Abstract Background/Aims Osteoarthritis (OA) is a major cause of disability worldwide, yet the systemic processes that precede disease onset remain ill-defined. Metabolomic profiling offers an opportunity to identify antecedent biological pathways linked to OA risk beyond traditional factors. The UK Biobank is a large prospective longitudinal research database (> 500,000 participants) with detailed clinical data and nuclear magnetic resonance (NMR)-based metabolomics assays from blood samples at enrolment. Methods A nested case-control analysis was conducted within UK Biobank. Participants who developed hip, knee or hand OA (n = 30,490) after metabolomic sampling were matched 1:1 to controls on age, sex, body mass index (BMI) and smoking status. Serum metabolomic data (up to 249 measures) were analysed using random survival forests (RSF), a machine learning approach for time-to-event data, with 500 trees per model. RSF accommodates high-dimensional inputs and non-linear effects, moving beyond single metabolite analyses. Results Participants had a median age of 68.6 years (IQR 62.9-73.3) and 57.9% were female, with a mean BMI 29.2 kg/m2 (SD 5.2) and 8.8% were current smokers, 40.3% previous smokers, and 50.4% never smokers. There was a median interval of 7.3 years (IQR 4.1 - 10.3) between metabolomic blood sampling and subsequent OA diagnosis. The RSF model had a Brier score at 5 years of 0.14 (Brier scores range from 0 to 1, with lower values indicating better accuracy) and highlighted three broad domains of metabolic variation associated with subsequent OA diagnosis: 1) Markers of energy metabolism (lactate, citrate, amino acids including alanine and valine); 2) Markers of inflammation (glycoprotein acetyls); and 3) Lipid measures (apolipoprotein B, VLDL cholesterol, omega-6 fatty acids). Increased risk of OA was associated with higher levels of glycoprotein acetyls, the amino acid alanine, and an increased ratio of saturated fatty acids to total fatty acids; whilst higher proportions of polyunsaturated fatty acids, particularly linoleic acid, were associated with reduced risk of OA. Conclusion This analysis demonstrates that systemic metabolic perturbations are detectable years before OA onset, reducing the likelihood of reverse causation, and supports the potential future role of metabolomic biomarkers. The findings converge on three mechanistic pathways: 1) impaired energy metabolism, 2) chronic low-grade inflammation, and 3) adverse lipid composition. These processes provide biologically coherent explanations for how systemic metabolism may influence joint vulnerability, whilst protective associations with polyunsaturated fatty acids highlight potentially modifiable risk factors. Previous cross-sectional studies have reported metabolite-OA associations, while a recent longitudinal study incorporated metabolomics into a predictive OA diagnosis model but focussed on short-term prediction and patient stratification. In contrast, our analysis places metabolomics at the centre, uses a larger incident OA cohort with longer follow-up, and applies RSF to model time-to-diagnosis whilst capturing non-linear interactions with clinical covariates Disclosure P. Saha: None. M. Defernez: None. G. Le Gall: None. P. Cardenas-Canto: None. J. Dainty: None. G. Wortley: None. M. Yates: None. R. Davidson: None. I. Clark: None. M. Traka: None. K. Kemsley: None. A. MacGregor: None.

Abstract Background/Aims Hydroxychloroquine (HCQ)retinopathy is a serious adverse event of long-term therapy, often irreversibleand asymptomatic until advanced. Royal College of Ophthalmologists (RCOphth)guidelines recommend annual retinopathy monitoring for patients receivinglong-term HCQ therapy (>5 years) due to heightened risk. Additionally, ifany of the following factors are present, including concurrent tamoxifen use, renal impairment (eGFR <60mL/min/1.73m2) or high dose HCQ use(>5mg/kg/day) retinal maculopathy screening is recommended after one year oftreatment with HCQ. The guideline recommends screening comprising bothspectral-domain optical coherence tomography (SD-OCT) and wide-field fundusautofluorescence imaging (FAF), the latter of which is usually only availablein hospital ophthalmology departments. This audit assesses national compliancewith these guidelines across England. Methods Data on prescriptions of hydroxychloroquine, along with clinical characteristics of those prescribed the medication were obtained from a large administrative dataset (ECLIPSE - Prescribing Services Limited) covering 28,689,113 patients in England. Using clinical coding frameworks and information on weight, renal function and concomitant tamoxifen prescription were linked to outpatient attendance at ophthalmology services via referral code on the e-referral service (e-RS). HCQ duration of treatment along with HCQ dose/weight ratios were calculated. Those meeting RCOphth criteria were checked against ophthalmic outpatient attendance within the previous 365 days. Results Date of data download was the 28th August 2025 with clinical codes for prescriptions available since the 1st January 2016. A total of 52,419 patients (79% of whom were female) had been prescribed hydroxychloroquine with a median age of 61 years (IQR 51 to 71 years) across 32 Integrated Care Boards (ICB). A total of 26,780 patients had an active hydroxychloroquine prescription within the last three months. The median dose/kg was 3.6 mg/kg (IQR 2.7 to 4.7mg/kg), with 5,338 patients prescribed >5mg/kg and of those 1,115 >6.5mg/kg. There were 7432 patients prescribed hydroxychloroquine for longer than six years. There were 3,702 patients with an eGFR <60mL/min/1.73m2 of whom 238 had an eGFR <30mL/min/1.73 m2. There were 125 patients co-prescribed tamoxifen. Those meeting the threshold for screening comprised 13,298 individuals of whom 2,808 (21.1%) had been seen in ophthalmology services in the previous 365 days. This proportion ranged from 0% to 45.9% according to ICB region. Conclusion There appears to be suboptimal compliance with suggested RCOphth guidelines for screening meaning patients are at risk. The variability by ICB region may be reflective of barriers to access or commissioning policy. Limitations include that coding frameworks were based on e-RS and therefore rejected referrals would not have been captured nor those that attended community opticians. Considering that risk factors may vary over time, patient records should be regularly updated to ensure that newly emerging risks are identified, to ensure timely monitoring is initiated. Disclosure M. Odunlami: None. E. Metcalf: None. S. Platt: None. J. Brown: None. P. Saha: None. A. MacGregor: None. M. Yates: None.

Background Fibromyalgia Syndrome (FMS) is highly prevalent with a significant associated morbidity and socioeconomic burden. Effective treatments for FMS remain elusive with pharmacological management (including use of opioids) often proving ineffective. There is a need to develop accessible, innovative management approaches to improve patient care. Virtual reality (VR) interventions have shown evidence of efficacy in the management of acute pain and chronic low back pain, but their feasibility in FMS has not hitherto been explored. Methods This feasibility study investigates the use of four different VR systems, four interactive VR activities and two virtual environments in patients with FMS. Acceptability (including adverse effects) and study engagement were the main outcomes investigated. Clinical outcome data on pain and mood were also collected to gather preliminary information for future studies. Results The results demonstrated good feasibility across VR systems, activities and virtual environments with high levels of acceptability, low frequency of adverse effects, and positive perceptions of VR in patients with FMS. Reporting of adverse effects (including fatigue) varied across different VR components, with system comfort and virtual environmental design being particularly important. Conclusions The findings increase our confidence with respect to the feasibility of using VR in people with FMS, help to inform future randomised controlled trials and emphasise the importance of tailored interventional design for future VR therapeutics.

Background/Aims The European Prospective Into Cancer (EPIC) - Norfolk Study is a large UK observational study, with health and blood marker data. The study design facilitates analysis of incident cases of rheumatoid arthritis (RA). Novel markers of inflammation including the neutrophil lymphocyte ratio (NLR) and the platelet lymphocyte ratio (PLR) have been assessed in other cohorts, and we sought to characterise the inflammatory phase preceding disease onset. Methods Incident RA cases were validated in the EPIC-Norfolk study: age, sex, CRP, ferritin and fibrinogen were from the 1st health check (1993-1998), whilst NLR and PLR were calculated from the 3rd health check (2004-2011) due to data availability. NLR and PLR are respectively calculated as absolute neutrophil count/absolute lymphocyte count, and absolute platelet count/absolute lymphocyte count. Results There are 415 cases of RA, including 206 incident cases, in a cohort of 25,636 participants. Linear model ANOVA testing on an unmatched population, identified a statistically significant difference between incident RA cases and non-RA cases for CRP (mean [SD] 6.32 [10.29] vs. 3.07 [6.23], n = 251), NLR (mean [SD] 1.02 [1.90] vs. 0.58 [1.32], n = 80), and PLR (mean [SD] 54.6 [94.2] vs 22.4 [70.8], n = 258); all p-values <0.001. There was also a significant difference in fibrinogen (mean [SD] 3.10 [0.95] vs 2.94 [0.99], p = 0.032), but no difference for ferritin (mean [SD] 78.31 [63.82] vs. 89.72 [75.34], p = 0.086). Higher CRP levels at enrolment were related to a shorter duration to diagnosis of incident RA (Table 1). A linear regression model demonstrates the negative association between CRP and duration till diagnosis of RA (p = <0.001); with a CRP reference level of 12.6, as the group variable increases, the expected CRP value decreases by 1.8. When adjusting for age and sex, this remains statistically significant (p < 0.001). Conclusion Our study adds to existing literature that there are additional blood markers of inflammation from routine blood tests, beyond CRP and ESR, which can be reviewed when assessing patients with possible RA. There exists a pre-disease inflammation risk that is detectable before disease onset, and our data suggest that there is a ‘window of opportunity’ 5-15 years before RA onset, where preventative strategies might be worthwhile implementing early. Disclosure P. Saha: None. J. Dainty: None. M. Yates: Grants/research support; NIHR, Versus Arthritis, Doris Hillier, PMRGCAuk, Health and Social Care Partners, Norfolk and Waveney Integrated Care Board, Norfolk and Norwich University Healthcare Trust Charitable Funds. Other; Advisory Board work - AbbVie, BioGen, Galapagos, Conference attendance - Lilly, AbbVie, UCB, Celltrion. A. MacGregor: None.

Despite advances in inflammatory arthritis management, long-term outcomes remain suboptimal; factors such as frailty and malnutrition may contribute but are under-investigated. Malnutrition, a modifiable environmental factor, can be assessed through validated indices such as the Prognostic Nutritional Index (PNI); low PNI levels suggest higher nutritional risks and poorer prognostic outcomes. The Norfolk Arthritis Register (NOAR) is a longitudinal inception cohort of ∼4,500 early inflammatory arthritis cases since 1989. Inclusion criteria were age >16 years with 2 swollen joints lasting ≥4 weeks. Data from NOAR were extracted using CogStack, including albumin and lymphocyte counts at years 1, 5, and 10. PNI is calculated as serum albumin (g/L) + 5 × total lymphocyte count (109/L), and values below the 10th percentile of the normal range are commonly taken as an indicator of malnutrition (∼45-49). Analyses employed a univariate linear regression model and a linear mixed effects multilevel model using lme4 and lmerTest packages in R. Among 4,488 participants (2,975 female, 1,513 male) with a mean age at diagnosis of 53.1 years (SD 14.7), PNI data were available for 40.4% at year 1, 46% at year 5, and 38.8% at year 10. Mean PNI values were 47.2 (SD 6.6) at year 1, 46.6 (SD 7.2) at year 5, and 45.3 (SD 8.8) at year 10. As outlined in table 1, the linear regression model showed that follow-up time significantly predicted PNI decline. The multilevel model revealed a significant interaction between age at onset and follow-up time, highlighting that those with a later age at onset experience a more accelerated decline in PNI over time, independent of disease activity. This suggests that later age at onset is a critical factor in the worsening of nutritional status. This analysis highlights that patients with incident inflammatory arthritis, particularly those with a later age at onset, face an elevated nutritional risk as indicated by declining PNI scores over time. These findings underscore the need for further studies to characterise malnutrition in the elderly with RA, and emphasises the importance of targeted nutritional interventions and policies to improve long-term outcomes. P. Saha: None. J. Dainty: None. M. Shemko: None. M. Yates: Grants/research support; NIHR, Versus Arthritis, Doris Hillier, PMRGCAuk, Health and Social Care Partners, Norfolk and Waveney Integrated Care Board, Norfolk and Norwich University Healthcare Trust Charitable Funds. Other; Advisory Board work - AbbVie, BioGen, Galapagos, Conference attendance - Lilly, AbbVie, UCB, Celltrion. A. MacGregor: None.

OBJECTIVES:This study aimed to investigate non-HLA genetic mechanisms underlying radiographic severity in rheumatoid arthritis (RA). METHODS:A systematic review of publications reporting non-HLA genetic associations with radiographic severity in RA across ancestries was undertaken. Experimental validation was performed in the Norfolk Arthritis Register, comprising 1407 patients with available genetic and treatment data followed prospectively for up to 10 years, with 2198 longitudinal radiographs. Genome-wide genotyping was performed with Illumina Human Core Exome Array. Radiographic outcomes (presence of erosions; Larsen score) were modelled longitudinally. Fine mapping and functional annotations to refine associations to potential causative loci were undertaken using FUMA, PolyPhen2, and RegulomeDB. RESULTS:The systematic review identified 102 publications reporting 139 independent associations with radiographic outcome. Association with 15 independent polymorphisms were replicated in the Norfolk Arthritis Register data set, implicating adaptive immune processes (Th1, Th2, and Th17 pathways), cytokine regulation, and osteoclast differentiation. Notably, we refined the association of rs59902911 at the CARD9 locus to an intronic polymorphism within an active enhancer (rs78892335), achieving genome-wide significance and with an effect size exceeding the minimal clinically important difference for each copy of the minor allele (4.78 Larsen units/copy; 95% CI, 3.15-6.41; p = 9.01 × 10-9). This polymorphism is associated with the expression of CARD9 in immune cells, including B cells. CONCLUSIONS:We provide a comprehensive list of validated genetic associations with RA outcome and demonstrate that non-HLA polymorphisms can associate with radiographic severity independently of disease susceptibility. This highlights the importance of dedicated genetic outcome studies for patient stratification in precision medicine for RA.

Background: Current guidelines recommend glucocorticoids (GCs) for the treatment of Polymyalgia Rheumatica (PMR) for 1 to 2 years. Surveys and administrative datasets from many European countries and the USA show that approximately 50% of patients are on GCs for at least 2 years. A previous study looking at individuals with PMR and GC dependence, found the neutrophil-lymphocyte ratio (NLR) at diagnosis to be associated with GC dependency. Objectives: To understand baseline factors and patient characteristics associated with continuing GCs at two years. Methods: Prescribing Services Limited (PSL) gained NHS Digital central assurance in 2017 and processes NHS data for 27 million people living in England, carrying out risk stratification for unplanned admission though the ECLIPSE Live system. We used de-identified data for patients not opted out of use of data for service development and research identifying all patients with a diagnosis of PMR between 1st Jan 2016 to 19th July 2021 and, linking to prescribing data, applied a previously used Clinical Practice Research Datalink (CPRD) definition of PMR (diagnosis of PMR, with issue of GCs within 6 months of diagnosis date and second prescription within 6 months of the first). Patient characteristics including blood test results were analysed using logistic regression to ascertain associations to GC prescription, with at least 2 years of follow-up. Results: There were 39,460 individuals with PMR who fulfilled these criteria; of these 22 were under the age of 40 years and excluded from analysis. The median age at diagnosis was 73.7 yrs (IQR 67.7 to 79.2 yrs), and 61.4% female. The median C-reactive protein (CRP) at diagnosis was 26 mg/L (IQR 11 to 53 mg/L), with 85.9% having a CRP of 6 mg/L or more, and 32.5% were anaemic at the time of diagnosis. The median NLR and platelet-to-lymphocyte (PLR) ratio was 3.36 (IQR 2.40 to 4.69) and 184 (IQR 137 to 248) respectively. GC prescription was assessed in 3-monthly brackets, observing those that managed to stay within a cumulative steroid exposure as recommended by the current BSR guidelines and those who were no longer prescribed steroids (See Table 1). Table 1. *The CPRD definition allowed individuals to start GCs within six months of diagnosis date and a second prescription within six months of the first prescription Logistic regression was carried out to assess factors associated with continuing GCs at 24 months. The multivariable model found an older age and higher initial steroid use were associated with lower probability of remaining on GCs, while females, smokers, those with anaemia, and a higher CRP and PLR were more likely to remain on steroids (see Table 2). Table 2. Conclusion: This is the largest study to assess baseline characteristics and their association to continuing GCs at 2 years. This community-based study of real-world evidence shows 86% have a raised CRP at diagnosis and is reflective of the demographic of what is expected for those diagnosed with PMR. Several baseline characteristics are associated with outcome at 2 years and exceeding the cumulative GCs dose in the first 3 months may allow individuals to come off steroids within 2 years. Replication of this study to other countries is of interest to inform GC tapering strategies. REFERENCES: [1] BSR guidelines for the management of Polymyalgia Rheumatica. Dasgupta et al. (2010) Rheumatology Oxford. [2] Incidence of diagnosed polymyalgia rheumatica and temporal arteritis in the United Kingdom, 1990-2001 Smeeth et al. (2006), Annals of Rheumatic Disease. [3] Neutrophil to lymphocyte ratio predicts glucocorticoid resistance in Polymyalgia Rheumatica. (2020) Owen et al. International Journal Rheumatic Disease. Acknowledgements: NIL. Disclosure of Interests: Max Yates Advisory Board for BioGen and Galapagos, Pratyasha Saha: None declared, Clare Aldus: None declared, Julian Brown: None declared, Alex MacGregor: None declared.

Background Plant-based diets may provide protection against cognitive decline and Alzheimer’s disease, but observational data have not been consistent. Previous studies include early life confounding from socioeconomic conditions and genetics that are known to influence both cognitive performance and diet behaviour. This study investigated associations between Mediterranean (MED) diet and MIND diets and cognitive performance accounting for shared genotype and early-life environmental exposures in female twins. Methods Diet scores were examined in 509 female twins enrolled in TwinsUK study. The Cambridge Neuropsychological Test Automated Battery was used to assess cognition at baseline and 10 years later (in n = 275). A co-twin case–control study for discordant monozygotic (MZ) twins examined effects of diet on cognitive performance independent of genetic factors. Differences in relative abundance of taxa at 10-year follow-up were explored in subsamples. Results Each 1-point increase in MIND or MED diet score was associated with 1.75 (95% CI : − 2.96, − 0.54, p = 0.005 and q = 0.11) and 1.67 (95% CI : − 2.71, − 0.65, p = 0.002 and q = 0.02) fewer respective errors in paired-associates learning. Within each MZ pair, the twin with the high diet score had better preservation in spatial span especially for MED diet ( p = 0.02). There were no differences between diet scores and 10-year change in the other cognitive tests. MIND diet adherence was associated with higher relative abundance of Ruminococcaceae UCG-010 (0.30% (95% CI 0.17, 0.62), q = 0.05) which was also associated with less decline in global cognition over 10 years (0.22 (95% CI 0.06, 0.39), p = 0.01). Conclusions MIND or MED diets could help to preserve some cognitive abilities in midlife, particularly episodic and visuospatial working memory. Effects may be mediated by high dietary fibre content and increased abundance of short-chain fatty acid producing gut bacteria. Longer follow-up with repeated measures of cognition will determine whether diet can influence changes in cognition occurring in older age.

Background/Aims Multiple long-term conditions (MLTC) refers to the co-existence of two or more chronic conditions. Rheumatic diseases (RMDs) are important long-term conditions which are common in individuals with MLTC. The effect of rheumatic disease on disability is well known and often requires support with activities of daily living, i.e. social care. However, while the presence of MLTC is known to be associated with social care need, the role of RMDs in driving social care need is unknown. This lack of data capture of social care resource use undermines healthcare strategy. We carried out two surveys in different parts of the country to ascertain the social care burden amongst those with RMDs. Methods An online cross-sectional survey to identify and characterise formal and informal social care use was sent to two different groups of patients with rheumatic and musculoskeletal disease. The first comprised participants within Norfolk Arthritis Register, who were emailed an invitation to complete the survey between March-July 2023. As part of Assembling the Data Jigsaw programme, the second group comprised patients attending the Salford Royal Hospital Rheumatology Department with text message invitations sent between October 2022-September 2023 via the DrDoctor app. All participants answered identical questions on formal and informal care provision and demographic characteristics. Logistic regression was used to identify factors associated with needing support. Results A total of 184 participants within the NOAR cohort completed the survey, comprising 128 (70%) women with a median age of 64 years. A total of 439 participants completed the survey in the Salford Royal Hospital cohort, comprising 326 (74%) women with a median age of 58 years. Rheumatoid arthritis was reported by 93 (73%) of the NOAR recruits and 136 (31%) of the Salford Royal Hospital cohort, with osteoarthritis the next most common with 37 (20%) and 112 (34%) of each cohort reporting this condition. Using self-reported data, 47% and 64% of the respective cohorts met the criteria for MLTC. For both cohorts, 30% reported needing four hours or fewer assistance per week, and 19% reported requiring >10hours a week. Although infrequent or occasional help was usually provided informally by spouses, partners, family members or friends, 42 (23%) of the NOAR cohort and 70 (16%) of the Salford Royal Hospital cohort arranged formal care provision themselves without involving the council. Multivariable logistic regression of both cohorts found that those who were younger were more likely to have reported needing assistance with ADLs in the last month. MLTC was also associated with requiring help. Conclusion The majority of individuals engage informal support or arrange care without help from local authorities. These data are important for describing social care resource use and for determining healthcare strategy. Disclosure M. Yates: None. M. Soomro: None. A. Onajole: None. A. MacGregor: None. W. Dixon: None. J. McBeth: None. J.H. Humphreys: None.

Objectives Disease-modifying antirheumatic drugs (DMARDs) are a first-line treatment in rheumatoid arthritis (RA). Treatment response to DMARDs is patient-specific, dose efficacy is difficult to predict and long-term results are variable. The gut microbiota are known to play a pivotal role in prodromal and early-disease RA, manifested by Prevotella spp. enrichment. The clinical response to therapy may be mediated by microbiota, and large-scale studies assessing the microbiome are few. This study assessed whether microbiome signals were associated with, and predictive of, patient response to DMARD treatment. Accurate early identification of those who will respond poorly to DMARD therapy would allow selection of alternative treatment (e.g. biologic therapy) and potentially improve patient outcome.Methods A multicentre, longitudinal, observational study of stool- and saliva microbiome was performed in DMARD-naive, newly diagnosed RA patients during introduction of DMARD treatment. Clinical data and samples were collected at baseline (n = 144) in DMARD-naive patients and at six weeks (n = 117) and 12 weeks (n = 95) into DMARD therapy. Samples collected (n = 365 stool, n = 365 saliva) underwent shotgun sequencing. Disease activity measures were collected at each timepoint and minimal clinically important improvement determined.Results In total, 26 stool microbes were found to decrease in those manifesting a minimal clinically important improvement. Prevotella spp. and Streptococcus spp. were the predominant taxa to decline following six weeks and 12 weeks of DMARDs, respectively. Furthermore, baseline microbiota of DMARD-naive patients were indicative of future response.Conclusion DMARDs appear to restore a perturbed microbiome to a eubiotic state. Moreover, microbiome status can be used to predict likelihood of patient response to DMARD.

Background: The ageing demographics of many European countries will result in a greater number of people diagnosed with disease, particularly those correlated with older age. In many European countries, people of older age tend to migrate from urban to rural and coastal areas. Polymyalgia Rheumatica (PMR) is strongly correlated to age, with rising incidence in each ten-year age band after the fifth decade of life. There is wide variation in glucocorticoid prescription and most patients are currently cared for by community-based practitioners. Whilst most patients can be managed effectively within the community, enhanced use of routinely collected data and better integration of data and services may result in improved outcome. Objectives: To improve support for community-based practitioners caring for people living with PMR. Methods: Working with Prescribing Services Limited (PSL) and key stakeholders, we developed a community-based practitioner support programme (PMR IMPROVE). PSL Advice and Guidance System (ECLIPSE Live) gained NHS Digital central assurance in 2017 and processes NHS data for 27 million people living in England, with risk stratification for unplanned hospital admissions. ECLIPSE is a data processor, fully integrated with all NHS England primary care records systems with bi-directional dataflow to and from NHS systems. We identified patients with a diagnosis of PMR within ECLIPSE and developed tools, including patient-portal, digital steroid card, structured medication review and Advice and Guidance recommendations to general practitioners (GPs) to pilot the new system in a Primary Care Network of 7 practices in West Norfolk, UK. Results: Over 100,000 patients had a diagnosis of PMR, resulting in a point prevalence of 2.2% in those greater than 55 years of age. Iterative development of patient-reported outcome measures (PROMs) based on current evidence was implemented within a region of Norfolk UK, thorough an enhanced Advice and Guidance programme on ECLIPSE Live. All patients with PMR (n=450) on steroids (n=150) were asked to complete PROMS, issued with a digital steroid card and given access to the patient-engagement portal. Recommendations on steroid tapering were given by partnering rheumatologists, based on PROMs and C-reactive protein (CRP) levels. Within the patient portal, patients could view a graphic display of their average weekly steroid use overlaid with CRP results and have access to exercise information. Suggestions by partnering rheumatologists included advising patient review and referral to secondary care. Conclusion: In many countries, the number of those with PMR are as large as those with rheumatoid arthritis; changes in demographics will mean many with PMR will live in relatively dispersed rural and coastal communities. This pilot demonstrated the benefits of this enhanced support programme, which could be expanded to other common conditions and reduce hospital admissions. GPs found the system easy to use and found the use of Structure Medication Reviews with a graphic display of steroid prescriptions useful. Patients liked the support given to their GPs and improvements to their care. REFERENCES: [1] Chief Medical Officer's Annual Report 2023: Health in an Ageing Society. Acknowledgements: This work was funded by PMRGCAuk. Disclosure of Interests: Max Yates Advisory board for BioGen and Galapagos, Pratyasha Saha: None declared, Clare Aldus: None declared, Catherine Green: None declared, Julian Brown: None declared, Alex MacGregor: None declared.

Objective The Broccoli in Osteoarthritis (BRIO Study) was conducted to determine whether dietary sulforaphane (SFN), consumed as broccoli, improves pain and/or physical function in participants with knee osteoarthritis (OA). This was a proof of principle study to test the feasibility of the trial to optimise the design of an appropriately powered study. Design Two-centre, double-blind, two-arm parallel, randomised placebo-controlled, dietary intervention feasibility trial. Patients with radiographic knee osteoarthritis (Kellgren-Lawrence score 2-3), with pain of at least 4 on a scale of 0-10 were recruited. The intervention was a high glucoraphanin broccoli, (source of SFN), or a matched placebo (no SFN) soup. Pain and measures of physical function were measured at baseline, 6 and 12 weeks. Results The mean WOMAC pain score (scale 0 - 20) was decreased by 4.2 (95% CI: 1.03,7.38) following intervention, Similar patterns of improvement were observed for other pain and function outcome measures. Study data, sample collections and intervention adherence were 100% compliant except where COVID restrictions applied. Acceptability for randomisation was 100% and acceptability for the intervention was 92%. There were three related adverse events, two of which were expected. Conclusions High glucosinolate broccoli soup is a novel approach to managing OA that is widely accessible and can be used on a large scale. This study shows that it is an acceptable way of delivering dietary bioactives and has potential for therapeutic benefit. The primary outcome of pain improved in the intervention group compared to the placebo and the confidence interval encompassed the minimal clinically important difference. The data provide justification for proceeding to a large scale, appropriately powered intervention trial.### Competing Interest StatementDisclosure of interest: Rose Davidson: None declared, Laura Watts: None declared, Gemma Beasy: None declared, Shikha Saha: None declared, Paul Kroon: None declared, Aedin Cassidy: None declared, Allan Clark: None declared, William Fraser Speakers bureau: Roche, Incstar/Diasorin, IDS, Sanofi, Siemens, Menarini, Abbott, Entera Bio, NPS pharmaceuticals and Alexis, Consultant of: Roche, Incstar/Diasorin, IDS, Sanofi, Siemens, Menarini, Abbott, Entera Bio, NPS pharmaceuticals and Alexis, Grant/research support from: Roche, Incstar/Diasorin, IDS, Sanofi, Siemens, Menarini, Abbott, Entera Bio, NPS pharmaceuticals and Alexis., Iain McNamara: None declared, Sarah Kingsbury: None declared, Philip G Conaghan Speakers bureau: AbbVie, Novartis, Consultant of: AbbVie, AstraZeneca, Biosplice, BMS, Eli Lilly, Galapagos, Genascence, GSK, Janssen, Merck, Novartis, Pfizer, Regeneron, Stryker, and UCB, Ian Clark: None declared, Alex MacGregor: None declared### Clinical TrialISRCTN 11629849, CPMS 40910, ClinicalTrials.gov [NCT03878368][1]### Funding StatementThis work was supported by grants from Versus Arthritis (Ref: 21772) and Action Arthritis.### Author DeclarationsI confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.YesThe details of the IRB/oversight body that provided approval or exemption for the research described are given below:The procedures followed were in accordance with the ethical standards of NHS Health Research Authority and Health and Care Research Wales (HCRW) and with the Helsinki Declaration of 1975, as revised in 2000. Ethical approval for the study was granted by East of England - Cambridge East Research Ethics Committee (ref 19/EE/0007), IRAS: 250371. All patients gave their full informed and written consent to participate in the study. All patient data was anonymised.I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.YesAll data produced in the present study are available upon reasonable request to the authors [1]: /lookup/external-ref?link_type=CLINTRIALGOV&access_num=NCT03878368&atom=%2Fmedrxiv%2Fearly%2F2024%2F06%2F21%2F2024.06.20.24309233.atom

Background The relationship between multiple long term conditions (MLTC) and rheumatoid arthritis (RA) is well known; however the trajectory of accumulation of MLTC and the factors that influence the rate of progression of MLTC following the onset of RA are yet to be fully established. The Norfolk Arthritis Register (NOAR) is a cohort study established in 1989 to record incident cases of early inflammatory arthritis, and is unique in its long follow-up period of up to 20 years. Here we examine the pattern of MLTC progression in the cohort, and apply a multistate model to examine the factors that influence MLTC progression. Objectives: •Describe the patten of MLTC progression among newly diagnosed cases with RA over the disease course.•Develop and apply a multistate model to examine the rate of accumulation of MLTC among people with RA following disease onset.•Examine predictors that influence the rate of MLTC at different timepoints in the disease course. Methods The Norfolk Arthritis Register (NOAR) is an inception cohort of early inflammatory arthritis established in 1989, with over 4500 cases of new onset inflammatory arthritis. Cases were recruited in primary care or from hospital clinics, and inclusion criteria were age >16 years with 2 swollen joints lasting ≥4 weeks. The analysis focused on 13 self-reported chronic diseases involving multiple organ systems ascertained at each follow up point over an interval of 20 years. A multistate model was constructed to examine the accumulation of the number of MLTCs over time (‘states‘) since the onset of disease. Death was included as a separate state. Demographic data at baseline and follow up was included in the model. The multistate model was implemented in the R package msm, which provides longitudinal information on the transfer rates from the lower to higher levels of MLTC-M and to death. Multivariate multiple imputation (nburn=1000, 5 datasets) was used to impute missing values using R packages mice and jomo. Results The analysis is based on 4674 patients observed 2 to 15 times; this provided 33,356 records in total. The number of comorbidities increases over time from onset of RA, rho=0.299 (p-value < 2.2e-16). This increase was more pronounced than the increase by age, rho=0.159 (p-value < 2.2e-16), suggesting that inflammation drives number of comorbidities more than age. There were 269 unique combinations of MLTC. The most frequent single morbidities were hypertension, depression and lung disease, with frequencies 0.325, 0.320, and 0.167, respectively (0.812 total probability). The multistate model (Figure 1) considered six defined MLTC states corresponding to total number of MLTC conditions: ranging from State-1 representing RA alone to State-6 representing death. The best fitting model proved to be an additive model including the 5 predictors of gender, age of onset, BMI, smoking status and logDMARDs. Males had significantly higher hazards of accumulation of MLTCs and mortality risk. The total number of MLTCs alone were not associated with the hazard of death. Being overweight significantly increases hazards of progression from State-2 to State-3, and 3 to 4, but also did not affect hazards of death. Ex and non-smokers had significantly lower hazards of accumulation of MLTCs and death, and the number of DMARDs reduced disease progression from State-2 to State-3, and 3 to 4, but did not affect mortality. Conclusion This study provides a comprehensive long-term picture of the rate and pattern of accumulation of multimorbidity in RA over the disease course. MLTC involved most patients with RA over time. Smoking, obesity and the use of disease modifying drugs were all found to influence the rate of MLTC accumulation across a range of organ systems. The data suggest that sustained control of inflammation in RA, both through drug interventions and lifestyle management is an important part of limiting the risk of MLTC development throughout the disease course. References: NIL. Acknowledgements: NIL. Disclosure of Interests None Declared.Figure 1State transitions (arrows) and transition rates (λ) between 6 states in NOAR

Abstract Background/Aims Recent translational advances in genetics report the ability to accurately predict the diagnosis of patients presenting with synovitis, providing potential to accelerate treatment and improve patient outcomes. One tool is G-PROB, which uses genetic information to calculate conditional probabilities, known as G-probabilities ranging from 0 to 100%, for defined diseases. In the original study, G-PROB was configured to discriminate between patients with rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), psoriatic arthritis (PsA), spondyloarthritis (SpA), gout and “other rheumatological diseases” using reported bias-adjusted odds ratios from 250 known single nucleotide polymorphisms and human leukocyte antigen variants of uncorrelated risk variants. The original study tested G-PROB on 243 patients with synovitis. Our aim was to assess whether G-PROB could aid diagnosis using data from the Norfolk Arthritis Register (NOAR), a large observational cohort of patients with early inflammatory arthritis. Methods Genotypes, and clinician diagnosis were obtained from NOAR. The same prevalence settings and risk variants as the original study were used. Six G-probabilities each corresponding to one disease were created for each patient and performance was assessed using linear regression without intercept, negative and positive predictive values (NPV, PPV), and receiver-operator-curve (ROC) analysis. Results From NOAR, 2031 genotyped patients were identified and underwent case note review to determine the clinician diagnosis. Clinician diagnoses included RA (n = 767), PsA (n = 106), SpA (n = 15), SLE (n = 14), gout (n = 5), and “other” (n = 65). For n = 1059, case notes were not available resulting in exclusion. The mean G-probability was 41% for those which corresponded to clinician-defined disease, which was significantly higher compared to 12% for those that did not (95%CI -0.30 to -0.28). As reported in the original study, G-probabilities were concordant with real disease status (β regression coefficient of 1.03 vs 0.99, where 1.00 is ideal). We found 42% of all G-probabilities were <5%, corresponding to a NPV of 99%, where it was possible to deprioritise >1 disease for 100% of patients, >2 diseases for 96% of patients, and >3 diseases for 69% of patients. We found 17.8% of patients had a single G-probability >50% corresponding to a PPV of 41%. This compared to 45% of patients, and PPV of 64% reported in the original study. Accuracy of G-probabilities to discriminate clinician-defined disease was similar in our cohort (AUC of 0.86 95% CI 0.84-0.87) compared to the original study (AUC 0.84 95% CI 0.81-0.86). In 57% of patients in our cohort, the disease with the highest G-probability corresponded to the clinician-defined disease compared to 53% in the original study. Conclusion We were able to replicate several findings of the original study in a large independent cohort including calibration, high NPV, but PPV was lower, suggesting that G-PROB is most valuable as a tool to rule out diagnoses. Disclosure R.M. Hum: None. S.D. Sharma: None. M. Stadler: None. N. Nair: None. S. Viatte: None. C. Yap: None. J.H. Humphreys: None. A. MacGregor: None. M. Yates: None. M. Soomro: None. S.M. Verstappen: None. P. Ho: None. A. Barton: None. J. Bowes: None.

OBJECTIVES:To compare the magnitude of cognitive impairment against age-expected levels across the immune mediated inflammatory diseases (IMIDs: systemic lupus erythematosus [SLE], rheumatoid arthritis [RA], axial spondyloarthritis [axSpA], psoriatic arthritis [PsA], psoriasis [PsO]).METHODS:A pre-defined search strategy was implemented in Medline, Embase and Psychinfo on 29/05/2021. Inclusion criteria were: (i) observational studies of an IMID, (ii) healthy control comparison, (iii) measuring cognitive ability (overall, memory, complex attention/executive function, language/verbal fluency), and (iv) sufficient data for meta-analysis. Standardised mean differences (SMD) in cognitive assessments between IMIDs and controls were pooled using random-effects meta-analysis. IMIDs were compared using meta-regression.RESULTS:In total, 65 IMID groups were included (SLE: 39, RA: 19, axSpA: 1, PsA: 2 PsO: 4), comprising 3141 people with IMIDs and 9333 controls. People with IMIDs had impairments in overall cognition (SMD: -0.57 [95% CI -0.70, -0.43]), complex attention/executive function (SMD -0.57 [95% CI -0.69, -0.44]), memory (SMD -0.55 [95% CI -0.68, -0.43]) and language/verbal fluency (SMD -0.51 [95% CI -0.68, -0.34]). People with RA and people with SLE had similar magnitudes of cognitive impairment in relation to age-expected levels. People with neuropsychiatric SLE had larger impairment in overall cognition compared with RA.CONCLUSIONS:People with IMIDs have moderate impairments across a range of cognitive domains. People with RA and SLE have similar magnitudes of impairment against their respective age-expected levels, calling for greater recognition of cognitive impairment in both conditions. To further understand cognition in the IMIDs, more large-scale, longitudinal studies are needed.

BackgroundSeveral long-term chronic illnesses are known to be associated with an increased risk of dementia independently, but little is known how combinations or clusters of potentially interacting chronic conditions may influence the risk of developing dementia. Methods447 888 dementia-free participants of the UK Biobank cohort at baseline (2006-2010) were followed-up until 31 May 2020 with a median follow-up duration of 11.3 years to identify incident cases of dementia. Latent class analysis (LCA) was used to identify multimorbidity patterns at baseline and covariate adjusted Cox regression was used to investigate their predictive effects on the risk of developing dementia. Potential effect moderations by C reactive protein (CRP) and Apolipoprotein E (APOE) genotype were assessed via statistical interaction. ResultsLCA identified four multimorbidity clusters representing Mental health, Cardiometabolic, Inflammatory/autoimmune and Cancer-related pathophysiology, respectively. Estimated HRs suggest that multimorbidity clusters dominated by Mental health (HR=2.12, p<0.001, 95% CI 1.88 to 2.39) and Cardiometabolic conditions (2.02, p<0.001, 1.87 to 2.19) have the highest risk of developing dementia. Risk level for the Inflammatory/autoimmune cluster was intermediate (1.56, p<0.001, 1.37 to 1.78) and that for the Cancer cluster was least pronounced (1.36, p<0.001, 1.17 to 1.57). Contrary to expectation, neither CRP nor APOE genotype was found to moderate the effects of multimorbidity clusters on the risk of dementia. ConclusionsEarly identification of older adults at higher risk of accumulating multimorbidity of specific pathophysiology and tailored interventions to prevent or delay the onset of such multimorbidity may help prevention of dementia.

Objective There is growing evidence that genetic data are of benefit in the rheumatology outpatient setting by aiding early diagnosis. A genetic probability tool (G‐PROB) has been developed to aid diagnosis has not yet been tested in a real‐world setting. Our aim was to assess whether G‐PROB could aid diagnosis in the rheumatology outpatient setting using data from the Norfolk Arthritis Register (NOAR), a prospective observational cohort of patients presenting with early inflammatory arthritis. Methods Genotypes and clinician diagnoses were obtained from patients from NOAR. Six G‐probabilities (0%–100%) were created for each patient based on known disease‐associated odds ratios of published genetic risk variants, each corresponding to one disease of rheumatoid arthritis, systemic lupus erythematosus, psoriatic arthritis, spondyloarthropathy, gout, or “other diseases.” Performance of the G‐probabilities compared with clinician diagnosis was assessed. Results We tested G‐PROB on 1,047 patients. Calibration of G‐probabilities with clinician diagnosis was high, with regression coefficients of 1.047, where 1.00 is ideal. G‐probabilities discriminated clinician diagnosis with pooled areas under the curve (95% confidence interval) of 0.85 (0.84–0.86). G‐probabilities <5% corresponded to a negative predictive value of 96.0%, for which it was possible to suggest >2 unlikely diseases for 94% of patients and >3 for 53.7% of patients. G‐probabilities >50% corresponded to a positive predictive value of 70.4%. In 55.7% of patients, the disease with the highest G‐probability corresponded to clinician diagnosis. Conclusion G‐PROB converts complex genetic information into meaningful and interpretable conditional probabilities, which may be especially helpful at eliminating unlikely diagnoses in the rheumatology outpatient setting. image