mechanism for this Parkinsonian-like ataxia. Most recently, this notion has received indirect confirmation by a study of lesion location relevant to gait dysfunction in 385 participants in the population-based Tasmanian Study of Cognition and Gait using high resolution MRI and computerized gait analysis (Srikanth et al., 2010). These investigators observed greatest covariance with poorer gait for lesions in white matter voxels that corresponded mainly to major anterior projection fibers (anterior and superior thalamic radiations, corticopontine, and corticospinal tracts), and adjacent parts of anterior association fibers (corpus callosum, superior longitudinal fasciculus, short association fibers). Together with our functional findings, this provides compelling evidence that WMH may contribute to age-related gait decline by disrupting frontally located afferent and efferent projection and association tracts that are known to be involved in the execution of planned movements. Among other factors, final exhaustion of such adaptive processes might lead to impaired gait and risk of falls associated with WMH.
When interpreting our results, some weaknesses of the study also need to be kept in mind. First, the general effect size of the fMRI results was weak and only noted with fixed effects analysis. This is not entirely surprising as we studied otherwise healthy subjects and given the histopathological substrates of WMH, that do not reflect areas of complete tissue destruction (Fazekas et al., 1993). Also, the findings appear physiologically plausible. Secondly, the incidence of WMH increases with age. As such, the study groups were unbalanced regarding age. We therefore corrected all analyses, including age as a covariate. Nonetheless, it is likely that the activation differences observed result from an interaction of aging processes of the brain and WMH rather than WMH effects only. In addition, our cohort consisted predominantly of women. While this might limit the generalizability of our findings, a major sex-related effect appears unlikely as the sex distribution was not significantly different between WMH subgroups and inclusion of sex as a covariate in the fMRI analyses did not alter the results. Thirdly, more sophisticated fMRI paradigms including different kinematics, more skilled, or bilateral movements might have led to functionally more relevant insights. However, to the best of our knowledge, this is the first study to probe the functional effects of WMH on motor network organization in healthy elderly individuals, which is why we chose simple standardized and widely used motor paradigms allowing straightforward interpretability. Stimulated by our findings, future studies might wish to use more complex fMRI motor paradigms, e.g., for testing motor skill learning in subjects with different grades of WMH. Finally, including tractography measures would allow more specific testing of the “disconnection hypothesis”, by investigating the relationship between the integrity of white matter pathways, WMH, and functional connectivity in frontosubcortical circuits.
In summary, using fMRI we here provide first evidence for covert functional motor network changes associated with more severe age-related WMH in the absence of overt deficits in mobility. We interpret these findings as compensatory mechanisms which serve to limit the functional consequences of small vessel disease. Among other factors, final exhaustion of this functional plasticity in the presence of WMH progression over a certain threshold might explain the rapid transition to disability noted recently in subjects with severe WMH (Inzitari et al., 2009). Besides prevention of WMH progression by adequate treatment of risk factors, specific training strategies might help to maintain compensation and thus contribute to successful aging.
Contributions
Study concept and design: Linortner, Fazekas, Schmidt, Ropele, Neuper, Enzinger; acquisition of data: Linortner, Schmidt, Ropele, Pendl, Petrovic, Loitfelder, Enzinger; analysis and interpretation of the data: Linortner, Fazekas, Schmidt, Ropele, Pendl, Petrovic, Loitfelder, Neuper, Enzinger; drafting of the manuscript: Linortner, Fazekas, Schmidt, Ropele, Enzinger; critical revision of the manuscript for important intellectual content: Linortner, Fazekas, Schmidt, Ropele, Pendl, Petrovic, Loitfelder, Neuper, Enzinger; statistical expertise: Linortner, Ropele, Petrovic, Loitfelder, Enzinger; obtained funding: Schmidt; technical and material support: Linortner, Fazekas, Schmidt, Ropele, Pendl, Petrovic, Loitfelder, Neuper, Enzinger; study supervision: Fazekas, Schmidt, Ropele, Enzinger.
We thank the subjects for participation in this study and Erich Flooh for help with database management.
Disclosure statement
The authors have no actual or potential conflicts of interest to disclose.
The study has been approved by the local ethics committee and all participants gave their informed consent.
Acknowledgements
We had no funding sources for this study.
References
- . Association of gait and balance disorders with age-related white matter changes (The LADIS study). Neurology. 2008;70:935–942
- . Deep frontal and periventricular age related white matter changes but not basal ganglia and infratentorial hyperintensities are associated with falls: cross sectional results from the LADIS study. J. Neurol. Neurosurg. Psychiatry. 2009;80:608–613
- . The special nature of human walking and its neural control. Trends Neurosci. 2002;25:370–376
- . Effector-specific fields for motor preparation in the human frontal cortex. Neuroimage. 2007;34:1209–1219
- . Frontal-subcortical circuits and human behavior. Arch. Neurol. 1993;50:873–880
- . Occipital lobe. In: Ramachandran VS editors. Encyclopedia of the Human Brain. San Diego: Academic Press; 2002;
- . Functions of the left superior frontal gyrus in humans: a lesion study. Brain. 2006;129:3315–3328
- . Brain activity changes associated with treadmill training after stroke. Stroke. 2009;40:2460–2467
- . Functional MRI correlates of lower limb function in stroke victims with gait impairment. Stroke. 2008;39:1–7
- . MR signal abnormalities at 15T in Alzheimer's dementia and normal aging. AJR Am. J. Roentgenol. 1987;149:351–356
- . Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology. 1993;43:1683–1689
- . The effect of white matter lesions on cognition in the elderly − small but detectable. Nat. Clin. Pract. Neurol. 2007;3:620–627
- . White matter lesions in the elderly: Pathophysiological hypothesis on the effect on brain plasticity and repair. J. Neurol. Sci. 2008;278:3–9
- . Changes in white matter as determinant of global functional decline in older independent outpatients: three-year follow-up of LADIS (leukoaraiosis and disability) study cohort. BMJ. 2009;339:b2477
- . Changes in connectivity profiles define functionally distinct regions in human medial frontal cortex. Proc. Natl. Acad. Sci. U. S. A. 2004;101:13335–13340
- . Anatomy of motor learning (I. Frontal cortex and attention to action). J. Neurophysiol. 1997;77:1313–1324
- . The architecture of cognitive control in the human prefrontal cortex. Science. 2003;302:1181–1185
- . Comparing brain activation associated with isolated upper and lower limb movement across corresponding joints. Hum. Brain Mapp. 2002;17:131–140
- . Functional role of the supplementary and pre-supplementary motor areas. Nat. Rev. Neurosci. 2009;9:856–869
- . Impact of age-related cerebral white matter changes on the transition to disability – the LADIS study: rationale, design and methodology. Neuroepidemiology. 2005;24:51–62
- . Motor areas of the medial wall: A review of their location and functional activation. Cereb. Cortex. 1996;6:342–353
- . Imaging the premotor areas. Curr. Opin. Neurobiol. 2001;11:663–672
- . Aging of the brain and its impact on cognitive performance: Integration of structural and functional findings. In: Craik FIM, Salthouse TA editor. The Handbook of Aging and Cognition. second ed.. Mahwah: Lawrence Erlbaum Associates; 2000;
- . Cortical adaptation in patients with MS: a cross-sectional functional MRI study of disease phenotypes. Lancet Neurol. 2005;4:618–626
- . Functional magnetic resonance imaging for the evaluation of the motor system: Primary and secondary brain areas in different motor tasks. Stereotact. Funct. Neurosurg. 2002;78:3–16
- . Progression of cerebral white matter lesions: 6-year results of the Austrian Stroke Prevention Study. Lancet. 2003;361:2046–2048
- . White matter lesion progression: a surrogate endpoint for trials in cerebral small-vessel disease. Neurology. 2004;63:139–144
- . Bilateral frontal periventricular locations of cerebral white matter lesions show greatest covariance with poorer gait. Cerebrovasc. Dis. 2010;29(suppl 2):(abstract)
- . Gait disorder of subcortical arteriosclerotic encephalopathy: Binswanger's disease. Mov. Disord. 1987;2:1–8
- . Relating functional changes during hand movement to clinical parameters in patients with multiple sclerosis in a multi-centre fMRI study. Eur. J. Neurol. 2008;15:113–122
- . An application of prefrontal cortex function theory to cognitive aging. Psychol. Bull. 1996;120:272–292
PII: S0197-4580(10)00257-5
doi:10.1016/j.neurobiolaging.2010.06.005
© 2012 Elsevier Inc. All rights reserved.
