Causal association between subtypes of osteoarthritis and common comorbidities: A Mendelian randomisation study
Will Thompson,Subhashisa Swain,Sizheng Steven Zhao,Anne Kamps,Carol Coupland,Changfu Kuo,Sita Bierma-Zeinstra,Jos Runhaar,Michael Doherty,Weiya Zhang
DOI: https://doi.org/10.1016/j.ocarto.2023.100414
2023-10-21
Abstract:Objective: To investigate the causal association between Osteoarthritis (OA) and five comorbidities: depression, tiredness, multisite chronic pain, irritable bowel syndrome (IBS) and gout. Design: This study used two-sample Mendelian Randomisation (MR). To select the OA genetic instruments, we used data from the largest recent genome-wide association study (GWAS) of OA (GO Consortium), with a focus on OA of the knee (62,497 cases, 333,557 controls), hip (35,445 cases, 316,943 controls) and hand (20,901 cases, 282,881 controls). Genetic associations for comorbidities were selected from GWAS for depression (246,363 cases, 561,190 controls), tiredness (449,019 participants), multisite chronic pain (387,649 participants), IBS (53,400 cases, 433,201 controls) and gout (6543 cases, 456,390 controls). We performed a bidirectional MR analysis using the inverse variance weighted method, for both joint specific and overall OA. Results: Hip OA had a causal effect on multisite chronic pain (per unit change 0.02, 95% CI 0.01 to 0.04). Multisite chronic pain had a causal effect on knee (odd ratio (OR) 2.74, 95% CI 2.20 to 3.41), hip (OR 2.12, 95% CI 1.54 to 2.92), hand (OR 2.24, 95% CI 1.59 to 3.16) and overall OA (OR 2.44, 95% CI, 2.06 to 2.86). In addition, depression and tiredness had causal effects on knee and hand, but not hip, OA. Conclusions: Apart from Hip OA to multisite chronic pain, other joint OA did not have causal effects on these comorbidities. In contrast, multisite chronic pain had a causal effect on any painful OA.