Table 4. Odds ratios (95% CIs) for environmental risk score (ERS) categorized by quintile^a (n = 3847).

Phenotypeb	Single-phenotype Approachc		Multi-phenotype Approachd
	ERS1e	ERS2f	ERS1e	ERS2f
Total cholesterol	1.450 (1.112, 1.892)	1.722 (1.317, 2.252)	1.781 (1.337, 2.374)	1.564 (1.191, 2.054)
HDL	1.372 (1.077, 1.748)	1.450 (1.144, 1.838)	1.471 (1.142, 1.894)	1.565 (1.230, 1.990)
LDL	1.824 (1.394, 2.386)	1.820 (1.391, 2.381)	1.357 (1.061, 1.735)	1.262 (0.973, 1.637)
Triglyceride	1.843 (1.366, 2.487)	1.536 (1.147, 2.056)	1.758 (1.275, 2.424)	2.027 (1.521, 2.703)

^aOdds ratios for dichotomized phenotype (high vs. low) comparing subjects with ERS in the top 20% to those in the bottom 20%, adjusted for covariates and micronutrients.

^bDichotomization thresholds: 200 mg/dL for total cholesterol, 40 mg/dL (male) or 50 mg/dL (female) for HDL, 130 mg/dL for LDL and 150 mg/dL for triglyceride.

^cPollutants selected by single-phenotype regression (n = 13, 9, 5 and 27 for total cholesterol, HDL, LDL and triglyceride, respectively) to construct ERS, adjusted for phenotype-specific micronutrients.

^dPollutants selected by multi-phenotype regression (n = 45) to construct ERS, adjusted for union of selected micronutrients (n = 14).

^eERS constructed with coefficient estimates from single-pollutant models as weights.

^fERS constructed with coefficient estimates from multi-pollutant models as weights.