Skip to main content

Table 3 Multiple Cox regression analysis on significant factors of developing dyslipidaemia in training set

From: Genetic factors increase the identification efficiency of predictive models for dyslipidaemia: a prospective cohort study

Variables

β

S.E.

Wald

P

HR (95%CI)

Conventional model

Age

0.005

0.003

3.017

0.082

1.005(0.999, 1.010)

Family history of diabetes

0.194

0.125

2.429

0.119

1.215(0.951, 1.551)

Physical activity

 Low

Reference

 Moderate

0.793

0.080

99.087

< 0.001

2.210(1.890, 2.583)

 High

0.324

0.071

20.810

< 0.001

1.383(1.203, 1.590)

BMI

0.016

0.010

2.777

0.096

1.016(0.997, 1.036)

TG

0.292

0.074

15.609

< 0.001

1.339(1.158, 1.548)

HDL-C

−2.103

0.196

114.907

< 0.001

0.122(0.083, 0.179)

LDL-C

0.284

0.052

29.792

< 0.001

1.329(1.200, 1.472)

Conventional + GRS model

Age

0.005

0.003

2.887

0.089

1.005(0.999, 1.010)

Family history of diabetes

0.198

0.125

2.517

0.113

1.219(0.954, 1.557)

Physical activity

 Low

Reference

 Moderate

0.802

0.080

101.097

< 0.001

2.230(1.907, 2.607)

 High

0.328

0.071

21.347

< 0.001

1.389(1.208, 1.596)

BMI

0.017

0.010

2.998

0.083

1.017(0.998, 1.037)

TG

0.281

0.074

14.410

< 0.001

1.325(1.146, 1.532)

HDL-C

−2.095

0.195

114.889

< 0.001

0.123(0.084, 0.180

LDL-C

0.286

0.052

29.968

< 0.001

1.330(1.201, 1.474)

Weighted GRS

0.276

0.088

9.925

0.002

1.318(1.110, 1.565)

  1. Note: The predictors of the conventional model are variables that are significantly associated with dyslipidaemia in simple Cox regression analysis. GRS is added to the conventional model to construct the conventional+GRS model
  2. Abbreviations: BMI body mass index, TG triglyceride, HDL-C high density lipoprotein, LDL-C low density lipoprotein, GRS genetic risk score