Patient characteristics associated with benefits from weight loss to prevent progression of knee osteoarthritis: data from the osteoarthritis initiative

Abstract

Purpose: The effect of weight loss on knee osteoarthritis (OA) development and severity has been subject to a number of previous studies that found significantly less progression of cartilage degeneration in subjects with weight loss compared to those with stable weight or weight gain. However, little is known about the extent to which subgroups with certain demographic characteristics and imaging findings benefit from weight loss, and if some groups could benefit more from weight loss compared to others. Since OA and overweight are both frequent disorders with a rising prevalence, identifying those that would particularly benefit from weight loss would be highly useful in clinical practice. Our aim was therefore to assess how different baseline subject characteristics influence the protective effect of weight loss on morphological knee joint degeneration over 4 years. Methods: In this study, 592 subjects (age 62.2 ± 8.9 years, sex: 39.5% females, BMI: 30.1 ± 3.6 kg/m2) from the Osteoarthritis Initiative (OAI) were included who were overweight or obese (BMI ≥ 25 kg/m2) and showed either weight loss or stable weight over 4 years. Weight loss was defined as ≥ 5% decrease in BMI over 4 years (n = 367), while stable weight was defined as a weight change of < 3% in BMI over 4 years (n = 225). Morphological knee abnormalities were assessed on 3T MRIs at baseline and 4 year follow-up in the right knee using the Whole-Organ Magnetic Resonance Imaging Score (WORMS). In addition Kellgren-Lawrence (KL) scores were assessed from weight bearing fixed flexion radiographs. Cartilage composition was obtained by measuring T2 relaxation time using 3T MRI and available in a subset of 370 knees. To identify the characteristics of subjects who would benefit most from weight loss, in a multivariable linear regression model we tested for the interaction of baseline demographics (age, sex, BMI), baseline radiographic findings (KL scores), and cartilage T2 with weight loss in predicting WORMS sum score changes over 4 years. Results: Our data showed that progression of WORMS sum scores over 4 years was significantly associated with weight loss (effect size -1.09; 95% confidence interval (CI) -2.00, - 0.18; p = 0.019) and KL scores at baseline (effect size 0.79; 95% CI 0.37, 1.2; p < 0.001). However, no significant interaction with weight loss was found for any of the subject demographics (age, sex, BMI), baseline radiographic imaging findings (KL score), and cartilage T2 at baseline (p>0.05 for all interaction terms). Overall, we could not identify a subcohort with specific baseline subject characteristics that were associated with significantly more benefits from weight loss as measured by progression of WORMS scores. Conclusions: Our findings highlight the importance of weight loss for the overweight and obese population to prevent progression of morphological knee joint degeneration as measured by WORMS. Moreover, our results suggest that weight loss is beneficial for all individuals suffering from obesity or overweight. We could not identify specific baseline characteristics that significantly influenced the protective effect of weight loss on morphologic knee joint degeneration over 4 years.