Table 5.

Results from two-sample summary-level Mendelian randomization^a.

Exposure and outcome GWAS	Method	N SNPs	Betab	SE	p
Migraine80 Years of schooling63	Inverse variance weighted	29	−0.002	0.011	0.824
	Weighted median	29	0.016	0.010	0.090
	Weighted mode	29	0.018	0.011	0.113
	MR Egger	29	−0.014	0.029	0.619
	MR Egger—Intercept	29	0.001	0.002	0.654
Asthma79 Years of schooling63	Inverse variance weighted	8	−0.023	0.005	<0.0001
	Weighted median	8	−0.018	0.007	0.009
	Weighted mode	8	−0.013	0.014	0.396
	MR Egger	8	−0.042	0.038	0.313
	MR Egger—Intercept	8	0.003	0.005	0.632
BMI68 Years of schooling85	Inverse variance weighted	945	−0.158	0.010	<0.0001
	Weighted median	945	−0.109	0.012	<0.0001
	Weighted mode	945	−0.032	0.031	0.306
	MR Egger	945	−0.042	0.028	0.137
	MR Egger—Intercept	945	−0.002	0.000	0.000

^aConducted in MR Base with the TwoSampleMR package. Details of GWAS used in Supplementary Table 3.

^bBetas from Two-Sample MR represent change in the outcome (years of schooling) per unit increase in BMI, or per unit increase in log-odds of having asthma or migraine.