Table 2.

Results of a Mendelian randomization study on the effect of plasma urate on CHD with different IV estimators

	Estimates
	Intercept	Slope
	OR (95% CI)	OR (95% CI)
IVW
		1.18 (1.03; 1.34)
MRE
	1.008 (0.998; 1.018)	1.05 (0.87; 1.27)
BMRE
${\sigma }_{\phi }^2 = {10}^{-6}$	1.001 (0.998; 1.003)	1.17 (1.02; 1.34)
${\sigma }_{\phi }^2 = {10}^{-5.8}$	1.001 (0.998; 1.004)	1.16 (1.01; 1.34)
${\sigma }_{\phi }^2 = {10}^{-5.6}$	1.001 (0.997; 1.006)	1.15 (1.00; 1.33)
${\sigma }_{\phi }^2 = {10}^{-5.4 }$	1.002 (0.997; 1.007)	1.14 (0.99; 1.33)
${\sigma }_{\phi }^2 = {10}^{-5.2 }$	1.003 (0.997; 1.009)	1.13 (0.97; 1.32)
WM		1.12 (0.99; 1.27)

Results presented as odds ratio per 1 SD increase in urate with 95% confidence (or credibility) interval (CI) in brackets. The intercept measures the amount of pleiotropy, the slope estimates the effect of plasma urate on CHD. IVW, inverse variance weighted method; MRE, MR-Egger method; BMRE, Bayesian MR-egger method; WM, weighted median method. ${\mu }_{\phi } = 0$, the slope mean and variance priors were 0 and 10 throughout, respectively.