Skip to main content

Table 3 Bias results for gene-environment interaction effect assessment comparing ordinary statistical methods (OLR and OLMM) to family-based methods (AQTDTM and ALMM) under homogeneous (HP) and stratified (SP) populations. Each time, n cases were simulated with parameters β3 = α7 = 1. Simulations are based on the additive genetic model. † = number that identifies each model in the paper.

From: Gene-environment interaction tests for family studies with quantitative phenotypes: A review and extension to longitudinal measures

 

Two-step models

Mixed models

n

β ^ 3 - β 3

α ^ 7 - α 7

 

OLR (21) †

AQTDT M (22)

OLMM (23)

ALMM (24)

 

HP

SP

HP

SP

HP

SP

HP

SP

100

0.001

-0.687

-0.013

0.066

-0.013

-0.719

0.008

0.029

200

0.001

-0.745

0.034

-0.008

0.007

-0.738

-0.008

-0.007

300

-0.008

-0.713

0.007

0.004

-0.005

-0.716

0.011

0.009

400

0.002

-0.719

0.012

0.01

0.005

-0.711

0.014

0.016

500

0.01

-0.729

0.000

-0.027

-0.001

-0.725

0.011

-0.006

600

0.011

-0.697

0.014

0.008

0.012

-0.704

0.016

0.005

1000

0.003

-0.728

0.007

-0.018

0.052

-0.730

0.011

-0.075

  1. OLR, ordinary linear regression; OLMM, ordinary linear mixed models; ALMM, adjusted linear mixed models; AQTDTM, adjusted quantitative transmission disequilibrium test with mating type indicators