Table 2.
R2 Change and p Values for the Change in FA Correlating With AD PRS for Each Threshold
| Training PT Value | Fornix R2 (p Value) | Cingulum L R2 (p Value) | Cingulum R R2 (p Value) | Cingulum R No APOER2 (p Value) | PHC L R2 (p Value) | PHC R R2 (p Value) |
|---|---|---|---|---|---|---|
| PT < 1 × 10−8 | <.001 (.791) | .009 (.164) | .030 (.011)a | .006 (.254) | .008 (.194) | .002 (.534) |
| PT < 1 × 10−7 | <.001 (.813) | .009 (.162) | .029 (.013)a | .006 (.268) | .006 (.233) | .002 (.515) |
| PT < 1 × 10−6 | <.001 (.813) | .005 (.270) | .023 (.026)a | .019 (.044)a | .006 (.255) | .002 (.538) |
| PT < 1 × 10−5 | <.001 (.846) | .004 (.344) | .025 (.020)a | .006 (.270) | .005 (.292) | .001 (.655) |
| PT < 1 × 10−4 | <.001 (.780) | .009 (.165) | .032 (.009)a,b | <.001 (.885) | .002 (.547) | <.001 (.949) |
| PT < .01 | <.001 (.952) | .001 (.610) | .006 (.273) | <.001 (.858) | .003 (.432) | .001 (.727) |
| PT < .1 | <.001 (.886) | .001 (.626) | .005 (.305) | .001 (.588) | .005 (.289) | .006 (.268) |
| PT < .3 | <.001 (.840) | .003 (.440) | .002 (.493) | <.001 (.777) | .003 (.388) | .004 (.357) |
| PT < .5 | <.001 (.806) | .002 (.486) | .002 (.471) | .001 (.731) | .003 (.387) | .007 (.240) |
Fornix (n = 157), cingulum left (n = 197) and right (n = 197) (with and without APOE SNPs), and parahippocampal cingulum left (n = 197) and right (n = 197) are shown here.
AD, Alzheimer’s disease; FA, fractional anisotropy; FDR, false discovery rate; L, left; PRS, polygenic risk score; PHC, parahippocampal cingulum; PT, polygenic threshold; R, right; SNP, single nucleotide polymorphism.
a
Nominally significant associations (p value < .05).
b
Associations that survive FDR correction.