Gluteus medius activation in females with patellofemoral pain syndrome by Ryan Unsworth

Is Patellofemoral Pain Syndrome Associated with Alterations in Gluteus Medius Neuromuscular Activation in the Female Population? A Literature Review By

Ryan Patrick Unsworth

OXFORD BROOKES UNIVERSITY

Supervisor: Stephen Castleton

Word Count: 20,550

This dissertation is submitted in part fulfilment of the regulations for a degree of Masters of Osteopathy

At Faculty of Health and Life Sciences Marston Road Oxford Brookes University Oxford January 2015

CONTENTS Acknowledgements ...................................................................................................................................................... iv Abstract......................................................................................................................................................................... v 1.0 - Introduction ...........................................................................................................................................................1 1.1 - Patellofemoral Pain Syndrome ..........................................................................................................................1 1.2 - Osteopathy ........................................................................................................................................................2 1.3 - Rationale ............................................................................................................................................................3 1.4 - Aims and Objectives ..........................................................................................................................................4 1.5 - Research Question .............................................................................................................................................4 2.0 - Preliminary Literature Review ...............................................................................................................................5 2.1 - Introduction .......................................................................................................................................................5 2.2 - Anatomy ............................................................................................................................................................5 2.3 - Patellofemoral Pain Syndrome ..........................................................................................................................8 2.3.1 - Finding a Definition.....................................................................................................................................8 2.3.2 - Causes of PFPS ............................................................................................................................................9 2.4 - Biomechanics ...................................................................................................................................................11 2.5 - Gender Differentiation ....................................................................................................................................13 2.6 - Psychological Implications ...............................................................................................................................14 2.7 - Conclusion .......................................................................................................................................................15 3.0 â&#x20AC;&#x201C; Methodology ......................................................................................................................................................16 3.1 - Introduction .....................................................................................................................................................16 3.2 - Research Paradigms.........................................................................................................................................16 3.2.1 - Positivism ..................................................................................................................................................17 3.2.2 - Post-Positivism .........................................................................................................................................17 3.2.3 - Constructivism ..........................................................................................................................................18 3.2.4 - Critical Theory ...........................................................................................................................................18 3.3 - Research Methods ...........................................................................................................................................19 3.3.1 â&#x20AC;&#x201C; Qualitative Research ................................................................................................................................19 3.3.2 Quantitative Research ................................................................................................................................19 3.4 - Research Design...............................................................................................................................................21 3.4.1 - Hierarchy of Evidence ...............................................................................................................................21 3.4.2 - The Literature Review...............................................................................................................................24 3.5 - Literature Search .............................................................................................................................................25 3.5.1 - Keywords ..................................................................................................................................................25 3.5.2 - Truncations, Wildcards and Boolean Operators. ......................................................................................26 3.5.3 - Inclusion exclusion criteria .......................................................................................................................27 3.5.4 - Snowball Sampling ....................................................................................................................................28

3.5.5 - Grey Literature .........................................................................................................................................28 3.6 - Databases ........................................................................................................................................................29 3.7 - Critical Appraisal ..............................................................................................................................................29 3.7.1 - Electromyographic Critical Appraisal ........................................................................................................30 3.8 - Evidence Analysis .............................................................................................................................................31 3.8.1 - Thematic Analysis .....................................................................................................................................31 3.8.2 - Meta Analysis ...........................................................................................................................................32 3.8.3 - Conclusion ................................................................................................................................................32 3.9 - Ethical Considerations .....................................................................................................................................32 3.10 - Conclusion .....................................................................................................................................................33 4.0 - Method ................................................................................................................................................................35 4.1 - Introduction .....................................................................................................................................................35 4.2 - Data Collection ................................................................................................................................................35 4.2.1 - Online Databases ......................................................................................................................................35 4.2.2 - Key Terms .................................................................................................................................................36 4.2.3 - Search Tools ..............................................................................................................................................38 4.2.4 - Inclusion and Exclusion Criteria ................................................................................................................40 4.2.5 - Snowballing ..............................................................................................................................................42 4.2.6 - Grey Literature .........................................................................................................................................42 4.5 - Critical Appraisal ..............................................................................................................................................42 4.5.1 - CASP Critical Appraisal ..............................................................................................................................42 4.5.2 - EMG Critical Appraisal ..............................................................................................................................43 4.5.3 â&#x20AC;&#x201C; Valuation of Evidence ..............................................................................................................................43 4.6 - Analysis ............................................................................................................................................................44 5.0 - Results .................................................................................................................................................................45 5.1 - Introduction .....................................................................................................................................................45 5.2 - Search Results ..................................................................................................................................................45 5.3 - Critical Appraisal ..............................................................................................................................................47 5.3.1 - Research Aim and Background .................................................................................................................48 5.3.2 - Research Design and Methodology ..........................................................................................................48 5.3.3 - Sample ......................................................................................................................................................49 5.3.4 - Bias ...........................................................................................................................................................51 5.3.5 - Confounding Factors.................................................................................................................................52 5.3.6 - Results/Conclusions ..................................................................................................................................54 5.3.7 - Precision and Validity ...............................................................................................................................57 5.3.9 - Clinical Relevance .....................................................................................................................................60 5.3.10 - Summary .................................................................................................................................................60

6.0 - Analysis ................................................................................................................................................................64 6.1 - Introduction .....................................................................................................................................................64 6.2 - Theme Identification .......................................................................................................................................64 6.2.1 - Sub-theme Justification ............................................................................................................................65 6.2.2 - Theme Classification .................................................................................................................................69 6.3 - Major Themes ..................................................................................................................................................72 6.3.1 - Gluteus Medius Neuromuscular Activation ..............................................................................................72 6.3.2 - Functional Tasks .......................................................................................................................................73 6.4 - Moderate Themes ...........................................................................................................................................78 6.4.1 - Electromyography Method .......................................................................................................................78 6.5 - Minor Themes..................................................................................................................................................80 6.5.1 - Leg dominance..........................................................................................................................................80 7.0 - Discussion ............................................................................................................................................................81 7.1 â&#x20AC;&#x201C; Introduction ....................................................................................................................................................81 7.1.1 - Research Question ....................................................................................................................................81 7.2 - Statement of Findings ......................................................................................................................................81 7.3 - Contextualising Themes...................................................................................................................................82 7.3.1 - Major Themes ...........................................................................................................................................82 7.3.2 - Functional Tasks .......................................................................................................................................85 7.3 - Moderate Themes ...........................................................................................................................................87 7.2.3 - Electromyography Measurements ...........................................................................................................87 7.4 - Minor Themes..................................................................................................................................................89 7.5 - Recommendations for Future Research ..........................................................................................................89 7.6 - Study Limitations .............................................................................................................................................90 7.7 - Ethical Considerations .....................................................................................................................................91 8.0 - Conclusion ...........................................................................................................................................................92 8.1 - Research Question ...........................................................................................................................................92 8.2 - Conclusion .......................................................................................................................................................92 8.3 â&#x20AC;&#x201C; Recommendations For Practice ......................................................................................................................93 8.4 - Reflection .........................................................................................................................................................93 References ...................................................................................................................................................................94 Appendices ................................................................................................................................................................108 Appendix A.............................................................................................................................................................108 Appendix B .............................................................................................................................................................113 Appendix C .............................................................................................................................................................128

iii

ACKNOWLEDGEMENTS I would like to take this chance to express my complete gratitude to my inspiring parents, who have taught me how to smile at a challenge; to my friends, who have filled my years with more laughter than I thought possible; to my girlfriend, for her unwavering support and care; to Alan Watts, for a better perspective; and of course to Stephen, who read more of this than I would wish upon anyone.

ABSTRACT Background: Patellofemoral pain syndrome (PFPS) is a prevalent and commonly misunderstood condition.

Often referred to as the â&#x20AC;&#x153;black hole of orthopaedicsâ&#x20AC;?; as it has no single definition, aetiological process, diagnostic criteria or treatment plan. Research into PFPS interventions has been predominately inconclusive, suggesting that the underlying factors which are causing PFPS are not being addressed. The

Patellofemoral Pain Research Retreat requested that future research addresses gender differences and/or proximal aetiological factors of PFPS. Recent research has been addressing GMed NMA, however the evidence was limited and the link to gender is yet to be made. Aim: To assess associations between the neuromuscular activation of gluteus medius and patellofemoral

pain syndrome in females. Methods: A critical literature review utilising an exhaustive systematic search process and strict eligibility criteria to address all relevant research between December 2004 and December 2014. The suitable studies will then be assessed using the CASP critical appraisal tool and subjected to a thematic analysis in order to extract the themes and conclusions of the collective data. After this the conclusions will be discussed and related to osteopathic practice. Results: Ten studies were selected for analysis; six studies found a statistical difference between groups.

There was a significant delay in GMed onset timing during balancing tasks (P=0.025) and during gait there was a significant delay in onset timing (P=0.028) and a shortening of duration (P=0.01). In jump tasks a significant increase in average activation (P=0.002) was found. Discussion: When comparing to Barton et al (2013) who used a mixed gender sample; both reviews

found significant associations. However their results were incongruous; showing both corroborative and conflicting evidence. The variability of the comparison reflects the multi-factorial nature of PFPS, and the tissue homeostasis model. Conclusion: There is limited evidence to suggest an increase in GMed amplitude during jump tasks, a delay of onset during balance stabilisation and both delayed and shortened GMed contraction during running.

1.0 - INTRODUCTION This chapter will introduce the topic of the literature review, along with the project’s rationale, aims, and objectives. In addition to this, the osteopathic philosophy and related models of health will stated so that their significance can be referenced later in the review.

1.1 - PATELLOFEMORAL PAIN SYNDROME Patellofemoral pain (PFP) is estimated to account for 20% of all knee pain presentations in sports clinics (Gobbi et al 2014, Rathleff et al 2013, Boling et al 2010). The number of NHS physiotherapy sessions for patients with PFP is usually 8 to 17 sessions, and only 46% are pain free when discharged (Brown 2000). Widely considered the most misunderstood, neglected, and problematic of the knee pathologies is patellofemoral pain syndrome (Van der Heijden et al 2015, Sanchis-Alfonso 2014, Witvrouw et al 2014, Dye 2001). It is often referred to it as the “black hole of orthopaedics”, due to it having no single diagnostic criteria or treatment plan (Motealleh et al 2011, Sanchis-Alfonso 2011). Due to the uncertainty around the aetiology of PFPS, the treatment is complex and varies in success. Misdiagnosis can lead to unsuccessful treatment and unnecessary surgery, causing psychological distress in patients and distrust for the practitioner (Sanchis-Alfonso 2011). Patellofemoral pain syndrome (PFPS) occurs most commonly in females during adolescence or early adulthood (Petersen et al 2013, Lankhorst et al 2012). It is characterised by retropatellar or peripatellar knee pain which is made worse for loading of the knee extensors; for example prolonged sitting, squatting, running and stair ambulation (Van der Heijden et al 2015, Hryvniak et al 2014, Witvrouw et al 2014, Petersen et al 2013, Lankhorst et al 2012). The diagnosis is based on pain during these activities in the absence of trauma or the following pathologies: Hoffa’s syndrome, OsgoodSchlatters, Sinding-Larsen-Johansson’s, iliotibial band syndrome, plica, osteoarthritis, rheumatoid arthritis, or any bursitis, tendinitis or neuromas (Van der Heijden et al 2015, Hryvniak et al 2014). Therefore it is a diagnosis based on exclusion, and due to the complexity in ruling out all other pathologies there is a trend in misdiagnosing and over diagnosing this syndrome (Ferrari et al 2014, Nunes et al 2013). The aetiology is widely thought to be multi-factorial in nature, involving influences 1

both proximal and distal to the knee joint (Van der Heijden et al 2015, Hryvniak et al 2014, Lack et al 2014, Witvrouw et al 2014, Petersen et al 2013, Lankhorst et al 2012).

1.2 - OSTEOPATHY Osteopathy is a manual therapy which follows the philosophy that health is dependent on the structure and function of the body, and the well-being of the mind. Osteopathic care addresses dysfunction holistically and specifies the treatment to the person rather than the pathology (Thomson et al 2013, General Osteopathic Council 2012, Parsons & Marcer 2006). The practice of osteopathy was founded in the late 19th century by Andrew Taylor Still (Parsons & Marcer 2006). The practice has developed significantly since then; utilising a broad range of manual techniques, and following an evidence based paradigm (Cotton 2013, General Osteopathic Council 2012, General Osteopathic Council 2006, Parsons & Marcer 2006). Although the practice of osteopathy has changed the philosophical principles remain the same:

●

The body is a unit

●

The body possesses self-regulatory mechanisms

●

Structure and function are reciprocally inter-related

●

Rational therapy is based on an understanding of body unity, self-regulatory mechanisms, and

the inter-relationship of structure and function. (Cotton 2013) There are two concepts which are not exclusive to osteopathy, yet central to osteopathic practice; the biopsychosocial model and the salutogenic model. Both are similar in that they view dysfunction holistically. The biopsychosocial model is comparable to the total osteopathic lesion concept, which states that the pathology is a product of all aspects of the individual and their environment, not simply of the body (Parsons & Marcer 2006). The salutogenic model agrees with this concept however advances it by focusing on the promotion of health, rather than the treatment of dysfunction.

1.3 - RATIONALE Selfe et al (2013) stated that there were 52 RCTs (involving 2667 subjects) that investigated interventions for PFP, 60% of these did not establish statistical significance. On the 20th of January 2015 a Cochrane systematic review was published that concluded; there is still insufficient evidence to conclude any single treatment plan for PFPS (Van der Heijden et al 2015). The variability in results of interventions suggests that the underlying factors which are causing PFPS are not being addressed (Witvrouw et al 2014). The research is attempting to find a treatment for a condition which has not yet been understood. If the aetiological and influential factors are not known, they cannot be controlled, and as seen in the research outcomes, they will produce insignificant findings (Dye 2005). For every 2-3 years since 2009 the worldâ&#x20AC;&#x2122;s leading PFP specialists meet, to discuss and combine their findings in order to further the understanding of PFP. From their collection of opinions and discussions they write a consensus, which summarises what the research has shown and where it needs to be focused (Witvrouw et al 2014, Powers et al 2012, Davis & Powers 2010). In the past three Patellofemoral Pain Research consensuses, there were requests for future research to address gender differences and/or proximal aetiological factors (Witvrouw et al 2014, Powers et al 2012, Davis & Powers 2010). In the most recent consensus a potential association of changes in gluteus medius neuromuscular activation was identified in subjects with PFPS. However the evidence was limited and the link to gender is yet to be made.

1.4 - AIMS AND OBJECTIVES It is the aim of this review to develop the understanding of PFPS, by following the recommendations made by the current patellofemoral pain consensus (Witvrouw et al 2014). In following these recommendations, this review will be addressing associations between the neuromuscular activation of gluteus medius and patellofemoral pain syndrome in females.

It is the objective of this review to initially further the education of osteopaths and manual practitioners on the current associative factors which should better equip them to diagnose and manage the pathology in their patients. Secondly; to contribute to the patellofemoral research community, in the progression of understanding the aetiological factors behind PFPS.

1.5 - RESEARCH QUESTION

â&#x20AC;&#x153;Is patellofemoral syndrome pain associated with alterations in gluteus medius neuromuscular activation in the female population?â&#x20AC;?

2.0 - PRELIMINARY LITERATURE REVIEW

2.1 - INTRODUCTION A preliminary literature review is used to develop an account of the current understanding of in a topic, so that the question can be defined by where the research is required (Ross 2012, Aveyard 2010). This initial review will include the relevant anatomy, biomechanics and current theories and definitions of PFPS.

2.2 - ANATOMY The gluteal muscles are a set of three large muscles positioned posterior to the ilium and attach to the femur. The three muscles have different functions due to their differing sizes and orientation (Bartlett et al 2014, Flack et al 2014) Gluteus Maximus (GMax) is the largest of the three muscles (see figure 2.2.1). It is a powerful hip extensor and is the primary muscle which aids humans to stand upright on two legs (Antonio et al 2013, Drake et al 2009, Lieberman et al 2006). The superficial GMax insertion feeds into the fascia lata, which forms the iliotibial tract. Therefore as a secondary function; the GMax tightens the fascia of the lower extremity (Bartlett et al 2014, Antonio et al 2013, Myers 2001).

Figure 2.2– Gluteus Medius and Minimus

Figure 2.1 – Gluteus Maximus

Figure 2.2 – Illustrating the posterior hip musculature when gluteus maximus has been removed. Taken from Palastanga et al (2012)

Figure 2.1 – Illustrating the gluteus maximus origin and insertion, and the posterior bony landmarks of the pelvis. Taken from Palastanga et al (2012)

The gluteus medius (GMed) muscle is covered with strong fascia and shares the posterior part of this with the GMax (see figure 2.2.2). It plays an essential part in pelvic movement during functional activities such as squatting, jumping walking and running (Barton et al 2013, Palastanga et al 2012, Stone & Stone 2011, Drake et al 2009). The gluteus minimus (GMin), is the smallest of the gluteal muscles, however its origin spans the largest portion of the pelvis (see figure 2.2.2) (Drake et al 2009). GMin is vital in functional stabilisation of the pelvis and is similar in function to GMed (Palastanga et al 2012, Stone & Stone 2011, Drake et al 2009).

Table 2.1 â&#x20AC;&#x201C; Gluteal Anatomical Characteristics Name Origin Insertion Gluteus Maximus Posterior aspect of Gluteal tuberosity ilium and and iliotibial band posterolateral aspect of sacrum and sacrotuberous ligament Gluteus Medius Lateral surface of Superiolateral side ilium between the of the greater posterior and trochanter of the anterior gluteal femur. lines Gluteus Minimus Gluteal surface of Anterosuperior ilium in front of aspect of the the anterior and greater trochanter above the inferior of the femur. gluteal lines.

Innervation Inferior gluteal nerve L2-5

Function Extends, externally rotates.

Superior gluteal nerve L4-S1

Hip abduction and medial rotation

Superior gluteal nerve L4-S1

Medial rotation and hip abduction

(Stone & Stone 2011)

The knee is a relatively immobile joint between two mobile joints, therefore when biomechanical alterations occur it is predominately the knee which becomes dysfunctional (Palastanga et al 2012, Ferber et al 2011, Fulkerson & Arendt 2000). The current research has already identified some biomechanical factors which are associated with PFPS; these are shown in section 2.4.

2.3 - PATELLOFEMORAL PAIN SYNDROME As stated in section 1.1 PFPS is described as one of the most difficult pathologies to manage and treat (Dye 2001). This clinical uncertainty is a product of the causes and definitions remaining undefined in the literature. The following sections will cover the definitions, and theories of aetiology which are currently used in the literature.

2.3.1 - FINDING A DEFINITION. The definition of patellofemoral pain syndrome differs throughout the literature (Petersen et al 2013, Witvrouw et al 2005). This is due to the patient often exhibiting a variety of physical characteristics and symptoms (Hryvniak et al 2014, Petersen et al 2013). PFPS is a type of anterior knee pain; which is an umbrella term used for all anterior knee pathologies for example: chondromalacia patella, Osgood-schlatters, Sinding-larson Johansson, plicae, prepatellar bursitis etc (Gobbi et al 2014). Because of the ambiguity in the pathology the term â&#x20AC;&#x2DC;syndromeâ&#x20AC;&#x2122; is used. A syndrome is a group of signs and symptoms which commonly occur in combination (Witvrouw et al 2005). According to Petersen et al (2013) PFPS is defined by anterior knee pain in patients with no structural abnormalities i.e. altered Q-angle, or pathological changes to the articular cartilage. Orthopaedic tests such as Clarks (patella compression) test were used for diagnosis, however more recently these tests have been discredited as too sensitive and being very unspecific (Doberstein 2008, Fredericson and Yoon 2006, Post 1999).

2.3.2 - CAUSES OF PFPS The current theories for the development of PFPS are as described below; in table 2.2. Table 2.2 â&#x20AC;&#x201C; Aetiological Theories of PFPS Theory of Explanation Aetiology Biological theory states that PFPS is a clinical misdiagnosis of idiopathic Biological Theory chondromalacia patella and synovial plica. Neurogenic theory states that PFPS is caused by repetitive microtrauma to the Neurogenic Theory

patellofemoral joint (PFJ), which causes ischemic degeneration of the joint retinaculum and local neural structures. Therefore producing an influx of pain mediators such as substance P; leading to hyperalgesia (sensitisation) and pain. Biomechanical theory states that PFPS is caused by anatomical and biomechanical alterations in postural alignment and dynamic kinematics (Huston 2008). This causes the increased loading of the PFJ and the release of cartilaginous fragments,

Biomechanical Theory

inducing inflammation and pain. This theory is widely supported by current research as it easily supplements most kinematic research (Gobbi et al 2014, Hryvniak et al 2014, Nakagawa et al 2014, Petersen et al 2013, Lankhorst et al 2012, Nakagawa et al 2012, Boling et al 2010, Powers 2010, Boling et al 2009, Waryasz & McDermott 2008). Tissue homeostasis (TH) theory states that joints are active metabolic systems, and that PFPS is caused by a disruption of joint homeostasis. The amount of force which the joint can withstand is called the envelope of function. This envelope is determined by anatomical structure, biomechanical function, physiological factors, and treatment given (Dye 2005, Dye 2001). If forces exceed the range of the

Tissue Homeostasis Theory

envelope, increases in osseous remodelling and peripatellar synovitis begin to develop, which causes the envelope to decrease (Dye 2001). When the envelope is reduced to the point that a regular activity such as squatting or descending stairs exceeds it, PFPS develops (Dye 2005). This theory is less concerned with the individual factors but instead keeping the exerted forces within the ranges of tissue homeostasis. The theory explains why the aetiological factors are clinically elusive (Gobbi et al 2014, Sanchis-Alfonso 2011, Dye 2005, Dye 2001). (Gobbi et al 2014)

There are parallels in TH theory to osteopathic practice, as a central principle of both is addressing the individual holistically to facilitate the body to heal itself (Sanchis-Alfonso 2014, SanchisAlfonso 2011, Parsons & Marcer 2006, Dye 2005, Dye 2001). The Biomechanical and tissue homeostasis (TH) theories are currently the most convincing and justified (Gobbi et al 2014, Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Powers 2010). The two theories can co-exist as the TH theory accepts the biomechanical theory explores the factors which could reduce the envelope of function. The TH theory gives context to the biomechanical model, as clinicians can tend to be focused on set rehabilitation plans, whereas the TH model focuses on the individualâ&#x20AC;&#x2122;s ability to heal through aiding the functional envelope to increase (Gobbi et al 2014, Sanchis-Alfonso 2014). Through researching the biomechanical factors, such as GMed NMA, and understanding the TH theory, a more holistic treatment approach can be developed (Gobbi et al 2014, Sanchis-Alfonso 2014, SanchisAlfonso 2011).

2.4 - BIOMECHANICS The biomechanical factors involved are classed as intrinsic and extrinsic. Both of which are thought to produce patella maltracking. Extrinsic factors are related to the environment and external strains placed on the knee, including training frequency, intensity, surfaces used and footwear. Intrinsic factors are forms of lower extremity malalignment and muscular imbalances, these factors are displayed with their supporting evidence in table 2.3 Table 2.3 â&#x20AC;&#x201C; Intrinsic Factors Associated with PFPS Intrinsic Factors Associated with PFPS

Evidence

Hip Abductor Weakness

Gobbi et al 2014, Lankhorst et al 2012, Powers 2010, Waryasz and McDermott 2008 Gobbi et al 2014, Petersen et al 2013, Lankhorst et al 2012, Radzimski et al 2012, Hollman et al 2006, Murphy et al 2003 Gobbi et al 2014, Lankhorst et al 2012, Boling et al 2009, Waryasz and McDermott 2008 Lankhorst et al 2012, Whyte et al 2010, Boling et al 2009, White et al 2009, Waryasz and McDermott 2008 Lankhorst et al 2012, Waryasz and McDermott 2008 Gobbi et al 2014, Hryvniak et al 2014, Nakagawa et al 2014, Petersen et al 2013, Lankhorst et al 2012, Nakagawa et al 2012, Massada et al 2011, Boling et al 2010, Powers 2010, Boling et al 2009, Waryasz & McDermott 2008, Tyler et al 2006, Ferber et al 2003

Excessive Foot Pronation Vasti Group Asymmetries Hamstring Tightness Iliotibial Tract Tightness Increased Dynamic Q-angle

All factors seen in table 2.3 have been linked to dynamic knee valgus (increased functional Qangle) (see figure 2.3). The Q-angle is most commonly measured statically; however it has been shown that the angle increases in some patients when performing certain functional tasks (Powers 2010). There is limited evidence to say that static valgus deformities are related to PFPS (Massada et al 2011). However dynamic Q-angle has been widely acknowledged as an associative factor to PFPS (see table 2.3)

Figure 2.3 - Knee Orientation

Figure 2.3 - Displaying a normal knee orientation (A) and genu valgus (increased Q-angle) (B).

Zazulak et al (2007) concluded that decreased neuromuscular control of the trunk is directly associated with an increased risk of knee injury. Saad et al (2011) theorised that this is due to a limitation in keeping postural control. If the activation of the postural musculature is reduced or delayed there are greater postural oscillations, as the body adjusts to stabilisation. As the knee is a relatively immobile joint these oscillations cannot be distributed as well as in the other more mobile joints (Palastanga et al 2012, Ferber et al 2011, Fulkerson & Arendt 2000). This microtrauma is repeated till the stress accumulates to the stage that the PFJ becomes symptomatic (Saad et al 2011). The TH model would state that it is at this point where the forces have exceeded the functional envelope (Dye 2005, Dye 2001).

2.5 - GENDER DIFFERENTIATION It is a widely documented that women are more likely to develop PFPS than men (Gobbi et al 2014, Petersen et al 2013, Lankhorst et al 2012, Nakagawa et al 2012, Pappas and Wong-Tom 2012, Massada et al 2011, Willson et al 2011, Boling et al 2010, Myer et al 2010, Powers 2010, Waryasz and McDermott 2008, Willson and Davis 2008, Zazulak et al 2007, Taunton et al 2002). Women are twice as likely as males to develop PFPS (Boling et al 2010, Taunton et al 2002), and Pappas and Wong-Tom (2012) found that 70% of PFPS cases occur in young females (16-25). However the factors which cause this gender bias are not conclusive (Boling et al 2010). Different theories have been suggested, with moderate evidence in each, the suggested causes are shown in table 2.4.

Table 2.4 â&#x20AC;&#x201C; Causes of Gender Bias in PFPS Causes of Justification Gender Bias

Supporting Research

Larger Q-angle

Females tend to have a wider pelvis than males, which increases the Q-angle to keep stability. It has been a suggested that this gender characteristic predisposes women to PFPS. However as stated in 2.4; the relevance of static Qangle in its association with PFPS is disputed in the literature (Fredericson & Yoon 2006, Lun et al 2004, Mizuno et al 2001, Livingston & Mandigo 1999).

Souza and Powers (2009) and Willson and Davis (2008).

Tendency for Dynamic Valgus

Women more frequently show increased Q-angle, relative weakness, and tendency for ligamentous laxity. These commonly cause greater amounts of hip adduction & internal rotation, knee external rotation and foot pronation; all of which increase PFJ loading.

Gobbi et al (2014), Nakagawa et al (2012), Radzimski et al (2012), Massada et al (2011), Boling et al (2010), Willson and Davis (2008), Hollman et al (2006), and Murphy et al (2003)

Weakness of LEx Musculature Activation Pattern

Menstrual Cycles

Females have a significant reduction in muscular strength when compared to males Quadriceps and hip abductor weakness has been correlated with the development of PFPS. Females show a significant increase in GMax activation compared to males, and tend to have less GMed control; relying more on the quadriceps musculature to control the knee. Knee valgus angles were found to be significantly increased in both follicular phases of menstruation. In addition to this females are at a significantly higher risk of injury during the early follicular phase due to altered joint proprioception. However women on the contraceptive pill showed no changes.

LEx=lower extremity

Lankhorst et al (2012), Pappas and Wong-Tom (2012), Bolgla et al (2010), and Boling et al (2010). Nakagawa et al (2012) and Zazulak et al (2007) Souza and Powers (2009) Zeller et al (2003)

Labato et al (2013), Fouladi et al (2012) and Cesar et al (2010).

2.6 - PSYCHOLOGICAL IMPLICATIONS The effects of PFPS are not only physical, but often affect the emotional state of the subject (Crossley 2014, Sanchis-Alfonso 2014). The disorder often occurs in young adulthood, an age when individuals are most likely to develop mental health issues (Patel et al 2007). It has been shown that young adults mental state is frequently dependent on body image and social ranking (Faith et al 2011, Kawachi & Berkman 2001). Physical activity is a major component in controlling weight and social gratification through sports (Crossley 2014). When this is no longer possible due to pain the psychological implications can often be just as difficult to maintain as the symptoms (Sanchis-Alfonso 2014). If inactivity is maintained a cycle can occur; as the weight increases, so does the stress on the joints, making physical activity more problematic. As the symptoms get worse, the patient’s tend to get more psychologically agitated (Doménech et al 2014). Additionally, a greater body weight has been associated with increased PFJ stress and loss of patella cartilage, which will aggravate PFPS (Crossley 2014, Hryvniak et al 2014, Petersen et al 2013, Yard & Comstock 2011). Psychological factors such as previously unsuccessful treatment can make the patients demotivated, depressed and anxious that symptoms will be made worse. These factors can make long term rehabilitation difficult due to an anxious and distrustful attitude (Selfe et al 2013, Jensen et al 2005, Clark et al 2000). Doménech et al (2014) and Sanchis-Alfonso (2014) found the effects of catastrophisation and kinesiophobia can reduce the patients’ ability to recover.

2.7 - CONCLUSION From the findings of this preliminary literature review and the documented ineffectiveness of standardised treatment plans, the evidence suggests that the osteopathic and salutogenic model could be an appropriate alternative to conventional treatment (Gobbi et al 2014, Witvrouw et al 2014, Selfe et al 2013, Parsons & Marcer 2006, Lindstrรถm & Eriksson 2005). However without understanding the biomechanical influences which are affecting the functional envelope, the practitioner is cannot to aid the body in its recovery (Gobbi et al 2014, Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Marks 2002). As stated in section 1.5, this review is in answer to the Patellofemoral Research Consensus recommendations (Witvrouw et al 2014, Powers et al 2012). By addressing associations between the neuromuscular activation of gluteus medius and patellofemoral pain syndrome in females, with the ambition to develop the collective understanding of PFPS.

3.0 – METHODOLOGY 3.1 - INTRODUCTION Bowling (2009) believed that the role of scientific research is to minimise the influence of external factors when gathering data. If this statement is true then the methodology is the process in which the researcher begins to understand and minimise these variables. This chapter will explore and contrast the methodological elements of thought and practice involved in research (Jesson et al 2011, Ridley 2008). The following chapter will address: research paradigms, qualitative and quantitative methods, search terms, databases, analytical tools and the role of ethics in research. Then in summation of all these considerations: the philosophical, methodological and ethical locus which this project will abide to.

3.2 - RESEARCH PARADIGMS A research paradigm is the fundamental philosophy which itself cannot be corroborated, yet is used as a foundation to develop theories and beliefs (Saks & Allsop 2012, Hart 2001, Guba and Lincoln 1994). Paradigms define the researcher’s perception of reality and their relationship to it; therefore how the studied phenomenon will be analysed (Polgar & Thomas 2013). There are three essential branches of philosophy which are used to define each paradigm; these are explained in table 3.1. 3.1. – Branches of Philosophy Branches of Philosophy Ontology

Explanation The branch of philosophy which concerns itself with the defining and questioning of reality. It seeks to define humanity’s perception of its environment; is the concept of real an absolute, or is reality dependent on the individual? (Polgar & Thomas 2013). The philosophy which questions the legitimacy, origins and essence of knowledge. Whether or not there is objective truth to research, or if all knowledge is subjective patterns of indiscernible phenomena (Polgar & Thomas 2013). The methodology is how can an understanding of reality be gained and understood. It is this philosophy which is most central to modern day research, and this review (Polgar & Thomas 2013, Ross 2012, Saks & Allsop 2012, Atkinson 2011).

Epistemology

Methodology

These three thought categories define each of the paradigms of research philosophy. These paradigms will be used in interpreting, synthesising and analysing the collected data (Polgar & Thomas 2013, Saks & Allsop 2012, Guba and Lincoln 1994). The four main paradigms are positivism, post-positivism, constructivism and critical theory.

3.2.1 - Positivism A positivist model states the environment is objective, reality is deterministic and thought is based on cause and effect (Hickson 2013, Hart 2001). The ontology is labelled as realism or foundationalism, due to the rejection of subjectivity when analysing data (Guba and Lincoln 1994). The model is concerned with what can be measured and any phenomena which cannot; does not exist (Saks & Allsop 2012, Guba and Lincoln 1994). Positivism states science is to observe and record what is seen, in order to relinquish experience and culture to record the phenomena with no discernible bias (Polgar & Thomas 2013). The epistemological perspective is objectivist; they are a non-interactive observer of the subject, one which records but does not alter. Positivism is interested in empirical data, which is paramount for understanding the static truths of the world (Bowling 2009). A positivist scientific method would be the development of a hypothesis which is confirmed with evidence. Positivism uses quantitative methods to allow for objective empirical data to be gathered and generalised, using research tools suitable for replication (Polgar and Thomas 2013).

3.2.2 - POST-POSITIVISM Post-positivism was modification to positivism which occurred during the second half of the 20th century (Letourneau & Allen 1999). Post positivism states that the phenomenon which is being studied is altered by its environment and the actions of the researcher (Bowling 2009). Which makes the ability to independently record data without introducing an effect impossible (Zammito 2004). Post-positivistic methods counteract this influence by using multiple designs and researchers in one subject area before the conclusions are made (Cody 2011). Post-positivism follows critical realism; that reality is objective, 17

but due to the fallibility of the observer reality can only be approximated rather than confirmed (Polgar and Thomas 2013). Therefore their epistemological perspective is modified objectivist. The post-positivist approach to methodology is to focus on falsifying the theory rather than proving it (Zammito 2004, Letourneau and Allen 1999).

3.2.3 - CONSTRUCTIVISM Constructivism is a theory that states truth and knowledge are generated through the individualâ&#x20AC;&#x2122;s past experiences and cultural nuances and that reality is actively synthesised and completely dependent on our choices and perceptions of our surroundings (Garner et al 2009). The ontological perspective of constructivism is relativism and the epistemological view is subjectivist, this means truth is only a concept of the human mind, one which has no independent universal criteria (Polgar and Thomas 2013). The constructivist interpretation of science was a contradiction to the positivist approach it superseded, as it redefined science as being subjective (Garner et al 2009). The methodological stance is hermeneutic, as the significance of the findings are believed dependent on the context and culture of the surroundings and participants (Polgar and Thomas 2013).

3.2.4 - CRITICAL THEORY Critical theory believes in realism in that truths can be achieved through studying the environment (Holloway and Wheeler 2013). However there is a subjective relationship between the researcher and the empirical evidence as it is believed to be mirroring their own orientation (Garner et al 2009). Habermas (1974) States that critical theory is a combination of the subjectivity and realism.

3.3 - RESEARCH METHODS There are two main contrasting methods in data collection; qualitative and quantitative. Both of which have characteristic advantages and disadvantages (Pope et al 2007).

3.3.1 â&#x20AC;&#x201C; QUALITATIVE RESEARCH Qualitative research is founded on constructivism, it is the social exploration of how individuals contextualise the world they live in (Holloway & Wheeler 2013, Guba and Lincoln 1994). The objective in qualitative data is to expose the behaviour, emotions or interpretations of groups, individuals and cultures on the researched phenomena (Petty et al 2012, Pope et al 2007, Pope & Mays 1995). Researchers will commonly use narrative, open questions, inclusive language and case studies as tools to achieve a more valid representation of their subject (Creswell 2013). Its advantages and disadvantages are displayed in table 3.1.

Table 3.1 - Qualitative Methods Advantages Ability to explore complex issues through communication and context. Data gathered has high internal validity

Disadvantages Usually expensive and time consuming. Difficult to synthesise data. Usually a smaller group is studied Data is difficult to generalise due to the effect of culture and individuality. (Saks & Allsop 2012)

3.3.2 QUANTITATIVE RESEARCH Quantitative research is founded on positivism; it is the systematic pursuit of replicable empirical data which can be analysed to represent the studied phenomena (Polgar & Thomas 2013, Burns & Grove 2010). The shared objective is to reveal the cause and effect relationships between previously undifferentiated variables (Sale et al 2002). Quantitative researcher commonly use laboratory conditioned studies (see 3.4) to systematically analyse empirical data, allowing for relatively little subjectivity and confounding variables (Creswell 2013, Burns & Grove 2010). Its advantages and disadvantages are displayed in table 3.2.

Table 3.2 - Quantitative Methods Advantages Usually less expensive and labour intensive than qualitative study methods The data gathered can be more easily generalised when using random sample populations, and repeating in different cultures Provides objective empirical data which can be easily synthesised and compared. Laboratory conditions can allow for a higher control of confounding variables and therefore a further isolation of the cause and effect relationship being studied. The influence of researcher bias is less due to the objectivity of numerical data Larger populations can be studied.

Disadvantages Weak internal validity as the artificial setting may induce Hawthorne effect which has been known to alter results in certain studies (Pope & Mays 1995. Restricted in exploration of complex topics (Bowling 2009) Less availability for researcher/participant communication

(Saks & Allsop 2012)

Both qualitative and quantitative methods should be considered when studying PFPS, as the emotive aspects in PFPS can be as influential on the pain as any mechanical factor (see 2.6) (Crossley 2014, Sanchis-Alfonso 2014). However the use of qualitative data would be inappropriate in studying a measurement as sensitive as neuromuscular activation; so a quantitative approach shall be used. Manual therapy research is predominately focused on quantitative research which has aided in the development of treatment plans, therapies, mechanisms and rehabilitation strategies (Petty et al 2012). It is important however to emphasise that although this review follows quantitative methods, future research could explore the psychological implications behind injury and qualitative aspects to living with the pathology (see 7.5).

3.4 - RESEARCH DESIGN The research design is the practical expression of the philosophical paradigms and methods (Cresswell 2013). It is the scaffolding on which the researcher can assemble their data and discussions in order to explore the subject area (DePoy & Gitlin 2011). There are two main types of research; primary or secondary: 1. Primary data collection represents the discovery of common themes and associations by directly studying the phenomena. 2. Secondary data is the development of conclusions through the collaboration of the current evidence (Aveyard 2010). The type selected is influenced by the researcher and the current state of knowledge in the area, whether there is already a large quantity of data out there, or whether it is an entirely unexplored subject (Jacobsen 2011). Both have their own merits in the scientific community (DePoy & Gitlin 2011).

3.4.1 - HIERARCHY OF EVIDENCE In the medical research community some research designs are valued more highly than others; this is due to the quality of the design in terms of rigorous controls, elimination of bias and replicability (Creswell 2013, Jesson et al 2011). The designs which are more susceptible to bias are at the bottom of the table and as the list goes up the designs increase in rigour. Although this is a useful tool to quickly evaluate research in health care, the hierarchy is not a universal concept (Creswell 2013, Aveyard 2010). As the value of research design is often relative to the question, and subject area. For example in this review, and others seeking aetiological factors of a pathology the highest quality designs would be large prospective cohort studies and the systematic reviews including them (Witvrouw et al 2014, Barton et al 2013, Powers et al 2012, Davis & Powers 2010). This review will be aimed at collecting highly controlled case-controlled, and cohort studies to determine associations (See table 4.3).

Table 3.3 â&#x20AC;&#x201C; Research Designs Research Design Systematic reviews and Meta-analyses

Randomised Control Trials (RCTs)

Cohort Studies and Case-controlled Studies

Description

A systematic review utilises a strict protocol to systematically analyse the medical literature. The review will typically use rigorous methods, critical appraisal, and comprehensive searching tools over multiple databases. Then by selecting data through standardised criteria a predefined question will be answered (Ross 2012, Aveyard 2010, Bowling 2009, Egger et al 2008). A Meta-analysis is a tool used in systematic reviews, specifically the combination of relevant results from a multitude of studies. Therefore, every meta-analysis should be contextualised through a systematic review, but not every systematic review necessarily contains a meta-analysis (Hickson 2013, Polgar & Thomas 2013). RCTs are a type of clinical trial. Participants are randomly distributed into groups, then typically the intervention is given to one group and nothing or a control is given to the others. After a predetermined timeframe the differences are compared. The results can then be associated with the corresponding variable given to that particular group (Polgar & Thomas 2013, Aveyard 2010). RCTs are the favoured design in primary data collection and thought to be the â&#x20AC;&#x2DC;gold standardâ&#x20AC;&#x2122; in judging the efficacy of a medical intervention (Polgar & Thomas 2013, Jacobsen 2011, Bowling 2009). A cohort study can be retrospective or prospective. Both observe a population who have adopted a specific lifestyle or exposed to a unique factor. The population is then followed up to determine any changes which could be attributed to the given variable (Aveyard 2010, Jacobsen 2011). A retrospective study looks in the past to identify causality through shared factors in relation to the populations common characteristic, in this case; patellofemoral pain. A prospective study looks at a healthy cohort or one which share a potential predisposing factor and then following them to assess how many develop the studied phenomena (Polgar & Thomas 2013, Jacobsen 2011, Dykacz 2005)

Type of Data Secondary

Primary

A case control study observes a sample which already possesses a specific condition and then contrasts them with a control population in order to deduce the cause through association (Polgar & Thomas 2013, Bowling 2009). They are usually cross-sectional, which means the data does not involve progression like the cohort studies. Both cohort and case-controlled studies are observational. These designs are most frequently used when an RCT is not possible or applicable (Polgar & Thomas 2013).

Surveys

Cross-sectional designs are reliable for determining associations (Polgar & Thomas 2013, Saks & Allsop 2012, Ross 2012). However prospective research is needed for establishing cause and effect (Polit & Beck 2013, Saks & Allsop 2012, Atkinson 2011). A survey is a versatile design used to gain a wide overview of a particular population. It possesses little control over variables but can reach a large sample with relatively little expense (Polgar & Thomas 2013). They can be used in conjunction with other designs to gather either quantitative or qualitative data. They are often rejected due to the fact they often have a low response rate due to subjects getting bored and the inability to explain any unclear questions (Bowling 2009). 22

Primary

Case Reports

Qualitative Studies Expert Opinion

Anecdotal Opinion

A case report is thorough summary of a specific case. Documenting its intricacies so as to understand all variables, aetiology and sequlae to a certain condition. These often involve interviews, reviews of medical records, observations, and personal written accounts (Garner et al 2009). It is impossible to generalise the data, as although it is thorough and focused it is only exploring one case rather than a large sample of subjects from different cultures (Garner et al 2009). Qualitative studies do not aim to quantify phenomena into statistics but rather explore themes within it. This is typically done through interviews, observations and focus groups. This data then tends to be subjective and interpretative (Hickson 2013, Polgar & Thomas 2013). Although experts are often the most informed and experienced in the field it can be the case that some expert opinions may not be substantiated by evidence (Aveyard 2010). The fallibility of opinion is due to uncontrollable external factors, preferences and biases which highlight why objective data commonly preferred (Bowling 2009). Anecdotal opinion is of little to no use in medical research, due to the fact it is often not substantiated by evidence (Hickson 2013, Aveyard 2010).

Primary

Secondary Primary

Primary

3.4.2 - THE LITERATURE REVIEW Systematic reviews are a type of literature review and as seen in table 3.3, are the highest in the Cochrane Collaboration hierarchy of evidence (Saks & Allsop 2012, Aveyard 2010, Egger et al 2008). Each type of literature review has characteristic specifications and a deemed quality within the research community (Aveyard 2010). A literature review is a comprehensive summation of data and research relevant to a particular subject. It involves the process of searching and analysing data in order to answer a specified research question (Saks & Allsop 2012, Garner et al 2009). Literature reviews are essential for health care professionals such as osteopaths who have an interest in delivering evidence based care, yet a limited time to search and synthesis research independently (Buetow 2007, General Osteopathic Council 2006). A review can present the current understanding on a clinically relevant subject in a way which is easily digestible and implementable for the health care professional (Ross 2012, Aveyard 2010, Egger et al 2008). In a systematic review the research method is methodical and bound by a strict process of selection and synthesis to minimise the influences of bias and random errors giving it the highest quality of the literature reviews (Machi & Envoy 2012, Ross 2012, Aveyard 2010, Hart 2001). Through displaying the method and references the research becomes open for analysis to support or dispute by the research community (Hickson 2013 Bowling 2009). If the literature review does not follow a systematic process it is labelled as a narrative review (Aveyard 2010, Hart 2001). Bowling (2009) stated that the narrative review is limited by its subjectivity in its inclusion and analysis of the literature. To complete a Cochrane systematic review requires; a high budget, an experienced team of researchers and a long time frame (Saks & Allsop 2012, Aveyard 2010). Therefore due to budget, equipment and task force available to an undergraduate researcher this design is unavailable. Instead this literature review will be executed with a systematic approach. This approach may not fulfil the requirements of the Cochrane Collaborationâ&#x20AC;&#x2122;s standard of a systematic review. However by following their standards to the extent of the available resources, the final review will still be a systematic presentation and analysis of the current research. 24

3.5 - LITERATURE SEARCH The search process is a crucial point in any literature review, the credibility of the conclusions made is directly dependent on the amount of rigour used in study selection (Hickson 2013, Machi & McEvoy 2012). The validity is determined by the use of exhaustive sampling, strict predefined inclusion and exclusion criteria and unbiased statistical analysis of the data acquired (Egger et al 2008, Aveyard, 2010, Saks & Allsop, 2012). This section explores the development and use of keywords in a literature search, the inclusion/exclusion criteria, the ‘snowballing’ technique and ‘grey literature’ and how they can be used in literature searching. Saks & Allsop (2012) highlighted the importance of organisation and documentation of the search terms and process, so that the findings can be verified through replication. This includes the set predefinition of the inclusion and exclusion criteria, which needs to be formed meticulously from the review question and preliminary literature review (Hickson 2013, Saks& Allsop 2012, Pope et al 2007). The databases used are often the first indicative factor for the quality of evidence used (Machi & McEvoy 2012). The use of electronic databases is the most efficient and time effective method of gathering relevant data (Egger et al, 2008; DePoy & Gitlin 2011). Saks & Allsop (2012) suggest that by utilising specialist journals and databases it is easier to gather relevant studies. Unfortunately due to osteopathy’s developing nature there is, as of yet, no specialist database. However as this reviews’ question is not specifically osteopathic the medical databases provide a sufficient service in the searching process.

3.5.1 - KEYWORDS Keywords must be broad enough to include all the applicable studies, and specific enough to exclude any irrelevant research, they are characterised by their close association to the research question (Bowling 2009, Hart 2001). Khan et al (2011) states that both free text terms and controlled terms need to be utilised to search databases effectively. Free text terms are gathered through use of a thesaurus and mind maps to collect and develop synonyms which can be used as keywords (Machi & 25

McEvoy 2012, Aveyard 2010). Controlled terms are pre-set keywords which are used as tags in the logging of citations, on MEDLINE and PubMed these are referred to as medical subject headings or MeSH terms. By using a combination of both free text and controlled terms the search becomes more sensitive to the research question (Khan et al 2011). Hickson (2013) states the advantages to using ‘PICO’ in the development of the question and search terms. PICO is a mnemonic which aids the development and specification of a question or terms in a systematic process, it can aid the researcher in refining their topic and in guiding their progress (Cook & West 2012). Schardt (2007) found searches utilising PICO retrieved a higher percentage of relevant citations than a control search. P =Problem/Patient; what is the issue in question which is being explored. I = Intervention/Exposure; what’s the intervention used or the factor used to test the patient. C = Comparison; what is the comparative factor, used as a reference. O = Outcome; what is the desired outcome. As this is a review focused on the aetiological associations of PFPS the PICO mnemonic cannot be used in its entirety, as the ‘outcome’ is not relevant, however it is helpful as seen above in developing new search terms. These key terms are used again in the following search tools (3.5.2) and development and execution of the inclusion and exclusion criteria (See 3.5.3 and 4.3).

3.5.2 - TRUNCATIONS, WILDCARDS AND BOOLEAN OPERATORS. Truncations are used to broaden the search by accommodating for alternative pluralities or suffixes (Ross 2012, Khan et al 2011). They will use either an asterisk or a dollar symbol at the end of the shared spelling to include all variations, for example “Glute* musc*” would include; gluteal musculature, gluteal muscles, gluteus muscle (Khan et al 2011, Aveyard 2010). Wildcards are similar to truncations in that they are used within the key terms to broaden the search; however wildcards replace one character usually in the middle of the term. Using wildcards can be very helpful in accommodating for Americanisations; the question mark is used as its symbol, for example: “sensiti?ation” will include sensitization and sensitisation (Ross 2012). 26

Boolean operators are used in conjunction with the truncations, key terms and wildcards to both broaden and specify the search depending on what is needed (Aveyard 2010, Bowling 2009). “AND”, “OR” and “NOT” are used to modify the input of the keywords and truncations. By placing “AND”

between search terms the literature gathered will be of a higher specificity as the results will include both terms which were linked by the operator (Hart 2001). “OR” will be used to broaden the search, as results will be identified on the account of it including just one of the terms linked with the operator rather than both. The exclusion process can also be reduced to one key term to narrow the literature search in specified direction. This allows for a more dynamic exclusion process, in this example “NOT” is used to prevent any result with the key term from being gathered (Machi & McEvoy 2012, Aveyard 2010).

3.5.3 - INCLUSION EXCLUSION CRITERIA The inclusion/exclusion criteria are used to filter the results of the literature search to ensure that the data found is of a credible and relevant nature (Ross 2012, Aveyard 2010). The main factors considered are normally the study type, date published and language. For example a review’s inclusion criteria might be only randomised control trials published in English in the last 5 years (Polgar & Thomas 2013). By specifying these criteria it allows for the researcher to refine the data collected in an unbiased predefined and systematic approach (Ross 2012). Hickson (2013) highlights how important it is to specify the target patient group, for example their age, the duration of their condition, the severity, their environment, social status, daily activities and any other ailments. However inclusion exclusion criteria which are too strict can restrict the data and produce a limited view of the studied phenomena (Ross 2012). In this review the criteria were of paramount importance in distinguishing the sample studied. As PFPS is so widely misunderstood there will be a high variability in the severity and conditions contained within the evidence. This variability is a central limitation in the development of secondary data in this area (see 7.6).

3.5.4 - SNOWBALL SAMPLING Snowball sampling is the process of discovering literature through the references of relevant studies and reviews, and can be used to identify high quality evidence that would usually be hard to find (Chapman et al 2010, Greenhalgh & Peacock 2005). Journals which have a large amount of relevant literature can be further scrutinised using this technique (Aveyard 2010). Greenhalgh & Peacock (2005) found that this method produced the highest yield compared to other searching strategies used. However if the reference used is biased itself then the references found may add to its biased view of the area (Horsley et al 2011). However this bias can be counteracted if the literature search is an exhaustive process which utilises a variety of other search strategies (Aveyard 2010, Greenhalgh & Peacock 2005).

3.5.5 - GREY LITERATURE Bowling (2009) states the importance of including unpublished or â&#x20AC;&#x2DC;greyâ&#x20AC;&#x2122; literature in the search; this is due to the reluctance of journals to publish inconclusive or contradictory research. This often results in the misrepresentation of the studied phenomenon through publication bias. The issue of publication bias in the skewing of results is often overlooked. John et al (2012) states that this misconduct is prevalent across the whole scientific community and is ultimately damaging the academic enterprise. However it is often more challenging to find grey literature, as it is not represented by many databases (Ross 2012). Unpublished health literature can be found in conference summaries, dissertations, and unpublished reports and studies. This is often where the most up to date data is found, as the researchers are not limited to the trials involved in publication (Hickson 2013). This literature can be found using Internet searches, however it must be noted that this method does not hold the same rigour as database searching and cannot be controlled and recorded in the same manner (Saks & Allsop 2012).

3.6 - DATABASES Almost all research which has been published in journals are indexed in electronic databases, these can be searched in order to gather relevant data (Ross 2012). The two types of databases are primary and secondary. The two types are named after the research they contain. Primary databases index original studies from a variety of journals, whereas secondary will contain systematic reviews, guidelines and summaries of previously collected data (Hickson 2013). Database bias should be considered when selecting studies, whether it is bias in what is published or the use of colloquialisms or American spellings in the search terms. The table below highlights databases which were searched (Polgar & Thomas 2013, Hart 2001).

3.7 - CRITICAL APPRAISAL Once the selected research has been filtered through the inclusion exclusion criteria the next stage is to assess each study for its quality through an in-depth critical appraisal (Khan et al 2011). A critical appraisal tool (CAT) can give a standardised structure to the evaluation of research, this can aid the researcher in the efficiency of their appraisal and in avoiding bias (Aveyard 2010, Ciliska et al 2008). The selected CAT will structure how the researcher evaluates the evidence, therefore has the potential to influence how the data is interpreted (Katrak et al 2004). Therefore the use of a CAT is not always necessary for data analysis, as it can hinder the quality of the analysis (Crowe & Sheppard 2011). Aveyard (2010) however strongly recommends the use of a carefully selected CAT to aid the amateur researcher. Across the literature emphasis is placed on choosing the CAT tailored to the research (Crowe & Sheppard 2011, Aveyard 2010, Katrak et al 2004). Machi and McEvoy (2012) suggest that when the researcher begins the analysis they should be consciously opposed to thoughts of selective bias. Sanderson et al (2007) conclude that a good CAT has four attributes: 1. A small amount of key factors. 2. It is specified to the given research i.e. study design and topic area. 3. Utilises a checklist, rather than a scale. 4. Gives reference to its validity, reliability and development.

Table 3.4 â&#x20AC;&#x201C; Critical Appraisal Tools Appraisal Tool Discussion Critical Appraisal Skill CASP is a non-profit based in Oxford which develops CATs and workshops Program (CASP) (2013). to aid researchers in evaluating a variety of studies and reviews in a systematic manner (Aveyard 2010). They provide eight CATs which are free to use, each of them specified to their own design; Systematic Reviews, Randomised Controlled Trials, Cohort Studies, Case Control Studies, Economic Evaluations, Diagnostic Studies, Qualitative studies and Clinical Prediction Rule (Smith et al 2009). McMaster (Law et al 1998a & The McMaster CAT was developed by Occupational Therapy Evidence Based Law et al 1998b) Practice Research Group at the McMaster University (Law et al 1998b). The CAT includes 8 sections which focus on different aspects of the research with simple guidelines. The tools are not individualised to the study design like CASP CATs, this means the criteria is not as defined. Only qualitative and quantitative options are available (Law et al 1998a & Law et al 1998b). An advantage to this CAT is that a supplementary guideline of use is given (Law et al 1998b). Although Aveyard (2010) encourages the use of design specified CATs, Law et al (1998b) provides sufficient support through the guidelines to counteract any reduction in validity. The Downs and Black This CAT is a checklist used for both randomised and non-randomised Checklist (Downs & Black research. Its internal validity and reliability is good, however its external 1998) validity is poor. It is a 24 item checklist which is easily modifiable and thorough (Downs & Black 1998). However Barton et al (2013) identified methodological limitations to their modified checklist in their review of gluteal muscle activation in a patellofemoral pain group. Finding difficulties in the matching of participant characteristics and addressing confounding variables. The Transparent Reporting The TREND statement is a CAT specified for quantitative non-randomised of Evaluations with controlled trials, such as in this case case-controlled studies. Des Jarlais et al Nonrandomized Designs (2004) developed the 22 item checklist to aid in standardising evaluation of (TREND) Statement research and the clinical development of practice through evidence based recommendations. However this CAT lends itself to more established researchers due to its exploratory nature (Des Jarlais et al 2004).

In this review the CASP (2013) CAT will be used as it is recommended by Aveyard (2010) and fulfils Sanderson et al (2007) criteria for an effective tool. Unlike the McMasters tool, CASP quantifies the quality of each paper allowing for a more objective evaluation and comparison.

3.7.1 - ELECTROMYOGRAPHIC CRITICAL APPRAISAL The recording of the data or more specifically the standard of EMG which will be used to address NMA must be evaluated by addressing the studies' method. During the preliminary literature review a search for specified EMG CAT was conducted; but nothing was found. Therefore a CAT was developed through the comparison of the recommendations from; the European standards for surface EMG

(SENIAM) (Stegeman & Hermens 2007, Day 2002). The check list was developed so that the studies could be scored and their methods readily compared in order to corroborate or discredit their findings relative to each other.

3.8 - EVIDENCE ANALYSIS The analysis allows the researcher to quantify the findings of each of the studies to develop an overall conclusion (Polgar & Thomas 2013, Ross 2012). The analysis of current research is central to evidence based practice, without it there can be no consensus and subsequently no clinical advances (Hickson 2013). The analysis serves to identify patterns, assess validity and bias and highlight anomalous data. There are other types of analysis such as meta-ethnography and meta-syntheses but these are typically used for a team research project in qualitative analysis so will not be considered further (Hickson 2013, Aveyard 2010, Hart 2001). The following section describes the two relevant types of analysis with their strengths and limitations;

3.8.1 - THEMATIC ANALYSIS The purpose of this analysis is to distinguish patterns emerging in the literature (Fereday & Muir-Cochrane 2008). These patterns are synthesised into codes and eventually themes which can link the conclusions of the data together (Aveyard 2010). Ideally this process of coding and theme selection should be undergone by two independent researchers to exclude bias and ensure methodological rigour (Hickson 2013, Aveyard 2010, Bowling 2009). Although this type of analysis is most suitable for qualitative data it is commonly used in quantitative (Fereday & Muir-Cochrane 2008). This is to explore the subjectâ&#x20AC;&#x2122;s conclusions and explore the developments in theory and findings (Bowling 2009). Rourke & Anderson (2004) call this form quantitative content analysis. Due to its nature, and simplicity it is recommended for novice researchers to aid in understanding the collective research and developing conclusions (Aveyard 2010).

3.8.2 - META ANALYSIS This method is a thorough breakdown and analysis of the data; it is typically used for assessing quantitative research. It is the combining of the statistical results from a variety of studies using similar methods and design to reanalyse the potential for a collective trend or conclusion. This method allows patterns to emerge that would be inconceivable in smaller trials alone (Hickson 2013, Aveyard 2010, Bowling 2009). However a meta-analysis usually requires a large amount of time and a team of researchers to complete (Aveyard 2010).

3.8.3 - CONCLUSION A meta-analysis of the data would be preferable in this review due to the tendency of PFPS studies to show relatively small sample sizes (Barton et al 2013, Selfe et al 2013). However due to the restrictions in time, finance, facilities and experience of the researcher; a systematic thematic analysis will be used instead. This method will allow exploration of the selected studies in order to develop conclusions and evaluate methodological validity.

3.9 - ETHICAL CONSIDERATIONS Atkinson (2011) states that ethics in research seek to protect participants from harm and ensure the research serves the interests of society. The first of these statements is termed as â&#x20AC;&#x153;nonmaleficenceâ&#x20AC;?, one of the core principles of medicine, research and osteopathy (General Osteopathic Council 2012, Emanuel et al 2011, Parsons & Marcer 2006, Hart 2001, Gillon 1985). A literature review is not bound to the same scrutiny that a primary study is, as it doesnâ&#x20AC;&#x2122;t need to be approved by a research ethics committee (Ross 2012, Emanuel et al 2011, Aveyard 2010). However there are still a number of ethical considerations which must be acknowledged when conducting a review, these considerations tend to be in the selection of studies. For manual health practitioners the only way that studies can be clinically relevant is through the analysis of human subjects. Whenever humans are used in studies it is vital to consider and maintain respect for their

safety and rights (Emanuel et al 2011). Lewith et al (2010) states that rigour is a fundamental factor to consider when producing ethically grounded research. Rigour refers to the elimination of bias, plagiarism and the respect of authorship. Bias can occur in many different forms, all of which can ultimately affect the validity of the research (Pannucci & Wilkins 2010). The main one which can affect a review is publisher bias, as discussed earlier in 3.5.5 this can be countered by the inclusion of grey literature. Aveyard (2010) adds that tools used, whether a theoretical framework or critical appraisal tool, can induce bias through guiding the researchers analysis of the data (Katrak et al 2004). It is the researcherâ&#x20AC;&#x2122;s obligation to minimise all biases to the best of their standard (Machi and McEvoy 2012). Plagiarism is the process of claiming false ownership of ideas or published material through not crediting the original author (Aveyard 2010). The respect for authorship is vital in the research community, for an author to publish their work they must commit to a long and challenging process of evaluation and refinement. Their progression in the research community is based upon how their work is acknowledged (Hickson 2013). Therefore it is important to fully credit any authors whose research has been quoted or paraphrased in the review(Ross 2012). It is imperative that all authors involved are named in the references in the correct format (Hickson 2013, Lewith et al 2010). In this review the Harvard author-date system is used, dictated by the Oxford Brookes referencing guide for health and social care students (Harper et al 2011).

3.10 - CONCLUSION To conclude, this review will analyse the current research in PFPS and GMed NMA in females. This body of research will take the form of a quantitative literature review completed in a systematic manner. The research will be explored through a post-positivist approach, utilising systematic searching of multiple databases while implementing a strict inclusion exclusion criterion for data selection. The inclusion/exclusion criteria and search terms have been specified and altered in respect of the preliminary literature review, thus ensuring the evidence is current and the analysis is relevant. After

the studies have been acquired the CASP (2013) CAT will be used to assess their quality and a thematic analysis completed to highlight themes and conclusions. The proposed question could have been answered through primary research, however due to the small sample sizes of PFPS research, the inaccessibility required equipment and the relative inexperience of the researcher; a literature review was selected instead. Hart (2001) concluded that literature reviews are an integral part of research, used to inform both researcher and reader and in the progressive narrowing and refining of the topic and gathered evidence. Although the review doesnâ&#x20AC;&#x2122;t need ethical approval it is vital as a member of the research community that studies included were executed in an ethical manner.

4.0 - METHOD 4.1 - INTRODUCTION This section will cover the method of which this review was completed. It will describe the complete search process, the inclusion exclusion criteria used, the studies included, and how they were appraised and analysed.

4.2 - DATA COLLECTION All data used in this review was collected through the systematic searching of online databases up to October 2014, predominately accessed via the Oxford Brookes University Shibboleth service (Oxford Brookes University 2011, Sinnott et al 2006). Initially while the question was still being formed a preliminary literature review was completed in order to develop the key terms and Boolean tools, and eventually refine the final search terms. These will be shown in figures 4.1, 4.2 and 4.3

4.2.1 - ONLINE DATABASES Table 4.1 - Databases Database Description BioMed Central A sub database of BioMed Central. BMCM is a specialised medical database which Medicine contains a variety of musculoskeletal journals which can be accessed freely online (BioMed Central 2014). Cochrane Evidence based health care focusing on the development and distribution of systematic Collaboration reviews (See 3.4) (The Cochrane Collaboration 2014) CINAHL Cumulative Index to Nursing and Health Literature (EBSCO 2014) PEDro Physiotherapy evidence database which includes a variety of reviews, studies and practice guidelines (Physiotherapy Evidence Database 2014) PubMed PubMed indexes citations from a vast array of biomedical literature covering MEDLINE, life science journals and online books (NCBI 2014). ScienceDirect ScienceDirect contains a large directory of scientific, technical and medical journals.(Elsevier 2014) Google Scholar Google scholar is a free online search index for international literature across a variety of disciplines (Google Scholar 2014). Web of Web of Knowledge is one of the largest online citation indexes, which includes a variety Knowledge of databases across many subject areas (Reuters 2008).

These databases (shown in table 4.1) were chosen due to their relevance to musculoskeletal disorders, simplicity of use, reputation in the health community and through recommendation of Oxford Brookes University (Oxford Brookes University 2011). It is advantageous to use multiple citation indexes, even if they are covering similar databases. For example both Web of Knowledge and Pubmed include MEDLINE in their search, but may provide

different results, due to their search process. For example Kulkarni (2009) found that different results were obtained when using the same search terms in both Google Scholar and Web of Knowledge. By utilising a range of sources a detailed and thorough search can be achieved (Machi & McEvoy 2012, Saks & Allsop 2012, Aveyard, 2010).

4.2.2 - KEY TERMS The search was first used in the preliminary literature search (chapter 2.0) and it was apparent during this stage that a thorough search is dependent on the use of the correct search terms (Machi & McEvoy 2012). These terms were developed from the utilisation of the PICO framework (See 3.5.1), in considering each aspect of the question in order to maximise the results, while keeping the content relevant (Aveyard 2010). Figure 4.1 shows the initial mind map used to consider synonyms to be included in the search.

Figure 4.1 - PICO Search Terms

Figure 4.2.2 - PICO framework in the development of search terms

4.2.3 - SEARCH TOOLS After using the PICO framework to gather search terms these need to be converted into a precise search which utilising the tools explained in section 3.5.2. The first tool to be utilised on the key terms were truncations and wildcards, these tools are used to broaden and specify the search and are illustrated in figure 4.2 (Hickson 2013). Figure 4.2 â&#x20AC;&#x201C; Search Tools

Figure 4.2 - Illustrating the use of truncations and wildcards in the literature search (See section 3.5.2 for definitions and justifications of these search tools.)

After the synthesis of the key terms seen in figure 4.1 and the identification of the wildcard and truncations appropriate (figure 4.2), the next stage is to use Boolean tools to construct the search. The final Boolean search is illustrated in figure 4.3 Figure 4.3 - Boolean Search

Boolean search including key terms and truncations used

Some minor alterations were made to the truncations and Boolean tools during each search according to the nuances of the database, for example whether the asterisk or dollar sign was used, or the use of quotation marks to separate a phrase from singular terms. This may affect the replicability of this search, especially if the replicator is unaccustomed to the databasesâ&#x20AC;&#x2122; individual variations; however the effectiveness of data collection is the priority. It must be noted that the key terms were constant throughout. The search criteria were conducted across the chosen databases to test its reliability. The results are shown in table 4.2

Table 4.2 â&#x20AC;&#x201C; Preliminary Search Results Databases BioMed Central Medicine Cochrane Collaboration CINAHL PEDro PubMed ScienceDirect Google Scholar Web of Knowledge

Number of Results 3 0 24 0 34 105 574 63

4.2.4 - INCLUSION AND EXCLUSION CRITERIA Once the search process had been completed study selection begins, this starts with filtering through the search results using the eligibility criteria to assess the abstracts of the papers. To counteract the variability described in 3.5.3 and 7.6 the criterion was developed from the findings of the preliminary literature review. The Petersen et al (2013) definition of PFPS defined in section 2.3.1 will be used in sample selection. This means that studies including chondromalacia patella and plica involvement will be excluded.

Table 4.3 â&#x20AC;&#x201C; Inclusion and Exclusion Criteria Inclusion Criteria Studies published within the last 10 years

Female participants only, or result differentiation between genders Studies which record electromyographic data of gluteus medius

Participants aged between 18-45

Studies which are of the following designs were included: Case-control, cohort studies or RCTs. Studies published in English

Studies which describe subjects having patellofemoral pain syndrome, patellofemoral pain or anterior knee pain of an insidious onset.

Studies which use a control group of healthy females, matched on; age, height and weight

Exclusion Criteria

Justification

As the understanding of patellofemoral pain has developed dramatically in the last decade itâ&#x20AC;&#x2122;s important that the studies gathered respect these developments. (Hryvniak et al 2014, Aveyard 2010). Otherwise the Studies published before 2004. conclusions drawn may be of little relevance to the understanding and treatment of PFPS. Secondly, due to the amount of time and resources a complete review of all data would require. The focus on the female population is due to the disproportionate Male only subjects or mixed gender groups with occurrence between sexes which is explored in section 2.5 (Hryvniak et al shared undifferentiated results and subject 2014, Petersen et al 2013, Boling et al 2010). If the gender's results are criteria. mixed it will reduce the overall specificity, and could ignore some of the underlying factors which cause the gender bias. EMG is the variable used to distinguish differences in neuromuscular Studies which do not record EMG activity of activation (Criswell 2010). Ferrari et al (2014) found EMG to be an effective gluteus medius. tool in the analysis of muscle activation in patients with PFP. By setting this age range the likelihood of alternative causes for the PFP is minimised. Younger participants are more likely to have undiagnosed Studies including subjects below the age of 18 and chondromalacia patellae, Osgood-Schlatters, Sinding-Larsen-Johansson above the age of 45 (Hryvniak et al 2014, Yen 2014, Petersen et al 2013). As the population gets older the effects of degenerative changes to the patellofemoral articulation could be causing the PFP (Karam et al 2012). It has all so been noted that age can affect EMG results, which cause misleading data trends (Criswell 2010). As stated in section 3.4.1 RCTs are not compatible with associative Studies using the following designs: questions, the most valuable are case-control and cohort studies. Other less Literature reviews, surveys, case reports, qualitative controlled designs were deemed to lack methodological rigour and validity studies or expert opinion. (Jesson et al 2011). More time and resources would be required to translate a text and check the translations reliability with another reference. This criterion is implemented Studies not published in English to prevent any mistranslation, which may result in breaches in the code of authorship (see ethics 3.9). Studies which include subjects previously diagnosed To effectively explore this review's question the inclusion of data must be with osteoarthritis, rheumatoid arthritis, specific and defined (Jesson et al 2011). By excluding other anterior knee chondromalacia patellae, Hoffaâ&#x20AC;&#x2122;s syndrome, Sinding-pathologies and underlying structural deformities the likelihood of gathering Larson Johansson, Osgood-Schlatters, iliotibial band representative data for PFPS is higher. As PFPS is not clearly defined in the syndrome, plica, bursitis, neuromas or other patella literature (see 2.3.1) it is difficult to acquire relevant data or verify the tendinopathies. In addition to if subjects have had diagnosis, this is discussed in further in 7.6 (Petersen et al 2013). knee surgery or experienced previous serious direct trauma to the knee. The control group gives a reference to the data found, as EMG methods vary between studies, comparison of results with other studies can significantly Studies with no control group reduce validity (Criswell 2010). By comparing against a matched control the effect of PFPS can be isolated (Bolgla et al 2010, Day 2002).

4.2.5 - SNOWBALLING The snowball sampling method was used on the reference lists of the studies already identified and relevant secondary data. Through the identification of a systematic review by Barton et al (2013) a further two studies were obtained for inclusion (Willson et al 2011, Saad et al 2011).

4.2.6 - GREY LITERATURE No unpublished literature could be identified which passed the inclusion and exclusion criteria, despite the use of internet searches and reading through dissertations. The analysis of PFPS is complex in the method and controls needed. The lack of valid unpublished research suggests that reason for the study to be unpublished is not publisher bias, but the lack of controls considered (Saks & Allsop 2012).

4.5 - CRITICAL APPRAISAL 4.5.1 - CASP CRITICAL APPRAISAL Each study was appraised fully in regard to each of the questions specified in the CASP CAT (2013), the results of these can be found in detail in section 5.3, and the full appraisal forms for each study is shown in the appendix C. The CASP (2013) checklists were used to quantify the methodological rigour and value of the data in the process of answering the reviewâ&#x20AC;&#x2122;s question. This tool was chosen as it adds more objectivity into the appraisal process than some of the other tools, and it is an established and recommended tool in the health care research community (Aveyard 2010, Bowling 2009). Each of the studies were analysed with the same checklist questions and by the same researcher. The checklist was scored as follows; if the question was answered yes a point was given, if the answer was not clear, no change in score, if the answer was no a point was subtracted. The summary checklist is seen in table 5.8. Both questions seven (what are the results of the study?) and eight (How precise are the results?) related to the results of the studies and their statistical precision, their answers are listed in

section 5.3.6 and illustrated in table 5.5. However in the checklist the two were scored as one, if the study seemed to of recorded the results in a controlled and systematic manner. Additionally question eleven (do the results of this study fit with other available evidence?), was omitted from the overall score due to the controversial and multifaceted aetiology of PFPS (see 1.1 & 1.3). If the results do not parallel the other research in the area it is likely due to the heterogeneity of the designs (Barton et al 2013). Therefore due to the rapidly changing methods and controls used in testing PFPS, this question was judged as premature; in the absence of a relevant data trend (Shirazi et al 2014, Bolgla 2010, Bolgla 2007).

4.5.2 - EMG CRITICAL APPRAISAL By using the guidelines recommended for EMG (SENIAM) a CAT was created to review each of the papers EMG method systematically (Stegeman & Hermens 2007, SENIAM 2006, Day 2002). The SENIAM standards are supported from recommendations by Bolgla et al (2010), Clarys et al (2010), Criswell (2010), and Bolgla et al (2007). The CAT was used by contrasting the EMG method of each study (see table 5.6) to the SENIAM recommendations (see appendix A & B), when the study was compliant with a methodological attribute they were given a point, when the attribute was not specified; no change, when the attribute differed; a point was subtracted. The results can be seen in table 5.7.

4.5.3 â&#x20AC;&#x201C; VALUATION OF EVIDENCE For ease of reference what this review classes as high, moderate and poor quality evidence is specified in table. The CASP score dictates the overall quality of the study. However when specifically comparing the variables found in each studies EMG method, the EMG score is referenced.

Table 4.4 - Valuing Evidence Value of Evidence CASP SCORE EMG SCORE High Quality 7-9 6-7 Moderate Quality 4-6 4-5 Poor Quality 0-3 0-3 Table â&#x20AC;&#x201C; Shows the relevance of the critical appraisal scores to the overall value of the evidence.

4.6 - ANALYSIS As stated in 3.8.3 a thematic analysis was chosen to explore the question. The data will then be grouped by theme and coded so that the results can be combined and compared (Bowling 2009). The thematic analysis will be driven by the methodological and philosophical decisions described in chapter 3 (Aveyard). The use of themes will allow the novice researcher to explore different associations in the results. For example if there is any activation characteristic which is specific to squatting but not jumping in females with PFPS. Fereday & Muir-Cochrane (2008) state that using a thematic analysis allows the researcher to easily explore trends in the collected data.

During the sections of analysis the summations of the evidence will be referenced to as strong, moderate and limited evidence. This reference is relative to the quality and quantity of data in this review. If three or more high quality studies conclude statistical significance in the absence of any conflicting data this would be regarded as â&#x20AC;&#x2DC;supported by strong evidenceâ&#x20AC;&#x2122;. However in comparison to the levels of significance required by the Cochrane collaboration or other large scale systematic reviews, this evidence would be relatively insignificant (Van der Heijden et al 2015, Ross 2012, Aveyard 2010). The grading of evidence is

shown in table 4.5.

Table 4.5 â&#x20AC;&#x201C; Grading of supporting evidence Evidence Grading Strong Evidence Moderate Evidence Limited Evidence

Quantity of Evidence Supported by 3< high quality papers or 4< moderate quality papers with no conflicting evidence of high or moderate quality. Supported by 2 high quality or three moderate quality studies with no conflicting evidence of high or moderate quality. 2< high quality studies with 1< moderate or 3< poor quality studies conflicting. Supported by 1< study of high or moderate quality in the absence of even amounts of conflicting evidence

5.0 - RESULTS 5.1 - INTRODUCTION This chapter will present the results of the search outlined in chapter 4. The search results and the studies which met the eligibility criteria are seen in section 5.2, the critical appraisals are explored in section 5.3, and finally the studies are summarised in table 5.9. Full details of the critical appraisals can be found in appendix C.

5.2 - SEARCH RESULTS The literature was searched for case-control or cohort studies which addressed EMG activity of gluteus medius in females with patellofemoral pain. Abstracts of 803 papers were analysed, 19 of these seemed eligible. The full texts of these 19 were located to determine any ambiguities in their abstracts. From these texts a further nine were excluded according to set the criteria, leaving 10 studies appropriate for analysis. The final nine were excluded due to ambiguities in their own PFPS criteria detailed in the abstract, when analysed they were found to include chondromalacia patella, previous cruciate surgeries and traumatic onsets, amongst others. Moher et al (2009) developed the PRISMA statement in order to improve the quality and rigour of systematic reviews, these guidelines aided the researcher to execute the inclusion and exclusion criteria systematically. The PRISMA search and selection is illustrated below in figure 5.1

Figure 5.1 - Flow Criteria

Figure 5.1 - Process of data identification, screening and inclusion taken from Moher et al (2009). The 10 studies chosen were all case-controlled laboratory studies using a cross-sectional design, which examined a variety of functional tasks. Although prospective research would have been preferable of all the search results, only cross-sectional studies passed the inclusion and exclusion criteria (see 3.4.1 & table 4.3). Some prospective case reports were found in the subject area however they were excluded due to the exclusion criteria (see table 4.3). The studies PFPS group ranged in size from 9-27 with a mean of 20, and the control group ranged from 8-27 with a mean of 17. Both groups were matched on age, height and weight. All sample details are shown in table 5.2. By addressing a range of functional tasks it is possible to gather a dynamic representation of PFPS, and how it can affect a range of different activities (Witvrouw et al 2014, Powers et al 2012).

There was no equalisation in the quantity of studies addressing each activity, but instead all studies of sufficient rigour were selected. This is because the quality of the studies will have a higher impact on the overall validity of this review (Witvrouw et al 2014, Davis & Powers 2010). The quality of the studies will ultimately dictate their representation in the conclusion (Atkinson 2011). Table 5.1 - Study selection Studies Included Bley et al (2014) - Propulsion Phase of the Single Leg Triple Hop Test in Women with Patellofemoral Pain Syndrome: A Biomechanical Study. Bolgla et al (2011) - Comparison of hip and knee strength and neuromuscular activity in subjects with and without patellofemoral pain syndrome Nakagawa et al (2011) - Electromyographic preactivation pattern of the gluteus medius during weightbearing functional tasks in women with and without anterior knee pain. Nakagawa et al (2012) - Trunk, pelvis, hip, and knee kinematics, hip strength, and gluteal muscle activation during a single-leg squat in males and females with and without patellofemoral pain syndrome. O'Sullivan et al (2012) - No difference in gluteus medius activation in women with mild patellofemoral pain. Saad et al (2011) - Analysis of the centre of pressure displacement, ground reaction force and muscular activity during step exercises. Shirazi et al (2014 ) - Comparative Evaluation of Core Muscle Recruitment Pattern in Response to Sudden External Perturbations in Patients With Patellofemoral Pain Syndrome and Healthy Subjects. Song et al (2014) - Effects of femoral rotational taping on pain, lower extremity kinematics, and muscle activation in female patients with patellofemoral pain. Souza and Powers (2009) - Differences in hip kinematics, muscle strength, and muscle activation between subjects with and without patellofemoral pain. Willson et al (2011) - Gluteal muscle activation during running in females with and without patellofemoral pain syndrome.

5.3 - CRITICAL APPRAISAL The value of a review is dependent on the validity of the data gathered and the rigour in which it is analysed (Ross 2012, Ridley 2008, Hart 2001). The PRISMA statement identifies that review quality is dependent on a strict search criteria and a thorough review of the studies to contextualise any biases and methodological limitations (Moher et al 2009). In the following section the studies are appraised to highlight their individual significance to the review (Ross 2012, Aveyard 2010, Bowling 2009). The CASP (2013) CAT has been used to structure the appraisal and the checklist summary can be seen in table 5.8. The full appraisals can be found in Appendix C.

5.3.1 - RESEARCH AIM AND BACKGROUND The introduction and background sections in research papers present the reader with context for the paper, and exhibits the researcherâ&#x20AC;&#x2122;s knowledge and considerations in the area (Ross 2012, Aveyard 2010). All of the studies included showed sufficient background knowledge of PFPS, such as; the aetiology, kinematic associations and gender differences. However Shirazi et al (2014) and Bolgla et al (2011) provided little reasoning for the gender differentiation compared to the other studies. The majority of references used were modern relative to the paper (within 5 years) and when accessed were found to be relevant and of good quality. However none of the studies included followed the recommendations from the 3rd patellofemoral pain consensus, which requested for future research to sub-classify subjects into whether their PFPS was highly correlated to one etiological factor (Witvrouw et al 2014 Selfe et al 2013). Despite some differences all studies included were clearly focused in their aim and acceptably informed in their background.

5.3.2 - RESEARCH DESIGN AND METHODOLOGY All studies were case-controlled laboratory studies using a cross-sectional design; which were, in the absence of prospective research, the most appropriate design for the question (see table 3.3). Due to the case-controlled design randomisation, participant blinding and placebo controls are not possible, as the participants would already know whether they had PFPS or not (Hickson 2013, Ridley 2008). However none of the papers chose to blind the researchers when recording the data; this is a methodological flaw which implications shall be discussed further in section 5.3.4 (Polit & Beck 2013). The EMG methodology is discussed in section 5.3.7 in terms of outcome measure precision. All studies included used a matched subjects design, where those in the PFPS group are matched by sex, height, age and weight to determine a difference between the two groups that were not affected by these factors (Polgar & Thomas 2013). An issue with using matched subjects design in

addressing PFPS is it is difficult to match all significant variables, for example exercise level, foot posture, muscular imbalances, leg dominance etc (Witvrouw et al 2014, Petersen et al 2013, Polgar & Thomas 2013, Polit & Beck 2013). These confounding variables are explored in 5.3.5. Despite the aforementioned factors all studies were deemed of sufficient methodological and design quality to answer the question.

5.3.3 - SAMPLE All studies included used the matched controls design, subjects were matched on sex, age, height and weight (see Table 5.2). Due to the eligibility criteria all the participants were female, apart from Nakagawa et al (2012) who split the analysis of both sexes, so that female data could be extracted separately from the males. The female only sample was central to this reviewâ&#x20AC;&#x2122;s question and is justified in section 2.5. An age range was set to minimise the likelihood of including patients with undiagnosed osteoarthritis or a traction apophysitis, see exclusion criteria (table 4.3). Eight out of ten studies set an age range of between 18-35 (Bley et al 2014, Song et al 2014, Nakagawa et al 2012, Oâ&#x20AC;&#x2122;Sullivan 2012, Bolgla et al 2011, Nakagawa et al 2011, Saad et al 2011, Willson et al 2011). So were less likely to include these potential cases however Shirazi et al (2014) set the range at 18-40 and Souza and Powers (2009) at 18-45, which increases their potential of an unrepresentative sample (Polit & Beck 2013). RodriguezFontenla et al (2012) state that although the peak of knee OA occurs between age 70-80 the first symptoms often start at 40. This suggests that patients in included in Shirazi et al (2014) and Souza and Powers (2009) could have only just developed knee OA and been misdiagnosed as PFPS.

The studies included all stated similar heights, their averages differed by a maximum of 10cm (see table 5.2). None of the studies specified subject leg length. As stated in 2.6 there is a relationship between PFPS and increased body weight (Crossley 2014, Hryvniak et al 2014, Petersen et al 2013, Yard & Comstock 2011). All studies included have a similar body weight, with an average which ranges from 52- 65kg. The difference was deemed acceptable for comparison, as previous research found no significant difference when associating body weight with 49

PFPS at these ranges (Foss et al 2012). The psychological and social effects of body weight and PFPS are discussed in section 2.6. The sample size varied between each study, a large sample is vital to make the study generalizable and to identify significance (Hickson 2013, Atkinson 2011, Bowling 2009). The samples ranged from 9 to 27 subjects in the patellofemoral groups. The full sample sizes are presented in below in table 5.2. Table 5.2 - Studies Participant Details Age

Sample Study

Height (m)

Weight (kg)

PFP

Control

PFP

Control

PFP

Control

PFP

Control

Bley et al (2014)

23.5 ± 2.1

23.1 ±3.3

1.65 ±0.04

1.62 ±0.06

55.3 ±4.8

55.9 ± 7.1

Bolgla et al (2011)

24.5 ± 3.2

23.9 + 2.8

1.7 ± 0.1

63.1 ± 9.1

62.1 ± 8.5

Nakagawa et al (2011)

23±5

23±2

1.65±0.07

1.63±0.06

61±10

56±4

Nakagawa et al (2012)

22.3± 3.1

21.8± 2.6

1.66 ± 0.6

1.63± 0.7

61.1± 7.5

59.4± 7.3

O'Sullivan et al (2012)

23 ± 4

21 ± 1

1.66 ± 0.06

1.65 ± 0.08

62.8 ± 7.6

62.6 ± 9.9

Saad et al (2011)

23±2

1.60±3

59±4

53±2

Shirazi et al (2014)

26.6± 3.6

26.4± 3.3

1.63± 0.05

1.64± 0.05

60.7± 6.3

61.6± 5. 9

Song et al (2014)

25.7 ± 6.1

28.6 ± 5.7

1.64 ± 0.05

1.61± 0.06

55.5 ± 5.8

52.1 ± 5.6

Souza and Powers (2009)

27±6

26±5

1.70±0.06

1.70±0.05

65±10

63±7

Willson et al (2011)

21±3

22±5

1.68±0.06

1.69±0.09

63±8

62±9

Table 5.2.1 - Displaying the sample’s quantity, age, height and weight in their respective groups showing standard deviation were appropriate. The samples were collected in local physical therapy clinics (Bley et al 2014, Shirazi et al 2014, Nakagawa et al 2012, Nakagawa et al 2011, Souza and Powers 2009) at Universities (Bley et al 2014, Nakagawa et al 2012, Willson et al 2011), or through word of mouth or advertisements in the local community (Song et al 2014, O’Sullivan et al 2012). Saad et al (2011) and Bolgla et al (2011) did not specify. See table 5.9 for full sampling details.

Bolgla et al (2011) and Saad et al (2011) did not specify their sampling methods so were not scored. Saad et al (2011) defined a good specific criterion to ensure the control group gathered was asymptomatic and representative. Nakagawa et al (2011) used a sample which was deemed too small to be generalised. Song et al (2014) used half the quantity of controls as the PFP group, this means the control sample isnâ&#x20AC;&#x2122;t as representative of healthy NMA as is required in assessing PFPS (Powers et al 2012, Bolgla et al 2010, Criswell 2010). All other studies obtained their samples in a suitable way.

5.3.4 - BIAS Pannuccu and Wilkins (2010) identified three stages where bias is introduced; this framework shall be used in conjunction with the CASP appraisal to identify bias in the research. Figure 5.3.4 - Sources of Bias

Figure 5.4.4 - Major sources of Bias in Clinical Research Taken from Pannuccu and Wilkins (2010).

As stated in 5.3.2 randomisation, placebo controls and blinding of participants were not possible due to the study designs chosen. All studies implemented a sufficient inclusion and exclusion criteria, so that selection bias could be minimised.

None of the studies blinded the researchers during EMG, which means the researchers may of consciously or subconsciously been able to altered the findings (Polit & Beck 2013). The tests were standardised and identical for both groups and participants were informed equally and given practice tests. However some studies gave verbal encouragement during EMG tasks (Bley et al 2014, Nakagawa et al 2012, O’Sullivan et al 2012, Souza and Powers 2009). Therefore the researcher may have given a disproportionate amount of encouragement between groups, which has been shown to affect performance (Andreacci et al 2002). Even so, the potential for performance bias is minimal throughout. The data collected is all objective, so there is very little bias that can be introduced in the analysis section of the studies. All studies took place over 1-2 days, and none of the papers stated any drop-outs. All studies were deemed of acceptable credibility and free from any major influences of bias.

5.3.5 - CONFOUNDING FACTORS All studies chose the most symptomatic leg to test if the subjects had bilateral pain. However they differed on which leg to test from the control group. O’Sullivan et al (2012) were the only study to match controls on leg dominance, three studies pre-defined the leg they were to test (Song et al 2014, Bolgla et al 2011, Willson et al 2011). Three studies matched the leg to the matched PFPS subject (Bley et al 2014, Nakagawa et al 2012, Nakagawa et al 2011). Souza and Powers (2009) matched by number of each leg. Activity level was addressed in 4 of the 10 studies. Three of these selected physically active subjects; Bley et al (2014) stated 20 minutes of physical activity 3 times a week, O’Sullivan et al (2012) requested a minimum of 30 minutes, 3 times a week and Willson et al (2011) specified 10 miles ran a week. Saad et al (2011) selected a sedentary sample, yet didn’t specify the subjects’ levels of activity. Pain and severity scales were used by 8 out of 10 studies in the inclusion and exclusion criteria. Five studies (Bley et al 2014, Song et al 2014, O’Sullivan 2012, Nakagawa et al 2011, Willson et al 2011) used the specified anterior knee pain scale (AKPS), a scale which has been widely researched and validated when addressing AKP and PFPS (Myer et al 2014, Kievit et al 2013, Pattyn et al 2012). Three 52

studies just recorded VAS (Shirazi et al 2014, Bolgla et al 2011, Souza and Powers 2009). Nakagawa et al (2012) and Saad et al (2011) did not control this variable. Studies which measure amplitude chose to average different phases. The NMA of GMed is different between these phases; therefore the conclusion of average amplitude may have been subjected to bias if left unspecified. Both Bolgla et al (2011) and Souza and Powers (2009) averaged stance phase during the stepping task, Nakagawa et al (2011) averaged swing phase and O’Sullivan et al (2012) recorded both, during a step up and over task. Saad et al (2011) did not specify phases (Discussed in 7.3.2). Although 8 out of 10 studies controlled speed of task, none of them acknowledged this factor during comparison of the data. Table 5.3 displays the speed of each task where appropriate, balance and jump tasks were left out as speed could not be controlled or compared. Table 5.3 - Task speeds Study Bolgla et al (2011) Nakagawa et al (2011) Nakagawa et al (2012)

O'Sullivan et al (2012)

Saad et al (2011) Song et al (2014)

Souza and Powers (2009) Willson et al (2011)

Squat

Gait

96 beats/minute (Step task) 154 steps/min (Step task) 4 seconds >60° degrees (SLS) 5 seconds (Step task) 4 Seconds 30° (SLS) 30 step/min (Step task) 3 seconds 45° (SLS) 2 seconds (Step task) -

1 minute on treadmill 117 step/min -

15m 3m/s 20m 3.52-3.89 m/s

In gait Nakagawa et al (2011) used a treadmill, Where Willson et al (2011) and Souza and Powers (2009) used a runway. In addition Willson et al (2011) and Souza and Powers (2009) were the only two studies which controlled footwear. Similar to speed activation timing can be affected by the choice of threshold used. The varying thresholds are shown in table 5.4, the effects are discussed in section 6.4.1.

Table 5.4 - Onset Thresholds Onset Threshold

Results

Bolgla et al (2011)

Deviation of 3 SDs for >25ms

Nakagawa et al (2011)

Deviation of 2 SDs for >25ms

Shirazi et al (2014)

Deviation of 3 SDs for >25ms

Willson et al (2011)

Deviation of 5 SDs for >25ms

Squat N/A Statistically Insignificant (P = 0.07) Squat Delayed 27ms Statistically Insignificant (P = 0.17) Jump Delayed 3ms Statistically Insignificant (P = 0.81) Balance Delayed 22ms Statistically Significant (P = 0.025) Gait Delayed 24ms Statistically Significant (P = 0.028)

Study

Table 5.4 - Onset Variables and results. SDs=standard deviations ms=milliseconds N/A=Not applicable P=probability value

As stated in section 2.5 there has been a link identified in changes to knee kinematics and joint proprioception observed in women across the phases of menstruation (Labato et al 2013, Fouladi et al 2012, Cesar et al 2010). This variable was not considered by any of the studies included and could affect

the NMA of GMed during the tasks examined. As stated in 1.1 and justified in section 2.3.2 PFPS is a multi-factorial pathology. Each factor will have a varying role in maintaining the dysfunction; therefore the influence of each confounding variable will likely differ between groups and tasks (Witvrouw et al 2014). The majority of studies showed sufficient control of variables, two studies were not rated due to a conflict of good controls and limited information (Bley et al 2014, Nakagawa 2012) one study showed little control of variables (Nakagawa et al 2011).

5.3.6 - RESULTS/CONCLUSIONS Nine studies investigated amplitude of GMed NMA in females with PFPS during a variety of functional tasks (Bley et al 2014, Song et al 2014, Nakagawa et al 2012, Oâ&#x20AC;&#x2122;Sullivan 2012, Bolgla et al 2011, Nakagawa et al 2011, Saad et al 2011, Willson et al 2011, Souza and Powers 2009). Four of the studies found significant results, suggesting increases of amplitude during jump (Bley et al 2014) and step activities (Bolgla et al 2011). The other two studies found decreased amplitudes in step (Saad et al 2011) and squat activities (Nakagawa et al 2012). 54

Two statistically insignificant studies showed increases by 4.5% during a SLS (Song et al 2014), and 25% during running (Willson et al 2011) with probability values close to statistical significance (P = 0.08)(P = 0.064). Four studies addressed onset timing (Shirazi et al 2014, Bolgla et al 2011, Nakagawa et al 2011, Willson et al 2011). The two studies which found significant results stated delayed onset by 22ms during SEP (Shirazi et al 2014) and 24ms during running (Willson et al 2011). The findings by Nakagawa et al (2011) were deemed insignificant. However it should be noted that the average onset of GMed during squat and jump activities in the PFPS group was delayed when compared to controls.

Two studies addressed duration of muscle contraction (Shirazi et al 2014, Willson et al 2011). One study found significant evidence to suggest that PFPS decreases the duration of GMed contraction during running (Willson et al 2011). The results from Shirazi et al (2014) on balance were insignificant (P = 0.132), but showed a potential increase in duration time. One study recorded the peak muscular activation of GMed in females with PFPS during running (Willson et al 2011). The results showed an insignificant (P = 0.87) increase in activation when compared to the control group. See table 5.5 for full results and values of significance.

Table 5.5 - Summary of results EMG Theme

Study

Squat

Gait

Jump

Balance

Propulsion: P = 0.002

Bley et al (2014)

▲ 50% Loading: P = 0.001

▲ 110%

Bolgla et al (2011)

Single leg stance: P = 0.002

▲ 140% Pre-swing: P = 0.6

▲ 2%

Amplitude Nakagawa et al (2011) Nakagawa et al (2012) O'Sullivan et al (2012)

P = 0.15

P = 0.41

P = 0.33

▲ 2.8%

▲ 0.9%

▲ 4.3%

P = 0.33

▲ 3%

▼ 3%

P = 0.017

▼5.7% P = 0.97 N/A P = 0.01

Saad et al (2011)

▼ 18%

Song et al (2014)

▲ 4.5%

Souza and Powers (2009)

P = 0.08

P = 0.064

Willson et al (2011) Bolgla et al (2011)

Nakagawa et al (2011)

▲ 25% P = 0.07 N/A

P = 0.17 Delayed 27ms

P = 0.81 Delayed 3ms

Onset Timing P = 0.025 Delayed 22ms

Shirazi et al (2014)

P = 0.028 Delayed 24ms

Willson et al (2011)

P = 0.132 Longer 90ms

Shirazi et al (2014) Duration P = 0.01 Shorter 42ms P = 0.87

Willson et al (2011) Peak

Willson et al (2011)

▲ 9.3%

Table 6.? - Summary of findings in relation to themes. Green cells = statistical significant results, plain cells = statistically insignificant results, grey cells = not tested. P = probability values, ▲= increase, ▼= decrease

Both Bolgla et al (2011) and O’Sullivan (2012) did not display their insignificant findings of onset timing (Bolgla et al 2011) and amplitude (O’Sullivan et al 2012). Therefore their results were marked N/A in table 5.5 and throughout the analysis.

All studies determined levels of significance by P values. A p value represents the statistical probability that the result is due to chance. In all papers statistical significance was valued below 0.05. The lower the p value the less likely the results were due to chance (Hickson 2013, Bowling 2009) Effect size was used by 3 of 10 studies (Bley et al 2014, Song et al 2014 Willson et al 2011), Similar to p values it quantifies the probability of the results, a value of 0.2-0.5 shows poor causality, 0.50.8 average and 0.8< high (Hickson 2013, Atkinson 2011). By comparing the groups averages, the effect size is determined by each of their standard deviations. If the SD does not differ by 0.2< then the difference is insignificant, even if individually significant. A range of different statistical tests were used to test various aspects of the studiesâ&#x20AC;&#x2122; samples methods and results. All statistical tests used were deemed appropriate for their assigned purpose by supporting evidence (Hollander et al 2013, Cohen 2013, Stevens 2012, Zhu et al 2012, Khuri et al 2011, Razali & Wah 2011).

5.3.7 - PRECISION AND VALIDITY The precision of the results is dependent on the rigour in the studies' method and the skill in which this is executed and recorded (Ross 2012, Aveyard 2010, Ridley 2008). All papers measured EMG of GMed. The measurements used to record NMA were amplitude, onset timing, duration of contraction and peak activation. Amplitude measures the average activation of the muscle over the period of contraction (Criswell 2010). Onset timing measures whether the muscle is contracting later than a control or relative to other muscles associated with it, which could suggest an altered firing pattern (O'Sullivan et al 2012, Criswell 2010). Duration of contraction is to test whether the pathological knee causes a reduction in the length of GMed contraction. Peak activation investigates whether PFPS causes the maximal point of activation to change (Criswell 2010). These are reiterated in table 6.1.

When referring to table 5.7 and 4.4, two of the studies showed high rigour in their EMG method (Bley et al 2014, Shirazi et al 2014), four moderate (Song et al 2014, Oâ&#x20AC;&#x2122;Sullivan et al 2012, Bolgla et al 2011, Willson et al 2011) and four poor (Nakagawa et al 2012, Nakagawa et al 2011, Saad et al 2011, Souza and Powers 2009) (see Table 5.9)

The details of each EMG method are displayed in table 5.6. These are discussed in relation to the overall precision of the data in section 7.2.3.

Table 5.6 - EMG Comparison sort table. Study Bolgla et al (2011) Nakagawa et al (2011)

Measurements Amplitude – RMS of 5 trials, expressed as % MVIC. Onset Timing – signal deviation of 3 SDs for >25 ms Amplitude – Mean value of 6 strides expressed as % MVIC Onset Timing – signal deviation of 2 SDs for >25ms

Electrodes

Grounding

Normalisation

Skin Preparation

Amplification

Other

5mm Bi-polar AgAgCl, 20mm between electrodes.

Ipsilateral clavicle

MVIC taken side-lying, average of three with variability less than 10%

Shaving, abrading and cleaning with alcohol

Band pass filter at 101000Hz. Sampling at 960Hz

3 minute warm up on exercise bike, submaximal speed.

30mm Bi-polar AgAgCl, 30mm between electrodes

Ipsilateral radial styloid process.

MVIC taken side-lying, average of two.

Shaving, abrading and cleaning with alcohol

Band pass filter at 20500Hz Amplifier gain of x1000

5 minute walking warm up on treadmill at 117 steps per minute.

Shaving, abrading and cleaning with alcohol

Band pass filter at 35500Hz Amplifier gain of x1000 60Hz notch filter

5 minute walking warm up on treadmill at 1.66 m/s

Shaving, abrading and cleaning with alcohol

Bandwidth at 500Hz Sampling at 1000Hz Amplifier gain of x2000

MVIC taken side-lying, after initial practice average of three with variability less than 10% MVIC taken side-lying, after 2 practice trials 3 were recorded, the highest being chosen.

Nakagawa et al (2012)

Amplitude – Mean value of 3 SLS expressed as % MVIC

10mm Bi-polar AgAgCl, 10mm between electrodes

Ipsilateral radial styloid process.

O'Sullivan et al (2012)

Amplitude – RMS of 3 trials, expressed as % MVIC.

144mm2 Bi-polar, 18mm between electrodes

Ipsilateral ulnar styloid process.

Saad et al (2011)

Amplitude – Mean value of 3 repetitions then averaged.

Differential surface electrodes (Not specified)

Not specified

Normalised by the average of the filtered results.

Shaving, abrading and cleaning with alcohol

Band pass filter at 20500Hz

Bley et al (2014)

Amplitude – RMS of 3 SLTHT, expressed as % MVIC.

10mm Bi-polar AgAgCl, 20mm between electrodes

Not Specified

MVIC taken side-lying, average of three.

Shaving, abrading and cleaning with alcohol

Band pass filter at 20500Hz Sampling at 1000Hz Amplifier gain of x1000

Warm up treadmill at 1.5m/s for 10 minutes.

Onset Timing – From perturbation till first signal deviation of 3 SDs for >25ms, average of three. Duration – was measured from onset till when EMG dropped below the threshold. Average of 3

Pre-gelled Bi-polar Ag-AgCl, 20mm between electrodes

Iliac Crest

Not applicable

Shaving, abrading and cleaning with alcohol

Band pass filter at 8500Hz Sampling at 1000Hz Amplifier gain of x1000

Amplitude – Mean value of 3 repetitions then averaged.

10mm Bi-polar AgAgCl, 20mm between electrodes

Not specified

Normalised by the RMS average of three SLS.

Shaving, abrading and cleaning with alcohol

Souza and Powers (200)

Amplitude – Mean value of 3 SLS expressed as % MVIC

Preamplified bipolar, grounded electrodes (Not specified).

Not specified

MVIC taken side-lying, average of three.

Shaving, abrading and cleaning with alcohol

Willson et al (2011)

Amplitude – Mean value of ... expressed as % MVIC Onset Timing – signal deviation of 5 SDs for >25ms Duration – was measured from onset till when EMG dropped below the threshold. Peak activation – Noted when EMG hits its peak value during contraction

Preamplified single differential electrodes, 10mm between electrodes

Ipsilateral clavicle

MVIC taken side-lying, two practices then one final recorded trial.

Shaving, abrading and cleaning with alcohol

Shirazi et al (2014)

Song et al (2014)

Band pass filter at 101000Hz Sampling at 1000Hz Amplifier gain of x1000 Band pass filter at 35500Hz Sampling at 1560Hz 60Hz notch filter

Band pass filter at 20450Hz Sampling at 1080Hz

RMS=root mean square, SD=standard deviation, MVIC=maximum voluntary isometric contraction.

5 minute warm up on treadmill at self-selected pace, and gentle lower limb stretches. No warm up specified. Tasks were completed at 30 steps /min.

Practice running at 180 m/min along a 15m walk way.

20m runway 3.523.88m/s

5.3.9 - CLINICAL RELEVANCE Four out of ten studies (Bley et al 2014, Bolgla et al 2011, Saad et al 2011, Willson et al 2011) stated that their results highlight the importance of rehabilitating the hip musculature when treating PFPS, as the current treatment modalities focus predominately on quadriceps strengthening (Hryvniak et al 2014, Petersen et al 2013). Oâ&#x20AC;&#x2122;Sullivan et al (2011) adds that GMed rehabilitation should include a variety of movement planes. Souza and Powers (2009) highlight the importance of testing hip strength and kinematics when diagnosing PFPS. Nakagawa et al (2012) stated the diagnostic potential of the SLS, and Nakagawa et al (2011) believed that observing the preactivation stages was key. Half of the studies involved were deemed clinically relevant (Bley et al 2014, Nakagawa et al 2012, Saad et al 2011, Willson et al 2011), three had poor control over variables or inadequate sampling to determine, so were left unscored (Song et al 2014 Bolgla et al 2011 Souza and Powers 2009). The remaining two studies presented inadequate rigour and were deemed of little clinical relevance.

5.3.10 - SUMMARY This section summarises the critical appraisals into tables which depict the overall quality of the included studies. The EMG appraisal is explored first as it influences some of the answers to the CASP checklist. The standard of EMG was evaluated by addressing the studies' method in comparison to the ideal recommended European standards of EMG (Stegeman & Hermens 2007, SENIAM 2006, Day 2002). The full EMG details have been compiled in table 5.6. A modified check list was created and a score given to each EMG method. The results of this checklist are seen in table 5.7, which was then used to score question eight of the CASP CAT. The complete study summaries are shown in table 5.9. The CASP CAT appraisal sheets can be found in appendix C.

Table 5.7 - EMG Appraisal Checklist - Song is wrong EDIT Study Bley et al (2014) Bolgla et al (2011) Nakagawa et al (2011) Nakagawa et al (2012) O'Sullivan et al (2012) Saad et al (2011) Shirazi et al (2014) Song et al (2014) Souza and Powers (2009) Willson et al (2011)

Electrodes Type Size IED Y Y Y Y N Y Y N N Y Y N Y Y Y Y Y Y Y Y Y Y N

Skin Preparation Y Y Y Y Y Y Y Y Y Y

Amplification Band Pass Notch Filtering Y Y Y Y Y Y Y N Y Y Y Y Y Y Y Y Y N Y Y

Normalisation

Score

Y Y Y Y Y N Y N Y Y

7 5 3 3 5 4 6 5 3 5

Table 5.2- EMG Appraisal Checklist developed from recommendations by Bolgla et al 2010, Clarys et al 2010, Criswell 2010, Bolgla et al 2007, Stegeman & Hermens 2007, SENIAM 2006, Day 2002. Scores and key: Yes (y) = +1 , No (n) = -1, Can't Tell (-)= 0. IED = Inter-electrode Distance.

Table 5.8 - CASP checklist Study Bley et al (2014) Bolgla et al (2011) Nakagawa et al (2011) Nakagawa et al (2012) O'Sullivan et al (2012) Saad et al (2011) Shirazi et al (2014) Song et al (2014) Souza and Powers (2009) Willson et al (2011)

1 Y Y Y Y Y Y Y Y Y Y

2 Y Y Y Y Y Y Y Y Y Y

3 Y Y Y N Y N Y

4 Y Y Y Y Y Y N Y Y

Question Number 5 6 7 8 Y Y Y Y Y Y N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

9 Y Y N N Y Y N Y

10 Y N Y N Y Y Y

11 -

CASP SCORE 8 6 1 6 3 5 9 5 3 9

5.1 - CASP Questions: 1) Did the study address a clearly focused issue? 2) Did the authors use an appropriate method to answer their question? 3) Were the cases recruited in an acceptable way? 4) Were the controls selected in an acceptable way? 5) Was the exposure accurately measured to minimise bias? 6) What confounding factors have the authors accounted for? + Have the authors taken account of potential confounding factors in the design or in their analysis? 7) What are the results of the study? 8) How precise are the results? 9) Do you believe the results? 10) Can the results be applied to the local population? 11) Do the results of this study fit with other available evidence? Scores and key: Yes (y) = +1 , No (n) = -1, Can't Tell (-)= 0

Table 5.9 - Study Summaries Author, Date and Sampling

Aims of study

Type of Study

Strengths and Weaknesses

To compare kinematic, kinetic and muscle activity in the trunk and lower limb during propulsion in the SLTHT using women with PFPS and pain free controls.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Clinically diagnosed PFPS 2) Includes specified AKP scale 3) Demanding functional task Weaknesses: 1) Doesn't state dominant leg 2) Addresses only one functional task 3) Pain was not assessed during SLTHT.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Analysis of three separate weight bearing phases of step task 2) Published linked research in PFPS methods, to improve validity (Bolgla et al 2010, Bolgla et al 2007). 3) Recorded subjects average pain and duration of condition. Weaknesses: 1) Primary examiner was not blinded to subjects conditions. 2) PFPS was not clinically diagnosed.

Bley et al (2014) Sampling: A local physiotherapy clinic and common areas of the university. Bolgla et al (2011)

Sampling: Not Reported

Nakagawa et al (2011) Sampling: A local physical therapy clinic

Nakagawa et al (2012) Sampling: Flyers in physical therapy clinics, athletic clubs and common areas in the university

To compare hip and knee strength, and EMG activity during stair descent in subjects with and without PFPS.

To compare GMED EMG preactivation pattern in walking, stair descent and single leg jumps in women with and without AKP.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Uses specified AKP pain scale 2) Variety of functional tasks. 3) Addresses preactivation. Weaknesses: 1) Small sample size 2) Only assessed preactivation rather than complete functional activity. 3) Subjects had minor AKP

To compare whether there are any differences between the sexes in trunk, pelvis, hip and knee kinematics, hip strength and gluteal muscle activation during the performance of a SLS in individuals with and without PFPS

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Large sample size, 2) Strict criteria and controls 3) Clinically diagnosed PFPS. 4) separate sex evaluation Weaknesses: 1) Focuses on one specific functional task 2) limited information on PFPS intensity and duration in each group.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Separate GMED subdivision analysis 2) Variety of functional tasks 3) Uses specified AKP pain scale 4) States dominant limbs of subjects. Weaknesses: 1) Small sample size 2) Placement of anterior electrodes over TFL 3) Unilateral squat too shallow 4) Subjects had minor AKP 5) All subjects had AKP in their dominant leg

O'Sullivan et al (2012)

Sampling: Word of Mouth via local community

To compare NMA in the 3 GMED subdivisions during 4 weight bearing exercises in women with and without PFPS.

Values of Significance (P values are significance between means)

Conclusions

Females with PFPS showed increased NMA of GMED during the propulsion phase of SLTHT (P = 0.002, Effect size = 1.0)

1) It is suggested by these results that the increase in GMED NMA is to compensate for a deficient alignment of trunk and lower extremities in females with PFPS.

Females with PFPS demonstrated significantly higher GMED EMG amplitudes during loading (P = 0.001) and single leg stance (P = 0.002) of stair descent. However the pre-swing phase (P = 0.602) and onset (P = 0.77) of GMED were insignificant.

CASP & EMG Scores CASP Score 8 EMG Score 7 CASP Score 6

1) PFPS rehabilitation models should not ignore hip and quadriceps NMA.

EMG Score 5

No significant differences between groups in GMED onset time and average preactivation amplitude during walking (P = 0.41), descending stairs (P = 0.17, P = 0.15), or single leg jumps (P = 0.81, P = 0.33).

1) These results do not support the association between females with AKP and altered GMED preactivation.

Females with PFPS had less GMED activation (P = .017) during the SLS.

1) This study showed a decrease in female GMED activation during the selected functional task. 2) There are specific sex differences when addressing PFPS which need to be considered in future studies and rehabilitation regimes.

No significant differences in NMA between groups (P = .97) in any GMED subdivisions (P = .36) and during all functional tasks tested (Wall press, pelvic drop, step-up and over and unilateral squat) (P = .19). Other Values: DF = 1 F Value = 0.001 Îˇp2 = .00

CASP Score 1 EMG Score 3 CASP Score 6 EMG Score 3 CASP Score

1) GMED NMA was similar in all subdivisions in both PFPS and control group in all tasks recorded

3 EMG Score 5

Saad et al (2011)

Sampling: Not reported

To evaluate the postural control during step up and down activities in patients with and without AKP.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Uses both step up and down tasks 2) States dominant limbs 3) Good criteria and controls Weaknesses: 1) Poor EMG description 2) All individuals had AKP in their dominant leg.

To evaluate the core muscles' EMG in response to unexpected perturbations to the pelvis in patients with and without PFPS

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Large sample size 2) Assesses functional postural stability 3) Clinically diagnosed PFPS 4) Very good criteria and controls. Weaknesses: 1) EMG amplitude was not recorded 2) leg dominance was not specified.

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Specified subjects' dominant leg and leg tested. 2) clinically diagnosed PFPS 3) Specific AKP scale was used. Weaknesses: 1) Small control group 2) All controls had dominant leg tested.

Shirazi et al (2014) Sampling: Convenience sampling through orthopaedic referral to physical therapy clinics.

Females with AKP showed decreased NMA of GMED during step up and down tasks. P Values for comparison of means were not presented, however findings were deemed significant to a P value of 0.01.

Females with PFPS showed delayed GMED onset (P=.025) but no difference to duration of activity (P=.132).

Song et al (2014)

Sampling: Local Community via advertisement.

Souza and Powers (2009) Sampling: Convenience sampling of local orthopaedic and physical therapy clinics

To explore the hip and knee kinematics as well as muscle activation between females with and without PFPS.

To determine whether females with PFP demonstrate differences in hip kinematics, muscle strength and activation patterns when compared to controls.

Strengths: 1) Large sample size 2) Multiple tests used. Weaknesses: 1) Some subjects not clinically diagnosed 2) Leg dominance not specified or controlled 4) Current pain level not specified

1) Results suggest females with PFPS use different core stability recruitment patterns when compared to controls 2) NMA rehabilitation should be considered in treatment of PFPS 1) This study showed no differences to GMED NMA during a SLS in females with or without PFPS 2) Femoral rotational taping could be beneficial in the treatment of PFPS.

CASP Score 6 EMG Score 4 CASP score 9 EMG Score 6 CASP Score 5 EMG Score 6 CASP Score

Females with PFPS showed no significant difference in GMED NMA during running, step-down and drop jump tasks when compared to controls (P=.332, F value 1.14, DF 38).

1) Although the NMA of GMED was unchanged this study supports the link between abnormal hip kinematics and PFP.

3 EMG Score 4

Willson et al (2011)

Sampling: Three Universities and two local fitness centres

Case controlled laboratory study using a cross-sectional design.

Females with PFPS showed no significant difference in GMED NMA during an SLS when compared to controls (p = 0.080) (effect size 0.59)

1) Females with AKP do not have muscular imbalances but instead show a reduced NMA of stabilising muscles I.e. GMED, in a step up and down task

To compare the hip kinematics and NMA of GMED and gluteus Maximus in females with and without PFPS

Case controlled laboratory study using a cross-sectional design.

Strengths: 1) Utilises a range of measurements for NMA 2) Large sample size 3) Clinically diagnosed PFPS 4) Specified AKP scale used 5) standardised neutral footwear. Weaknesses: 1) Dominant legs were not specified in either group 2) only the controls right legs were tested.

Females with PFPS showed a significantly delayed (P=0.028, effect size=0.76) and shorter (P=0.01, effect size=0.88) GMED activation during running when compared to controls. No significant difference in GMED amplitude or peak activation during running when compared to controls.

1) Hip NMA differences exist in females with and without patellofemoral pain

CASP Score

2) Future rehabilitation programmes should focus on this dysfunction.

EMG Score

6.0 - ANALYSIS 6.1 - INTRODUCTION This section will analyse the data of the 10 chosen articles summarised in chapter 5. As described in section 3.8.1 and 4.6 a thematic analysis will be used to identify significant themes in the literature and draw conclusions from the collective research (Ross 2012, Aveyard 2010).

6.2 - THEME IDENTIFICATION Themes were developed through the process of sorting the data into common codes. Then by analysing and grouping the codes, themes and their sub-themes can be synthesised (Hickson 2013, Aveyard 2010, Fereday & Muir-Cochrane 2008). The themes are not necessarily the findings of the collective research, but the subjects which are most discussed, or most relevant to the question (Ross 2012, Aveyard 2010). Figure 6.1 shows a mind map of how the themes were created; starting from the boarders of the map the sub-themes can be seen in the blue ovals. These were the first grouping of codes which were noted for their frequency in the papers. After the development of sub-themes, connections were made between them to allocate each into a suitable group. These final groups became the themes, and are shown in the green ovals in figure 6.1.

Figure 6.1 - Mind Map of Themes and Sub-themes

Figure 6.1 - Major, Moderate and Minor Themes. Themes are displayed in the green ovals, sub-themes are shown in dark blue ovals. EMG=Electromyography, N=Number of studies

6.2.1 - SUB-THEME JUSTIFICATION Due to the variety of the functional tasks used, sub-themes were created so that tasks with similar functional demands could be more easily compared. However it is recognised that the grouping of tasks could lead to an introduction of researcher bias, therefore the justifications are given below (Pannuccu and Wilkins 2010). The Gait sub-theme includes three studies: one which analyses walking (Nakagawa et al 2011) and two running (Willson et al 2011, Souza and Powers 2009). The biomechanics of walking and running were analysed by Lohman III et al (2011) who commented that although running is the natural progression to walking there are characteristic kinematic differences to be noted. For example the additional phases found in running which are not found in walking, most notably the early and late float phases. This is coupled with the reduction of stride width; this reduction in the runnerâ&#x20AC;&#x2122;s centre of 65

gravity is to increase their stability due to the introduction of float phases and the reduction in stance phase. This compensation increases femoral adduction and internal rotation which as described in section 2.2 recruits gluteus medius (Lohman III et al 2011). Chumanov et al (2008) found that females showed a significant increase in GMed activity as speed of gait increased from walking to running when compared to the male comparisons. This increase must be understood when evaluating and drawing conclusions from the collected data. However as the study which addressed walking (Nakagawa et al 2011) was considering preactivation and onset of GMed, the Chumanov et al (2008) findings do not apply. As other research shows that the precontraction phase of GMed during gait is similar irrespective of speed (Bartlett et al 2014, Kim et al 2013, Lieberman et al 2006, Sasaki & Neptune 2006). Therefore group data was recorded on both walking and running together in the gait theme.

The second sub-theme; Jump, combines three different studies which record EMG data on the Single leg triple hop test (SLTHT) (Bley et al 2014), Single leg jumps (Nakagawa et al 2011) and drop jumps (Souza and Powers 2009). Bley et al (2014) states in their discussion that findings from the SLTHT should be akin to jumping related activities and that the test itself is to specify the leg with AKP (Halabchi et al 2013).

The third sub-theme is Squat, which is the largest of the three groups and represents a large proportion of the PFP literature. The sub-theme includes both SLS and step tasks as the stair descent tasks are the functional parallel to the SLS. Stair descent tasks produced knee pain in 57 of 77 PFPS patients (74%) (Selfe et al 2001). Although the SLS shows more hip flexion, anterior pelvic tilt and less pelvic rotation on the transverse plane, recent research addressing trunk, pelvis, hip, and knee kinematics in the comparison of SLS and stepping tasks, shows that both of these tests can be used to assess PFPS effectively, with a non-significant degree of inter test variability (Nakagawa et al 2014, Whatman et al 2013). Therefore both of these tasks have been used to generate the Squat theme.

The final and smallest sub-theme is Balance, this is due to recent innovate research by Shirazi et al (2014) who recorded GMed NMA in response to sudden external perturbations (SEP). This gives insight into the core stability recruitment patterns with little anticipation bias of the patients (Stensdotter et al 2009). Addressing the recruitment pattern and speed in force production can be a definitive factor in maintaining stability and preventing injury (Saad et al 2011). Assessing NMA in functional tasks is beneficial for noticing adaptive changes, but SEP can be used to identify the fundamental control system and highlight modifications made by pathology (Shirazi et al 2014, Stensdotter et al 2009, Park et al 2004). Although the use of SEP in testing patients with PFPS is a modern and promising method of analysis, the literature available is currently limited. The theme was created due to the high rigour of the study, which received a nine on the CASP score and a six on the EMG checklist. Although the study was of high quality, the conclusions made will be limited in their validity do to the lack of supporting research (Shirazi et al 2014). See Table 5.9 for the study summary.

The ‘EMG method’ theme is a more complicated than other themes. The sub-themes shown in figure 6.1 are Average Amplitude, Onset Threshold and SENIAM compliance, the first two are factors which are not standardised by the SENIAM guidelines yet have a large effect on the results given (Stegeman & Hermens 2007, SENIAM 2006, Day 2002). These two subthemes are explored later in section 7.2.3. The last sub-theme is the compliance of the EMG method (see table 5.6) in contrast with the SENIAM EMG guidelines (see Appendix A & B). The variables were then combined for the ease of illustrating and comparing the themes. The frequency scoring in table 6.3 and the value of the Y-axis in figure 6.2 display how many studies discussed their EMG method in their papers. Both the scores in figure 6.1 and table 6.3 are relevant and shall be discussed in section 7.2.3. The ‘GMed NMA’ and ‘leg selection’ themes contained definitive sub-themes which did not warrant the same level of justification as the sub-themes under the ‘Functional Tasks’ and ‘EMG Method’ themes. The sub-themes and their explanations are shown in table 6.2.2 and 6.2.3

Table 6.1 - Functional Tasks Sub-themes Subthemes Amplitude

Onset Timing

Duration

Peak

Study Bley et al (2014) Bolgla et al (2011) Nakagawa et al (2011) Nakagawa et al (2012) O'Sullivan et al (2012) Saad et al (2011) Song et al (2014) Souza and Powers (2009) Willson et al (2011) Bolgla et al (2011) Nakagawa et al (2011) Shirazi et al (2014) Willson et al (2011) Shirazi et al (2014) Willson et al (2011) Willson et al (2011)

Explanation

Amplitude was the largest of the groups, including nine different studies which addressed the extent that GMed activated in females with PFPS during a functional task when compared to healthy controls.

Onset included four different studies which addressed whether GMed activated earlier or later during a functional task in females with PFPS when compared to healthy controls.

Duration included two studies which recorded the length of the contraction, to address whether females with PFPS had a reduced or prolonged activation in a functional task when compared to controls. Peak was the smallest theme, including only one study (see table 5.9). The study analysed the peak activation to see whether females with PFPS could not achieve as great of GMed NMA as that of healthy controls.

The ‘leg selection’ sub-theme addresses if and how the controls’ legs were matched to the PFPS subjects, see table 6.2. Table 6.2 – Leg selection Sub-themes Sub-themes

Explanation

Matched to Subjects Matched on Dominance Matched on Numbers Pre-defined Leg Choice

The tested leg of the control was matched to the most symptomatic leg to their matched PFPS subject. The most symptomatic leg of the PFPS subject was chosen, whether this leg was dominant or nondominant determined which of the controls legs were matched. The most symptomatic leg was chosen, the number of each side was counted, and then the number was matched in the control group selection. Irrespective of the PFPS group, all controls were tested on one side, which was selected before examination.

6.2.2 - THEME CLASSIFICATION The next stage is to stratify the themes into classes; the class of each theme is determined by two variables: the frequency of theme and its relevance to the question (Ross 2012, Aveyard 2010). The thematic frequency is shown below in Table 6.3. This data is then plotted on a graph with the theme relevance (X-axis) to determine whether the themes are major, moderate or minor; the differentiation is seen in figure 6.2. The value of each theme in its contribution to this reviewâ&#x20AC;&#x2122;s conclusion is dependent on its class (Aveyard 2010, Fereday & Muir-Cochrane 2008).

Table 6.3 - Thematic Frequency

Figure 6.2 - Valuation of Themes

Figure 6.2.3 - Valuation of Themes X-axis displaying relevance to question, Y-axis frequency of theme in studies

6.3 - MAJOR THEMES

6.3.1 - GLUTEUS MEDIUS NEUROMUSCULAR ACTIVATION Nine studies addressed GMed amplitude across functional tasks. Two high quality studies (Bley et al 2014, Willson et al 2011), four moderate quality (Song et al 2014, Nakagawa et al 2012, Bolgla et al 2011, Saad et al 2011) and three poor quality (Oâ&#x20AC;&#x2122;Sullivan 2012, Nakagawa et al 2011, Souza and Powers 2009). There was conflict in the results with Nakagawa et al (2012) and Saad et al (2011) reporting a decrease and Bolgla et al (2011) and Bley et al (2014) stating an increase. The collective results of the five insignificant studies (Song et al 2014, Oâ&#x20AC;&#x2122;Sullivan 2012, Nakagawa et al 2011, Willson et al 2011, Souza and Powers 2009) showed a trend increase in amplitude over the functional tasks, however with this data alone no conclusions can be drawn to whether changes in GMed amplitude is associated with PFPS in females. The Onset sub-theme included two high quality studies (Shirazi et al 2014, Willson et al 2011), one moderate (Bolgla et al 2011) and one poor (Nakagawa et al 2011). There is moderate evidence to suggest that the onset timing of GMed in females with PFPS is delayed during balance tasks and gait. This is supported by two studies of high quality evidence (Shirazi et al 2014, Willson et al 2011) and through addressing the trend in other lesser quality and otherwise insignificant results (Nakagawa et al 2011). Two high quality studies were included in the duration sub-theme (Shirazi et al 2014, Willson et al 2011). Willson et al (2011) concluded a significant reduction in GMed contraction during running. Shirazi et al (2014) did not find a statistically significant difference between groups during the SEP. The studies give limited evidence to suggest that contraction duration of GMed is reduced during gait in females with PFPS. This statement is supported by one high quality study (Willson et al 2011). There is at present insufficient evidence to suggest that PFPS affects GMed peak muscular activation in females when compared to healthy controls during functional tasks. One high quality study 72

(Willson et al 2011) deemed there to be an insignificant association between females with PFPS and healthy controls.

6.3.2 - FUNCTIONAL TASKS The analysis of each subtheme will be separated into two parts. The first will be analysing the results found in relation to each task, so as to determine if the PFP NMA characteristics are specific to the activity. The second part will compare the differences between the functional task procedures; to identify potential extraneous variables and links in the results, which will be discussed in section 7.3.

The squat sub-theme contains eight sets of data from seven different studies (Song et al 2014, Nakagawa et al 2012, O'Sullivan et al 2012, Bolgla et al 2011, Nakagawa et al 2011, Saad et al 2011, Souza and Powers 2009). All studies recorded GMed amplitude and two studies recorded onset (without significance) (Nakagawa et al 2012, Bolgla et al 2011). As shown in table 6.4 Bolgla et al (2011) found a significant increase in amplitude during the loading and single leg stance phases of the step task. Saad et al (2011) and Nakagawa et al (2012) stated a significant decrease during SLS and step task. The insignificant data developed a trend toward an increase (see table 6.4). Bolgla et al (2011) found a far larger increase than any other study, potential reasons for this are explored in the second part of this subtheme, and section 7.3.1.1. There is insufficient evidence at present to determine whether females with PFPS show altered GMed NMA during a squat activity.

Table 6.4 - Squat Results Studies

Amplitude

Bolgla et al (2011)

Loading: P = 0.001 ▲ 110% Single leg stance: P = 0.002 ▲ 140% Pre-swing: P = 0.6 ▲ 2% Nakagawa et al (2011) Amplitude P = 0.15 ▲ 2.8% Onset P = 0.17 Delayed 27ms Nakagawa et al (2012) P = 0.017 ▼5.7% O'Sullivan et al (2012) P = 0.97 N/A Saad et al (2011) P = 0.01 ▼ 18% Song et al (2014) P = 0.08 ▲ 4.5% Souza and Powers (2009) P = 0.33 ▲ 3% Colours over P-values show significance level: dark green=very significant, olive=significant, orange=close to statistical significance, red=not significance. ms=milliseconds

All studies included in the squat theme address unilateral contraction and stabilisation on a flexed and loaded knee, five of these being stepping tasks and three using single leg squat tasks. The procedures in each study differed in height/depth and speed of the task. O’Sullivan et al (2012) used a 15cm step and Bolgla et al (2011), Nakagawa et al (2011), and Saad et al (2011) used a 20cm step. Souza and Powers (2009) standardised the step height as 10% of the individual, to reduce the effect of height on the results. In addition, Saad et al (2011) and O’Sullivan et al (2012) used significantly slower step pace than all other studies, see table 5.3 for details (Bolgla et al 2011, Nakagawa et al 2011, Souza and Powers 2009). Song et al (2014) used a pace of 30°/s to perform a SLS to 45°, this study only included eccentric (quadriceps) squat data, whereas Nakagawa et al (2012) and O’Sullivan et al (2012) recorded the EMG till back to standing. Nakagawa et al (2012) required the SLS to be greater than 60° and over a 4 second

period; 2 seconds into the squat and 2 back to standing. Oâ&#x20AC;&#x2122;Sullivan (2012) required the same timing but into a 30Â° SLS. The data from Song et al (2014) is most comparable to Bolgla et al (2011) and Souza and Powers (2009) due to their focus on eccentric contraction. Whereas data from Nakagawa et al (2012) and Oâ&#x20AC;&#x2122;Sullivan et al (2012) should be compared to Saad et al (2011) as they include both eccentric and concentric contractions. When these considerations are made the data suggests a new theory. In which there is an increase in GMed amplitude during loaded flexion phase, which is followed by a larger decrease in amplitude in the extension stage of the SLS and step tasks. (See table 6.4, colour coding has been added for ease of reference). This theory is discussed in section 7.3.1.1. The Gait sub-theme used three studies, one of high quality (Willson et al 2011) and two of poor quality (Nakagawa et al 2011, Souza and Powers 2009). Most data and all significant data came from Willson et al (2011). There were no significant differences found for amplitude or peak, however the insignificant data shows a collective increase in both (Nakagawa et al 2011, Souza and Powers 2009). There is limited evidence to suggest a delayed onset and shortened duration of GMed in females with PFPS during running when compared to controls (Willson et al 2011). However due to all this data coming from one study the results could be unrepresentative, this is discussed in section 7.3. All results are shown in table 6.6. As addressed in 6.2.1 the gait sub-theme includes both walking and running. However as Nakagawa et al (2011) recorded pre-stance phase EMG, and used a treadmill rather than a runway; their results should be analysed independently. Willson et al (2011) and Souza and Powers (2009) used similar length runways (20m and 15m), similar speeds (3.52-3.89m/s and 3m/s) and both standardised neutral footwear (see table 5.3). These two studies are easily comparable, and show good standardisation of procedure. However the considerations do not have any change on the previous conclusions. The influence of these factors is discussed in section 7.3.2.

Table 6.6 - Gait Results Studies

Gait

Nakagawa et al (2011) Souza and Powers (2009) Willson et al (2011)

Amplitude P = 0.41 ▲ 0.9% Amplitude P = 0.33 ▲ 3% Amplitude P = 0.064 ▲ 25% Onset Timing P = 0.028 Delayed 24ms Duration P = 0.01 Shorter 42ms Peak P = 0.87 ▲ 9.3%

Three studies were included in the jump sub-theme, one high quality study (Bley et al 2014) and two poor (Nakagawa et al 2011, Souza and Powers 2009). There is moderate evidence to suggest an increase in amplitude in females with PFPS during the jump task. The onset timing was insignificant (Nakagawa et al 2011). Nakagawa et al (2011) recorded the single leg jumps amplitude from time of GMed onset to moment the foot touches the floor, so essentially ignoring the landing phase, focusing on the propulsion phase. This means that the two studies can be easily compared, as Bley et al (2014) specifies each phase of the SLTHT, so that the propulsion phase can be compared with the results from Nakagawa et al (2011). The drop jump task is largely the same mechanically, but with a higher stability and lower muscular demand. This is due to the task using both legs and therefore does not require as much GMed NMA as the single leg tasks (Souza and Powers 2009). This difference in biomechanical requirements of Souza and Powers (2009) jump could be the reason for their (statistically insignificant) decrease in amplitude. As both studies which used unilateral jump tasks stated an increase. There is insufficient data to determine any effect on onset.

Table 6.7 – Jump Results Studies

Jump

Bley et al (2014) Nakagawa et al (2011)

Souza and Powers (2009)

Amplitude P = 0.002 ▲ 50% Amplitude P = 0.33 ▲ 4.3% Onset P = 0.81 Delayed 3ms Amplitude P = 0.33 ▼ 3%

The balance sub-theme included one high quality study (Shirazi et al 2014) which stated a significant delay in GMed onset during SEP. The duration findings were deemed insignificant. There is moderate evidence to suggest a delay in GMed NMA in females with PFPS during a balance task. As only Shirazi et al (2014) used this procedure there was no issue with data comparison.

6.4 - MODERATE THEMES

6.4.1 - ELECTROMYOGRAPHY METHOD 6.4.1.1 â&#x20AC;&#x201C; SENIAM GUIDELINES Of the studies included two showed high rigour in their EMG method (Bley et al 2014, Shirazi et al 2014), five moderate (Song et al 2014, Oâ&#x20AC;&#x2122;Sullivan et al 2012, Bolgla et al 2011, Saad et al 2011, Willson et al 2011) and three poor (Nakagawa et al 2012, Nakagawa et al 2011, Souza and Powers 2009) (see Table 5.9). As seen in table 5.6 all studies selected used bipolar electrodes and all where specified stated they were Ag/AgCl, however some studies differ in electrode sizes used. Nakagawa et al (2011) used electrodes larger than recommended (30mm) and Bolgla et al (2011) used a smaller alternative (5mm). Saad et al (2011), Shirazi et al (2014), Souza and Powers (2009) and Willson et al (2011) did not specify. The inter-electrode distance is recommended at 20mm (Stegeman & Hermens 2007, SENIAM 2006, Day 2002). Nakagawa et al (2011) again exceeded using 30mm, and Nakagawa et al (2012) and Willson et al (2011) used an IED of 10mm. These influences could potentially cause a further decrease in Nakagawa et al (2012) amplitude and an increased delay in Willson et al (2011) onset. However Criswell (2010) state that changes less than 20mm are likely too minor to alter significance. Notch filtering was used in Nakagawa et al (2012) and Souza and Powers (2009), and is not recommended by the used guidelines (Day 2002). Its effect on their results cannot be determined due to the variety of other variables used in other studies of similar design. The influence of notch filtering is discussed in section 7.2.3. Normalisation of all studies but Saad et al (2011) and Song et al (2014) used maximum voluntary isometric contraction (MVIC) normalisation. Saad et al (2011) and Song et al (2014) used the average of the filtered results to normalise the data. This decreases the comparability of the research. However even when all the MVIC normalised data is analysed separately there is not enough evidence to support a definitive association for amplitude alterations in females with PFPS.

6.4.1.2 - SPECIFIED PATELLOFEMORAL PAIN VARIABLES Studies recording amplitude in the squat sub-theme differ in the phases of task which they record. The differences during the SLS tasks were stated in section 6.3.2, as the differences were found in the task procedure. The step tasks however were determined by changes to the process of amplitude averaging during the EMG method, so the differences are analysed below. During the step tasks, Bolgla et al (2011) and Souza and Powers (2009) calculated their average from the stance phase. Nakagawa et al (2011) recorded GMed preactivation (swing phase) and excluded stance phase. Oâ&#x20AC;&#x2122;Sullivan et al (2012) recorded a continuous 5 second period of a step up and down, not differentiating the switch between the step up and step down phases. And Saad et al (2011) did not specify which stages were averaged, so the assumption is made that it was the entirety of the task. Due to the aforementioned factors these studies (Oâ&#x20AC;&#x2122;Sullivan et al 2012, Nakagawa et al 2011, Saad et al 2011) were difficult to compare with Bolgla et al (2011) and Souza and Powers (2009). This lack of specification may have caused Oâ&#x20AC;&#x2122;Sullivan et al (2012), Nakagawa et al (2011) and Saad et al (2011) results to be reduced when compared to Bolgla et al (2011). However even when analysing these studies separately there is no significant association, due to the conflicting results of Nakagawa et al (2012) and Bolgla et al (2011). See table 6.4 for results and specifics on probability values. Another variable which differed between studies was the onset threshold, see table 5.4. The significant results from Shirazi et al (2014) and Willson et al (2011) showed an average delay of 23ms. Willson et al (2011) found the highest delay of GMed; however their threshold was the highest of all of the studies. This could suggest that the delay found was less significant, however more research is needed. When analysing the data addressing only the studies with high electromyographic quality; a significant increase in amplitude during propulsion phase of jumping jump and delay in activation during balancing is seen. However as only two papers reached this standard no definitive conclusions can be made.

6.5 - MINOR THEMES

6.5.1 - LEG DOMINANCE Because of the differences between limbs (explained in 7.4) studies analysing GMed NMA should account for limb dominance asymmetries through the matching of the dominant or non-dominant leg with the controls. Only one study (O'Sullivan et al 2012) considered this factor in their method, and unfortunately all participants presented with dominant limb pain, which may affect the validity of the findings. When patientsâ&#x20AC;&#x2122; presented with bilateral symptoms the majority of studies selected the most symptomatic leg (Bley et al 2014, Song et al 2014, Nakagawa et al 2012, O'Sullivan et al 2012, Bolgla et al 2011, Nakagawa et al 2011, Willson et al 2011, Souza and Powers 2009). The two studies which didnâ&#x20AC;&#x2122;t either had only unilateral participants (Saad et al 2011) or stated bilateral symptoms as part of their exclusion criteria (Shirazi et al 2014). Two studies standardised their methods through using the control groupsâ&#x20AC;&#x2122; right leg, however they did not specify dominance (Bolgla et al 2011, Willson et al 2011). Other studies matched the controls leg to the most affected on the patient with PFPS; this is a more accurate method than the last but still ignores the potential for dominant/non-dominant asymmetries (Bley et al 2014, Nakagawa et al 2012, Nakagawa et al 2011). Song et al (2014) did not specify the legs chosen for the PFPS group, and tested only the dominant legs of the control group. Souza and Powers (2009) matched the controls by numbers of left or right; there were 13 right and 8 left symptomatic legs, so 13 right and 7 left legs were selected from the controls. Explanation of differences and discussion is found in section 7.4. As only one study controlled leg dominance, and all of the participants presented with dominant sided pain no conclusions could be drawn from the results. Future studies should control this variable, see section 7.5.

7.0 - DISCUSSION 7.1 – INTRODUCTION This chapter will discuss the analysis of the included studies with relevant existing research in order to develop the conclusions. It will then acknowledge the limitations, ethical considerations, and clinical implications throughout the studies, and in this literature review.

7.1.1 - RESEARCH QUESTION “Is patellofemoral pain syndrome associated with alterations in gluteus medius neuromuscular activation in the female population?”

7.2 - STATEMENT OF FINDINGS Of the ten studies addressing the relationship of GMed and PFPS in females; six studies found a statistical difference between groups. The studies varied in: methodological rigour, controls used, samples, outcome measures, EMG measurements and the use of functional tasks. This amount of variability can make the conclusions and assimilation of data problematic (Aveyard 2010, Hart 2001). The suggestions made from the analysis section are as follows: There is insufficient evidence to conclude that PFPS is associated with changes to female GMed amplitude during gait and balance tasks. In addition there were no significant associations of PFPS and peak activation in all functional tasks, and no definitive trend during the squat task across all outcome measurements. However there is limited evidence to suggest: that onset timing is delayed during balancing tasks (P=0.025); there is a delay in onset timing and a shortening of duration during gait tasks (P=0.028) (P=0.01); and an increased average activation in jump tasks (P=0.002).

7.3 - CONTEXTUALISING THEMES This section will bring in the theories and conclusions of other research in the field, to add context and explain the results found in this review. The context of background literature will allow the research question to be answered more effectively (Polgar & Thomas 2013, Aveyard 2010).

7.3.1 - MAJOR THEMES 7.3.1.1 - GLUTEUS MEDIUS NEUROMUSCULAR ACTIVATION

The research is currently limited by: a lack of prospective studies, low sample sizes and high levels of heterogeneity; in the study designs, functional tasks used, and EMG method (van der Heijden et al 2015, Barton et al 2013, Lankhorst et al 2012). All of these factors are applicable to the included studies (see table 5.2 and 5.9) and is a conceivable reason for the large percentage of insignificant results (40%) found. The previous research in NMA in PFPS has been predominantly on mixed gender samples (Lack et al 2014, Barton et al 2013, Lankhorst et al 2012, Mills et al 2012, Aminaka et al 2011, Ott et al 2011, Cowan et al 2009, Waryasz & McDermott 2008, Boling et al 2006, Earl et al 2005, Brindle et al 2003). Most gender specified research addresses differences in hip and knee frontal plane kinematics, which has been linked to the development of PFPS (Willson et al 2012, Willson et al 2011, Chumanov et al 2008, Ferber et al 2003). A recent systematic review of gluteal NMA in individuals with PFPS was completed by Barton et al (2013). Although they used mixed gender groups, it is the closest review to use in comparison. When comparing the two reviews, the conclusions were incongruous; as some would corroborate, while others conflict. The first example was that this review and Barton et al (2013) used Willson et al (2011) to suggest a delay of GMed onset during gait. No other studies (apart from those included) addressed GMed onset during gait, even with mixed samples. Mills et al (2012) recorded GMed onset during running with different orthoses, however no control group was included, and the controls were too heterogeneous to compare to the data from Willson et al (2012, 2011). The current data in gait

NMA and PFPS is predominately focused on the quadriceps (Gobbi et al 2014, Lankhorst et al 2012, Boling et al 2009, Waryasz and McDermott 2008). Showing that vastus medialis obliquus (VMO) is often weaker and delayed (Hryvniak et al 2014, Petersen et al 2013, Toumi et al 2013). Therefore although both reviews made the same conclusion the validity of this statement cannot yet be supported by complementary research. Barton et al (2013) stated moderate evidence to support shorter and delayed GMed activation during stair ambulation in the PFPS group when compared to controls. However this review found no significant delay between groups during stair ambulation (see table 5.5). This contradiction may be due to the gender differences found in GMed NMA (see 2.5) or the methodological differences in Bolgla et al (2011) (table 5.9 and section 6.3.2) and the poor rigour displayed by Nakagawa et al (2011)(See table 5.9). Research from Bowling (2008) found an earlier onset timing of GMed during step tasks; this again conflicts with the findings of Barton et al (2013) and the conclusions of this review. The reason for the variation could be due to the lower average threshold in Bowling (2008); who used <3SD whereas Barton et al (2013) and this review included thresholds of <5SD (The influence of threshold value is explained in section 7.2.3). Or that the nature of NMA in PFPS is based on the biomechanics of the individual rather than a set system of dysfunction (Dye 2005, Dye 2001). These findings suggest that the right controls have not yet been implemented to accurately test GMed NMA during step tasks (see 7.5). The shortened duration found could not be supported or contested as this outcome measurement was not recorded in any of the included studies. The current evidence on GMed contraction duration and PFPS is limited (Witvrouw et al 2013). Only one study could be found which suggests a shortened GMed activation in patients with anterior cruciate ligament injuries. However due to the study including not addressing PFPS, including both genders, and being 15 years old; no conclusions can be made (Griffin et al 2000). No other study could be found which used SEP for GMed NMA in subjects with PFPS. Stensdotter et al (2009) used SEP to address PFPS stabilisation kinematics, he found that patients showed avoidance behaviour on the afflicted limb but did not record EMG. Pain avoidance behaviour has

been linked to changes in EMG, however research is yet to specify those with PFPS (Hermens & Hutten 2002). When considering the task variables analysed in the squat sub-theme (6.3.2), the evidence suggests an amplitude oscillation. This is an increase in amplitude when the knee is in a state of loaded flexion, followed by a larger decrease in amplitude during the extension phase. If the entire EMG recording was averaged then the findings would show a slight decrease, as seen stated by Nakagawa et al (2012) and Saad et al (2011). This theory is supported by Zeller et al (2003) who suggest that women tend to have less GMed control and rely more on the quadriceps musculature to control the knee. As the quadriceps contract eccentrically, a greater load is placed on the PFJ (Wallace et al 2002). As the load increases; the quadriceps are inhibited due to reflex inhibition at the spinal level, and then GMed compensates to keep dynamic control of the knee (Hart et al 2010, Grenholm et al 2009, Fagan & Delahunt 2008, Moseley & Hodges 2005). During the (concentric) second phase the decrease is caused by the quadriceps regaining dynamic control of the knee, due to the decreased loading of the PFJ (Distefano et al 2009, Fagan & Delahunt 2008, Wallace et al 2002). This peak of gluteal activation occurring during the first loading stages is corroborated by Bolgla et al (2011), Willson et al (2011), and Zeller et al (2003), however more research is needed. The peak EMG measure is often replaced by average amplitude, as it is disputed in the literature to whether it is necessary to record both measures (Hibbs et al 2011, Criswell 2010, Burden 2007). Only one study included this in the analysis, no other research is available to support this association (Willson et al 2011). Willson et al (2011) suggested that the NMA showed similar characteristics over a variety of functional tasks. However this research and the research from Barton et al (2013) disagree; as the current evidence suggests that functional tasks should be analysed separately, due to the high variability in the outcome measures.

7.3.2 - FUNCTIONAL TASKS The functional tasks chosen were based on a variety of weight bearing activities which are commonly symptomatic for patients with PFPS, and ones which are essential for sport or daily human activity (Hryvniak et al 2014, Petersen et al 2013). The analysis section highlighted some of the interstudy differences between functional task methods and their implemented controls. In this section their effect on the outcome measures will be discussed. Each task included several variables which made drawing conclusions problematic. As stated in 6.3.2 the step height and squat depths were often heterogeneous between studies. Lankhorst et al (2012) states there is no current association between subject height and PFPS. However the height of the subject can affect the precontraction length of the musculature during stepping tasks, if the step height is not relative to the individual. The precontraction length of the musculature has been shown to affect muscle demand (Selfe et al 2001). In addition during a step down task if their legs are shorter the amount of hip flexion required will increase, again increasing the muscular demand (Zeller et al 2003, Wallace et al 2002). Research has shown that the speed at which the functional task is completed directly effects the NMA of GMed (Bovi et al 2011, KyrĂślĂ¤inen et al 2005). This is a factor which affects both the squat and gait tasks. There are currently no specified studies addressing EMG differences due to squat or running speeds in subjects with PFPS. However findings by Heiderscheit et al (2011) states that an increase in step rate can substantially reduce the loading at the hip and knee joints during running. Reducing loading at the PFJ will reduce the inhibition of quadriceps, and therefore gluteus medius (Hart et al 2010, Grenholm et al 2009, Fagan & Delahunt 2008, Moseley & Hodges 2005). Petersen et al (2015) found that running at slow speeds over a set distance increases the cumulative loading of the knee joint when compared to faster speeds. As explained in section 6.2.1, the data from this review would not be affected due to Nakagawa et al (2011) assessing swing phase. However these findings do apply as a rule for future research considerations. Although no definitive GMed characteristics related to task speed can be made, it is suggested by the given evidence that there is an influence.

The standardisation of footwear when assessing kinematics or EMG is vital to achieving reliable data in all tasks. Research has shown that changes in footwear can affect the outcomes of all the functional tasks measured in this review (Harry et al 2015, Sinclair et al 2014, Burke 2012, Scott et al 2012). A consideration for the gait sub-theme is that recent research has highlighted kinematic differences between treadmill and over ground running (Lindsay et al 2014, Sinclair et al 2013, Van Caekenberghe et al 2013, Hong et al 2012). These findings suggest that data collected on and off a treadmill cannot be compared to a high degree of validity, as the kinematic differences are likely to affect the outcome measures if left uncontrolled. Barton et al (2009) reviewed 24 studies which analysed gait characteristics in subjects with and without PFPS. Patients with PFPS exhibited increased hip adduction, knee external rotation, delayed and reduced rear foot eversion. This review included both treadmill and over ground tasks and a definitive trend was still found in the data; this suggests that the option of running ground has less influence than the pathological mechanics. However to minimise bias future studies should standardise this task (Sinclair et al 2013). As identified in the analysis section; jump tasks differed as to whether the landing was on one leg or both. Pappas et al (2007) found an increase in muscular demand and amplitude of quadriceps and hamstrings during the single legged jump task, in addition the difference was much great in women. However research from Carcia & Martin (2007) state there is no significant difference in EMG between genders in drop jump tasks. Willson & Davis (2009) found that women with PFPS present with altered kinematics and GMed weakness during jump tasks. Ogasawara et al (2014) supported this data and added that women showed increased quadriceps and gluteus medius activation during a one legged landing. The current literature supports the notion that single leg tasks require greater activation, and could be a good diagnostic tool for PFPS (Ogasawara et al 2014, Willson & Davis 2009, Pappas et al 2007).

7.3 - MODERATE THEMES

7.2.3 - ELECTROMYOGRAPHY MEASUREMENTS 7.2.3.1 â&#x20AC;&#x201C; SENIAM GUIDELINES The SENIAM standards recommend bipolar Ag/AgCl electrodes with a maximum size of 10mm, as larger sizes have been shown to give higher amplitudes than normal (Stegeman & Hermens 2007, Day 2002). 10mm is deemed large enough to detect a representative pool of motor units, but small enough to reduce the effects of crosstalk from the neighbouring musculature (Day 2002). Therefore Bolgla et al (2011) would find more of an increase during amplitude than stated and Nakagawa et al (2011) would find a small decrease. Clarys et al (2010) researched the reliability of EMG and stated that electrode size can have an influence of the overall recording. However Criswell (2010) states that the change is negligible in comparison to the differences being measured. As stated in 6.4.1 Nakagawa et al (2012) and Souza and Powers (2009) used notch filtering to filter their EMG. The issue with notch filtering is that important EMG signal information can be lost at the notched ranges (Criswell 2010, Day 2002). The literature agrees that although the resulting data is rarely affected, notch filtering should be avoided (Criswell 2010, Stegeman & Hermens 2007, Day 2002). A limitation of using surface EMG is that there is the potential for cross-talk of musculature (Clarys et al 2010). Some research recommends that you separately analyse TFL to exclude its influence on GMed Semciw et al (2013). None of the studies included followed this. However they did all standardise the electrode positioning with the SENIAM (2006) recommendations, prepared the skin in the appropriate manner and used tape to prevent electrode slippage. Dieterich et al (2014) states that surface EMG is the most reliable method when compared to ultrasound and fine wire EMG in predicting GMed NMA. Finally the different methods of EMG normalisation must be addressed as it can affect the variability (Clarys et al 2010, Bolgla et al 2010, Bolgla et al 2007). All studies but Saad et al (2011) and Song et al (2014) used MVIC normalisation. Although Clarys et al (2010) states that isometric forces 87

cannot accurately represent dynamic movements, the SENIAM standards (Stegeman & Hermens 2007 Day 2002), Burden (2010) and work by Bolgla et al (2007) specifically in addressing GMed EMG and PFPS has shown MVIC to be an effective normalisation method. Saad et al (2011) justifies their choice by referencing Yang and Winter (1984), a study which did not address GMed, used 11 subjects and was published 31 years ago. For these reasons the MVIC method is preferred in the data analysis. Of the ten studies included only two completed the SENIAM guidelines (Bley et al 2014, Shirazi et al 2014) to satisfactory requirements (Stegeman & Hermens 2007, Day 2002).

7.2.3.2 â&#x20AC;&#x201C; SPECIFIED PATELLOFEMORAL PAIN VARIABLES The guidelines presented in the SENIAM recommendations are not specified for PFPS data collection (Stegeman & Hermens 2007, Day 2002). Because of this there are still a variety of variables relevant to PFPS which are not standardised in the guidelines (Barton et al 2013). This is largely a product of the conditionâ&#x20AC;&#x2122;s own ambiguity. One variable which differs between studies is; when the amplitude be recorded and averaged. This variable can introduce bias into the data, as the phases chosen tend to show different amplitudes. The method most frequently used is averaging stance phase, as it is the most symptomatic in patients with PFPS, due to the process of loading a flexed knee (Hryvniak et al 2014, Sanchis-Alfonso 2014, Halabchi et al 2013, Petersen et al 2013). This method is used frequently and favoured in addressing kinematics and NMA in subjects with PFPS (Bolgla et al 2011, Willson et al 2011, Bolgla et al 2010). Onset threshold was analysed in section 6.4.1; showing a high threshold will show a delayed onset. However a relatively low threshold level will result in an early onset; and tends to give more false activations (Xu et al 2013, Rasool et al 2012, Wong 2009). As EMG onset is often differentiated by milliseconds the change in threshold could determine the results (Xu et al 2013, Criswell 2010). Wong (2009) stated that there is no current standardisation for the threshold values, which has caused high variation in the literature. Similar to the phase averaging, the freedom of choosing the threshold value allows the researcher to influence the results.

7.4 - MINOR THEMES A consideration which was overlooked in the majority of the literature was the effect that leg dominance has on NMA and proprioception. A number of papers from the Experimental Brain Research Journal found that the non-dominant arm is more effective at dynamic proprioceptive feedback during functional tasks (Goble & Brown 2009, Goble & Brown 2007, Goble et al 2006). Han et al (2013) then extended this theory to other joints and found a similar increase in the non-dominant sides relative to the dominant. It is theorised that this asymmetry is due to the non-dominant limb commonly being used to control objects, allowing the dominant side to interact with it them. For example the non-dominant leg will stabilise the player while the other strikes a ball, one hand will hold a nail while the other strikes it (Han et al 2013, Sainburg 2005, Sainburg 2002). Then over time neuroplasticity will cause these sides to specialise, causing a reinforcement of the asymmetries (Dayan & Cohen 2011, Jacobs et al 2005). Jacobs et al (2005) researched the NMA, strength and endurance asymmetries of gluteus medius to compare dominance in healthy subjects, and found a significant increase in strength on the dominant side. Research from Simon & Ferris (2008) adds that lower leg dominance is caused by neural factors rather than muscular asymmetries between the limbs. These findings support the suggestion that it would be clinically advantageous to include proprioceptive training into current rehabilitation models so as not to focus too exclusively on muscular asymmetries (Witvrouw et al 2014, Petersen et al 2013).

7.5 - RECOMMENDATIONS FOR FUTURE RESEARCH It has become apparent from evaluating the literature that the most urgent advancements need to be made in development of sub-groups in PFPS research, and the standardisation of data collection and analysis (Witvrouw et al 2014, Selfe et al 2013). These must be completed before the evidence can be reliably compared. Once this framework has been developed, future research should focus on completing large longterm prospective cohort studies, which utilise sub-group classification into the aetiological factors of PFPS (Selfe et al 2013).

The psychological and qualitative aspects in PFPS have been largely ignored in the literature, leaving a reductionist view of the emotive sides of pain. Future studies should include the role of psychology in PFPS. In addition to this the squat task amplitude oscillations theorised in section 6.3.2 should be explored, by a study that separately averages each stage of the squat, including a large sample and strictly controlled. Future research should control variables such as leg dominance, menstruation phases and footwear when gathering NMA data from women (Harry et al 2015, Sinclair et al 2014, Cesar et al 2010, Simon &

Ferris 2008)

7.6 - STUDY LIMITATIONS The choice of a literature review was seemingly the most appropriate for the subject area in order to answer the question at the start of this review. However upon evaluation of the literature exploring the NMA, there was a lack of standardized primary evidence. This in turn affected the conclusions which could be drawn from this review. In addition to this only cross-sectional research was analysed which means that no cause and effect of aetiological factors could be confirmed. As stated in 7.5, future research should focus on creating prospective evidence, specifying sub-groups and a standardized method. Limitations of individual studies are explained in section 5.3. The data included was all quantitative, which has been criticized for ignoring the emotive side to research (Bowling 2009). As the discussed in section 2.6, PFPS is affected by the individualâ&#x20AC;&#x2122;s psychological state, therefore it could be suggested that this review has neglected the qualitative aspects to PFPS (DomĂŠnech et al 2014, Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Jensen et al 2005). The use of a thematic analysis in a quantitative review is a limitation in the ability for the review to synthesize the collective findings of the studies. However this type of analysis proved more appropriate due to the level of heterogeneity in the selected studies. This review failed to gather any grey literature, which means it could be under the influence of publication bias. However this was not a conscious decision, but a product of the already limited evidence available.

The evaluation of the EMG method was flawed in that a CAT was created by a novice researcher. As the researcher has had no practical experience with EMG the evaluation was based on the literature available. All aspects of the SENIAM guidelines were placed in a checklist and compliance was scored (Stegeman & Hermens 2007, Day 2002). However in practice certain aspects of the method would have more significance on the final result than others (Criswell 2010). Due to the inexperience of the researcher all aspects of the method were rated with the same value. The systematic process of this review was limited compared to others due to the time, resources, and inexperience of the solo researcher involved in the project. Despite the aforementioned limitations, the review explored the associative relationship of GMed NMA and PFPS, and was able to elucidate differentiation between genders.

7.7 - ETHICAL CONSIDERATIONS All studies included stated that ethical approval was obtained, and no injuries were reported during the process of the functional tasks. This review was developed in complete compliance of the Research Framework for Osteopaths (NCOR 2007) which states strict ethical guidelines for osteopathic research. In practice this review has facilitated ethical treatment, as the findings and discussion topics in this review can be implemented straight into clinical practice. Through the development of the practitionerâ&#x20AC;&#x2122;s knowledge base, their treatment can become safer and more effective (General Osteopathic Council 2006). The osteopathic standards state that an osteopath should provide a safe and ethical treatment protocol based on evidence and experience (General Osteopathic Council 2012)

8.0 - CONCLUSION 8.1 - RESEARCH QUESTION â&#x20AC;&#x153;Is patellofemoral pain syndrome associated with alterations in gluteus medius neuromuscular activation in the female population?â&#x20AC;?

8.2 - CONCLUSION The current understanding PFPS is limited by the lack of; prospective research, standardisation and homogeneity in the methods, sample sizes, and a definitive definition. The ambiguity of the current evidence cannot be distinguished as a product of these factors or the nature of PFPS. The conclusions of this review state that there is limited evidence to suggest shortened and delayed GMed activation timings (onset and duration) during balance tasks and gait, and increases in muscle activation amplitude during the propulsion phase of jumping in females. A potential amplitude oscillation was found during squat tasks; showing an increase during loaded knee flexion and decrease in the extension phase. This was identified through analysing the squat sub-theme data, and should be explored further in future research. Although theses association are suggested, the relationship between the findings cannot be established due to the limitations of cross-sectional studies; prospective research is needed to determine any cause and effect. When contrasting the findings with mixed gender PFPS research, both state significant alterations in GMed NMA during functional tasks, however the observed effects are different between genders (Barton et al 2013). This supports the need for future research to be gender specific, and for practitioners to specify their assessment and treatment to the individual. Further research is needed in the identification and potential aetiological significance of GMed NMA in females with PFPS.

8.3 – RECOMMENDATIONS FOR PRACTICE The findings of this review and the trend seen in the current research support the tissue homeostasis theory of PFPS. This is an important theory to be taken into clinical practice, and until research can elucidate a direct aetiology, this theory could determine the treatment and management of the syndrome (Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Lindström & Eriksson 2005). In understanding that the condition is a consequence of varying individualistic influences, the focus comes from the pathology and to the person. The current research states that the best treatment for PFPS is conservative therapy and exercise which is specified to the individual’s structural, functional, psychological and social state (Van der Heijden et al 2015, Hryvniak et al 2014, Sanchis-Alfonso 2014, Petersen et al 2013, Ferber et al 2011, Fagan & Delahunt 2008). It is here where the salutogenic model of care and osteopathic practice has the potential to be beneficial (Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Parsons & Marcer 2006, Lindström & Eriksson 2005, Antonovsky 1996). From the findings and discussions of this review, it is recommended that when assessing anterior knee pain, the biomechanics of the entire LEx during functional tasks should be addressed. Both hip and knee strengthening along with proprioceptive training should be considered in the rehabilitation plan, which should be dependent on the individual. Finally the practitioners should specify their approach and techniques to the social and psychological factors of the person behind the pathology (Sanchis-Alfonso 2014, Sanchis-Alfonso 2011, Parsons & Marcer 2006).

8.4 - REFLECTION The completion of this literature review has aided the researcher in understanding the process, skillset and editing of a piece of research. Before taking this project, research seemed like a definitive answer to the question it pursued. However by evaluating this subject area and the literature within it, the fallibility of evidence has become much more apparent. Any piece of research that is used to justify changes in practice or advice to patients must be fully explored. ‘A study says…’ will no longer suffice, as the amount of bias or extraneous variables in a method could be the only reason why the results are significant. In completing this piece of research I feel that I am now better equipped to evaluate the quality of the evidence, and the significance of their conclusions. In addition to this I am thankful for choosing such a complicated and common dysfunction, as it has already benefited my own practice in the diagnosis, treatment and management of individuals with patellofemoral pain. Word Count Excluding References, Tables and Appendices: 20,550

REFERENCES Aminaka, N., Pietrosimone, B. G., Armstrong, C. W., Meszaros, A., & Gribble, P. A. (2011). Patellofemoral pain syndrome alters neuromuscular control and kinetics during stair ambulation. Journal of Electromyography and Kinesiology, 21(4), 645-651. Andreacci, J. L., Lemura, L. M., Cohen, S. L., Urbansky, E. A., Chelland, S. A., & Duvillard, S. P. V. (2002). The effects of frequency of encouragement on performance during maximal exercise testing. Journal of sports sciences, 20(4), 345-352. Antonio, S., Wolfgang, G., Robert, H., Fullerton, B., & Carla, S. (2013). The anatomical and functional relation between gluteus maximus and fascia lata. Journal of bodywork and movement therapies, 17(4), 512-517. Antonovsky, A. (1996). The salutogenic model as a theory to guide health promotion. Health promotion international, 11(1), 11-18. Atkinson, M. (2011). Key concepts in sport and exercise research methods. Sage. Aveyard, H. (2010). Doing A Literature Review In Health And Social Care: A Practical Guide: A Practical Guide. McGraw-Hill International. Bartlett, J. L., Sumner, B., Ellis, R. G., & Kram, R. (2014). Activity and functions of the human gluteal muscles in walking, running, sprinting, and climbing. American journal of physical anthropology, 153(1), 124-131. Barton, C. J., Lack, S., Malliaras, P., & Morrissey, D. (2012). Gluteal muscle activity and patellofemoral pain syndrome: a systematic review. British journal of sports medicine, 47(4), 207. Barton, C. J., Levinger, P., Menz, H. B., & Webster, K. E. (2009). Kinematic gait characteristics associated with patellofemoral pain syndrome: a systematic review. Gait & posture, 30(4), 405-416. BioMed Central (2014) About us Available at: http://www.biomedcentral.com/about Bley, A. S., Correa, J. C. F., Dos Reis, A. C., Rabelo, N. D. D. A., Marchetti, P. H., & Lucareli, P. R. G. (2014). Propulsion Phase of the Single Leg Triple Hop Test in Women with Patellofemoral Pain Syndrome: A Biomechanical Study. PloS one, 9(5), e97606. Bolgla, L. A., Earl-Boehm, J., Emery, C., Hamstra-Wright, K., & Ferber, R. (2014). Comparison of hip and knee strength in males with and without patellofemoral pain. Physical Therapy in Sport. Bolgla, L. A., Malone, T. R., Umberger, B. R., & Uhl, T. L. (2011). Comparison of hip and knee strength and neuromuscular activity in subjects with and without patellofemoral pain syndrome. International journal of sports physical therapy, 6(4), 285. Bolgla, L. A., Malone, T. R., Umberger, B. R., & Uhl, T. L. (2010). Reliability of electromyographic methods used for assessing hip and knee neuromuscular activity in females diagnosed with patellofemoral pain syndrome. Journal of Electromyography and Kinesiology, 20(1), 142-147. Bolgla, L. A., & Uhl, T. L. (2007). Reliability of electromyographic normalization methods for evaluating the hip musculature. Journal of Electromyography and Kinesiology, 17(1), 102-111. Boling, M. C. (2008). A Prospective Investigation of Biomechanical Risk Factors for Anterior Knee Pain. ProQuest. Boling, M., Padua, D., Marshall, S., Guskiewicz, K., Pyne, S., & Beutler, A. (2010). Gender differences in the incidence and prevalence of patellofemoral pain syndrome. Scandinavian journal of medicine & science in sports, 20(5), 725-730. Boling, M. C., Padua, D. A., Marshall, S. W., Guskiewicz, K., Pyne, S., & Beutler, A. (2009). A prospective investigation of biomechanical risk factors for patellofemoral pain syndrome the joint undertaking to 94

monitor and prevent ACL injury (JUMP-ACL) cohort. The American journal of sports medicine, 37(11), 21082116. Boling, M. C., Bolgla, L. A., Mattacola, C. G., Uhl, T. L., & Hosey, R. G. (2006). Outcomes of a weight-bearing rehabilitation program for patients diagnosed with patellofemoral pain syndrome. Archives of physical medicine and rehabilitation, 87(11), 1428-1435. Bovi, G., Rabuffetti, M., Mazzoleni, P., & Ferrarin, M. (2011). A multiple-task gait analysis approach: kinematic, kinetic and EMG reference data for healthy young and adult subjects. Gait & posture, 33(1), 6-13. Bowling, A. (2009). Research methods in health: investigating health and health services. McGraw-Hill International. Brindle, T. J., Mattacola, C., & McCrory, J. (2003). Electromyographic changes in the gluteus medius during stair ascent and descent in subjects with anterior knee pain. Knee Surgery, Sports Traumatology, Arthroscopy, 11(4), 244-251. Brown, J. (2000). Physiotherapists' knowledge of patello-femoral pain syndrome. British Journal of Therapy and Rehabilitaiton, 7(8), 346-354. Buetow, S. (2007). Health research methods: A tabular presentation. Nova Science Pub Inc. Burden, A. (2010). How should we normalize electromyograms obtained from healthy participants? What we have learned from over 25years of research. Journal of Electromyography and Kinesiology, 20(6), 10231035. Burden, A. (2007). Surface electromyography. Biomechanical Evaluation of Movement in Sport and Exercise, 77. Burke, J. R. (2012). Effects of footwear comfort perception on the neuromuscular control of balance. International Journal of Neuroscience, 122(4), 209-220. Burns, N., & Grove, S. K. (2010). Understanding nursing research: Building an evidence-based practice. Elsevier Health Sciences. Carcia, C. R., & Martin, R. L. (2007). The influence of gender on gluteus medius activity during a drop jump. Physical Therapy in Sport, 8(4), 169-176. Chapman, A. L., Morgan, L. C., & Gartlehner, G. (2010). Semiâ&#x20AC;?automating the manual literature search for systematic reviews increases efficiency. Health Information & Libraries Journal, 27(1), 22-27. Chumanov, E. S., Wall-Scheffler, C., & Heiderscheit, B. C. (2008). Gender differences in walking and running on level and inclined surfaces. Clinical biomechanics, 23(10), 1260-1268. Ciliska, D., Thomas, H., & Buffett, C. (2008). A Compendium of Critical Appraisal Tools for Public Health Practice. links. Clark, D. I., Downing, N., Mitchell, J., Coulson, L., Syzpryt, E. P., & Doherty, M. (2000). Physiotherapy for anterior knee pain: a randomised controlled trial. Annals of the rheumatic diseases, 59(9), 700-704. Clarys, J. P., Scafoglieri, A., Tresignie, J., Reilly, T., & Van Roy, P. (2010). Critical appraisal and hazards of surface electromyography data acquisition in sport and exercise. Asian journal of sports medicine, 1(2), 69. The Cochrane Collaboration (2014) About us. Available at: https://www.cochrane.org/about-us Cody, W. K. (Ed.). (2011). Philosophical and theoretical perspectives for advanced nursing practice. Jones & Bartlett Publishers. Cohen, J. (2013). Statistical power analysis for the behavioral sciences. Routledge Academic. 95

Cook, D. A., & West, C. P. (2012). Conducting systematic reviews in medical education: a stepwise approach. Medical education, 46(10), 943-952. Cotton, A. (2013). Osteopathic principles in the modern world. International Journal of Osteopathic Medicine, 16(1), 17-24. Coughlan, M., Cronin, P., & Ryan, F. (2007). Step-by-step guide to critiquing research. Part 1: quantitative research. British Journal of Nursing, 16(11), 658-663. Cowan, S. M., Crossley, K. M., & Bennell, K. L. (2009). Altered hip and trunk muscle function in individuals with patellofemoral pain. British journal of sports medicine, 43(8), 584-588. Creswell, J. W. (2013). Research design: Qualitative, quantitative, and mixed methods approaches. Sage. Critical Appraisal Skills Programme (CASP)(2013) Case Control Study Checklist 11 questions to help you make sense of case control study. Available at: http://www.casp-uk.net/#!casp-tools-checklists/c18f8 Criswell, E. (2010). Cram's introduction to surface electromyography. Jones & Bartlett Publishers. Crossley, K. M. (2014). Is patellofemoral osteoarthritis a common sequela of patellofemoral pain?. British journal of sports medicine, 48(6), 409-410. Crowe, M., & Sheppard, L. (2011). A review of critical appraisal tools show they lack rigor: alternative tool structure is proposed. Journal of Clinical Epidemiology, 64(1), 79-89. Davis, I. S., & Powers, C. (Eds.). (2010). Patellofemoral Pain Syndrome: Proximal, Distal, and Local Factors— International Research Retreat, April 30–May 2, 2009, Baltimore, Maryland. Journal of Orthopaedic & Sports Physical Therapy, 40(3), A1-A48. Day, S. (2002). Important factors in surface EMG measurement. Calgary: Bortech Biomedical Ltd. Dayan, E., & Cohen, L. G. (2011). Neuroplasticity subserving motor skill learning. Neuron, 72(3), 443-454. DePoy, E., & Gitlin, L. N. (2011). Introduction to research understanding and applying multiple strategies. 4th edition. St.Louis missouri: Mosby. Des Jarlais, D. C., Lyles, C., Crepaz, N., & Trend Group. (2004). Improving the reporting quality of nonrandomized evaluations of behavioral and public health interventions: the TREND statement. American journal of public health, 94(3), 361. Dieterich, A. V., Pickard, C. M., Strauss, G. R., Deshon, L. E., Gibson, W., & McKay, J. (2014). Muscle thickness measurements to estimate gluteus medius and minimus activity levels. Manual Therapy. Distefano, L. J., Blackburn, J. T., Marshall, S. W., & Padua, D. A. (2009). Gluteal muscle activation during common therapeutic exercises. journal of orthopaedic & sports physical therapy, 39(7), 532-540. Doberstein, S. T., Romeyn, R. L., & Reineke, D. M. (2008). The diagnostic value of the Clarke sign in assessing chondromalacia patella. Journal of athletic training, 43(2), 190. Doménech, J., Sanchis-Alfonso, V., & Espejo, B. (2014). Changes in catastrophizing and kinesiophobia are predictive of changes in disability and pain after treatment in patients with anterior knee pain. Knee Surgery, Sports Traumatology, Arthroscopy, 1-6. Downs, S. H., & Black, N. (1998). The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. Journal of epidemiology and community health, 52(6), 377-384. Drake, R., Vogl, A. W., & Mitchell, A. W. (2009). Gray's anatomy for students. Elsevier Health Sciences.

Dye, S. F. (2005). The pathophysiology of patellofemoral pain: a tissue homeostasis perspective. Clinical orthopaedics and related research, 436, 100-110. Dye, S. F. (2001). Patellofemoral pain current concepts: an overview. Sports Medicine and Arthroscopy Review, 9(4), 264-272. Dykacz, J. M. (2005). Prospective and Retrospective Studies. Encyclopedia of Statistics in Behavioral Science. Earl, J. E., Hertel, J., & Denegar, C. R. (2005). Patterns of Dynamic Malalignment, Muscle Activation, Joint Motion, and Patellofemoral-Pain Syndrome. Journal of sport rehabilitation, 14(3). Earley, B. E., & Luce, H. (2010). An introduction to clinical research in osteopathic medicine. Primary Care: Clinics in Office Practice, 37(1), 49-64. EBSCO (2014) Nursing Resources from EBSCO Health. Available at: http://www.ebscohost.com/nursing/products/cinahl-databases/cinahl-complete Egger, M., Smith, G. D., & Altman, D. (Eds.). (2008). Systematic reviews in health care: meta-analysis in context. John Wiley & Sons. Elsevier (2014) Who uses ScienceDirect. Available at: http://www.elsevier.com/onlinetools/sciencedirect/who-uses-sciencedirect Emanuel, E. J., Grady, C. C., Crouch, R. A., Lie, R. K., Miller, F. G., & Wendler, D. D. (Eds.). (2011). The Oxford textbook of clinical research ethics. Oxford University Press. Fagan, V., & Delahunt, E. (2008). Patellofemoral pain syndrome: a review on the associated neuromuscular deficits and current treatment options. British journal of sports medicine, 42(10), 789-795. Faith, M. S., Butryn, M., Wadden, T. A., Fabricatore, A., Nguyen, A. M., & Heymsfield, S. B. (2011). Evidence for prospective associations among depression and obesity in populationâ&#x20AC;?based studies. Obesity Reviews, 12(5), e438-e453. Ferber, R., Kendall, K. D., & Farr, L. (2011). Changes in knee biomechanics after a hip-abductor strengthening protocol for runners with patellofemoral pain syndrome. Journal of athletic training, 46(2), 142. Ferber, R., McClay Davis, I., & Williams Iii, D. S. (2003). Gender differences in lower extremity mechanics during running. Clinical Biomechanics, 18(4), 350-357. Fereday, J., & Muir-Cochrane, E. (2008). Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International journal of qualitative methods, 5(1), 80-92. Ferrari, D., Kuriki, H. U., Silva, C. R., Alves, N., & MĂcolis de Azevedo, F. (2014). Diagnostic Accuracy of the Electromyography Parameters Associated With Anterior Knee Pain in the Diagnosis of Patellofemoral Pain Syndrome. Archives of physical medicine and rehabilitation. Flack, N. A. M. S., Nicholson, H. D., & Woodley, S. J. (2014). The anatomy of the hip abductor muscles. Clinical Anatomy, 27(2), 241-253. Foss, K. D. B., Hornsby, M., Edwards, N. M., Myer, G. D., & Hewett, T. E. (2012). Is body composition associated with an increased risk of developing anterior knee pain in adolescent female athletes?. The Physician and sportsmedicine, 40(1), 13. Fouladi, R., Rajabi, R., Naseri, N., Pourkazemi, F., & Geranmayeh, M. (2012). Menstrual cycle and knee joint position sense in healthy female athletes. Knee Surgery, Sports Traumatology, Arthroscopy, 20(8), 16471652.

Fredericson, M., & Yoon, K. (2006). Physical examination and patellofemoral pain syndrome. American journal of physical medicine & rehabilitation, 85(3), 234-243. Fulkerson, J. P., & Arendt, E. A. (2000). Anterior knee pain in females. Clinical orthopaedics and related research, 372, 69-73. Gardiner, P. (2011). Advanced neuromuscular exercise physiology. Human Kinetics. Garner, M., Wagner, C., & Kawulich, B. (Eds.). (2009). Teaching research methods in the social sciences. Ashgate Publishing, Ltd.. General Osteopathic Council (2006). Continuing professional development guidelines for osteopaths. London: GOsC. Available at: http://www.osteopathy.org.uk/practice/Continuing-professionaldevelopment/ General Osteopathic Council (2012). Osteopathic Practice Standards. London: GOsC Gillon, R. (1985). " Primum non nocere" and the principle of non-maleficence. British medical journal (Clinical research ed.), 291(6488), 130. Gobbi, A., Espregueira-Mendes, J., & Nakamura, N. (Eds.). (2014). The Patellofemoral Joint: State of the Art in Evaluation and Management. Springer. Goble, D. J., & Brown, S. H. (2009). Dynamic proprioceptive target matching behavior in the upper limb: effects of speed, task difficulty and arm/hemisphere asymmetries. Behavioural brain research, 200(1), 7-14. Goble, D. J., & Brown, S. H. (2007). Task-dependent asymmetries in the utilization of proprioceptive feedback for goal-directed movement. Experimental brain research, 180(4), 693-704. Goble, D. J., Lewis, C. A., & Brown, S. H. (2006). Upper limb asymmetries in the utilization of proprioceptive feedback. Experimental Brain Research, 168(1-2), 307-311. Google Scholar (2014)Stand on the shoulders of giants Available at: http://scholar.google.co.uk/intl/en/scholar/about.html Greenhalgh, T., & Peacock, R. (2005). Effectiveness and efficiency of search methods in systematic reviews of complex evidence: audit of primary sources. Bmj, 331(7524), 1064-1065. Grenholm, A., Stensdotter, A. K., & H채ger-Ross, C. (2009). Kinematic analyses during stair descent in young women with patellofemoral pain. Clinical Biomechanics, 24(1), 88-94. Griffin, L. Y., Agel, J., Albohm, M. J., Arendt, E. A., Dick, R. W., Garrett, W. E., ... & Wojtys, E. M. (2000). Noncontact anterior cruciate ligament injuries: risk factors and prevention strategies. Journal of the American Academy of Orthopaedic Surgeons, 8(3), 141-150. Guba, E. G., & Lincoln, Y. S. (1994). Competing paradigms in qualitative research. Handbook of qualitative research, 2, 163-194. Habermas, J. (1974) Theory and Practice. London, Heinemann. (cited in Holloway and Wheeler 2013) Halabchi, F., Mazaheri, R., & Seif-Barghi, T. (2013). Patellofemoral pain syndrome and modifiable intrinsic risk factors; how to assess and address?. Asian journal of sports medicine, 4(2), 85. Han, J., Anson, J., Waddington, G., & Adams, R. (2013). Proprioceptive performance of bilateral upper and lower limb joints: side-general and site-specific effects. Experimental brain research, 226(3), 313-323. Harry, J. R., Paquette, M. R., Caia, J., Townsend, R. J., Weiss, L. W., & Schilling, B. K. (2015). The Effects of Footwear Condition on Maximal Jumping Performance. The Journal of Strength & Conditioning Research. Hart, C. (2001). Doing a literature search: a comprehensive guide for the social sciences (1st edition). Sage 98

Hart, J. M., Pietrosimone, B., Hertel, J., & Ingersoll, C. D. (2010). Quadriceps activation following knee injuries: a systematic review. Journal of athletic training, 45(1), 87. Heiderscheit, B. C., Chumanov, E. S., Michalski, M. P., Wille, C. M., & Ryan, M. B. (2011). Effects of step rate manipulation on joint mechanics during running. Medicine and science in sports and exercise, 43(2), 296. Hermens, H. J., & Hutten, M. M. (2002). Muscle activation in chronic pain: its treatment using a new approach of myofeedback. International Journal of Industrial Ergonomics, 30(4), 325-336. Hibbs, A. E., Thompson, K. G., French, D. N., Hodgson, D., & Spears, I. R. (2011). Peak and average rectified EMG measures: Which method of data reduction should be used for assessing core training exercises?. Journal of Electromyography and Kinesiology, 21(1), 102-111. Hickson, M. (2013). Research Handbook for health care professionals. John Wiley & Sons. Hollander, M., Wolfe, D. A., & Chicken, E. (2013). Nonparametric statistical methods (Vol. 751). John Wiley & Sons. Holloway, I., & Wheeler, S. (2013). Qualitative research in nursing and healthcare. John Wiley & Sons. Hollman, J. H., Kolbeck, K. E., Hitchcock, J. L., Koverman, J. W., & Krause, D. A. (2006). Correlations between hip strength and static foot and knee posture. JSR, 15(1). Hong, Y., Wang, L., Li, J. X., & Zhou, J. H. (2012). Comparison of plantar loads during treadmill and overground running. Journal of Science and Medicine in Sport, 15(6), 554-560. Horsley, T., Dingwall, O., & Sampson, M. (2011). Checking reference lists to find additional studies for systematic reviews. Cochrane Database Syst Rev, 8. Hryvniak, D., Magrum, E., & Wilder, R. (2014). Patellofemoral Pain Syndrome: An Update. Current Physical Medicine and Rehabilitation Reports, 2(1), 16-24. Huston, R. (2008). Principles of biomechanics (1st edition). CRC press. Ireland, M. L., Willson, J. D., Ballantyne, B. T., & Davis, I. M. (2003). Hip strength in females with and without patellofemoral pain. Journal of orthopaedic & sports physical therapy, 33(11), 671-676. Jacobs, C. A., Uhl, T. L., Mattacola, C. G., Shapiro, R., & Rayens, W. S. (2007). Hip abductor function and lower extremity landing kinematics: sex differences. Journal of athletic training, 42(1), 76. Jacobs, C., Uhl, T. L., Seeley, M., Sterling, W., & Goodrich, L. (2005). Strength and fatigability of the dominant and nondominant hip abductors. Journal of athletic training, 40(3), 203. Jacobsen, K. (2011). Introduction to health research methods. Jones & Bartlett Publishers. Jensen, R., Hystad, T., & Baerheim, A. (2005). Knee function and pain related to psychological variables in patients with long-term patellofemoral pain syndrome. Journal of Orthopaedic & Sports Physical Therapy, 35(9), 594-600. Jesson, Matheson and Lacey (2011). Doing your literature review: traditional and systematic techniques. Sage. John, L. K., Loewenstein, G., & Prelec, D. (2012). Measuring the prevalence of questionable research practices with incentives for truth telling. Psychological science, 0956797611430953. Katrak P, Bialocerkowski A, Massy-Westropp N, Kumar V, Grimmer K (2004) A systematic review of the content of critical appraisal tools. BMC Medical research Methodology. 4 (22) 99

Karam, J. A., Higuera, C. A., Smith, E. B., & Sharkey, P. F. (2012). Patellofemoral Arthritis. Operative Techniques in Orthopaedics, 22(4), 196-202. Kawachi, I., & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban health, 78(3), 458-467. Khan, K., Kunz, R., Kleijnen, J., & Antes, G. (2011). Systematic reviews to support evidence-based medicine. CRC Press. Khuri, A. I., Mathew, T., & Sinha, B. K. (2011). Statistical tests for mixed linear models (Vol. 906). John Wiley & Sons. Kievit, A. J., Breugem, S. J., Sierevelt, I. N., Heesterbeek, P. J., van de Groes, S. A., Kremers, K. C., ... & Haverkamp, D. (2013). Dutch translation of the Kujala Anterior Knee Pain Scale and validation in patients after knee arthroplasty. Knee Surgery, Sports Traumatology, Arthroscopy, 21(11), 2647-2653. Kim, T. Y., Yoo, W. G., An, D. H., Oh, J. S., & Shin, S. J. (2013). The Effects of Different Gait Speeds and Lower Arm Weight on the Activities of the Latissimus Dorsi, Gluteus Medius, and Gluteus Maximus Muscles. Journal of physical therapy science, 25(11), 1483. Kinney, A. L., Besier, T. F., Silder, A., Delp, S. L., D'Lima, D. D., & Fregly, B. J. (2013). Changes in in vivo knee contact forces through gait modification. Journal of Orthopaedic Research, 31(3), 434-440. Kulkarni, A. V., Aziz, B., Shams, I., & Busse, J. W. (2009). Comparisons of citations in Web of Science, Scopus, and Google Scholar for articles published in general medical journals. JAMA, 302(10), 1092-1096. Kyröläinen, H., Avela, J., & Komi, P. V. (2005). Changes in muscle activity with increasing running speed. Journal of sports sciences, 23(10), 1101-1109. Lack, S., Barton, C., Woledge, R., Laupheimer, M., & Morrissey, D. (2014). The immediate effects of foot orthoses on hip and knee kinematics and muscle activity during a functional step-up task in individuals with patellofemoral pain. Clinical Biomechanics, 29(9), 1056-1062. Lankhorst, N. E., Bierma-Zeinstra, S. M., & van Middelkoop, M. (2012). Risk factors for patellofemoral pain syndrome: a systematic review. journal of orthopaedic & sports physical therapy, 42(2), 81-A12. Law, M., Stewart, D., Pollock, N., Letts, L., Bosch, J., & Westmorland, M. (1998a). Critical review formQuantitative studies. McMaster University: Occupational Therapy Evidence-Based Practice Research Group. Law, M., Stewart, D., Pollock, N., Letts, L., Bosch, J., & Westmorland, M. (1998b). Guidelines for critical review form–quantitative studies. McMaster University: Occupational Therapy Evidence-based Practice Research Group. Letourneau, N., & Allen, M. (1999). Post positivistic critical multiplism: a beginning dialogue. Journal of advanced nursing, 30(3), 623-630. Lewith, G. T., Jonas, W. B., & Walach, H. (2010). Clinical research in complementary therapies: principles, problems and solutions. Elsevier Health Sciences. Licciardone, J. C. (2007). Osteopathic research: elephants, enigmas, and evidence. Osteopathic medicine and primary care, 1(1), 7. Lieberman, D. E., Raichlen, D. A., Pontzer, H., Bramble, D. M., & Cutright-Smith, E. (2006). The human gluteus maximus and its role in running. Journal of Experimental Biology, 209(11), 2143-2155. Lindsay, T. R., Noakes, T. D., & McGregor, S. J. (2014). Effect of treadmill versus overground running on the structure of variability of stride timing. Perceptual and motor skills, 118(2), 331. Lindström, B., & Eriksson, M. (2005). Salutogenesis. Journal of Epidemiology and community health, 59(6), 440-442. 100

Livingston, L. A., & Mandigo, J. L. (1999). Bilateral Q angle asymmetry and anterior knee pain syndrome. Clinical biomechanics, 14(1), 7-13. Lobato, D. F. M., de Marche Baldon, R., Wun, P. Y. L., Santiago, P. R. P., & Serrão, F. V. (2013). Effects of the use of oral contraceptives on hip and knee kinematics in healthy women during anterior stair descent. Knee Surgery, Sports Traumatology, Arthroscopy, 21(12), 2823-2830. Lohman III, E. B., Balan Sackiriyas, K. S., & Swen, R. (2011). A comparison of the spatiotemporal parameters, kinematics, and biomechanics between shod, unshod, and minimally supported running as compared to walking. Physical Therapy in Sport, 12(4), 151-163. Lun, V., Meeuwisse, W. H., Stergiou, P., & Stefanyshyn, D. (2004). Relation between running injury and static lower limb alignment in recreational runners. British journal of sports medicine, 38(5), 576-580. Machi, L. A., & McEvoy, B. T. (2012). The literature review: Six steps to success (2nd Edition). Sage. de Marche Baldon, R., Nakagawa, T. H., Muniz, T. B., Amorim, C. F., Maciel, C. D., & Serrão, F. V. (2009). Eccentric hip muscle function in females with and without patellofemoral pain syndrome. Journal of athletic training, 44(5), 490. Marks, D. F. (2002). Perspectives on Evidence-Based Practice, Health Development Agency of the Public Health Evidence Steering Group, London. Massada, M., Aido, R., Magalhães, C., & Puga, N. (2011). Dynamic quadriceps angle: a comparison of female elite volleyball players with and without patellofemoral pain syndrome. British Journal of Sports Medicine, 45(6), 537-537. McCrory, J. L., Quick, N. E., Shapiro, R., Ballantyne, B. T., & McClay Davis, I. (2004). The effect of a single treatment of the Protonics™ system on biceps femoris and gluteus medius activation during gait and the lateral step up exercise. Gait & posture, 19(2), 148-153. McGinnis, P. (2013). Biomechanics of sport and exercise. Human Kinetics. Mills, K., Blanch, P., & Vicenzino, B. (2012). Comfort and midfoot mobility rather than orthosis hardness or contouring influence their immediate effects on lower limb function in patients with anterior knee pain. Clinical Biomechanics, 27(2), 202-208. Mizuno, Y., Kumagai, M., Mattessich, S. M., Elias, J. J., Ramrattan, N., Cosgarea, A. J., & Chao, E. (2001). Q-angle influences tibiofemoral and patellofemoral kinematics. Journal of Orthopaedic Research, 19(5), 834-840. Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Annals of internal medicine, 151(4), 264-269. Moseley, G. L., & Hodges, P. W. (2005). Are the changes in postural control associated with low back pain caused by pain interference?. The Clinical journal of pain, 21(4), 323-329. Moseley, G. L., Nicholas, M. K., & Hodges, P. W. (2004). Does anticipation of back pain predispose to back trouble?. Brain, 127(10), 2339-2347. Motealleh, A., Maroufi, N., Sarrafzadeh, J., & Sanjari, M. A. (2011). The test-retest reliability of the onset of core and vasti electromyographic activity while ascending and descending stairs in healthy controls and patellofemoral pain patients. Medical Journal of The Islamic Republic of Iran (MJIRI), 24(4), 221-231. Murphy, D. F., Connolly, D. A. J., & Beynnon, B. D. (2003). Risk factors for lower extremity injury: a review of the literature. British Journal of Sports Medicine, 37(1), 13-29. Myer, G. D., Ford, K. R., Barber Foss, K. D., Goodman, A., Ceasar, A., Rauh, M. J., ... & Hewett, T. E. (2010). The incidence and potential pathomechanics of patellofemoral pain in female athletes. Clinical Biomechanics, 25(7), 700-707. 101

Myer, G. D., Foss, K. D. B., Gupta, R., Hewett, T. E., & Ittenbach, R. F. (2014). Analysis of patient-reported anterior knee pain scale: implications for scale development in children and adolescents. Knee Surgery, Sports Traumatology, Arthroscopy, 1-8. Myers, T. (2001) Anatomy Trains Myofascial Meridians for Manual and Movement Therapists. Churchill Livingstone. Nakagawa, T. H., Moriya, É. T. U., Maciel, C. D., & Serrão, F. V. (2014). Test–retest reliability of threedimensional kinematics using an electromagnetic tracking system during single-leg squat and stepping maneuver. Gait & posture, 39(1), 141-146. Nakagawa, T. H., Moriya, É. T., Maciel, C. D., & Serrão, F. V. (2012). Trunk, pelvis, hip, and knee kinematics, hip strength, and gluteal muscle activation during a single-leg squat in males and females with and without patellofemoral pain syndrome. journal of orthopaedic & sports physical therapy, 42(6), 491-501. Nakagawa, T. H., Muniz, T. B., Baldon, R. M., Maciel, C. D., Amorim, C. F., & Serrão, F. V. (2011). Electromyographic preactivation pattern of the gluteus medius during weight-bearing functional tasks in women with and without anterior knee pain. Brazilian Journal of Physical Therapy, 15(1), 59-65. NCBI (2014) Pubmed. Available at: http://www.ncbi.nlm.nih.gov/pubmed. National Council for Osteopathic Research (2007). Research Governance Framework for Osteopaths. NCOR [Online]. Available at: www.ncor.org.uk Nunes, G. S., Stapait, E. L., Kirsten, M. H., de Noronha, M., & Santos, G. M. (2013). Clinical test for diagnosis of patellofemoral pain syndrome: Systematic review with meta-analysis. Physical Therapy in Sport, 14(1), 5459. O'Sullivan, K., Herbert, E., Sainsbury, D., McCreesh, K., & Clifford, A. (2012). No difference in gluteus medius activation in women with mild patellofemoral pain. Journal of sport rehabilitation, 21(2), 110. Ogasawara, I., Miyakawa, S., & Wakitani, S. (2014). Gender difference in neuromuscular hip and knee joint control during single-leg landing. Mukogawa Women's University. Ott, B., Cosby, N. L., Grindstaff, T. L., & Hart, J. M. (2011). Hip and knee muscle function following aerobic exercise in individuals with patellofemoral pain syndrome. Journal of Electromyography and Kinesiology, 21(4), 631-637. Oxford Brookes University (2011a) Databases for Health and Social Care. Available at: http://www.brookes.ac.uk/library/health/healthdb.html Harper A, Hunt J, Williams K and Woolliams M (2011) How to reference for health and social care students (7th edition). Oxford: Oxford Brookes University. Palastanga, N., Field, D., & Soames, R. (2006). Anatomy and human movement: structure and function (Vol. 20056). Elsevier Health Sciences. Pannucci, C. J., & Wilkins, E. G. (2010). Identifying and avoiding bias in research. Plastic and reconstructive surgery, 126(2), 619. Pappas, E., Hagins, M., Sheikhzadeh, A., Nordin, M., & Rose, D. (2007). Biomechanical differences between unilateral and bilateral landings from a jump: gender differences. Clinical Journal of Sport Medicine, 17(4), 263-268. Pappas, E., & Wong-Tom, W. M. (2012). Prospective Predictors of Patellofemoral Pain Syndrome A Systematic Review With Meta-analysis. Sports Health: A Multidisciplinary Approach, 4(2), 115-120. Park, S., Horak, F. B., & Kuo, A. D. (2004). Postural feedback responses scale with biomechanical constraints in human standing. Experimental Brain Research, 154(4), 417-427. 102

Parsons, J., & Marcer, N. (2006). Osteopathy: models for diagnosis, treatment and practice. Elsevier Health Sciences. Patel, V., Flisher, A. J., Hetrick, S., & McGorry, P. (2007). Mental health of young people: a global publichealth challenge. The Lancet, 369(9569), 1302-1313. Pattyn, E., Mahieu, N., Selfe, J., Verdonk, P., Steyaert, A., & Witvrouw, E. (2012). What predicts functional outcome after treatment for patellofemoral pain?. Medicine and science in sports and exercise, 44(10), 18271833. Petersen, W., Ellermann, A., Gösele-Koppenburg, A., Best, R., Rembitzki, I. V., Brüggemann, G. P., & Liebau, C. (2013). Patellofemoral pain syndrome. Knee Surgery, Sports Traumatology, Arthroscopy, 1-11. Petersen, J., Sørensen, H., & Nielsen, R. O. (2015). The cumulative loads increase in the knee joint at slowspeed running compared with faster running: A biomechanical study. Journal of Orthopaedic & Sports Physical Therapy, (Early Access), 1-22. Petty, N. J., Thomson, O. P., & Stew, G. (2012). Ready for a paradigm shift? Part 1: Introducing the philosophy of qualitative research. Manual therapy, 17(4), 267-274. Physiotherapy Evidence Database (2014) Available at: http://www.pedro.org.au/ Polgar, S., & Thomas, S. A. (2013). Introduction to research in the health sciences. Elsevier Health Sciences. Polit, D. F., & Beck, C. T. (2013). Essentials of nursing research. Lippincott Williams & Wilkins. Pope, C., & Mays, N. (1995). Qualitative research: reaching the parts other methods cannot reach: an introduction to qualitative methods in health and health services research. Bmj, 311(6996), 42-45. Pope, C., Mays, N., & Popay, J. (2007). Synthesising Qualitative and Quantitative Health Evidence: A Guide to Methods: A Guide to Methods. McGraw-Hill International. Post, W. R. (1999). Current concepts clinical evaluation of patients with patellofemoral disorders. Arthroscopy: The Journal of Arthroscopic & Related Surgery, 15(8), 841-851. Powers, C. M., Bolgla, L. A., Callaghan, M. J., Collins, N., & Sheehan, F. T. (2012). Patellofemoral Pain: Proximal, Distal, and Local Factors—2nd International Research Retreat, August 31–September 2, 2011, Ghent, Belgium. journal of orthopaedic & sports physical therapy, 42(6). Powers, C. M. (2010). The influence of abnormal hip mechanics on knee injury: a biomechanical perspective. J Orthop Sports Phys Ther, 40(2), 42-51. Radzimski, A. O., Mundermann, A., & Sole, G. (2012). Effect of footwear on the external knee adduction moment・A systematic review. The Knee, 19(3), 163-175. Rasool, G., Iqbal, K., & White, G. A. (2012). Myoelectric activity detection during a Sit-to-Stand movement using threshold methods. Computers & Mathematics with Applications, 64(5), 1473-1483. Rathleff, C. R., Baird, W. N., Olesen, J. L., Roos, E. M., Rasmussen, S., & Rathleff, M. S. (2013). Hip and knee strength is not affected in 12-16 year old adolescents with patellofemoral pain-a cross-sectional population-based study. PloS one, 8(11), e79153. Razali, N. M., & Wah, Y. B. (2011). Power comparisons of shapiro-wilk, kolmogorov-smirnov, lilliefors and anderson-darling tests. Journal of Statistical Modeling and Analytics, 2(1), 21-33.

103

Reuters, T. (2008). Web of science. Online factsheet Thomson Reuters, Philadelphia, Pennsylvania (Available from: www. thomsonreuters. com/content/PDF/scientific/Web_of_Science_factshe et. pdf). Ridley D (2008) The Literature Review: A step-to- step guide for students (1st Edition). London: SAGE publications Ltd. Rodriguez-Fontenla, C., López-Golán, Y., Calaza, M., Pombo-Suarez, M., Gómez-Reino, J. J., & González, A. (2012). Genetic risk load and age at symptom onset of knee osteoarthritis. Journal of Orthopaedic Research, 30(6), 905-909. Ross, T. (2012). A survival guide for health research methods. McGraw-Hill International. Rourke, L., & Anderson, T. (2004). Validity in quantitative content analysis. Educational Technology Research and Development, 52(1), 5-18. Saad, M. C., Felício, L. R., Masullo, C. D. L., Liporaci, R. F., & Bevilaqua-Grossi, D. (2011). Analysis of the center of pressure displacement, ground reaction force and muscular activity during step exercises. Journal of Electromyography and Kinesiology, 21(5), 712-718. Sainburg, R. L., (2005) Handedness: differential specializations for control of trajectory and position. Exerc Sport Sci Rev 33:206–213 Sainburg, R. L., (2002) Evidence for a dynamic-dominance hypothesis of handedness. Exp Brain Res 142:241–258 Saks, M., & Allsop, J. (Eds.). (2012). Researching health: qualitative, quantitative and mixed methods. Sage. Sale, J. E., Lohfeld, L. H., & Brazil, K. (2002). Revisiting the quantitative-qualitative debate: Implications for mixed-methods research. Quality and quantity, 36(1), 43-53. Sanchis-Alfonso, V. (2014). Holistic approach to understanding anterior knee pain. Clinical implications. Knee Surgery, Sports Traumatology, Arthroscopy, 1-11. Sanchis-Alfonso, V. (2011). Background: Patellofemoral Malalignment versus Tissue Homeostasis. In Anterior knee pain and patellar instability (pp. 3-19). Springer London. Sanderson, S., Tatt, I. D., & Higgins, J. P. (2007). Tools for assessing quality and susceptibility to bias in observational studies in epidemiology: a systematic review and annotated bibliography. International journal of epidemiology, 36(3), 666-676. Sasaki, K., & Neptune, R. R. (2006). Differences in muscle function during walking and running at the same speed. Journal of biomechanics, 39(11), 2005-2013. Schardt, C., Adams, M. B., Owens, T., Keitz, S., & Fontelo, P. (2007). Utilization of the PICO framework to improve searching PubMed for clinical questions. BMC medical informatics and decision making, 7(1), 16. Scott, L. A., Murley, G. S., & Wickham, J. B. (2012). The influence of footwear on the electromyographic activity of selected lower limb muscles during walking. Journal of Electromyography and Kinesiology, 22(6), 1010-1016. Selfe, J., Callaghan, M., Witvrouw, E., Richards, J., Dey, M. P., Sutton, C., ... & Turner, D. (2013). Targeted interventions for patellofemoral pain syndrome (TIPPS): classification of clinical subgroups. BMJ open, 3(9), e003795. Selfe, J., Harper, L., Pedersen, I., Breen-Turner, J., & Waring, J. (2001). Four outcome measures for patellofemoral joint problems: part 1. development and validity. Physiotherapy, 87(10), 507-515.

104

Semciw, A. I., Pizzari, T., Murley, G. S., & Green, R. A. (2013). Gluteus medius: An intramuscular EMG investigation of anterior, middle and posterior segments during gait. Journal of Electromyography and Kinesiology, 23(4), 858-864. SENIAM (2006) Recommendations for sensor locations in hip or upper leg muscles. Available at: http://www.seniam.org/ Shirazi, Z., Biabani Moghaddam, M., & Motealleh, A. (2014). Comparative Evaluation of Core Muscle Recruitment Pattern in Response to Sudden External Perturbations in Patients With Patellofemoral Pain Syndrome and Healthy Subjects. Archives of Physical Medicine and Rehabilitation. Simon, A. M., & Ferris, D. P. (2008). Lower limb force production and bilateral force asymmetries are based on sense of effort. Experimental brain research, 187(1), 129-138. Sinclair, J., McCarthy, D., Bentley, I., Hurst, H. T., & Atkins, S. (2014). The influence of different footwear on 3-D kinematics and muscle activation during the barbell back squat in males. European journal of sport science, (ahead-of-print), 1-8. Sinclair, J., Richards, J., Taylor, P. J., Edmundson, C. J., Brooks, D., & Hobbs, S. J. (2013). Three-dimensional kinematic comparison of treadmill and overground running. Sports Biomechanics, 12(3), 272-282. Sinnott, R., Jiang, J., Watt, J., & Ajayi, O. (2006). Shibboleth-based access to and usage of grid resources. In Proceedings of the 7th IEEE/ACM International Conference on Grid Computing (pp. 136-143). IEEE Computer Society. Smith, C., Andre, K., Lewis, K., & Parker, B. (2009). Evidence to inform nursing practice: An applied approach. Using Evidence to Guide Nursing Practice, 151. Song, C. Y., Huang, H. Y., Chen, S. C., Lin, J. J., & Chang, A. H. (2014). Effects of femoral rotational taping on pain, lower extremity kinematics, and muscle activation in female patients with patellofemoral pain. Journal of Science and Medicine in Sport. Souza, R. B., & Powers, C. M. (2009). Differences in hip kinematics, muscle strength, and muscle activation between subjects with and without patellofemoral pain. journal of orthopaedic & sports physical therapy, 39(1), 12-19. Stegeman, D., & Hermens, H. (2007). Standards for surface electromyography: The European project Surface EMG for non-invasive assessment of muscles (SENIAM). Stensdotter, A. K., Guerra, J. B., & H채ger-Ross, C. (2009). Limb support in response to balance provocations in women with patellofemoral pain. Advances in Physiotherapy, 11(2), 97-103. Stevens, J. P. (2012). Applied multivariate statistics for the social sciences. Routledge. Stone, R. J., & Stone, J. A. (2003). Atlas of skeletal muscles.(7th Ed) McGraw-Hill. Taunton, J. E., Ryan, M. B., Clement, D. B., McKenzie, D. C., Lloyd-Smith, D. R., & Zumbo, B. D. (2002). A retrospective case-control analysis of 2002 running injuries. British journal of sports medicine, 36(2), 95101. Thomson, O. P., Petty, N. J., & Moore, A. P. (2013). Reconsidering the patient-centeredness of osteopathy. International Journal of Osteopathic Medicine, 16(1), 25-32. Toumi, H., Best, T. M., Pinti, A., Lavet, C., Benhamou, C. L., & Lespessailles, E. (2013). The role of muscle strength & activation patterns in patellofemoral pain. Clinical Biomechanics, 28(5), 544-548. Tyler, T. F., Nicholas, S. J., Mullaney, M. J., & McHugh, M. P. (2006). The role of hip muscle function in the treatment of patellofemoral pain syndrome. The American journal of sports medicine, 34(4), 630-636. 105

Van Caekenberghe, I., Segers, V., Willems, P., Gosseye, T., Aerts, P., & De Clercq, D. (2013). Mechanics of overground accelerated running vs. running on an accelerated treadmill. Gait & posture, 38(1), 125-131. Van Cant, J., Pineux, C., Pitance, L., & Feipel, V. (2014). Hip Muscle Strength and Endurance In Females With Patellofemoral Pain: A Systematic Review With Meta-synthesis. International journal of sports physical therapy, 9(5), 564. Van der Heijden, R. A., Lankhorst N. E., van Linschoten R., Bierma-Zeinstra S.M., & van Middelkoop M. (2015) Exercise for treating patellofemoral pain syndrome. Cochrane Database of Systematic Reviews, (1). Wallace, D. A., Salem, G. J., Salinas, R., & Powers, C. M. (2002). Patellofemoral joint kinetics while squatting with and without an external load. Journal of Orthopaedic & Sports Physical Therapy, 32(4), 141-148. Waryasz, G. R., & McDermott, A. Y. (2008). Patellofemoral pain syndrome (PFPS): a systematic review of anatomy and potential risk factors. Dynamic medicine, 7(1), 9. Whatman, C., Hume, P., & Hing, W. (2013). Kinematics during lower extremity functional screening tests in young athletesâ&#x20AC;&#x201C;Are they reliable and valid?. Physical Therapy in Sport, 14(2), 87-93. White, L. C., Dolphin, P., & Dixon, J. (2009). Hamstring length in patellofemoral pain syndrome. Physiotherapy, 95(1), 24-28. Whyte, E. F., Moran, K., Shortt, C. P., & Marshall, B. (2010). The influence of reduced hamstring length on patellofemoral joint stress during squatting in healthy male adults. Gait & posture, 31(1), 47-51. Willson, J. D., & Davis, I. S. (2008). Lower extremity mechanics of females with and without patellofemoral pain across activities with progressively greater task demands. Clinical Biomechanics, 23(2), 203-211. Willson, J. D., & Davis, I. S. (2009). Lower extremity strength and mechanics during jumping in women with patellofemoral pain. Willson, J. D., Kernozek, T. W., Arndt, R. L., Reznichek, D. A., & Scott Straker, J. (2011). Gluteal muscle activation during running in females with and without patellofemoral pain syndrome. Clinical Biomechanics, 26(7), 735-740. Willson, J. D., Petrowitz, I., Butler, R. J., & Kernozek, T. W. (2012). Male and female gluteal muscle activity and lower extremity kinematics during running. Clinical Biomechanics, 27(10), 1052-1057. Witvrouw, E., Callaghan, M. J., Stefanik, J. J., Noehren, B., Bazett-Jones, D. M., Willson, J. D., ... & Crossley, K. M. (2014). Patellofemoral pain: consensus statement from the 3rd International Patellofemoral Pain Research Retreat held in Vancouver, September 2013. British journal of sports medicine, 48(6), 411-414. Witvrouw, E., Werner, S., Mikkelsen, C., Van Tiggelen, D., Berghe, L. V., & Cerulli, G. (2005). Clinical classification of patellofemoral pain syndrome: guidelines for non-operative treatment. Knee Surgery, Sports Traumatology, Arthroscopy, 13(2), 122-130. Wong, Y. M. (2009). Recording the vastii muscle onset timing as a diagnostic parameter for patellofemoral pain syndrome: Fact or fad?. Physical Therapy in Sport, 10(2), 71-74. Xu, Q., Quan, Y., Yang, L., & He, J. (2013). An adaptive algorithm for the determination of the onset and offset of muscle contraction by EMG signal processing. Neural Systems and Rehabilitation Engineering, IEEE Transactions on, 21(1), 65-73. Yang, J. F., & Winter, D. A. (1984). Electromyographic amplitude normalization methods: improving their sensitivity as diagnostic tools in gait analysis. Arch Phys Med Rehabil, 65(9), 517-21. 106

Yard, E., & Comstock, D. (2011). Injury patterns by body mass index in US high school athletes. Journal of physical activity and health, 8(2), 182. Yen, Y. M. (2014). Assessment and Treatment of Knee Pain in the Child and Adolescent Athlete. Pediatric Clinics of North America, 61(6), 1155-1173. Zammito, J. H. (2004). A nice derangement of epistemes: Post-positivism in the study of science from Quine to Latour. University of Chicago Press. Zazulak, B. T., Hewett, T. E., Reeves, N. P., Goldberg, B., & Cholewicki, J. (2007). Deficits in neuromuscular control of the trunk predict knee injury risk a prospective biomechanical-epidemiologic study. The American journal of sports medicine, 35(7), 1123-1130. Zeller, B. L., McCrory, J. L., Kibler, W. B., & Uhl, T. L. (2003). Differences in Kinematics and Electromyographic Activity Between Men and Women during the Single-Legged Squat*. The American journal of sports medicine, 31(3), 449-456. Zhu, B., Xu, Z., & Xia, M. (2012). Hesitant fuzzy geometric Bonferroni means. Information Sciences, 205, 7285.

107

APPENDICES APPENDIX A

Standards for surface electromyography: the European project "Surface EMG for non-invasive assessment of muscles (SENIAM)” D.F. Stegeman1,3, H.J. Hermens2 1

Institute of Neurology, Department of Clinical Neurophysiology, University Medical Centre Nijmegen, and Graduate Institute for Fundamental and Clinical Human Movement Sciences, 2Roessingh Research and Development, Enschede, The Netherlands, 3FB Motorik, Institute for Pathophysiology, Friedrich-SchillerUniversity Jena, Germany

The SENIAM (surface EMG for non-invasive assessment of muscles) project was organised as a European concerted action, financed in the context of the Biomed 2 program of the European Community (1996 – 1999). The objectives of the project were to integrate basic and applied research on surface EMG (sEMG) at a European level, to establish European cooperation and to solve key items that presently prevent a useful exchange of data and clinical experience. The key topics of this joint effort were: (1) sEMG sensors and sensor placementprocedures, (2) sEMG signal processing and (3) sEMG modelling. These three topics were handled according to a specific scheme in which first an inventory was made by literature searches or exchanges, from which the state of the art was defined. When necessary, additional evaluation tasks were defined. Sets of recommendations and/or test facilities were the final result for each of these topics. The paradoxical problem with sEMG is that it is one of the easiest electrophysiological signals to measure, but also one of the hardest to interpret quantitatively. With an oscilloscope and two metal objects connected to the oscilloscope’s input one will often display sEMG without notice. “To its detriment, electromyography is too easy to use and consequently too easy to abuse” (DeLuca, 1997). The SENIAM project brought together the expertise of 16 European groups working on developments and applications of sEMG. The initiative came from and the project coordination was done by Roessingh Research and Development (Hermie Hermens and Bart Freriks, Enschede, NL). The project management group further consisted of Roberto Merletti (Torino, IT), Cathy Disselhorst-Klug and Günther Rau (Aachen, D), Goran Hägg (Solna, S) and Dick Stegeman (Nijmegen, NL). The project has resulted in: (1) recommendations for electrode design and placement procedures, (2) recommendations for recording and processing of sEMG signals, (3) a set of 4 computer models for better insight in sEMG generation, (4) a set of reference signals, (5) eight books with various contributions of the participants of the project, (6) the SENIAM CD-ROM, and (7) the SENIAM club, a European network on sEMG with over 100 members. The last book (SENIAM 8, ISBN 90-754562-15-2) and the CD-ROM (ISBN 90-754552-144) contain the European SENIAM recommendations for sEMG especially in kinesiological applications and are easily available. In the following sections, important recommendations with respect to (only) bipolar sensor design , placement and signal conditioning principles are given. The recommendations are partly based on sEMG model simulation results. The modelling task is not treated separately

108

in this contribution. Details can be found in the SENIAM deliverables listed at the end of this contribution. Spatial sensor characteristics and sensor placement procedures The term sensor, instead of electrode, is used to stress the fact that each sEMG measurement needs an ensemble of (at least 2) single electrodes, often housed together, whereby the preamplifier might be integrated in the housing. In most practical situations, a bipolar electrode montage is used which means that a single sEMG signal is recorded as the electric potential difference between two electrodes over the muscle which are relatively closely spaced in comparison to the muscle’s dimensions. A bipolar sEMG recording is influenced by the electrode shapes, sizes, positions, orientations and the inter electrode distance (IED). Electrode shape: To begin with a “weak” recommendation: with respect to electrode shape no clear and well defendable standard was recommended by SENIAM, apart from what is said below on electrode size in different directions with respect to the muscle fibres. No major influence on the sEMG signal is expected from taking different shapes (e.g. square vs. circular). Users should however clearly indicate the shape of the electrodes used. Electrode size: In current sEMG practice, electrode size varies between 1 mm2 until a few cm2. Different from the electrode shape, the electrode size clearly influences sEMG signals. It should be realised that the potential recorded by a finite size electrode can be conceived of as the average over the electrode surface of the potential actually present under that electrode surface. Upon an increase of the size perpendicular to the muscle fibres (e.g. in case of a rectangular bar electrode transversally over the muscle), the view of the electrodes increases. No absolute quantitative data on this effect per muscle have been studied. Upon an increase of the size in the direction of the muscle fibres, it can be shown theoretically and experimentally that this mainly has a low pass filtering effect on the sEMG signal. It is recommended that the size of the electrodes in the direction of the muscle fibres should not exceed 10 mm. The European inventory showed that circular electrodes with a diameter of 10 mm are preferred. Figure 1: Example of SENIAM recommendation. Placement of bipolar electrodes and reference electrode for the tibial anterior muscle (from: SENIAM 8)

Sensor position: In SENIAM 8 and 9, recommendations for the location of sensors for a number of often studied muscles are given (e.g. Figure 1). They are derived from the following principle: with respect to the longitudinal (in fibre direction) location of the sensor on the muscle, it is recommended to place the sensor halfway the (most) distal motor endplate zone and the distal tendon. With respect to the transversal location of the sensor on the muscle, it is recommended to locate the sensor at the surface away from the ‘edge’ with other subdivisions or other muscles so that the geometrical distance to other muscles is maximised.

109

Sensor orientation and IED: It is advisedto place the bipolar sEMG electrodes around the optimal sensor location directed parallel to the muscle fibres. The recommendations for sensor locations for individual muscles in SENIAM 8 also give an advise for the orientation of the bipolar sEMG electrodes for each individual muscle. It is preferred to apply bipolar sEMG sensors with an IED of 20 mm, because a maximal sEMG amplitude is expected with this IED. This recommendation is, among others, based on model simulation studies. If bipolar electrodes are applied on relatively small muscles, IED should not exceed 1/4 of the muscle fibre length. In this way unstable recordings, due to tendon and motor endplate effects can be avoided. Concluding remark: The above recommendations for electrode size, sensor orientation and IED are derived from considerations on sEMG signal amplitude and the representativeness of the signal for a muscle as a whole. In case of other aims, like an increased selectivity of the sensor for superficial motor units, other IED (e.g. 3-5 mm) and electrode size values may be optimal. Electrode material and sensor construction The electrode material, which forms the contact layer with the skin, needs to realise a good electrode skin contact, a low electrode-skin impedance and a â&#x20AC;&#x2DC;stationaryâ&#x20AC;&#x2122; behaviour in time (that is with respect to impedance and chemical reactions at the skin interface). In case of sEMG , the electrode material is less critical than in case of brain electric signals (EEG) where lower signal frequencies (around or below 1 Hz, see next section) must be detected. An inventory has shown that different types of material are used, mostly Ag/AgCl, AgCl, Ag, Au, of which the Ag/AgCl electrodes are most common. They provide a stable transition with low noise and are easily available commercially. Electrodes are mostly combined with electrode gel. Both pre-gelled and non-gelled electrodes are commercially available. Electrode gel and paste are used to reduce the electrode-skin impedance. A low impedance gives stable recordings and low electrode noise levels. Since use of non-gelled electrodes is cumbersome and time consuming, pre-gelled electrodes are recommended, although the performance of pre-gelled and non-gelled electrodes is similar. It is not expected that the sensor construction (and its mass) do directly effect sEMG characteristics. There are nevertheless two important indirect effects which can disturb or interfere with the recorded sEMG pattern. First, if the construction of the sensor is such that IED can vary during muscle contraction, this will artificially modulate the amplitude, shape and width of the action potentials and will consequently affect both sEMG amplitude and frequency characteristics. Second, if the construction of the sensor is such that electrodes and cables can move due to pulling of cables or inertia of the construction, there is the potential risk for movement artefacts, because of destabilisation of the electric layer and changes of impedances and magnetically induced currents in the cables. It is therefore recommended that a sensor construction with fixed inter electrode distance is used, made of light weight material. Cables need to be fixed using (double sided) tape or elastic band in such a manner that pulling artefacts can be avoided. If in fast dynamic contractions the sensor causes too much (movement) artefacts (due to the inertia of the construction), it is recommended to fix the inter electrode distance using (double sided) tape or rings.

110

SEMG signal conditioning and processing In Figure 2 the basic principles of sEMG signal conditioning and digital computer acquisition are depicted. At the ‘low frequency’ side of the signal spectrum the choice for a high pass filter is mostly determined by the need to remove slow variations in the signal caused by the movement artefacts mentioned. It is obvious that the situation in case of a dynamic movement studies is much more vulnerable to these disturbances. In most cases a high pass filter between 10 and 20 Hz will preserve the important frequencies in the sEMG. However, in the 5-20 Hz frequency range, the sEMG spectrum contains information

Electrod es

Low noise, high input impedance amplifier

Low-pass (500-1000 Hz) and high-pass (10-20 Hz) filter

Sample and 16 bits A/D convertion

Figure 2: schematic outline of an sEMG recording system

concerning the firing rates of the active motor units, but in many cases (e.g. in movement analysis) this information may not be of great interest. Note that sudden signal changes due to movements may not be completely attenuated by a 10 – 20 Hz filter. With the exception of a few cases, about 95% of sEMG power is accounted for by harmonics up to 400 Hz and most of the remaining percents by electrode and equipment noise. A low pass filter has to be applied to the signal to further attenuate these unwanted components. The cut-off frequency is usually chosen near to 500 Hz and the sampling frequency must then be 1000 samples/second or higher (sampling theorem: sampling must be >2 times the highest frequency in the signal). The digitisation requirement for the sEMG signal is dependent on the smallest and the highest amplitudes expected. The delectability of low amplitude details of the sEMG signal is limited by the noise level of the amplifier system. Modern amplifiers have a noise level of a few Vs. Therefore, a digitisation of about 0.5 V/bit is sufficiently accurate. A 16 bit A/D convertor has 216-1 = 65535 levels, meaning that without gain adaptations the range of measurements with such a convertor is ± 16 mV, which is enough for almost any sEMG application. SEMG signal analysis In the context of this contribution it is not possible to go into details with respect to signal processing after the sEMG has been captured in digital form. The possibilities are virtually unlimited. Most often used are amplitude estimation (RMS, mean after rectification), spectral analysis (after Fourier transformation) and the measurement of muscle fibre velocities. An important distinction again has to be made between dynamic and non-dynamic muscle contractions. Different classes of signal processing tools have to be selected because many well known tools are based on the assumption of stationarity, which means that the signal characteristics are assumed constant over time. For dynamic conditions such requirement is per definition not met. Especially in case of the much applied spectral analysis of the sEMG signal, the stationarity condition has to be evaluated. Conclusion The above gives a summary of the SENIAM concerted action within the limited space of this contribution. The reader should feel free to contact the authors for further information or to

111

acquire a selection of the SENIAM deliverables (www.rrd.nl; h.hermens@rrd.nl). We are convinced that the goal of reaching full maturity of sEMG as a scientifically founded set of experimental methods in the study of human movement came closer by the SENIAM initiative. References [SENIAM 1] European Activities on Surface ElectroMyoGraphy, Proceedings of the first general SENIAM workshop, Torino, Italy, September 1996, eds. H. J. Hermens, R. Merletti and B. Freriks, Roessingh Research and Development b.v., 1996, ISBN 90-75452-05-5. [SENIAM 2] European Applications on Surface ElectroMyoGraphy, Proceedings of the second general SENIAM workshop, Stockholm, Sweden, June 1997, eds. H.J. Hermens, G. Hägg, B. Freriks, Roessingh Research and Development b.v., 1997, ISBN 90-75452-06-3. [SENIAM 3] Surface ElectroMyoGraphy Application Areas and Parameters, Proceedings of the third general SENIAM workshop, Aachen, Germany, May 1998, eds. H.J. Hermens, G. Rau, C. Disselhorst-Klug, B. Freriks, Roessingh Research and Development b.v., 1998, ISBN 90-75452-10-1. [SENIAM 4] Future applications of Surface ElectroMyoGraphy, Proceedings of the fourth general SENIAM workshop, ’s Hertogenbosch, September 1999, eds. H.J. Hermens, B. Freriks, Roessingh Research and Development b.v., 1999, ISBN: 90-75452-16-0. [SENIAM 5] The State of the Art on Sensors and Sensor Placement Procedures for Surface ElectroMyoGraphy: A proposal for sensor placement procedures, deliverable of the SENIAM project, eds. H.J. Hermens, B. Freriks, Roessingh Research and Development b.v., 1997, ISBN 90-75452-09-8. [SENIAM 6] The state of the Art on Modelling Methods for Surface ElectroMyoGraphy, deliverable of the SENIAM project, eds. H.J. Hermens, D. Stegeman, J. Blok and B. Freriks, Roessingh Research and Development b.v., 1998, ISBN 90-75452-11-X. [SENIAM 7] The state of the Art on Signal Processing Methods for Surface ElectroMyoGraphy, deliverable of the SENIAM project, eds. H.J. Hermens, R. Merletti, B. Freriks, Roessingh Research and Development b.v., 1999, ISBN: 90-75452-17-9. [SENIAM 8] European Recommendations for Surface ElectroMyoGraphy, deliverable of the SENIAM project, authors: H.J. Hermens, B. Freriks, R. Merletti, G. G. Hägg, D. Stegeman, J. Blok, G. Rau, C. Disselhorst-Klug, Roessingh Research and Development b.v., 1999, ISBN: 90-75452-15-2. [SENIAM 9] European Recommendations for Surface ElectroMyoGraphy, results of the SENIAM project, authors: B. Freriks, H.J. Hermens, Roessingh Research and Development b.v., 1999, ISBN: 90-75452-14-4 (CD-rom). De Luca CJ. The use of electromyography in biomechanics. J Appl Biomech. 1997; 13:135-163.

Contact address Dick F. Stegeman (d.stegeman@czzoknf.azn.nl) 314 Department of Clinical Neurophysiology Institute of Neurology, University Medical Centre Nijmegen P.O. Box 9101, NL-6500HB Nijmegen

112

APPENDIX B

Important Factors in Surface EMG Measurement By Dr. Scott Day Bortec Biomedical Ltd 225, 604-1st ST SW Calgary, AB T2P 1M7 Ph +1 403.237.8144 Email info@bortec.ca Website www.bortec.ca

113

Table of Contents EMG Introduction ............................................................................................................... 3 Sources of Noise ................................................................................................................. 4 Ambient Noise ................................................................................................................ 4 Transducer Noise ............................................................................................................ 4 The Importance of Skin â&#x20AC;&#x201C; Electrode Impedance................................................................. 5 Skin Preparation.................................................................................................................. 6 Cross Talk ........................................................................................................................... 6 sEMG Normalization .......................................................................................................... 7 Types of Electrodes............................................................................................................. 7 Dry Electrodes ................................................................................................................ 7 Gelled Electrodes ............................................................................................................ 8 SENIAM Recommendations for Bipolar sEMG Electrodes .............................................. 8 Electrode Shape .............................................................................................................. 8 Electrode Size ................................................................................................................. 9 Inter-Electrode Distance ................................................................................................. 9 Electrode Material........................................................................................................... 9 Electrode Construction.................................................................................................. 10 Electrode Placement.......................................................................................................... 10 Signal Conditioning and Amplification ............................................................................ 11 Properties of an ideal pre-amplifier .................................................................................. 12 Common Mode Rejection Ratio (CMRR) .................................................................... 13 Input Impedance............................................................................................................ 13 Distance from Signal Source ........................................................................................ 14 D/C Signal Suppression ................................................................................................ 14 References ......................................................................................................................... 16

114

EMG Introduction Small electrical currents are generated by muscle fibres prior to the production of muscle force. These currents are generated by the exchange of ions across muscle fibre membranes, a part of the signaling process for the muscle fibres to contract. The signal called the electromyogram (EMG) can be measured by applying conductive elements or electrodes to the skin surface, or invasively within the muscle. Surface EMG is the more common method of measurement, since it is non-invasive and can be conducted by personnel other than Medical Doctors, with minimal risk to the subject. Measurement of surface EMG is dependent on a number of factors and the amplitude of the surface EMG signal (sEMG) varies from the uV to the low mV range (Basmajian & DeLuca, 1985). The amplitude and time and frequency domain properties of the sEMG signal are dependent on factors such as (Gerdle et al., 1999): • the timing and intensity of muscle contraction • the distance of the electrode from the active muscle area • the properties of the overlying tissue (e.g. thickness of overlying skin and adipose tissue) • the electrode and amplifier properties • the quality of contact between the electrode and the skin In most cases, information on the time and intensity of muscle contraction is desired. The remainder of the factors only exacerbates the variability in the EMG records, making interpretation of results more difficult. Nevertheless, there are methods to reduce the impact that non- muscular factors have on the properties of the EMG signal. For example, much of this variability in the sEMG signal can be minimized through: • using the same electrodes and amplifier (i.e. same signal conditioning parameters) • ensuring consistency in the quality of contact between the electrodes and the skin Within subjects, the variability of the sEMG signal can also be reduced in consecutive recording sessions by placing the electrodes over the same skin location. In addition, there are other methods of normalizing the EMG signal to reduce the variability both within and between subjects. Many of the most important issues relating to the acquisition and analysis of the sEMG signal were recently addressed in a multi-national consensus initiative called SENIAM: Surface EMG for the Non-Invasive Assessment of Muscles (www.rrd.nl/projects/content/file_100.htm) (Freriks and Hermens, 2000). Measuring and accurately representing the sEMG signal depends on the properties of the electrodes and their interaction with the skin, amplifier design, and the conversion and subsequent storage of the EMG signal from analog to digital form (i.e. A/D conversion). The quality of the measured EMG is often described by the ratio between the measured EMG signal and unwanted noise contributions from the environment. The goal is to maximize the amplitude of the signal while minimizing the noise. Assuming that the amplifier design and process of A/D conversion exceed acceptable standards (see below and Gerdle et al., 1999), the signal to noise ratio is determined almost exclusively by the

115

electrodes, and more specifically, the properties of the electrode – electrolyte – skin contact. The remainder of this document outlines the factors influencing the characteristics of the EMG signal, with an emphasis on mechanisms to increase the consistency and accuracy of the sEMG signal. For more detailed information please refer to the sEMG methods chapter “Acquisition, Processing and Analysis of the Surface Electromyogram”, in the recent textbook (1999) entitled “Modern Techniques in Neuroscience Research” (Eds. U. Windhorst & H. Johansson).

Sources of Noise Before we can develop strategies to eliminate unwanted noise we must understand what the sources of noise are. The two types of noise are ambient noise and transducer noise. Ambient Noise Ambient noise is generated by electromagnetic devices such as computers, force plates, power lines etc. Essentially any device that is plugged into the wall A/C (Alternating Current) outlet emits ambient noise. This noise has a wide range of frequency components, however, the dominant frequency component is 50Hz or 60Hz, corresponding to the frequency of the A/C power supply (i.e. wall outlet). Transducer Noise Transducer noise is generated at the electrode – skin junction. Electrodes serve to convert the ionic currents generated in muscles into an electronic current that can be manipulated with electronic circuits and stored in either analog or digital form as a voltage potential. There are two types of noise sources that result from this transduction from an ionic to an electronic form: • D/C (Direct Current) Voltage Potential: caused by differences in the impedance between the skin and the electrode sensor, and from oxidative and reductive chemical reactions taking place in the contact region between the electrode and the conductive gel (Gerdle et al., 1999) • A/C (Alternating Current) Voltage Potential: generated by factors such as fluctuations in impedance between the conductive transducer and the skin. One effective method to decrease impedance effects is to use Ag-AgCl electrodes. This electrode consists of a silver metal surface plated with a thin layer of silver chloride material. (Duchene & Goubel, 1993) The goal with EMG measurements is to maximize the signal to noise ratio. Technological developments have decreased the level of noise in the EMG signal. The most important development was the introduction of the bipolar recording technique. Bipolar electrode arrangements are used with a differential amplifier, which functions to suppress signals common to both electrodes. Essentially, differential amplification subtracts the potential at one electrode from that at the other electrode and then amplifies the difference. 116

Correlated signals common to both sites, such as from power sources and electromagnetic devices, but also EMG signals from more distant muscles are suppressed. Moreover, the D/C components such as the over-potential generated at the electrode skin junction will be detected with similar amplitude (see below) and will therefore be suppressed. In contrast, signals from muscle tissue close to the electrodes will not be correlated and will be amplified (Gerdle et al., 1999). The advent of bipolar recordings with differential pre-amplification has enabled the recording of the full EMG bandwidth while increasing the spatial resolution (i.e. the size of the recording area). This also has the effect of increasing the signal to noise ratio. One remaining factor is how the quality of the electrode – skin contact impacts the process of differential amplification in bipolar EMG measurements. The electrode – skin contact is quantitatively defined by the resistance of the skin and underlying tissues, in addition to the capacitance of the electrodes. It is commonly called electrode – skin impedance. The electrode – skin impedance can be measured quantitatively, such as with the BISIM impedance measurement device offered by Bortec.

The Importance of Skin – Electrode Impedance Consistency in impedance is critical for the reliability of EMG measurements. Modern pre-amplifier design (i.e. high input impedance) has reduced the importance of measuring EMG with a low level of electrode – skin impedance. While the absolute level of muscle impedance is not a critical factor, the stability in impedance over time and the balance in impedance between electrode sites have a considerable effect on the signal to noise ratio of the measured EMG signal (Freriks and Hermens, 2000), both in terms of noise levels and spatial resolution. The balance in impedance between electrode sites is important to minimize noise components. The impedance at each site does not have to be perfectly balanced, however, they should be relatively similar. The level of impedance balance is rather arbitrary, depending on the properties of the differential pre-amplifier in use, among other factors. The impedance determines the energy levels of the electrical signal measured at each electrode site (i.e. the view of the muscle and environment that the electrode measures). As the impedance becomes increasingly different between electrode sites, so too does the signal strength entering the process of differential amplification. Differential amplification only cancels common signal components. For example, if the energy of power-line noise is different, some of the noise will remain in the signal following the process of differential amplification (Gerdle et al., 1999). Similarly, the D/C voltage potential will be different and part of it will not be cancelled. If the pre-amplifier does not have sufficient D/C noise suppression in the residual D/C component, once amplified, can lead to pre-amplifier instability, inaccuracy and saturation. The general rule is the more balanced the electrode – skin impedance between electrode sites, the lower the noise and as a result the higher the signal to noise ratio. Given the 117

complexity of the EMG signal, it is difficult to predict or measure to what extent imbalanced impedance alters the properties of the EMG signal. Suffice it to say that by achieving a similar balance in impedance is the best way to increase the reliability of EMG measurements. It is also important that the impedance remains consistent over the duration of the measurement session. For the reasons similar to those above, the signal to noise ratio will wander if the impedance drifts during measurement, as will the spatial resolution of the recorded EMG signal. Recent evidence demonstrates that the relative level of electrode – skin impedance has a significant effect on the energy of the measured EMG signal. That is, low impedance (<10 kOhm) resulted in a high level of energy for EMG frequency components under 100 Hz as compared to high electrode – skin impedance (>100 kOhm).

In contrast, for EMG signal frequency components between 100 and 150 Hz, low electrode – skin impedance resulted in a lower signal energy level than the measurements with high impedance (Hewson et al., 2002). Other findings support these results (Duff et al., 2002), indicating the spatial resolution (i.e. the electrical view into the muscle) is altered with changes in electrode – skin impedance. Therefore, it is very important that the impedance remains consistent throughout the measurement session. For more information, see Gerdle et al., 1999.

Skin Preparation The electrode – skin interface generates a D/C voltage potential, mainly caused by a large increase in impedance from the outermost layer of skin, including dead skin material and oil secretions. This D/C potential, common to all electrodes, can be minimized with proper skin preparation. In fact, the quality of contact is typically reduced by at least a factor of 10 with proper preparation (Merletti and Migliorini, 1998). The SENIAM initiative recommends that the skin be adequately prepared (www.rrd.nl/projects/ content/fil_100.htm ).

Cross Talk It is important to recognize that the bipolar sEMG is not always a selective representation of the electrical activity of a single muscle directly underlying the recording electrodes. With smaller muscles the electrodes may overlook the electrical activity of one or more neighboring muscles and their signals may crosstalk with the sEMG from the desired muscle. While signal sources close to the electrode will dominate the recorded sEMG signal, more distant sources from other muscles may experience crosstalk (Gerdle et al., 1999). The distance for effective electrode measurement is the radius about the electrode where the amplitude of signal contributions is larger than the standard deviation of the signal noise (Gerdle et al., 1999). The amplitude of the bipolar sEMG signal decays 118

exponentially with increased distance from the recording electrode (Day, 1997). This is due to the fact that muscle fibres, subcutaneous fat and skin are anisotropic and act as a spatial filter with low pass frequency properties, where an increase in the distance between the muscle fibre and electrode increases the filtering effect. Effectively this means that less and less signal will be measurable from progressively more distant electrical sources; and similarly, that the frequency of sEMG contributions become progressively lower (i.e. shift in frequency spectrum to lower frequencies) ( Lindstrom & Magnusson, 1977). Crosstalk can be avoided by choosing the appropriate size of the electrodes conductive area and the appropriate inter-electrode distance. Decreasing the size of the conductive area reduces the effective sEMG measurement distance (i.e. depth). Similarly, decreasing the inter-electrode distance decreases the effective recording distance and shifts the EMG bandwidth to higher frequencies (Lindstrom & Magnusson, 1977). The SENIAM initiative has information on recommended electrode placement procedures for different muscles and muscle areas (www.rrd.nl/projects/content/fil_100.htm ).

sEMG Normalization The voltage potential of the sEMG signal detected by the electrodes strongly depends on several factors, varying between individuals and also over time within an individual. Thus, the amplitude of the sEMG itself is not useful in group comparisons, or to follow events over a long period of time (Mathiassen, 1997). The fact that the recorded EMG amplitude is never absolute is mainly due to the fact that the impedance varies between the active muscle fibres and electrodes and is unknown (Gerdle et al., 1999). Therefore, when comparing amplitude variables between measurements, normalization of some kind is required, i.e. the sEMG converted to a scale that is common to all measurement occasions. Normalizing the signal amplitude with respect to force or torque is a commonly used technique. Typically, the EMG is related to a maximal contraction, or a submaximal contraction at a known level of force. For other examples and types of signal normalization please see the papers by Mathiassen et al. (1995, 1997) and Merletti et al. (1995).

Types of Electrodes Two types of surface electrodes are commonly in use: â&#x20AC;˘ Dry electrodes in direct contact with the skin â&#x20AC;˘ Gelled electrodes using an electrolytic gel as a chemical interface between the skin and the metallic part of the electrode Dry Electrodes

119

“Dry electrodes are mainly used in applications where geometry or size of electrodes does not allow gel” (Gerdle et al., 1999). Bar electrodes and array electrodes are examples of dry electrodes. With dry electrodes it is common to have the pre-amplifier circuitry at the electrode site, due in part to the high electrode – skin impedance associated with dry electrodes. As such, dry electrodes (typically >20g) are considerably heavier than gelled electrodes (<1g). This increased inertial mass increases the difficulty in maintaining electrode fixation, as compared to gelled electrodes. Gelled Electrodes Gelled electrodes use an electrolytic gel as a chemical interface between the skin and the metallic part of the electrode. Oxidative and reductive chemical reactions take place in the contact region of the metal surface and the gel. Silver - silver-chloride (Ag- AgCl) is the most common composite for the metallic part of gelled electrodes. The AgCl layer allows current from the muscle to pass more freely across the junction between the electrolyte and the electrode. This introduces less electrical noise into the measurement, as compared with equivalent metallic electrodes (e.g. Ag). Due to this fact, Ag-AgCl electrodes are used in over 80% of surface EMG applications (Duchene & Goubel, 1993). Gelled electrodes can either be disposable or reusable. Disposable electrodes are the most common since they are very light. Disposable electrodes come in a wide assortment of shapes and sizes, and the materials comprising the patch and the form of the conductive gel varies between manufacturers. With proper application, disposable electrodes minimize the risk of electrode displacement even during rapid movements.

SENIAM Recommendations for Bipolar sEMG Electrodes The SENIAM initiative (www.rrd.nl/projects/content/fil_100.htm , Freriks and Hermens, 2000) has recommendations for construction of bipolar sEMG electrodes (sensors) including: • Electrode shape • Electrode size • Inter-electrode distance • Electrode material • Electrode construction Electrode Shape • The “electrode shape is defined as the shape of the conductive area”. Electrodes come in several shapes from circular to square to bar shaped electrodes. The practical implication is that the surface area for each electrode site (i.e. bipolar) should be the same, so the input impedance at each site is similar and common

120

mode disturbance is limited. “SENIAM has found no clear and objective criteria for a recommendation for electrode shape”. Electrode Size • The “electrode size is defined as the size of the surface of a conductive area of a sEMG electrode”. SENIAM recommends that the “size of electrodes in the direction of the muscle fibres is maximum 10mm”. • “Upon an increase of the size perpendicular to the muscle fibres (bar electrodes perpendicular to the muscle fibres), it is expected that the view of the electrodes increases. No quantitative data on the extent of this effect on the sEMG characteristics is available at the moment.” • “Upon an increase of the size in the direction of the muscle fibres, it can be shown that this has an integrative effect on the sEMG signal, increasing the detected amplitude and decreasing the high frequency contents.” • “The European inventory showed that circular electrodes with a circular diameter of 10mm are most preferred.” • “For bipolar sensors, in general, the size of the electrodes should be large enough to record a reasonable pool of motor units, but small enough to avoid crosstalk from other muscles.” Inter-Electrode Distance • “Inter-electrode distance is defined as the center to center distance between the conductive areas of electrodes.” • “The influence of the inter-electrode distance on the pick up area and crosstalk is a relevant item in the literature.” • SENIAM recommends applying “the bipolar sEMG electrodes around the recommended sensor location with an inter-electrode distance of 20mm”. • “When bipolar electrodes are being applied on relatively small muscles the interelectrode distance should not exceed ¼ of the muscle fibre length. In this way unstable recordings, due to tendon and motor end-plate effects can be avoided.” Electrode Material • “SENIAM recommends to use pre-gelled Ag/AgCl electrodes.” • “The electrode material which forms the contact layer with the skin needs to realize a good electrode skin contact, a low electrode-skin impedance, and a ‘stable’ behavior in time (that is, with respect to impedance and chemical reactions at the skin interface).” • “An inventory has shown that different types of materials are used: Ag/AgCl (silver/silver-chloride), AgCl (silver-chloride), Ag, Au etc.” • “Electrodes are mostly combined with electrode gel.” • “Electrode gel and paste are used to reduce the electrode-skin impedance.”

121

• Disposable and reusable electrodes give similar performance; reusable electrodes have an increased risk of improper and inconsistent measurement characteristics through change impedance. Electrode Construction • Defined as “the (mechanical) construction which is used to integrate the electrodes, the cables and (if applicable) the pre-amplifier”. • “SENIAM recommends a construction with a fixed inter electrode distance, built from light weight material.” • “Cables need to be fixed using tape or an elastic band in such a manner that pulling artifacts can be avoided.” • “If in fast dynamic contractions the sensor construction causes too much (movement) artifact due to the inertia of the construction, SENIAM recommends to fix the inter electrode distance using (double sided) tape or rings.” • It is expected that the construction (and its mass) do not directly affect the sEMG characteristics. These are nevertheless some potential indirect effects which can disturb or interfere with the recorded sEMG pattern. • “If the construction of the sensor is such that inter-electrode distance can vary during muscle contraction, this will modulate the amplitude, shape and width of the action potentials, and will consequently affect the interference pattern both with respect to its amplitude and frequency characteristics.” • “If the construction of the sensor is such that electrodes and cables can move, there is the potential risk for movement artifacts (due to pulling of cables or inertia of the construction).”

Electrode Placement The EMG signal provides a view of the electrical activity in a muscle(s) during contraction. The electrical view is highly dependent on where the electrode is overlying the muscle of interest. Since electrode placement determines the electrical view of a muscle, then it is important in EMG measurements to be consistent in the placement of the electrodes for a subject over consecutive recording sessions and between different subjects. When determining electrode placement, the use of the guidelines set forth in the international SENIAM initiative is highly recommended. One of the deliverables of this recent consensus initiative was the electrode placement procedures for 27 different muscle areas. Please see the SENIAM web-site (www.rrd.nl/projects/content/fil_100.htm ) for further details or contact Bortec Biomedical Ltd. The sensor location is defined as the position of the two bipolar sites overlying a muscle in relation to a line between two anatomical landmarks. The goal of sensor placement is to achieve a location where a good and stable surface EMG signal can be obtained. There are two general strategies for placement of electrodes. From the skin surface the electrode can be arranged longitudinally with respect to the long axis of the muscle and transversely, perpendicular to the long axis. 122

â&#x20AC;˘ Longitudinal: the recommendation is to place the bipolar electrode arrangement halfway from the distal motor end-plate (approximation â&#x20AC;&#x201C; muscle mid-line) zone and distal tendon. The goal is to avoid the sensor overlying the innervation zone or tendon during the whole range of motion. â&#x20AC;˘ Transverse: the recommendation is to place the bipolar electrode on the muscle so that each sensor is away from the boundary of the muscle recording area of interest. This can consist of the compartments of a large muscle and neighboring muscles underlying the area of the electrode. Typically, this means that the line between the centers of the electrode sensors is roughly parallel to the long axis of the muscle.

Signal Conditioning and Amplification Advent of modern electronics and the process of differential amplification have enabled the measurement of EMG signals of low noise and high signal fidelity (i.e. high signal to noise ratio). With differential amplification, it is now possible to measure the full effective bandwidth of the EMG signal. Typical bandpass frequency ranges are from between 10 and 20Hz (high pass filtering) to between 500 and 1000Hz (low-pass filtering). High-pass filtering is necessary because movement artifacts are comprised of low frequency components (typically <10Hz). Low pass filtering is desirable to remove high-frequency components to avoid signal aliasing (see Gerdle et al., 1999). In the past, it was common to remove power-line (A/C) noise components (i.e. either 50 or 60Hz) by using a sharp notch filter. See Figure 1 below. There are problems with notch filtering because EMG has large signal contributions at these and neighboring frequencies. The result of notch filtering is the loss of important EMG signal information, so notch filtering should be avoided as a general rule. Amplification is also necessary to optimize the resolution of the recording or digitizing equipment (for more information see Gerdle et al., 1999). Amplifiers of high quality have adjustable gains of between, at least, 100 and 10 000 to maximize the signal to noise ratio of the EMG signal during each recording. This range of gains provides the sufficient range of amplifications for surface EMG signals which can range typically from 0 to 6mV peak to peak (Basmajian & DeLuca, 1985). The AMT-8 EMG System from Bortec Biomedical Ltd. has a gain range of between 100 and 15 000. The quality of the EMG signal, in part, depends on the characteristics of the amplification process. While there may be several stages of amplification, the most important stage is often described as pre-amplification. Pre-amplification implies the first stage of amplification, close to the signal source. There are several important parameters in preamplifier signal conditioning of the EMG signal.

123

Properties of an ideal pre-amplifier There are several important properties to consider in a pre-amplifier: • High common mode rejection ratio • Very high input impedance • Short distance to the signal source • Strong DC signal suppression Common Mode Rejection Ratio (CMRR) Bipolar electrode arrangements are used with a differential amplifier, which functions to suppress signals common to both electrodes. Essentially, differential amplification subtracts the potential at one electrode from that at the other electrode and then amplifies the difference. Correlated signals common to both sites, such as from power sources and electromagnetic devices, but also EMG signals from more distant muscles are suppressed. The common mode rejection ratio provides an index on the extent to which common signal components are attenuated from the signal. As such it is desirable to have the highest common mode rejection (CMR) possible. At the moment the best CMRR that can be realistically achieved with current technology is about 120 (dB/Octave @ e.g. 60Hz). The CMR is expressed in logarithmic form, thus an increase in CMRR from 90 (dB) to 110 (dB) is an increase of 100 x. See Table 1 for a conversion of the CMR in (dB) to a linear scale. The Bortec Biomedical pre-amplifiers achieve CMRR near 115 (dB @ 60Hz). Formula: A(dB) = 20 log(10) (A0/A1) dB Linear Scale 3 1.41 5 10 1.78 20 50 3.16 70 10 80 316 90 100 3160 110 120 10 000 31 600 100 000 316 000 1 000 000 124

Table 1: Comparison of logarithmic and linear scaling, which indicates the level of suppression of signal components common to the bipolar leads input into the differential amplifier.

Input Impedance It is very important that EMG pre-amplifiers have high input impedance. Input (i.e. source) impedance is typically less than 50 kOhms with gel electrodes and proper skin preparation. In order to measure a voltage accurately, the input resistance of the measurement device should be considerably larger than the impedance at the skin1. If not, the signal will be attenuated and distorted due to the effects of process called input loading. For gel electrode recordings an input impedance in the tens of MOhms range is sufficient. For example, the pre-amplifiers from Bortec Biomedical Ltd. use active components with input impedance reaching 10GOhm. In contrast, for dry electrode recordings with a much higher skin-electrode impedance, into the thousands of KOhms (i.e. few MOhms), input impedances in the GOhm range (i.e. 1010 Ohms) are required to achieve an adequate signal to noise ratio. Distance from Signal Source One drawback with high input impedance is that power line noise, RF (radio frequency) noise and movement artifacts are introduced in the lead wires by means of capacitive coupling: the higher the input impedance of the pre-amplifier, the greater the impact of noise and movement artifact from lead wires. That is, with increased length of the leads the parasitic capacitance increases, thus the resulting coupled noise is increased. In other words, the long lead length combined with high impedance of the pre-amplifier results in reduced signal to noise ratio. For gel electrode recordings with low electrode â&#x20AC;&#x201C; skin impedance levels, these effects are minimal, provided a short lead length from the electrode to the pre-amplifier is used. Bortec Biomedical pre-amplifiers are specifically designed for a short lead length, to minimize the mass of the electrode assembly to the patch and the pre-amplifier clip attachment. For dry electrode recordings, the pre-amplifier is typically housed with the electrode sensors minimizing the lead distance. This increases the mass of the electrode assembly attached to the subject, increasing the risk of inadequate fixation. Inadequate fixation can cause common mode disturbance in amplifiers and movement artifacts, which will effectively reduce the signal to noise ratio of the recorded EMG. D/C Signal Suppression It is important that pre-amplifier circuits have strong D/C component suppression circuitry. There are D/C components caused by factors involving skin impedance and the chemical reactions between the skin and the electrode and gel. Any difference in the D/C potential measured at the each of the electrode sensors will be amplified, which can lead to pre-amplifier instability or saturation. Such instabilities are often referred to as common mode disturbances. 125

The practical test of the D/C suppression capacity of a pre-amplifier is to bombard the pre-amplifier with D/C potential fields, by using measuring EMG in conjunction with D/C based electromagnetic motion capture equipment or when stimulating nerve or muscle with a magnetic stimulator. The AMT-8 EMG System from Bortec Biomedical Ltd. is often used in the presence of equipment emitting high levels of D/C electromagnetic radiation. The decrease in the signal to noise ratio is surprisingly small indicating that the D/C suppression circuitry is effective 1

The typical rule of thumb in electronics is for the input impedance of the pre-amplifiers to be at least 10 times the electrode â&#x20AC;&#x201C; skin impedance: at least 100 times is required for high precision measurements.

126

References Basmajian JV, De Luca CJ (1985) Muscles Alive. Their Function Revealed by Electromyography. Williams & Wilkens, Baltimore. Day SJ (1997) The Properties of Electromyogram and Force in Experimental and Computer Simulations of Isometric Muscle Contractions: Data from an Acute Cat Preparation. Dissertation, University of Calgary, Calgary. Duchêne J and Gouble F (1993) Surface electromyogram during voluntary contraction: Processing tools and relation to physiological events. Critical Reviews in Biomedical Engineering 21(4):313–397 Duff R, Nolan P, Rybansky M, O’Malley M. (2002) Evolution in impedance at the electrode-skin interface of two types of surface EMG electrodes during long-term recordings. Proceedings: XIVth Congress of the International Society of Electrophysiology and Kinesiology 175-176. Eds: Kollmitzer J and Bijak M. University of Vienna, Austria. Freriks B and Hermens H (2000) European Recommendations for Surface ElectroMyoGraphy, Results of the SENIAM Project (CD-rom). Roessingh Research and Development, The Netherlands Gerdle B, Karlsson S, Day S, Djupsjöbacka M (1999) Acquisition, Processing and Analysis of the Surface Electromyogram. Modern Techniques in Neuroscience. Chapter 26: 705-755. Eds: Windhorst U and Johansson H. Springer Verlag, Berlin. Hewson DJ, Hogrel J-Y, Langeron Y, Duchêne J (2002) Evolution in impedance at the electrode-skin interface of two types of surface EMG electrodes during long-term recordings. Proceedings: XIVth Congress of the International Society of Electrophysiology and Kinesiology 173-174. Eds: Kollmitzer J and Bijak M. University of Vienna, Austria. Lindström LH, Magnusson RI (1977) Interpretation of myoelectric power spectra: a model and its applications. Proceedings of the IEEE 65: 653–662 Mathiassen SE, Winkel J, Hägg GM (1995) Normalization of surface EMG amplitude from the upper trapezius muscle in ergonomic studies – a review. J Electromyogr Kinesiol 5: 197–226 Mathiassen SE (1997) A checklist for normalisation of surface EMG amplitude. Proceedings of the Second General SENIAM Workshop Chapter 2. Eds: Hermens H, Hagg G, Freriks B. Stockholm, Sweden. Merletti R, Gulisashvili A, Lo Conte LR (1995) Estimation of shape characteristics of surface muscle signal spectra from time domain data. IEEE Trans Biomed Eng 42: 769–776 Merletti R and Migliorini M (1998) Surface EMG electrode noise and contact impedance. Proceedings of the third general SENIAM workshop.

127

APPENDIX C CASP Critical Appraisals forms for: 1.

Bley, A. S., Correa, J. C. F., Dos Reis, A. C., Rabelo, N. D. D. A., Marchetti, P. H., & Lucareli, P. R. G. (2014). Propulsion Phase of the Single Leg Triple Hop Test in Women with Patellofemoral Pain Syndrome: A Biomechanical Study. PloS one, 9(5),

Bolgla, L. A., Malone, T. R., Umberger, B. R., & Uhl, T. L. (2011). Comparison of hip and knee strength and neuromuscular activity in subjects with and without patellofemoral pain syndrome. International journal of sports physical therapy, 6(4), 285.

Nakagawa, T. H., Muniz, T. B., Baldon, R. M., Maciel, C. D., Amorim, C. F., & Serrão, F. V. (2011). Electromyographic preactivation pattern of the gluteus medius during weight-bearing functional tasks in women with and without anterior knee pain. Brazilian Journal of Physical Therapy, 15(1), 5965.

Nakagawa, T. H., Moriya, É. T., Maciel, C. D., & Serrão, F. V. (2012). Trunk, pelvis, hip, and knee kinematics, hip strength, and gluteal muscle activation during a single-leg squat in males and females with and without patellofemoral pain syndrome. journal of orthopaedic & sports physical therapy, 42(6), 491-501.

O'Sullivan, K., Herbert, E., Sainsbury, D., McCreesh, K., & Clifford, A. (2012). No difference in gluteus medius activation in women with mild patellofemoral pain. Journal of sport rehabilitation, 21(2), 110.

Saad, M. C., Felício, L. R., Masullo, C. D. L., Liporaci, R. F., & Bevilaqua-Grossi, D. (2011). Analysis of the center of pressure displacement, ground reaction force and muscular activity during step exercises. Journal of Electromyography and Kinesiology, 21(5), 712-718.

Shirazi, Z., Biabani Moghaddam, M., & Motealleh, A. (2014). Comparative Evaluation of Core Muscle Recruitment Pattern in Response to Sudden External Perturbations in Patients With Patellofemoral Pain Syndrome and Healthy Subjects. Archives of Physical Medicine and Rehabilitation.

Song, C. Y., Huang, H. Y., Chen, S. C., Lin, J. J., & Chang, A. H. (2014). Effects of femoral rotational taping on pain, lower extremity kinematics, and muscle activation in female patients with patellofemoral pain. Journal of Science and Medicine in Sport.

Souza, R. B., & Powers, C. M. (2009). Differences in hip kinematics, muscle strength, and muscle activation between subjects with and without patellofemoral pain. journal of orthopaedic & sports physical therapy, 39(1), 12-19.

10. Willson, J. D., Kernozek, T. W., Arndt, R. L., Reznichek, D. A., & Scott Straker, J. (2011). Gluteal muscle activation during running in females with and without patellofemoral pain syndrome. Clinical Biomechanics, 26(7), 735-740.

128