Skip to main content
. 2020 Apr 21;11:250. doi: 10.3389/fneur.2020.00250

Figure 3.

Figure 3

Path models of WMH on memory tested by the mediation analysis. Significant pathways were highlighted in bold characters. For each connection, the standard coefficient (a, b, c, c′) and P-value were shown. The indirect mediation effect (a × b) and its 95% confidence interval was also shown. (A) Association between TWMH and memory deficit was significantly mediated by NLp of IFGoperc.L (Indirect effect: −0.481; 95% confidence interval: −1.329, −0.027). (B) Association between PWMH and memory deficit was significantly mediated by NLp of IFGoperc.L (Indirect effect: −0.480; 95% confidence interval: −1.341, −0.023). (C) Association between TWMH and VR-DR was significantly mediated by NLp of IFGoperc.L (Indirect effect: −0.513; 95% confidence interval: −2.129, −0.019). (D) Association between PWMH and VR-DR was significantly mediated by NLp of IFGoperc.L (Indirect effect: −0.737; 95% confidence interval: −2.092, −0.013). WMH, white matter hyperintensity; IFGoperc, opercular part of inferior frontal gyrus; L, left; NLp, nodal path length; PWMH, periventricular white matter hyperintensity; TWMH, total white matter hyperintensity; VR-DR, Visual Reproduction–Delay Recall.