Skip to main content

Table 4 Association between principal components and subject variables in EPIC

From: Epigenome-wide association study reveals decreased average methylation levels years before breast cancer diagnosis

  First PC Second PC Third PC
% of variance explained 0.064 0.034 0.022
Minimum p value chips 0.327 0.124 0.516
Covariates    
 Case/control status 0.02 0.0005 0.23
 Age 0.00012 0.63 0.16
 Weight 0.84 0.10 0.01
 Height 0.73 0.59 0.60
 BMI (continuous) 0.74 0.14 0.01
 BMI (categorical)    
  Underweight 0.50 0.11 0.44
  Overweight 0.86 0.11 0.34
  Obese 0.96 0.20 0.01
 Physical activity (cat)    
  Moderately inactive 0.05 0.17 0.57
  Moderately active 0.03 0.01 0.38
  Active 0.43 0.73 0.76
 Red meat 0.02 0.08 0.90
 Alcohol 0.23 0.27 0.40
 Folate 0.91 0.0006 0.72
 Smoking    
  Former 0.77 0.72 0.72
  Current 0.15 0.20 0.93
 Age at menarche (cat)    
  12–14 0.53 0.03 0.16
  ≥15 0.20 0.41 0.88
 Age at menopause 0.45 0.73 0.25
 Menopausal state 0.002 0.25 0.78
 Ever pill 0.21 0.62 0.18
 Ever HRT 0.34 0.09 0.60
 ER status 0.42 0.14 0.39
 PR status 0.76 0.44 0.08
  1. To demonstrate that there was no batch effect for the chip, we report the smallest p value for the association between the PCs and all chips.
  2. Italics = p-values <0.01