Knee symptomatic osteoarthritis, walking disability and all-cause mortality in the population-based longitudinal Wuchuan Osteoarthritis Study

Abstract

Purpose: Several studies have reported that knee symptomatic osteoarthritis (SxOA) increases the risk of all-cause mortality. While the underlying biological mechanisms linking knee SxOA to high mortality are not fully understood, it has been postulated that walking disability from knee SxOA may partially account for such an association. To date no study has quantitatively evaluated to what extent an increased mortality among individuals with knee SxOA is mediated through walking disability. Methods: Between 8/2005-10/2005, 1025 residents aged ≥ 50 years were recruited in randomly selected rural communities in Wuchuan, China. Participants completed a home interview and had weight-bearing posteroanterior semiflexed view of radiographs at tibiofemoral (TF) joints and skyline view of radiographs at patellofemoral (PF) joints. During the home-interview participants were asked if they could perform several daily activities, including current ability of walking for 1 kilometer with no-difficulty, some difficulty, and very difficult/unable to do. We defined a knee as having whole ROA if either K/L score at TF joint ≥ 2 or presence of PFOA based on OARSI criteria. Knee SxOA was recorded if both pain (i.e., knee pain occurred on most days in past month) and whole ROA were present at the same knee. Subjects were followed until November 31, 2013. First, we examined the relation of knee SxOA to the risk of all-cause mortality using the Cox-proportional hazards model adjusting for age, sex, body mass index (BMI), education levels, income levels, history of occupational physical activity, and comorbidities. We then partitioned the total effect of knee SxOA on mortality into indirect, i.e., an effect of SxOA on mortality via current ability of walking for 1 kilometer, and direct effect, i.e., an effect of SxOA not through current ability of walking for 1 kilometer using a marginal structural model. Results: Over 8 years of follow-up period, 99 subjects died. As shown in Table 1, the mortality rates were higher among subjects with knee SxOA (32.6/1000 person-years) than those without SxOA (10.9/1000 person-years), with an adjusted hazard ratio of 1.98 (95% confidence interval: (CI): 1.09-3.62). The indirect effect of knee SxOA on mortality was 1.92 (95% CI: 0.86-4.26) whereas the direct effect was 1.08 (95% CI: 0.55-2.12), suggesting that effect of knee SxOA on all-cause mortality was almost entirely mediated through its effect on current ability of walking in 1 kilometers. Conclusions: In this population-based longitudinal study, we found that an increase in all-cause mortality among subjects with knee SxOA was mainly mediated through its effect on walking disability.Tabled 1Table 1. Total, Direct and Indirect Effect of Knee SxOA on All-Cause Mortality Mediated Through its Effect on Walking DisabilitySxOAStatusSubjectsTotal effectHR (95% CI)*Indirect effectHR (95% CI)*Direct effectHR (95% CI)*No631.01.01.0Yes9601.98 (1.09, 3.62)1.92 (0.86, 4.26)1.08 (0.55, 2.12)*Adjusting for age, sex, body mass index (BMI), education levels, income levels, levels of occupational physical activity, and comorbidities. Open table in a new tab *Adjusting for age, sex, body mass index (BMI), education levels, income levels, levels of occupational physical activity, and comorbidities.