Objective:The Osteoarthritis Initiative (OAI) is a multicentre study targeted at identifying biomarkers for evaluating the progression and risk factors of symptomatic knee OA. Here cartilage loss using 3 Tesla (3 T) MRI is analysed over 1 year in a subset of the OAI, together with its association with various risk factors.Methods:An age- and gender-stratified subsample of the OAI progression subcohort (79 women and 77 men, mean (SD) age 60.9 (9.9) years, body mass index (BMI) 30.3 (4.7)) with both frequent symptoms and radiographic OA in at least one knee was studied. Coronal FLASHwe (fast low angle shot with water excitation) MRIs of the right knee were acquired at 3 T. Seven readers segmented tibial and femoral cartilages blinded to order of acquisition. Segmentations were quality controlled by one expert.Results:The reduction in mean cartilage thickness (ThC) was greater (p = 0.004) in the medial than in the lateral compartment, greater (p = 0.001) in the medial femur (−1.9%) than in the medial tibia (−0.5%) and greater (p = 0.011) in the lateral tibia (−0.7%) than in the lateral femur (0.1%). Multifactorial analysis of variance did not reveal significant differences in the rate of change in ThC by sex, BMI, symptoms and radiographic knee OA status. Knees with Kellgren–Lawrence grade 2 or 3 and with a BMI >30 tended to display greater changes.Conclusions:In this sample of the OAI progression subcohort, the greatest, but overall very modest, rate of cartilage loss was observed in the weight-bearing medial femoral condyle. Knees with radiographic OA in obese participants showed trends towards higher rates of change than those of other participants, but these trends did not reach statistical significance.
Application of machine learning for the prediction of cartilage loss: data from the osteoarthritis initiative & multicenter osteoarthritis study