We have demonstrated that the prevalence of lateral PFJ cartilage damage and BMLs and medial PFJ cartilage damage was higher in weaker compared to stronger knees. We have previously reported that knees with patella alta are at increased risk for structural damage in the PFJ [7
], however we did not find an interaction between quadriceps weakness and patella alta, indicating that the presence of both weakness and patella alta did not further increase the risk of patellofemoral cartilage damage or BMLs in the lateral PFJ beyond the risk imposed by quadriceps weakness or patella alta alone. Interestingly, there was a significant interaction for medial PFJ BMLs, but it was not in the hypothesized direction. The prevalence of BMLs decreased across tertiles of quadriceps strength (strong to weak) in knees with patella alta (i.e. weakest knees with patella alta had the lowest prevalence).
Periarticular muscles act to provide stability and shock absorption to joints [11
]. Amin et al. [12
] found that knees in the lowest (weakest) tertile of quadriceps strength had 2.5 times the odds of cartilage loss over 30 months compared to those in the highest (strongest) tertile. In subjects with patella alta, quadriceps weakness may be even more detrimental to a patella that is already unstable and tracking superior to the trochlea. Although we did not find an interaction between patella alta and quadriceps weakness, our findings strongly suggested that each of these factors independently increased the risk of PFJ structural damage so that there was an absolute prevalence increase in structural outcomes (cartilage damage and BMLs) in the lateral PFJ on MRI in those knees with both patella alta and weak quadriceps compared to those with without patella alta and weak quadriceps. For example in knees with both patella alta and quadriceps weakness (weakest tertile), the prevalence of lateral PFJ cartilage damage was 51% vs. only 26% in knees with neither of these factors, adjusted prevalence difference 20.0% (95% CI 9–31). A similar pattern was seen for lateral PFJ BMLs. These results suggest that subjects with patella alta and a weak quadriceps may experience increased instability that could lead to increased shear forces and PFJ stress, thereby increasing the prevalence of structural damage in the lateral PFJ.
Alternatively, quadriceps weakness may be a result of structural damage in the PFJ (or TFJ). Once a knee develops structural damage, pain is likely to follow, and some individuals may change their movement patterns to avoid contraction of the quadriceps. It is known that individuals with patellofemoral pain ambulate with decreased quadriceps activity [18
], this quadriceps avoidance movement pattern may decrease loading of the PFJ potentially decreasing pain and weakening the quadriceps over time. Because of the cross sectional nature of this study, we are unable to determine if weakness or structural damage occurred first. However, Segal et al has demonstrated that among knees with no pain or OA at baseline, knees in the highest tertile of quadriceps strength were less likely to develop symptomatic whole knee (TFJ or PFJ) radiographic OA at 30 month follow up [19
]. This data suggests that weakness may precede the OA disease process. It is also possible that these subjects' strength may further decline due to pain and structural damage. Additionally, we performed analyses adjusting for radiographic TFJ OA (Kellgren Lawrence ≥2) and pain (WOMAC pain scale), as TFJ OA and pain may contribute to quadriceps weakness, and results were similar to our main analyses.
Knees with patella alta that were weak had had a lower prevalence of BMLs in medial PFJ subregions compared to knees that were strong (adjusted prevalence difference −14%), while there was no association between quadriceps weakness and structural damage in those without patella alta. BMLs represent areas of bone damage in response to traumatic forces, and histologically represent areas of necrosis and fibrosis [20
]. One potential mechanism for greater medial PFJ BMLs in knees with patella alta and strong quadriceps may be that patellae in these knees are more unstable and a strong quadriceps may pull the patella more laterally during contraction and as the patella returns to the midline it may compress the medial PFJ.
In the absence of patella alta, a strong quadriceps muscle potentially acts as a dynamic stabilizer and decreases forces and stress in the PFJ, reducing the prevalence of structural damage. Additionally, it is known that young females with patellofemoral pain have greater femur internal rotation [21
] and weak hip abductors and extensors compared to those without pain [22
]. It is hypothesized that hip weakness leads to internal rotation of the femur during movement and compression of the lateral PFJ. It is unknown if individuals with PFJ OA also have weak hip muscles, but this would suggest a focus on strengthening the hip extensors as a non-invasive treatment of PFJ OA.
There are limitations to this study. First, our measure of knee extensor strength was a concentric measure in an open chain position (i.e. seated extending the knee from a flexed position). We acknowledge that this is not a functional measure of strength and that our results might have been different if we had used a measure of eccentric quadriceps strength in a closed chain position (i.e. when the foot is on the ground). Concentric strength assessment was selected over eccentric for safety reasons with an older population who were at risk for cardiovascular events. Additionally, the cross-sectional design of this study limits our ability to infer the direction of causation between patella alta, structural damage, and quadriceps weakness.
In summary, in addition to our previous findings demonstrating that patella alta is associated with PFJ structural damage on MRI, we report here that quadriceps weakness is associated with similar structural damage. However, we did not confirm our hypothesis that quadriceps weakness and patella alta interact to produce more patellofemoral damage. In fact, the association between quadriceps weakness and lateral PFJ structural damage was similar among those knees with and without patella alta. Future research is needed to determine the exact mechanism by which a strong or weak quadriceps affects the mechanics of PFJ potentially leading to OA and to further illuminate how this mechanism differentially affects the lateral and medial PFJ.