Table 1. Meta-analysis results from survey years 1999–2006 with a threshold of FDR < 1% for systolic blood pressure (SBP).

SBP	Random Effects Meta-Analysis 1999–2006 (training dataset)
Variable Name (change per 1 SD of the logarithm of factor)	Effect Size (mmHg)	p-value	Total N	FDR
Alcohol	0.04	1.4 × 10−11	22986	2.0 × 10−9
Enterolactone	−0.08	2.0 × 10−11	7850	2.2 × 10−9
Equol	−0.05	1.9 × 10−10	7597	1.3 × 10−8
Mercury, urine	−0.08	2.2 × 10−8	6986	1.3 × 10−6
Mono-(3-carboxypropyl) phthalate	−0.05	3.1 × 10−6	6081	0.0001
Folate, serum	−0.05	4.1 × 10−6	22269	0.0002
2,5-dichlorophenol	−0.06	4.4 × 10−6	7547	0.0002
Cadmium, urine	−0.08	1.1 × 10−5	7457	0.0003
4-fluoro-3-phenoxybenzoic acid	−0.02	1.5 × 10−5	3647	0.0004
Vitamin A	−0.03	7.0 × 10−5	17829	0.002
Retinol	−0.03	9.0 × 10−5	17829	0.002
Thallium, urine	−0.07	0.0002	7549	0.003
Cesium, urine	−0.07	0.0003	7613	0.004
Cadmium	0.03	0.0003	22493	0.005

Variables were log transformed and z-standardized therefore results reflect a change per 1 standard deviation (SD). FDR = False Discovery Rate. N = number of participants measured.