PROJECT SUMMARY/ABSTRACT Osteoarthritis affected at least 27 million US adults in 2008, and that number is expected to rise to about 50 million by 2020 because of growing obesity and the aging population. A leading cause of disability in the world, osteoarthritis has contributed to about 30% increase in Years Lived with Disability (YLDs) according to Global Burden of Disease 2016 estimates. Yet, no definitive treatment exists for osteoarthritis. The number of total knee replacements, the ultimate end-point of osteoarthritis when medical and rehabilitative efforts fail, is expected to rise to 3.5 million annually by 2030. Medical interventions so far have focused on pain-relieving interventions, with nonsteroidal anti-inflammatory drugs (NSAIDs) and non-NSAID analgesics being the primary non-surgical treatment choices. Despite the widespread use of these medications, their potential adverse effects on the disease in the long-term is not well characterized. Specifically, limited data exist on how reducing chronic pain in patients suffering from osteoarthritis would affect the risk of disease progression or the risk of knee replacement. The long-term goal of this project is to facilitate decision making for clinicians and patients of a more optimal treatment strategy for chronic pain relief interventions in osteoarthritis. The central hypothesis of this proposal is that non-NSAID analgesics users, compared to NSAIDs users, are at an increased risk of long- term osteoarthritis adverse outcomes, including disease progression (determined radiographically) and knee replacement. While both types of pain relievers may, with effective pain relief, produce higher levels of injurious activity, only NSAIDs may mitigate this damage by reducing intra-articular inflammation. Non-NSAID analgesics, unlike NSAIDs, lack anti-inflammatory properties that could reduce intra-articular inflammation sufficiently enough to minimize potential damage. We propose to use the Osteoarthritis Initiative (OAI) cohort, a longitudinal study of persons with or at risk of knee osteoarthritis studied annually over 10 years, with extensive clinical, imaging, and medication use data. Our aims are: to characterize the time-varying effects on osteoarthritis outcomes of 1) non-NSAID pure analgesics, 2) prescription NSAIDs, and 3) to compare the effects of non-NSAID analgesics with those from NSAIDs. The contribution of this work is significant because these two types of medications are among the most widely used for osteoarthritis. Our novel analytic approach is a departure from the traditional approaches of a restrictive definition of exposure that did not allow medication use to be intermittent (i.e. time-varying) or did not allow treatment switching. Unlike time-invariant approaches that were indifferent to changes in pain, our time-varying approach will allow us to account for time-dependent effect of pain, the main indicator for intervention or treatment switching. Our work will provide new evidence about effects of widely used medication of disease outcomes and will inform treatment choices.

PROJECT NARRATIVE No definitive treatment exists for osteoarthritis, a leading cause of disability worldwide which is expected to affect at least 50 million Americans by 2020. Limited knowledge exists on potential long-term adverse effects of pain- relieving interventions in knee osteoarthritis, which includes disease worsening that ultimately may result in total knee replacement. This proposal utilizes modern epidemiologic research methods to bring new insights into potential adverse effects of commonly used pain-relieving medications on outcomes of patients with osteoarthritis, which could inform treatment options for chronic pain relief in millions of Americans affected by osteoarthritis. !