Table 4A.

Unstandardized parameter estimates for phenotypic and biometric models estimating the effects of self-report change in alcohol use on anxiety.

		Do not use vs. use more			Do not use vs. use the same			Do not use vs. use less
		Est	OR [95% CI]	p	Est	OR [95% CI]	P	Est	OR [95% CI]	p
Phenotypic model
	bp	.351	1.42 [1.29, 1.56]	<.001	−.006	.99 [.92, 1.07]	.884	.087	1.09 [.99, 1.20]	.086
Quasi-causal model
	bA	.444	1.56 [1.10, 2.21]	.013	.076	1.08 [.85, 1.36]	.521	.090	1.09 [.78, 1.23]	.599
	bp	.119	1.13 [.96, 1.32]	.135	−.047	.95 [.86, 1.06]	.386	.038	1.04 [.88, 1.23]	.660
Quasi-causal model (with covariates)
	bA	.178	1.19 [.77, 1.86]	.430	.092	1.10,.81, 1.48]	.549	.138	1.15 [.76, 1.74]	.518
	bp	.139	1.15 [.98, 1.35]	.086	−.046	.96 [.86, 1.06]	.388	.038	1.04 [.88, 1.22]	.649
	Age	−.238	.79 [.72,.86]	<.001	−.042	.96 [.91, 1.01]	.131	−.014	.99 [.92, 1.06]	.682
	Sex (F)	.228	1.26 [.95, 1.67]	.114	−.134	.87 [.72, 1.06]	.168	−.070	.93 [.72, 1.21]	.592
RMSEA [90%CI]		.026 [0,.046]			.012 [0,.037]			.015 [0,.039]

Phenotypic model does not include controls for between-pair confounds, whereas quasi-causal model include controls for between-pair confounds. Perceived stress is square root transformed; age is divided by 10.

OR, odds ratio; b_A, amount of variance in perceived stress attributable to additive genetic influences; b_C, amount of variance in perceived stress attributable to shared environmental influences; b_P, phenotypic association between predictor and outcome; RMSEA, root mean square error of approximation.