Fall risk is associated with amplified functional connectivity of the Central Executive Network in patients with Parkinson's disease
Journal of Neurology
Keren Rosenberg-Katz, PhD, Talia Herman, MscPT, Yael Jacob, MSc, Anat Mirelman, PhD, Nir Giladi, MD, Talma Hendler MD, PhD, Jeffrey M Hausdorff, PhD*
*Correspondence:
Prof. Jeffrey M Hausdorff
Center for the study of Movement, Cognition, and Mobility, Department of Neurology
Tel Aviv Sourasky Medical Center
6 Weizmann street Tel Aviv, 64239, Israel.
Telephone number: +97236973081
Fax number: +97236977514
Supplementary Content:
1. Data quality assurance
2. Assessment of gait
3. MRI data acquisition and analysis
4. rs-fMRI data preprocessing
5. Functional connectivity analysis
6.VBM analysis
7. Mediation analysis
8. Assessments of GM volume in regions showing increased functional connectivity
Supplementary Fig. 1.
Supplementary Fig. 2.
Supplementary Fig. 3.
Table A.1
Table A.2
References
Appendix
1.Data quality assurance
Particular care was taken to minimize the impact of the subject’s head motion at the individual and at the group levels. For each subject, we evaluated mean head motion, as determined by the realignment procedure, as well as the whole-brain mean BOLD signal standard deviation (SD) averaged over the whole brain [10]. Quality assurance-based exclusion criteria were empirically determined with the objective of maximizing the number of included subjects while simultaneously excluding subjects with motion artifacts [10]. In order to account for head movements within the scanner, the six directions rigid body head motion correction parameters which include translational movement in millimeters (x, y, z) and rotational motion in degrees (pitch, roll, yaw), were analyzed. Subjects with mean head movement >2mm, rotation>1 degree or with functional connectivity signal SD >3.5 standard deviations from their group mean were excluded. The finally obtained head movement measures were comparable across the non-fallers and fallers with mean cumulative translations: non-fallers: mean = 0.32 mm, SD = 0.12; fallers: 0.28 mm, SD = 0.09 (p = 0.17) and mean rotation <0.006 degrees in both groups (p=0.18).
2. Assessment of gait
Participants were asked to walk back and forth in a 30 meter corridor while wearing a small, lightweight body-fixed sensor (McRoberts, DynaPort Hybrid system, the Netherlands) attached with a belt to their lower back.Gait speed was determined by measuring the average time the subject walked the middle 10 meters of the corridor under single and dual tasking conditions. Stride time variability was computed using coefficient of variation of stride time [4].
.
Due to the large variance of stride time variability, we used the normalized z-scores of this measurement.
3. MRI data acquisition and analysis
To evaluate alterations in functional connectivity of the CEN, participantswere instructed to relaxinthe scanner and stay awake with eyes open without fixation. rs-fMRI data consisted of 266 volumes of a repeated gradient-echo planar imaging T2*-weighted sequence with the following parameters: 30 axial slices, TR = 1,680 milliseconds, TE = 35 milliseconds, flip angle = 90, slice thickness = 3.5 mm, interslice gap=0 mm, matrix = 64x64, FOV = 200x200 mm2.
Three-dimensional high-resolution T1-weighted axial images (GE sequence BRAVO, TR = 9000 milliseconds, TE = 3.6 milliseconds, flip angle = 90, voxel size = 1x1x1 mm3, matrix = 256x256, FOV = 250X250 mm2) were acquired for registration and normalization of the functional images and to evaluate striatal gray matter loss.
4. rs-fMRI data preprocessing
rs-fMRI data were pre-processed using Statistical Parametric Mapping (SPM8) software[12] running on MATLAB 7.9 (Math-Works, Natick, MA, USA). Data were realigned for motion correction, spatially normalized into the standard Montreal Neurological Institute (MNI) space, slice-time corrected and smoothed with an 8-mm full-width at half-maximum Gaussian kernel to decrease spatial noise before statistical analysis.
5. Functional connectivity analysis
To define the regions of interest for functional connectivity analysis, we used the AAL atlas tool in WFU pick atlas [6]. A Gaussian mask with an 8-mm radius was created around the peaks of previously validated ROIs.
For the CEN, the peaks of the 8 mm radius sphereswere selected based on previously established baseline set of functionally relevant intrinsic connectivity (ICs) networks [2]. These networks (ICs 34 and 60) included superior temporal gyrus (STG), inferior parietal lobe (IPL) and posterior parietal lobe (PPL). In addition, we included coordinates corresponding to thedorsolateral prefrontal cortex (DLPFC) based on a meta-analysis using the N-back paradigm [8].
For the sensorimotorcortico-striatal network[1], the thalamus, globuspallidus internal and putamenwere defined based on the WFU pick atlas anatomical mask as we aimed to include the entire region while keeping minimal overlap with other regions. Theposterior putamen was further defined by selecting the part of the putamenwhich is posterior to the line passing through the anterior commissure[5, 9]. The precentral gyrus [2] and the supplementary motor area (SMA) [7]8 mm sphereswere selected based on meta-analyses. Pick coordinates of the selected ROIs are presented in supplementary table 1.
For each network, the signal was extracted from all ROIs, and a set of pairwise Pearson correlation values were calculated for each subject. After a Fisher Z transformation, a comparison between faller and non-fallers was made using two-tailedtstatistics. For these comparisons, significant differences between the groups were defined at theP< 0.005 for the CEN(18 possible pairs- 6 pairs within each hemisphere (total 12) + 6 inter-hemispheric connections between frontal, parietal or temporal regions) and P< 0.007 for the sensorimotor network (10possible pairs), using Bonferroni corrections for multiple comparisons.
Supplementary Table 1. Regions of interestCentral executive Network / x,y,z peak coordinate (MNI) / Sensorimotor network / x,y,z peak coordinate (MNI)
R DLPFC / 36,33,24 / Thalamus / WFUPickAtlas
L DLPFC / -36,33,24 / globuspallidus / WFUPickAtlas
R STG / 56,0,2 / Posterior Putamen / WFUPickAtlas
L STG / -61,-2,0 / R precentral / 36,33,24
R IPL / 42,-56,42 / L precentral / -36,33,24
L IPL / -47,-57,39 / SMA / 0,-10,60
R PPL / 6,-52,37
L PPL / -6,-52,37
DLPFC, dorso-lateral prefrontal cortex; STG, superior temporal gyrus; IPL, inferior parietal lobe; PPL, posterior parietal lobe; SMA, supplementary motor area;
As we were more interested on the changes within the sensorimotor network which are related to the disease related sensory-motor changes, we looked for each patient at the more affected hemisphere, which is contralateral to the body side presented a higher score in the motor UPDRS. For healthy controls and patients which had bilateral motor symptoms we looked at the average between the right and left hemispheres.On the contrary, as laterality within the CEN is less constrained, we examined both hemispheres within this network.
6. VBM analysis
Voxel-based-morphometry (VBM) analysis was conducted using SPM8 software package ( running on MATLAB 7.9 (MathWorks, Natick, MA, USA).ROIs for the VBM analysis included the posterior putamen and caudate head. The posterior putamen was defined by selecting the part of the putamen which is posterior to the line passing through the anterior commissure [5, 9].
For the caudate head and posterior putamen we looked for each patient at the more affected hemisphere which is contralateral to the body side which presented a higher score in the motor UPDRS. For healthy controls and patients which had bilateral motor symptoms we looked at the average between the right and left hemispheres.
First, MR images were segmented into gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF) using the standard unified segmentation model in SPM8. To remove non-brain tissue, the ‘clean-up’ procedure was applied to the segmented GM images. Then, GM and WM segmented images were normalized, respectively, to the GM and WM population templates generated from the complete image set using the Diffeomorphic Anatomical Registration using Exponentiated Lie algebra (DARTEL) registration method [3]. This non-linear warping technique minimizes between-subject structural variations. Subsequently, images were modulated to ensure that relative volumes of GM and WM were preserved following spatial normalization. Finally, spatially normalized images were modulated to ensure that the overall amount of each tissue class was not altered by the spatial normalization procedure, and smoothed with an 8-mm full-width at half maximum Gaussian kernel. The values presented represent the mean voxels probability belonging to gray matter class.
7. Mediation analysis
The mediation analysis tests the hypothesis that a proportion of the variance in a dependent variable (i.e. faller or non-faller) that is predicted by variance in an independent or predictor variable (i.e. caudate head GM volume) can be accounted for by the mediator variables (i.e. level of CEN functional connectivity between left PPl and right IPL), in the sense that GM volume accounts for variance of CEN connectivity and in turn, this variance in of CEN connectivity accounts for a proportion of the variance of fallers or non-fallers. In other words, the mediation analysis partitions the variance explained by gray matter volume into a part that is independent of CEN connectivity, and a part that is accounted for via CEN connectivity. This analysis differs from multiple regression which estimates how several independent predictor variables account for the variance in the dependent variable[11]. To test the statistical significance of the indirect effect, we used bootstrapping with1,000replicates for the mediation effect.
8. Assessments of GM volume in regions showing increased functional connectivity
For each of the nodes within the CEN that were significantly different between the groups, we tested the level of GM volume. Using ANCOVA model corrected for age, we found no significant difference in GM loss in the left PPL (F = 0.29, P > 0.75) or the right ILP (F = 0.22, P > 0.80) (see supplementary Figure 3).
Supplementary Fig. 1: Fisher Z-transformed correlation coefficients, which represent the level of functional connectivity between the nodes of the CEN in fallers and in non-fallers, are plotted as a function of thecaudate head GM volume estimate.
Supplementary Fig. 2: No differences in VBM gray matter volume estimate were found in the nodes within the CEN showing changes in functional connectivity between fallers and non-fallers.
Supplementary Table 2. Summary of main findingscontrast / adjustment / p value
Caudate head gray matter volume
Fallers vs. non fallers / Age and disease duration / F=2.7, P < 0.05
Fallers vs. non fallers / Age, disease duration, Hoehn and Yahr and levodopa equivalent dosage / F=6.74, p<0.01
ANCOVA including fallers, non-fallers and healthy controls / Age / F=4.0, P < 0.01
Post hoc- Fallers vs. non-fallers / Age / P Fisher'sLSD < 0.03
Post hoc- Fallers vs. healthy controls / Age / P Fisher'sLSD < 0.01
Connectivity between left posterior parietal lobe (PPL) and right inferior parietal lobe nodes of the CEN
Fallers vs. non-fallers / - / p<0.005, Bonferroni corrected
Fallers vs. non-fallers / Age and disease duration / p<0.03, uncorrected
Fallers vs. non-fallers / Age, disease duration, Hoehn and Yahr and levodopa equivalent dosage / F= 4.85, p<0.03
ANCOVA including fallers, non-fallers and healthy controls / Age / p<0.04, uncorrected
Post hoc- Fallers vs. non-fallers / Age / P Fisher'sLSD < 0.008
Post hoc- fallers vs. healthy controls / Age / P Fisher'sLSD = 0.06
Reference List
1. Alexander GE, DeLong MR, Strick PL (1986) Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu Rev Neurosci 9:357-381
2. Allen EA, Erhardt EB, Damaraju E, Gruner W, Segall JM, Silva RF, Havlicek M, Rachakonda S, Fries J, Kalyanam R, Michael AM, Caprihan A, Turner JA, Eichele T, Adelsheim S, Bryan AD, Bustillo J, Clark VP, Feldstein Ewing SW, Filbey F, Ford CC, Hutchison K, Jung RE, Kiehl KA, Kodituwakku P, Komesu YM, Mayer AR, Pearlson GD, Phillips JP, Sadek JR, Stevens M, Teuscher U, Thoma RJ, Calhoun VD (2011) A baseline for the multivariate comparison of resting-state networks. Front Syst Neurosci 5:2-5
3. Ashburner J (2007) A fast diffeomorphic image registration algorithm. Neuroimage 38:95-113
4. Hausdorff JM (2005) Gait variability: methods, modeling and meaning. J NeuroEng Rehabil 2:doi:10.1186/1743-0003-2-19
5. Lehericy S, Bardinet E, Tremblay L, Van de Moortele PF, Pochon JB, Dormont D, Kim DS, Yelnik J, Ugurbil K (2006) Motor control in basal ganglia circuits using fMRI and brain atlas approaches. Cereb Cortex 16:149-161
6. Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH (2003) An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 19:1233-1239
7. Mayka MA, Corcos DM, Leurgans SE, Vaillancourt DE (2006) Three-dimensional locations and boundaries of motor and premotor cortices as defined by functional brain imaging: a meta-analysis. Neuroimage 31:1453-1474
8. Owen AM, McMillan KM, Laird AR, Bullmore E (2005) N-back working memory paradigm: a meta-analysis of normative functional neuroimaging studies. Hum Brain Mapp 25:46-59
9. Postuma RB, Dagher A (2006) Basal ganglia functional connectivity based on a meta-analysis of 126 positron emission tomography and functional magnetic resonance imaging publications. Cereb Cortex 16:1508-1521
10. Power JD, Mitra A, Laumann TO, Snyder AZ, Schlaggar BL, Petersen SE (2014) Methods to detect, characterize, and remove motion artifact in resting state fMRI. Neuroimage 84:320-341
11. Preacher KJ, Hayes AF (2004) SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav Res Methods Instrum Comput 36:717-731
12. The MathWorks I (2009) Natick, MA, USA.
1