Evaluating the impact of glucokinase activation on risk of cardiovascular disease: a Mendelian randomisation analysis

Table 2 Associations of genetically proxied GK activation with CAD and HF risks using different MR methods

Exposure	Method	OR (95% CI)	P
Main analysis: Genetically proxied GK activation (per 1% lower HbA_1c) instrumented by 17 SNPs	CAD (122,733/424,528)
	IVW (P for heterogeneity = 0.292)	0.38 (0.29, 0.51)	8.77 × 10⁻¹¹
	Weighted median	0.39 (0.27, 0.57)	9.56 × 10⁻⁷
	MR-Egger regression	P for intercept = 0.067
	MR-PRESSO (no outliers detected)	P for global test = 0.266
	HF (47,309/930,014)
	IVW (P for heterogeneity = 0.752)	0.54 (0.41, 0.73)	3.55 × 10⁻⁵
	Weighted median	0.51 (0.34, 0.76)	8.62 × 10⁻⁴
	MR-Egger regression	P for intercept = 0.498
	MR-PRESSO (no outliers detected)	P for global test = 0.777
Sensitivity analysis: Genetically proxied GK activation (per 1% lower HbA_1c) instrumented by 2 SNPs	CAD (122,733/42,4528)
	IVW (P for heterogeneity = 0.915)	0.43 (0.26, 0.71)	0.001
	Weighted median	NA
	MR-Egger regression	NA
	MR-PRESSO	NA
	HF (47,309/930,014)
	IVW (P for heterogeneity = 0.868)	0.56 (0.31, 1.02)	0.056
	Weighted median	NA
	MR-Egger regression	NA
	MR-PRESSO	NA

The population was restricted to European ancestry. 1% lower HbA_1c equals to 11 mmol/mol lower
OR, odds ratio; CI, confidence interval; GK, glucokinase; SNP, single-nucleotide polymorphism; CAD, coronary artery disease; HF, heart failure; IVW, inverse variance weighted; MR-PRESSO, MR Pleiotropy RESidual Sum and Outlier; NA, not applicable

ISSN: 1475-2840