BACKGROUND: Preclinical evidence suggests a potential protective effect of statins on the risk of hand osteoarthritis (HOA), presumably via antiinflammatory mechanisms. Since evidence from large observational studies remains scarce, we aimed to investigate the association between new statin use and incident HOA.
METHODS: We performed a propensity scorematched cohort study using data from the UKbased Clinical Practice Research Datalink. Statin users had ≥1 statin prescription between 1996 and 2015 and were matched 1:1 on their propensity score to non-users within 10 sequential 2-year cohort entry blocks. Patients were aged 45-84 years and had a ≥3 years statin-free active history prior to cohort entry. After a 180-day run-in period (allowing statin users to reach maintenance dose and excluding all patients with a followup ≤180 days), patients were followed in an “as treated” approach until a recorded diagnosis of HOA or until censoring (change in exposure status, maximum follow-up 5.5 years). We applied Cox proportional hazard regression to calculated hazard ratios (HR) with 95% confidence intervals (CI) overall and in subgroups of age, sex, daily statin dose, and treatment duration.
RESULTS: Among 280522 statin users and the same number of non-users, we observed an overall HR of 0.98 (95% CI 0.91-1.04) which remained unchanged in subgroups of age and sex. HRs slightly decreased with increasing treatment duration (HR 0.90, 95% CI 0.810.99, 4-5.5 years of follow-up) and with increasing daily statin dose (HR 0.65, 95% CI 0.44-0.97, >40 mg simvastatin equivalents per day).
CONCLUSIONS: The results of this large observational cohort study do not suggest an overall protective effect of statins on the risk of HOA. However, a slight protective effect in association with longterm use or with use of high daily statin doses is possible, and will have to be followed up in future research.
