Causal Associations Between Serum Bilirubin Levels and Decreased Stroke Risk: A Two-Sample Mendelian Randomization Study

Abstract

Objective: A number of epidemiological studies have reported that decreased serum bilirubin, an endogenous antioxidant, is associated with cardiovascular disease. However, previous Mendelian randomization analyses conducted using a single sample have shown no evidence of association. Approach and Results: A 2-sample summary Mendelian randomization study was performed by obtaining exposure and outcome data from separate nonoverlapping samples. We utilized data from the KoGES (Korean Genome and Epidemiology Study; n=25 406) and KCPS-II (Korean Cancer Prevention Study-II; n=14 541) biobank for serum bilirubin and stroke, respectively. Using KoGES, a total of 1784 single nucleotide polymorphisms associated with serum bilirubin levels were discovered using a genome-wide significance threshold (P<5×10^-8), of which 10 single nucleotide polymorphisms were identified as independent (R²<0.005) and adopted as genetic instruments. From KCPS-II, total and ischemic stroke cases were identified (n=1489 and n=686), with 12 366 acting as controls. Various 2-sample summary Mendelian randomization methods were employed, with Mendelian randomization estimates showing an inverse causal association between serum bilirubin levels and total stroke risk (odds ratio, 0.481 [95% CI, 0.234-0.988]; P=0.046). This association increased in magnitude when restricting the analysis to ischemic stroke cases (odds ratio, 0.302 [95% CI, 0.105-0.868]; P=0.026).

Conclusions: Our findings provide evidence of significant causal relationship between high levels of bilirubin and decreased stroke risk in Korean population in agreement with observational approaches. This highlights the potential for bilirubin to serve as a therapeutic target for oxidative stress-related diseases such as stroke and suggests that previous findings were not a consequence of unmeasured confounding.

Keywords: bilirubin; cardiovascular diseases; causality; single nucleotide polymorphism; stroke.

Figure 1.

Scatter plot to visualize causal effect of serum bilirubin on total stroke risk (A) and ischemic stroke risk (B). The slope of the straight line indicates the magnitude of the causal association. IVW indicates inverse-variance weighted; and MR, Mendelian randomization.

Figure 2.

Funnel plots to visualize overall heterogeneity of Mendelian randomization (MR) estimates for the effect of bilirubin on the total stroke (A) or ischemic stroke risk (B). IVW indicates inverse-variance weighted.

Figure 3.

Radial plots to visualize individual outlier single nucleotide polymorphisms (SNPs) in the Mendelian randomization (MR) estimates for total stroke (A) or ischemic stroke risk (B). Black dots show valid SNPs and green dots display invalid outlier SNPs. There is no significant outlier SNP in these present plots. IVW indicates inverse-variance weighted.