Skip to main content
. 2018 Mar 15;8:4637. doi: 10.1038/s41598-018-22984-5

Table 2.

Multiple linear regressions controlling for group membership with two dichotomous variables (D) for the Lewy body dementia groups.

Network variable Clinical variable Linear regression
[AD + DLB + PDD]
HC AD DLB PDD
Β 1 Β1 p-value ρ (p-value) ρ (p-value) ρ (p-value) ρ (p-value)
PLI high-theta mean NPI hallucinations −9.87 0.0463 — (–) 0.047 (0.824) −0.498 (0.011) −0.127 (0.584)
PLI theta mean Animal naming −31.47 0.0400 0.22 (0.398) −0.36 (0.075) −0.31 (0.129) 0.05 (0.842)
PLI theta mean CAMCOG memory −39.83 0.0204 −0.04 (0.866) −0.23 (0.261) −0.31 (0.137) −0.33 (0.147)
PLI theta mean CAMCOG total −119.70 0.0153 0.06 (0.834) −0.33 (0.105) −0.32 (0.116) −0.21 (0.361)
Leaf theta mean Animal naming −48.14 0.0221 0.02 (0.938) −0.43 (0.027) −0.32 (0.120) −0.02 (0.933)
Leaf theta mean CAMCOG memory −58.91 0.0125 0.19 (0.460) −0.31 (0.130) −0.28 (0.174) −0.36 (0.113)
Leaf theta mean CAMCOG total −169.77 0.0124 0.21 (0.430) −0.37 (0.061) −0.33 (0.110) −0.21 (0.373)
Root theta mean CAMCOG memory −43.24 0.0196 −0.03 (0.914) −0.17 (0.395) −0.29 (0.148) −0.43 (0.054)
Root theta mean CAMCOG total −126.38 0.0178 0.07 (0.788) −0.25 (0.220) −0.32 (0.124) −0.32 (0.161)
Dominant frequency Animal naming 1.36 0.0131 0.3 (0.244) 0.41 (0.040) 0.25 (0.230) 0.03 (0.902)
Dominant frequency CAMCOG attention 0.59 0.0317 −0.27 (0.292) 0.34 (0.088) 0.38 (0.062) −0.38 (0.091)
Dominant frequency CAMCOG executive 1.54 0.0071 0.22 (0.393) 0.38 (0.056) 0.32 (0.124) 0.02 (0.918)
Dominant frequency CAMCOG total 5.24 0.0029 0.08 (0.756) 0.5 (0.010) 0.33 (0.104) −0.05 (0.819)
Dominant frequency MMSE 1.19 0.0294 −0.25 (0.336) 0.45 (0.021) 0.30 (0.141) −0.2 (0.389)
Dominant frequency Trail A −21.88 0.0266 −0.28 (0.270) −0.42 (0.034) −0.27 (0.220) −0.39 (0.112)
Dominant frequency Verbal fluency 6.93 3.69E-05 0.20 (0.439) 0.59 (0.002) 0.52 (0.008) 0.16 (0.559)

Model: Clinical variable ~ β1*network variable + β2*DDLB + β2*DPDD. Shown p-values correspond to the β1 coefficient for the network variable in the regression model and are shown uncorrected. Within group Pearson’s correlation and p-values are also given; full model’s coefficients are given in Supplementary Table S5.