This study aimed at exploring the potential causal effects of leisure sedentary behavior (LSB) on common types of arthritis.

Two-sample Mendelian randomization (MR), including both univariable MR (UVMR) and multivariable MR (MVMR) analysis, was performed to explore the effects of LSB on the risk of several common types of arthritis, including osteoarthritis, rheumatoid arthritis (RA), and psoriatic arthritis (PsA). Genetic variants from genome-wide association studies (GWAS) of LSBs for time spent on television watching, computer use, and driving were obtained from the UK Biobank. Summarized GWAS data of OA [overall, OA of the hip (HOA), and OA of the knee (KOA)], RA [overall, seronegative RA (nRA) and seropositive RA], and PsA was also acquired from the FinnGen Biobank Analysis. Causal Analysis Using Summary Effect Estimates (CAUSE) were further applied to verify the causality.

UVMR results provided evidence for the causal relationship of time spent on watching TV with overall OA [odds ratio (OR) = 1.80, 95% confidence interval (CI) = 1.45–2.23], KOA (OR = 1.86, 95% CI = 1.45–2.39) and HOA (IVW-fixed: OR = 1.65, 95% CI =1.20–2.26). Similar associations were observed in the TV-overall RA and TV-pRA, and TV-PsA, but the CAUSE method results only supported the causal association of time spent TV watching with OA and KOA. Moreover, MVMR results showed indicated an independent causal effect of TV watching on OA (overall, KOA, and HOA).

This study demonstrated the genetic causal association of prolonged TV watching time with overall OA and KOA risks.