Methodologic Remedies for the Risk Factor Paradox in Osteoarthritis Progression

Massachusetts General Hospital, Boston MA

Investigators

Linked publications & trials

Abstract

? DESCRIPTION (provided by applicant): This proposal is in response to NIH Research Grant Opportunity PA-13-237 using the Osteoarthritis Initiative (OAI) database. Knee osteoarthritis (OA) constitutes a tremendous disease burden in the US, as it causes pain and decreased mobility, and its progression leads to a loss of independence and functional capacity, disability, and total knee replacement. Thus, OA progression represents a compelling target for secondary prevention; however, unlike the risk factors for incident knee OA, the results of studies on risk factors for the progression of knee OA have been paradoxical. For example, a number of risk factors for incident knee OA have been consistently identified, including female sex, obesity, high bone mineral density, knee injury, and repetitive occupational stress on joints. In contrast, none of these factors have been consistently associated with the risk of OA progression. Further, studies have failed to find a consistent association even between obesity or aging (two well-established risk factors for incident knee OA) and the risk of knee OA progression. Although biological explanations for these counterintuitive results may exist, an enticing methodologic explanation is a type of selection bias called index event bias, which can affect research on disease sequelae (e.g., progression) when there are multiple risk factors for sequelae that are also risk factors for having the disease in the first place (e.g., the effect of MI on the risk of severe OA among mild to moderate OA patients [i.e., those with index events]). To date, little research has methodologically investigated the paradoxical phenomena of risk factors for OA progression, leaving a crucial gap in knowledge on this important topic. Unless an appropriate design and analytic method are used to measure the germane impact of purported risk factors in OA (i.e., the most common form of arthritis), ample research funds, time, and effort can be depleted without providing useful evidence for clinical recommendations and biological insight. To validly characterize and overcome the methodological challenges associated with assessing the risk conferred by purported risk factors for OA progression, we will investigate previously reported paradoxical findings using the OAI cohort for the following specific aims: (1) to determine and quantify the index event bias and to clarify the core mechanisms underlying the risk factor paradoxes; (2) to provide methodological remedies, including sensitivity analyses, assessment of the impact of a change in exposure status, and causal modeling approaches for the unbiased impact of purported risk factors on knee OA progression. Furthermore, we will demonstrate inherent limitations of certain exposures in this key research context. By leveraging our expertise in the relevant advanced methodology and the clinical topic, as well as the strengths and our hands-on knowledge of the OAI database, the proposed project will fill crucial gaps in our understanding of the germane impact of risk factors for knee OA progression, and provide key methodologic advances needed for evidence-based medicine in rheumatology and many other disciplines.