Table 4.

Association of low-density lipoprotein cholesterol (LDL-C)-lowering and high density lipoprotein cholesterol (HDL-C)-raising with coronary stenosis severity change. Univariate (models 1 and 2) and multivariate (model 3) linear regression.

	Change in % stenosis (Δ%S)
	All 3 studies N=445		FATS N=120	HATS N=146	AFREGS N=130	Interaction with study**
Model	Regr. Coef*. (95% CI)	p-value	Regr. Coef. (95% CI)	Regr. Coef. (95% CI)	Regr. Coeff (95% CI)	p-value
1. LDL-C-lowering (per 10%)	−0.34 (−0.49 – −0.18)	<0.001	−0.60 (−0.97 – −0.23)	−0.23 (−0.47 – 0.01)	−0.24 (−0.45 – −0.03)	0.3
2. HDL-C-raising (per 10%)	−0.29 (−0.46 – −0.12)	<0.001	−0.60 (−1.02 – −0.18)	−0.35 (−0.71 – 0.00)	−0.11 (−0.30 – 0.07)	0.3
3. LDL-C-lowering (per 10%)	−0.29 (−0.45 – −0.13)	<0.001	−0.55 (−0.95 – −0.16)	−0.18 (−0.42 – 0.07)	−0.22 (−0.44 – 0.00)	0.6
HDL-C-raising (per 10%)	−0.21 (−0.39 – −0.04)	0.02	−0.48 (−0.90 – −0.06)	−0.29 (−0.66 – 0.08)	−0.05 (−0.25 – 0.14)	0.3

Outliers excluded: 5 for models including data from all three studies (including testing for interaction with study), no outliers for FATS (except for LDL model with one outlier), one outlier for HATS, 3 outliers for AFREGS

^*Adjusted for the study effect,

^**interaction of the predictor with the study variable (ANOVA)