Skip to main content
. 2021 Dec 2;33:102904. doi: 10.1016/j.nicl.2021.102904

Table 3.

Explanation of concurrent disability using SPACE-MS metrics.

Clinical variable at baseline# (dependent variable)
EDSS score Inverse of TWT Inverse of 9HPT PASAT score SDMT score
Best a priori models
Model R2 0.1156 0.2071 0.2193 0.2102 0.3414
Main predictors (below)
Lesion load at baseline in mLRC (95%CI), p-value 0.009 (0.003 to 0.014), p = 0.002 −0.0004 (−0.0007 to −0.0001), p = 0.009 −0.0002 (−0.0003 to −0.00015), p < 0.001 −0.371 (-0.471 to −0.271), p < 0.001 −0.405 (-0.492 to −0.317), p < 0.001
WM volume at baseline in mLRC (95%CI), p-value −0.001 (−0.004 to 0.002), p = 0.439 0.0001 (−0.00001 to 0.0003), p = 0.077 0.00004 (0.00001 to 0.00007), p = 0.006 0.003 (−0.041 to 0.048), p = 0.885 0.087 (0.049 to 0.125), p < 0.001
GM volume at baseline in mLRC (95%CI), p-value −0.0004 (−0.003 to 0.002), p = 0.752 −0.00003 (−0.0001 to 0.0001), p = 0.670 6.41•10-06 (−0.00002 to 0.00003), p = 0.635 0.041 (0.001 to 0.081), p = 0.042 −0.008 (−0.045 to 0.030), p = 0.683
Age at baseline in yearsRC (95%CI), p-value 0.006 (−0.005 to 0.017), p = 0.261 −0.0004 (−0.001 to 0.0002), p = 0.178 0.0002 (0.0001 to 0.0003), p = 0.004 0.272 (0.074 to 0.469), p = 0.007 0.048 (-0.123 to 0.219), p = 0.581
Male sexRC (95%CI), p-value −0.222 (-0.370 to −0.073), p = 0.003 0.012 (0.004 to 0.021), p = 0.004 −0.003 (−0.005 to −0.001), p < 0.001 3.974 (1.292 to 6.655), p = 0.004 −0.938 (-3.275 to 1.399), p = 0.430
Disease duration at baseline in yearsRC (95%CI), p-value 0.009 (0.001 to 0.017), p = 0.028 −0.0006 (−0.0011 to −0.0002), p = 0.007 −0.0001 (−0.00020 to −8.33•10-06), p = 0.033 −0.011 (-0.156 to 0.134), p = 0.879 −0.040 (-0.160 to 0.081), p = 0.514



SPACE-MS models of concurrent disability&
Baseline SPACE-MS metrics (below)$
NCIRC (95%CI), p-value
Model R2 (%I)
−0.135 (−1.151 to 0.880), p = 0.793
0.1158 (0.2%)
0.031 (−0.025 to 0.086), p = 0.276
0.2092 (1.0%)
−0.016 (−0.028 to −0.004), p = 0.010
0.2308 (5.2%)
20.106 (2.061 to 38.152), p = 0.029
0.2185 (3.9%)
−7.103 (-21.637 to 7.430), p = 0.337
0.3431 (0.5%)
Maximum lesion NCIRC (95%CI), p-value
Model R2 (%I)
0.593 (0.046 to 1.140), p = 0.034
0.1238 (7.1%)
−0.030 (−0.061 to 0.001), p = 0.058
0.2135 (3.1%)
−0.009 (−0.015 to −0.002), p = 0.011
0.2305 (5.1%)
−0.423 (−10.478 to 9.632), p = 0.934
0.2102 (0%)
−10.978 (−19.161 to −2.796), p = 0.009
0.3540 (3.7%)
MCIRC (95%CI), p-value
Model R2 (%I)
0.0004 (−0.0003 to 0.001), p = 0.245
0.1181 (2.2%)
−7.16•10-06(−0.00005 to 0.00003), p = 0.712
0.2073 (0.1%)
−0.00001(−0.00002 to−3.95•10-06), p = 0.004
0.2336 (6.5%)
−0.001 (−0.014 to 0.011), p = 0.857
0.2102 (0%)
0.0003 (−0.010 to 0.011), p = 0.949
0.3414 (0%)
CAIRC (95%CI), p-value
Model R2 (%I)
−0.780 (−1.484 to −0.076), p = 0.030
0.1242 (7.4%)
0.049 (0.011 to 0.088), p = 0.013
0.2181 (5.3%)
0.012 (0.003 to 0.020), p = 0.006
0.2323 (5.9%)
10.248 (−2.599 to 23.095), p = 0.118
0.2145 (2.0%)
14.667 (4.500 to 24.834), p = 0.005
0.3559 (4.2%)
CPIRC (95%CI), p-value
Model R2 (%I)
−0.519 (−1.088 to 0.049), p = 0.073
0.1215 (5.1%)
0.024 (−0.007 to 0.055), p = 0.131
0.2111 (1.9%)
0.008 (0.001 to 0.015), p = 0.023
0.2281 (4.0%)
1.196 (−9.090 to 11.483), p = 0.819
0.2103 (0.1%)
6.921 (−1.65248 15.49546), p = 0.113
0.3460 (1.3%)
CSIRC (95%CI), p-value
Model R2 (%I)
0.960 (0.343 to 1.578), p = 0.002
0.1323 (14.4%)
−0.053 (−0.087 to −0.019), p = 0.002
0.2237 (8.0%)
−0.013 (−0.021 to −0.006), p < 0.001
0.2408 (9.8%)
−8.953 (−20.189 to 2.283), p = 0.118
0.2145 (2.0%)
−15.562 (−24.667 to −6.458), p = 0.001
0.3616 (5.9%)



Best SPACE-MS model of concurrent disability
Best SPACE-MS metric(s) at baselineRC (95%CI), p-value
Model R2 (%I)
CSI: 0.960(0.343 to 1.578), p = 0.002
0.1323 (14.4%)
CSI: −0.053(−0.087 to −0.019), p = 0.002
0.2237 (8.0%)
CSI: −0.013(−0.020 to −0.005), p = 0.001
NCI: −0.014(−0.026 to −0.002), p = 0.019
0.2500 (14.0%)
NCI: 20.106 (2.061 to 38.152), p = 0.029
0.2185 (3.9%)
CSI: −15.562(−24.667 to −6.458), p = 0.001
0.3616 (5.9%)

The ‘best a priori model’ included, as covariates, baseline lesion volume, baseline WM volume, baseline GM volume, age in years at baseline, male sex, disease duration at baseline, and study centre. Of note, the regression coefficients for ‘study centre’ are not shown (for simplicity). The ‘SPACE-MS models’ included all the variables of the ‘best a priori model’ plus one SPACE-MS metric at a time. See methods for full details. #: the EDSS score is measured in EDSS score units; the inverse of TWT and the inverse of 9HPT, in 1/s; and the PASAT and SDMT scores, in number of correct answers; &: all ‘SPACE-MS models’ included, as covariates, the variables included in the ‘best a priori models’. However, for simplicity, only the RC (95%CI) of the SPACE-MS metric is shown; $: all SPACE-MS metrics are measured in dimensionless units except for MCI, which is measured in mm2. Significant results are highlighted. Abbreviations (in alphabetical order): %I: percentage of model improvement based on R2; 9HPT: nine-hole peg test; CAI: covariance anisotropy index; CI: Confidence Interval; CPI: covariance planarity index; CSI: covariance sphericity index; EDSS: expanded disability status scale; GM: grey matter; Max: maximum; MCI: mean covariance index; NAWM: normal-appearing white matter; NCI: neuraxis caudality index; PASAT: paced auditory serial addition test (measured in number of correct answers); RC: regression coefficient; SD: standard deviation; SDMT: symbol digit modalities test (measured in number of correct answers); SPACE-MS: spatial patterns (of MS lesions) assessed through covariance estimations of lesional voxels; TWT: 25-foot timed walk test.