Skip to main content
. 2024 Feb 15;59(2):342–370. doi: 10.1080/00273171.2023.2283634

Table 4.

Parameter estimates in the best-fitting IV-CLPM model applied to smoking status and alcohol use.

Parameter Description Estimate SE Wald’s T p-Value
Regression paths
  bx1 PGS of Smoking (Smk) → Smk at T1 0.146 0.016 8.901 0.0000
  bx2 PGS of Smk → Smk at T2 0.014 0.013 1.085 0.2777
  by1 PGS of Alcohol (Alc) → Alc at T1 0.169 0.021 8.054 0.0000
  by2 PGS of Alc → Alc at T2 0.037 0.018 2.036 0.0418
  bx2x1 Smk at T1 → Smk at T2 0.808 0.022 36.461 0.0000
  by2x1 Smk at T1 → Alc at T2 (Distal) 0.065 0.027 2.404 0.0162
  bx1y1 Alc at T1 → Smk at T1 (Proximal) 0.196 0.094 2.095 0.0362
  by2y1 Alc at T1 → Alc at T2 0.662 0.015 44.032 0.0000
(Co)variances
  Vpx Variance of the PGS of Smk 1.000 0.020 49.472 0.0000
  Cixiy Covariance of the PGSs 0.196 0.015 13.441 0.0000
  Vpy Variance of the PGS of Alc 1.000 0.020 49.472 0.0000
  Vx1 Residual variance of Smk at T1 0.726 0.040 18.248 0.0000
  Cx1y1 Covariance of residuals of Smk and Alc at T1 −0.119 0.158 −0.751 0.4526
  Vy1 Residual variance of Alc at T1 1.679 0.040 41.973 0.0000
  Vx2 Residual variance of Smk at T2 0.179 0.010 18.803 0.0000
  Cx2y2 Covariance of residuals of Smk and Alc at T2 0.047 0.013 3.611 0.0003
  Vy2 Residual variance of Alc at T2 0.819 0.025 33.486 0.0000
Covariates
  b_smk_age Age → Smk 0.007 0.001 8.026 0.0000
  b_alc_age Age → Alc 0.009 0.001 8.014 0.0000
  b_smk_female Female sex → Smk −0.088 0.030 −2.974 0.0029
  b_alc_female Female sex → Alc −0.812 0.039 −21.041 0.0000
Intercepts
  b0_resSmkPRS Intercept (mean) of the PGS of Smk 0.000 0.014 0.000 1.0000
  b0_resAlcPRS Intercept (mean) of the PGS of Alc 0.000 0.014 0.000 1.0000
  b0_smk1 Intercept of Smk at T1 −0.347 0.049 −7.007 0.0000
  b0_alc1 Intercept of Alc at T1 3.818 0.062 61.984 0.0000
  b0_smk2 Intercept of Smk at T2 −0.442 0.052 −8.466 0.0000
  b0_alc2 Intercept of Alc at T1 3.580 0.065 54.752 0.0000

Note. Rows shown in bold indicate estimates of the causal paths between smoking status (latent liability scale of an ordinal variable with levels: never–former–current smoking) and alcohol use (drinks per week).