A number of factors that alter the local biomechanics of the knee may contribute to the observed “wear and tear” of articular degeneration in osteoarthritic (OA) joints (1–3). Evidence suggests that abnormal loading in the frontal plane of the knee, in particular, is associated with the progression of OA. Indeed, modulations in frontal plane joint laxity, alignment, and joint kinematics and kinetics during gait have been demonstrated in populations of patients with knee OA (4–8), which may lead to abnormal loading on the articular cartilage. Understanding the underlying physiologic factors that could contribute to these alterations in frontal plane joint stability may be useful in developing therapies and interventions designed to slow the progression of knee OA.
In a healthy joint, frontal plane knee stability is maintained not only by passive restraints (i.e., ligaments, bone/cartilaginous contact forces) but also through coordinated voluntary and reflexive muscle activation (9–11). The latter, neuromuscular control, is especially dependent on proprioceptive feedback to enable protective muscular stabilization in situations of normal and abnormal loading. Indeed, proprioceptive deficits have been associated with pain and functional limitations in patients with knee OA (3, 12–14). Given the importance of frontal plane joint loading on the pathomechanics of knee OA (1, 15), examination of the proprioceptive characteristics in this plane may elucidate the potential neuromechanical underpinnings of the progression of the disease.
Joint proprioception in the sagittal plane of the knee has been extensively studied in both healthy populations and those with pathology (16–18). In contrast, little is known about proprioceptive characteristics of the joint in the frontal plane and how these characteristics are modulated by OA. Technical difficulties associated with the ability to isolate varus and valgus movements at the knee may have contributed to the lack of data on frontal plane proprioception in the literature. We recently developed a testing paradigm to characterize knee joint proprioception in the varus/valgus direction in young, healthy individuals (19). This methodology can be used to evaluate whether the previously reported sagittal plane proprioceptive deficits in knee OA represent a global decline in sensation at the joint or are direction specific.
Accordingly, the aim of this study was to quantify proprioceptive acuity in both the frontal and sagittal planes in healthy and OA knees. We hypothesized that proprioceptive acuity in both planes will be decreased in patients with knee OA. However, given the disease-induced changes in frontal plane joint mechanics (7, 20), we propose that proprioceptive differences between healthy and arthritic knees will be greater in the frontal plane compared with the sagittal plane. Because frontal and sagittal plane rotations at the joint primarily target different joint structures, direction-specific variations in proprioception in the setting of knee OA may indicate that the disease differentially affects the contribution of specific joint tissues and sensory afferents. We propose that knowledge of proprioceptive acuity in the frontal plane in knee OA may help to elucidate some of the neuromechanical factors that contribute to progression of the disease.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
The results of the TDPM assessment indicated poorer proprioceptive acuity in patients with knee OA compared with control subjects (Figure 2). Repeated-measures ANOVA showed significant main effects of subject group (P = 0.002) and movement direction (P = 0.003). Post hoc comparisons using Tukey–Kramer tests revealed that TDPM values were significantly higher (indicating reduced proprioceptive acuity) in the group with knee OA compared with the control group (mean difference 0.97° [95% confidence interval (95% CI) 0.39–1.56°]). The difference between the TDPM in healthy subjects and patients with knee OA was consistent across all movement directions, with mean differences as follows: for valgus, 0.94° (95% CI 0.20–1.65°); for varus, 0.92° (95% CI 0.18–1.68°); for extension, 0.93° (95% CI 0.19–1.66°); for flexion, 1.11° (95% CI 0.38–1.85°).
Figure 2. Results of assessment of the threshold to detection of passive movement (TDPM) in the frontal (valgus and varus) and sagittal (extension and flexion) planes of the knees of patients with osteoarthritis (OA) and control subjects. In the control group, the TDPM in valgus, varus, and extension was significantly less than that in flexion (∗∗), and the TDPM in valgus was significantly less than that in extension (#). Values are the mean ± SD. ∗ = P < 0.05.
Download figure to PowerPoint
As a secondary outcome, we also examined differences in proprioception across movement directions for each group. Within the control group, the TDPM was significantly lower in the frontal plane of the knee compared with the sagittal plane (P < 0.001). Post hoc comparisons indicated that the varus TDPM value was lower than the flexion value (mean difference 0.44° [95% CI 0.20–0.67°]), and the valgus TDPM value was lower than the values for both flexion and extension (for flexion–valgus, 0.49° [95% CI 0.25–0.72°]; for extension–valgus, 0.24° [95% CI 0.004–0.48°]). In addition, results from this analysis indicated that the mean TDPM value in extension was significantly lower than that in flexion (0.25° [95% CI 0.02–0.48°]). While, on average, the TDPM value was lower in the frontal plane compared with the sagittal plane in the group with knee OA, the difference was not statistically significant (P > 0.05), perhaps due to the greater variance within this group. Indeed, post hoc power analysis indicated that the subject sample size provided only 60% power to detect differences across movement directions within the group with knee OA at a level of α = 0.05.
Linear regression analyses indicated a positive correlation between all pairs of TDPM values for both study groups (Figure 3). As shown in Figures 3A and B, TDPM values measured within the same plane of movement (i.e., varus/valgus and flexion/extension) were significantly correlated (P < 0.05) within both study groups. However, whereas both the OA and control groups demonstrated similar trend lines for the relationship between varus and valgus, the trends in flexion/extension were group specific. As seen in Figure 3B, the slope of the regression line for the OA group was nearly 1 but was <1 in the control group, indicating that TDPM values in extension tended to be less than those in flexion for the control group.
Figure 3. Correlation of threshold to detection of passive movement estimates within and across planes of movement in patients with knee osteoarthritis (OA) and control (Con.) subjects. Pearson's correlation coefficients (r) are reported to indicate the strength of the correlation. Associations within the same plane of movement, i.e., varus and valgus (A) and flexion and extension (B), were stronger than associations across planes of movements (C–F). The dashed lines represent the unity line with a slope of 1. ∗ = significant linear correlation at the P = 0.05 level.
Download figure to PowerPoint
Associations between TDPM values across planes of movement (Figures 3C–F) were weaker than those within the same plane of movement. The only significant across-plane associations were between extension and varus/valgus in the control group and between flexion and valgus in the OA group (P < 0.05). Examination of the scatter plots indicated greater variability within the OA group compared with the control group. Furthermore, it can be seen that all control subjects demonstrated lower TDPM values in the frontal plane compared with the sagittal plane (Figures 3C–F). In contrast, the relative sensitivity in the OA group was variable, with some patients demonstrating higher TDPM values in the frontal plane.
Separate linear regression analyses were performed to determine whether the TDPM was associated with subjective outcome measures in the group with knee OA. As shown in Table 2, although no significant correlations were observed, the TDPM in all movement directions was negatively correlated with WOMAC stiffness and physical function scores and positively correlated with WOMAC pain scores. However, Pearson's correlation coefficients were very low, and none of these correlations were significant (P > 0.05).
Table 2. Associations between the threshold to detection of passive movement and WOMAC scores in patients with knee osteoarthritis*
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
The purpose of this study was to examine proprioceptive acuity in both the frontal and sagittal planes of the knee in persons with and those without knee OA. Similar to previous reports (17, 18, 26, 27), our results indicated that patients with knee OA exhibited decreased proprioceptive acuity compared with age- and sex-matched healthy control subjects, in both the sagittal and frontal planes. Furthermore, the amplitude of the proprioceptive deficit in patients with knee OA was consistent across all movement directions (on average 0.97°). Because we believe that testing proprioception in each plane of movement targeted different tissues and joint afferents (19), a consistent decline in proprioception may indicate a global loss of proprioception at the joint rather than a tissue-specific loss. When comparing proprioceptive acuity across movement directions, we observed lower TDPM values in the frontal plane than in the sagittal plane in the control group, a finding that is consistent with our previous observations in young, healthy individuals (19). In contrast, the difference in TDPM values between planes was statistically insignificant in the OA group; however, this may likely reflect the greater variability within this group. Overall, the results of this study help to characterize factors that contribute to the multidimensional stability of the knee in OA.
The exact mechanisms underlying the observed proprioceptive differences in the patients with knee OA are unknown. It is possible that some level of proprioceptive impairment predated the disease and may have contributed to the initiation of knee OA. However, because this was a cross-sectional study, the degree to which proprioceptive deficits preceded knee OA cannot be determined. Conversely, proprioceptive acuity may have declined as a consequence of the disease. Multiple neural and mechanical factors associated with knee OA may influence peripheral and central sensorimotor processes, leading to a reduction in proprioceptive acuity (18). It is likely that a combination of factors contributed to the differences in proprioception observed in both planes of movement. Broadly, these factors can be grouped into 3 categories: peripheral changes in mechanoreceptor function, altered mechanical input to receptors, and potential central mechanisms.
In the periphery, specialized mechanoreceptors located in joint, muscle, skin, ligaments, and tendon convert mechanical stimuli (strain or stress) into neural signals (28). The results of previous investigations suggest that the afferent outflow of each of these may be altered in knee OA. Histologic examinations have demonstrated a loss of neural elements in the ligaments of arthritic knees (29), which would result in decreased afferent outflow. Furthermore, potential disease-related changes in the output of articular mechanoreceptors may hinder the sensitivity of muscle spindles via the γ-motor neurons, reducing proprioceptive acuity (30). Early evidence demonstrates that proprioceptive acuity can be somewhat improved by enhancing afferent outflow with subthreshold electrical stimulation (31). These findings support the notion that some of the proprioceptive deficits in knee OA are attributable to peripheral neural mechanisms.
In addition to potential changes in mechanoreceptor sensitivity or density, alterations in joint mechanics may have contributed to a decreased mechanical stimulus. With the progressive deterioration of articular cartilage and joint space narrowing, the attachment sites of the collateral ligaments of the knee move closer together, and varus/valgus laxity often develops (20). Increased laxity implies that a larger joint rotation is necessary to produce the threshold level of tension on the ligamentous structures necessary to signal joint movement. Previous examinations revealed no correlation between varus/valgus laxity and flexion/extension proprioceptive acuity (14, 26). However, given the different joint tissues targeted with each rotation, a lack of association is not surprising. Assessment of proprioceptive acuity in the frontal plane provides a more direct neurologic correlate to frontal plane joint mechanics, and this association will be the topic of future investigations.
Variations in the central processing of afferent feedback may also play a role in the proprioceptive deficits observed in the patients with knee OA. Recent evidence demonstrates impaired proprioceptive acuity in the unaffected knee and elbows of patients with unilateral knee OA (32) as well as deficits in vibratory sensation at the lower extremity (33). These results suggest generalized sensory impairments that cannot be attributed solely to local changes at the arthritic joint. As suggested by Lund and colleagues (32), changes in central sensitivity to proprioceptive feedback may be influenced by persistent nociceptive input associated with knee OA. The gate control theory of pain suggests that the transmission of painful stimuli can be blocked by various inhibitory pathways in the spinal cord and cortex (34, 35). It is possible that other mechanoreceptor input is also blocked, leading to a general reduction in proprioceptive acuity. In support of this notion, prior investigations (12, 13, 26) have noted a correlation between increased pain (as measured by the WOMAC score) and worse proprioception. In the current study, we observed a weak correlation between proprioception and pain (as well as other WOMAC scores). However, our analysis may have been limited by the small sample size and the relatively moderate pain scores reported in the group with OA. Nonetheless, further examinations are necessary to fully understand the relationship between pain and proprioceptive function in knee OA.
The results of our group analysis revealed a consistent difference in TDPM values between patients with knee OA and control subjects across all directions of movement. Thus, it could be asserted that assessment of the TDPM in the sagittal plane is also sufficient to detect proprioceptive deficits in the frontal plane when comparing groups. However, when data were analyzed on an individual level, only weak associations between frontal and sagittal plane TDPM measurements were observed (Figure 3), suggesting that sagittal plane proprioceptive acuity cannot accurately predict frontal plane acuity. This may be especially true in the OA group, which demonstrated greater variability than the control group, underscoring the subject-specific nature of proprioceptive metrics across movement directions. Given the associations between altered frontal plane joint kinetics and kinematics and knee OA (4–8), characterization of frontal plane proprioceptive acuity may provide a useful metric in the context of therapeutic interventions designed to improve joint loading during gait.
In our study population, the OA group had a significantly greater body mass index (BMI) than the control group, and it is possible that this may have influenced our results for proprioception. To our knowledge, only one study has examined the effect of BMI on joint proprioceptive capabilities, and that study showed that proprioception was poorer in obese young boys compared with nonobese boys in knee flexion but was not significantly different in knee extension or at the ankle (36). However, because these results are from a younger population, it is unclear whether they can be generalized to an older population. Furthermore, the authors provided little mechanistic explanation for the correlation between BMI and proprioception. Upon closer inspection of the current data, it was observed that BMI was significantly correlated to the TDPM across all movement directions for both study groups, indicating that subjects with a higher BMI were more likely to have poorer proprioception. When accounting for the variance due to BMI using an analysis of covariance, the TDPM in varus and valgus remained significantly different between groups (P = 0.03 for varus and P = 0.005 for valgus); however, the TDPM in flexion and in extension were no longer significantly different (P = 0.22 for flexion and P = 0.13 for extension).
These results potentially suggest that the effect of obesity is greater on proprioceptive measures in the sagittal plane than in the frontal plane. It is interesting to speculate that if, indeed, muscle spindles primarily contribute to sagittal plane proprioceptive capability, then excess body fat may have a greater impact on the sensitivity of muscle spindles than on periarticular mechanoreceptors. Nonetheless, given that patients with knee OA typically have a higher BMI than healthy individuals, future investigations may seek to better control for BMI and the effects of obesity on proprioception.
It is important to note that the results of this study may have been dependent on the metric used to characterize proprioception. Several other metrics have been used in proprioception studies, most notably the assessment of joint position sense (12, 13, 25). It has been demonstrated that results from different proprioceptive tests are not well correlated (37), likely because these tests assess different proprioception characteristics (i.e., movement sensing versus position sensing). The TDPM was chosen in the current study because it has been demonstrated to be a reliable metric to assess group differences, and it is simple to implement (17). Nonetheless, it remains to be seen whether the use of a different proprioception metric would yield similar results.
Mechanistically, proprioceptive deficits may limit the development of appropriate motor control strategies to reduce loading on the articular cartilage, which may then lead to the progression of OA (27, 38). Thus, enhancing proprioceptive acuity, particularly in the constrained frontal plane of the knee, may be important in mitigating abnormal loading on the articular cartilage. However, although proprioceptive training has long been advocated for use in sports injury rehabilitation (39), the effects of proprioceptive training paradigms on improving symptoms and function in knee OA have only recently been reported. Early evidence indicates that training programs have a positive effect on proprioceptive acuity as well as walking ability and self-reported functional outcomes (40–42). However, further investigation is necessary to develop optimal training paradigms to improve local biomechanics of the joint, particularly in the constrained frontal plane, and determine their effect on slowing disease progression (43).
The initiation and progression of knee OA are complex, multifactorial processes, and the role that proprioceptive impairments play in this process is not yet fully understood. Although cross-sectional studies have demonstrated sagittal plane proprioceptive deficits in knee OA and associations to functional limitations (14, 26), relatively few longitudinal studies have been conducted to examine the long-term consequences of these deficits. Sharma et al (3) demonstrated that poorer proprioception at baseline was associated with poorer functional abilities at 18-month and 3-year followup. Similarly, recent reports from the Multicenter Osteoarthritis Study indicated that poorer proprioception was associated with worse pain and physical functioning at a 30-month followup (12, 13). However, proprioceptive deficits were not strongly associated with the initiation or progression of radiographic knee OA (12, 13). In the context of the current study and given the importance of frontal plane joint loading in the pathomechanics of knee OA (1, 15), it may be informative to further explore the role of frontal plane proprioceptive acuity in the progression of the disease.
To our knowledge, this was the first assessment of the effects of knee OA on proprioceptive acuity in the frontal plane, a direction that may have a direct mechanical impact on the stability of the joint and the potential progression of OA. We observed similar proprioceptive deficits in patients with knee OA across all movement directions tested, suggesting that more generalized, rather than direction-specific, sensory losses are present in knee OA. Clinically, our results suggested that assessment of proprioceptive acuity in the sagittal plane is also sufficient to characterize frontal plane proprioceptive deficits associated with OA, when comparing groups of subjects. However, our correlation analysis underscored the variability and subject-specific nature of proprioceptive capabilities across movement directions. Thus, given the recent emphasis on improving neuromuscular control specifically within the frontal plane of the knee (43), understanding the improvements in frontal plane proprioceptive acuity with training may be important for the development of new therapies. Furthermore, because frontal plane joint loading has been implicated in the progression of knee OA, examining the associations between frontal plane proprioception and disease progression may shed light on the pathomechanics of knee OA.