Skip to main content

Table 5 Bivariate and univariate model fitting comparisons under the competing ACE, AE, CE, and E models

From: Association of neurofilament light chain with renal function: mechanisms and clinical implications

Bivariate

Model

ep

− 2LL

df

Δ− 2LL

Δdf

p

AIC

Correlated factors

ACE

11

4980.17

1885

   

5002.17

AE

8

4981.28

1888

1.11

3

0.7374

4997.28

CE

8

5001.96

1888

21.79

3

< 0.001

5017.96

E

5

5173.55

1891

193.38

6

< 0.001

5183.55

Univariate

Model

ep

− 2LL

df

Δ− 2LL

Δdf

p

AIC

Creatinine

ACE

4

2536.65

932

   

2544.65

AE

3

2537.12

933

0.47

1

0.4945

2543.12

CE

3

2546.23

933

9.58

1

0.0020

2552.23

E

2

2655.25

934

118.60

2

< 0.001

2659.25

Plasma NfL

ACE

4

2633.55

956

   

2641.55

AE

3

2633.62

957

0.07

1

0.7952

2639.62

CE

3

2645.60

957

12.05

1

< 0.001

2651.60

E

2

2723.36

958

89.81

2

< 0.001

2723.36

  1. Best fitting model is in bold font
  2. A additive genetic, C common or shared environment, E non-shared environment, ep number of estimated parameters, − 2LL − 2 × log-likelihood, Δ− 2LL change in − 2 × log-likelihood, Δdf change in degrees of freedom, AIC Akaike Information Criteria