Derek J. Rutherford

Co-activation differences in lower limb muscles between asymptomatic controls and those with varying degrees of knee osteoarthritis during walking

BACKGROUND Increased muscle co-activation during gait has been identified as a neuromuscular alteration associated with knee osteoarthritis, however levels of co-activation among different osteoarthritis severity have not been established. The purpose of this study was to determine if differences in co-activation could be detected among asymptomatic controls, those with moderate and those with severe osteoarthritis using a co-activation index and a pattern recognition technique. METHODS Surface electromyograms from vastus lateralis and medialis, lateral and medial hamstring and gastrocnemius pairs were recorded from 63 asymptomatic, 59 moderate and 48 severe osteoarthritic subjects during self-selected walking. A co-activation index was calculated over the initial stance for four medial and lateral muscle pairs. The four co-activation indices were tested among groups using a one factor ANOVA (alpha=0.05). Gait waveform pattern recognition procedures were applied to yield a principal pattern, scored for each muscle site and subject. A mixed model ANOVA (group-muscle) tested for principal pattern score differences. FINDINGS A significant group effect was found (P<0.05) for all four co-activation indices. Principal pattern one captured the amplitude and general shape of activity throughout the entire stance phase. ANOVA revealed a significant (P<0.05) group by muscle interaction for the principal pattern scores. Significant differences were found among all three groups and between the two osteoarthritic groups for both measures. INTERPRETATION The co-activation indices and principal patterns identified that lateral site differences occurred among all three groups with medial site differences between the two osteoarthritic groups. These findings suggest that measures of muscle co-activity provide additional information related to knee osteoarthritis severity.

Foot progression angle and the knee adduction moment: a cross-sectional investigation in knee osteoarthritis

OBJECTIVE To test the hypothesis that an association exists between the characteristics of the knee adduction moment and foot progression angle (FPA) in asymptomatic individuals and those with mild to moderate and severe knee osteoarthritis (OA). DESIGN Fifty asymptomatic individuals, 46 patients with mild to moderate and 44 patients with severe knee OA were recruited. Maximum knee adduction moment during late stance and principal component analysis (PCA) were used to describe the knee adduction moment captured during gait. Multiple regression models were used for each of the three group assignments to analyze the association between the independent variables and the knee adduction moment. RESULTS FPA explained a significant amount of the variability associated with the shape of the knee adduction moment waveform for the asymptomatic and mild to moderate groups (P<0.05), but not for the severe group (P>0.05). Walking velocity alone explained significant variance associated with the shape of the knee adduction moment in the severe OA group (P<0.05). CONCLUSION A toe out FPA was associated with altered knee adduction moment waveform characteristics, extracted using PCA, in asymptomatic individuals and those with mild to moderate knee OA only. These findings are directly implicated in medial knee compartment loading. This relationship was not evident in those with severe knee OA.

Explaining the hip adduction moment variability during gait: Implications for hip abductor strengthening

BACKGROUND There is emerging interest in hip abductor function during gait and its potential relationship to knee joint pathology. During gait, the hip abductor muscles are primarily responsible for generating moments of force to control frontal plane movement. The current study investigated the relationship between hip abductor muscle function and frontal plane hip moments of force during gait. METHODS Frontal plane hip moments of force and electromyographic features of gluteus medius were measured during walking in 22 healthy individuals. Hip abductor strength, subject anthropometrics and gait velocity were recorded. Multiple regression models were used to evaluate the relationship between the anthropometric, velocity, strength and electromyographic variables and the initial and mid-stance magnitude of the hip adduction moment. FINDINGS A positive relationship was found between the initial peak moment (Nm), and both body mass and gait speed (R(2)=90%). Body mass (positive) and hip abductor strength (negative) explained significant levels of mid-stance magnitude variability (R(2)=62.5%). Gait speed (positive) explained significant levels of variability in the normalized initial peak moment (Nm/kg) (R(2)=52%). No variables were included in the normalized mid-stance moment model (P>0.05). INTERPRETATION Body mass was the key factor associated with high hip adduction moments during initial and mid-stance of the gait cycle. Increased gait velocity was associated with higher initial peaks and higher muscle strength was associated with lower mid-stance magnitude of the external hip adductor moment during walking. These findings suggest that in a healthy adult population, hip abductor strength and activation were not directly related to the hip adduction moment magnitude during gait.

Neuromuscular alterations exist with knee osteoarthritis presence and severity despite walking velocity similarities

BACKGROUND Neuromuscular strategies during walking in individuals with knee osteoarthritis are being explored for diagnostic information; however, isolating differences to disease progression is difficult given walking velocity decreases with osteoarthritis severity. This study investigated lower extremity electromyograms during walking in asymptomatic individuals and individuals with different severities of knee osteoarthritis who walked with similar self-selected velocities. METHODS Muscle activity in lateral and medial gastrocnemius, vastus lateralis and medialis, rectus femoris and the lateral and medial hamstrings was monitored during self-selected walking in 230 subjects with asymptomatic knees, moderate and severe knee osteoarthritis. Sixteen asymptomatic individuals, 16 individuals with moderate and 15 individuals with severe knee osteoarthritis were identified based on similarities in average walking velocity. Principal component analysis was employed to derive amplitude and temporal characteristics of the electromyographic (EMG) waveforms. Analysis of variance models tested for group and muscle differences in principal pattern scores (α=0.05). Bonferroni post hoc testing was utilized on all significant findings. FINDINGS Despite similar walking velocities, individuals with moderate knee OA had elevated and prolonged quadriceps and elevated lateral hamstring activity compared to asymptomatic individuals (P<0.05). A diminished phase shift between medial and lateral gastrocnemius muscle activation, greater and prolonged lateral compared to medial hamstring activation were found in the severe group compared to asymptomatic and moderate knee OA groups (P<0.05). INTERPRETATION Lower extremity neuromuscular function during walking is altered with the presence and severity of knee osteoarthritis and not simply a direct function of walking velocity.

Changes in knee joint muscle activation patterns during walking associated with increased structural severity in knee osteoarthritis

PURPOSE To determine whether alterations in knee joint muscle activation patterns during gait were related to structural severity determined by Kellgren-Lawrence (KL) radiographic grades, for those with a moderate knee OA classification. SCOPE Eighty-two individuals with knee OA, classified as moderate using a functional and clinical criterion were stratified on KL-grade (KL II, KL III and KL IV). Thirty-five asymptomatic individuals were matched for age and walking velocity. Lower limb motion and surface electromyograms from rectus femoris plus lateral and medial sites for the gastrocnemii, vastii and hamstring muscles were recorded during self-selected walking. Gait velocity and characteristics from sagittal plane knee angular displacement waveforms were calculated. Principal component analysis extracted amplitude and temporal features from electromyographic waveform. Analysis of variance models tested for main effects (group, muscle) and interactions (α=0.05) for these features. No differences in anthropometrics, velocity, knee muscle strength and symptoms were found among the three OA groups (p>0.05). Specific features from medial gastrocnemius, lateral hamstring and quadriceps amplitude and temporal patterns were significantly different among OA groups (p<0.05). CONCLUSIONS Systematic alterations in specific knee joint muscle activation patterns were associated with increasing structural severity based on KL-grades whereas other alterations were associated with the presence of OA.

Maximal voluntary isometric contraction exercises: a methodological investigation in moderate knee osteoarthritis

PURPOSE The objectives were, (i) to determine whether differences exist in relative activation amplitudes for participants with asymptomatic knees and participants with moderate medial compartment knee osteoarthritis during a series of maximal effort contractions and (ii) to determine whether maximum activations occurred on similar exercises for both groups. SCOPE Sixty-eight participants with asymptomatic knees and 68 participants with moderate medial compartment knee osteoarthritis completed eight standardized 3-s maximal voluntary isometric exercises. Maximal electromyographic amplitudes were identified for a 100 ms window from three quadriceps, two gastrocnemius and two hamstring muscle sites for each exercise. For each exercise, amplitudes were normalized to percent of the absolute maximum activation (%MVIC). Frequency counts for exercises eliciting absolute maximum amplitudes were recorded. Analysis of variance models determined exercise and group main effects and interactions in relative amplitudes (%MVIC) for each muscle. CONCLUSION The exercises produced similar relative activation amplitudes between groups. The highest relative amplitude occurred for gastrocnemius during standing plantarflexion (86-93%MVIC), for the vasti during knee extension (45°) and (15°) (81-86%MVIC), for rectus femoris during knee extension (15°) (89%MVIC) and for hamstring muscles during knee flexion (15°) and prone knee flexion (55°) (81-94%MVIC). No single exercise elicited absolute maximum activation for every participant for each muscle, supporting the value of using an exercise series for normalization purposes.

Knee effusion affects knee mechanics and muscle activity during gait in individuals with knee osteoarthritis.

OBJECTIVE To test the hypothesis that knee effusion presence in those with knee osteoarthritis (OA) alters knee joint muscle activation patterns and sagittal plane mechanics during gait. METHODS Thirty-five patients with medial compartment knee OA were assessed for the presence of effusion using a brush test. Based on the results, they were assigned to the knee effusion (n = 17) and no knee effusion (n = 18) groups. Electromyograms from seven lower extremity muscles (lateral and medial gastrocnemius, vastus lateralis and medialis, rectus femoris and the lateral and medial hamstrings), leg motion and ground reaction forces were recorded during self-selected walking. Isometric knee extensor, plantar flexor and knee flexor strength were measured. Discrete measures from angular knee motion and net external moment of force waveforms were identified. Principal component analysis extracted electromyographic waveform features. Analysis of variance models tested for main effects (group, muscle) and interactions (α = 0.05). Bonferroni post-hoc testing was employed. RESULTS No differences in age, body mass index, knee pain, Western Ontario McMaster Osteoarthritis Index scores, gait velocity and muscle strength were found between groups (P > 0.05). Individuals with effusion had a greater overall quadriceps activation and prolonged hamstring activation into mid-stance (P < 0.05). Knee joint flexion angles were higher (P < 0.05) and net external knee extension (KE) moments in mid to late stance lower in the effusion group. CONCLUSION Quadriceps and hamstrings activation during walking were altered when effusions were present. Increased knee flexion (KF) angles and decreased KE moment in mid-late stance provide a mechanical explanation for the effect of joint effusion on muscle activation in those with knee OA.

Reliability of surface electromyographic recordings during walking in individuals with knee osteoarthritis

To determine test-retest reliability of a surface electromyographic protocol designed to measure knee joint muscle activation during walking in individuals with knee osteoarthritis (OA). Twenty-one individuals with moderate medial compartment knee OA completed two gait data collections separated by approximately 1month. Using a standardized protocol, surface electromyograms from rectus femoris plus lateral and medial sites for the gastrocnemii, vastii and hamstring muscles were recorded during walking. After full-wave rectification and low pass filtering, time and amplitude normalized (percent of maximum) waveforms were calculated. Principal component analysis (PP-scores) and co-contraction indices (CCI) were calculated from the waveforms. Intraclass correlation coefficients (ICC2,k) were calculated for PP-scores and CCIs. No differences in walking speed, knee muscle strength and symptoms were found between visits (p>0.05). The majority of PP-scores (17 of 21) and two of four CCIs demonstrated ICC2,k values greater than 0.81. Remaining PP-scores and CCIs had ICC2,k values between 0.61 and 0.80. The results support that reliable EMG characteristics can be captured from a moderate knee OA patient population using a standardized protocol.

Reliability of principal components and discrete parameters of knee angle and moment gait waveforms in individuals with moderate knee osteoarthritis

Gait measures are used to evaluate change in patients with knee osteoarthritis (OA), but reliability has not been fully established in this population. This study examined test-retest reliability of knee angle and moment gait waveform characteristics captured using discrete parameters and principal component analysis (PCA) in individuals with moderate knee OA. Participants (n=20) underwent three-dimensional gait analysis on two occasions. Motion and force data were captured using two camera banks, infrared light emitting diodes and force plate during self-selected walking. Knee angle and moment waveforms were calculated and analyzed using discrete parameters and by identifying waveform characteristics using PCA. Intraclass correlation coefficients (ICC2,k) examined test-retest reliability of discrete parameters and PCA derived scores (PC-scores). ICC2,k values ranged from 0.57 to 0.93 for discrete parameters, 0.52-0.86 for knee angle PC-scores and 0.30-0.94 for the knee moment PC-scores. However, 10 of 13 discrete parameters, six of nine knee angle PC-scores and seven of nine knee moment PC-scores had ICC2,k values greater than or equal to 0.70. Discrete parameters and PC-scores from flexion angles and adduction moments had the highest ICC2,k values while adduction angles, rotation angles, and rotation moments had the lowest. Most knee angle and moment waveform characteristics demonstrated ICC2,k values that could be interpreted as acceptable. Caution should be used when examining adduction and rotation angle magnitudes and early/mid-stance rotation moment magnitudes due to lower ICC2,k values.

The neuromuscular demands of altering foot progression angle during gait in asymptomatic individuals and those with knee osteoarthritis

OBJECTIVES To determine the immediate effects of a toe-out foot progression angle modification during gait on the major lower limb muscle activation characteristics and to establish whether asymptomatic individuals and those with moderate knee OA have similar responses. DESIGN Seventeen patients with knee OA and 20 asymptomatic control subjects participated. Informed consent was obtained. Electromyographic (EMG) recordings were acquired from the lateral and medial gastrocnemii, vastus lateralis, vastus medialis, rectus femoris and the lateral and medial hamstrings during neutral and toe-out walking conditions. The EMG waveforms were amplitude normalized to maximal voluntary isometric contractions and time normalized to the gait cycle. Principal component analysis extracted principal waveform features. Analysis of variance models tested for main effects and interactions. Bonferroni post hoc testing was employed (alpha=0.05). RESULTS Both groups altered foot progression angle by approximately 15 degrees during toe-out walking (P<0.05). A shift in gastrocnemius activation towards later stance (P<0.05) and increased magnitude and duration of quadriceps activation (P<0.05) was found. A differential activation occurred in the overall magnitude and principal shape of the lateral and medial hamstring musculature in the asymptomatic group only (P<0.05). Significant group differences were shown in each muscle analysis (P<0.05). CONCLUSION Neuromuscular demands of adopting a toe-out gait differ from a neutral foot progression angle. Demands also differ between asymptomatic controls and patients with moderate knee OA. These findings have relevance for altered joint loading and changes in metabolic cost of this gait modification in individuals with knee OA.