Impact of neuromuscular function on pain during walking in individuals with knee osteoarthritis

Abstract

Purpose: The neuromuscular mechanisms for pain-related mobility deficits in patients with symptomatic knee osteoarthritis (OA) are not well understood. Alterations in muscle activity including prolonged activation and greater co-activation of muscles surrounding the knee have been reported and vary with radiographic severity; however, their relevance for symptom onset, severity and mobility impairment is not clear. The first study aim was to quantify differences in muscle activation patterns for individuals who do or do not experience increased pain with walking. We hypothesized that those who experience pain with walking would have greater activation intensity and total activation power as compared to those who do not. The second aim was to quantify the impact of increased pain on muscle activity during a bout of walking. We hypothesized with an increase in pain there would be a decreased mean activation intensity of medial muscles and total activation power. Methods: Twenty-two individuals with knee OA performed a 20-minute treadmill walk at a preferred pace: 11 reported an increase in self-reported pain (IP) in the affected limb, and 11 who reported no change in pain (NC)(Table1). Electromyography (EMG) was collected at 2000Hz on 8 lower extremity muscles of the most affected limb (Rectus Femoris(RF), Tibialis Anterior(TA), and the medial and lateral Vastii(VM, VL), gastronemii(MG,LG), and hamstrings(MH,LH). Nine walking strides from minute 2 and minute 20 of the treadmill walk were resolved in time and frequency space simultaneously using 20 non-linearly scaled wavelets ranging from 0-250Hz. EMG for each muscle was normalized to mean activation intensity in minute 2 and time normalized to 100% of the gait cycle. To test for group and time effects in EMG intensity a curve analysis with statistical parametric mapping (SPM) was used to apply 2-way repeated measures ANOVA over the gait cycle. EMG power (sum of intensity over time) for high (cfs: 108.7-246.1) and low (cfs: 12.2-93.0) frequency bands, and total power were also calculated. Group and time effect were assessed with a 2-way repeated measures ANOVA. Post-hoc tests were applied where appropriate.Tabled 1Table 1 DemographicsAge (yrs)BMI(kg/m2)Gait Speed (m/s)ΔvNRSKOOS painNC62.8 (6.3)25.8 (3.1)1.16 (0.6)-0.09 (0.3)69.9 (12.6)IP67.9 (3.3)25.4 (4.2)1.39 (0.4)2.27 (1.4)71.45 (11.6) Open table in a new tab Results: No significant group effects for EMG intensity were found. Main effects of time were found with a decrease of EMG intensity in the LG from 79-84% (p<0.001), and MG from 69-73% of the gait cycle (p<0.001)(Fig.1). A group*time interaction effect was found for the TA in late-stance (p<0.001; 35-43% gait cycle). Post-hoc testing found greater TA EMG intensity from 32-47% (p=0.006) of the gait cycle for NC compared to IP at minute 2 only as NC had a decrease pre to post for 10-56% gait cycle (p<0.001). Significant main effect of time was found with ∼12% decrease in total power in LG (p=0.04) and a group*time interaction in the TA (p=0.009). Post-hoc testing found at the end of the walk IP had greater TA EMG power compared to NC (p=0.04) as the NC decreased total EMG power in the walk by ∼20%, while IP had no change. No other significant differences were found for total or lower frequency power. For high frequency EMG power, main effects of time were found with decreases in LG, MH, LH, RF, TA, and VL (p<0.05). Group*time interactions were found for the TA (p=0.028), VL (p=0.029), VM (p=0.046). IP had lower EMG power compared to NC in the VL (p=0.02) and VM (p=0.002) at minute 2, and VM (p=0.04) at minute 20 (Fig.2). NC decreased EMG power in the VL (p=0.003) and VM(p=0.021) with walking. No significant post hoc differences were found for the TA. Conclusions: The study findings suggest there are differences at baseline in the EMG power within higher frequency bands of the Vastii muscles for knee OA patients who do vs do not experience pain with a bout of walking. Those who reported an increase in pain during the walk had 40% lower EMG power in the Vastii muscles at the walk start compared to those who reported no change in pain. Greater TA activation intensity at baseline was also found for those with no pain change compared to pain increase suggesting there may be differences in stride length. In response to the treadmill walk, those who had no change in pain had a reduction in high frequency EMG power in the Vastii muscles. Yet, at the end of the walk, high frequency power in the VM remained lower in those with a pain increase as compared to those without. Several possible physiologic changes may alter the higher frequency EMG including a modulation of motor unit recruitment, firing rate, and/or slowing of conduction velocity and should be further explored to understand the neuromuscular mechanisms for pain-related mobility deficits.View Large Image Figure ViewerDownload Hi-res image Download (PPT)