A biopsychosocial approach to understanding the contributors to chronic pain

Background: The biopsychosocial model suggests that chronic pain is the result of a dynamic interplay between biological, psychological, and social factors. Psychological variables, such as anxiety, are recognised as key contributors to chronic pain. However, their role, and particularly that of cognitive ability, in chronic pain from osteoarthritis (OA) ¬- the main cause of chronic pain worldwide - is less well understood. Social stressors, such as social deprivation may further explain the pain experience and may predict increased functional disability (limitations in activities of daily living), but their interaction with psychological factors has not been studied much, especially in the early stages of OA. Meanwhile, anxiety and social isolation extending over many months during the coronavirus pandemic may exacerbate pain in chronic pain patients, but this has not been investigated over time. Chronic pain in OA is underpinned by neuropathic and pain sensitization mechanisms that may predict resistance to therapy. Considering biological factors, such as host genetics, may enhance our understanding of these key mechanisms in OA. Lastly, little is known about the gut microbiome’s role in chronic pain. Identifying the gut microbiome’s link with anxiety may give us some understanding of how it might be linked to pain. Therefore, the aim of the current thesis was to explore how psychosocial factors (cognitive ability, anxiety and social deprivation/isolation), host genetics and the gut microbiome, interact in their contribution to chronic pain. Objectives: [1] to design a study aimed at investigating the specific neuro-cognitive domains that are impaired in OA pain, [2] to explore the dynamic interplay between key socioeconomic (i.e. social deprivation, education) and psychological factors (i.e. cognitive function, anxiety) and limitations in activities of daily living in OA pain; [3] to explore the potential impact of the first coronavirus lockdown on social isolation, anxiety, pain and overall quality of life in chronic pain patients; [4] to identify genetic markers associated with pain sensitisation and neuropathic-like pain in OA; and [5] to identify bacterial taxa reproducibly associated with anxiety and the role of serum levels of microbial metabolites and neuromodulators. Methods: [1] We recruited eight participants from the Nottinghamshire area to conduct a feasibility study. Distinct cognitive functions were measured using the Cambridge Neuropsychological Test Automated Battery (CANTAB). Pain measures included Quantitative Sensory Testing and pain questionnaires, such as painDETECT. Confounders were measured, including depression/anxiety, sleep quality, quality of life and physical function. Additional cognitive measures were taken, along with the National Adult Reading Test that served as a proxy of general premorbid IQ. Data on prosociality were also collected. For identifying molecular biomarkers, gut microbiome measures and genetic markers, blood, urine and faecal samples were collected. We assessed the feasibility of the study based on the experiences of the patients. Descriptive statistics were reported and correlational analysis was used to investigate the relationship between cognitive function, anxiety and pain. [2] We used data from waves 4, 5, 6 and 7 of the Survey of Health, Ageing and Retirement in Europe (SHARE) (n = 971) and selected a subsample of respondents who initially did not report a diagnosis of OA until wave 6. We used path models to test how social deprivation, education and anxiety, before diagnosis (waves 4 and 5), affect the relationship between cognitive ability, pain and limitations in activities of daily living following diagnosis (waves 6 and 7). [3] We analysed data from a sample of chronic pain patients (N = 305 of which 28.9% reported having OA) from Investigating Musculoskeletal Health and Wellbeing cohort thrice during the pandemic (May 2020, July 2020, September 2020). We employed repeated measures ANOVAs (and Kruskal–Wallis where appropriate) to compare measures of affect, social isolation, exercise changes due to COVID-19, quality of life and pain at the various time points. [4] We performed a genome-wide gene-based association study using pressure pain detection thresholds (PPTs) from distal pain-free site (anterior tibia), a measure of distal sensitisation, and from proximal pain-affected site (lateral joint line), a measure of local sensitisation, in 320 knee OA participants from the Knee Pain and related health in the Community (KPIC) cohort. We next performed gene-based fixed-effects meta-analysis of PPTs and a neuropathic-like pain phenotype (painDETECT) using genome-wide association study data from KPIC and from an independent cohort of 613 OA participants, respectively. [5] We meta-analysed data from two cohorts (WebEx: n = 37 and Omega 3/Fibre: n = 58) and published literature (n = 60) in order to identify bacterial taxa reproducibly associated with anxiety. We further investigated the role of serum levels of SCFAs and indoles on anxiety and the links between anxiety-associated operational taxonomic units (OTUs) and neuromodulators involved in anxiety. Gut microbiota was characterized using 16S rRNA sequencing in stool samples. OTU data were transformed and normalized. Serum metabolites were measured using tandem mass spectrometry. Results: [1] The study was feasible and all practical aspects were executed as planned, and the study was well received by the patients. Our preliminary findings suggested that anxiety, pain and cognitive ability are correlated and that global cognitive impairment and impairments in certain aspects of cognitive ability might be associated with poor physical function, increased neuropathic-like pain and increased central pain processing. [2] We showed that high social deprivation before diagnosis predicted greater limitations in activities of daily living after diagnosis, with this effect partly mediated by impaired cognitive ability. We also found that higher educational attainment before diagnosis may protect against limitations in activities of daily living after diagnosis via better cognitive ability and lower anxiety. [3] Covid-19 lockdown increased anxiety in May and September, whereas pain levels were unaffected. The highest levels of social isolation and depression occurred in the early stages of lockdown but declined thereafter. The at least a bit anxious group had heightened pain in May and September compared to the not at all anxious group. Doing at least an hour of exercise per day (on a typical day) was a protective factor against increased anxiety during lockdown, and quality of life improved by September when lockdown rules were eased. [4] The most significant genes associated with distal and local sensitisation were OR5B3 and BRDT, respectively. We also found previously identified neuropathic pain-associated genes – KCNA1, MTOR, ADORA1, and SCN3B – associated with PPT at the anterior tibia and an inflammatory pain gene – PTAFR – associated with PPT at the lateral joint line. Meta-analysis results of anterior tibia and neuropathic-like pain phenotypes revealed genes associated with bone morphogenesis, neuro-inflammation, obesity, type 2 diabetes, cardiovascular disease and cognitive function. [5] We identified reproducible microbiome associations of anxiety with Escherichia-Shigella (fixed-effect beta (95% CI) = 0.28 (0.11, 0.45), False Detection Rate (FDR)-adj. p = 3 x 10−3), Christensenellaceae R-7 group (fixed-effect beta (95% CI) = -0.32 (-0.48, -0.15), FDR-adj. p = 1 x 10−3) and Ruminococcaceae UCG-010 (fixed-effect beta (95% CI) = -0.24 (-0.40, -0.08), FDR- adj. p = 6 x 10−3). We found no significant associations between anxiety and gut-derived metabolites nor between anxiety-associated bacteria and anxiety-related markers after meta-analysis and correcting for multiple testing. Conclusion: This thesis showed that improving cognitive ability and reducing anxiety early in the disease course might protect against the negative impact of socioeconomic factors on subsequent functional disability in painful OA. Furthermore, regular exercise during lockdown and a balanced gut microbiome may contribute to reduced anxiety levels in chronic pain. Therefore, interventions like cognitive behavioural therapy, exercise, and probiotic supplementation might alleviate the influence of pain and anxiety on poor health outcomes in older adults with OA. Future research will investigate the role of inflammation, and the gut microbiome in the link between pain sensitisation and psychosocial factors, such as cognitive ability