The purpose of this study was to explore, in patients before and 6-months following a TKA, whether knee extension ROM had an indirect effect on SF-36 physical function through the putative pathway of knee extensor strength. Our results suggest that changes (pre to post TKA) in knee extensor strength mediated the association between changes in knee extension ROM and self-report physical function. To our knowledge, these findings have not been previously described in patients with TKA.
Our results, without considering mediating effects, are consistent with cross-sectional studies
[7, 9, 12] of a positive association between knee extension ROM and physical function in TKA (path c in Figure
1). Furthermore, case series
[28, 29] and small trials
 are available demonstrating that in patients before and following a TKA, rehabilitation interventions – for example, such as manual therapy and splinting – increased knee ROM and improved physical function. Overall, our large study using longitudinal change scores extends the previous literature to suggest that knee extension ROM is an important correlate of physical function in TKA.
Consistent with theoretical expectations, changes in knee extension ROM were associated with changes in knee extensor strength (Figure
2) which, in turn, were associated with changes in physical function (Figure
1). Furthermore, the results from the sensitivity analysis using a reversed mediation model support the direction of the proposed mediation. How do we explain our results? As mentioned in the Introduction, a knee flexion contracture (knee extension ROM > 0°) may potentially compress the force-length relationship
[14, 15] of the knee extensors such that strength production diminishes at lesser degrees (~30°) of knee flexion. As a corollary, our results indicate that deficits in knee extension ROM are a putative multivariate cause of knee extensor weakness in TKA. And if future studies can substantiate this notion, the discussion in the literature about whether knee ROM or knee muscle strength should be given greater prominence in TKA rehabilitation
[31–33] would become moot.
Our study has limitations. First, although our use of change scores permitted a more rigorous exploration of mediation than do cross-sectional data, our study variables were measured concurrently at each time point which precluded an examination of temporal associations between variables
. Accordingly, to better assess the criterion of temporality, future longitudinal studies should evaluate the changes in the independent variable, mediator, and outcomes sequentially over different time points. Second, and relatedly, because we did not have multiple sequential measurements at successive time intervals, as pointed out by the reviewer, we were unable to evaluate whether the rate of change in knee extension ROM influenced (moderated) the proposed mediation effects. Third, we did not have any follow-up strength measurements later than 6 months post TKA; hence, we were unable to evaluate the effects of improving knee extension ROM post 6 months on knee extensor strength and functional ability. Finally, whilst we used a self-report measure of physical function to facilitate data collection, we acknowledge that performance-based measures of physical function are necessary because they provide important, complementary information about functional status