Skip to main content
. 2016 Sep 29;6:34051. doi: 10.1038/srep34051

Table 1. Association of the methylation at the human bace-1 promoter with β-amyloid load and PHFtau tangle density.

Predictor β-Amyloid Load
PHFtau Tangle Density
Parameter Estimate Std Error p-value Parameter Estimate Std Error p-value
cg01025770 −0.0113 0.0405 0.7807 −0.0102 0.0506 0.8405
cg02062003 −0.0111 0.0404 0.7829 −0.0367 0.0504 0.4673
cg07119404 0.0451 0.0412 0.2736 0.0369 0.0514 0.4735
cg07619960 0.0132 0.0403 0.7425 −0.0450 0.0502 0.3710
cg14112985 0.0148 0.0422 0.7263 −0.0564 0.0527 0.2847
cg15427448 −0.0215 0.0419 0.6082 −0.0091 0.0523 0.8627
cg16822189 −0.0742 0.0393 0.0591 −0.0714 0.0490 0.1459
cg17007365 0.0131 0.0392 0.7382 0.1582 0.0486 0.0012*
cg21048949 −0.0970 0.0404 0.0167 −0.0779 0.0506 0.1239
cg22261612 0.0220 0.0394 0.5771 0.0424 0.0491 0.3882
cg23435082 −0.0539 0.0400 0.1787 0.0599 0.0499 0.2308
cg26462656 −0.0733 0.0425 0.0848 −0.0452 0.0531 0.3944

The estimate represents the change in pathology (amyloid or tangles) with one unit increase in methylation. Estimates, standard errors, and p-values at each row came from separate multivariable linear regression models with either β-amyloid load or PHFtau tangle density as the continuous outcome and DNA methylation as the predictor, adjusted for age, sex, bisulfite conversion efficiency and batch. Note that although methylation at cg21048949 was nominally associated with lower β-amyloid, the result did not survive Bonferroni correction for multiple testing.