Table 1.

Effect estimates (β) and standard errors (SE) of the instrumental variables (SNPs) on 25(OH)D concentrations and refractive error (RE)

		Effect on 25(OH)D (10 nmol/l)a			RE Europeans N = 37382b				RE Europeans young cohorts N = 7523c				RE Asians N = 8376b
SNP (gene)	EA/ NEA	βzx	SEzx	R2	βzy	SEzy	βiv	SEiv	βzy	SEzy	βiv	SEiv	βzy	SEzy	βiv	SEiv
Based on SNPs used in Afzal et al.27
rs7944926 (DHCR7)	A/G	−0.20	0.019	0.4%	−0.011	0.018	0.06	0.09	−0.038	0.025	0.19	0.13	0.020	0.042	−0.1	0.21
rs12794714 (CYP2R1)	A/G	−0.30	0.022	0.6%	−0.015	0.016	0.05	0.05	0.036	0.021	−0.12	0.07	0.015	0.042	−0.05	0.14
Combined	−	−	−	1%	−	−	0.05	0.05	−	−	−0.05	0.06	−	−	−0.06	0.12
Based on SNPs used in Mokry et al.28
rs10741657 (CYP2R1)	C/T	−0.28	0.07	0.6%	0.007	0.017	−0.02	0.06	0.034	0.021	−0.12	0.08	0.024	0.050	−0.09	0.18
rs12785878 (DHCR7)	G/T	−0.31	0.07	0.6%	−0.009	0.019	0.03	0.06	−0.035	0.028	0.12	0.09	0.025	0.042	−0.08	0.14
rs2282679 (GC)	C/A	−0.26	0.07	0.5%	0.021	0.018	−0.09	0.07	−0.044	0.023	0.17	0.09	−0.038	0.039	0.15	0.15
rs6013897 (CYP24A1)	A/T	−0.15	0.08	0.1%	0.012	0.020	−0.08	0.13	0.008	0.027	−0.05	0.19	−0.032	0.042	0.22	0.29
Combined	−	−	−	1.8%	−	−	−0.02	0.03	−	−	0.03	0.05	−	−	0.01	0.09

β_zx refers to the effect of the SNP on 25(OH)D, β_zy refers to the effect on RE and β_iv shows the causal effect estimates [dioptres (D) per 10 nmol/l 25(OH)D increase].

EA/NEA: effect allele/non-effect allele.

^aEffect estimates were extracted from Azfal et al.²⁷ or Mokry et al.²⁸, respectively. Standard errors and R² estimates based on Mokry et al. were computed from the reported F-statistics.

^bEffect estimates are based on those from the large RE GWAS meta-analysis from CREAM.³⁴

^cEffect estimates were computed using data from young individuals from the TEST, BATS and ALSPAC cohorts (mean age 16).