ACTIVATION DEFICITS DO NOT LIMIT QUADRICEPS STRENGTH TRAINING GAINS IN PATIENTS AFTER TOTAL KNEE ARTHROPLASTY

ABSTRACT

Purpose/Background:

Patients after total knee arthroplasty (TKA) are known to exhibit deficits in quadriceps muscle activation. The purpose of this study was to determine if quadriceps activation levels in patients after TKA at the beginning of rehabilitation would influence quadriceps strength after rehabilitation.

Design:

A secondary analysis of data from a prospective, randomized, longitudinal clinical trial.

Setting:

Institutional clinic and research laboratory.

Participants:

Patients who underwent unilateral TKA (Men= 102; Female= 84).

Main Outcome:

Voluntary activation of the quadriceps during maximal voluntary isometric contractions (MVIC) was measured using the central activation ratio (CAR). Hierarchical multivariate regression analysis was used to determine if CAR prior to treatment could predict MVIC after the strength training intervention.

Results:

After controlling for age, sex, and initial strength levels (R²= 0.548; p<0.001), the predictability of quadriceps strength after the 6‐week intervention did not change when pain during MVIC (R²= 0.551; p= 0.317) and pre‐rehabilitation activation levels (R²= 0.551; p= 0.818) were introduced into the regression.

Conclusions:

Initial quadriceps activation levels, for patients who underwent TKA, did not predict the quadriceps strength following a strength training intervention. Therefore, deficits in voluntarily activation post‐operatively should not be considered as a rate‐limiting factor in recovering quadriceps strength after TKA.

Level of Evidence:

Retrospective cohort study. Level IIb

INTRODUCTION

Total knee arthroplasty (TKA) is the surgical procedure of choice for many with end‐stage knee osteoarthritis. The 2010 National Statistics for the United States indicate that over 600,000 TKAs were performed,¹ with projections of 3.48 million TKA procedures to be performed annually by 2030.² The projected increases may be attributed in part to the ability of the TKA to relieve pain and lead to improvements in lower extremity function,^3,4 but also to the number of younger and more active patients undergoing TKA.^5,6 While the recommendations and activity options for mature athletes become clearer, deficits in functional ability for patients who undergo TKA persist for years after surgery and may remain unresolved, especially when compared to age‐matched individuals without knee pathology.^7‐10 In individuals with knee osteoarthritis, pain is positively correlated with disability and strength is negatively correlated with disability.^11‐13 However, following TKA, when the arthritic pain is typically resolved, strength is often substantially reduced.^7,14,15 The weakening of the quadriceps following TKA is attributed to muscle atrophy and reduced levels of voluntary activation,^14,15 with activation deficits explaining more of the strength deficits than atrophy,¹⁴ primarily in the early stages after TKA.¹⁶ The factors that typically contribute to reductions in voluntary activation include pain^11,17 and swelling.^18,19 Joint damage can also lead to alterations in the sensory information from the periphery (e.g. afferent feedback), which influences the excitability of alpha motor neuron pool, in turn reducing the voluntary activation, even in the presence of maximal descending commands.²⁰ In patients with end‐stage, unilateral knee osteoarthritis, muscle activation has also been shown to be a better predictor of quadriceps weakness than lean muscle cross‐sectional area,²¹ which further supports the suggestion that quadriceps activation failure is a moderator of the relation between quadriceps strength and physical function.¹² Therefore, in individuals with weak quadriceps, the level of voluntary activation influences their functional abilities. Large activation deficits may, in turn, impede the recovery of quadriceps strength if the intensity of the strength training exercises falls below the necessary threshold for inducing strength gains.²² The benefits of high‐resistance strength training, as discussed in the review by Folland and Williams, can be attributed to a combination of both neural and muscular adaptations that are more likely to be achieved at higher resistance intensities, including hypertrophy of type II fibers and improvements in muscle activation.²³ Therefore, if activation deficits influence the effectiveness of the strength training intervention because insufficient resistances can be utilized, clinicians need to question the whether the strength training intervention will be an effective treatment for these individuals.

A recent clinical trial incorporated progressive strength training of the lower extremity musculature as a post‐surgical treatment regimen that was based‐on a model described by Stevens et al.²⁴ The training comprised a 6‐weeks regimen designed to reduce pain, improve range of motion, and focused on increasing quadriceps strength through progressive increases in exercise resistances. The findings from this study demonstrated that substantial recovery of quadriceps strength, as well as significantly greater recovery of functional performance, were observed compared to standard of care treatments that focus on range of motion and pain alone.²⁵ However, it is unclear if the quadriceps activation levels of patients after TKA, prior to starting a progressive strength training rehabilitation regimen, can predict the recovery in quadriceps strength in response to rehabilitation. Quadriceps activation is often assessed by the burst superimposition technique, where a train of electrical stimuli are applied to a muscle during a maximal voluntary contraction and the increase in muscle force, in response to the stimulation, is quantified.²⁶ Quadriceps muscle activation is often assessed in patients after TKA.^27‐30 The theory is that if the neural drive to a muscle is disrupted, the magnitude of disruption or activation deficit can be quantified and expressed relative to the maximal force output generated with the addition of a peripheral stimulation provided to the muscle. Scopaz and colleagues recently demonstrated that pre‐exercise therapy levels of the quadriceps activation, in patients with knee OA, was not predictive of the change in quadriceps strength following therapy.³¹ However, the surgical procedures involved in TKA lead to substantial changes to the knee joint, especially with respect to joint receptors and reductions in pain. Therefore, the objective of this study was to determine if the post‐surgical levels of voluntary quadriceps activation assessed prior to the start of out‐patient rehabilitation for patients after TKA could predict quadriceps strength in patients that participated in a strength training intervention. The authors' hypothesized that quadriceps activation levels prior to participating in a progressive strength training rehabilitation protocol would be predictive of quadriceps strength observed following a strength training regimen, such that individuals with lower levels of voluntary activation pre‐treatment would demonstrate less strength than those with higher activation levels after treatment.

METHODS

Subjects

The data from patients at the initial evaluation (IE) for outpatient rehabilitation and at the end of a strength training program (∼3mos) were examined as a secondary analysis of a cohort enrolled in a prospective, randomized, longitudinal clinical trial. Subjects underwent unilateral TKA (ages ranged from 48‐84 years; mean age 64.8 ± 8.6 years) and participated in a clinical trial investigating two rehabilitation protocols; patients who received conventional treatment and those who received progressive strength training (with an imbedded cohort of patients with [Exercise + NMES] and without neuromuscular electrical stimulation [Exercise Only]). The progressive strength training group (including those in the imbedded cohort) exhibited significantly larger improvements in strength and functional measures compared to the standard of care group.²⁵ The NMES protocol is described in detail elsewhere,²⁵ however, in brief the treatment involved 10 electrically evoked contractions of the quadriceps that lasted 10 seconds each. The intensity of the stimulations was determined based on patient tolerance, with a targeted minimum of 30% of the subjects' maximum voluntary isometric contraction (MVIC). There were no differences at any time period between the two progressive strength training cohorts (Exercise+NMES and Exercise Only); subjects combined from both groups comprise the study group for this analysis. Patients scheduled for unilateral TKA were invited to participate in the study by mail; interested participants were screened through a telephone interview. Individuals were subsequently excluded if they had uncontrolled hypertension, diabetes, body mass index (BMI) ≥40, symptomatic osteoarthritis in the contralateral knee (>4 out of 10 pain on verbal analog scale), other neurologic impairments or lower extremity orthopedic problems that limited function. The surgical procedure was cemented tricompartmental TKA with medial parapatellar approach. All patients received both inpatient and home physical therapy treatment prior to enrolling. The Human Subject Review Board at the University of Delaware approved the study, with all participants providing informed consent prior to participation.

Progressive Strength Training Intervention

The treatment programs were administered at the University of Delaware's Physical Therapy Clinic. The progressive strength training programs focused on improving strength in the quadriceps, hamstrings, triceps surae, and the hip flexors, extensors, and abductors.²⁴ The programs were individualized, with the type and intensity of the exercises determined and adjusted by a licensed physical therapist based each individual's clinical evaluation and follow‐up assessments. In general, subjects completed two sets of the prescribed exercises bilaterally with the resistance set to achieve a maximum of 10 repetitions. Training was progressed in all exercises to achieve a similar maximum of 10 repetitions over three sets. All subjects completed two or three sessions each week for six weeks (average of 17 sessions). The participants in the progressive strength training combined with neuromuscular electrical stimulation group also completed a set of electrically stimulated contractions to the quadriceps. The electrical stimulation protocol, which did not result in any additive benefit over the progressive strengthening program, is described in detail elsewhere.²⁵

Strength and Activation Assessments

Quadriceps strength of the operated knee was quantified as the peak force generated during a maximal voluntary isometric knee extensor contractions (MVIC). The MVICs were completed while the subject sat on an electromechanical dynamometer (KinCom; Chattecx, Chattanooga, TN) with the knee flexed to 75°, the axis of rotation of the knee was aligned with the axis of rotation of the dynamometer, and the force transducer affixed to the leg approximately 2 cm above the lateral malleolus (Figure 1). Subjects first completed warm‐up contractions at 50, 75, and 100% of maximum and were provided visual feedback on a computer screen and received verbal encouragement to maximize effort. Voluntary activation was examined with the burst superimposition test²⁶ while the subject remained seated and fitted to the dynamometer with same set‐up as the MVIC. A custom written program (LabVIEW 4.01; National Instruments, Austin, TX) delivered a 100‐Hz, 12‐pulse, stimulus train at 135 Volts (Grass S8800 stimulator in series with a Grass model stimulus isolation unit; Grass Instruments, West Warwick, RI) during the peak force of a maximal voluntary contraction through two stimulating electrodes (7.2‐ 3 12.7‐cm; CONMED Corp, Utica, NY) positioned proximally over the rectus femoris muscle and distally over the vastus medialis muscle (Figure 1). Any increased force in response to the electrical stimulation was used to quantify the level of activation. The Central Activation Ratio (CAR) was used to quantify voluntary activation by dividing the maximum force output produced voluntarily by the maximum force generated in response to the electrical stimulation.²⁶ The CAR was expressed as a percentage, such that a CAR of 1.0 is equivalent to 100% activation and a CAR of less than 95% indicating incomplete activation. If incomplete activation was observed, the procedure was repeated up to 3 times for each leg, with the subject provided with sufficient rest between trials (∼5‐minutes). The maximum MVIC and CAR from the two to three trials was retained for analyses. Pain was assessed during the MVIC using a Verbal Analog Scale (PainMVIC; 0= no pain, 10= worst pain). Subjects were also asked about their pain via the Knee Outcomes Survey by identifying to the degree to which pain affects their daily activity. The options are assigned numeric scores ranging from “I do not have pain”, which is assigned a zero, to “pain prevents me from all daily activities” which is assigned a score of five.

Figure 1.

Experimental set‐up for measuring maximal isometric knee extensor strength (MVIC) and voluntary activation (CAR) levels, as well as treatments using neuromuscular electrical stimulation (NMES).

Data Analysis

All data analyses were performed with SPSS statistical software (IBM Inc., Somers, NY). Descriptive statistics were initially evaluated to identify outliers and to assess data distribution. Bivariate correlation coefficients were then determined as factors that covaried with MVIC and/or CAR, including age, sex, BMI, pain during MVIC (PainMVIC), and group randomization (Exercise Only or Exercise +NMES). Spearman Rho coefficients were used for categorical variables (sex, group) and continuous variables that were not normally distributed (age, BMI, MVIC, Pain during MVIC, CAR). Hierarchical multivariate regression analysis was then completed to examine the predictability of quadriceps strength (MVIC) at the end of the strength training program (∼3mos; dependent variable) from variables associated with voluntary activation (CAR) or strength (MVIC) at initial evaluation (IE), prior to beginning the training program. The initial steps of the analysis controlled for those factors other than strength and activation that were correlated with either MVIC or CAR at IE. The next step of the analysis controlled for MVIC at IE and the final steps of the analysis examined the influence of PainMVIC and CAR at IE on MVIC at 3mos. Statistical significance was set at α= 0.05.

RESULTS

The current study examined the associations between data from prior to (IE) and after the strength training program (∼3mos)(N=186). Data from 186 of the original 199 subjects enrolled in the clinical trial were included in the analyses; the 13 subjects not included were not tested at 3mos. The physical activity levels for this cohort ranged from sedentary to heavy work/ manual labor. Descriptive statistics for the included subjects are presented in Table 1. Correlation analyses identified age and sex as covariates (Table 2). Sex was significantly correlated with MVIC at both IE (ρ= ‐0.206, p= 0.006 and 3mos (ρ= ‐0.490, p<0.001). Age was not correlated with MVIC at either IE (ρ= ‐0.006, p=0.930) or 3mos (ρ= ‐0.137, p=0.062), but was correlated with CAR at IE (0.202; p= 0.006). BMI was not correlated with MVIC at IE (ρ= 0.024, p= 0.747) or 3mos (ρ= 0.084; p=0.252) nor was BMI correlated with CAR at IE (‐0.015, p= 0.839). Upon completing the correlation analyses, age and sex, but not BMI, were entered into the regression analyses prior to MVIC at IE to assess the ability of using CAR at IE to predict MVIC at 3mos.

Table 1.

Descriptive statistics for the 186 patients tested. (Men= 102, Women= 84, of which Group Exercise= 106; Exercise + NMES 80)

	Mean ± SD	Range / Median
Age (yrs)	64.8 ± 8.6	48 ‐ 84 / 64
Height (m)	1.71 ± 0.10	1.47‐1.93 / 1.73
Weight (kg)	89.2 ± 17.2	47.6 ‐ 162.8 / 88.4
Pain MVICIE *	2.0 ± 2.9	0 ‐ 10 / 2.9
BMIIE (kg/m2)	30.3 ± 4.8	21.2 ‐ 53.8 / 30.3
MVICIE (N)	296.1 ± 118.1	33 – 698 / 280.5
MVIC3mos (N)	506.7 ± 189.9	138 – 1195 / 467.5
CARIE (%)	79.6 ± 16.5	20 – 100 / 84
CAR3mos (%; N=181)	88.8 ± 11.2	43 – 100 / 92.6
£PainADL at IE	2.5 ± 1.0	1 ‐ 5 / 3
£PainADL at 3mos	1.2 ± 1.0	1 ‐ 4 / 1

MVIC=Maximum voluntary isometric contraction; BMI= Body mass index; CAR= Central activation ratio; IE=initial evaluation; 3mos=3 months post‐operative

Note:

^*Pain MVIC_IE was the subjects’ pain during MVIC at initial evaluation and was assessed with a Verbal Analog Scale. (0‐10, 0= no pain; 10= worst pain)

£Pain_ADL was the degree of pain reported at IE (initial evaluation) and at 3months (3months post TKA) on the Knee Outcome Survey. (0= I do not have pain; 5= Pain prevents me from all daily activities)

Table 2.

Spearman Rho correlation coefficients between potential covariates and both voluntary activation (CAR_IE) and strength of the quadriceps (MVIC_IE & MVIC_3mos)

	CARIE	MVICIE	MVIC3mos
Sex	0.115	−0.206	−0.490
p=	0.120	0.005	<0.001
Age	0.216	0.001	−0.107
p=	0.003	0.988	0.146
BMI	−0.028	−0.004	0.075
p=	0.708	0.951	0.309
Group	−0.042	0.030	−0.031
p=	0.565	0.681	0.672
PainMVIC	−0.329	−0.075	0.099
p=	<0.001	0.310	0.177
CARIE	‐	0.555	0.214
p=	‐	<0.001	0.003

MVIC= Maximal Voluntary Isometric Contraction

CAR= Central Activation Ratio

PainMVIC was the pain felt during the MVIC at IE.

Findings from the linear regression analyses of CAR at IE predicting MVIC at 3mos are presented in Table 3. After controlling for MVIC at IE, age, and sex (R²= 0.548; p=<0.001), the addition of voluntary activation levels prior to treatment (CAR at IE) did not significantly improve the ability to predict MVIC at 3mos (R²= 0.548; p= 0.911).

Table 3.

Hierarchical regression analysis of quadriceps strength after the strength training intervention

Factor	R2	Adj. R2	df	β	F Change	Sign. F Change
Sex	0.237	0.233	1,184	−0.487	57.244	<0.001
Sex				−0.511
+Age	0.276	0.268	1, 183	−0.199	9.857	0.002
Sex				−0.385
+Age				−0.180
+MVICIE	0.548	0.541	1, 182	0.536	109.495	<0.001
Sex				−0.378
+Age				−0.166
+MVICIE				0.543
+PainMVICIE	0.551	0.541	1, 181	0.052	0.971	0.317
Sex				−0.382
+Age				−0.168
+MVICIE				0.543
+PainMVICIE				0.056
+CARIE	0.551	0.538	1, 180	0.016	0.053	0.818

β = standardized coefficients; CAR = central activation ratio; MVIC= maximal voluntary contraction, IE= initial evaluation (post surgery), 3mos= 3 months after surgery

DISCUSSION

The post‐surgical impairments in quadriceps activation observed here were not predictive of the improvements in quadriceps strength achieved through progressive strength training for patients after TKA. In that, the ability to predict patients' strength after a progressive strength training rehabilitation protocol was not dependent upon pre‐treatment voluntary activation. As the goal of the progressive strength training intervention is to more aggressively address muscle weakness after TKA than in standard care, the authors were concerned that activation deficits would have a confounding effect on the ability to strengthen the quadriceps. While the findings in this study do not agree with the stated hypothesis, they are in agreement with previous work by Scopaz and colleagues,³¹ who found that pretreatment quadriceps activation levels were not predictive of the change in quadriceps strength in response to rehabilitation for individuals with knee OA. The current work extends the findings of Scopaz et al³¹ by demonstrating that idea that activation deficits are not an impediment to strength gains for individuals free from arthritic knee pain (after TKA), compared to the patients with knee OA who do have pain.

Reductions in voluntary activation are attributed to arthrogenic muscle inhibition; a process that reduces the excitability of the alpha motor neuron pool responsible for activating a muscle.³² Various mechanisms have been linked to reductions in activation levels including joint swelling, inflammation, joint laxity, and damage to the joint's sensory receptors, all of which influence the sensory information transmitted to the nervous system.³² While TKA reliably reduces pain,³³ activation impairments persist,^14,34 whereas individuals with moderate knee OA who complete exercise therapy without undergoing TKA exhibit improvements in both quadriceps strength and voluntary activation³¹ presumably without any significant changes in pain, which was already low (0.5 ± 1.4 out of 10). Therefore, while pain may contribute to activation deficits, it is not likely to be the primary mechanism responsible for the deficits after TKA. Conversely, damage to the knee joint sensory apparatus and subsequent joint swelling are to be expected with the surgical procedures involved in TKA, but arthrogenic muscle inhibition is present for patients after TKA and individuals with end‐stage knee OA, therefore, the impairments in voluntary activation after TKA cannot be solely attributed to the surgical procedure. The average activation levels observed here (80.0 ± 16.6%), when a medial parapatellar surgical approach was used, were comparable to a previous report on patients 3‐months after undergoing bilateral TKA with either a midvastus (76.1 ± 20.0%) or subvastus (76.7 ± 14.5%) approach,³⁵ further suggesting that surgical approach is not likely a factor contributing to the impairments in activation.

Irrespective of the mechanism leading to post‐surgical weakness, resolving quadriceps weakness following TKA remains a primary focus for clinicians, as quadriceps strength and functional ability are so strongly related. Moreover, as quadriceps weakness is more strongly attributed to voluntary activation deficits in the early post‐operative period¹⁴ compared to muscle atrophy, it seems reasonable that resolving activation deficits, early in the recovery process after surgery, may lead to more substantial improvements in strength.³⁶

Regardless of the etiology, the importance of restoring quadriceps strength after TKA should remain a primary goal for treatment of individuals who undergo TKA. And while the findings here do not suggest that voluntary activation levels before rehabilitation can predict strength gains, activation failure has been shown to contribute more to quadriceps weakness after TKA than muscle atrophy, in the short‐term.¹⁴ Similar to Scopaz et al.,³¹ the patients in the current study demonstrated a broad range of changes in quadriceps strength from IE to 3mos (‐29.6% to 85.4% MVIC) and pretreatment activation was moderately correlated with pretreatment strength (r= 0.555; p< 0.001). However, in the long‐term treatment of individuals after TKA (e.g. more than one year after TKA), muscle volume, as estimated using MRI, contributes more substantially to quadriceps strength than activation.¹⁶ Therefore, future work should consider evaluating if even earlier resolution of quadriceps activation deficits after TKA can influence the long‐term recovery of quadriceps strength and lower extremity functional performance after TKA. Additionally, as pre‐surgical (prior to TKA) exercise therapy has been shown to attenuate post‐surgical weakness and activation impairments,³⁷ factors other than voluntary activation must be influencing the response to treatment; these factors may include the frequency, type, and duration of the rehabilitation protocol, patient compliance with therapist directed home exercises, or management of other impairment related factors (e.g. effusion).

CONCLUSION

Activation deficits of the quadriceps in patients who undergo TKA are not predictive of quadriceps strength three months after surgery. These findings should provide clinicians with confidence that with patients who exhibit poor voluntary activation of the quadriceps after TKA are capable of achieving meaningful gains in quadriceps strength in response to strength training and that the magnitude of recovery of quadriceps strength is not limited by pretreatment activation levels.