Skip to main content
. 2021 Nov 10;11:580. doi: 10.1038/s41398-021-01703-1

Table 4.

F1, F2, and Fiso group comparisons*.

Covariates: age, sex, motion Covariates: age, sex, motion, CSF volume Covariates: age, sex, motion, WM volume
F1
F-test 36% of the WM skeleton 32% of the WM skeleton 14% of the WM skeleton
 22q-del versus HC 22q-del > HC: 8% 22q-del > HC: – 22q-del > HC: –
22q-del < HC: – 22q-del < HC: – 22q-del < HC: –
 22q-dup versus HC 22q-dup > HC: – 22q-dup > HC: – 22q-dup > HC: –
22q-dup < HC: 26% 22q-dup < HC: 29% 22q-dup < HC: 21%
 22q-del versus 22q-dup 22q-del > 22q-dup: 43% 22q-del > 22q-dup: 31% 22q-del > 22q-dup: 21%
22q-del < 22q-dup: – 22q-del < 22q-dup: – 22q-del < 22q-dup: –
F2
F-test 40% of the WM skeleton 26% of the WM skeleton 23% of the WM skeleton
 22q-del versus HC 22q-del > HC: 28% 22q-del > HC: 12% 22q-del > HC: 2%
22q-del < HC: – 22q-del < HC: – 22q-del < HC: –
 22q-dup versus HC 22q-dup > HC: – 22q-dup > HC: – 22q-dup > HC: –
22q-dup < HC: 21% 22q-dup < HC: 21% 22q-dup < HC: 19%
 22q-del versus 22q-dup 22q-del > 22q-dup: 46% 22q-del > 22q-dup: 34% 22q-del > 22q-dup: 28%
22q-del < 22q-dup: – 22q-del < 22q-dup: – 22q-del < 22q-dup: –
Fiso
F-test 66% of the WM skeleton 55% of the WM skeleton 50% of the WM skeleton
 22q-del versus HC 22q-del > HC: – 22q-del > HC: – 22q-del > HC: –
22q-del < HC: 49% 22q-del < HC: 18% 22q-del < HC: 12%
 22q-dup versus HC 22q-dup > HC: 48% 22q-dup > HC: 52% 22q-dup > HC: 47%
22q-dup < HC: – 22q-dup < HC: – 22q-dup < HC: –
 22q-del versus 22q-dup 22q-del > 22q-dup: – 22q-del > 22q-dup: – 22q-del > 22q-dup: –
22q-del < 22q-dup: 72% 22q-del < 22q-dup: 64% 22q-del < 22q-dup: 47%

CSF cerebrospinal fluid, F1 fractional volume of primary fiber population, F2 fractional volume of secondary fiber population, Fiso fractional volume of the isotropic compartment, HC healthy controls, WM white matter, 22q-del 22q11.2 deletion carriers, 22q-dup 22q11.2 duplication carriers.

*All group comparisons were conducted utilizing non-parametric voxel-wise permutation tests for each voxel on the WM skeleton in FSL’s Randomize [81]. We tested data against a null distribution generated with 5000 permutations for each contrast using threshold-free cluster enhancement [82] and family wise error correction at a significance level of p < 0.05.