Near infrared spectroscopy for assessing oxygenation and hemodynamics in the upper extremities of healthy subjects and patients

with work-related muscle pain

Guilherme Hilgert Elçadi

Centre for Musculoskeletal Research, Department of

Occupational and Public Health Sciences, University of Gävle

Department of Community Medicine and Rehabilitation,

Rehabilitation Medicine, Umeå University

Umeå 2012

Responsible publisher under Swedish law: the Dean of the Medical Faculty

This work is protected by the Swedish Copyright Legislation (Act 1960:729)

New series No. 1527

ISBN: 978-91-7459-493-5

ISSN: 0346-6612

Electronic version available at http://umu.diva-portal.org/

Printed by: Print & Media

Umeå, Sweden 2012

To all of those who helped to create and to make this dream to become a

reality.

i

Table of Contents

Table of Contents i Abstract iii List of abbreviations v Original papers vi Introduction 1

General background 1 The Extensor Carpi Radialis and Trapezius muscles 2 Blood flow and its regulation 4 The Hemoglobin molecule 4 Mechanisms of WRMP 5 Near infrared spectroscopy (NIRS) 5

The modified Lambert-Beer Law 7 Validity of NIRS for muscle tissue 8 Reliability of NIRS 8 Factors affecting the NIRS signal 9

Electromyography (EMG) 9 NIRS and EMG 10 NIRS and WRMP 10

Aims of the Thesis 11 General aims 11 Specific aims 11

Materials and methods 12 Subjects 12

Study 1 12 Study 2 13

Methods (Studies 1 and 2) 14 Work model (Studies 1 and 2) 18 Statistics (papers I, II, III and IV) 20

Paper I 20 Paper II 20 Paper III 20 Paper IV 21

Results 22 Paper I 22 Paper II 27 Paper III 29 Paper IV 33

Discussion 36 Methodological considerations 36 Oxygenation and myoelectric activity for males and females (Paper I) 38

ii

Isometric contractions 38 Sustained contraction 38

Interday reliability of NIRS variables (Paper II) 39 Short-term submaximal contractions for WRMP patients and healthy controls

(Paper III) 42 Sustained contraction for WRMP patients and healthy controls (Paper IV) 43

Conclusions 44 Acknowledgements 45 References 47

iii

Abstract The prevalence of work-related muscle pain (WRMP) is large in the general population in the industrialized world. Despite significant advances over recent years in some research areas, the mechanisms of why WRMP occurs and the pathophysiological mechanisms behind the disorders are still unclear. One suggested explanation is that WRMP is caused initially by a limitation of the local muscle circulation and oxidative metabolism. There is a lack of objective methods to gauge the development and diagnosis of WRMP.

Near infrared spectroscopy (NIRS) is a non-invasive technique that allows for determinations of oxygenation and blood flow. The purpose of this thesis was to evaluate NIRS (1) as a method for measuring muscle oxygenation and hemodynamics for the extensor carpi radialis (ECR) and trapezius descendens muscles (TD), and (2) to investigate whether variables measured by NIRS differed between patients diagnosed with WRMP and healthy subjects.

Several variables of NIRS were produced and investigated. These included muscle oxygenation (StO2%), changes during contractions (∆StO2%) and StO2% recovery (Rslope), total hemoglobin (HbT) as an indication of blood volume and its changes during contractions (∆HbT). In addition, for the ECR, by applying an upper arm venous occlusion (VO) HbTslope increase as a surrogate of blood flow, and for both VO and arterial occlusion (AO) HHbslope increase (i.e. deoxyhemoglobin slope) as a surrogate of oxygen consumption were variables of interest.

A first objective was to determine how StO2% and HbT responded to various contraction forces and how it related to muscle activation measured by electromyography (EMG). For both muscles isometric contractions of 10, 30, 50 and 70% of maximal voluntary contraction (MVC) were maintained for 20 s each by healthy males and females; additionally a 10% MVC contraction was sustained for 5 min. For the different contraction levels, predictable relationships were seen between ∆StO2% and force, and between ∆StO2% and EMG RMS amplitude. The general trend was a decrease in ∆StO2% with increasing force and increasing EMG. Females showed a tendency for a higher oxygen use (i.e., drop in StO2%) for the ECR over force levels than males and a higher RMS% MVC for the TD. For the 10% MVC contraction sustained for 5 min gender specific changes over time for HbT and RMS for the ECR, and for StO2% for the TD muscle were seen.

A second objective was to determine the day-to-day reliability of NIRS variables for the ECR and TD muscles at group level (Pooled data) and at gender level (males and females). Measurements were performed on two occasions separated by 4-6 days and intraclass correlation coefficients (ICC) and limits of agreement (LOA) were determined as reliability and reproducibility indicators, respectively. Variables tested were ∆StO2% during

iv

submaximal isometric contractions of 10, 30, 50 and 70% MVC and StO2% recovery (Rslope) after contractions and after AO. For the ECR, HbTslope as an indication of blood flow (using VO) and HHbslope as a surrogate of oxygen consumption for both VO and AO were computed. For ∆StO2% for the ECR the highest ICC was at 30% MVC for both the pooled data and at gender level. For the TD ICCs were comparably high for 30, 50, 70 % MVC (for both muscles the ∆StO2% at 10% MVC showed the lowest ICC). Further, females showed a higher ICC than males for contraction levels of 50 and 70% MVC. For both muscles, LOA for ∆StO2% was lowest at 10% and highest at 50 and 70% MVC. For the ECR Rslope ICCs were high for all contraction levels, but was lower for AO; LOA was lowest at 70% MVC. For the TD, Rslope ICCs were also high for all contraction levels and LOA was lowest at 30 % MVC. ICC for HbTslope was the lowest of all variables tested. For HHbslope ICC was higher for AO than for VO, and LOA was lower for AO.

A third objective was to determine if there were differences between healthy subjects and patients diagnosed with WRMP in ∆StO2% and ∆HbT responses during varying submaximal contractions (10, 30, 50 and 70% MVC), and StO2% recovery (Rslope) immediately after contractions and AO. Additional variables tested in the ECR at rest were HHbslope to indicate oxygen consumption (using AO) and HbTslope as an indication of blood flow. There were no differences between groups in ∆StO2% and ∆HbT variables during the contractions or Rslope in the recovery after contractions or AO. Furthermore, HbTslope was not different between groups However, oxygen consumption for the ECR and StO2% for the TD at rest were significantly greater for healthy subjects compared to patients.

A fourth objective was to determine if there were differences in StO2% and HbT between healthy subjects and WRMP patients during a 12 min sustained contraction of 15 % MVC. In addition, the protocol included a recovery period of 30 min. Prior to contraction, as well as during the recovery period, HbTslope as a surrogate of blood flow was determined for the ECR. Neither the ECR nor the TD exhibited significant differences between groups for StO2% and HbT during the contraction. For the TD patients showed a lower StO2% value at rest and throughout the contraction than healthy subjects. For the ECR HbT during the sustained contraction the general trend was an initial decrease with gradual increase throughout the contraction for both groups. For HbTslope no differences were seen between patients and healthy subjects before the sustained contraction and during the recovery period for both muscles.

NIRS is deemed a suitable technique for assessing physiological measurements of the upper extremity, including for day-to-day testing.

NIRS was not able to distinguish between the patients with WRMP and controls. A concern in the thesis is the characteristics of the patient group in being equally active in recreational sports, actively working, and similar in muscle strength as controls. Thus, applying NIRS for studying a more severe patient group could yield different results.

v

List of abbreviations

AO Arterial occlusion

ATT Adipose tissue thickness

BMI Body mass index

ECR Extensor carpi radialis

EMG Electromyography

ICC Intraclass correlation coefficient

LDF Laser Doppler flowmetry

LOA Limits of agreement

MPF Mean power frequency

MSDs Musculoskeletal disorders

MTh Muscle thickness

NIRS Near infrared spectroscopy

HHb Deoxyhemoglobin

HbO2 Oxyhemoglobin

HbT Total hemoglobin

RMS Root mean square

S2bs Between subjects variance component

S2ws Within subjects variance component

StO2% Oxygen percent saturation

TD Trapezius descendens

VO Venous occlusion

WRMD Work-related musculoskeletal disorders

WRMP Work-related muscle pain

vi

Original papers This Thesis is founded on the original papers listed below. In the following text these papers will be referred to by their Roman numerals.

I. Elcadi GH, Forsman M, Crenshaw AG (2011) The relationship between oxygenation and myoeletric activity in the forearm and shoulder muscles of males and females. Eur J Appl Physiol 111(4): 647-658

II. Crenshaw A, Elcadi GH, Hellstrom F, Mathiassen SE (2012) Reliability of near-infrared spectroscopy for measuring forearm and shoulder oxygenation in healthy males and females. Eur J Appl Physiol 112(7): 2703 – 2715

III. Elcadi GH, Forsman M, Aasa U, Fahlstrom M, Crenshaw AG (2012) Shoulder and forearm oxygenation and myoeletric activity in patients with work-related muscle pain and healthy subjects. Eur J Appl Physiol – (DOI 10.1007/s00421-012-2530-6)

IV. Elcadi GH, Forsman M, Aasa U, Fahlstrom M, Hallman D, Crenshaw AG. No differences in oxygenation in the forearm and shoulder of patients with work-related muscle pain and healthy subjects during a low-load sustained contraction. (Manuscript)

1

Introduction

General background

Work-related musculoskeletal disorders (WRMD) affect a large number of people in the world today. WRMD have been reported to account for two thirds of all occupational disorders in industrialized countries, for instance France [3]. Although accurate data and statistics on the incidence and prevalence of WRMD are not easy to obtain for comparisons between countries, these disorders are known to have a high incidence in countries such as Canada, Japan, Sweden, Finland and England [10, 88]. It is not, however, a new condition. In the 18th century in one of the earliest accounts for the disorder Italian doctor Bernardino Ramazzini reported health disorders related to office workers and their repetitive work and mental stress [98, 111]. Since then several reports have been produced describing WRMD in many different occupational activities and under a variety of different names (e.g. Work-related musculoskeletal disorders, Work-Related Myalgia, Neck-Shoulder Myalgia, Cervicobrachial syndrome, Repetitive Strain Injury, etc.). In every case the symptoms are about the same and include pain in the muscle that starts in a narrower location that soon spreads to a larger area, and the symptoms often are increased by work. For these conditions an adjunct to the pain may be local swelling, muscle tenderness and functional loss. In spite of the acknowledgement of the magnitude of the problem and the efforts of many researchers, the pathophysiological mechanisms behind WRMD disorders are still unclear. Furthermore it is known that females are more prone to develop WRMD than males in occupations such as light repetitive assembly work or office work [81, 82]. As mentioned one of the inherent symptoms of WRMD disorders is muscle pain, thus throughout this thesis these disorders will be referred to as work-related muscle pain (WRMP).

It is acknowledged that the aetiology of WRMP is multifactorial and that understanding these mechanisms is pivotal for diagnosis, development of prevention and rehabilitation [11]. Known risk factors such as prolonged low level work and psychosocial stressors may interact affecting a complex manifold of mechanisms at different levels of the neuromuscular system which interact to cause or maintain WRMP. These interactions may become positive feedback loops and act as a vicious cycle that may reinforce the cause-effect relationships within these loops; however, the causal links between these risk factors are not fully understood.

It has been suggested that important feedback loops which play a crucial role in WRMP involve chemo- and nociceptors [42, 52]. Furthermore, it has been suggested that the accumulation of metabolites and inflammatory substances in the muscles during low level repetitive or sustained work may cause pain and motor control disturbances and that increased activity of the sympathetic central nervous system may worsen the problem [85] [86] [108]. These factors may have the effect of causing vasoconstriction, changes

2

in muscle fiber contractility and motor control [43], and disturbances in the oxidative metabolism system [47]. In a review by Visser and Van Dieen [108] that included a number of studies aimed at investigating the pathophysiological mechanisms, disturbances in blood flow and/or oxygen metabolism were concluded to be associated to WRMP. As examples, by the use of Laser-Doppler flowmetry probes (LDF), placed percutaneously within the muscle, Larsson and co-workers [62] and Larsson and co-workers [61] reported impaired regulation of the microcirculation for chronic trapezius myalgia patients and lowered blood flow for patients suffering cervico-brachial pain syndrome. In one of these studies the reduced blood flow was shown to be correlated with mitochondrial changes [62].

To date objective methods to assess physiological disturbances leading to or maintaining WRMP do not exist. An attractive method that is non-invasive and allows for determinations of oxygenation and blood flow is near infrared spectroscopy (NIRS). However, NIRS has not been thoroughly evaluated in regard to WRMP. The purpose of this thesis was to evaluate NIRS for measuring muscle oxygenation and hemodynamics (1) for the extensor carpi radialis and trapezius descendens muscles – two muscles prone for WRMP, for healthy males and females, and (2) to investigate whether patients diagnosed with WRMP and healthy subjects differed in variables detected by NIRS. Specifically, a number of NIRS variables were produced that included muscle oxygenation and blood volume changes during and after contractions, and blood flow and oxygen consumption in the resting muscles.

The Extensor Carpi Radialis and Trapezius muscles

The extensor muscles of the forearm, i.e. the extensor carpi radialis brevis (ECRB), along with the extensor carpi radialis longus (ECRL) and the shoulder trapezius pars descendens (TD) will be investigated because they are prone to WRMP [55, 68, 90, 107].

The ECRB belongs to the radial wrist extensors group of muscles. Its action is along with the ECRL of the extension of the wrist and the radial deviation (abduction) of the wrist. Their main actions are to act as an extensor and abductor of the wrist and mid carpal joints. Its origin is the lateral epicondyle of the humerus by a tendon that is shared with other forearm extensor muscles, the radial collateral ligament and adjacent intramuscular septa. Its belly ends at about mid forearm in a flat tendon. The tendon passes under the extensor retinaculum where it lies in a shallow groove on the back of the radius and is attached to the dorsal surface of the base of the third metacarpal on its radial side, distal to its styloid process and on adjoining parts of the second metacarpal base. The ECRB tendon shares the same synovial sheath with the ECRL. The ECRL and ECRB have been suggested to be composed by types I and IIA and IIB fibers but by a larger number of type II muscle fibres [66, 91]. To our knowledge differences between males and females have not yet been reported. The ECRB is irrigated by mainly two pedicles. One single branch from the brachial artery (1/3 of the way down the

3

forearm) and there is an additional proximally blood supply from branches from the radial collateral branch of the profunda brachii. The ECRB is innervated by the posterior interosseous nerve, C7 and C8. The muscle belly and the tendon of the ECRB can be palpated when the wrist is extended and abducted against resistance with the forearm pronated.

The ECRL originates mainly from the distal third of the lateral supracondylar ridge of the humerus and the front of lateral intermuscular septum and some fibres come from the common origin of the forearm extensors Its belly ends at the junction of the proximal and middle thirds of the fore arm in a flat tendon that passes under the extensor retinaculum and lies in a groove on the back of the radius behind the styloid process. It inserts on the radial side of the dorsal surface of the base of the second metacarpal bone, sending slips to the first or third metacarpals The ECRL is irrigated mainly by a single branch of the radial recurrent artery and additional blood supply comes from branches of the radial collateral branch of the profunda brachii and directly from the radial artery in the distal part of the muscle.

The trapezius muscle is a flat triangular muscle that extends over the back of the neck and upper thorax. The actions of the trapezius are to coordinate with other muscles to stabilize the scapula and control the scapula during arm movements. Together with the levator scapulae the trapezius upper fibres elevate the scapula and with the serratus anterior it rotates the scapula so that the arms can be raised above the head. Acting with the rhomboids the transverse part of the trapezius retracts the scapula and with the shoulders fixed it can bend the head backwards and laterally. The trapezius, levator scapulae, rhomboids and serratus anterior combine to produce a series of scapular movements. The lateral part of the muscle occurs at the shoulder tips and the superior part at the occipital protuberance and superior nuchal lines and the inferior part at the level of the 12th thoracic vertebrae at the spine. On both sides the muscle attaches to the medial third of the superior nuchal line, external processes and their supraspinous ligaments from C7 to T12. Superior fibres descend. Inferior fibres ascend, and the middle fibres between them advance horizontally. The lower third of the descending portion, the transverse, and the ascending portions of the muscle have a predominance of type I fibers, whereas the most superior parts of pars descendens have a higher proportion of type II fibers [65]. Lindman et al [64] have investigated the trapezius of healthy males and females. Females tended to show fewer type I fibers and more type IIB fibers in the descending portion of the muscle, and the fibers of the lower regions of the descending portion were somewhat larger, also the mean cross-sectional area of the fibers in the female muscle was considerably smaller. The upper part of the trapezius is supplied by a transverse muscular branch which comes from the occipital artery at the level of the mastoid process. The middle portion of the muscle is irrigated by the superficial artery or by a superficial branch of the transverse cervical artery. The lower third of the trapezius is irrigated by a muscular branch of the dorsal scapular artery. The trapezius is innervated by the spinal part of the accessory nerve. Sensory (proprioceptive) branches are derived from the ventral rami of C3 and C4.

4

Blood flow and its regulation

By varying sympathetic and parasympathetic activity, the central nervous system (CNS) has a direct influence on the heart and vascular tone and consequently muscle blood flow. With exercise blood flow is rapidly increased by increased heart rate and decreased vascular tone in order to meet the metabolic demands of the contracting muscle. The increase in blood flow in the muscle tissue has been shown to be linearly related to the increase in exercise intensity and blood flow can reach a level 50 times higher than at rest [2, 93]. This increase in blood flow is called active hyperaemia and the distribution of blood flow depends on the level of arteriolar smooth muscle tension.

At the capillary level, when a muscle is active, arterioles, metarterioles (arterial capillary) and precapillary sphincters are dilated by the action of vasodilator metabolites produced in the active muscle. Also neural and hormonal factors may affect vasodilation but to a lesser extent during exercise [12, 93]. This dilation allows for blood to flow through the capillaries, in contrary to what happens at rest when the capillaries are collapsed.

Metabolites produced in the active muscle include nitric oxide (NO), carbon dioxide (CO2), acetylcholine (Ach), potassium ions (K+) and adenosine [16, 22, 89]; the latter being a primary candidate as the most important vasodilator in skeletal muscle [93]. These metabolites, however, originate from muscle fibres, endothelium cells and erythrocytes (red blood cells) and along with muscle tissue activity and muscle mechanical factors (e.g. intramuscular pressure) compile the mechanisms for skeletal muscle blood flow regulation [89, 93].

The Hemoglobin molecule

The hemoglobin molecule is a protein to which oxygen, and to a lesser extent, carbon dioxide, reversibly combines. It consists of four subunits bound together. Each subunit has a molecular group called heme and a polypeptide attached to the heme. The four polypeptides in a hemoglobin molecule are together named globin. Each of the four heme groups contains an iron (Fe) atom to which oxygen binds. Since each iron atom can bind one molecule of oxygen a single hemoglobin molecule can bind four oxygen molecules. For simplicity the reaction equation can be written in terms of the reaction of single polypeptide-heme chain as:

O2 + Hb < = > HbO2

Thus, this chain can exist in two forms, deoxyhemoglobin (Hb) and oxyhemoglobin (HbO2) which will be extensively discussed in the course of this Thesis. Also in a blood sample containing many hemoglobin molecules, the portion of hemoglobin containing oxygen is expressed as a percent

5

saturation and will also be thoroughly discussed in the course of this Thesis. The human systemic arterial blood at sea level binds approximately 98 % of the available oxygen to hemoglobin molecules. The human body has the capacity of binding 1.34 ml of oxygen per gram of hemoglobin. The average concentration of hemoglobin is 140 g/L of blood in women and 160 g/L in men. Thus since the total amount of blood in the human body is about 5 L we are able to carry approximately 200 ml of oxygen to the tissues/min.

The hemoglobin molecule contains also chromophores. Chromophores are the parts of a molecule that gives it its color, in the case of hemoglobin, the heme part of the molecule. The kind of chromophore in the hemoglobin molecule is called a metal chromophore. Chromophores absorb certain wavelengths of light. This characteristic is the basis for the methodology used in this Thesis and will be further discussed in detail.

Mechanisms of WRMP

Several hypothetical mechanisms have been put forth to explain the occurrence (pathogenesis) and/or the maintenance (pathophysiology) of WRMP. The so-called ‘Cinderella hypothesis’ was first proposed by Hägg [39] and it claims that selected low-threshold (type 1) motor units are continually turned on during low level work that over long-term leads to degenerative changes. Ultimately pain is believed to result from sensitization of muscle nociceptors due to metabolic overloading of these fibers. Vollestad and Roe [110] bring up the idea about shear forces resulting from friction between muscle fibers during work. With prolonged activity the shear forces increase and nociceptors located between the muscle fibers are affected. Knardahl [52] proposes an interaction of blood vessels and nociceptor to be the site of pain origin, i.e. not the muscle cell. An integrated model that accounts for both the occurrence (pathogenesis) and the maintenance of WRMP was proposed by Johansson et al. [43]. The model has been coined the ‘Brussels model’ because it was originated as a consensus of a group of prominent researchers in a symposium in Brussels, Belgium in 2000. The model states that continuous low-level work leads to an accumulation of metabolites and/or inflammatory substances in the muscles that in turn activate chemosensitive nerve fibers. This ultimately causes alterations in muscle spindle activity. The reactions of the muscle spindles may also be affected by sympathetic nervous system activity. This effect may be direct via sympathetic nerve innervation of the spindle, or indirect due to sympathetically enhanced vasoconstriction, which may hinder the removal and thereby continued buildup of metabolites and inflammatory substances in the muscles; thus, a vicious cycle of chronic pain is demonstrated.

Near infrared spectroscopy (NIRS)

NIRS is a non-invasive method for the continuous monitoring of changes in the local muscle oxygenation, which represents the dynamic balance between oxygen delivery and consumption (c.f. Ferrari et al. [29]). A probe placed on

6

the skin enables monitoring of NIRS signals from small vessels, i.e. arterioles, capillaries and venules, deep within the muscle for a distinct volume of tissue as determined by the size of the probe. Using estimates for the increase in optical path length beyond probe spacing, known as the differential path length factor, continuous wave (cw) spectrometers measure changes in the attenuation of 2–6 wavelengths of light, allowing for algorithms based on the modified Lambert-Beer Law to provide good estimates of tissue deoxyhemoglobin (HHb) and oxyhemoglobin (HbO2) concentration changes [21]. The ultimate goal is to accurately measure chromophore concentrations in living tissue. It is not, however, a very simple task to achieve accurately. The difficulty in this accurate measurement has led to the development of several types of NIRS methods. Pulse modulated spectroscopy equipment averages the amount of time and therefore distance that photons travel by modulating the intensity of emitted light [14]. Time resolved spectroscopy equipments detects the amount of time photons remain in the tissue by the emission of pulsed laser with synchronized detection allowing for picoseconds resolution [101]. Spatially resolved spectroscopy equipments utilizes multiple probe space distances to analyze the attenuation of the light signal using an estimated or calibrated tissue scattering coefficient [73]. Also combinations of these methods are available and require the use of equations based on diffusion theory for an estimate of the tissue absorbance coefficient [18].

A further goal with NIRS however, is to measure chromophores accurately without the need of tissue scattering assumptions. The methods described above all use algorithms based on the modified Lambert-Beer law [21, 53] which in turn is based on two assumptions: (1) the absorption of the tissue changes homogeneously, and (2) the scattering of light is constant. It is known, however, that absorption changes are usually inhomogeneous, and therefore the modified Lambert-Beer law underestimates the changes in concentrations (partial volume effect) and every calculated value is influenced by the change in the concentration of other chromophores (cross-talk between chromophores). Three biological compounds show oxygen dependent absorption spectra in the near infrared range of light: Hb (Hemoglobin), Mb (myoglobin) and (Cyt aa3) Cytochrome c Oxydase [96]. To accommodate the accurate measurement of hemoglobin forms (chromophores) the NIRS equipment utilized in this thesis was an Inspectra 325 spectrometer, Hutchinson Technology, MN, USA. The Inspectra 325 is a continuous wave spectrometer that uses derivative spectroscopy to establish the chemical composition of biological fluids. The Inspectra 325 applies a “wide gap” (40 nm wave length) second derivative algorithm method. At one end of the probe light is emitted at wavelengths of 680, 720, 760, and 800 nm into the muscle, then sensors at the other end measure the amount of that light absorbed by the different hemoglobin forms. These wavelengths cover the spectrum that includes those that are sensitive for oxy-hemoglobin (HbO2) and deoxy-hemoglobin (HHb) (see caption). This method allows for the quantification of hemoglobin oxygen saturation (StO2%) and requires fewer wave lengths which simplifies the instrument and allows for application in a variety of tissues exhibiting different scattering

7

characteristics [80]. The depth of the measurements is a function of the size of the probe with the emitted light and the receiving sensor, and is believed to be half the size of the probe [17]. The Inspectra 325 can be used with a number of probe sizes. For this thesis we used a 25 mm probe which allows for a maximum penetration depth of 23 mm. The bulk of light penetration and measurement however is half the size of the probe – i.e. 12.5 mm [17]. The probe was inserted into a sticky shield placed onto the skin. This facilitated placement for repeated measures over different days. Below is a schematic figure of a NIRS probe placed over the muscle of interest to demonstrate the principle:

Fig 1. Near infrared light covering the wavelength spectrum from 680-800 nm traverses into the muscle and is reflected or absorbed. Oxyhemoglobin (HbO2) and deoxyhemoglobin (HHb) absorb light at different wavelengths. In this way the percent saturation (StO2%) is calculated as HbO2/(HbO2 + HHb) x 100. Total haemoglobin (HbT) is calculated as HbO2 + HbO. In this figure the 25-mm probe emits light that penetrates to a muscle tissue depth of approximately 12.5 mm (i.e. half of the distance between sending and receiving).

The modified Lambert-Beer Law The Lambert-Beer law also known as Beer’s law states that light intensity decreases exponentially when sent through a sample if the thickness of the sample is increased or if the concentration of the medium is increased. In other words it states that the transmission of light through a medium is dependent on a logarithmic function between the transmission of light and the product of the absorption coefficient of the substance and the distance the light travels through the material (or its path length). The absorption coefficient can be, sequentially computed as a product of the molar

Skeletal Muscle

Epidermis

Detection

23mm95% SignalThreshold

Patient InterfaceIllumination

25mm

8

absorptive (extinction coefficient) of the absorber and the concentration of absorbing species in the material.

The so called Modified Lambert-Beer Law and is central to the operation of functional NIRS. It makes it possible for NIRS to be applied without the need for extensive pre-processing of the sample. The modified Beer-Lambert law states that the change in light attenuation is proportional to the changes in the concentrations of tissue chromophores, mainly oxy- and deoxyhemoglobin in human tissue. If attenuation changes are measured at two or more wavelengths, concentration changes can be calculated.

Validity of NIRS for muscle tissue

The validity of NIRS to measure and evaluate oxygenation and hemodynamics in muscle tissue has been well recognized and tested under a diversity of experimental conditions in humans. Mancini et al. [72] validated NIRS for the flexor muscles of the forearm against antecubital vein blood samples. Also by the use of laser flow Doppler no influence of skin blood flow was found. Furthermore using arterial infusions of nitroprusside and NE NIRS signals showed improved blood perfusion with nitroprusside and significant deoxygenation with NE. In the same study Mancini et al. [72], also showed NIRS to be a better measurement method than plethysmography. Similarly De Blasi et al [19] validated NIRS against plethysmography during venous and arterial occlusions and concluded that NIRS was simpler and better method that can provide more precise information without discomfort for the subjects. In the same line Homma et al [37] validated NIRS comparing to strain-gauge plethysmography and a blood gas analyser for O2 ,and Hamaoka et al [34] and Sako et al [92] have also validated the NIRS method during arterial occlusion and a dynamic hand grip exercise, respectively. Furthermore Van Beekvelt et al [104] compared NIRS with the Fick method and Plethysmography and showed that NIRS is an appropriate method to provide information on local muscle oxygenation and hemodynamics in the muscle.

Reliability of NIRS

Currently, there are but a few studies that have looked at the reliability of NIRS measurements on an intraday [79, 104] and interday basis [13, 51, 54, 58, 71, 79, 105]. There is no agreement, however, among the studies on the metrics for reliability and on the practical interpretability (i.e. application) of the statistical metrics proposed in these studies to address different aspects of reliability (e.g. intraclass correlation coefficient, coefficient of variation, etc.). The term reliability may refer to the consistency of a measurement (i.e. the random dispersion of error repeated measurements within a subject), but also to the ability of a method to differentiate among subjects; metrics for these interpretations are different and should be discussed as so [38, 99].

9

Hence it is an important precondition in the application of NIRS is to understand the test retest reliability of NIRS variables and to assess appropriate metrics for measurement and what extent changes can be expected between measurements for our NIRS variables.

Factors affecting the NIRS signal

As shown by several investigators [28, 63, 103, 109] an important factor that may influence greatly in oxygenation and hemodynamics responses to NIRS signal is adipose tissue thickness (ATT). Homma et al [37] also found an influence of ATT while measuring changes in optical densities in the hallucis longus muscle. For our studies all subjects underwent a preparation process where among other procedures ATT was carefully assessed and recorded by the use of an ultrasound imaging (Aloka SSD-2000, Aloka Co., Japan). Skin blood flow and myoglobin are also factors to consider. Both however, have been shown to account for very little of the NIRS signal [8, 72].

Electromyography (EMG)

Muscle force is produced when a group of muscle fibers are activated by their common motor nerve, in a so-called motor unit (MU). Surface electromyography (EMG) registers the electrical activity of a number of motor units being activated and the signal amplitude, typically measured through the root mean square (RMS) value of the EMG signal, is a measure of the extent of muscle activity. Thus, RMS is a commonly used metric of surface EMG and its variations with force can be used to represent changes in the number of active MU and/or their mean firing rate [76].

The mean power frequency (MPF) is another commonly used variable derived from surface EMG. MPF is reportedly useful for determining muscle fiber type arrangement showing, in general, higher values for muscles with greater percentages of type II fibers [31, 32] or greater relative areas of type II fibers [56]. Unlike RMS, MPF responses with force are not straightforward. For example, for increasing mid-range submaximal contractions MPF was shown to increase with force for the vastus lateralis muscle [6, 31], but to decrease with force for the biceps brachii. This may suggest that MPF changes are muscle dependent.

RMS appears to respond in a consistent manner for sustained contractions. During sustained submaximal contractions the motor unit discharge rates usually decrease over time. To compensate for this, additional motor units are recruited [26], which means that RMS tends to increase over time. MPF, on the other hand, although considered, together with the increase in RMS, as a classical sign of fatigue development with a decrease over time mostly attributed to changes in the muscle fiber membrane and MU control properties causing a decrease in the myoelectric conduction velocity [4, 69]. MPF decrease with sustained contractions has also been shown to not be straight forward and may depend on contraction level and muscle [35, 40].

10

NIRS and EMG

Over recent years there has been growing interest to combine NIRS and EMG to investigate oxygenation changes with varying muscular activity levels [87]. Both methodologies are non-invasive, presenting an advantage in experimental and clinical settings, and combining NIRS, as an indication of the energy consumed for muscle contraction, with surface EMG, as an estimate of the level of muscle activity, can provide important insight into the functional and metabolic properties of muscle tissue. In healthy subjects the general finding is that an inverse linear relationship exists depicted as decreasing oxygenation and increasing EMG amplitude in response to increasing contraction force levels. However, very few studies [77]; [87] have addressed the relationship between EMG amplitude and oxygenation for varied contraction loads. Miura and co-workers [77] investigated EMG amplitude and oxygenation for the vastus lateralis during cycling exercises at different loads and Praagman and co-workers [87] investigated the biceps breve and brachioradialis muscles for isometric contractions from 10% to 70% of maximum voluntary contraction (MVC). For both studies good correlations depicted as increasing EMG amplitude (RMS) and decreasing oxygenation were demonstrated. Thus, these variables may share a common mechanism or are causally related. Furthermore, recently, Felici and co-workers [27] showed good correlations between the slopes of oxygenation change at onset and the median power frequency (EMG signal) slopes during sustained submaximal contractions.

These previous studies investigated NIRS-EMG for the biceps, brachioradialis and vastus lateralis muscles and our concern is to examine the relationship for the ECR and TD muscles – both prone to work-related muscle pain.

NIRS and WRMP

Some studies have found differences between patients with WRMP and healthy subjects. Brunnekreef and co-workers [9] used NIRS to report a decrease in oxygen consumption and reduced blood flow in the forearm patients with WRMP. In addition, Sjogaard et al. [97] showed that oxygenation patterns during pegboard work differed between patients with work-related trapezius myalgia and healthy subjects. However, Flodgren et al. [30] found no differences in oxygen saturation patterns between trapezius myalgia patients and healthy controls during low-level repetitive arm work. Discrepancies between studies are not clear. For this thesis we will used a comprehensive approach employing a number of NIRS derived variables to try to study WRMP.

11

Aims of the Thesis

General aims

With a reference point of studying work-related muscle pain, two general aims were addressed in this thesis: (1) evaluation of the NIRS technique for measuring in the upper extremity, and (2) using NIRS to test for differences between myalgic and healthy muscles.

Specific aims

To assess the relationship between oxygenation and muscle activation at different levels of force production for males and females for the extensor carpi radialis and trapezius muscles.

To determine the reliability of NIRS for inter-day measurements of oxygenation and hemodynamics for the extensor carpi radialis and trapezius muscles.

To investigate differences in oxygenation and hemodynamics during acute submaximal contractions in the upper extremities of patients with work-related muscle pain and healthy subjects.

To investigate differences in oxygenation and hemodynamics during a sustained low level muscle contraction and recovery in the upper extremities of patients with work-related muscle pain and healthy subjects.

12

Materials and methods

In this thesis two studies were performed to produce four papers (Fig 2). For

study 1 healthy males and females were recruited and used for the

production of papers I and II. For study 2 patients with WRMP sent from

occupational health care clinics and healthy controls were used for the

production of papers III and IV.

Fig 2. Shows the studies performed during the thesis and the respective papers derived.

Subjects

Study 1

For study 1 30 healthy subjects, consisting of 15 males and 15 females were

selected. Prior to testing subjects completed a questionnaire concerning their

level of physical activity and general health and none reported any pain or

other medical condition that required any kind of regular medication. All

subjects reported to be in good health and free of discomfort related to the

upper body. All subjects were right handed and non-smokers. Several

subjects reported to regularly practice sports only 1 reported competing at

local elite level. All subjects were asked to not workout at the gym or practice

any heavy activity that would affect the target muscles the day prior to

testing. Also, female subjects were asked to select dates for testing that did

not coincide with their menstrual cycles.

Study 1 – Two testing occasions Healthy males and females

Occasion 1 Occasion 2 (4-6 days separation)

Paper I Paper II

First part of the protocol

Study 2 Patients and Controls

Second part of the protocol

Paper III Paper IV

13

Table 1. Summary of the inclusion and exclusion criteria for study 1

Inclusion criteria Exclusion criteria

Healthy Chronic pain or injury in the upper extremities)

Right handed Medication (some exceptions)

Age (20-40 years old) Heavy smokers (>10 cig/day)

Body mass index (BMI) > 29

High blood pressure

Hemoglobin content under normal values

Study 2

For study 2 19 patients were referred from several health care clinics in the

Umeå and Västerbotten county regions in Sweden. Their diagnosis was

muscle pain or discomfort considered to be chronic (longer than 3 months)

and related to work. Patients were matched in sex and age with a healthy

control group free of any symptoms of MSDs. To further confirm patient

diagnosis and inclusion criteria both patients and healthy controls were sent

to a physiotherapy clinic and were reexamined by two separate

physiotherapists that independently performed an assessment analysis of the

probable cause of the pain or discomfort. Patients who fit the inclusion

criteria were then called back for testing. It is also relevant to point that for

this study, although it was not included in the inclusion criteria, patients and

controls had a similar level of recreational physical activity.

14

Table 2. Summary of inclusion criteria and exclusion criteria for study 2.

Healthy controls

Inclusion criteria Exclusion criteria

Healthy Chronic pain or injury in the upper extremities

Right handed Medication some exceptions

Age (20-65 years old) Heavy smokers (>10 cig/day)

Adipose tissue thickness > 9 mm

Hemoglobin content under normal values

High blood pressure

Patients (work-related chronic muscle pain)

Inclusion criteria Exclusion criteria

Right handed Medication with some exceptions

Age (20-65 years old) Heavy smokers

Pain or discomfort in the upper extremities

related to work > 3 months

Adipose tissue thickness > 9 mm

Hemoglobin content under normal values

High blood pressure

Rheumatologic pain

Peripheral or central neuropathic pain

Traumatic pain

Generalized pain

Disc herniation

Hyperalgesia

Neurological illnesses

Autonomic or affective pain

Low arterial systemic oxygenation

Methods (Studies 1 and 2)

For studies I and II NIRS was used to measure the local muscle tissue oxygenation and hemodynamics. In both studies NIRS measurements were collected from the ECR and TD muscles. For study 1 (first occasion only) and study 2 EMG was also used to assess muscle activity. Fig 3 depicts the placement of NIRS shields and EMG electrodes for the ECR and TD. EMG electrodes were attached to the skin perpendicularly to the NIRS probe shields in line with the muscle fibers as shown in the Fig 3.

For both studies data collection was performed at the Centre for Musculoskeletal Research (CBF) in Umeå Sweden. For both studies subjects underwent an array of preparation procedures prior to testing. Blood pressure measurements were first taken for each subject to ensure normal

15

blood pressures. Then fingertip blood samples were collected and analyzed for hemoglobin content and systemic arterial oxygenation was measured with a finger oxymeter to ensure normal levels. Next, was to identify sites of NIRS measurement for the forearm ECR and shoulder TD muscles for the right side only. For the ECR the measurement site was 2 cm distal to a point that was half-way between the supracondylar ridge and the inner bend of the elbow. The site on the TD was one-third the distance (in a lateral direction) between C7 and the acromion process of the scapula. Ultrasound imaging (Aloka SSD-2000, Aloka Co., Ltd., Japan) was then performed at each respective site to get a measurement of the distance from skin to muscle (i.e. adipose tissue thickness, ATT) and muscle thickness (Mth). NIRS and EMG probes were placed at the appropriate muscle sites. The NIRS probe was placed over the ECR and TD in the direction of the muscle fibers. An adhesive shield with a window was attached to the skin to secure the probe. The shield was positioned by the same investigator on all occasions with the same landmark calculations used for all subjects. Thus, the NIRS probe was placed over the muscle paths of the ECR and TD in accordance to the sites described above. For papers I, III and IV the EMG electrodes were positioned just under the lateral flap of the NIRS shield and care was taken so they did not block the window where the probe was inserted. This assured that NIRS measurements were in close proximity to EMG recordings. For all studies the distance between light emitter and detector for the NIRS probe was 25 mm, and the distance between EMG electrodes was ~20 mm. For all studies the NIRS sampling frequency was 0.35 Hz. For paper IV also EKG was used and electrodes were placed at the level of the sixth left rib and the distal end of the sternum.

ECR TD Fig 3. Placement of NIRS shields and EMG electrodes on the ECR and TD muscles just before a testing occasion.

16

Figure 4 shows the protocol used for study 1 that originated papers I and II.

Fig 4. Shows the protocols used for study 1 which originated papers I and II. Data collected in day 1 were used for paper I and data collected in days 1 and 2 for paper II

In paper I (study 1) NIRS and EMG data were collected during submaximal isometric contractions of 10%, 30%, 50% and 70% MVC for 20 s each and during a sustained contraction of 10% MVC for 5 min. During the submaximal isometric contractions a value for each NIRS variable (i.e. ΔStO2% and ΔHbT) was computed as the difference of the average of the last three data points just prior to relaxation and the average baseline value before each contraction. For the sustained 10% MVC contraction StO2% and HbT values were sampled continuously (i.e. sampling at 0.35 Hz) for the duration of the contraction, i.e. 5 min. For EMG amplitude, the signal was computed as root mean square (RMS) as a percentage of EMG at MVC; for frequency, the mean power frequency (MPF) was computed as one value for each submaximal contraction. For the sustained contraction RMS and MPF were computed in 10 s windows yielding 30 values for the 5 min.

Paper II (study 1) was a reliability study so testing was conducted on two days separated by 4 to 6 days. NIRS data were collected on both days during submaximal isometric contractions of 10%, 30%, 50% and 70% MVC and during the recovery period for the ECR and TD muscles. For submaximal isometric contractions NIRS data were processed as described above. For oxygenation recovery after contractions (30%, 50% and 70% MVC) the slope of recovery (Rslope) of StO2% was calculated for the values recorded for the initial 25 s following the cessation of contractions and regressed over time. For the ECR only venous (VO) and arterial (AO) occlusions were performed by applying an upper arm pneumatic cuff (Fig 5). VO was used to obtain an estimate of muscle blood flow and oxygen consumption and AO was used to

Study I

Day 1

ECR / TRAP

Rest1min

Rest2min

Rest5min

10%5min

Max 15sec

Max 25sec

Max 35sec

30%20sec

50%20sec

70%20sec

PreparationRest2min

Rest2min

Rest2min

Rest2min

EMG

NIRS

Venous/arterialocclusion

Day 2

Rest1min

Rest2min

Rest5min

10%5min

Max 15sec

Max 25sec

Max 35sec

30%20sec

50%20sec

70%20sec

Preparation Rest2min

Rest2min

Rest2min

Rest2min

NIRS

Venous/arterialocclusion

Paper I

Paper II

17

obtain oxygen consumption as well. For VO, after subject positioning and a rest period of 1 min, a pneumatic cuff was rapidly inflated to 50 mmHg. The time to inflation was 3-5 s. At 50 mmHg venous outflow is blocked, whereas arterial inflow is not affected [19]. Pressure was maintained for 1 min and for this time period the rate of increase in total hemoglobin (HbT) was used as an indication of blood flow, and the rate of increase in deoxyhemoglobin (HHb) was used to indicate oxygen consumption. For AO, the cuff was inflated to a suprasystolic pressure of 250 mmHg and maintained at this level for 1 min during which full occlusion occurred. Again, the rate of increase in HHb (just as for VO) was used to indicate oxygen consumption, while we confirmed that HbT remained constant. These procedures have been validated against standard methodologies in previous studies [19] [37] [104] and were used in this thesis for studies II (VO, blood flow and oxygen consumption and AO, oxygen consumption), III (VO, blood flow and AO, oxygen consumption) and IV (VO, blood flow). Furthermore after AO the cuff pressure was abruptly released and the subjects rested quietly for 30 s to gauge oxygenation recovery.

Figure 5 shows the protocol used for study 2 that originated papers III and IV.

Fig 5. Shows the protocols used for study 2 which originated papers III and IV. Data collected in the first part of the protocol were used for paper III and data collected in the second part of the protocol were used for paper IV

For paper III (study 2) participants performed submaximal isometric contractions of 10%, 30%, 50% and 70% MVC randomized in a Latin square design for 20 s each while NIRS and EMG were collected from the ECR and TD muscles. Also for the ECR VO and AO were performed. NIRS data were analyzed as described above for in papers I and II for oxygenation during contractions, VO, AO and oxygen recovery after contractions and AO. The EMG signal was analyzed as described for study 1.

For paper IV (study 2) participants performed a sustained contraction at 15% MVC for a maximum time of 12 min, or for as long each participant

Study 2

Randomized

10% - 30% - 50% - 70% MVC20 sec each

EMG

NIRS

MVC x 35 sec each

Venous/Arterialocclusion

15% MVCto fatigue

0min

15min

30min

Venousocclusion

Preparation

Paper III

Paper IV

18

could hold the contraction. Also if force level dropped 20% from the original 15% MVC level for more than 3 s, it was considered a sign that the participant could no longer maintain the established force level and the contraction was stopped. Before the contraction VO and AO were applied, as well as a 10 s test contraction. During the recovery period three VOs were applied one immediately after the contraction and subsequently two more every 15 min. Also three 10 s test contractions (TC) were used to gauge recovery after the sustained contraction. Furthermore subjective ratings of fatigue sensation were collected before the sustained contraction and during the recovery period. NIRS, EMG and EKG data were collected continuously.

Work model (Studies 1 and 2)

For both studies the work model was the same. It was designed to focus as much as possible on activity in the ECR and TD muscles, and, at the same time, allow for simultaneous measurements of NIRS and EMG. Fig 6 shows the work set up for both the ECR and TD muscles and the positioning for both VO and AO for the ECR.

For the ECR submaximal isometric and sustained contractions, subjects sat comfortably with the back supported in a sturdy upright chair. The forearm was placed on a special designed arm support with the wrist in a pronated position, palm facing downwards, and placed at the edge of a height adjustable table. For each subject the height of the arm support was adjusted to keep the shoulders horizontally lined up avoiding trunk bending sideways. The subjects were instructed to sit in an upright position to obtain an elbow joint angle of ~100 degrees. A vertical strap was positioned over the dorsum of the hand; the other end of the strap was attached to a strain gauge dynamometer that was bolted to the floor next to the chair. The task was to extend the wrist joint in an attempt to pull the strap upwards while producing force. Forearm and wrist remained in contact with the testing table for the whole duration of procedures, thereby minimizing the action of upper arm and shoulder muscles on the exerted force.

For the TD submaximal isometric and sustained contractions, participants were again seated in the testing chair as described above. The arms were positioned vertically along the body so as to minimize the involvement of biceps and deltoid muscles during force production [102]. With the right hand, subjects grabbed a handle securing a firm grip (Eleiko – 22 in. handle straps). The handle was connected to a strap attached to the strain gauge dynamometer used also for the ECR. Also the same type of strap with handle arrangement was used on the subject’s left side but this was attached to a fixture on the floor, i.e. force measurements were only done for the right side. Both handles were aligned with the straight arms and adjusted at heights that made shoulders line up horizontally. Participants were instructed to produce isometric force by attempting to elevate the shoulder joints on both sides of the body simultaneously.

19

For VO and AO participants were moved to a hospital gurney (Sjöbloms AB, Sweden) and were seated in a reclined position with the forearm at heart level. Following the pneumatic cuff was placed around the upper arm and rapidly inflated to a pressure 50 mmHg to attain VO and 250 mmHg to attain AO and maintained for 1 min.

ECR TD

VO/AO occlusions

Fig 6 Work set up for both the ECR and TD muscles and VO and AO for the ECR.

20

Statistics (papers I, II, III and IV)

Statistical analyses and computing were performed with IBM SPSS statistics (IBM Inc. Chicago, IL, USA) versions 15.0 for papers I and II and 20.0 for papers III and IV. For papers I, III and IV a P-value of < 0.05 was considered statistically significant. For papers III and IV a P-value of 0.10 was considered a limit for tendencies.

Paper I

For the graded contraction data, linear regression analyses were used to determine correlations for ΔStO2% to RMS and to force at group level, and for males and females separately for each muscle. In addition, full factorial repeated measures analyses of variance (ANOVA) were used to determine if changes over force were gender dependent. A two-way ANOVA was used with force level (i.e. 10-70%) as the within subject factor and gender as the between subject factor. For the sustained contraction an ANOVA was used to test differences with time and gender as factors.

Paper II

Variance components were determined using a standard linear, hierarchical residual maximum likelihood model giving estimates of the between subjects (s2

bs) and within subjects between days (s2ws) variability. To discriminate

between subjects, standard intraclass correlation coefficients, ICC (1,1), were

computed as:

For appreciation of the minimum detectable

significant difference of measurements between trials, limits of agreement (LOA) were computed [38]. The LOA were calculated as √s2

ws · tα · √2, where tα is the Student’s t-value with degrees of freedom as determined by the population size (tα for total population 1.96, tα for each gender group 2.18). Confidence intervals (CI) on ICC estimates were calculated using the reliability function in SPSS 15.0. CI for the s2

bs and s2ws were calculated in

MatLab (MatLab works 7.4 - R2007a) using customary algorithms and negative lower bounds for variance component CIs were replaced by 0 [95].

Paper III

To determine possible differences for the between controls and patients for all NIRS and EMG variables during submaximal isometric contractions and NIRS recovery data for the ECR and TD, a mixed model with the function mixed in SPSS was used with group as fixed effects, subjects as random effects and force as a covariate [95]. For the ECR VO data an independent Student’s t-test was performed to test for possible differences between controls and patients for HbT slopes. For AO because the data were not normally distributed a Mann-Whitney test was used to assess possible differences in HHb slopes between controls and patients. For assessing

21

possible differences between controls and patients for StO2% recovery after AO an independent Student’s t-test was performed.

For the ECR VO data a One-Way ANOVA test was performed to assess possible differences in HbT slopes between controls and patient groups. For AO data a One-Way ANOVA test was performed to assess possible differences in HHb slopes and StO2% recovery slopes between controls and patient groups. In the cases where significance was found between groups a Post Hoc test was performed and Sidak corrections of the P values were applied.

Paper IV

For NIRS and EMG variables during the sustained 12 min contraction full factorial repeated measures analyses of variance (ANOVA) were used to assess differences between patients and controls over time. The dependent variables StO2%, HbT, RMS and MPF were analyzed in the ANOVA model with time as the categorical independent variable and group (patients and controls) as the between-subjects factor.

For VO and fatigue ratings full factorial repeated measures ANOVAs were used to assess differences between patients and controls over measuring occasions with group (patients and controls) as the between-subjects factor. If the assumption of sphericity was not met, a Greenhouse-Geisser correction was applied.

22

Results

For papers I and II male subjects showed significantly higher MVC and MTh for both the ECR and TD whereas, ATT was significantly higher for females in paper I for both the ECR and TD. ATT in paper II was not significantly different between males and females for both muscles, despite of a tendency for the TD, because 1 male and 3 female subjects were removed from analyses due to high ATT (i.e. over 9 mm) or fault in data collection in the second day of testing which influence ATT differences. Furthermore, for both papers male subjects showed higher resting levels of StO2% for both the ECR and TD.

For papers III and IV for the both the ECR and TD muscles MVC and MTh were not significantly different between patients and controls. ATT was significantly higher for the TD of patients. For the TD controls showed a higher level of StO2% at rest. Endurance time was significantly lower for patients compared to controls for both the ECR and TD. For the ECR endurance times were 346.80 s and 582.10 s (P = 0.01), for patients and controls, respectively, and for the TD endurance times were 430.00 s for patients and 723.50 s for the controls (P = 0.03).

Paper I

For paper I at group level a good correlation was shown as an increasing muscle activity (EMG – RMS) and decreasing oxygenation (∆StO2%) for the ECR and a moderate correlation for the TD muscle as depicted by the equations below. At gender level (Fig 7) also good correlations were found for males and females for the ECR. For the TD a good correlation was found for males and a moderate correlation was found for females.

ECR: ∆StO2% = - 10.28 – 0.56 x %RMSmax (r = - 0.53, P <0.001)

TD: ∆StO2% = - 1.81 – 0.38 x %RMSmax (r = - 0.44, P <0.001)

For the ECR (Fig 8 left) over submaximal isometric contractions, the increasing levels of force showed a significant effect on ∆StO2% (P < 0.001), reflected as a decrease in ∆StO2% with increasing force. There was no significant impact of gender in this response, however there was a tendency for difference between genders with females demonstrating larger changes than males (P = 0.13). There were significant differences in ∆HbT over force (P < 0.001) and these were significant between gender (F = 3.67; P = 0.02).

For RMS a significant difference was shown over force (P < 0.001) but there was no impact of gender (P = 0.41). Also for MPF, there was a significant difference over force (P < 0.001) but there was no impact of gender (P = 0.18).

23

For the TD (Fig 8 right), there was a significant effect of force on ∆StO2% (P < 0.001) but there was no impact of gender (P = 0.60). For ∆HbT, there was no significant change over force, albeit a tendency, (P = 0.07); however, gender was not a factor (P = 0.49).

For RMS a significant difference was shown over force (P < 0.001) but there was no impact of gender (P = 0.09) although there was a tendency with females demonstrating a larger RMS response than males. For MPF, there was a significant difference over force (P < 0.01) but there was no impact of gender (P = 0.29).

For the sustained contractions for the ECR (Fig 9 left) there were significant changes in StO2% over time (P < 0.001); however, these responses were not gender dependent (P = 0.91). Also, HbT changed significantly over time (P < 0.001) and the responses were different between genders (P = 0.04).

RMS increased significantly over time (P < 0.001) and the responses were different between genders (P < 0.01). The increase over time for females was larger than that for males. MPF decreased significantly over time (P < 0.001) but this was not impacted by gender (P < 0.57).

For the trapezius (Fig 9 right), StO2% changed significantly over time (P < 0.001) and these responses were gender dependent (P < 0.001). HbT did not change significantly over time (P = 0.99)

RMS increased significantly over time (P < 0.01) but the responses were not different between genders (P = 0.84). For MPF there were no significant changes over time, albeit a tendency, (P < 0.06); however there were no differences between genders (P = 0.44).

24

Fig 7 shows the correlations between ∆StO2% and RMS obtained from 20 s submaximal isometric contractions with increasing force levels of 10, 30, 50 and 70% MVC for Males and Females ECR (left) and Males and Females TD (right).

-100

-80

-60

-40

0

20

-20

20 40 60 80 100 120

0

StO

%2

RMS

ECR Males

r = -0.52

-100

-80

-60

-40

0

20

-20

20 40 60 80 100 120

0

StO

%2

RMS

Trapezius Males

r = -0.56

-100

-80

-60

-40

0

20

-20

20 40 60 80 100 120

0

StO

%2

RMS

ECR Females

r = -0.56

-100

-80

-60

-40

0

20

-20

20 40 60 80 100 120

0

StO

%2

RMS

Trapezius Females

r = -0.36

25

Fig 8 shows the mean values for ∆StO2%, HbT, RMS and MPF for males (●) and females (○ ) during graded isometric contractions of 10, 30, 50 and 70% MVC.

ECR

0

-60

-50

-40

-30

-20

-10

-70

S

tO%

2

Males

Females

0

-20

-15

-10

-5

-25

Hb

T

5

60

0

20

40

-20

RM

S

80

100

60

70

80

0 7010 30 50

MP

F

Force (%MVC)

90

100

110

120

130

140

Trapezius

0

-60

-50

-40

-30

-20

-10

-70

Males

Females

S

tO%

2

0

-20

-15

-10

-5

-25

Hb

T

5

60

0

20

40

-20

RM

S

80

100

60

70

80

0 7010 30 50

MP

F

Force (%MVC)

90

100

110

120

130

140

26

Fig 9. Mean ∆StO2%, HbT, RMS and MPF recorded during the 5 min isometric contractions (left) represents the ECR and (right) the TD.

ECR

48

Hb

T (

A.U

.)

60

10

RM

S (

%M

VC

)

500

MP

F (

Hz)

Time (minutes)

150

Trapezius

50

S

tO%

)2(

Males

Females55

60

65

70

75

80

85

90

50

Males

Females55

60

65

70

75

80

85

90

50

52

54

56

58

48

60

50

52

54

56

58

11

12

13

14

15

16

17

18

10

11

12

13

14

15

16

17

18

70

90

110

130

1 2 3 4 550

0

Time (minutes)

150

70

90

110

130

1 2 3 4 5

Hb

T (

A.U

.) M

PF

(H

z)

RM

S (

%M

VC

)

StO

%)

2(

27

Paper II

Table 3 shows the reliability parameters for the group data for the ECR.

For resting StO2% ICC was 0.65 and LOA was 16.6.

For ΔStO2% in isometric contractions ICCs ranged from 0.58 (10% MVC) to 0.95 (30% MVC) and the lowest ICC was at 10% MVC. LOA was lowest at 10% MVC and highest at 50% and 70% MVC.

For HbTslope ICC was the overall lowest 0.17. For HHbslope ICC was higher for AO than for VO. LOA for HHbslope was lower for AO than for VO.

For Rslope in the submaximal isometric contractions ICCs were quite similar for all contraction levels but the highest was for the 70% MVC. ICC for Rslope after AO was lower than for isometric contractions. The lowest Rslope LOA was for the 70% MVC and the highest for the 50% MVC.

Table 4 shows the reliability parameters for the group data for the TD.

For resting StO2% ICC was 0.82 and LOA was 16.1.

For ΔStO2% during isometric contractions ICCs ranged from 0.55 (10% MVC) to 0.79 (70% MVC). The lowest ICC was at 10% MVC. For Rslope in submaximal isometric contractions ICC was higher for 30% MVC and lowest for the 70% MVC. The lowest LOA was for the 30% MVC and the highest LOA for the 70% MVC.

28

Table 3. ECR data for pooled subjects for resting oxygen saturation (StO2%), saturation changes during submaximal isometric contractions (∆StO2%), venous (VO) and arterial (AO) occlusion parameters and recovery slopes (Rslopes) post isometric contractions and AO. Data are presented for mean values of Day 1 and Day 2, grand mean value of both days, between subjects variance (s2

bs), within subject variance (s2ws), intraclass correlation coefficient (ICC),

and limits of agreement (LOA). 95% confidence intervals (C.I.) are included.

† Outlier excluded from all calculations.

ECR Day 1 Day 2 Days mean [95% C.I.]

s2

bs [95% C.I.]

s2ws

[95% C.I.]

ICC [95% C.I.]

LOA

Rest StO2% 66.8 70.4 68.6

)

58.6 32.3 0.65 16.6

[64.9;72.3]

[18.7;149.6] [19.7;62.4] [0.34;0.83]

Contractions (∆StO2%)

10% MVC -8.6 -9.4 -8.9 20.9 15.2 0.58 11.4

[-11.2;-6.7] [5.0;55.8] [9.3;29.5] [0.24;0.80]

30% MVC -36.4 -37.0 -36.7 651.4 31.5 0.95 16.4

[-47.6;-25.8] [353.2;1444.1] [19.2;61.1] [0.90;0.98]

50% MVC -44.9 -46.9 -45.9 473.5 111.6 0.81 30.9

[-55.6;-36.2] [217.3;1109.3] [68.1;216.0] [0.61;0.91]

70% MVC -50.6 -53.4 -52.0 457.9 100.7 0.82 29.4

[-61.5;-42.5] [213.4;1068.0] [61.4;194.9] [0.63;0.92]

VO/AO Occlusion (au∙3.5∙sec-1)

HbTslopeVO † 0.23 0.22 0.22 0.0015 0.0073 0.17 0.25

[0.19;0.25] [0.0;0.0085] [0.0044;0.0144] [0.0;0.54]

HHbslopeVO 0.11 0.13 0.12 0.0032 0.0011 0.73 0.10

[0.10;0.15] [0.0013;0.0078] [0.0007;0.0022] [0.49;0.88]

HHbslopeAO 0.11 0.12 0.12 0.0018 0.0004 0.83 0.06

[0.10;0.14] [0.0009;0.0043] [0.0002;0.0007] [0.64;0.92]

Recovery (StO2∙3.5∙sec-1)

Rslope30 1.61 1.68 1.65 2.0475 0.2483 0.89 1.46

[1.03;2.27] [1.0439;4.6397] [0.1514;0.4806] [0.77;0.95]

Rslope50 2.83 2.67 2.75 1.7137 0.5416 0.88 2.15

[2.15;3.34] [0.8601;3.9038] [0.1454;0.4614] [0.74;0.95]

Rslope70 2.72 2.91 2.82 1.3872 0.1676 0.90 1.20

[2.30;3.33] [0.7076;3.1430] [0.1021;0.3242] 0.77;0.95]

RslopeAO 0.86 0.91 0.88 0.3868 0.7898 0.33 2.60

[0.51;1.26] [0.0;0.1376] [0.0481;0.1528] [0.00;0.64]

29

Table 4. TD data for pooled subjects for resting oxygen saturation (StO2%), saturation changes during submaximal isometric contractions (∆StO2%), and recovery slopes (Rslopes) post isometric contractions. Data are presented for mean values of Day 1 and Day 2, grand mean value of both days, between subjects variance (s2

bs), within subject variance (s2ws), intraclass

correlation coefficient (ICC), and limits of agreement (LOA). 95% confidence intervals (C.I.) are included.

Paper III

For the ECR during submaximal isometric contractions (Fig 10) there was a significant effect of force on ∆StO2%, ∆HbT, RMS%max and MPF, all P values < 0.001. For ∆StO2% there was no significant differences between patients and controls (P = 0.32). We did observe that the ∆StO2% responses during the 30% and 50% MVC were more marked for controls. ∆HbT was also not significant different between groups although there was a tendency for differences (P = 0.07) with controls showing a greater drop over the contractions. For RMS%max and MPF there were no significant differences between patients and controls (P = 0.82) and (P = 0.34) respectively.

TD Day 1 Day 2 Days mean [95% C.I.]

s2

bs [95% C.I.]

s2ws

[95% C.I.]

ICC [95% C.I.]

LOA

Rest StO2% 70.8 71.2 71.0 149.1 33.9 0.82 16.1

[65.5;76.4] [69.0;348.4] [20.6;65.5] [0.47;0.87]

Contractions (∆StO2%)

10% MVC -0.7 -2.9 -1.9 11.0 9.0 0.55 8.8

[-3.5;-0.1] [2.2;30.1] [5.5;17.5] [0.20;0.78]

30% MVC -14.7 -17.7 -16.2 226.6 68.7 0.77 23.0

[-23.0;-9.4] [97.2;541.1] [41.9;132.9] [0.54;0.89]

50% MVC -26.1 -28.8 -27.4 314.7 106.5 0.75 30.2

[-35.6;-19.3] [130.0;759.2] [64.9;206.2] [0.50;0.88]

70% MVC -31.8 -35.7 -33.7 372.5 96.9 0.79 28.8

[-42.4;-25.1] [166.9;878.9] [59.1;187.5] [0.58;0.91]

Recovery ( StO2∙3.5∙sec-1)

Rslope30 0.74 0.72 0.73 0.5372 0.1948 0.73 1.29

[0.39;1.07] [0.2161;1.3044] [0.1188;0.3771] [0.48;0.87]

Rslope50 1.36 1.40 1.38 0.7521 0.3139 0.71 1.64

[0.98;1.78] [0.2843;1.8539] [0.1914;0.6074] [0.44;0.86]

Rslope70 1.55 1.44 1.49 0.6615 0.3866 0.63 1.82

[1.10;1.88] [0.2009;1.7050] [0.6074;0.2843] [0.32;0.82]

30

Fig 10 The mean data for ∆StO2%, ∆HbT, RMS%max and MPF for controls (●) and patients (○) during submaximal isometric contractions of 10%, 30%, 50% and 70% MVC for the ECR (2a and the trapezius (2b). For the sake of clarity data are presented as means and SE.

For the ECR oxygenation recovery after submaximal isometric contractions there was a significant effect of force on Rslope (P < 0.001), but no significant differences between patients and controls (P = 0.14); however, we did observe that controls had higher average values, i.e. faster recovery than patients.

Fig 11 Oxygenation recovery slopes (Rslope) for the ECR after submaximal isometric contractions of 10, 30 ,50 and 70% MVC.

StO

2%

Rslo

pe

(S

tO2%

* 3

.5 s

-1)

31

For the ECR during VO for the pooled data showed no significant differences in HbTslope between patients and controls (P = 0.58).

Fig 12 Total hemoglobin slopes during VO for the ECR.

For HHbslope there was a significant difference between patients and controls (P = 0.03), with controls demonstrating a steeper slope than patients, i.e. more oxygen consumption during AO.

Fig 13 Deoxyhemoglobin slope (HHb slopes) for the ECR during AO for controls and patients

For the TD during submaximal isometric contractions there was as expected a significant effect of force on ∆StO2% (P < 0.001), ∆ HbT (P = 0.001), RMS%max (P < 0.001) and MPF (P = 0.04). For ∆StO2% there was no significant effect of group (patients and controls) and therefore there were no significant differences between patients and controls (P = 0.67). For ∆HbT there were no significant differences between patients and controls (P = 0.92). For RMS%max and MPF there were no significant differences between patients and controls (P = 0.83) and (P = 0.31), respectively.

*

*Shows significance

32

Fig 14 The mean data for ∆StO2%, ∆HbT, RMS%max and MPF for controls (●) and patients (○) during submaximal isometric contractions of 10%, 30%, 50% and 70% MVC for the TD.

For the TD oxygen recovery after submaximal isometric contractions there were no significant differences between patients and controls (P = 0.96), although on average controls showed a slightly steeper slope for the higher contraction levels of 50% and 70% MVC.

Fig 15 Oxygenation recovery slopes (Rslope) for the TD after submaximal isometric contractions of 10, 30, 50 and 70% MVC.

StO

2%

Rslo

pe

(S

tO2%

* 3

.5 s

-1)

33

Paper IV

For the ECR figure 16 shows the changes in StO2%, HbT, RMS and MPF during the sustained contraction.

For StO2% there were significant changes over time (P = 0.01), however these changes were not significantly different between patients and controls (P = 0.14). We observed a general trend for the changes with an initial decrease for both groups with controls showing a larger drop. Patients there off remained unchanged throughout the contraction while controls showed an increase and subsequent decrease pattern. Also, HbT changed significantly over time (P < 0.001) and differences were not significantly different between patients and controls (P = 0.26). As with StO2% the general trend was an initial decrease with gradual increase throughout the contraction and a similar pattern as of StO2% for both groups.

RMS increased significantly over time (P < 0.001) and the responses were not significantly different between patients and controls, however, there was a tendency for difference between patients and controls with controls showing a larger increase over time than patients (P = 0.09). MPF decreased significantly over time (P < 0.001) but there was no significant difference between patients and controls (P = 0.76).

Fig 16 Shows mean changes in StO2%, HbT, RMS and MPF for the ECR during the sustained contractions.

34

For the TD Figure 17 shows changes in StO2%, HbT, RMS and MPF during the sustained contractions.

StO2% did not change significantly over time (P = 0.40) and there was no difference between patients and controls (P = 0.37) although patients showed a lower StO2% value at the start and throughout the contraction. HbT did not change significantly over time (P = 0.18) and there were no differences between patients and controls (P = 0.38). RMS increased significantly over time (P < 0.001) but there was no difference between patients and controls (P = 0.46). For MPF there were significant changes over time (P = 0.04) but no significant differences between patients and controls (P = 0.44).

Fig 17 shows the mean relative changes in StO2%, HbT, RMS and MPF for the TD during the sustained contractions.

For the ECR subjective ratings there were significant changes over time (P < 0.001), but these changes were not significantly different between patients and controls (P = 0.32). The general trend however was for patients to report higher values then controls.

For the TD subjective ratings there were significant changes over time (P < 0.001), but these changes were not significantly different between patients

35

and controls (P = 0.19). There was, however, a trend for patients to report higher values then controls.

Fig 18 shows the changes for subjective ratings of fatigue prior to the sustained contraction, immediately after and during the recovery period for the ECR (top) and TD (bottom) for patients and controls.

36

Discussion An underlying idea for this thesis was to determine the ability of near infrared spectroscopy (NIRS), a non-invasive easy to apply technique that provides information on the local muscle oxygenation and hemodynamics, for studying work-related muscle pain. Based on previous findings using biopsies and direct measurements of blood flow, NIRS appeared as a relevant technique in the field that could allow continuous monitoring, repeatability and for a number of variables. Thus, the first approach was to extensively evaluate the method for its use for the upper extremity, and the second approach to test for differences between healthy and painful muscles. The discussion of the thesis will begin with a consideration of the methodology, e.g. protocols, NIRS equipment, etc., and then will be structured around the four specific aims.

Methodological considerations

The protocols used for studies 1 and 2 included isometric contractions, low-level sustained contractions, and upper arm cuff occlusions. This design was patterned to generate specific variables of interest fueled by previous studies. The short-term (20 sec) isometric contractions allowed determination of acute oxygenation use (∆StO2%) and recovery (Rslope). A concern here was not to produce fatigue. Both of these variables were shown to respond differently for painful versus non-painful muscles in previous studies that also used NIRS [9, 50]. Furthermore, using graded isometric contractions allowed for linearity assessment of oxygen use with myoelectric activity. The sustained contraction was intended to mimic a low-load activity, e.g. computer work for monitoring oxygen saturation (StO2%) and blood volume (HbT) patterns over time, i.e. variables of interest in a number of previous studies (e.g. [7, 30]). Fatigue generation was an acceptable outcome but was not a controlled outcome for the sustained contraction; instead the contraction was standardized by time. Upper arm cuff occlusions allowed for the estimations of blood flow (HbTslope) and oxygen consumption (HHbslope) since blood flow and oxygen consumption have been implicitly and explicitly associated with WRMP in previous studies (for review see Visser & Van Dieen [108]).

The NIRS equipment used did not provide absolute quantitative units for HbT or HHb but instead arbitrary units. This means that these variables were used as surrogates of blood flow and oxygen consumption, respectively. In a number of NIRS studies, blood flow and oxygen consumption data are reported in actual units, thus allowing for comparison of these measures across studies, which cannot be done with data in this thesis. Another limitation of the NIRS equipment was the relatively low sampling frequency – 0.35 Hz. This fact impeded performing some analyses. As examples, ½ recovery times and reoxygenation rates for StO2% after provocations could not be computed.

37

The healthy subjects recruited for study 1 were mostly university students, which meant that they were somewhat younger than the normal working population. This was a matter of convenience since the study was conducted in a small University town in the north of Sweden. This may be a consideration for generalizing our data since NIRS variables are known to be influenced by age [57]. The patients for study 2 were recruited from health clinics around the Umeå area. Patients had been diagnosed by their physicians as having pain that was related to their work. An additional screening was done by our research group to confirm diagnosis, which resulted in excluding some patients. Further criteria were that patients suffered pain on the right side (because of our measurement equipment) and for the neck/shoulder and forearm. These requirements presented some limitations on availability of patients. Thus, the number of subjects in the thesis may be a concern regarding statistical power; this is especially so when a number of different variables are being assessed each requiring a separate power calculation. An attribute of the patients is worthy to consider for the context of this thesis. The patients regularly participated in some type of recreational physical activity, were currently working, and were similar in muscle strength to the healthy control subjects. This may imply that this group was not as severe as groups in other studies, which could impact comparing studies.

As presented previously an important consideration is the impact of ATT on NIRS which is known to influence the signal during exercise [103, 109]. Our data showed significantly higher mean ATT for the ECR and TD of females than males (paper I) and for the TD of patients in papers III and IV. The NIRS light penetration depth is considered to be 12.5 mm (half the distance between optodes for the 25 mm probe) [17]. For our studies ATT was measured by ultrasound. For paper I although ultrasound results were not used ATT values showed that NIRS light was well within the muscle tissue for both the ECR and TD muscles (cf paper I). For papers II, III and IV participants with ATT higher than 9 mm were excluded, assuring that NIRS light was well within the muscle tissue. Therefore we believe that an adequate depth of muscle tissue was measured and our results were not influenced by the ATT differences. If influence of ATT had occurred, we would have seen a blunting affect such that of a lesser response in ∆StO2% during the contractions for females and patients, but this was not the case. Skin blood flow and myoglobin are also factors to consider.

38

Oxygenation and myoelectric activity for males and females (Paper I)

Isometric contractions

That we found significant negative correlations for the ECR and TD muscles for ∆StO2% and RMS for submaximal isometric contractions could indicate that these variables share a common mechanism or are causally related. These findings are in general agreement with previous studies for other muscles [77, 87]. Considering that ΔStO2% and RMS reflect different aspects of muscle contraction, establishing relationships between variables for the ECR and TD could be a point of reference for future studies of upper extremity disorders.

Gender specific patterns for ∆StO2% were observed for the ECR with females tending to show a larger drop than males. RMS, however, was the same for males and females. An explanation for the gender differences in ∆StO2% is not straightforward, but since females are more at risk for WRMP [82], a speculation could be that the larger usage of oxygen could cause a larger production of metabolites [43].

For the TD NIRS showed no differences in ∆StO2% over contraction levels; however, RMS tended to be higher for females. In consideration that females had a lower MVC, the absolute forces generated throughout the contraction would be therefore be lower for females than males. This implies that females use more motor units at relative force levels than do males. Lindman et al. [64] showed dysfunctional and more abundant type II muscle fibres in the TD of females than males which may comply with our finding of higher RMS in the muscles of females.

Sustained contraction

During the sustained contraction at 10% MVC for 5 min ECR StO2% changed significantly over time, but the changes were not different between genders. The initial drop in oxygenation continued unchanged for the whole duration of the contraction although a metabolic vasodilatation was seen allowing for more blood to come into the tissue (supported by our HbT data) (Fig 8). This is in contrast to Blangsted et al. [7] for a 10% MVC sustained contraction of the biceps where oxygenation dropped initially to quickly return to levels similar to the beginning of the contraction. An explanation may lie in the fact that the muscles studied between studies were not the same. In the present study there was a clear increase in RMS indicating an increase in motor unit activation, whereas in the Blangsted study [7] RMS remained constant throughout the contraction. The difference in genders in ECR HbT, exhibited as an increase for males that was larger than for females, suggest more vasodilatation for males; this could allow for a better metabolites removal. Our EMG data for the ECR showed an increase in RMS and decrease in MPF over the duration of the contraction indicating the development of fatigue

39

[84] [70]. There was a larger increase in RMS for females than males meaning that a larger number of motor units were recruited by female during the sustained contraction in order to maintain the same relative force level.

For the TD the changes in StO2% were to a certain extent different between males and females, although a similar pattern of decrease in StO2% was seen for both genders in the beginning of the contractions. Females continued to decrease throughout the contraction while males gradually recovered after the first 30 sec. RMS increased during the contraction similarly indicating the recruitment of more motor units to maintain contraction force. Although the circulation seems to be able to deliver enough blood to maintain demand since the HbT amount remains unchanged during contraction, females seemed to use more oxygen than males.

Interday reliability of NIRS variables (Paper II)

In paper II of this thesis, we have shown a comprehensive reliability chart for several NIRS variables during contraction and recovery as well as variables derived from arterial and venous occlusions. We have used intraclass correlation coefficients (ICC (1,1)) as a metric showing the ability of a method to differentiate between subjects, since an ICC (1,1) measures the relative size of within-subject variance s2

ws (test-retest reliability) to between-subject variance s2

bs, and thus the ability of a method to correctly identify that two (or more) individuals differ in the assessed variable. As a measure of absolute reliability or reproducibility of measurements we have used the limits of agreement (LOA). LOA are expressed in the actual units of the variable.

While some studies report interday test-retest reliability for a single contraction [51, 71], we found one recently published study of particular relevance that examined forearm oxygenation responses during handgrip exercise at various intensities ranging from 20% MVC to maximal adjusted in 10% MVC increments [13]. Celie and co-workers [13] reported ICC for 20% MVC to be the highest, and that ICC values decreased with increasing contraction intensity. This finding may somewhat concord with our data. Two other studies investigated interday reliability of oxygenation responses for multiple contraction forces [79, 105]. With both studies using CV as a generalized metric of reliability, Muthalib and co-workers [79] reported a lower CV at 100% MVC compared to 30% MVC and thus recommended the higher contraction force for future clinical applications. Van Beekvelt and co-workers [105] found uniform CVs ranging from 16% to 23% for isometric contractions at seven intensities (10% to 90% MVC) concluding good reliability over a wide range but did not give preference to which intensity was ‘most reliable’.

For our data the highest ICC for ECR-ΔStO2%, i.e. the best for discriminating between subjects, was at 30% MVC, and the lowest ICC was at 10% MVC.

40

From a mathematical point, and as mentioned above, it is clear that the relative association between s2

ws and s2bs supports the ICC magnitudes for

these contraction levels. For that matter a further generalization, encompassing data for both the ECR and TD muscles, can be made that ICC values were comparably high for the 30%, 50% and 70% MVCs thereby routing the discussion as to why the 10% MVC contractions had lower ICCs. One explanation might be based on differences in intramuscular pressures (IMP) between contractions. NIRS signals reflect the dynamic balance between oxygen delivery and consumption and a change in signal is not clearly distinct for either delivery or consumption. At contraction levels > 25% MVC IMP is sufficiently high to disrupt blood flow and impact local muscle oxygenation [20, 41]. Thus, in our data for the 30%, 50% and 70% MVCs factors influencing ΔStO2% are presumed to include disrupted circulation and oxygen consumption, whereas at 10% MVC circulation may be only minimally impaired (also evidenced by a small decrease in total haemoglobin paper I). The more ’mechanical’ impact on the data afforded by the higher contractions may have contributed to better repeatability between days (i.e. higher ICCs) than for the 10 % MVC.

For VO the ICC value was 0.17, and is due to the low s2bs(s) suggesting that

blood flow measured with NIRS is not a suitable method for discriminating among our group of healthy subjects, while at the same time indicating a physiologically interesting consistency among subjects. The situation might be different for a patient group exhibiting a more varied hemodynamic pattern due to various stages of their anomalies. Since the variability within subject was rather high for HbTslope, it may be possible to increase the reliability in terms of ICC by increasing the number of repeated determinations on each subject. We found no other study in the literature that investigated between-days reliability of NIRS-derived forearm blood flow; however, there are some data on within-day (”intraday”) reproducibility. Van Beekvelt and co-workers (2001) interpreted the lack of a significant difference in the group mean, albeit with relatively high CVs (20-30%), as an indication of good reproducibility suggesting that NIRS-derived forearm blood flow is suitable for intraday testing. However, the current authors take issue with a reliability approach based on testing for systematic differences between measurements since this does not show whether random variability is large or small, let alone quantify the size of this variability. This adds to the importance of our data.

For AO, the generated data for oxygen consumption and delivery, ICCs for oxygen consumption (estimated by HHbslope) were higher for AO than VO. To look at oxygen consumption for both occlusion types was of a practical point considering that if the VO was equally as or more reliable than AO one could prefer using VO in future applications due to it being less inconvenient, especially a concern for patients. As far as we know ICCs for oxygen consumption are not available from other studies. However, Kragelj et al [54] and Van Beekvelt et al [105] (2002) found no significant difference between days for oxygen consumption obtained by AO and thus claimed good day-to-day reproducibility. The argument we give above against using a

41

test for systematic differences between measurements to indicate test-retest reliability applies here as well.

Interestingly, it was shown previously that oxygen consumption measured with AO was more reproducible than VO for intraday testing (Van Beekvelt and co-workers [104]). The authors attributed this to VO being more prone to ever occurring variations in flow due to changes in blood pressure and local vasoreactivity, whereas these influences would be negligible for AO due to the full occlusion of blood flow. Thus this explanation may account for the higher ICC values for AO compared to VO in our data.

Determining oxygen recovery (Rslope) with NIRS after exercise or total occlusion reflects both the influx of oxygenated arterial blood and the continued consumption during the recovery [15]. Approaches for doing this include calculating the time for the half recovery of HbO2 from maximum deoxygenation [15], or by determining the rate of recovery of oxygen saturation post exercise or occlusion (McCully and Natelson [75]; Muthalib et al. [78]). That previous studies show impaired oxygen delivery following exercise and cuff occlusion for chronic fatigue syndrome patients [75], and following exercise for low back pain subjects [50] fuel the appeal of this variable for our research of work related muscle pain. In the present study we calculated the slope of recovery of StO2 (Rslope) following submaximal contraction and arterial occlusion as an expression of reoxygenation.

One can appreciate from the data tables for both muscles that Rslopes for the 50% and 70% MVC were 2-fold larger than at 30% MVC. Chance and co-workers [15] studying rowers at different workloads reported that re-saturation times (i.e. time for the half recovery of HbO2 from maximum deoxygenation) were found to increase with the intensity of work, which concords with our data. We also see that for the ECR, Rslope was two times larger for 30% MVC than for AO. In contrast to our data, McCully and Natelson [75] reported a significantly lower rate of recovery of oxygen saturation after exercise than after cuff ischemia. Differences between studies may be due to different durations of exercise/contractions – 20 sec contractions in our study versus 5 min of exercise in the McCully and Natelson [75] study contributing a substantial metabolic build-up.

As was the case for ΔStO2%, ICC values for Rslope were comparable across contraction levels. In other words, the consistency seen with day-to-day oxygenation change during contraction appears to be reflected for the reoxygenation post contraction. However, whether the reliability between these data sets is physiologically linked is uncertain. The lower ICC for Rslope following AO (compared to the contractions) can be attributed to the large within subject variation (s2

ws). It could be that the reoxygenation post AO is more sensitive to the individual’s daily variations in blood pressure (i.e. perfusion pressure) than for contractions. For most of the NIRS variables reasonable reliabilities were shown. For many variables the method demonstrated to be capable of discriminating

42

between subjects and reasonable absolute reproducibility were demonstrated by the LOA.

Short-term submaximal contractions for WRMP patients and healthy controls (Paper III)

For patients and healthy control subjects, maximal voluntary contraction force (MVC) was similar between the groups, which was somewhat surprising. This is in contrast to previous studies that showed lower MVC for patients with muscle pain for the ECR [9], trapezius muscle [1, 94, 97] and cranio-cervical flexor muscles [83]. This means that relative forces for the contractions were similar in magnitude for the patient and healthy groups in this thesis. StO2% rest value for the TD for patients was lower than controls. This may indicate reduced microcirculation in the TD of patients, and although TD ∆StO2% was similar over contractions for the two groups, patients would have started from a lower saturation level. Thus, the result during contractions would be a lower absolute saturation level for patients.

The lack of significant differences between WRMP patients and controls in ∆StO2% or ∆HbT during acute submaximal isometric contractions for both the ECR and the TD muscles means that patients and controls presented rather similar physiological responses during these provocations – i.e. oxygen consumption and delivery were in equilibrium for both groups. Also, StO2% recovery (Rslope) after contractions or arterial occlusion, and blood flow (HbTslope) for the ECR were not different between groups. Previous studies employing other methodologies such as capillary density from muscle biopsies and laser Doppler flowmetry (LDF) measured other aspects of muscle structure and physiology and found significant differences between WRMP patients and controls in muscle capillarization, fiber typing, oxidative metabolism and blood flow [1, 46, 47, 59-61, 64, 66]. We chose NIRS to specifically monitor oxygen saturation and hemodynamics and did not find any corresponding differences. It could be that (1) respective conditions (anomalies) for detecting differences between patients and controls for other techniques are not synonymous with what can be detected with NIRS, (2) methods are different in their sensitivities, (3) physiological attributes of NIRS-derived variables are not reflective of the pain sensations, and/or (4) differences exist in patients’ conditions between studies. In line with the latter, Kadi et al [47] found more signs of disturbances in blood flow and oxidative metabolism in women with more severe pain due to WRMP than in those with less pain.

Oxygen consumption (HHbslope) for the ECR during AO was higher for controls compared to WRMP patients. This is in line with Brunnekreef et al. [9]. These authors gave no explanation to their findings, and we are also not sure about the implications for our data. It may be important to point out that Van Beekvelt et al [106] found similar differences in consumption between healthy subjects and patients with mitochondrial myopathies in the muscle at rest, and even more pronounced differences during low intensity

43

exercise (10% MVC). We, however, found no group differences in the drop in saturation at this contraction magnitude.

There were no differences between patients and controls in myoelectric activity, which is in contrast to previous studies. Several researchers have shown differences for both RMS and MPF for patients with painful forearm extensor and neck/shoulder muscles compared to healthy controls [23-25, 44, 45, 100]. It appears then that for the patients and controls in the thesis for both muscles similar motor unit recruitment patterns and motor control strategies were employed.

Sustained contraction for WRMP patients and healthy controls (Paper IV)

The patients terminated the 15% sustained earlier than the 12 min maximum allotted time for the ECR and TD, and subsequently were significantly different than controls. This is in line with previous studies showing that patients fatigue faster than controls [5, 33, 48, 59]. As previously mentioned, MVCs were similar between patients and controls, and so the relative force was also similar. Although there was no difference in oxygenation response (StO2%) between groups during the contraction, the lower rest StO2% for patients (significant for the TD) may have contributed to the early fatigue. It has been suggested that oxygen availability is associated with muscle fatigability [36, 49]. The lack of differences between groups in StO2% during the contraction appears in contrast to a study by Sjogaard et al. [97]. These authors reported differences in TD oxygenation patterns between controls and patients during a 40 min peg board work displayed as controls recovering toward baseline but patients did not. The dynamic work task in their study is different than the sustained isometric work in this thesis and also the patient groups likely differ between studies. Although signs of fatigue development were clear with an increase in RMS amplitude and decrease in MPF of the EMG [4, 69], those variables did not show differences between patients and controls; this is in agreement with Hansson et al. [35]. For the ECR, however, a tendency toward a difference in RMS (with controls being higher) was seen and could be explained by the longer duration of the contraction among controls. The lower RMS values for patients may comply with Schulte et al. [94] who compared groups during a 6 min sustained contraction. This may be indicative of a motor control adaptation to pain – the Pain adaptation model [67]. It has been suggested that endurance time differences may be attributed to differences in individual psychosocial factors, such as pain tolerance, motivation and mood [74]. In this thesis subjects were not required to comment on their pain levels during contractions, which may be seen as a limitation.

44

Conclusions

The first part of the thesis (Study 1) dealt with an extensive evaluation of the NIRS technique for measuring in the forearm and shoulder of healthy males and females.

Combining NIRS and electromyography (EMG) it was shown that NIRS-derived oxygenation responses (aerobic demands) to varied contractions correlated with EMG myoelectric responses (an established physiological measurement) for both muscles. For a sustained contraction responses were of a reasonable physiological nature for both muscles. Some responses for oxygenation and myoelectric activity were gender specific.

By implementing maneuvers (e.g. contractions, cuff occlusions) and data processing (e.g. oxygenation recovery) a number of NIRS-derived variables could be generated that are physiologically relevant. Overall, these variables showed reasonable between day reliability for both the forearm and shoulder muscles.

Based on the findings in Study 1 NIRS is deemed a suitable technique for assessing physiological measurements of the upper extremity, including for day-to-day testing.

The second part of the thesis (Study 2) dealt with comparing NIRS variables between patients diagnosed with work-related muscle pain (WRMP) and healthy controls.

A number of NIRS variables were tested for varied short-term isometric contractions. For most variables there were no differences between WRMP and controls for both the forearm and shoulder muscles. The exception was a lower resting oxygenation in the trapezius muscle and lower oxygen consumption in the forearm extensor muscle of patients.

For a 15% MVC contraction sustained for 12 minutes (or until fatigue), WRMP patients had a shorter endurance time than controls, but oxygenation responses during the contraction were similar for both groups.

Based on the findings in Study 2 NIRS was not able to distinguish between the patients with WRMP and controls. A concern in the thesis is the characteristics of the patient group in being equally active in recreational sports, actively working, and similar in muscle strength as controls. Thus, applying NIRS for studying a more severe patient group could yield different results.

45

Acknowledgements

I would like to express my sincere gratitude to all at the Centre for Musculoskeletal Research for your friendship and support during my PhD studies. It was a true privilege to work with you all. I would like to extend my sincere gratitude to:

Albert Crenshaw, my supervisor, mentor and friend, for first of all giving me the opportunity and for guiding me through these PhD years with brilliancy and expertise. Thank you for teaching me logic, scientific thinking, writing and, well, basically everything! Thank you for our innumerous physiological discussions and for years of friendship that made my PhD studies a fun, inspirational and amazing ride.

Mikael Forsman, my co-supervisor and friend, for all the help and support in innumerous occasions when it was needed most. Thank you for your friendship, for all the visits to Stockholm, for teaching me electromyography and for all the articles we wrote together.

Martin Fahlström, my co-supervisor, for the support whenever it was needed always with great expertise, wisdom and enthusiasm. Thank you for all the outstanding work on the papers we wrote together, and for the very important clinical discussions.

Svend Erik Mathiassen, Professor and head of research at the Centre for Musculoskeletal Research and my co-supervisor, for opening the doors of the Centre for me, and for invaluable scientific teachings over the years, when you shared a little of your brilliancy and expertise with me.

My co-authors: Fredrik Hellström, Ulrika Aasa and David Hallman for all the work in the completion of our articles.

My fellow PhD students at CBF: Thomas, Jenny, Jennie, Camilla, Mahmoud, David, Eva, Hasse, Åsa, and the next generation Thomas and Linda for a great collaboration over these years, success to all of you.

Per Liv for helping me in several occasions and actually teaching me statistics over the years in many short but effective lessons.

Margaretha Marklund for all the cheerfulness you brought every day to CBF and of course thanks to all the exceptional graphic work you have done.

Per Gandal for the outstanding and extensive assistance you gave me with software programming and with data collection. Thank you Per for your friendship and for all the conversations every time I walked across the corridor.

46

All my fellow CBF colleagues and friends, researchers, administrative personal you were all pivotal to the completion of my thesis. I would like to thank all of you.

I would like to acknowledge the friendly and competent staff of the Department of Community Medicine and Rehabilitation at Umeå University. It was a privilege doing my PhD in cooperation with you.

To my family for always being there for me, my mom, Norma, my sister, Adriana, step father Saldanha and my girlfriend Sara.

And to my father Dr Vicente J.F. Elçadi who is no longer here with us but I am sure would be very proud of this book.

47

References

1. Andersen, L.L., C. Suetta, J.L. Andersen, M. Kjaer, and G. Sjogaard, Increased proportion of megafibers in chronically painful muscles. Pain, 2008. 139(3): p. 588-593.

2. Andersen, P. and B. Saltin, Maximal perfusion of skeletal muscle in man,

in J. Physiol. (London)1985. p. 233-249.

3. Aptel, M., A. Aublet-Cuvelier, and J.C. Cnockaert, Work-related musculoskeletal disorders of the upper limb, in Joint, Bone, Spine : Revue du Rhumatisme2002. p. 546-555.

4. Basmajian, J.V. and C.J. De Luca, Muscles alive. Their functions revealed by electromyography. Fifth edition, J. Butler, Editor 1985, Williams & Wilkins: Baltimore, London, Sydney. p. 1-561.

5. Beneck, G.J., L.L. Baker, and K. Kulig, Spectral analysis of EMG using

intramuscular electrodes reveals non-linear fatigability characteristics in persons with chronic low back pain. J Electromyogr Kinesiol, 2012. 9(12): p. 001.

6. Bilodeau, M., S. Schindler-Ivens, D.M. Williams, R. Chandran, and S.S. Sharma, EMG frequency content changes with increasing force and during fatigue in the quadriceps femoris muscle of men and women. J Electromyogr Kinesiol 2003. p. 83-92.

7. Blangsted, A.K., P. Vedsted, G. Sjogaard, and K. Sogaard, Intramuscular pressure and tissue oxygenation during low-force static contraction do not underlie muscle fatigue. Acta Physiol Scand, 2005. 183(4): p. 379-88.

8. Boushel, R. and C.A. Piantadosi, Near-infrared spectroscopy for

monitoring muscle oxygenation. Acta Physiol Scand, 2000. 168(4): p. 615-22.

9. Brunnekreef, J.J., J. Oosterhof, D.H.J. Thijssen, W.N.J.M. Colier, and C.J.T. Van Uden, Forearm blood flow and oxygen consumption in patients with bilateral repetitive strain injury measured by near-infrared spectroscopy. Clin Physiol Funct I, 2006. 26(3): p. 178-184.

10. Buckle, P., Ergonomics and musculoskeletal disorders: overview. Occup Med-c, 2005. 55(3): p. 164-167.

48

11. Buckle, P.W. and J. Jason Devereux, The nature of work-related neck and upper limb musculoskeletal disorders. Appl Ergon, 2002. 33(3): p. 207-217.

12. Buckwalter, J.B., P.J. Mueller, and P.S. Clifford, Autonomic control of

skeletal muscle vasodilation during exercise. J Appl Physiol, 1997. 83(6): p. 2037-2042.

13. Celie, B., J. Boone, R. Van Coster, and J. Bourgois, Reliability of near infrared spectroscopy (NIRS) for measuring forearm oxygenation during incremental handgrip exercise. Eur J Appl Physiol, 2012. 112(6): p. 2369-74.

14. Chance, B., M. Cope, E. Gratton, N. Ramanujam, and B. Tromberg, Phase measurement of light absorption and scatter in human tissue. Rev. Sci. Instrum., 1998. 69(10): p. 3457-3481.

15. Chance, B., M.T. Dait, C. Zhang, T. Hamaoka, and F. Hagerman,

Recovery from exercise-induced desaturation in the quadriceps muscles of elite competitive rowers. Am J Physiol, 1992. 262(3 Pt 1): p. 766-75.

16. Clifford, P.S. and Y. Hellsten, Vasodilatory mechanisms in contracting skeletal muscle. J Appl Physiol, 2004. 97(1): p. 393-403.

17. Cui, W., C. Kumar, and B. Chance, Experimental study of migration depth for the photons measured at sample surface. Vol. 1431. 1991: SPIE. 180-191.

18. De Blasi, R., S. Fantini, M. Franceschini, M. Ferrari, and E. Gratton,

Cerebral and muscle oxygen saturation measurement by frequency-domain near-infra-red spectrometer. Med Biol Eng Comput, 1995. 33(2): p. 228-230.

19. De Blasi, R.A., M. Ferrari, A. Natali, G. Conti, A. Mega, and A. Gasparetto, Noninvasive measurement of forearm blood flow and oxygen consumption by near-infrared spectroscopy. J Appl Physiol, 1994. 76(3): p. 1388-93.

20. De Ruiter, C.J., J.F. Goudsmit, J. Van Tricht, and A. De Haan, The Isometric Torque at Which Knee-Extensor Muscle Reoxygenation Stops. Med Sci Sports Exerc, 2007. 39(3): p. 443-453.

21. Delpy, D.T. and M. Cope, Quantification in tissue near–infrared

spectroscopy. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 1997. 352(1354): p. 649-659.

49

22. Dyke, C.K., D.N. Proctor, N.M. Dietz, and M.J. Joyner, Role of nitric oxide in exercise hyperaemia during prolonged rhythmic handgripping in humans. J Physiol, 1995. 488(Pt 1): p. 259-65.

23. Falla, D., G. Bilenkij, and G. Jull, Patients With Chronic Neck Pain

Demonstrate Altered Patterns of Muscle Activation During Performance of a Functional Upper Limb Task. Spine, 2004. 29(13): p. 1436-1440 10.1097/01.BRS.0000128759.02487.BF.

24. Falla, D. and D. Farina, Muscle fiber conduction velocity of the upper trapezius muscle during dynamic contraction of the upper limb in patients with chronic neck pain. Pain, 2005. 116(1-2): p. 138-45.

25. Falla, D., A. Rainoldi, R. Merletti, and G. Jull, Myoelectric manifestations of sternocleidomastoid and anterior scalene muscle fatigue in chronic neck pain patients. Clin Neurophysiol, 2003. 114(3): p. 488-495.

26. Fallentin, N., K. Jörgensen, and E.B. Simonsen, Motor unit recruitment

during prolonged isometric contractions. Eur J Appl Physiol O, 1993. p. 335-341.

27. Felici, F., V. Quaresima, L. Fattorini, P. Sbriccoli, G.C. Filligoi, and M. Ferrari, Biceps brachii myoelectric and oxygenation changes during static and sinusoidal isometric exercises. J Electromyogr Kinesiol, 2009. 19(2): p. e1-11.

28. Feng, W., D. Haishu, T. Fenghua, Z. Jun, X. Qing, and T. Xianwu, Influence of overlying tissue and probe geometry on the sensitivity of a near-infrared tissue oximeter. Physiol Meas, 2001. p. 201-8.

29. Ferrari, M., L. Mottola, and V. Quaresima, Principles, techniques, and

limitations of near infrared spectroscopy. Can J Appl Physiol, 2004. 29(4): p. 463-87.

30. Flodgren, G., A.G. Crenshaw, F. Hellström, and M. Fahlström, Combining microdialysis and near-infrared spectroscopy for studying effects of low-load repetitive work on the intramuscular chemistry in trapezius myalgia. J Biomed Biotechnol, 2010. p. 9.

31. Gerdle, B., K. Henriksson-Larsén, R. Lorentzon, and M.L. Wretling, Dependence of the mean power frequency of the electromyogram on muscle force and fibre type. Acta Physiol Scand, 1991. p. 457-465.

50

32. Gerdle, B., S. Karlsson, A.G. Crenshaw, J. Elert, and J. Friden, The influences of muscle fibre proportions and areas upon EMG during maximal dynamic knee extensions. Eur J Appl Physiol, 2000. p. 2-10.

33. Hagberg, M. and S. Kvarnström, Muscular endurance and

electromyographic fatique in myofascial shoulder pain. Arch Phys Med Rehab, 1984. p. 522-525.

34. Hamaoka, T., T. Katsumura, N. Murase, S. Nishio, T. Osada, T. Sako, H. Higuchi, Y. Kurosawa, T. Shimomitsu, M. Miwa, and B. Chance, Quantification of ischemic muscle deoxygenation by near infrared time-resolved spectroscopy. Vol. 5. 2000: SPIE. 102-105.

35. Hansson, G.A., U. Stromberg, B. Larsson, K. Ohlsson, I. Balogh, and U. Moritz, Electromyographic fatigue in neck/shoulder muscles and endurance in women with repetitive work. Ergonomics, 1992. 35(11): p. 1341-52.

36. Hogan, M.C., R.S. Richardson, and S.S. Kurdak, Initial fall in skeletal

muscle force development during ischemia is related to oxygen availability. J Appl Physiol, 1994. 77(5): p. 2380-4.

37. Homma, S., H. Eda, S. Ogasawara, and A. Kagaya, Near-infrared estimation of O2 supply and consumption in forearm muscles working at varying intensity. J Appl Physiol, 1996. 80(4): p. 1279-84.

38. Hopkins, W.G., Measures of Reliability in Sports Medicine and Science. Sports Med, 2000. 30(1): p. 1-15.

39. Hägg, G. Lack of relation between maximal force capacity and muscle

disorders caused by low level static loads. A new explanation model. in IEA. 1991.

40. Hägg, G.M., Interpretation of EMG spectral alterations and alteration indexes at sustained contraction. J Appl Physiol, 1992. 73(4): p. 1211-7.

41. Jensen, B.R., K. Jorgensen, A.R. Hargens, P.K. Nielsen, and T. Nicolaisen, Physiological response to submaximal isometric contractions of the paravertebral muscles. Spine, 1999. p. 2332-2338.

42. Johansson, H., L. Arendt-Nilsson, M. Bergenheim, S. Blair, J. van Dieen,

M. Djupsjöbacka, N. Fallentin, J.E. Gold, G. Hägg, N. Kalezic, S.E. Larsson, M. Ljubisavljevic, E. Lyskov, T. Mano, M. Magnusson, M. Passatore, F. Pedrosa-Domellöf, L. Punnett, S. Roatta, L.E. Thornell, U. Windhorst, and Z. Zukowska, Epilogue:An integrated model for chronic work-related myalgia

51

”Brussels Model”, H. Johansson, et al., Editors. 2003, Gävle University Press. p. 291-300.

43. Johansson, H., U. Windhorst, M. Djupsjöbacka, and M. Passatore, eds.

Chronic Work-Related Myalgia. Neuromuscular Mechanisms behind Work-Related Chronic Muscle Pain Syndromes. ed. H. Johansson, et al.2003, Gävle University Press: "". 1-310.

44. Johnston, V., G. Jull, R. Darnell, N. Jimmieson, and T. Souvlis, Alterations in cervical muscle activity in functional and stressful tasks in female office workers with neck pain. Eur J Appl Physiol, 2008. 103(3): p. 253-264.

45. Johnston, V., G. Jull, T. Souvlis, and N.L. Jimmieson, Neck Movement and Muscle Activity Characteristics in Female Office Workers With Neck Pain. Spine, 2008. 33(5): p. 555-563.

46. Kadi, F., G. Hägg, R. Håkansson, S. Holmner, G.S. Butler-Browne, and

L.E. Thornell, Structural changes in male trapezius muscle with work-related myalgia. Acta Neuropathol, 1998. 95: p. 362-360.

47. Kadi, F., K. Waling, C. Ahlgren, G. Sundelin, S. Holmner, G.S. Butler-Browne, and L.E. Thornell, Pathological mechanisms implicated in localized female trapezius myalgia. Pain, 1998. p. 191-196.

48. Kallenberg, L.A., E. Schulte, C. Disselhorst-Klug, and H.J. Hermens, Myoelectric manifestations of fatigue at low contraction levels in subjects with and without chronic pain. J Electromyogr Kinesiol, 2007. 17(3): p. 264-74.

49. Katayama, K., M. Amann, D.F. Pegelow, A.J. Jacques, and J.A. Dempsey,

Effect of arterial oxygenation on quadriceps fatigability during isolated muscle exercise. Am J Physiol Regul Integr Comp Physiol, 2007. 292(3): p. R1279-86.

50. Kell, R.T. and Y. Bhambhani, Relationship between erector spinae static endurance and muscle oxygenation-blood volume changes in healthy and low back pain subjects. Eur J Appl Physiol, 2006. 96(3): p. 241-8.

51. Kell, R.T., M. Farag, and Y. Bhambhani, Reliability of erector spinae oxygenation and blood volume responses using near-infrared spectroscopy in healthy males. Eur J Appl Physiol, 2004. 91(5-6): p. 499-507.

52

52. Knardahl, S., Psychophysiological mechanisms of pain in computer work: the blood vessel-nociceptor interaction hypothesis. Work & Stress, 2002. 16(2): p. 179-189.

53. Kocsis, L., P. Herman, and A. Eke, The modified Beer-Lambert law

revisited. Phys Med Biol, 2006. 51(5): p. 15.

54. Kragelj, R., T. Jarm, and D. Miklavcic, Reproducibility of parameters of postocclusive reactive hyperemia measured by near infrared spectroscopy and transcutaneous oximetry. Ann Biomed Eng, 2000. 28(2): p. 168-73.

55. Kuorinka, I. and P. Koskinen, Occupational rheumatic diseases and upper limb strain in manual jobs in a light mechanical industry. Scand J Work Environ Health, 1979. 3: p. 39-47.

56. Kupa, E.J., S.H. Roy, S.C. Kandarian, and C.J. De Luca, Effects of muscle

fiber type and size on EMG median frequency and conduction velocity. J Appl Physiol, 1995. 79(1): p. 23-32.

57. Kutsuzawa, T., S. Shioya, D. Kurita, M. Haida, and H. Yamabayashi, Effects of age on muscle energy metabolism and oxygenation in the forearm muscles. Med Sci Sports Exerc, 2001. 33(6): p. 901-6.

58. Lacroix, S., M. Gayda, V. Gremeaux, M. Juneau, J.-C. Tardif, and A. Nigam, Reproducibility of near-infrared spectroscopy parameters measured during brachial artery occlusion and reactive hyperemia in healthy men. J Biomed Opt, 2012. 17(7): p. 077010-1.

59. Larsson, B., J. Bjork, K.G. Henriksson, B. Gerdle, and R. Lindman, The

prevalences of cytochrome c oxidase negative and superpositive fibres and ragged-red fibres in the trapezius muscle of female cleaners with and without myalgia and of female healthy controls. Pain, 2000. 84(2-3): p. 379-87.

60. Larsson, B., J. Bjork, F. Kadi, R. Lindman, and B. Gerdle, Blood supply and oxidative metabolism in muscle biopsies of female cleaners with and without myalgia. Clin J Pain, 2004. 20(6): p. 440-6.

61. Larsson, R., P.A. Öberg, and S.E. Larsson, Changes of trapezius muscle blood flow and electromyography in chronic neck pain due to trapezius myalgia. Pain, 1999. 79: p. 45-50.

62. Larsson, S.E., L. Bodegård, K.G. Henriksson, and P.Å. Öberg, Chronic

trapezius myalgia. Morphology and blood flow studied in 17 patients. Acta Orthop Scand, 1990. 61: p. 394-398.

53

63. Lin, L., M. Niwayama, T. Shiga, N. Kudo, M. Takahashi, and K. Yamamoto, Influence of a fat layer on muscle oxygenation measurement using near-IR spectroscopy: quantitative analysis based on two-layered phantom experiments and Monte Carlo simulation. Front Med Biol Eng, 2000. 10: p. 43-58.

64. Lindman, R., A. Eriksson, and L.E. Thornell, Fiber type composition of

the human female trapezius muscle: Enzyme-histochemical characteristics. American Journal of Anatomy, 1991. 190: p. 385-392.

65. Lindman, R., A. Eriksson, and L.E. Thornell, Fiber type composition of the human male trapezius muscle: enzyme-histochemical characteristics. Am J Anat, 1990. 189(3): p. 236-44.

66. Ljung, B.O., R.L. Lieber, and J. Friden, Wrist extensor muscle pathology in lateral epicondylitis. J Hand Surg [Br], 1999. 24(2): p. 177-83.

67. Lund, J.P., R. Donga, C.G. Widmer, and C.S. Stohler, The pain-

adaptation model - a discussion of the relationship between chronic musculoskeletal pain and motor activity. Can J Physiol Pharmacol, 1991. p. 683-694.

68. Luopajarvi, T., I. Kuorinka, M. Virolainen, and M. Holmberg, Prevalence of tenosynovitis and other injuries of the upper extremities in repetitive work. Scand J Work Environ Health, 1979. 3: p. 48-55.

69. Madeleine, P., D. Farina, R. Merletti, and L. Arendt-Nielsen, Upper trapezius muscle mechanomyographic and electromyographic activity in humans during low force fatiguing and non-fatiguing contractions. Eur J Appl Physiol O, 2002. 87: p. 327-336.

70. Madeleine, P., L.V. Jorgensen, K. Sogaard, L. Arendt-Nielsen, and G.

Sjogaard, Development of muscle fatigue as assessed by electromyography and mechanomyography during continuous and intermittent low-force contractions: effects of the feedback mode. Eur J Appl Physiol, 2002. 87(1): p. 28-37.

71. Maikala, R. and Y. Bhambhani, Microvascularity of the lumbar erector spinae muscle during sustained prone trunk extension test, in Oxygen transport to tissue XXX, P. Liss, et al., Editors. 2009, Springer: New York. p. 67-73.

72. Mancini, D.M., L. Bolinger, H. Li, K. Kendrick, B. Chance, and J.R. Wilson, Validation of near-infrared spectroscopy in humans. J Appl Physiol, 1994. 77: p. 2740-2747.

54

73. Matcher, S.J., P. Kirkpatrick, K. Nahid, M. Cope, and D.T. Delpy, Absolute quantification methods in tissue near infrared spectroscopy, in Optical Tomography, Photon Migration, and Spectroscopy of Tissue and Model Media: Theory, Human Studies, and Instrumentation, Proceedings of, Pts 1 and 2, B. Chance, R.R. Alfano, and A. Katzir, Editors. 1995, Spie - Int Soc Optical Engineering: Bellingham. p. 486-495.

74. Mathiassen, S.E. and E. Ahsberg, Prediction of shoulder flexion

endurance from personal factors. Int J Ind Ergonom, 1999. 24: p. 315-329.

75. McCully, K.K. and B.H. Natelson, Impaired oxygen delivery to muscle in chronic fatigue syndrome. Clin. Sci., 1999. 97(5): p. 603-608.

76. Merletti, R., M. Knaflitz, and C.J. De Luca, Myoelectric manifestations of fatigue in voluntary and electrically elicited contractions. J Appl Physiol, 1990. 69(5): p. 1810-20.

77. Miura, H., H. Araki, H. Matoba, and K. Kitagawa, Relationship among

oxygenation, myoelectric activity, and lactic acid accumulation in vastus lateralis muscle during exercise with constant work rate. Int J Sports Med, 2000. p. 180-184.

78. Muthalib, M., M. Jubeau, G.Y. Millet, N.A. Maffiuletti, M. Ferrari, and K. Nosaka, Biceps brachii muscle oxygenation in electrical muscle stimulation. Clin Physiol Funct Imaging, 2010. 30(5): p. 360-8.

79. Muthalib, M., G.Y. Millet, V. Quaresima, and K. Nosaka, Reliability of near-infrared spectroscopy for measuring biceps brachii oxygenation during sustained and repeated isometric contractions. J Biomed Opt, 2010. 15(1): p. 017008.

80. Myers, D.E., L.D. Anderson, R.P. Seifert, J.P. Ortner, C.E. Cooper, G.J.

Beilman, and J.D. Mowlem, Noninvasive method for measuring local hemoglobin oxygen saturation in tissue using wide gap second derivative near-infrared spectroscopy. J Biomed Opt, 2005. 10.

81. Nordander, C., K. Ohlsson, I. Akesson, I. Arvidsson, I. Balogh, G.-A. Hansson, U. Stromberg, R. Rittner, and S. Skerfving, Risk of musculoskeletal disorders among females and males in repetitive/constrained work. Ergonomics, 2009. 52(10): p. 1226 - 1239.

82. Nordander, C., K. Ohlsson, I. Balogh, G.A. Hansson, A. Axmon, R. Persson, and S. Skerfving, Gender differences in workers with identical repetitive industrial tasks: exposure and musculoskeletal disorders. Int Arch Occup Environ Health, 2008. 81(8): p. 939-47.

55

83. O'Leary, S., G. Jull, M. Kim, and B. Vicenzino, Cranio-cervical flexor muscle impairment at maximal, moderate, and low loads is a feature of neck pain. Man Ther, 2007. 12(1): p. 34-9.

84. Orizio, C., R. Perini, and A. Veicsteinas, Changes of muscular sound

during sustained isometric contraction up to exhaustion. J Appl Physiol, 1989. 66(4): p. 1593-1598.

85. Passatore, M. and S. Roatta, Influence of sympathetic nervous system on sensorimotor function: whiplash associated disorders (WAD) as a model. Eur J Appl Physiol, 2006. 98(5): p. 423-49.

86. Passatore, M. and S. Roatta, Modulation operated by the sympathetic nervous system on jaw reflexes and masticatory movement. Arch Oral Biol, 2007. 52(4): p. 343-6.

87. Praagman, M., H.E.J. Veeger, E.K.J. Chadwick, W.N.J.M. Colier, and

F.C.T. van der Helm, Muscle oxygen consumption,determined by NIRS,in relation to external force and EMG. J Biomech,2003. p. 905-912.

88. Punnett, L. and D.H. Wegman, Work-related musculoskeletal disorders: the epidemiologic evidence and the debate. J Electromyogr Kinesiol, 2004. 14(1): p. 13-23.

89. Radegran, G., Human skeletal muscle hyperaemia: its magnitude and regulation during exercise. Dan Med Bull, 2003. 50(1): p. 39-63.

90. Ranney, D., R. Wells, and A. Moore, Upper limb musculoskeletal

disorders in highly repetitive industries: precise anatomical physical findings. Ergonomics, 1995. 38(7): p. 1408-23.

91. Rodrigues, A.D.E.C., M.D.P. Silva, and V.D. Pai, Distribution of Different Fibre Types of M. extensor carpi radialis longus of the Rat. Anatomia, Histologia, Embryologia, 1994. 23(4): p. 352-356.

92. Sako, T., T. Hamaoka, H. Higuchi, Y. Kurosawa, and T. Katsumura, Validity of NIR spectroscopy for quantitatively measuring muscle oxidative metabolic rate in exercise. J Appl Physiol,2001. p. 338-344.

93. Saltin, B., G. Radegran, M.D. Koskolou, and R.C. Roach, Skeletal muscle

blood flow in humans and its regulation during exercise. Acta Physiol Scand, 1998. 162(3): p. 421-36.

94. Schulte, E., L.A. Kallenberg, H. Christensen, C. Disselhorst-Klug, H.J. Hermens, G. Rau, and K. Sogaard, Comparison of the electromyographic

56

activity in the upper trapezius and biceps brachii muscle in subjects with muscular disorders: a pilot study. Eur J Appl Physiol, 2006. 96(2): p. 185-93.

95. Searle, S.R., G. Casella, and C.E. McCulloch, Variance components 2006,

Hoboken Wiley, cop. 1-501.

96. Simonson, S.G. and C.A. Piantadosi, Near-infrared spectroscopy. Clinical applications. Crit Care Clin, 1996. 12(4): p. 1019-29.

97. Sjogaard, G., L. Rosendal, J. Kristiansen, A. Blangsted, J. Skotte, B. Larsson, B. Gerdle, B. Saltin, and K. Sogaard, Muscle oxygenation and glycolysis in females with trapezius myalgia during stress and repetitive work using microdialysis and NIRS. Eur J Appl Physiol, 2010. 108(4): p. 657-669.

98. Sluiter, J.K., K.M. Rest, and M.H. Frings-Dresen, Criteria document for

evaluating the work-relatedness of upper-extremity musculoskeletal disorders. Scand J Work Environ Health, 2001. 27 Suppl 1: p. 1-102.

99. Stratford, P., Reliability: consistency or differentiating among subjects? Physical Therapy, 1989. 69(4): p. 299-300.

100. Szeto, G.P.Y., L.M. Straker, and P.B. O’Sullivan, A comparison of symptomatic and asymptomatic office workers performing monotonous keyboard work—1: Neck and shoulder muscle recruitment patterns. Man Ther, 2005. 10(4): p. 270-280.

101. Takahashi, E., D.F. Bruley, H. Wabnitz, M. Moeller, A. Liebert, H. Obrig,

J. Steinbrink, and R. Macdonald, Time-Resolved Near-Infrared Spectroscopy and Imaging of the Adult Human Brain, in Oxygen Transport to Tissue XXXI, Springer US. p. 143-148.

102. Troiano, A., F. Naddeo, E. Sosso, G. Camarota, R. Merletti, and L. Mesin, Assessment of force and fatigue in isometric contractions of the upper trapezius muscle by surface EMG signal and perceived exertion scale. Gait Posture, 2008. 28(2): p. 179-86.

103. Van Beekvelt, M.C., M.S. Borghuis, B.G. van Engelen, R.A. Wevers, and W.N. Colier, Adipose tissue thickness affects in vivo quantitative near-IR spectroscopy in human skeletal muscle. Clin. Sci., 2001. 101(1): p. 21-28.

104. Van Beekvelt, M.C., W.N. Colier, R.A. Wevers, and B.G. Van Engelen,

Performance of near-infrared spectroscopy in measuring local O(2)

57

consumption and blood flow in skeletal muscle. J Appl Physiol, 2001. 90: p. 511-519.

105. Van Beekvelt, M.C., B.G. Van Engelen, R.A. Wevers, and W.N. Colier, In

vivo quantitative near-infrared spectroscopy in skeletal muscle during incremental isometric handgrip exercise. Clin Physiol Func I, 2002. 22(3): p. 210-7.

106. Van Beekvelt, M.C.P., B.G.M. Van Engelen, R.A. Wevers, and W.N.J.M. Colier, Quantitative near-infrared spectroscopy discriminates between mitochondrial myopathies and normal muscle. Annals of Neurology, 1999. 46(4): p. 667-670.

107. Viikari-Juntura, E., Neck and upper limb disorders among slaughterhouse workers. An epidemiologic and clinical study. Scand J Work Environ Health, 1983. 9(3): p. 283-90.

108. Visser, B. and J.H. van Dieen, Pathophysiology of upper extremity

muscle disorders. J Electromyogr Kinesiol, 2006. 16(1): p. 1-16.

109. Wolf, U., M. Wolf, J.H. Choi, L.A. Paunescu, L.P. Safonova, A. Michalos, and E. Gratton, Mapping of hemodynamics on the human calf with near infrared spectroscopy and the influence of the adipose tissue thickness, in Oxygen Transport to Tissue Volume Xxiii: Oxygen Measurements in the 21st Century: Basic Techniques and Clinical Relevance, D.F. Wilson, et al., Editors. 2003, Kluwer Academic/Plenum Publ: New York. p. 225-230.

110. Vollestad, N. and R. C., Metabolic and mechanical changes during low-intensity work and their relation to work-related pain, in Chronic work-related myalgia. Neuromuscular mechanisms behind work-related chronic muscle pain syndromes, H. Johansson, et al., Editors. 2003, Gävle University Press: Gävle. p. 117-126.

111. Yassi, A., Repetitive strain injuries. Lancet, 1997. 349(9056): p. 943-

947.