Table 4.
Mediation Analysis Showing the Proportion of the Mediated Effect of FA on the Total Effect of 15q11.2 BP1-BP2 CNV on Reaction Time, Fluid Intelligence, Symbol Substitution, and Trail Making B
| WM Tracts |
Path B |
Mediation Effects |
||||
|---|---|---|---|---|---|---|
| β (SE) |
pFDR |
Proportion Mediated (CI) |
p Value |
pFDR |
Comparison |
|
| Reaction Time | ||||||
| ALIC_L | .03 (.006) | 3.67 × 10−7a | −0.029 (−0.09 to −0.01) | .004 | .03a | Del vs. NoCNV |
| PLIC_L | .003 (.006) | .6 | −0.003 (−0.02 to 0.01) | .51 | .6 | |
| PTR_R | .03 (.006) | 6.35 × 10−5a | 0.024 (0.008 to 0.07) | .003 | .03a | |
| Cing_HIP_L | .02 (.006) | .003a | −0.022 (−0.07 to −0.01) | .002 | .03a | |
| Cing_CG_R | .04 (.006) | 1.32 × 10−11a | −0.07 (−0.57 to 0.26) | .07 | .1 | Dup vs. NoCNV |
| Cing_HIP_R | .03 (.006) | 2.6 × 10−5a | −0.052 (−0.41 to 0.2) | .07 | .1 | |
| Cing_HIP_L | .02 (.006) | .003a | −0.03 (−0.3 to 0.16) | .09 | .2 | |
| Fluid Intelligence | ||||||
|---|---|---|---|---|---|---|
| ALIC_L | .03 (.006) | 4.34 × 10−7a | −0.02 (−0.07 to 0) | .06 | .1 | Del vs. NoCNV |
| PLIC_L | .02 (.006) | .006a | −0.012 (−0.05 to 0) | .03 | .09 | |
| PTR_R | .05 (.006) | 2.97 × 10−12a | 0.044 (0.015 to 0.14) | .004 | .03a | |
| Cing_HIP_L | .02 (.006) | .004a | −0.023 (−0.07 to −0.01) | .007 | .03a | |
| Cing_CG_R | .01 (.006) | .06 | 0.01 (−0.59 to 0.6) | .9 | .9 | Dup vs. NoCNV |
| Cing_HIP_R | .03 (.006) | 2.75 × 10−5a | 0.037 (−1.33 to 1.66) | .9 | .9 | |
| Cing_HIP_L | .02 (.006) | .004a | −0.03 (−0.88 to 1.05) | .8 | .9 | |
| Symbol Substitution | ||||||
|---|---|---|---|---|---|---|
| ALIC_L | .04 (.007) | 3.65 × 10−9a | −0.05 (−0.19 to −0.01) | .009 | .03a | Del vs. NoCNV |
| PLIC_L | .009 (.007) | .2 | −0.006 (−0.04 to 0) | .2 | .3 | |
| PTR_R | .06 (.007) | 8.94 × 10−13a | 0.06 (0.02 to 0.2) | .006 | .03a | |
| Cing_HIP_L | .02 (.008) | .001a | −0.02 (−0.09 to 0) | .01 | .04a | |
| Cing_CG_R | .05 (.008) | 2.7 × 10−10a | −0.13 (−2.31 to 1.6) | .3 | .4 | Dup vs. NoCNV |
| Cing_HIP_R | .02 (.008) | .04a | −0.03 (−0.55 to 0.35) | .3 | .4 | |
| Cing_HIP_L | .02 (.008) | .001a | −0.05 (−0.75 to 0.79) | .3 | .4 | |
| Trail Making B | ||||||
|---|---|---|---|---|---|---|
| ALIC_L | .04 (.008) | 4.11 × 10−7a | −0.06 (−0.25 to −0.02) | .01 | .04a | Del vs. NoCNV |
| PLIC_L | .01 (.008) | .1 | −0.01 (−0.05 to 0) | .09 | .2 | |
| PTR_R | .05 (.008) | 3.65 × 10−9a | 0.04 (−0.007 to 0.15) | .02 | .05a | |
| Cing_HIP_L | .01 (.008) | .2 | −0.01 (−0.05 to 0.01) | .2 | .3 | |
| Cing_CG_R | .03 (.008) | .0001a | −0.06 (−0.63 to 0.61) | .2 | .3 | Dup vs. NoCNV |
| C_HIP_R | .003 (.008) | .7 | −0.003 (−0.07 to 0.04) | .7 | .8 | |
| Cing_HIP_L | .01 (.008) | .2 | −0.01 (−0.28 to 0.18) | .4 | .5 | |
Negative proportions indicate opposite signs between the mediator (FA) and the total effect (CNV effect on cognitive measure). The linear regression (Path B) shows the overall effects of white matter on cognition (including all deletion and duplication carriers). Both uncorrected and corrected (FDR) p values are shown.
CI, confidence interval; CNV, copy number variant; Del, deletion; Dup, duplication; FA, fractional anisotropy; FDR, false discovery rate; NoCNV, no pathogenic copy number variant; WM, white matter. See Table 2 for white matter tract abbreviations.
p < .05 after FDR correction.