Table 3:

Mendelian randomization estimates for the effect of circulating vitamin D on delirium

Method *	Estimate^	95% CI		P
IVW	−0.48	−0.66	−0.30	2×10−7
Weighted median	−0.51	−1.02	0.00	0.048
MR-Egger	−0.41	−0.73	−0.10	0.010
MR-Egger (intercept)	0.00	−0.03	0.02	0.755
Radial IVW	−0.48	−0.76	−0.20	0.001
Radial MR-Egger	−0.49	−1.01	0.04	0.144
Radial MR-Egger (intercept)	0.01	−0.87	0.90	0.977
IVW (excluding rs3755967)	−0.41	−1.01	0.18	0.173

^*IVW = penalized robust inverse-variance weighted regression (assumes there is no unbalanced horizontal pleiotropy); Weighted median = penalized weighted median estimate (assumes less than 50% of the weight in the analysis comes from invalid instruments); MR-Egger = penalized robust Egger regression (assumes the genetic variants’ effect is not correlated with any pleiotropic effect on the outcome); MR-Egger (intercept) = like IVW but the MR-Egger incept is not fixed, as deviation from the null is used to test for possible horizontal pleiotropy; Radial IVW = radial inverse-variance weighted regression using modified second-order weights (no significant outliers detected); Radial MR-Egger = intercept in unconstrained and assumes that pleiotropic effects are independent of the Radial weights.

^{^}ln(HR) per ln(vitamin D). See Supplementary Table S6 for full details.