Skip to main content

Table 2 Statistically significant results reflecting the effect of genetically predicted longer telomere length on CSF biomarkers

From: Genetically predicted telomere length and Alzheimer’s disease endophenotypes: a Mendelian randomization study

 

APOE-ɛ4 non-carriers (N= 157)

High genetic AD risk (N= 106)

Aβ ratio

NfL

p-tau

Causal inference methods

β

SE

p-value

q-value*

β

SE

p-value

q-value*

β

SE

p-value

q-value*

Inverse-variance weighted

−0.007

0.002

3.26E−04

6.52E−04

13.267

2.604

3.50E−07

1.40E−06

−9.186

4.534

4.28E−02

1.48E−01

Maximum likelihood

−0.007

0.002

2.19E−03

4.37E−03

13.044

3.312

8.22E−05

3.29E−04

−8.932

4.563

5.03E−02

1.68E−01

Weighted median

−0.007

0.002

4.41E−03

8.83E−03

13.627

3.568

1.34E−04

5.37E−04

−8.571

6.224

1.69E−01

3.07E−01

Weighted mode

−0.007

0.002

2.62E−03

5.24E−03

13.327

3.259

4.33E−05

1.73E−04

−7.867

7.917

3.20E−01

5.38E−01

Sensitivity methods

p-value

p-value

p-value

Cochran Q, heterogeneity test

0.412

0.254

0.921

MR-PRESSO, global test

0.602

0.505

0.911

MR-Egger, intercept test

0.170

0.742

0.543

  1. amyloid-β, AD Alzheimer’s disease, CSF cerebrospinal fluid, NfL neurofilament light, PRESSO Pleiotropy RESidual Sum and Outlier, SE standard error
  2. *q-value refers to the false discovery rate-adjusted p-value