Risk Factors for End-Stage Glenohumeral Osteoarthritis

Project Summary/Abstract Glenohumeral osteoarthritis (OA) can be as debilitating as hip or knee OA, and in the US, shoulder arthroplasty rates for OA are rising more rapidly than hip and knee arthroplasty. Yet few studies have investigated glenohumeral OA risk factors, resulting in insufficient insights to guide evidence-based prevention

and treatment strategies. For OA at other joints, many risk factors are known, including sex, occupation, and genetics, but effects can vary considerably by joint type. Therefore effects on glenohumeral OA cannot be inferred without studies specific to the shoulder. While occupational risk factors for shoulder pain have been

identified, there has been limited investigation of factors that specifically effect glenohumeral OA. Moreover, the genetics of glenohumeral OA is an unexplored area. Shoulder arthroplasty is also strongly correlated with hip and knee arthroplasty, and is known to be more common in patients with a history of shoulder instability.

But estimates of the absolute risk of shoulder arthroplasty and how this risk differs across patient populations has not been well-defined. This information could improve patient counseling, and could indicate opportunities for early evaluation of OA in high-risk patient groups. Our central hypothesis is that glenohumeral OA results

from a mix of modifiable (occupation) and non-modifiable factors (age, sex, genetics), and that these factors can be used alongside surgical history to produce informative risk predictions for glenohumeral OA. As no sufficiently large studies systematically capture radiographic measures of glenohumeral OA, our study will

focus on shoulder arthroplasty as a measure of end-stage glenohumeral OA. Our objectives are to determine associations of occupational exposures with OA-related shoulder arthroplasty risk (Aim 1), identify genetic variants associated with risk of OA-related shoulder arthroplasty (Aim 2), and develop prediction models for 5-

year and 10-year risk of OA-related shoulder arthroplasty (Aim 3). Across all aims we will use the UK Biobank (UKB), a population-based, 500,000-person cohort with more than a decade of prospectively linked hospital records capturing diagnoses and procedures, including >1200 OA-related shoulder arthroplasty cases. For Aim

1, we will link a novel job exposure matrix to UKB job titles to measure shoulder-specific occupational tasks. For Aim 2, we will combine whole-genome sequencing data available in UKB and the US-based All of US cohort to conduct an innovative sequencing-based genome-wide association study for OA-related shoulder

arthroplasty. For Aim 3, we will leverage both the UKB and the US Healthcare Cost and Utilization Project to develop prediction models for us in both the general population, and specific high-risk patient populations. An understanding of glenohumeral OA risk factors is critically needed to allow informed design of prevention

approaches. As shoulder arthroplasty is strongly correlated with OA at other joints, better understanding of genetic influences of glenohumeral OA can provide deeper insights regarding OA etiology. Findings from this study can enable better prediction of glenohumeral OA, improving patient counseling in the future.