Understanding the mechanisms by which weight change affects progression of knee osteoarthritis in obese and overweight individuals: An analysis of the Osteoarthritis Initiative Dataset

项目摘要 骨关节炎（OA）是一种退行性关节疾病，在美国影响超过2700万人， 是导致老年人残疾的最常见原因。受症状性OA影响的人数 由于人口老龄化和肥胖人数的增加， 一个确定的OA风险因素。肥胖已经成为美国的一种“流行病”，预测表明， 到2030年，美国将有86.3%的成年人超重或肥胖。肥胖是一个可改变的风险因素， OA伴体重减轻为肥胖和超重患者的疾病预防提供了潜在的非侵入性治疗 个体研究表明，体重减轻减缓OA进展，体重增加加剧 进展然而，这些研究并没有具体评估哪些因素或途径将负责 改善或恶化的结果，如相关的炎症，局部身体成分和肌肉减少症。 在目前的建议中，我们将全面审查与机械相关的机制， 导致膝关节退行性变化的负荷（体重减轻和增加），包括 软骨损失，即炎症，局部身体组成（关节周围脂肪）， 以及肌肉形态和力量。确定哪些机制最有利于减缓 体重减轻期间的OA进展将导致靶向治疗，以实现有效和优化的治疗 在疾病的早期阶段，可以预防疾病的进展。通过通径分析， 调解分析和机器学习，我们将确定潜在的预防措施（如肌肉 加强和抗炎措施），这可能会放大减少机械的积极影响， 正在加载OA。因此，我们的项目的临床影响是开发一个特定于对象的预测模型， 临床医生可以个性化定制治疗计划，减缓关节崩溃，并降低侵入性关节炎的可能性。 和昂贵的手术，如全膝关节置换术。 提出了三个具体目标： 具体目标1：我们将描述体重变化与膝关节变化之间的相关性 炎症、局部身体组成、肌肉横截面积（CSA）、脂肪浸润和肌肉强度， 并研究这些参数与膝关节退行性病变进展之间的相关性。 具体目标2：使用路径分析，我们将探索体重变化影响OA的机制 进展，并量化这些因素介导这种关系的程度。 具体目标3：最后，我们将使用机器学习开发一个预测模型，以确定 人口统计学、临床和MRI特征（包括局部身体成分、炎症、肌肉变化） 可以预测OA在8年内的进展。

Project Summary Osteoarthritis (OA) is a degenerative joint disease which affects more than 27 million people in the US and is the single most common cause of disability in older adults. The number of people affected with symptomatic OA will increase due to the aging of the population and the growing number of people with obesity, which represents an established risk factor for OA. Obesity has become a US “epidemic,” and projections have suggested that 86.3% of adults in the United States will be overweight or obese by 2030. Obesity is a modifiable risk factor for OA with weight loss offering a potential non-invasive therapy for disease prevention in obese and overweight individuals. Research has shown that weight loss slows OA progression and weight gain exacerbates progression. However, these studies did not specifically assess factors or pathways which would be responsible for improved or worse outcomes, such as associated inflammation, local body composition and sarcopenia. In the current proposal, we will comprehensibly examine the mechanisms associated with mechanical loading (weight loss and gain) that are responsible for driving knee joint degenerative changes including cartilage loss, namely concurrent changes in inflammation, local body composition (periarticular fat), and muscle morphology and strength. Identifying which mechanism(s) are most beneficial for slowing OA progression during weight loss will lead to targeted therapies for effective and optimized treatment of OA at early stages of disease during which progression may be prevented. Through pathway analysis, mediation analysis, and machine learning, we will identify potential preventive measures (such as muscle strengthening and anti-inflammatory measures) that could amplify the positive effects of reduced mechanical loading on OA. Thus, the clinical impact of our project is development of a subject-specific prediction model for clinicians to individually-tailor treatment plans that slow joint breakdown, and decrease probability for invasive and costly surgeries such as total knee arthroplasty. Three specific aims are proposed: Specific Aim 1: We will characterize the associations between changes in weight with changes in knee joint inflammation, local body composition, muscle cross-sectional area (CSA), fat infiltration and muscle strength, and investigate the associations between these parameters and progression of knee degenerative changes. Specific Aim 2: Using a path analysis we will explore the mechanisms by which weight change impacts OA progression, and quantify the degree to which these factors mediate this relationship. Specific Aim 3: Finally, we will develop a prediction model using machine learning to determine which demographic, clinical, and MRI features (including changes in local body composition, inflammation, muscle) can predict progression of OA over 8 years.