Causal relationship between smoking and level of C-reactive protein： a two-sample Mendelian randomization study

BackgroundPrevious investigations have illuminated the correlation between smoking and C-reactive protein （CRP）， but previous research findings may be influenced by other confounding factors. The causal relationship of CRP in smoking-related pathological process requires further exploration.ObjectiveTo investigate the causal relationship between smoking behavior and CRP by utilizing the cumulative statistical data from existing genome-wide association studies （GWAS）， so as to provide references for formulating relevant public health policies and smoking intervention measures.MethodsThis research utilized the GWAS summary statistics for CRP and four smoking phenotypes： age of initiation of regular smoking， smoking initiation， smoking cessation and cigarettes per day-selecting independent genetic loci correlated with smoking and CRP as instrumental variables. The study employed the inverse variance weighted method （IVW） and the weighted median approach for two-sample Mendelian randomization （MR） analysis to explore the bidirectional causal relationship between smoking and CRP. The Cochran's Q test was applied to assess heterogeneity among single nucleotide polymorphisms （SNPs）. MR pleiotropy residual sum and outlier was used to detect SNP outliers. MR-Egger intercept test examined the horizontal pleiotropy of SNPs. Leave-one-out sensitivity analysis assessed the impact of individual SNP on the Mendelian randomization results.ResultsThe MR analysis revealed a bidirectional causal relationship between CRP and smoking initiation （β=0.170， P=0.01） （with smoking initiationas the exposure）， （β=0.040， P=0.001） （with CRP as the exposure）.ConclusionSmoking may lead to alterations in CRP levels， while changes in CRP levels could also influence individual's propensity to initiate smoking.