Predictive value of adipose to muscle area ratio based on MRI at knee joint for postoperative functional outcomes in elderly osteoarthritis patients following total knee arthroplasty

Abstract

Background: The current research used a new index—adipose to muscle area ratio (AMR)—to measure fatness compared with body mass index (BMI) in elderly osteoarthritis (OA) patients following total knee arthroplasty. Our study aimed to test the relationship between the two indexes (AMR and BMI) and to examine whether AMR was a predictive factor of patient-reported outcome measures (PROMS) for elderly OA patients following total knee arthroplasty (TKA). Methods: The retrospective data of 78 OA patients (older than 60 years) following TKA was included in our study. Clinical features of patients included age, BMI, sex, AMR, side of the implant, time of follow-up, complications, the Knee Society Score (KSS score), and the Hospital for Special Surgery knee score (HSS score). The area of adipose tissue and muscle tissue was measured on the cross section (supra-patella, midline of the patella, joint line of the knee) of the knee magnetic resonance imaging (MRI). AMR was calculated as the average of adipose to muscle area ratio at the three levels. The Pearson correlation analysis, simple linear regression, and multiple linear regression were used to study the relationship between BMI, AMR, and PROMS (KSS total-post score and HSS-post score) in the study. Results: Of all patients, the mean (± standard deviations (SD)) of age was 67.78 ± 4.91 years. For BMI and AMR, the mean (± SD) were 26.90 ± 2.11 and 2.36 ± 0.69, respectively. In Pearson correlation analysis, BMI had a good correlation with AMR (r = 0.56, p = 0.000), and AMR (r = − 0.37, p = 0.001, HSS-post score; r = − 0.43, p = 0.000, KSS total-post score) had better correlations with PROMS postoperatively compared with BMI (r = − 0.27, p = 0.019, HSS-post score; r = − 0.33, p = 0.003, KSS total-post score). In multivariate linear regression analysis, AMR was negatively correlated with KSS total-post score as well as HSS-post score, while BMI was not. As for patients with complications, AMR values were between the 3rd quartile and 4th quartile of the AMR value in the entire study cohort. Conclusions: In this study, the new obesity evaluation indicator—AMR, which was well related with BMI, was found to be a predictor of PROMS (KSS total-post score and HSS-post score) in elderly OA patients following TKA.