J. Runhaar1 M. van Middelkoop1 M. Reijman1 D. Vroegindeweij2 E. H. Oei1 S. M. Bierma-Zeinstra1;
Purpose. The present study evaluates the effects of malalignment and its interaction with BMI on the onset of clinical and radiographic knee osteoarthritis (OA) over a 2.5 year follow-up period in a high risk group of middle-aged overweight and obese women.
1Erasmus MC Rotterdam NETHERLANDS 2Maasstad Hosp. Rotterdam NETHERLANDS.
Methods. Data of the PROOF study (ISRCTN 42823086) were used. In total 407 women between 50 and 60 years with a BMI ≥ 27 kg/m2 and without clinical and radiological knee OA at baseline were included in this study. Both knees of all 351 women (86%) with baseline knee alignment data and the primary outcome available were selected. At baseline body weight and height were measured and standardized semi-flexed PA radiographs of both knees were taken according to the MTP protocol. All subjects filled in a questionnaire with questions on knee complaints and number of days with knee pain. All measurements were repeated after 2.5 years of follow-up. Minimal joint space width (medial and lateral) K&L grade and anatomical knee alignment angle were digitally assessed on all radiographs. Varus alignment was defined as an anatomical angle 184°. The predefined primary outcome measure was the incidence of knee OA defined as onset of K&L ≥ 2 or the onset of clinical knee OA (according to the ACR criteria) or joint space narrowing (JSN) ≥ 1.0 mm in the medial or lateral compartment. Using Generalized Estimated Equations which takes into account the correlation between knees within subjects effects of varus and valgus alignment on the primary outcome and on the items separately were evaluated with neutrally aligned knees serving as reference. The interaction between malalignment and baseline BMI was also studied by adding BMI and the interaction term with malalignment to the analysis. If a significant interaction was found overweight (BMI < 30 kg/m2) and obese subjects (BMI ≥ 30 kg/m2) were analysed separately. All analyses were adjusted for K&L grade at baseline and the randomized groups of the interventions of the PROOF study.
Results. Varus alignment was found in 40% and valgus alignment in 13% of all knees. Baseline characteristics are presented in Table 1. Overall only varus alignment had a significant effect on the incidence of K&L ≥ 2 (9% vs. 3% in neutral knees. OR 2.8 95% CI 1.3 - 5.9). For the primary outcome and for medial JSN a significant interaction between malalignment and baseline BMI was found (p < 0.01). In obese subjects varus alignment had a significant effect on the primary outcome (22% vs. 13% in neutral knees. OR 1.8 95% CI 1.1 - 3.1) and on medial JSN (9% vs. 4% in neutral knees. OR 2.6 95% CI 1.1 - 6.3). These associations were not found in non-obese subjects.
Conclusions. In women at high risk for developing knee OA varus aligned knees had a significant increased risk for the development of radiographic knee OA. Within obese women varus aligned knees also had a significantly increased risk for incidence of knee OA according to the primary outcome and for joint space narrowing in the medial compartment. Since varus alignment is a potentially modifiable factor results from the present study suggest that varus alignment might be a target for the prevention of knee OA in middle-aged overweight and obese women.