Background: Identification of low back pain subgroups has been highlighted as a .priority for secondary prevention of persistent problems in primary care. The aim of this PhD was to develop and evaluate a new clinical tool (called the STarT Back Tool) to allocate patients with back pain into initial treatment subgroups based on treatmentmodifiable prpgnostic indicators. Methods: The setting was, UK primary care adults consulting with non-specific pack ··pain:Constructs that were independent prognostic indicators for persistence were identified from secondary analysis of 2 existing cohorts and published literature and single screening questions identified. Psychometric properties of the tool, including concli.rrent and discriminant validity, internal consistency, repeatabili~y, were assessed . within a 'development' cohort (n=131) and tool score cut-offs established to enable alloc~tion to three subgrOlJPs ('low', 'medium' and 'high' risk). Predictive and ext~rnal . validity were evaluated within an 'independent external' cohort (n=500). The tool was . compared to an existing musculoskeletal screening tool (Orebro Musr:;uloskeletal Pain Screening Questionnaire) and to clinical acumen using video taped clinical interviews, . reviewed by primary care clinical-experts. Finally, the practical application of the tool was tested in. a pilot study and demonstrated to be feasible and acceptable to both . patients and clinicians. Results: The Tool included nine items: referred leg pain; comorbid pain; disability (two items); bothersomeness; catastr~phising; fear; anxiety, and depression. The latter 5 items were identified as a psychosocial subscale. The Tool demonstrated good reliability and validity and was acceptable to patients and clinicians. Conclusions: Identifying treatment subgroups with distinct clinical characteristics among patients in primary care presenting wi.th low back pain is pospible and feas'ible in clinical practic~. The tool shows potential for enabling early targeted interventions for the prevention of persistent disabling back problems.

This thesis consisted of five interlinking studies to examine the role of expectations in exercise behaviour in individuals with knee osteoarthritis (OA). In a systematic review on the role of expectations in interventions for chronic musculoskeletal joint pain, positive expectations were consistently related to positive outcomes. However, only three studies were suitable and a number of methodological problems with measuring expectations were identified. A cross-sectional qualitative study explored outcome and treatment expectations and findings were used to develop a questionnaire to measure OA-related beliefs. In a second qualitative study a small number of participants were re-interviewed a year later to examine how the condition and perceptions of the conditions had changed. Findings suggest expectations are malleable and respond to new information. The Knee Pain Beliefs Questionnaire (KPBQ) was developed from the qualitative data and consisted of four factors, including two on expectations (future state/illness expectations and defensive optimism). The measure showed satisfactory internal reliability, test re-retest reliability and discriminant validity. The KPBQ was used to measure expectations in a subsample of primary care OA patients participating in a randomised controlled trial of a self-management intervention. Participants were randomised to one of three arms; control (n=30), individual intervention (n=27), group intervention (n=33). Questionnaire were completed at baseline, post-intervention (approximately 6-8 weeks after baseline) and follow-up (approximately 6 months after post-intervention). In the cross-sectional analyses of these data (N=80), illness expectations and defensive optimism were more strongly related to activity levels than outcome or self-efficacy expectations. In the longitudinal analyses (N=57, 19 per arm), current beliefs increased in both intervention arms and defensive optimism increased in the group arm.

No description available.

In the arthritides, the breakdown of articular cartilage and the erosion of subchondral bone lead to loss of efficient joint function. Matrix metalloproteinases (MMPs) can collectively degrade the extracellular matrix of cartilage, and have been strongly implicated in cartilage destruction. MMPs are produced as inactive precursors requiring activation. Addition of interleukin-1 (IL-1) and oncostatin M (OSM) to bovine cartilage in explant culture results in a synergistic loss of the collagen matrix, accompanied by a dramatic increase in pro-MMP synthesis and activation. Increasing evidence implicates serine proteinases in pathologic tissue turnover and the transmembrane serine proteases fibroblast activation protein-α (FAPα), dipeptidyl peptidase IV (DPPIV) and matriptase were recently found to be up-regulated in osteoarthritic (OA) cartilage compared to normal cartilage. The aim of this work was to identify whether these transmembrane serine proteases have roles in cartilage collagen breakdown. Inhibition of IL-1+OSM-mediated collagen breakdown was not observed using selective FAPα inhibitors, indicating that FAPα enzyme activity per se plays no direct role in collagen breakdown. In contrast, DPPIV inhibition showed significant protection of IL-1+OSM-mediated collagen breakdown. However, the mechanism of action of DPPIV remains to be elucidated. Matriptase was demonstrated to activate pro-MMPs as well as induce MMP expression in OA cartilage via protease-activated receptor 2 activation. This makes matriptase a key protease in the pathology of OA. Overall, this study has identified matriptase as a key serine protease in OA pathology and confirmed a potential role for DPPIV in cytokine-mediated cartilage degradation. These findings support the need for further research in order that therapeutic interventions targeting these enzymes may be realised.

Rheumatoid arthritis (RA) is an autoimmune disease characterised by chronic inflammation at multiple joints. Because of the association of the condition with certain alleles of the DRB1 gene which encodes the β-chains of human leukocyte antigen (HLA)-DR molecules, it has been postulated that auto-reactive CD4+ T-cells - specific for joint-associated autoantigens presented in the context of these RA-associated HLA-DR molecules – are involved in the initiation or perpetuation of the disease process. In addition, citrullination of autoantigen-derived peptides may enhance their binding affinity to RA-associated HLA-DR molecules, implicating citrullinated autoantigens in RA pathogenesis. Since RA primarily affects the joints, proteins of articular origin and those expressed in the joints such as human cartilage glycoprotein 39 (HCgp39), type II collagen (CII), aggrecan, α-enolase, fibrinogen and vimentin are candidate autoantigens in RA. However, it has been difficult to consistently detect T-cell responses to these candidate RA autoantigens in RA patients. In this project, I hypothesised that T-cell responses to candidate autoantigens are heightened in RA patients compared to healthy subjects and that the T-cell responses of RA patients are pro-inflammatory while those of healthy subjects are not. I first optimised an experimental system for detecting autoantigen-specific T-cell responses and applied the protocol to measure proliferative and cytokine production responses of peripheral blood mononuclear cells (PBMC) from RA patients, healthy subjects and disease controls to unmodified and citrullinated whole protein and peptide forms of the candidate RA autoantigens mentioned above. Responses to the candidate autoantigens were readily detectable in RA patients, healthy subjects and disease controls. Furthermore, I found no evidence of increased immunogenicity of citrullinated peptides versus their unmodified counterparts. Thus the study failed to provide evidence that T-cell responses to either unmodified or citrullinated autoantigens are important in RA pathogenesis.

In order to maintain visual fixation, the oculomotor muscles are controlled by reflexes arising from the complex interaction between the vestibular apparatus, the perception of slippage of the target image on the retina and, significantly, by the proprioceptive mechanisms of the cervical spine; the cervico-ocular reflex (COR). Chronic neck dysfunction, particularly in whiplash patients, has been associated with modifications in eye movement behaviour. The nature of this association is poorly defined, although previous studies on individuals with post-whiplash syndrome have noted such things as delayed high speed eye movements following immobilisation in a cervical spine collar; altered capabilities of pursuit when following a target; increased gain (amplitude of movement with reference to input movement) of the COR and alterations in the ability to accurately fixate on target positions. Much of the previous literature has reported qualitative changes in passing, sometimes hampered by the limits of technology employed in the experimentation, and often by repeated measures on a few individuals; accordingly, the precise nature of the COR in the presence of cervical dysfunction requires further description. The purpose of this study was to define the precise nature of the influence of the COR on eye movement in a large cohort and to describe any changes that may be present in patients with whiplash or other types of neck dysfunction.

Articular stiffness is an important symptom in most arthritic diseases and appears to be a useful marker of disease activity in rheumatoid arthritis. Attempts to obtain a reliable objective measure of articular stiffness span the last 30 years but a meaningful measure of this symptom remains elusive. A number of reasons have been suggested to explain the discrepancy between objective and subjective stiffness in arthritis and these can be summarised as: a semeiological confusion, aberrant mechano-receptor thresholds and concurrent muscle wasting. This thesis examines each of these hypotheses. Some patients may confuse pain and stiffness or may wish to use other words to describe their joint symptoms. A questionnaire was developed which enabled patients to express their joint symptomatology using a wide range of descriptors. No differences were found between health professionals and patients in their definition of each of the descriptors. The questionnaire discriminated clearly between groups of patients with rheumatoid arthritis, ankylosing spondylitis and non-articular rheumatism. Movement perception threshold was measured in the finger but it was found that subjects relied on cutaneous information. Vibration perception threshold was used as an alternative measure of mechano-receptor thresholds: no abnormalities were found in 50 patients with rheumatoid arthritis. Muscle cross-sectional area was calculated from anthropometric data and the results compared with measurements obtained from computed tomographic scans. A significant decrease in forearm muscle cross-sectional area was found in rheumatoid arthritis but the decrease was not sufficient to explain the reduction in grip strength observed, some of the variation being explained by deformity and pain in the joints. From this study it was possible to make a correction for muscle wasting in previously published stiffness data, revealing significant increases in metacarpo-phalangeal joint stiffness in rheumatoid arthritis. This result was confirmed in new data based on the resonant frequency of the wrist. Further data on the qualitative aspects of muscle were obtained by relating dynamic angular wrist stiffness to level of contraction of forearm muscles. Although arthritic subjects differed significantly from normals at maximum activation, when the results were expressed in terms of absolute grip strength no differences were found, suggesting inhibition of muscle activation in rheumatoid arthritis. It is concluded that symptomatic stiffness is objectively quantifiable in arthritis providing measurements are made in relationship to the equilibrium position of the joint and providing a correction is made for muscle wasting.

The purpose of this study was to examine the concept of the physiological management of pain following acute Grade 2 whiplash injury while facilitating recovery and maintenance of a full range of cervical spine movement. In a series of studies a statistically significant effect was demonstrated on the improved and sustained range of cervical movement p<0.0005 and on the effective control of pain p<0.0005 generated by movement to relieve post injury stiffness. In an Audit study there was a beneficial effect on the improvement in patient confidence in selfmanagement of their condition 84% following attendance at an educational neck school. The effective use of soft collars on the available range of cervical movement p<0.025 and on the balance of the body showed a statistically significant result p<O.OOl. The study results indicated that there are inbuilt systems of pain control which can be effectively used to control pain following Grade 2 whiplash injury while facilitating the recovery of full range of cervical movement. This study emphasised the importance of not allowing post injury stiffness to develop and that post injury stiffness prejudices the survival of the joint neuroreceptors. The study showed the relevance of the Quebec Task Force Guidelines in the management of whiplash injury.

Finding genes that lead to an increased susceptibility towards complex genetic disorders would be of enormous benefit to both individuals and society. These genes would give an insight into disease mechanism and could potentially lead to improvements in drug selection for patients, reveal novel biology pathways for drug targeting, allow clinical prediction of likely disease progression and reduce the inappropriate treatment of patients. Here I present four publications that document the progressive search for candidate susceptible disease genes for rheumatoid arthritis (RA). Starting with a whole genome screen (WGS), locating de novo regions of the genome likely to harbour casual genes, this is followed by a replication study, confirming the regions of interest. The discovery study confirmed the known association between RA and the HLA region, and generated 10 other putative candidate regions worthy of further investigation. Six of these regions overlapped with other published RA WGS (1q, 3p, 6q, 1Oq, 14q and 16p) and four were specific to my UK study (4p, 7p, 21p and Xq). Replication analysis in the 10 candidate regions, genotyped in an independent cohort, failed to validate any of the initial findings. This is probably due to type 1, false positive, errors in the discovery study 6 and a lack of power in the replication cohort. Stratification analysis did, however, reveal regions of the genome linked to a more severe, younger age of onset disease. Two of these regions, on 1q and 16p, contained strong candidate genes, FCRL3 and MHC2TA, with reported positive association to RA in Japanese and Swedish cohorts respectively. Two of my subsequent papers presented here investigate the associations between these 2 strong candidate genes and RA in our UK cohort. No association was found for any polymorphism tested, including the reported functional changes, for either FCRL3 or MHC2TA in our UK cohort. This may indicate population specificity for RA causal genes, or that the initial reported association was a false positive.

No description available.