Purpose: To show the differences in lateral and anterior femoral shaft bowing (FSB) between radiographic and computed tomography (CT) images and to assess whether FSB is associated with various predictive factors in women with knee osteoarthritis (OA) using CT images.
Methods: We enrolled 135 Japanese women with medial compartment knee OA. Lateral and anterior FSB were measured by radiography and reconstructed CT. Age, body mass index (BMI), femorotibial angle (FTA), femoral length, and lumbar spine and femoral neck bone mineral density (BMD) were set as predictive factors for progression of knee OA. We compared predictive factors in the lateral FSB group having lateral angulations of >5° with those in the nonbowing group and compared predictive factors in the anterior FSB group having anterior angulations of >11° with those in the nonbowing group. Binomial logistic regression modeling was applied to determine independent predictors of both FSB.
Results: There were significant differences in both FSB between the radiographic and reconstructed CT images (P = 0.005 and P = 0.047, respectively). In binomial logistic regression analyses for lateral FSB on CT, age, BMI, and lumbar spine BMD were significant predictors, with odds ratios of 1.16, 1.22, and 0.03, respectively; for anterior FSB on CT, age was a significant predictor, with an odds ratio of 1.06.
Conclusions: It is preferable to measure both FSB on reconstructed CT when planning reconstructive knee surgeries. Age, BMI, and lumbar spine BMD were predictors of lateral FSB progression, and age was a predictor of anterior FSB progression. Level of evidence Level III.
Keywords: Bowing; Computed tomography; Femur; Knee osteoarthritis; Radiograph.