Table 2 CpGs with significant causal effects on serum urate or gout from Mendelian randomization analysis.

Probe ID	Chr	Position (b37)	Nearest gene	No. of meQTLs used for MR analysis	Odds ratio	Effect size^c	SE	p value	p value for pleiotropy	Q statistic for heterogeneity	p value for heterogeneity
Causal effects of CpGs on serum urate
cg02387843	4	9892887	SLC2A9	6	–	−0.52	0.14	2.14E–04	2.40E–01	1109.2	7.54E–239
cg13841979^a	4	9990048	SLC2A9	10	–	−0.46	0.12	1.38E–04	1.49E–01	1836.5	<1E–300
cg03725404^b	4	9998017	SLC2A9	8	–	−0.65	0.14	4.36E–06	1.95E–01	1461.0	1.53E–312
cg11266682^a,b	4	10021025	SLC2A9	11	–	0.21	0.06	8.80E–04	3.33E–01	3173.1	<1E–300
Causal effects of CpG on gout
cg03725404^b	4	9998017	SLC2A9	8	0.43	−0.85	0.14	2.98E–09	1.01E–01	105.8	1.52E–20

Odds ratio and log(odds ratio) of CpGs on gout were estimated per SD in rank-based transformed DNA methylation beta levels.
P values from MR analysis were two-sided and obtained from inverse-variance multiplicative random effect method.
P value for heterogeneity was based on Cochran’s Q test.
The cis methylation QTLs (meQTLs) used as instruments for the urate-associated CpGs were provided by GoDMC (N ≤ 27,750). The summary statistics of serum urate and gout were from individuals of European ancestry (N = 288,649 for urate and N = 692,537 for gout). In total, 27 of the urate-associated CpGs had ≥4 meQTLs for MR analysis (Methods). The significance level was set at 1.9E–3 (=0.05/27).
Chr Chromosome, MR Mendelian randomization, SE standard error, SD standard deviation.
^aCpGs with significant mediating effect for two urate GWAS index SNPs (Supplementary Data 12 and 13).
^bCpGs associated with gene expression of SLC2A9 in monocytes (Supplementary Data 14).
^cEffect size of CpGs on serum urate estimated in mg/dL per SD in rank-based transformed DNA methylation beta levels.

Quick links

Search