Serum liver enzymes and risk of stroke: Systematic review with meta-analyses and Mendelian randomization studies

Abstract

Background and purpose: Previous observational studies have identified correlations between liver enzyme levels and stroke risk. However, the strength and consistency of these associations vary. To comprehensively evaluate the relationship between liver enzymes and stroke risk, we conducted meta-analyses complemented by Mendelian randomization (MR) analyses.

Methods: Following the PRISMA guidelines, we performed meta-analyses of prospective studies and conducted subgroup analyses stratified by sex and stroke subtype. Subsequently, adhering to the STROBE-MR guidelines, we performed two-sample bidirectional univariable MR (UVMR) and multivariable MR (MVMR) analyses using the largest genome-wide association studies summary data. Finally, the single-nucleotide polymorphisms associated with liver enzymes on sex differences underwent gene annotation, gene set enrichment, and tissue enrichment analyses.

Results: In the meta-analyses of 17 prospective studies, we found the relative risks for serum γ-glutamyl transferase (GGT) and alkaline phosphatase (ALP) were 1.23 (95% CI: 1.16-1.31) and 1.3 (95% CI: 1.19-1.43), respectively. Subgroup analyses revealed sex and stroke subtype differences in liver enzyme-related stroke risk. Bidirectional UVMR analyses confirmed that elevated GGT, alanine aminotransferase, and aspartate aminotransferase levels were associated with increased stroke occurrence. The primary results from the MVMR analyses revealed that higher ALP levels significantly increased the risk of stroke and ischemic stroke. Gene set and tissue enrichment analyses supported genetic differences in liver enzymes across sexes.

Conclusions: Our study provides evidence linking liver enzyme levels to stroke risk, suggesting liver enzymes as potential biomarkers for early identification of high-risk individuals. Personalized, sex-specific interventions targeting liver enzymes could offer new strategies for stroke prevention.

Keywords: alanine transaminase; alkaline phosphatase; aspartate aminotransferase; gamma‐glutamyltransferase; ischemic stroke.

FIGURE 1

PRISMA flow diagram for study selection for the systematic review (a) and workflow for Mendelian randomization analyses (b). ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, γ‐glutamyl transferase; MR, Mendelian randomization; MVMR, multivariable Mendelian randomization; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta‐Analyses; SNP, single‐nucleotide polymorphism; UVMR, univariable Mendelian randomization.

FIGURE 2

Forest plot for the associations of serum γ‐glutamyl transferase (n = 13), alkaline phosphatase (n = 5), alanine aminotransferase (n = 6), and aspartate aminotransferase (n = 4) with risk of stroke. Squares represent the estimate of relative risk for each study; the horizontal lines represent the 95% confidence interval (95% CI), and diamonds represent the overall estimate and its 95% CI. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CI, confidence interval; D + L, DerSimonian and Laird random effects model; GGT, γ‐glutamyl transferase; I‐V, inverse variance fixed effects model; RR, relative risk.

FIGURE 3

Subgroup analyses on the associations of serum γ‐glutamyl transferase level with stroke stratified by sex, stroke subtype, and stroke subtype‐sex. Squares represent the estimate of relative risk for each study; the horizontal lines represent the 95% confidence interval (95% CI), and the diamonds represent the overall estimate and its 95% CI. CI, confidence interval; D + L, DerSimonian and Laird random effects model; GGT, γ‐glutamyl transferase; HS, hemorrhagic stroke; ICH, intracerebral hemorrhage; IS, ischemic stroke; I‐V, inverse variance fixed effects model; RR, relative risk; SAH, subarachnoid hemorrhage.

FIGURE 4

Subgroup analyses on the associations of serum alkaline phosphatase level with stroke stratified by sex, stroke subtype, and stroke subtype‐sex. Squares represent the estimate of relative risk for each study; the horizontal lines represent the 95% confidence interval (95% CI), and the diamonds represent the overall estimate and its 95% CI. ALP, alkaline phosphatase; CI, confidence interval; D + L, DerSimonian and Laird random effects model; HS, hemorrhagic stroke; ICH, intracerebral hemorrhage; IS, ischemic stroke; I‐V, inverse variance fixed effects model; RR, relative risk.

FIGURE 5

Scatterplots assess the effects of aspartate aminotransferase on subarachnoid hemorrhage (a) and intracerebral hemorrhage (b), the effect of alanine aminotransferase on stroke (c) and ischemic stroke (d), and the effect of γ‐glutamyl transferase on cardioembolic stroke (e). ALT, alanine aminotransferase; AST, aspartate aminotransferase; CES, cardioembolic stroke; GGT, γ‐glutamyl transferase; ICH, intracerebral hemorrhage; IS, ischemic stroke; MR, Mendelian randomization; SAH, subarachnoid hemorrhage; SNP, single nucleotide polymorphism.

FIGURE 6

Multivariable Mendelian randomization analyses of the association between liver enzymes and the risk of stroke and its subtypes. Squares represent the estimate of odds ratio, and the horizontal lines represent the 95% confidence interval. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CI, confidence interval; GGT, γ‐glutamyl transferase; MVMR, multivariable Mendelian randomization; OR, odds ratio; SNP, single‐nucleotide polymorphism.