Are risk factors for patellofemoral and tibiofemoral knee osteoarthritis different

Objective. To detennine whether patellofemoral (PF), tibiofemoral (TF), and combined patterns of knee osteoarthritis (OA) differ in their strengths of associations with any of the known risk factors for knee OA, and especially to evaluate whether body mass index (BMI) correlates with all 3 patterns, or only with tibiofemoral disease, as previously suggested. Methods. We obtained anteroposterior and lateral knee radiographs on 608 participants at the 22nd biennial examination of the Framingham cohort study (1992-3). The presence or absence of OA in the TF and PF compartments of each knee was scored, and subjects were classified on the basis of the pattern of compartmental involvement in their 2 knees. The strength of association of age, sex, BMI, chondrocalcinosis, and knee injury was computed for PF, TF, and combined pattern of knee OA using multiple logistic regression. Results. The mean age and BMI of the sample were 80.7 yrs (SD 5.0) and 25.4. kg -2 (SD 3.7), respectively. PF, TF, and combined patterns of knee OA were present in 5.3, 23.0 and 19.7%, respectively. Elevated BMI was a risk factor for all 3 patterns of disease (adjusted OR for highest vs lowest tertile of BMI 3.7, 1.9, and 7.0 for PF, TF, and combined pattern, respectively). Risk factor profiles were broadly similar for TF and PF OA, with the possible exception of knee injury in men (adjusted OR = 2.0 for PF, 3.7 for TFOA). Risk factors were generally more strongly associated with the combined pattern of OA. Conclusion. Obesity is an important risk factor for PF, TF, and combined patterns of knee OA. The relationships of these patterns with the risk factors investigated here appear similar and are strongest for the combined pattern.