Osteoarthritis (OA) occurs in approximately 55% of individuals aged 70 years or older. Symptoms of OA include pain, swelling and stiffness of joints, leading to impaired mobility. To date it is not possible to heal OA, except by means of surgery. The cost of OA treatment such as pain relief or joint replacement is substantial and increases with the increasing average age of our society.
Statins are drugs, which lower lipid levels in the blood and which are prescribed to prevent heart attacks and stroke. Approximately 11% of the population in the United Kingdom (UK) are prescribed statins. Experiments in cells and animals have demonstrated that statins reduce inflammation and alter the immune response in OA-affected areas of the body. Results from studies in humans are contradictory. Finger joints are non weight bearing; therefore, they present an ideal joint in which to study the effect of statins on development of OA since pressure on joints by body weight only, as occurs in the hip or knee, does not apply. To date, no study has been published focussing on this association.
We plan to perform a propensity score (PS)-matched sequential cohort study to assess the relative risk of incident hand OA associated with new statin use. Statin initiators will be matched to non-initiators on their PS within 2 year enrolment blocks (to avoid time trend bias) after completion of a 180 day run-in period (to allow statin users to reach maintenance dose). After concatenating all blocks into one dataset, we will follow all patients from the day after they complete the run-in period to the first of the following: a maximum of 5.5 years of follow up, a record for hand OA, or end of the patient record. In the statin exposed cohort person time will be accumulated as exposed or non-exposed time according to prescriptions received. We will perform subgroup analyses by sex, age groups, subgroups of daily statin dose (in simvastatin equivalents), the specific statin agent used, and by indication for statin use. Cox proportional hazard analyses will be performed to calculate hazard ratios (HR) with 95% confidence intervals (CI). This will be the first observational study to specifically assess the association of statin initiation and incident hand OA, while controlling for confounding by indication by means of PS matching.
Health Outcomes to be Measured: