Purpose: Obesity and overweight are recognized risk factors for osteoarthritis (OA). There is a strong and consistently found relationship between high body mass index (BMI) and knee OA, but there is ongoing uncertainty as to whether higher BMI is a risk factor for hand OA. As with most OA research, the studies on hand OA consider this joint in isolation (i.e. without consideration of other joints). Given that hand OA often occurs in conjunction with OA of other joints, a failure to consider other joints could lead to misleading results. The Canadian Longitudinal Study on Aging (CLSA) asked separately about the presence of knee, hip, and hand OA, providing the opportunity to investigate whether there is an independent association between hand OA and BMI, accounting for the association of BMI with other concurrent joints, particularly the knee. Methods: The CLSA began in 2013 and is following a representative sample of respondents aged 45-85 years at baseline to gain insight into the development of disease and disability through the aging process. Respondents in the baseline assessment were asked separately about OA in the knee, in the hip, and in the hand, with the lead-in question, “Has a doctor ever told you that you have⋯” Participant height and weight was measured and BMI (kg/m2) calculated, and sociodemographic information was collected. Participants reporting arthritis other than OA or without information on OA or BMI were excluded from the present study. Descriptive analyses compared BMI distributions between those without OA, and those with OA of the knee, hip and hand, individually and with each possible combination of OA at these joint sites. Visual inspection was used to determine whether the shape of the BMI distributions differed for combinations of hand OA with knee and hip OA. The relationship between BMI and joint site was quantified by linear regression, adjusted for age and sex. The relationship between obesity and joint site was quantified by logistic regression, adjusted for age and sex. Results: The analytic sample included 25,459 individuals. Overall, 24.9% of the sample reported OA at any site. Among these, 56.6% reported knee OA, 31.2% hip OA and 46.9% hand OA. Forty-seven percent of individuals reporting hand OA also reported knee and/or hip OA. There was minimal difference in mean BMI between the OA groups (Table, Column A), but the percent obese, especially for the combinations of individual joint sites, showed greater differences across groups (Table, Column B). The distributions of BMI for the hand OA and hand and hip OA groups were similar, but distinct from the distributions that included knee OA, which were likewise similar but had a higher proportion of individuals in the obese I and II/III categories (Figure). Findings for males and females and different age groups were similar. Regression results for OA overall and for OA at individual joint sites (without consideration of other joints), showed a positive and significant association between each joint group and higher BMI and, in particular, likelihood of being obese (Table, Columns C and D). However, when all combinations of joint sites were simultaneously considered, neither hand OA alone nor hand and hip OA together were significantly associated with higher BMI or obesity, whereas any combinations that included knee OA were significantly associated with higher BMI and, in particular, likelihood of being obese (Table, Column C and D). Conclusions: Hand OA assessed without regard to the involvement of other joints appears to be significantly associated with higher BMI and obesity. However, when considering OA in other joints, and in particular knee OA, an independent association between hand OA and BMI/obesity was not found. This suggests that positive associations seen in previous studies are likely due to the presence, but lack of consideration, of concomitant knee OA. Furthermore, the results suggest that a simple comparison of mean BMI values, even when age and sex adjusted, between OA groups may miss important differences that exist in the upper tails of the BMI distribution (i.e. the obese portion of the distribution) where there is greater risk for OA. The findings also raise questions as to whether obesity is a risk factor for OA in general, or only for knee OA.View Large Image Figure ViewerDownload Hi-res image Download (PPT)
Read full abstract