Weight Loss in Patients With Overweight or Obesity and Knee Osteoarthritis

How much weight loss is needed to improve pain and quality of life for patients with overweight or obesity and knee osteoarthritis (OA)?

A total body weight loss of 5.1%, or 0.24% per week, results in clinically significant improvement in pain and disability in patients with overweight or obesity and knee OA. (Strength of Recommendation [SOR]: A, meta-analysis of four randomized controlled trials [RCTs].) Although weight loss improves pain caused by knee OA, there is no known minimum level of weight loss at which pain reduction is expected to occur. Greater weight loss leads to greater pain reduction.

A 2007 meta-analysis evaluated the effect of weight change on symptoms associated with knee OA in four RCTs that included 454 patients.¹ The mean body mass index (BMI) levels in the trials were 36, 36, 34, and 29 kg/m², respectively. Inclusion criteria were diagnosis of knee OA, RCT design, specification of comparative treatment, and relevant outcome data.

Two of the RCTs used the self-administered Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) to measure patient pain and disability. The WOMAC subscale for pain includes five items (range = 0–20 total score). The WOMAC subscale for disability asks patients to rate the degree of difficulty in performing activities on a scale from 0 (none) to 4 (extreme). The third RCT used the Fitness and Arthritis in Seniors Trial (FAST) Functional Performance Inventory to measure disability; the fourth study used an unnamed survey to assess pain. Two RCTs used the Lequesne index of severity for osteoarthritis of the hip (LISOH) score to measure severity.

Treatment effect sizes were determined for pain (417 patients), disability (417 patients), and LISOH score (117 patients). There were statistically significant but small reductions in pain (effect size = 0.2; 95% CI, 0–0.39; P = .05) and self-reported disability (effect size = 0.23; 95% CI, 0.04–0.42; P = .02) following a mean weight loss of 6.1 kg (13.4 lb). No statistically significant changes were noted in the LISOH score (effect size = 0.58; 95% CI, −0.4–1.56; P = .25) after a mean weight loss of 4.7 kg (10.4 lb). Meta-regression analysis assessed for a dose-dependent response. A response was shown for self-reported disability but not for pain. A minimum of 5.1% overall weight loss or 0.24% weight loss per week each predicted a significant decrease in level of disability.

A 2018 secondary analysis of an RCT of older adults with overweight or obesity and knee OA correlated long-term weight loss with fewer OA symptoms.² The analysis included 240 adults who were 55 years and older (mean age = 65.8 years; standard deviation = 6.0) with a baseline mean BMI of 33.4 kg/m² (standard deviation = 3.8).

Patients were assessed using the WOMAC pain and disability subscales. They were sorted into four groups based on body weight lost: less than 5%, between 5% and 10%, between 10% and 20%, and 20% or greater.

At 18 months, the group who lost 20% or more of their body weight had significantly less pain than the groups who lost less than 5% (mean difference = 1.68; 95% CI, 0.57–2.78; P ≤ .0125) and between 5% and 10% (mean difference = 1.81; 95% CI, 0.69–2.93; P ≤ .0125). The group who lost 20% or more of their body weight had 25% less pain than the group who lost between 10% and 20% of their body weight, although this was not statistically significant. Further, the group who lost between 10% and 20% of their body weight had significantly less pain than the group who lost between 5% and 10% of their body weight (mean difference = 0.89; 95% CI, 0.01–1.18; P < .05). Although this RCT did not investigate a predictable, dose-dependent relationship, it found that pain improved significantly with a greater amount of weight loss.