. 2021 Dec 2;52(2):92–107. doi: 10.1007/s10519-021-10094-4

Table 2.

Relationships between the PGSs and measures of educational attainment and neighborhood-SES, controlled for genetic covariates (10 PCs in both samples as well as genotyping batch in NTR) and sex and age

	NTR (N = 8989 for EA and N = 12,584 for neighborhood)			NEMESIS-2 (N = 3090)
	PGS_allsmok	PGS_EA	PGS_smok-noEA	PGS_allsmok	PGS_EA	PGS_smok-noEA
EA^a
b	− 0.059	0.221	− 0.016	− 0.08	0.228	− 0.031
SE	0.009	0.008	0.009	0.015	0.014	0.014
p	6.24E − 12	2.12E − 144	0.070	4.80E − 08	2.28E − 59	0.030
R²*	2.3%	7.7%	1.8%	1.0%	8.1%	0.2%
Neighborhood^b
b	− 0.039	0.126	− 0.011	0.027	0.001	0.025
SE	0.010	0.0116	0.0104	0.08	0.019	0.018
p	1.63E − 04	< 5E − 300	0.281	0.147	0.950	0.177
R²	0.6%	2.0%	0.5%	0.3%	0.2%	0.3%

The relationships were tested in separate models, so that these models do not control for overlap between the PGSs

^aEducational attainment. In NTR, 4-level variable with 1 = primary school, 2 = lower vocational/ lower secondary school, 3 = intermediate vocational/intermediate and high secondary school, and 4 = higher vocational/ university; in NEMESIS-2, 3-level variable with 1 = primary/lower secondary, 2 = higher secondary, 3 = higher professional education

^bIn NTR, a measure of neighborhood-level income; in NEMESIS-2, a survey-based measure of neighborhood quality

*R² is given for the model excluding age and sex