Skip to main content

Table 2 PGS association with eGFR in the elderly (AugUR) and general adults (KORA S4)

From: Polygenic scores for estimated glomerular filtration rate in a population of general adults and elderly – comparative results from the KORA and AugUR study

Model

Residual variance

[ml/min/1.73 m2]2

Beta (PGS) [ml/min/1.73 m2] [95% CI]

p-value (PGS)

R2 (PGS) [%] [95% CI]

eGFRcrea AugUR

    Model 1

228.93

-0.23 [-0.28, -0.19]

4.80 × 10–25

4.6 [3.95, 5.25]

    Model 2

218.72

-0.22 [-0.27, -0.18]

2.90 × 10–24

4.4 [3.80, 5.00]

    Model 3

213.33

-0.23 [-0.27, -0.19]

1.07 × 10–25

4.7 [4.02, 5.37]

    Model 4

198.70

-0.23 [-0.27, -0.19]

5.04 × 10–29

5.3 [4.47, 6.13]

eGFRcrea KORA S4

    Model 1

157.65

-0.29 [-0.33, -0.26]

5.44 × 10–66

9.6 [7.62, 11.58]

    Model 2

157.18

-0.29 [-0.32, -0.26]

1.94 × 10–65

9.6 [7.62, 11.58]

    Model 3

157.14

-0.29 [-0.32, -0.26]

1.89 × 10–65

9.6 [7.62, 11.58]

    Model 4

156.37

-0.29 [-0.32, -0.26]

5.76 × 10–66

9.6 [7.62, 11.58]

eGFRcys AugUR

    Model 1

234.68

-0.35 [-0.42, -0.27]

2.70 × 10–20

3.6 [3.19, 4.01]

    Model 2

216.08

-0.31 [-0.39, -0.24]

3.60 × 10–18

3.2 [2.87, 3.53]

    Model 3

211.93

-0.32 [-0.39, -0.25]

8.77 × 10–19

3.3 [2.95, 3.65]

    Model 4

191.64

-0.31 [-0.37, -0.24]

2.34 × 10–19

3.5 [3.11, 3.89]

eGFRcys KORA S4

    Model 1

186.38

-0.35 [-0.41, -0.29]

3.42 × 10–32

4.7 [4.10, 5.30]

    Model 2

177.57

-0.35 [-0.40, -0.29]

1.24 × 10–32

4.7 [4.10, 5.30]

    Model 3

177.29

-0.35 [-0.40, -0.29]

7.56 × 10–33

4.8 [4.18, 5.42]

    Model 4

175.89

-0.34 [-0.40, -0.29]

1.14 × 10–32

4.7 [4.10, 5.30]

  1. In AugUR (70–95 years, n = 2,272) and KORA S4 (20–69 years, n = 2,900), we derived the PGS association with eGFRcrea and eGFRcys via linear regression,\(Y_i\;=\;\beta_0+\;\beta_1\;PGS_i\;+\;\varepsilon_i,\;i\;=\;1,\dots,n\)with \(\varepsilon_i\sim\left(0,\sigma^2\right)\) independent and identically distributed. Yi denotes the residuals of individual i adjusted for i) age and sex (model 1), ii) additionally for BMI (model 2), iii) additional for diabetes and hypertension (model 3), and iv) additionally for CAD and high-ceiling diuretics intake (model 4). All models were further adjusted for 10 principal components. Shown are the residual eGFR variance, the regression coefficients (beta) per one unit increase in the PGS with 95% confidence interval (CI) and p-values, and the R2 of the PGS. One unit in the PGS corresponds to one eGFR-lowering allele of average eGFR-effect