Clinical Biomechanics 30 (2015) 521–527

Contents lists available at ScienceDirect

Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech

Review

The effect of exercise therapy on knee adduction moment in individuals with knee osteoarthritis: A systematic review Giovanni E. Ferreira a,⁎, Caroline Cabral Robinson b, Matheus Wiebusch c, Carolina Cabral de Mello Viero a, Luis Henrique Telles da Rosa a, Marcelo Faria Silva a,b a b c

Masters Program in Rehabilitation Sciences, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Brazil Doctoral Program in Health Sciences, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Brazil Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Brazil

a r t i c l e

i n f o

Article history: Received 18 August 2014 Accepted 16 March 2015 Keywords: Knee osteoarthritis Knee adduction moment Exercise therapy Systematic review

a b s t r a c t Background: Exercise therapy is an evidence-based intervention for the conservative management of knee osteoarthritis. It is hypothesized that exercise therapy could reduce the knee adduction moment. A systematic review was performed in order to verify the effects of exercise therapy on the knee adduction moment in individuals with knee osteoarthritis in studies that also assessed pain and physical function. Methods: A comprehensive electronic search was performed on MEDLINE, Cochrane CENTRAL, EMBASE, Google scholar and OpenGrey. Inclusion criteria were randomized controlled trials with control or sham groups as comparator assessing pain, physical function, muscle strength and knee adduction moment during walking at self-selected speed in individuals with knee osteoarthritis that underwent a structured exercise therapy rehabilitation program. Two independent reviewers extracted the data and assessed risk of bias. For each study, knee adduction moment, pain and physical function outcomes were extracted. For each outcome, mean differences and 95% confidence intervals were calculated. Due to clinical heterogeneity among exercise therapy protocols, a descriptive analysis was chosen. Findings: Three studies, comprising 233 participants, were included. None of the studies showed significant differences between strengthening and control/sham groups in knee adduction moment. In regards to pain and physical function, the three studies demonstrated significant improvement in pain and two of them showed increased physical function following exercise therapy compared to controls. Muscle strength and torque significantly improved in all the three trials favoring the intervention group. Interpretation: Clinical benefits from exercise therapy were not associated with changes in the knee adduction moment. The lack of knee adduction moment reduction indicates that exercise therapy may not be protective in knee osteoarthritis from a joint loading point of view. Alterations in neuromuscular control, not captured by the knee adduction moment measurement, may contribute to alter dynamic joint loading following exercise therapy. To conclude, mechanisms other than the reduction in knee adduction moment might explain the clinical benefits of exercise therapy on knee osteoarthritis. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Knee osteoarthritis (KOA) is a highly disabling condition, accounting for high social and economic burden in the western society (Neogi et al., 2009). Pain, stiffness and reduced physical capacity are the most common clinical presentations and are responsible for disability and activities limitation (Juhl et al., 2014).

⁎ Corresponding author at: Rua Sarmento Leite, 245, Porto Alegre, Rio Grande do Sul, Brazil. E-mail addresses: [email protected] (G.E. Ferreira), [email protected] (C.C. Robinson), matheusw.fi[email protected] (M. Wiebusch), [email protected] (C.C.M. Viero), [email protected] (L.H.T. da Rosa), [email protected] (M.F. Silva).

http://dx.doi.org/10.1016/j.clinbiomech.2015.03.028 0268-0033/© 2015 Elsevier Ltd. All rights reserved.

The role of biomechanical factors in KOA onset and progression has been investigated for more than a decade (Mills et al., 2013; Miyazaki et al., 2002). When it comes to biomechanics of the knee joint in individuals with KOA, the external knee adduction moment (KAM) is a widely used surrogate measure of the medial tibiofemoral contact force that reflects the relative medial-to-lateral distribution of forces across the joint (Andriacchi, 2013). Although strongly correlated to the medial tibiofemoral contact forces, the KAM varies considerably across subjects (Kutzner et al., 2013; Trepcznski, 2014). To date, there is no consensus regarding the association between the amount of structural pathology and pain intensity (Finan et al., 2013; Szebenyi et al., 2006). Besides, a recent systematic review showed that there is very limited evidence and low-quality evidence to support for a causal link between knee joint loading and structural progression of KOA (Henriksen et al., 2014).

522

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527

Exercise therapy (ET) is an evidence-based key intervention for pain relief and functional restoration based on strong recommendations from systematic reviews and expert consensus (Fransen et al., 2008; Juhl et al., 2014; McAlindon et al., 2014). Despite several factors that might account for the observed pain reduction and improved function in patients with KOA that underwent ET programs, such as cardiovascular, metabolical, neurophysiological and psychological effects, exercise regimens, such as quadriceps, hip abductors and adductors strengthening as well as neuromuscular training (Bennell et al., 2014; Fernandes et al., 2013; Thorstensson et al., 2007) have been designed with the aim of reducing joint loading in the knee. The biomechanical rationale beyond ET prescription depends on the focused muscle group. While the lateral orientation of the patellar tendon line of pull may provide abduction moments that help to stabilize the knee in the frontal plane and hence explain the mechanical effect of quadriceps strengthening, strengthening of the hip abductors is designed to improve pelvic imbalance, which results in drop of the contra-lateral pelvis and displacement in the center of mass away from the stance limb, which in turn could increase the KAM (Farrokhi et al., 2013). Irrespective of the adopted ET modality, the main rationale is to restore a proper lower limb biomechanics (Farrokhi et al., 2013). Thus, the reduction of pain and disability could be explained, among other factors, by the reduction in the KAM (Amin et al., 2009; Farrokhi et al., 2013; Foroughi et al., 2011a). Despite its recognizable clinical benefits, it is not known whether ET can influence the KAM. Our first intention was to investigate if clinical benefits of exercise therapy in individuals with KOA are associated with changes of KAM. However, given the paucity of studies that performed the assessment of the necessary outcomes to conduct an adequate explanatory analysis, we chose to verify the effect of ET on KAM in individuals with KOA in studies that assessed pain and physical function and conduct a qualitative analysis considering the effects of ET on such clinical outcomes. Muscle strength was considered too but as a secondary clinical outcome. 2. Methods 2.1. Protocol and registration This systematic review was performed in accordance with the Cochrane Collaboration and the Preferred Reporting Items for Systematic Review and Meta-Analysis, the PRISMA statement (Moher et al., 2009). The study protocol was prospectively registered on PROSPERO (registration number: CRD42013005410). 2.2. Data sources and search strategy Two reviewers (GF and CV) performed, independently, a comprehensive electronic search on MEDLINE, Cochrane CENTRAL and EMBASE from their inception to November, 2013. The search strategies for MEDLINE and EMBASE are depicted in Table 1. Additionally, a manual search within relevant systematic reviews as well as citation tracking was conducted in order to identify potential eligible studies. There were no language restrictions. Gray literature was investigated through

Google Scholar and OpenGrey, which is a multidisciplinary database which includes technical or research reports, doctoral dissertations, conference papers, official publications and other types of gray literature. 2.3. Study eligibility Inclusion criteria were randomized controlled trials (RCTs) which evaluated the effects of ET in patients with KOA by assessing physical function, pain, muscle strength and KAM, regardless of other outcomes. Trials with a control or sham groups as comparators were mandatory designs, regardless of other comparators. Participants should have KOA diagnosed according to the American College of Rheumatology criteria (Altman, 1991). ET was defined as a regimen or plan of physical activities designed and prescribed for specific therapeutic goals, whose purpose is to restore normal musculoskeletal function or to reduce pain caused by diseases or injuries. This is the Medical Subject Headings (MeSH) definition found at the National Library of Medicine controlled vocabulary thesaurus for indexing articles (www.nlm.nih.gov/mesh). Definition of ET allowed the inclusion of ET protocols irrespective of its intensity, volume and ET selection (e.g., high-load strengthening exercises and low-load motor control exercises). Studies that did not assess one of the three aforementioned outcomes, assessed the effects of a single bout of exercise, or studies with multimodal therapies (e.g., manual therapy, foot orthosis and ET) were excluded. 2.4. Study selection and data extraction Two reviewers (GF and CV) independently screened the titles and abstracts of the studies identified in the initial searches. A standard screening checklist based on the eligibility criteria above was employed for each study. Studies that did not meet the criteria, according to their titles or abstracts, were excluded. Full text versions of the remaining studies, including those potentially eligible and uncertain, were retrieved for a second review to determine eligibility. Disagreements regarding study eligibility were discussed between reviewers. When consensus was not reached, a third reviewer (CCR) arbitrated. For studies without sufficient information to evaluate the eligibility criteria, the authors were contacted via email to obtain clarifications. Studies with insufficient information after this contact were excluded. For studies with more than one publication reporting the results from the same population, the review team decided to include only the publication with the most commonly reported outcome. From the studies selected after eligibility screening, the following data were extracted: number of subjects, mean age, gender, body mass index (BMI), intervention protocol, baseline and post-intervention data of primary outcome, as well as standard deviations. Two review authors (GF and CCR) separately and independently extracted the data from the eligible studies. Disagreements regarding the data extraction between authors were resolved by discussion. When consensus was not reached, a third author (MW) arbitrated. When data were missing for synthesis or assessment of study quality, the study authors were contacted via email at least two times. When possible, estimation of

Table 1 Literature search strategy used for PubMed database. #1 Patient # 2 Intervention

# 3 Type of study

Knee osteoarthritis OR Osteoarthritis OR Knee OR Knee, Osteoarthritis of Knees OR osteoarthritis, Knee Strength Training OR Weight-Bearing Exercise Program OR Weight-Bearing Strengthening Program OR Weight-Lifting Exercise Program OR Weight-Lifting Strengthening Program OR Plyometric Drill OR Plyometric Training OR Stretch-Shortening Cycle Exercise OR Stretch-Shortening Drill OR Stretch-Shortening Exercise OR Exercise Therapy OR Exercise, Physical OR Physical Exercise OR Resistance Training OR Exercise OR Aerobic Exercise OR Exercise, Physical OR Isometric Exercise OR Exercise Movement Techniques OR Warm-up Exercise (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [mh] OR random allocation [mh] OR double-blind method [mh] OR single-blind method [mh] OR clinical trial [pt] OR clinical trials [mh] OR (“clinical trials”[tw]) OR ((singl*[tw] OR doubl*[tw] OR trebl*[tw] OR tripl*[tw]) AND (mask*[tw] OR blind*[tw])) OR (“latin square”[tw] OR placebos [mh] OR placebo*[tw] OR random*[tw] OR research design [mh:noexp] OR comparative study [mh] OR evaluation studies [mh] OR follow-up studies [mh] OR prospective studies [mh] OR crossover studies [mh] OR control*[tw] OR prospective*[tw] OR volunteer*[tw]) NOT (animal[mh] NOT human[mh]))

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527

missing data was performed (Higgins and Green, 2011). The study was excluded if data were still insufficient after this process. 2.5. Risk of bias assessment and level of evidence The risk of bias assessment was done by using The Cochrane Collaborations' tool for assessing risk of bias (Higgins and Green, 2011). Two independent reviewers (GF and MW) rated the included studies on five different items (sequence generation, allocation concealment, blinding, incomplete outcome data and selective outcome reporting) using thee different categories (high risk of bias, low risk of bias or unclear risk of bias). In the absence of consensus between both of them, a third researcher (CCR) arbitrated. Given the small number of included studies, it was considered inappropriate to present publication bias through funnel plot. The level of the evidence for each outcome was determined based on the recommendations of the Cochrane Collaboration Back Review Group (van Tulder et al., 2003). According to these criteria, the level of evidence can be divided into five categories: strong evidence (consistent findings among multiple high quality trials); moderate evidence (consistent findings among multiple low quality trials and/or one high quality trial); limited evidence (one low quality trial); conflicting (inconsistent findings among multiple trials) and no evidence from trials (no trials performed). The review team decided a priori that trials would be considered high quality if all five items but patient and therapist blinding (due to the nature of the intervention) were attended. In the presence of other biases, trials were deemed to be of low quality. The “unclear” classification was considered to be detrimental and, thus, downgraded the level of evidence. 2.6. Outcome measures The KAM is derived from kinematic and kinetic analysis, and its value is normalized to body weight. In all included studies the

523

participants walked barefoot at a self-selected normal pace. Pain was measured with the subscale of the WOMAC questionnaire, and physical function was assessed with the physical function subscale of the WOMAC questionnaire. The numeric scale used in each study varied both for the pain and physical function subscales, but given that pooling was not performed, transformations were not required and data were presented in its raw format. 2.7. Data analysis Due to the biomechanical variances between the combinations of exercises performed by the included studies, the presence of clinical heterogeneity precluded data pooling. Hence, a qualitative synthesis was the chosen method for presentation of results. 3. Results 3.1. Description of studies Data regarding study inclusion are depicted in Fig. 1. The electronic and manual searches were conducted from October to November of 2013 and yielded 1917 records. Gray literature investigation yielded 1850 citations on Google Scholar and zero citations on OpenGrey. No relevant articles could be retrieved from these two databases except from duplicates already included. After removal of duplicates, 1803 registers were screened for title and abstract, among which 1770 were excluded. The remaining 33 were assessed for eligibility through full-text read. Reasons for exclusions were: eight studies were nonrandomized controlled trials (Aalund et al., 2013; Ageberg et al., 2010; Astephen Wilson et al., 2011; Baker and McAlindon, 2000; Bechard et al., 2012; Bennell and Lim, 2007; Bennell et al., 2008; Bruce-Brand et al., 2012; Durmus et al., 2007; Foroughi et al., 2010; Forsyth et al., 2011; Hiyama et al., 2012; Hurley and Scott, 1998; Hurley et al., 2012; King et al., 2008; Riley, 2004; Shull et al., 2011; Williamson et al., 2007); one study

Fig. 1. PRISMA flow chart of inclusion procedure.

No intervention Quadriceps strengthening KL2 (4/3); KL3 (8/4); KL4 (14/19) KL2 (12/15); KL3 (7/10); KL4 (8/3) 28.2 (3.7)/ 30.3 (5.3) 29.0 (5.2)/ 28.4 (5) 13/14 10/11 67.2 (6.7)/ 66.6 (8.9) 64.1 (9.3)/ 60.8 (7.8) KOA by ACR criteria; medial knee pain; osteophytes; space narrowing; malaligned knee KOA by ACR criteria; medial knee pain; osteophytes; space narrowing; aligned knee Lim et al. (2008)

R: 42 (26/26) A: 42 (26/26) R: 55 (27/28) A: 55 (27/28)

Sham Hip abductors/adductors; knee extensors; ankle plantar flexors strengthening Omo (42.9/23.1); Omi (10.7/19.2); Os (25.0/19.2); Ovs (21.4/38.5) 33.2 (8.1)/ 31.9 (5.2) 100% woman 64 (8)/ 64 (7) R: 54 (28/26) A: 37 (18/19) KOA by ACR criteria Foroughi et al. (2011a,b)

IC, intervention group; CG, control group; Masc, masculine; BMI, body mass index; KOA, knee osteoarthritis; KAM, knee adduction moment; ACR, American College of Rheumatology; R, randomized; A, analyzed; KAM, knee adduction moment; KL, Kellgren–Lawrence (grades 1–4); O, Outerbrigde classification (mi, mild; mo, moderate; s, severe; vs, very severe); WOMAC, The Western Ontario and McMaster Universities Arthritis Index.

Strengthening did not alter the KAM in either the more malaligned or the more neutral groups but significantly improved knee pain in only the more neutrally aligned group. Any significant effect of quadriceps strengthening on physical function was found in either alignment group, despite a substantial quadriceps strength increase post intervention.

Strengthening of the lower extremity muscles was not associated with improvement in KAM. Pain and function improved significantly after both resistance training and sham exercises.

Hip strengthening had no significant effect on the external KAM but did lead to significant improvements in pain and physical function.

Peak of KAM (3-D inverse dynamics) Pain (WOMAC) Function (WOMAC) Muscle strength Peak of KAM (3-D inverse dynamics) Pain (WOMAC) Function (WOMAC) Muscle strength Peak of KAM (3-D inverse dynamics) Pain (WOMAC) Function (WOMAC) Muscle strength No intervention Hip abductors/adductors strengthening KL2 (15/15); KL3 (15/14); KL4 (15/15) 27.5 (4.7)/ 28.4 (4.1) 22/24 64.5 (9.1)/ 64.6 (7.6) KOA by ACR criteria; pain on walking ≥3; knee alignment ≤ 182°

R: 89 (45/44) A: 76 (39/37)

Outcome measurements Comparator Intervention protocol Classification (IC/CG) BMI (IC/CG) Masc. gender (IC/CG) Mean age (IC/CG) Total of randomized and analyzed (IC/CG)

Bennell et al. (2010)

3.2.1. Effect of exercise on KAM Results of each trial are summarized in Table 3. The Bennell et al. (2010) trial found no differences between strengthening and control groups in KAM [0.148 (95% confidence interval (CI) 0.039 to 0.335) Nm/BW × HT%] over the 12-week program. Indeed, a tendency toward increased KAM (4.6% adjusted) was noted in strengthening versus control group (0.5% adjusted increase). Foroughi et al. (2011a) did not find statistically significant differences in KAM between strengthening and sham-exercise group. There was no effect of strengthening on KAM in both more neutrally aligned and more malaligned individuals in the Lim et al. (2008) trial.

Participants

3.2. Effects of interventions

Author (year)

included individuals with knee arthroplasty (Aalund et al., 2013); one study did not meet the established criteria for the intervention with ET (Bechard et al., 2012); 16 studies did not assess the KAM (Aaboe et al., 2012; Aglamis et al., 2008, 2009; Bruce-Brand et al., 2012; Durmus et al., 2007; Foroughi et al., 2010; Forsyth et al., 2011; Hiyama et al., 2012; Hurley and Scott, 1998; Hurley et al., 2012; Knoop et al., 2013; Kudo et al., 2013; McQuade and de Oliveira, 2011; van Baar et al., 1998, 2001; Williamson et al., 2007). Three special cases were detected: two studies of Foroughi et al. (2011a,b) were eligible; however, both studies were performed on the same population. Therefore, we opted for exclusion of the Foroughi et al. (2011b) trial, since the KAM calculation presented in this study was divided in two peaks, which would hamper the data pooling. Sled et al. (2010) was excluded because the control group consisted of matched, healthy subjects. The study by Hunt et al. (2013) did not assess self-reported measures (i.e., pain and disability) and was excluded. Three studies were, therefore, eligible for analysis (Bennell et al., 2010; Foroughi et al., 2011a; Lim et al., 2008). A total of 233 patients were included in the qualitative analysis. All subjects had KOA diagnosed according to the American College of Rheumatology (Altman, 1991). Age ranged from 60.8 to 67.2; excepting the study by Foroughi et al. (2011a), which recruited only females, there was homogeneity between genders among the other studies. BMI ranged from 27.5 to 33.2. Two studies (Bennell et al., 2010; Lim et al., 2008) adopted the Kellgren-Lawrence classification of KOA severity, while the other (Foroughi et al., 2011a) used the Modified Outerbridge classification (Table 2). The training protocol varied among the included studies. Bennell et al. (2010) aimed to strengthen hip adductor and abductor muscles, in side-lying and standing, using elastic bands and ankle cuff weights, five times a week in a 12-week program. Patients included in this study were instructed to perform home exercises and attended a physiotherapy clinic on seven occasions in order to receive instructions and to determine load progression. The ET sessions were programmed to last 30 min. In the study by Foroughi et al. (2011a), individuals allocated to the intervention arm of the trial performed high-intensity progressive resistance training at 80% of their peak muscle strength using Keiser machines for a period of six months, three times per week. The ET protocol included unilateral knee extension, standing hip adduction and abduction, as well as bilateral knee flexion, leg press and plantar flexion strengthening and was always performed under supervision of an exercise physiologist. Three sets of eight repetitions were performed. Once a fortnight the patients were reassessed to define the new exercise intensity. The study by Lim et al. (2008) focused on quadriceps strengthening. Patients were taught five home-based exercises to be performed with ankle cuff and elastic bands. They were instructed to perform the ET protocol five times a week for 12 weeks. No details regarding training volume were reported. Individuals allocated to the control groups of the studies by Bennell et al. (2010) and Lim et al. (2008) did not perform any ET protocol and were advised to avoid starting any new treatment. In contrast, in the study by Foroughi et al. (2011a), the individuals trained with low intensity and volume and without load progression.

Conclusion

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527

Table 2 Literature search strategy used for PubMed database.

524

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527

525

Table 3 Results and conclusions. All values are reported in the original scale. Author (year)

Knee adduction moment

Bennell et al. (2010)

KAM1: MD 0.13 (95%CI -0.12 to 0.38)

Foroughi et al. (2011)

KAM1: MD -0.12 (95% CI -0.36 to 0.80)

Lim et al. (2008)

1 2 3

Clinical outcomes

Conclusions

Pain walking (0-10): MD -1.37 (95%CI -2.16 to -0.59) Pain (0-20)2: MD -2.40 (95% CI -3.25 to -1.54) Function (0-68)3: 6.17 (95% CI -9.41 to -2.93)

Strengthening improved symptoms but did not affect KAM.

Pain (0-20)2: MD -0.67 (95% CI -2.03 to 0.69) Function (0-68)3: MD -2.99 (95% CI -7.77 to 1.79)

High intensity resistance training did not improve KAM relative to controls.

Pain(0-20)2 malaligned: MD -1.6 (95% CI -7.06 to 3.86) 2 Quadriceps strenghtening did not have any significant KAM1 malaligned: MD 0.18 (95% CI -0.06 to 0.42) Pain(0-20) aligned: MD -13.9 (95% CI -19.24 to -8.55) effect on KAM in participants with either more 3 KAM1 aligned: MD -0.02 (95% CI-0.38 to 0.34) Function(0-68) malaligned: MD -4.10 (95% CI -9.94 to 1.74) malaligned or more neutrally aligned 3 Function(0-68) aligned: MD -5.40 (95% CI -10.90 to 0.10)

KAM, knee adduction moment. Measured with the WOMAC pain subscale. Measured with the WOMAC function subscale.

3.2.2. Effect of exercise on pain and physical function Results of each trial for both clinical variables are summarized in Table 3. Individuals allocated to the strengthening group of the Bennell et al. (2010) trial significantly improved in pain and physical function in comparison to control subjects. Patients in the Foroughi et al. (2011a) trial improved in pain and physical function, without significant differences among groups at 6-months follow-up. In the more neutrally aligned group, strengthening participants demonstrated a significant pain reduction in comparison to controls in Lim et al. (2008). The more malaligned group submitted to the strengthening intervention did not improve pain in comparison to malaligned controls. In this trial, alignment exerted effect on unadjusted scores of physical function, but the effect was not present after adjusting for covariates. Thus, patients submitted to strengthening in the Lim et al. (2008) trial did not improve physical function in comparison to controls.

3.2.3. Effect of exercise on muscle strength Regarding muscle strength, individuals submitted to a hip strengthening program in the trial by Bennell et al. (2010) significantly increased hip joint torques (flexion, extension, abduction, adduction, external and internal rotation), as well as the knee extension torque, in comparison with patients in the control group. Similarly, patients enrolled to the strengthening group in the Foroughi et al. (2011a) trial had better results regarding muscle strength of the knee extensors, knee flexion, plantar flexion, hip abduction and hip adduction in comparison with the sham-exercise group. Finally, both aligned and malaligned who underwent a strengthening program in the study by Lim et al. (2008) significantly increased quadriceps strength in comparison with control individuals.

3.3. Risk of bias assessment and level of evidence Risk of bias is shown in Fig. 2. Overall, none of the studies performed therapist and patient blinding, due to the nature of the intervention. In the Foroughi et al. (2011a) trial, participants were blinded to the investigator's hypothesis and were asked to inform to which group they believed to be enrolled. However, no details regarding the blinding success were available. There is moderate level of evidence (based on one high quality trial (Bennell et al., 2010) and two low quality trials (Foroughi et al., 2011a; Lim et al., 2008) that pain, physical function and muscle strength are positively affected by ET and that KAM is not significantly influenced by ET. 4. Discussion The present systematic review showed that ET is effective in reducing pain, improving physical capacity, and enhancing muscle strength but has no effect on the KAM. Considering such results, clinical efficacy of different protocols of ET was not followed by any alteration in the KAM in individuals with KOA. Several robust systematic reviews and practice guidelines endorse the positive clinical effects of ET in this population (Nelson et al., 2014; Szebenyi et al., 2006; Uthman et al., 2013), but this is the first systematic review demonstrating that the dynamic KAM was not reduced by ET, even when clinical benefits were evident. Furthermore, a tendency favoring KAM increase was shown. Although these results need to be interpreted with caution due to the small number of included studies, other studies that did not attend the inclusion criteria of this present review support our findings.

Fig. 2. Risk of bias: review author's judgment about each risk of bias item presented as percentages across all included studies.

526

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527

Sled et al. (2010) did not find significant changes in the KAM following an 8-week home program of hip abductors strengthening despite the improvement in pain and hip abductor muscle strength. King et al. (2008), with a test–retest design, found clinical improvement with no alteration in dynamic joint loading following a 12-week high intensity isokinetic resistance training program. By comparing neuromuscular exercise and quadriceps strengthening, Bennell et al. (2014) showed that both interventions had comparable beneficial effects on pain and physical function, regardless of changes in the KAM for both groups. There is a consistent body of evidence (Ageberg et al., 2010; Bennell et al., 2010; Foroughi et al., 2011a; Lim et al., 2008; Sled et al., 2010; Thorstensson et al., 2007) showing that the biomechanical mechanisms that theoretically support the efficacy of exercise, due to reduction in the KAM, appear to have no support from the literature. The clinical approaches for reducing the KAM consider the premise that vertical ground reaction force and its frontal plane lever arm are independent variables that could be manipulated through offloading strategies (Amin et al., 2009). As previously reported in individual studies and now evidenced through our descriptive analysis, the improvement in clinical outcomes such as pain and function, due to ET, was not followed by a reduction in the KAM. Notwithstanding, the rationale that supports the role of quadriceps strengthening as an offload strategy is that the contraction of such muscle group could provide abduction moments contributing to balance the KAM (Lloyd and Buchanan, 2001). Moreover, better control of the knee flexion motion provided by a stronger quadriceps could lead to reductions in the knee flexion moment, decreasing the compressive loading of the medial tibiofemoral compartment (Walter et al., 2010). Unfortunately, such mechanisms need further investigation on individuals with KOA to be clinically considered as an offloading factor. To date, this theoretical construct has not been proven to be related to clinical improvements. Furthermore, it is suggested that the hip abductors also have a role as an offload strategy. Since the drop of the contralateral pelvis due to weakness of the stance-limb hip abductors increases the external KAM lever arm by shifting the center of mass of the body away from the stance limb and toward the swing limb (Takacs and Hunt, 2012), the strengthening of such muscles would act as an offloading strategy by reducing the frontal plane lever arm of the KAM. However, the effect of such mechanism on the decrease of the KAM seems to happen only in cases of significant weakness of the hip abductors when contralateral hip drop is present (Takacs and Hunt, 2012). The only included study that assessed this issue has found an increase in pelvic drop following 12-weeks of hip abductors and adductors strengthening (Bennell et al., 2010). This review did not address whether ET is a protective intervention from a joint loading viewpoint. Even if the KAM did not significantly change after ET in all included studies, other factors, such as the role of muscle strength and neuromuscular control, must be considered, although it is controversial whether they can exert influence on disease progression (Bennell et al., 2013). Though, the inability of ET in reducing the KAM might explain why the long-term outcomes of ET are not encouraging, with benefits declining and disappearing over time (Pisters et al., 2007; van Baar et al., 2001). High BMI, which is a risk factor for KOA, was a common feature among the included studies. This is in line with previous studies that detected strong association between KOA and high BMI (Manek et al., 2003). A cross-sectional study by Harding et al. (2012) found that increased BMI was associated with changes in biomechanical pattern of the knee joint during gait when moderate KOA was present. Interestingly, weight loss has been associated with clinical improvement, such as reduction on pain and disability and increased walking speed and ambulatory knee function, but with increased dynamic joint load, which could accelerate joint degeneration (Amin et al., 2009). However, a 16-week weight loss program yielded positive clinical results despite the increased joint loading, but showing no change in structural markers of disease progression over a 1-year follow-up (Henriksen et al., 2013).

These results raise concern about the value of a biomechanicallyoriented framework for the management of KOA. In this sense, other mechanisms that explain the clinical effectiveness of ET need to be addressed in future trials (Stanton et al., 2012). Our study has limitations. Firstly, the few studies included may not allow our results to be generalized. Clinical heterogeneity between ET protocols among studies prevented pooling. Our decision to include only randomized controlled trials with control groups might have reduced the available evidence. However, this is the only study design that permits to establish an effect estimation of ET on dynamic knee loading. Nevertheless, future randomized controlled trials should focus on multiple rather than single outcomes in order to strengthen the consistency among studies. Enrolling patients with specific biomechanical alterations (e.g., increased trunk lean and increased contralateral pelvic drop) to tailored programs of ET (e.g., strengthening of hip muscles) may provide a better scenario for testing the biomechanical effects of ET rather than considering patients with KOA as a homogenous population. This scenario will further enhance our understanding about the therapeutic mechanisms of ET as well as its relative contribution to the observed clinical effects of a given intervention. We suggest that hereafter trials should present a better control of possible biases. Selective reporting bias and outcome assessor blinding are extremely feasible features that were absent in some of the included studies, which lowered the quality of the evidence. ET improved pain and physical function, but did not reduce the KAM. Mechanisms other than the reduction in dynamic joint load might explain the benefits of ET on KOA.

References Aaboe, J., Bliddal, H., Danneskiold-Samsøe, B., Christensen, P., Christensen, R., Henriksen, M., 2012. Muscle strength and muscle mass one year after an initial 16 week intense weight loss: a randomized controlled trial. Osteoarthr. Cartil. 20 (Suppl. 1), S170–S171. Aalund, P.K., Larsen, K., Hansen, T.B., Bandholm, T., 2013. Normalized knee-extension strength or leg-press power after fast-track total knee arthroplasty: which measure is most closely associated with performance-based and self-reported function? Arch. Phys. Med. Rehabil. 94 (2), 384–390 (Epub 2012/10/23). Ageberg, E., Link, A., Roos, E.M., 2010. Feasibility of neuromuscular training in patients with severe hip or knee OA: the individualized goal-based NEMEX-TJR training program. BMC Musculoskelet. Disord. 11, 126 (Epub 2010/06/23). Aglamis, B., Toraman, N.F., Yaman, H., 2008. The effect of a 12-week supervised multicomponent exercise program on knee OA in Turkish women. J. Back Musculoskelet. Rehabil. 21 (2), 121–128 (Epub 2008/6/15). Aglamis, B., Toraman, N.F., Yaman, H., 2009. Change of quality of life due to exercise training in knee osteoarthritis: SF-36 and WOMAC. J. Back Musculoskelet. Rehabil. 22 (1), 1878–6324 (Electronic). Altman, R.D., 1991. Criteria for classification of clinical osteoarthritis. J. Rheumatol. Suppl. 27, 10–12 (Epub 1991/02/01). Amin, S., Baker, K., Niu, J., Clancy, M., Goggins, J., Guermazi, A., et al., 2009. Quadriceps strength and the risk of cartilage loss and symptom progression in knee osteoarthritis. Arthritis Rheum. 60 (1), 189–198 (Epub 2009/01/01). Andriacchi, T., 2013. Valgus alignment and lateral compartment knee osteoarthritis: a biomechanical paradox of new insight into knee osteoarthritis? Arthritis Rheum. 65 (2), 310–313. Astephen Wilson, J.L., Deluzio, K.J., Dunbar, M.J., Caldwell, G.E., Hubley-Kozey, C.L., 2011. The association between knee joint biomechanics and neuromuscular control and moderate knee osteoarthritis radiographic and pain severity. Osteoarthr. Cartil. 19 (2), 186–193 (Epub 2010/11/16). Baker, K., McAlindon, T., 2000. Exercise for knee osteoarthritis. Curr. Opin. Rheumatol. 12 (5), 456–463 (Epub 2000/09/16). Bechard, D.J., Birmingham, T.B., Zecevic, A.A., Jones, I.C., Giffin, J.R., Jenkyn, T.R., 2012. Toeout, lateral trunk lean, and pelvic obliquity during prolonged walking in patients with medial compartment knee osteoarthritis and healthy controls. Arthritis Care Res. 64 (4), 525–532 (Epub 2012/01/04). Bennell, K., Lim, B.W., 2007. Does strength training affect the incidence and progression of knee osteoarthritis? Nat. Clin. Pract. Rheumatol. 3 (3), 134–135. Bennell, K.L., Hunt, M.A., Wrigley, T.V., Lim, B.W., Hinman, R.S., 2008. Role of muscle in the genesis and management of knee osteoarthritis. Rheum. Dis. Clin. N. Am. 34 (3), 731–754 (Epub 2008/08/09). Bennell, K.L., Hunt, M.A., Wrigley, T.V., Hunter, D.J., McManus, F.J., Hodges, P.W., et al., 2010. Hip strengthening reduces symptoms but not knee load in people with medial knee osteoarthritis and varus malalignment: a randomised controlled trial. Osteoarthr. Cartil. 18 (5), 621–628. Bennell, K.L., Kyriakides, M., Metcalf, B., Egerton, T., Wrigley, T.V., Hodges, P.W., et al., 2014. Neuromuscular versus quadriceps strengthening exercise in people with medial knee osteoarthritis and varus malalignment: a randomised controlled trial. Arthritis Rheum. 66 (4), 950–959 (Epub 2014/01/01).

G.E. Ferreira et al. / Clinical Biomechanics 30 (2015) 521–527 Bennell, K.L., Wrigley, T.V., Hunt, M.A., Lim, B.W., Hinman, R.S., 2013. Update on the role of muscle in the genesis and management of knee osteoarthritis. Rheum. Dis. Clin. N. Am. 39 (1), 145–176 (Epub 2013/01/15). Bruce-Brand, R.A., Walls, R.J., Ong, J.C., Emerson, B.S., O'Byrne, J.M., Moyna, N.M., 2012. Effects of home-based resistance training and neuromuscular electrical stimulation in knee osteoarthritis: a randomized controlled trial. BMC Musculoskelet. Disord. 13, 118 (Epub 2012/07/05). Durmus, D., Alayli, G., Canturk, F., 2007. Effects of quadriceps electrical stimulation program on clinical parameters in the patients with knee osteoarthritis. Clin. Rheumatol. 26 (5), 674–678 (Epub 2006/08/10). Farrokhi, S., Voycheck, C.A., Tashman, S., Fitzgerald, G.K., 2013. A biomechanical perspective on physical therapy management of knee osteoarthritis. J. Orthop. Sports Phys. Ther. 43 (9), 600–619 (Epub 2013/06/13). Fernandes, L., Hagen, K.B., Bijlsma, J.W., Andreassen, O., Christensen, P., Conaghan, P.G., et al., 2013. EULAR recommendations for the non-pharmacological core management of hip and knee osteoarthritis. Ann. Rheum. Dis. 72 (7), 1125–1135 (Epub 2013/04/19). Finan, P.H., Buenaver, L.F., Bounds, S.C., Hussain, S., Park, R.J., Haque, U.J., et al., 2013. Discordance between pain and radiographic severity in knee osteoarthritis: findings from quantitative sensory testing of central sensitization. Arthritis Rheum. 65 (2), 363–372 (Epub 2012/09/11). Foroughi, N., Smith, R.M., Lange, A.K., Baker, M.K., Fiatarone Singh, M.A., Vanwanseele, B., 2010. Dynamic alignment and its association with knee adduction moment in medial knee osteoarthritis. Knee 17 (3), 210–216 (Epub 2009/11/10). Foroughi, N., Smith, R.M., Lange, A.K., Singh, M.A., Vanwanseele, B., 2011a. Progressive resistance training and dynamic alignment in osteoarthritis: a single-blind randomised controlled trial. Clin. Biomech. 26 (1), 71–77 (Epub 2010/09/28). Foroughi, N., Smith, R.M., Lange, A.K., Baker, M.K., Fiatarone Singh, M.A., Vanwanseele, B., 2011b. Lower limb muscle strengthening does not change frontal plane moments in women with knee osteoarthritis: a randomized controlled trial. Clin. Biomech. 26 (2), 167–174 (Epub 2010/10/05). Forsyth, A.L., Quon, D.V., Konkle, B.A., 2011. Role of exercise and physical activity on haemophilic arthropathy, fall prevention and osteoporosis. Haemophilia 17 (5), e870–e876 (Epub 2011/03/26). Fransen, M., McConnell, L., Bell, M., 2008. Exercise for osteoarthritis of the knee. Cochrane Database Syst. Rev. 8 (4), CD004376. Harding, G.T., Hubley-Kozey, C.L., Dunbar, M.J., Stanish, W.D., Astephen Wilson, J.L., 2012. Body mass index affects knee joint mechanics during gait differently with and without moderate knee osteoarthritis. Osteoarthr. Cartil. 20 (11), 1234–1242 (Epub 2012/08/21). Henriksen, M., Hunter, D.J., Dam, E.B., Messier, S.P., Andriacchi, T.P., Lohmander, L.S., et al., 2013. Is increased joint loading detrimental to obese patients with knee osteoarthritis? A secondary data analysis from a randomized trial. Osteoarthr. Cartil. 21, 1865–1875. Henriksen, M., Creaby, M.W., Lund, H., Juhl, C., Christensen, R., 2014. BMJ Open 4 (7), e005368 (Epub 2014/07/15). Higgins, J.P.T., Green, S. (Eds.), 2011. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration (Available from www.cochrane-handbook.org). Hiyama, Y., Yamada, M., Kitagawa, A., Tei, N., Okada, S., 2012. A four-week walking exercise programme in patients with knee osteoarthritis improves the ability of dual-task performance: a randomized controlled trial. Clin. Rehabil. 26 (5), 403–412 (Epub 2011/10/07). Hunt, M.A., Pollock, C.L., Kraus, V.B., Saxne, T., Peters, S., Huebner, J.L., et al., 2013. Relationships amongst osteoarthritis biomarkers, dynamic knee joint load, and exercise: results from a randomized controlled pilot study. BMC Musculoskelet. Disord. 14, 115 (Epub 2013/03/28). Hurley, M.V., Scott, D.L., 1998. Improvements in quadriceps sensorimotor function and disability of patients with knee osteoarthritis following a clinically practicable exercise regime. Br. J. Rheumatol. 37 (11), 1181–1187 (Epub 1998/12/16). Hurley, M.V., Walsh, N.E., Mitchell, H., Nicholas, J., Patel, A., 2012. Long-term outcomes and costs of an integrated rehabilitation program for chronic knee pain: a pragmatic, cluster randomized, controlled trial. Arthritis Care Res. 64 (2), 238–247 (Epub 2011/09/29). Juhl, C., Christensen, R., Roos, E.M., Zhang, W., Lund, H., 2014. Impact of exercise type and dose on pain and disability in knee osteoarthritis: a systematic review and metaregression analysis of randomized controlled trials. Arthritis Rheum. 66 (3), 622–636 (Epub 2013/12/20). King, L.K., Birmingham, T.B., Kean, C.O., Jones, I.C., Bryant, D.M., Giffin, J.R., 2008. Resistance training for medial compartment knee osteoarthritis and malalignment. Med. Sci. Sports Exerc. 40 (8), 1376–1384 (Epub 2008/07/11). Knoop, J., Dekker, J., van der Leeden, M., van der Esch, M., Thorstensson, C.A., Gerritsen, M., et al., 2013. Knee joint stabilization therapy in patients with osteoarthritis of the knee: a randomized, controlled trial. Osteoarthr. Cartil. 21 (8), 1025–1034 (Epub 2013/06/01). Kudo, M., Watanabe, K., Otsubo, H., Kamiya, T., Kaneko, F., Katayose, M., et al., 2013. Analysis of effectiveness of therapeutic exercise for knee osteoarthritis and possible factors affecting outcome. J. Orthop. Sci. 18 (6), 932–939 (Epub 2013/10/03). Kutzner, I., Trepczynski, A., Heller, M.O., Bergmann, G., 2013. Knee adduction moment and medial contact force — facts about their correlation during gait. PLoS One 8 (12), e81036 (Epub 2013/12/07).

527

Lim, B.W., Hinman, R.S., Wrigley, T.V., Sharma, L., Bennell, K.L., 2008. Does knee malalignment mediate the effects of quadriceps strengthening on knee adduction moment, pain, and function in medial knee osteoarthritis? A randomized controlled trial. Arthritis Rheum. 59 (7), 943–951 (Epub 2008/06/26). Lloyd, D.G., Buchanan, T.S., 2001. Strategies of muscular support of varus and valgus isometric loads at the human knee. J. Biomech. 34, 1257–1267. Manek, N.J., Hart, D., Spector, T.D., MacGregor, A.J., 2003. The association of body mass index and osteoarthritis of the knee joint: an examination of genetic and environmental influences. Arthritis Rheum. 48 (4), 1024–1029 (Epub 2003/04/11). McAlindon, T.E., Bannuru, R.R., Sullivan, M.C., Arden, N.K., Berenbaum, F., Bierma-Zeinstra, S.M., et al., 2014. OARSI guidelines for the non-surgical management of knee osteoarthritis. Osteoarthr. Cartil. 22, 363–388 (Epub 2014/01/28). McQuade, K.J., de Oliveira, A.S., 2011. Effects of progressive resistance strength training on knee biomechanics during single leg step-up in persons with mild knee osteoarthritis. Clin. Biomech. 26 (7), 741–748 (Epub 2011/04/26). Mills, K., Hunt, M.A., Ferber, R., 2013. Biomechanical deviations during level walking associated with knee osteoarthritis: a systematic review and meta-analysis. Arthritis Care Res. 65 (10), 1643–1665 (Epub 2013/04/05). Miyazaki, T., Wada, M., Kawahara, H., Sato, M., Baba, H., Shimada, S., 2002. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann. Rheum. Dis. 61 (7), 617–622 (Epub 2002/06/25). Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., 2009. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 339, b2535 (Epub 2009/ 07/23). Nelson, A.E., Allen, K.D., Golightly, Y.M., Goode, A.P., Jordan, J.M., 2014. A systematic review of recommendations and guidelines for the management of osteoarthritis: the Chronic Osteoarthritis Management Initiative of the U.S. Bone and Joint Initiative. Semin. Arthritis Rheum. 43, 701–712 (Epub 2014/01/07). Neogi, T., Felson, D., Niu, J., Nevitt, M., Lewis, C.E., Aliabadi, P., et al., 2009. Association between radiographic features of knee osteoarthritis and pain: results from two cohort studies. BMJ 339, b2844 (Epub 2009/08/25). Pisters, M.F., Veenhof, C., van Meeteren, N.L., Ostelo, R.W., de Bakker, D.H., Schellevis, F.G., et al., 2007. Long-term effectiveness of exercise therapy in patients with osteoarthritis of the hip or knee: a systematic review. Arthritis Rheum. 57 (7), 1245–1253 (Epub 2007/10/02). Riley, D., 2004. Quadriceps strength and osteoarthritis progression in malaligned and lax knees. Ann. Intern. Med. 140 (2), 149 (Epub 2004/01/22). Shull, P.B., Lurie, K.L., Cutkosky, M.R., Besier, T.F., 2011. Training multi-parameter gaits to reduce the knee adduction moment with data-driven models and haptic feedback. J. Biomech. 44 (8), 1605–1609 (Epub 2011/04/05). Sled, E.A., Khoja, L., Deluzio, K.J., Olney, S.J., Culham, E.G., 2010. Effect of a home program of hip abductor exercises on knee joint loading, strength, function, and pain in people with knee osteoarthritis: a clinical trial. Phys. Ther. 90 (6), 895–904 (Epub 2010/04/10). Stanton, T.R., Lin, C.W., Smeets, R.J., Taylor, D., Law, R., Lorimer, Moseley G., 2012. Spatially defined disruption of motor imagery performance in people with osteoarthritis. Rheumatology (Oxford) 51 (8), 1455–1464 (Epub 2012/04/03). Szebenyi, B., Hollander, A.P., Dieppe, P., Quilty, B., Duddy, J., Clarke, S., et al., 2006. Associations between pain, function, and radiographic features in osteoarthritis of the knee. Arthritis Rheum. 54 (1), 230–235 (Epub 2005/12/31). Takacs, J., Hunt, M.A., 2012. The effect of contralateral pelvic drop and trunk lean on frontal plane knee biomechanics during single limb standing. J. Biomech. 45, 2791–2796. Thorstensson, C.A., Henriksson, M., von Porat, A., Sjodahl, C., Roos, E.M., 2007. The effect of eight weeks of exercise on knee adduction moment in early knee osteoarthritis — a pilot study. Osteoarthr. Cartil. 15 (10), 1163–1170 (Epub 2007/05/01). Trepcznski, 2014. Modulation of the relationship between external knee adduction moments and medial joint contact forces across subjects and activities. Arthritis Rheum. 66 (5), 1218–1227. Uthman, O.A., van der Windt, D.A., Jordan, J.L., Dziedzic, K.S., Healey, E.L., Peat, G.M., et al., 2013. Exercise for lower limb osteoarthritis: systematic review incorporating trial sequential analysis and network meta-analysis. BMJ 347, f5555 (Epub 2013/09/24). van Baar, M.E., Dekker, J., Oostendorp, R.A., Bijl, D., Voorn, T.B., Lemmens, J.A., et al., 1998. The effectiveness of exercise therapy in patients with osteoarthritis of the hip or knee: a randomized clinical trial. J. Rheumatol. 25 (12), 2432–2439 (Epub 1998/ 12/19). van Baar, M.E., Dekker, J., Oostendorp, R.A., Bijl, D., Voorn, T.B., Bijlsma, J.W., 2001. Effectiveness of exercise in patients with osteoarthritis of hip or knee: nine months' follow up. Ann. Rheum. Dis. 60 (12), 1123–1130 (Epub 2001/11/16). van Tulder, M., Furlan, A., Bombardier, C., Bouter, L., 2003. Updated method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group. Spine 28 (12), 12909 (Epub 2003/06/18). Walter, J.P., D'Lima, D.D., Colwell Jr., C.W., Fregly, B.J., 2010. Decreased knee adduction moment does not guarantee decreased medial contact force during gait. J. Orthop. Res. 28, 13481354. Williamson, L., Wyatt, M.R., Yein, K., Melton, J.T., 2007. Severe knee osteoarthritis: a randomized controlled trial of acupuncture, physiotherapy (supervised exercise) and standard management for patients awaiting knee replacement. Rheumatology (Oxford) 46 (9), 1445–1449 (Epub 2007/06/29).