Exploring the bidirectional causal associations between pain and circulating inflammatory proteins: A Mendelian randomization study

Abstract Multisite chronic pain (MCP) and site‐specific chronic pain (SSCP) may be influenced by circulating inflammatory proteins, but the causal relationship remains unknown. To overcome this limitation, two‐sample bidirectional Mendelian randomization (MR) analysis was used to analyse data for 91 circulating inflammatory proteins, MCP and SSCP encompassing headache, back pain, shoulder pain, hip pain, knee pain, stomach abdominal pain and facial pain. The primary MR method used was inverse variance weighting, sensitivity analyses included weighted median, MR pleiotropy residual sum and outlier and the Egger intercept method. Heterogeneity was also detected using Cochrane's Q test and leave‐one‐out analyses. Finally, a causal relationship between 29 circulating inflammatory proteins and chronic pain was identified. Among these proteins, 14 exhibited a protective effect, including MCP (T‐cell surface glycoprotein cluster of differentiation 5), headache (4E‐binding protein 1 [4EBP1], cluster of differentiation 40, cluster of differentiation 6 and C‐X‐C motif chemokine [CXCL] 11), back pain (leukaemia inhibitory factor), shoulder pain (fibroblast growth factor [FGF]‐5 and interleukin [IL]‐18R1), stomach abdominal pain (tumour necrosis factor [TNF]‐α), hip pain (CXCL1, IL‐20 and signalling lymphocytic activation molecule 1) and knee pain (IL‐7 and TNF‐β). Additionally, 15 proteins were identified as risk factors for MCP and SSCP: MCP (colony‐stimulating factor 1, human glial cell line‐derived neurotrophic factor and IL‐17C), headache (fms‐related tyrosine kinase 3 ligand, IL‐20 receptor subunit α [IL‐20RA], neurotrophin‐3 and tumour necrosis factor receptor superfamily member 9), facial pain (CXCL1), back pain (TNF), shoulder pain (IL‐17C and matrix metalloproteinase‐10), stomach abdominal pain (IL‐20RA), hip pain (C‐C motif chemokine 11/eotaxin‐1 and tumour necrosis factor ligand superfamily member 12) and knee pain (4EBP1). Importantly, in the opposite direction, MCP and SSCP did not exhibit a significant causal impact on circulating inflammatory proteins. Our study identified potential causal influences of various circulating inflammatory proteins on MCP and SSCP and provided promising treatments for the clinical management of MCP and SSCP.