Table 3 Improvement in risk prediction for ACS by methylation risk score and polygenic risk score beyond traditional risk factors

From: Genome-wide DNA methylation profiling in blood reveals epigenetic signature of incident acute coronary syndrome

Index

DFTJ cohort (n = 1502)

CKB cohort (n = 952)

Difference in estimate

95% CI

P valuea

Difference in estimate

95% CI

P valuea

MRS+ referenceb

 AUC

0.058

0.035–0.080

<0.001

0.104

0.077–0.131

<0.001

 NRI

0.364

0.259–0.469

<0.001

0.430

0.306–0.554

<0.001

 IDI

0.051

0.039–0.062

<0.001

0.075

0.058–0.092

<0.001

PRSBBJ+ referenceb

 AUC

0.011

−0.002–0.023

0.101

0.002

0.0001–0.004

0.035

 NRI

0.227

0.121–0.334

<0.001

0.115

0.012–0.217

0.028

 IDI

0.011

0.005–0.016

<0.001

0.002

0.000–0.003

0.004

PRSCHN+ referenceb

 AUC

0.001

−0.008–0.010

0.896

0.001

−0.0007–0.002

0.373

 NRI

0.077

−0.030–0.184

0.160

0.155

0.045–0.265

0.006

 IDI

0.002

−0.001–0.006

0.208

0.001

0.000–0.002

0.001

  1. ACS acute coronary syndrome, AUC area under ROC curve, CKB cohort the China Kadoorie Biobank cohort, DFTJ cohort Dongfeng-Tongji cohort, IDI integrated discrimination improvement, MRS methylation risk score, NRI net reclassification improvement, PRS polygenic risk score, PRSBBJ PRS constructed based on a method published in BioBank Japan cohort, PRSCHN PRS constructed based on a method published in Chinese populations.
  2. aP values were determined using two-sided statistical tests.
  3. bThe reference risk model included age, sex, BMI, smoking status, drinking status, hypertension, dyslipidemia, and diabetes.
Back to article page