The association between alcohol consumption and osteoarthritis: a meta-analysis and meta-regression of observational studies

There is conflicting evidence for the association between alcohol consumption and common joint conditions such as Osteoarthritis (OA), which affects millions of people. We sought to determine the true association between alcohol intake and OA. We conducted a PRISMA systematic review and meta-analysis of observational studies that reported associations between alcohol consumption and OA. Pooled estimates of association were represented through odds ratios (ORs). Publication bias was assessed with Funnel and Galbraith plots, and risk of bias was assessed with the Newcastle Ottawa Scale. We included 29 studies and 25,192 subjects with OA and reported an OR between any alcohol consumption and OA of 0.79 (0.68-0.93), suggesting a protective effect. OR of weekly or more frequent use was 0.79 (0.65-0.97). When grouped by covariates, alcohol consumption was negatively associated with radiographic (0.83, 0.70-0.98), hand (0.80, 0.66-0.95) and knee OA (0.85, 0.72-0.99), North American ethnicity and female gender. Subgroup analysis of unadjusted data resulted in an OR of 0.70 (0.55-0.89) but this disappeared upon analysis of studies with data adjusted for any covariate (0.93, 0.78-1.10). Whilst our pooled analysis suggest that weekly or more frequent alcohol consumption was negatively associated with OA, this was not observed when adjusted for confounding factors. Reasons for this include selection bias and lack of longitudinal exposure and adjustment for confounding variables. Therefore, this meta-analysis provides evidence to dispel notions that alcohol use may be protective against OA.

Automated summary

The meta-analysis conducted on the association between alcohol consumption and OA found a negative correlation, especially with weekly or more frequent alcohol intake. However, this association was less significant after adjusting for confounding factors, possibly due to selection bias and lack of appropriate adjustment for confounding variables.