In June 2001 the title ‘chiropractor’ came to be protected under British law and those who called themselves chiropractors attained a position of increased legitimacy within British society. Yet the details of chiropractic’s journey to statutory recognition have not been thoroughly explored in contemporary literature. The origins and development of chiropractic in Britain have received meagre attention from historical scholars. This thesis uses a neo-Weberian approach to explore the history of chiropractic in Britain through the lens of ‘professionalisation’. It investigates the emergence of chiropractic in Britain, and details how and why chiropractic developed in the way that it did, assessing the significance of processes and events in respect to chiropractic’s professionalisation, and examining intra- and inter-occupational tensions. The thesis is primarily a product of documentary research, but is also informed by interviews undertaken to provide oral testimonies. Although the origins of chiropractic are usually traced back to the 1890s, to Davenport, Iowa, and to the practice of Daniel David Palmer, it is argued in this thesis that it is misleading to claim that chiropractic was ‘discovered’ by Daniel Palmer, or that chiropractic in Britain was entirely an ‘import’ from the United States. Instead, chiropractic’s origins were complex and multifarious and form part of a broader history of manipulative practices. With regard to the development of chiropractic in Britain, chiropractic’s history is intertwined with that of osteopathy, and has involved medicalisation. This study demonstrates that through the course of its evolution chiropractic was subject to processes that can usefully be described in terms of professionalisation, sharing features in common with the professionalisation of other occupational groups described in historical and sociological literature. Even so, chiropractors did not attain the social presence or cultural authority of archetypal professionals such as medical doctors or lawyers. Although protection of title was achieved, many problems have remained, including divisions within the occupation.

In the early days of PKU, micronutrient deficiencies, undernutrition and growth failure were common features of patients with the condition. This was mainly due to the limited availability of engineered special low protein foods (SLPF) and PKU protein substitutes to use. Nowadays, SLPF foods and micronutrient-enriched PKU protein substitutes have become widely available and are free on prescription in most countries. These SLPF are high in carbohydrates and often in fat content, have a higher glycaemic index and provide more energy per weight compared with the protein-containing equivalent normal foods. Advancement in the PKU management, including dietary practices, led to nutritional problems that had never been reported before, but have become more frequent in the recent years. Overweight and obesity, rather than undernutrition, have become increasingly reported in patients with PKU, with some studies suggesting higher prevalence in females than males. Data on body composition in patients with PKU are inconsistent with some studies showing that patients with PKU have higher FM and lower FFM compared to healthy controls. This suggests that for a given body weight and height, patients with PKU might be fatter and look bigger. However, there is very little research looking at the determinants of nutrient status, body composition and obesity in patients with PKU. Hence, among the aims of this thesis was to investigate the impact of a PKU SLPF-based meal on appetite ratings, gut appetite hormones, thermic effect of feeding (TEF) and fat oxidation (Chapter 2). Twenty-three healthy adults (mean ± SD age: 24.3 ± 5.1 years; BMI: 22.4 ± 2.5 kg/m2) participated in a randomised crossover study. Each participant conducted two (PKU and Control) experimental trials which involved consumption of a PKU SLPF-based meal and protein substitute drink or an isocaloric and weight matched ordinary meal and protein-enriched milk drink. Appetite, metabolic rate, fat oxidation measurements and blood collections were conducted for the duration of 300 minutes. On completion of the measurements, an ad libitum buffet dinner was served. Responses of appetite ratings, plasma concentrations of GLP-1 and PYY (P > 0.05, trial effect, two-way ANOVA) and energy intake during ad libitum buffet dinner (P > 0.05, paired t-test) were not significantly different between the two trials. The TEF (PKU, 10.2 ±1.5%; Control, 13.2 ± 1.0%) and the total amount of fat oxidised (PKU, 18.90 ± 1.10 g; Control, 22.10 ± 1.10 g) were significantly (P < 0.05, paired t-tests) lower in the PKU than in the Control trial. The differences in TEF and fat oxidation were significant (P < 0.05, paired t-tests) for the post-meal period. Therefore, from this first study we concluded that consumption of a meal composed of SLPF has no detrimental impact on appetite and appetite hormones but produces a lower TEF and postprandial fat oxidation than an ordinary meal. We hypothesised that these metabolic alterations may contribute to the increased prevalence of obesity reported in patients with PKU on contemporary dietary management. In the second experimental chapter, we tested the hypothesis generated from the study above and measured TEF, fasting and postprandial fat and CHO oxidation in 13 patients with PKU and 10 healthy controls of similar age and BMI. Participants in the PKU group were provided an SLPF-based meal while those in the Control group were provided an isocaloric normal meal. It was found that TEF, and postprandial fat and CHO oxidation were not signifcantly different between the PKU and the Control groups. In addition, this study compared body composition characteristics (measured by Deuterium Oxide dilution technique) between PKU patients and healthy controls and revealed that differences in body composition are not significant between the two groups despite a tendency of PKU patients having higher percentage of body fatness (P=0.08). However, data generated from this study should be interpreted with caution and requires confirmation from studies with larger sample size. Micronutrient imbalance has been noted in patients with PKU despite their high provision through the PKU protein substitutes. Recent studies showed high blood levels of vitamin B12 and folate, but simultaneously deficient plasma levels of selenium and zinc in PKU patients prescribed with micronutrient-enriched PKU protein substitutes. Factors associated with micronutrient imbalance have rarely been studied in the literature. Therefore, the last chapter of this thesis aimed to evaluate the micronutrient status of children with PKU and explore factors associated with micronutrient imbalances and deficiencies. This was analysis of a large clinical dataset with serial measurements obtained from PKU children (≤16 years) attending the metabolic medicine clinic at the Glasgow Royal Hospital for Sick Children between 1990 and 2013. The study included 81 patients who provided a total of 512 blood samples for their routine annual micronutrient screening. Data on blood micronutrient measurements was available for vitamins A, B12, D, E, serum folate, and erythrocyte folate and the trace elements copper, selenium, zinc and serum ferritin as a biomarker of iron stores. Status of vitamin B12, E, and serum and erythrocyte folate measurements were above the normal range (NR) in 27%, 54%, 46% and 35% of the blood samples, respectively. However, 44% of selenium and 14% of zinc measurements were below the NR. Moreover, when we compared results with those from the UK National Dietary and Nutritional Survey, selenium and zinc deficiencies were specific to PKU condition and not a reflection of the epidemiology in the general UK population. In our PKU sample, poor metabolic control, PKU severity, and low adherence to PKU protein substitutes predicted low selenium status; while deficient zinc status was solely predicted by low adherence to PKU protein substitutes. Yet, these predictors, collectively, explained a small (5.8 – 8.8 %) variation in the status of selenium and zinc in this group of patients. Selenium and zinc deficiencies are common in PKU patients despite high levels of other nutrients including vitamin B12, E and folate. The findings of this study suggest that selenium and zinc deficiencies reported in patients with PKU may be attributed to other factors which we were unable to measure in this retrospective study, such as low bioavailability of these nutrients from the artificial PKU protein substitutes.

Background & Aim: Osteoarthritis (OA) sufferers frequently turn to complementary and alternative medicine (CAM), of which the most common are acupuncture and homeopathy, to improve manageability of their condition. However, there is little extant evidence of effectiveness for these treatments for OA, particularly for homeopathy. One criticism of homeopathic studies is that treatment protocols do not reflect true homeopathy. The nature of true homeopathy is not documented in extant literature. In the current study a mixed methods approach was used to investigate the use of homeopathy for osteoarthritis using a survey, conducted with a parallel acupuncture survey for comparison, follow up interviews with homeopaths and a patient-centred study in a homeopathic department offering treatment on the NHS, in order to inform future studies. Method: The current study involved three phases; (1) A descriptive survey conducted on n=362 medical and non-medical homeopaths and acupuncturists, was used to , investigate practice of the therapies (2) Follow up interview of n=28 of the homeopathic practitioners. (3) A patient-centred study of n=11 patients with OA receiving homeopathy in the primary care setting. Results: (1) Most commonly encountered conditions were chronic diseases. Medical and non-medical acupuncturists practised very different forms of acupuncture particularly in terms of diagnostic techniques used and theoretical underpinning. Homeopathic practitioners used individualised treatments, abiding by classical homeopathy. Differences between medical and non-medical homeopaths included time spent in the consultation (p= 0.01), strength of confidence in homeopathy for asthma (p=0.01), musculo-skeletal (p=0.046) and acute conditions (p=0.01)., and confidence in conventional medicine (p=0.01). There was a belief amongst acupuncturists and homeopaths that the treatments may work on electrodynamic fields in the body. (2) A similar approach was taken by participants during a detailed initial consultation. However, irrespective of medical status, varied approaches were used to identify the remedy, potency, and remedy f6mi, and the source of remedy also varied. Main themes regarding the modus operandi of homeopathy included stimulation of self healing mechanisms and identifying in detail events at the point where the initial health imbalance occurred. Identification of this point together with the patient was considered a potential trigger for the healing process to begin, adding a particular importance to the role of the consultation. (3) OA patients in the primary care setting identified pain or stiffness as the most common primary complaint., with an emotional factor such as anxiety and limitations caused by their condition as a secondary complaint. A desire to reduce their medication or to improve the manageability of their condition was a common theme for interest in receiving homeopathy, with access to NHS homeopathic treatment and perceived safety of receiving treatment from medical doctors being important factors. Following 6 months of' homeopathic treatment, most participants reported an improvement in the manageability of their condition. This, however was not supported by results from VAS pain, VAS stiffness, MYCAW scores or SF36 subscore, or salivary concentration of substance P results which were not found to be significant. Few correlations were found between outcome measures. Substance P level was strongly correlated with the functional limitations sub-score of the SF36 (p=0.01), indicating a potential role for this biochemical measure in future studies. Conclusion: Findings from the current study can inform future studies on how to enhance the evidence base for homeopathic and acupuncture treatment, and inform the integration debate. Future advances in the understanding of subtle processes in the body, the placebo response, and the nature of cure may add to our understanding of' CAM treatments. However, it is likely that in order to advance the evidence base on the effectiveness of homeopathy for OA, more effective tools that are sensitive to changes in biopsychosocial dimensions of health will be necessary. Future research on combination therapies is also warranted.

The aim of the present thesis was to examine imitation of biological motion in adults with autism spectrum disorders. Using a novel behavioural protocol, adults with autism and matched neurotypical control adults imitated models that displayed distinctly different, but biological believable kinematics. In Chapter Two it was observed that adults with autism exhibited low-fidelity imitation of atypical biological motion. In Chapter Three it was observed that when selective-attention instructions were provided, although eye movements recorded during action- observation was similar to controls, imitation of atypical biological motion was still impaired. In Chapter Four across three experiments it was shown that adults with autism exhibit reasonably high-fidelity imitation of atypical biological motion. This was achieved by presenting the to-be-imitated biological models in a fixed presentation structure which is known to facilitate greater integration and consolidation of sensorimotor information. This suggestion was supported by a further study where firstly participants were required to complete a secondary motor task during the inter-trial delay, and when the presentation structure was randomised (similar to Chapters Two and Three) resulting in low-fidelity imitation of atypical biological motion. These findings across the present thesis will be discussed in light of a critical evaluation with respect to current literature on imitation in autism, as well as implications for theoretical accounts of impaired imitation in autism and related sensorimotor control processes. Future considerations and translational research will be discussed, with the intention of offering prospective social rehabilitation protocols in autism.

The goal of this thesis is to identify the underlying impairments in aphasic disorders of auditory comprehension and repetition. The findings are interpreted within a cognitive neuropsychological framework. Models of normal language processing are discussed in the light of this evidence. Information processing models of the lexicon attempt to specify the stages of processing necessary for auditory comprehension, as well as different routes by which words can be repeated. Twenty fluent aphasic patients were used in the study. It was found that the patients did show qualitative differences in auditory word comprehension. Five levels of impairment were identified: word-sound deafness, word-for deafness, word-meaning deafness, a central semantic disorder and a disorder specific to abstract words. It was concluded that abstract words are more sensitive to impairment than concrete words. Word imageability was investigated in more detail in a number of experiments with a word meaning deaf patient (DRB). It was shown that his impairment is one of access from the input lexicon to the semantic system. The impairment results in under-specification in the semantic system, and an extremely robust effect of imageability in DRB's ability to comprehend and repeat auditorily presented words. This effect is not item-specific. Intriguingly, the results also suggest that DRB has an anomia for words of low imageability. In a subsequent section, the patients' abilities in repetition are investigated. Two routes for repetition are identified, a sub-lexical and a lexical/semantic route. Phonologically related errors arising in the former route tend to be non-words, occur particularly on longer words, and the errors tend to be in the final position of the string. Phonological errors arising in the lexical/semantic route are real words, tend to be higher in frequency than the stimulus items, and occur particularly on shorter words. The relationship between repetition and auditory short term memory is considered by further experiments with DRB. It is argued that sub-lexical repetition utilises the auditory short term memory system. DRB's sub-lexical repetition and his immediate serial recall are enhanced by lip read information. A model of repetition and auditory short-term memory is presented. It is argued that the system requires different input and output phonological codes, suggesting separate input and output lexicons. With the specification of how lexical information supports immediate serial recall, it is argued that there is no requirement for a direct, lexical, non-semantic route in repetition.

The ligaments of the trapeziometacarpal joint (TMC) are complex and highly varied. Several studies have reported different patterns of ligament arrangement. Which ligaments are responsible for governing the stability of the TMC is still a source of controversy; the very naming of some of the ligaments is also in dispute. The overall aim of the experiments in this thesis is to explore the stabilization of the TMC ligaments during a specific positions (the neutral and full abduction positions), the mechanics of the attachment, the relaxing and stretching of the TMC ligaments, and the responsibilities of the TMC ligaments to prevent the TMC joint subluxations such as dorsal subluxation, palmar subluxation, and lateral subluxation. An additional aim is to describe the orientation of the TMC ligaments (origin, insertion) and the ligament fibres’ directions. The study used development devices to render the TMC ligaments into 2D reconstruction and 3D modelling in the 3D virtual environment such as Rhinoceros V5, 3D Landmark software, and 3D Amira software. Length, width, area, volume, thickness, and cross-sectional measurements were assessed in the neutral and full abduction positions. The ligament stretcher was designed to stretch the TMC ligaments after the ligaments were cut away from the joint. Also used in the thesis is a new technique to develop staining procedures by combining two different stains with accurate timing and re-timing the procedures of the Miller’s Elastin Staining protocol. Also, the entheses investigations have been applied on the attachments of the TMC ligaments with trapezuim and first metacarpal bones (proximal and distal). Moreover, the technique of the high resolution episcopic microscopy (HREM) was used to identify the relation of the TMC ligaments fibres to the flexor retinaculum, especially at dorso-ulnar trapeziometacarpal ligament (DUTML). The layers of the palmar trapeziometacarpal ligament (PTML) were also investigated to identify the nature of this ligament. The osteological descriptions were noted for the first metacarpal and trapezium bones, the geometrical measurement experiments on the articular surface of both bones allowed for a description of the effects of the degenerative disease, especially osteioarthritis disease (OA), and comparisons between both genders, the right and left specimens, embalmed and fresh cadavers were investigated to achieve the most accurate and precise results. The index procedures against the third metacarpal assisted in describing the relation between the actual reading and index reading. Seven ligaments were shown in the 3D virtual environment: radial trapeziometacarpal ligament (RTML), superficial palmar trapeziometacarpal ligament (sPTML), deep palmar trapeziometacarpal ligament (dPTML), palmo-ulnar trapeziometacarpal ligament (PUTML), dorso-ulnar trapeziometacarpal ligament (DUTML), palmar intermetacarpal ligament (PIML), and dorsal intermetacarpal ligament (DIML). Also, the results revealed that the PIML and DIML were the main stabilizers through the experiments, and the DUTML and PUTML served as associated ligaments to prevent TMC joint dislocation. The RTML prevented radial subluxation, while the superficial and deep layers of the PTML acted as a pivot for the movement of the TMC joint and assisted in preventing a palmar subluxation. In addition, there were no significant differences (p > 0.05) between embalmed and fresh cadavers, but there were significant differences (p < 0.05) between male and female cadavers as well as between the right and left hands, especially in the measurement investigations. Moreover, the TMC ligament attachments were found to be of the fibrocartilagenous type; this was found at both attachments, proximal and distal. The combination beginning with Miller Elastin stain (ME) was better than that starting with Modified Masson Trichrome stain (MME) and the colours were nearer to those of the MMT results; however, the MMT alone was clearest. Also, a re-timing of the ME stain revealed that the experiment involving 2.5 hours of Miller’s Elastin and 15 seconds of Van Gieson Solution was the best of all experiments. The HREM technique revealed no connection between the collagen fibres of both the DUTML and the flexor retinaculum. The single coordination points of each of the first metacarpal (1st MC) and trapezium (TM) bones revealed varieties of prominences and declines in the bones’ surfaces, especially the articular surface. New names of the TMC ligaments as following: Old Name New Name Ulnar collateral ligament (UCL). Radial trapeziometacarpal ligament (RTML). Superficial anterior oblique ligament (sAOL). Superficial palmar trapeziometacarpal ligament (sPTML). Deep anterior oblique ligament (dAOL). Deep palmar trapeziometacarpal ligament (dPTML). Posterior oblique ligament (POL). Palmo-ulnar trapeziometacarpal ligament (PUTML). Dorso-radial ligament (DRL). Dorso-ulnar trapeziometacarpal ligament (DUTML). New osteological definitions were noted, such as the distal border of the 1st MC facet, the distal and proximal ridge of the palmar surface of the1st MC, and eminences of the trapezial ridge of TM. Overall, the results of my studies suggested that the importance of the PIML, DIML, and RTML should be considered when planning surgeries involving ligament reconstruction of the TMC joint.

Short chain fatty acids (SCFA), including propionate, are produced by the bacterial fermentation of carbohydrates in the colon. Propionate has many potential roles in health, including inhibiting cholesterol synthesis, de novo lipogenesis and increasing satiety. The profile of SCFA produced is determined by both the substrate available and the bacteria present and may be influenced by environmental conditions within the lumen of the colon. Whilst it may be beneficial to increase colonic propionate production, dietary strategies to achieve this are unproven. Adding propionate to food leads to poorer organoleptic properties, and oral propionate is absorbed in the small intestine. The optimum way to selectively increase colonic propionate would be to select fermentable carbohydrates that selectively promote propionate production. To date, few studies have undertaken a systematic assessment of the factors leading to increased colonic propionate production making the selection of propiogenic carbohydrates challenging. The aim of this thesis was to identify the best carbohydrates for selectively increasing propionate production, and to explore the factors which control propionate production. This work started with a systematic review of the literature for evidence of candidate carbohydrates, which led to a screen of ‘propiogenic’ substrates using in vitro batch fermentations and mechanistic analysis of the impact of pH, bond linkage and orientation using a range of sugars, polysaccharides and fibre sources. A new unit for SCFA production was developed to allow comparison of results from in vitro studies encompassing a range different methodologies found in the literature. The systematic review found that rhamnose yielded the highest rate and proportion of propionate production whereas, for polysaccharides, β-glucan ranked highest for rate and guar gum ranked highest for molar production, but this was not replicated across all studies. Thus, no single NDC was established as highly propiogenic. Some substrates appeared more propiogenic than others and when these were screened in vitro. Laminarin, and other β-glucans ranked highest for propionate production. Legume fibre and mycoprotein fibre were also propiogenic. A full complement of glucose disaccharides were tested to examine the role glycosidic bond orientation and position on propionate production. Of the glucose disaccharides tested, β(1-4) bonding was associated with increased proportion of propionate and α(1-1) and β(1-4) increased the rate and proportion of butyrate production. In conclusion, it appears that for fibre to affect satiety, high intakes of fibre are needed, and which a major mechanism is thought to occur via propionate. Within this thesis it was identified that rather than selecting specific fibres, increasing overall intakes of highly fermentable carbohydrates is as effective at increasing propionate production. Selecting carbohydrates with beta-bonding, particularly laminarin and other β(1-4) fermentable carbohydrates leads to marginal increases in propionate production. Compared with targeted delivery of propionate to the colon, fermentable carbohydrates examined in this thesis have lesser and variable effects on propionate production. A more complete understanding of the impact of bond configurations in polysaccharides, rather than disaccharides, may help selection or design of dietary carbohydrates which selectively promote colonic propionate production substrates for inclusion in functional foods. Overall this study has concluded that few substrates are selectively propiogenic and the evidence suggests that similar changes in propionate production may be achieved by modest changes in dietary fibre intake.

Approximately, 2/3 of patients with a spinal cord injury (SCI) suffer from chronic pain, leading to a reduction in quality of life. The prevalence of chronic central neuropathic pain (CNP) in the SCI population is 40%. Recent neuroimaging studies provided evidence that CNP is accompanied by modified brain activity at surface and deep cortical levels and that CNP is resistant to different pharmacological and non-pharmacological treatments. Our current knowledge on how CNP affects the brain activity of SCI patients is mainly based on fMRI studies. Although these studies provide precise spatial localisation of brain regions most affected by CNP, they indirectly measure brain activity through measuring blood oxygenation. Therefore they lack information specific to neuronal activity such as dynamic, time and frequency dependant oscillatory activity of cortical structures. Therefore, in Phase 1 of this study, electroencephalogram (EEG) activity of paraplegic patients with CNP (PWP) is compared with the EEG activity of able-bodied (AB) participants and paraplegic patients without CNP (PNP). It was found that CNP leads to frequency dependant EEG signatures both in the relaxed state and during motor tasks that are not restricted to the cortical representation of the body part perceived as being painful. The pharmacological treatment of CNP has a number of side effects and does not provide significant pain relief. The effect of non-pharmacological treatments is inconsistent. Neurofedback (NF) is a non-pharmacological treatment, based on the voluntarily modulation of brain activity to control pain intensity. Using NF training the patient can learn and apply a mental strategy to control pain, without the need for an external device. However, NF requires a large number of training sessions to learn the necessary mental strategy. Therefore, in Phase 2 of this study, the effect on pain intensity of a large number of NF sessions, using different NF training protocols, was assessed. The clinically and statistically significant reduction of pain observed in this study demonstrates that NF training has the potential to manage chronic CNP in paraplegic patients. This study also provides evidence that the reduction of pain achieved using NF training may not be due to a placebo effect. Furthermore, the study demonstrates the immediate global effect of NF training on power and coherence. To date, no neuroimaging studies that have applied NF training with patients with CNP have shown changes in brain activity before the first and after the last training session. Therefore, in phase 3 of this study, the long-term neurological effect of NF training was assessed using EEG. This study provides evidence that NF training does not only induce an effect on spontaneous EEG activity, but also induces changes on evoked EEG activity. In conclusion, this study compared the EEG activity of three groups (AB, PWP, and PNP) and found that CNP (PWP group) leads to frequency dependant dynamic oscillatory signatures. The study also reported that NF training has a potential to reduce pain and this reduction of pain might not be an effect of placebo. Furthermore, it was found that NF training induce long-term changes in the EEG activity recorded in relaxed state and during motor tasks. This long-term change in EEG activity was noticed at the surface and deep cortical structures.

Introduction: Low back pain remains a burden for society, since it can lead to sickness absence and work disability. Physical occupational risk factors can contribute to the development of back pain, yet little is known about any risks in sedentary jobs posed by sitting. The influence of psychosocial factors on back pain and sickness absence amongst sedentary workers is also unclear. The aim of this study was to measure work activities, lumbar movement characteristics, symptoms and psychosocial factors in order to determine associations with low back pain and sickness absence. Methods: Phase 1: involved validation of a fibre-optic goniometer system that attaches to the lumbar spine and hip to continuously measure: (1) activities (sitting, standing, walking); and (2) lumbar movement characteristics (notably sitting postures and kinematics). New questionnaires were also validated to measure aspects of low back discomfort. Phase 2: consisted of a cross-sectional survey of call centre workers (n=600) to collect data on: demographics, clinical and occupational psychosocial factors, and symptoms. An experimental sample (n=140) wore the goniometer system during work. Phase 3: involved a 6-month follow-up survey to collect back pain and sickness absence data (n=367). Logistic regression was used to determine associations (P<0.05) between data. Results: Workers spent 83% of work-time sitting, 26% of which was spent adopting a lordotic lumbar posture. Current back pain (>24hrs: yes/no) was associated with a kyphotic sitting posture (time spent with a lumbar curve ≥180°) (R2 0.05), although future back pain was not. Using multivariable models: limited variety of lumbar movement whilst sitting was associated with future (persistent) LBP, dominating other variables (R2 0.11); yet high levels of reported back discomfort, physical aggravating factors and psychological demand at work were stronger predictors of sickness absence, and dominated other variables (R2 0.24). Interpretation: Workers do not follow the advice from employers to maintain a lumbar lordosis whilst sitting, as recommended by statutory bodies. Furthermore, sitting with a kyphotic posture did not increase the risk of back pain, although a relative lack of lumbar movement did. Thus, ergonomic advice encouraging lumbar movement-in-sitting appears to be justified. Predictors of sickness absence were multi-factorial, and consideration of work-relevant biomedical and psychosocial factors would be more useful than adopting more narrow screening approaches.

Biosensors are often used to detect biochemical species either in the body or from collected samples with high sensitivity and specificity. Those based on piezoelectric sensing methods employ mechanically induced changes to generate an electrical response. Reliable collection and processing of these signals is an important aspect in the design of these systems. To generate the electrical response, specific recognition layers are arranged on piezoelectric substrates in such a way that they interact with target species and so change the properties of the device surface (e.g. the mass or mechanical strain). These changes generate a change in the electrical signal output allowing the device to be used as a biosensor. The characteristics of piezoelectric biosensors are that they are competitively priced, inherently rugged, very sensitive, and intrinsically reliable. In this study, a compound label-free biosensor was developed. This sensor consists of two elements: a Love wave sensor and an electrochemical impedance sensor. The novelty of this device is that it can work in both dry and wet measurement conditions. Whilst the Love wave sensor aspect of the device is sensitive to the mass of adsorbed analytes under both dry and wet conditions with high sensitivity, the sensitivity coefficients in these two conditions may be different due to the different (mechanical) strengths of interaction between the adsorbed analyte and the substrate. The impedance sensor element of the device however is less sensitive to the mechanical strength of the bond between the analyte and the sensing surface and so can be used for in-situ calibration of the number of molecules bound to the sensing surface (with either a strong or weak link): conventional Love wave sensors are not sensitive to material loosely bound to the surface. Thus, a combination of results from these two sensors can provide more information about the analyte and the accuracy of the Love wave sensor measurements in a liquid environment. The device functions with label-free molecules and so special reagents are not needed when carrying out measurements. In addition, the fabrication of the device is not too complicated and it is easy to miniaturise. This may make the system suitable for point-of-care diagnostics and bio-material detection. The substrate used in these sensors is 64°Y–X lithium niobate (LiNbO3) which is a kind of piezoelectric material. On the substrate, there is a pair of interdigital transducers (IDTs) which are composed of 100 Ti/Au split-finger pairs with a periodicity (λ) of 40μm. The acoustic path length, between both IDTs, is 200λ and the aperture between the IDTs is 100λ. On top of the substrate and IDTs, there is a PMMA guiding layer with an optimised thickness ranging from 1000 nm to 1300 nm. In addition, a gold layer with thickness 100 nm is deposited on the guiding layer to act as the electrodes for the electrochemical impedance sensor. The biosensor in this study has been used to measure Protein A, IgG, and GABA molecules. Protein A is often coupled to other molecules such as a fluorescent dye, enzymes, biotin, and colloidal gold or radioactive iodine without affecting the antibody binding site. In addition, the capacity of Protein A to bind antibodies with such high affinity is the driving motivation for its industrial scale use in biologic pharmaceuticals. Therefore, measuring Protein A binding is a useful method with which to verify the function of the biosensor. IgG is the most abundant antibody isotype found in the circulation. By binding many kinds of pathogens including viruses, bacteria, and fungi, IgG protects the body from infection. Also, IgG can bind with Protein A well so the biosensor here could also measure IgG after a Protein A layer is immobilised on the sensing area. GABA is the main inhibitory neurotransmitter in the mammalian central nervous system. It plays an important role in regulating neuronal excitability throughout the nervous system. The conventional method to measure concentrations of GABA under the extracellular conditions is by using liquid chromatography. However, the disadvantages of chromatographic methods are baseline drift and additions of solvent and internal standards. Therefore, it is necessary to develop a simple, rapid and reliable method for direct measurement of GABA, and the sensor here is an attractive choice. When the Love wave sensor works in the liquid media, it can only be used to measure the mass of analytes but does not provide information about the conditions of molecules bound with the sensing surface. In contrast, electrochemical impedance sensing based on the diffusion of redox species to the underlying metal electrode can provide real-time monitoring of the surface coverage of bound macromolecular analytes regardless of the mechanical strength of the analyte-substrate bond: the electrochemical impedance measurement is sensitive to the size and extent of the diffusion pathways around the adsorbed macromolecules used by the redox species probe i.e. it is sensitive to the physical area of the surface covered by the macromolecular analyte and not to the mass of material that is sensed through a mechanical coupling effect (as in a Love wave device). Although electrochemical impedance measurements under the dry state are quite common when studying batteries and their redox/discharge properties, these are quite different sorts of systems to the device in this study. Therefore, integrating these two sensors (Love wave sensor and electrochemical impedance sensor) in a single device is a novel concept and should lead to better analytical performance than when each is used on their own. The new type of biosensor developed here therefore has the potential to measure analytes with greater accuracy, higher sensitivity and a lower limit of detection than found when using either a single Love wave sensor or electrochemical impedance sensor alone.