Obesity Outcomes in Disease Management: Clinical Outcomes for Osteoarthritis

Osteoarthritis (OA) is the most common joint disorder, affecting over 25 million Americans ((1)), and one of the leading causes of disability among older Americans ((2), (3)). The knee and hip are the most disabling sites of OA, resulting in loss of mobility and lower extremity physical function and leading to over 250,000 total joint replacements each year. Surveys indicate that 10% to 13% of men and women age 65 and over have symptomatic knee OA, and approximately one-third have radiographic findings of knee OA ((4)). The prevalence of symptomatic OA of the hip is ∼3% and radiographic OA of the hip is ∼5% to 10% ((1), (5)). The spine is also a common site of disabling OA, but spine OA has not been well characterized in population surveys. OA is a slowly evolving degenerative disease that causes excessive breakdown and loss of articular cartilage and proliferation and destruction of periarticular bone. Pain during joint use, stiffness, and altered joint biomechanics contribute to loss of joint function and disability ((6)). The pathogenesis of OA involves the interplay of joint-specific, local biomechanical factors and general susceptibility that is related to age, genetics, and systemic and biochemical abnormalities ((2)). Both types of factors probably play a role in the relationship between obesity and OA in weightbearing joints such as the knee and hip. Excess weight increases the biomechanical load on the hip and knee joint during activity. Forces across the hip and knee during walking and stair- climbing, for example, are two to four times normal body weight ((7)), and there is probably a multiplier for the effect of each pound of extra weight on the joint. Obesity may also increase the risk of OA through the relationship of increased adiposity with abnormal levels of hormones and growth factors, high bone density, and other intermediary factors that may play a role in the development of OA ((8)). The weak association between obesity and OA in some nonweightbearing joints, such as the hand and fingers ((9)), is thought to be evidence implicating these pathways. That obesity is strongly associated with an increased risk of OA has been widely studied and confirmed for knee OA. Analyses of the National Health and Nutrition Examination Survey data show that adults in the United States with a body mass index (BMI) ≥30 kg/m2 have over a 4-fold higher prevalence of radiographic knee OA than those with a BMI <30 kg/m2 ((10)). In one study ((11)), the risk of developing both radiographic and symptomatic knee OA was increased in elderly men and women who were obese an average of 37 years earlier in life. The risk of developing disabling knee OA over a 10-year period is strongly and directly correlated with initial BMI ((12)). Obesity has a consistently strong effect on the risk of developing knee OA in prospective studies, with relative risks for a BMI of 30 kg/m2 or more ranging from 2 to 10. The association of overweight with knee OA may be stronger in women than in men ((2), (11)). Spector et al. ((13)) reported that 50% of healthy, middle-aged obese women with unilateral radiographic knee OA developed OA in the contralateral knee within 2 years, a 5-fold higher rate than in non-obese women. Among women ages 20 to 89 years in a health plan population, the incidence of new diagnoses of symptomatic knee OA were increased 9-fold in those with a BMI >30 kg/m2 and 4-fold in those with a BMI between 25 and 30 kg/m2 ((9)). Recently, Gelber et al. ((14)) reported that male medical students who had a BMI ≥25 kg/m2 between the ages of 20 and 29 years had a 3-fold increased risk of developing symptomatic knee OA by the time they reached age 60 years. Studies also find that being overweight is associated with an increased risk of radiographic progression of existing knee OA ((13), (15)) and may worsen joint pain and disability in those with the disease ((16)). Studies of obesity and radiographic hip OA are conflicting, but bilateral radiographic disease and symptomatic hip OA are more consistently associated with obesity ((17)). Bilateral knee OA is also especially strongly related to obesity, and bilateral disease in either the hips or knees is particularly disabling. Certain types of heavy physical activity (e.g., kneeling, squatting, and possibly stair-climbing) increases the risk of knee OA, and obese individuals have an especially high risk from these activities. According to observational studies of risk factors, obesity is the number one preventable cause of knee OA in women and ranks second in men after knee injury ((18), (19)). Weight loss is recommended for overweight persons with knee OA by the American College of Rheumatology treatment guidelines ((20)). Several studies suggest that weight loss may prevent the development or worsening of knee and hip OA. However, these studies have methodological limitations, and there are no randomized controlled trials of weight loss that show a reduction in knee or hip OA in treatment compared with control groups. In an observational study, Felson et al. ((19)) found that older women who had lost ∼5 kg during a 10-year period had a 50% reduction in the risk of new symptomatic knee OA. McGooey et al. ((21)) reported a 4- to 5-fold reduction in the prevalence of knee and hip pain in morbidly obese subjects who lost an average of 45 kg after gastric stapling. Another study of obese women who lost an average of 5 kg during a 6-month treatment with phentermine or placebo found that reductions in clinical symptoms of hip and knee OA were correlated with the amount of weight loss ((22)). Two more recent trials have tested lifestyle modification weight-loss interventions. In one uncontrolled study, overweight women with knee pain who lost an average of 7 kg during a 6-month diet and walking intervention showed improvements in knee pain, lower extremity disability, Vo2max, and 6-minute walk distance ((23)). In a recently published study ((24)) of 24 obese (BMI ≥ 28 kg/m2) older adults with symptomatic knee OA randomized to diet plus exercise or exercise alone, both groups showed significant improvements in knee pain (24% to 44% reductions), disability, and physical performance after 6 months. The diet plus exercise group lost 15 lbs more than the exercise alone group and had greater reductions in knee-related disability than the exercise group, although this did not reach significance. Together, these studies demonstrate that overweight and obese older persons with painful knee OA can lose 10 to 20 lbs over 6 months with a combined diet and exercise program, and that this amount of weight loss is associated with improvements in knee pain and disability. Disease management for obesity should include efforts to reduce the risk of knee and hip OA, manage knee and hip symptoms, and prevent lower extremity disability. Weight loss and avoiding certain physical activities may prevent the development or worsening of the disease and its symptoms, whereas exercise and pain medications can reduce its impact ((25)). There are two types of measurement appropriate for assessing the outcome of these efforts: 1) direct assessment of symptoms and disability and 2) monitoring of clinical event rates. Several disease-specific instruments are available that assess activity-related knee and hip pain and related lower extremity disability ((26), (27)). In addition, some generic health status instruments include assessments that tap lower extremity disability ((28)). Clinical event rates that could be monitored include new diagnoses of knee and hip OA and total knee and hip replacement, the standard treatments for end-stage disease. Of the disease-specific instruments, the Western Ontario McMasters University OA Index (WOMAC) is the most thoroughly validated and widely used. Several studies have documented WOMAC's validity, reliability, and sensitivity to change ((26), (29)). It is recommended for use in knee and hip OA clinical trials by most guidelines ((30)) and is in use in many ongoing trials. WOMAC has three components: pain with activity (5 items), lower extremity disability (17 items), and stiffness (2 items). Of these, we recommend the pain and disability components, which can each be analyzed separately, for obesity outcome assessment. WOMAC can be self-administered in either Likert or visual analog scale scoring formats. The time frame for reporting has several options: the past 48 hours, 7 days, or 30 days ((31)). Validation studies for WOMAC have been performed in patients with clinically diagnosed knee and hip OA. Joint symptom and disability items were derived from open-ended questioning of patients. The final selection of items took into account their prevalence and frequency of occurrence, as well as the importance of these items to patients. The pain and disability scales have demonstrated acceptable internal consistency and test–retest reliability (Table 1). In the absence of a gold standard measure of pain and disability, construct validity has been demonstrated by higher correlations of the scales with indices probing the same dimension compared with those probing different dimensions (Table 1). Sensitivity to change is important for an instrument's ability to assess the effect of an intervention. The WOMAC pain and function scales both show significant improvement in patients undergoing total hip or knee replacement and in patients starting nonsteroidal anti-inflammatory drugs for pain management ((26), (32)). In a trial comparing two nonsteroidal anti-inflammatory drugs, the WOMAC pain scale detected a significant difference in pain improvement between the treatments ((29)). The two WOMAC scales have also shown increased statistical efficiency for detecting treatment effects in these trials compared with standard clinical measures (walk time and range of joint motion) ((33)), an alternative lower extremity pain and function index for knee and hip OA ((34)), and two mobility indexes (arthritis impact measurement scale and health assessment questionnaire) designed for patients with a variety of arthritic conditions ((32)) (Table 2). Of particular relevance for obesity outcomes, a recent trial of weight loss in overweight women with knee pain showed significant changes in both the pain (46% reduction) and disability (38% reduction) scales of WOMAC after an average loss of 7 kg over a 6-month period ((23)). A potential limitation of disease-specific measures is that several instruments could be needed to assess a broad range of outcomes in obese individuals, whereas a single generic instrument might be sensitive to a greater spectrum of obesity impacts. An acceptable generic instrument, however, would need to be sensitive to important obesity outcomes like knee OA. The SF-36 Physical Function subscale is a potential alternative to the disease-specific WOMAC measures. Nine of the 10 items on the SF-36 subscale tap a dimension of lower extremity function likely to be affected by activity-related knee or hip pain. The validity and reliability of the Physical Function subscale have been extensively documented in patient groups with diverse medical conditions, including hypertension, diabetes, congestive heart failure, heart attack, and depression ((35)). Of greater importance for our purposes, three studies have examined the performance of the SF-36 Physical Function subscale in persons undergoing total hip or knee replacement for OA. None of these studies compared the SF-36 to the WOMAC instrument. However, one study found that the SF-36 was more sensitive (a greater standardized response mean) to change than an alternative generic instrument (the Sickness Impact physical function subscale) ((28)), whereas a second study suggested that the responsiveness of the SF-36 was similar to the physical function subscales of the arthritis-specific arthritis impact measurement scale and health assessment questionnaire ((36)). A third study found that the SF-36 and an orthopedic clinical rating system for knee OA had equivalent specificity for detecting change after total-knee replacement ((37)). Clinical events have face validity and a direct link to medical costs. New diagnoses of hip and knee OA can be monitored in defined populations of health plan members using computerized medical care and administrative data bases. Oliveria et al. ((38)) have demonstrated the feasibility of computerized search and diagnostic classification algorithms to screen for new cases of knee, hip, and hand OA. Using incident cases identified in this way, the authors have demonstrated a strong association between obesity and the risk of symptomatic knee, hip, and hand OA ((9)). Whether the rate of incident symptomatic OA would be responsive to weight loss in an overweight group of health plan members is not known. The method may also be costly, because human review of radiology reports and other medical records of potential cases is required. Total joint replacement of the hip and knee also have face validity and may be more accessible and easier to monitor than new diagnoses. Because of their greater risk of knee and hip OA, overweight individuals have an increased risk of undergoing total-hip or -knee replacement. However, the criteria for and rates of total-joint replacement vary by time and place. Low rates may sometimes reflect decreased access to medical attention. The short-term responsiveness of joint replacement to weight loss in an obese population is uncertain. Because obesity is often a contraindication for knee and hip replacement, weight loss could increase likelihood of surgery in those with disease. Obesity is an important risk factor for lower extremity joint symptoms and disability caused by knee and hip OA. As many as one-half of obese older individuals have clinically significant knee or hip symptoms. Validated and practical measures of lower extremity joint symptoms and disability, such as the WOMAC, are likely to be useful in monitoring the effect of weight- and obesity-related disease management on knee and hip OA in defined populations. Generic instruments such as the SF-36 physical function subscale also tap lower extremity functions likely to be impaired by OA and may be useful to identify obese individuals for more intensive follow-up with disease-specific instruments. Further research is needed using existing lower extremity pain and disability measures to identify weight loss treatment thresholds for those with knee and hip OA and to define clinically significant improvement. Monitoring rates of clinical events, such as new OA diagnoses and total joint replacement, may be feasible in some health plan settings, but their responsiveness to obesity management is uncertain.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.