Skip to main content

Table 2 Instrumental variable and direct regression estimates of effect size

From: Mendelian randomization in family data

Effect←Cause

Age (yr)

Regression Estimate (95% CI)a

IV Estimate (95% CI)b

BMI←LDL

0-29

0.102 (0.074, 0.130) c

0.319 (-0.143,1.454)

 

30-44

0.102 (0.085, 0.120)

0.065 (-0.293,0.401)

 

45-60

0.054 (0.035, 0.073)

0.452 (-0.011,1.925)

 

>60

-0.008(- 0.039, 0.023)

0.156 (-0.574,1.406)

SBP←LDL

0-29

0.014 (-0.004, 0.032)

0.008 (-0.303,0.313)

 

30-44

0.024 (0.013, 0.036)

0.026 (-0.187,0.247)

 

45-60

0.013 (-0.001, 0.028)

0.369 (-0.035,2.268)

 

>60

-0.009(- 0.036, 0.017)

0.085 (-0.730,0.963)

HDL←LDL

0-29

-0.151 (-0.195, -0.107)

-0.185(- 1.103,0.865)

 

30-44

-0.122 (-0.149, -0.096)

-0.014(- 0.397,0.393)

 

45-60

-0.106 (-0.136, -0.077)

0.250 (-0.552,2.800)

 

>60

0.085 (0.031, 0.139)

-0.712(- 3.072,0.118)

TG←LDL

0-29

0.428 (0.331, 0.524)

1.770 (0.500,6.230)

 

30-44

0.313 (0.257, 0.369)

-0.029(- 1.073,0.806)

 

45-60

0.252 (0.193, 0.311)

0.954 (-0.767,4.892)

 

>60

-0.178 (-0.273, -0.083)

1.038 (-0.755,7.229)

  1. aThe mean centered year of birth, age, and sex were used as covariates.
  2. bThe mean centered year of birth, age, sex and cholesterol treatment were used as covariates.
  3. cBold indicates statistical significance at α = 0.05.