Heart disease in osteoarthritis: hip versus knee and the influence of multiple joint involvement and sex

Abstract

Purpose: Inflammation has been implicated in osteoarthritis (OA) and cardiovascular disease (CVD) pathogenesis, with obesity playing a precursor role. Given known differences in the association between obesity and hip and knee OA, we investigated whether a differential likelihood of reporting CVD existed between these OA cohorts, adjusting for BMI. Further, we investigated whether this likelihood was modified by multiple joint involvement and sex. Methods: Participants (437 hip; 494 knee) with moderate to severe OA completed measures including CVD, diabetes, high blood pressure, symptomatic joints, height and weight, activity level, age, sex, and education. Sequential logistic regression models were utilized to investigate the association between study measures and reporting CVD. Interactions between knee/hip OA and symptomatic joint count and sex were evaluated. Results: Knee: age 35-88 years, 65% female, 4% CVD, 14% diabetes, 49% HBP. Hip: age 30-91 years, 55% female, 5% CVD, 8% diabetes, 41% HBP. Overall mean symptomatic joint count was 5 (±4). Adjusted for demographics and health status, no difference in CVD likelihood was detected between cohorts. However, with interactions considered, an increased likelihood was observed for knees among women (Odds Ratios: knee/hip 19.5; symptomatic joint count 1.4; knee/hip*symptomatic joint count 0.6; all p<0.006). Further, symptomatic joint count was associated with the likelihood of reporting CVD among hips; odds increased by 25% with each unit increase in symptomatic joint count (p<0.001). For an individual with a symptomatic joint count equivalent to the mean among the hip cohort (mean=4), this equates to a nearly 2.5 times greater odds. Conclusions: The variability in association with CVD between hip and knee OA alongside modifying effects of multiple symptomatic joint involvement and sex likely suggests different OA phenotypes, with potentially varying roles for inflammation and associations with metabolic syndrome. As well, this has implications for OA care, suggesting a need to evaluate CVD and associated risk factors in this population.