Inflammatory rheumatic diseases (IRDs) represent a significant risk factor for cerebrovascular events, independent of traditional cardiovascular risk factors. The elevated risk of stroke in individuals with IRDs arises from chronic systemic inflammation, endothelial dysfunction, accelerated atherosclerosis, prothrombotic effects of antiphospholipid antibodies, and cumulative corticosteroid exposure. Post-stroke rehabilitation in patients with IRDs involves distinct clinical challenges that extend beyond standard neurorehabilitation, including joint involvement, chronic pain, fatigue, and treatment-related limitations. Assessment tools that focus exclusively on neurological recovery are inadequate for this population. Dual-layered protocols that integrate both neurological and rheumatological criteria are required. Although wearable sensors, telemonitoring, and electronic patient-reported outcome measures provide a foundation for continuous monitoring, barriers such as limited digital literacy, insufficient infrastructure, and the absence of IRD-specific content addressing symptom overlap persist. Post-stroke cognitive impairment is frequent and multifactorial in this group, necessitating cognitive assessment approaches tailored to IRD subgroups, as disease-specific manifestations may confound the interpretation of standard screening tools. Physiotherapy, occupational therapy, telerehabilitation, and virtual reality-based interventions may be beneficial when specifically adapted to accommodate joint limitations and pain. While artificial intelligence-based tools hold promise, there remains a need for clinically validated, domain-specific expert platforms. Nurses are integral to this process, providing patient education, monitoring medication adherence, offering psychosocial support, and coordinating remote care. Optimal management of stroke associated with IRDs requires multidisciplinary, technology-enabled, and patient-centered care models that integrate neurology, rheumatology, rehabilitation medicine, and digital health. Given the predominantly indirect nature of current evidence, prospective studies employing validated assessment tools are urgently needed.

To investigate associations between (human leukocyte antigen) HLA alleles and overall lupus nephritis (LN) susceptibility, biopsy-defined histologic classes and reduced renal function in Taiwanese patients with systemic lupus erythematosus (SLE). We conducted a retrospective case-control study using data from the Taiwan Precision Medicine Initiative. The matched cohort included 388 SLE patients with biopsy-confirmed LN and 381 SLE patients without LN. HLA alleles were imputed from single nucleotide polymorphism (SNP) array data using HIBAG. Multivariable logistic regression was employed to evaluate associations of candidate HLA alleles with LN overall, individual LN histologic classes and reduced renal function within LN, adjusting for age, gender, hypertension and diabetes mellitus. The study included 388 biopsy-confirmed LN cases and 381 matched SLE controls without LN. Four alleles were independently associated with overall LN susceptibility, including HLA-B*13:01 (adjusted OR 1.68, 95% CI 1.09-2.59), HLA-B*58:01 (1.69, 1.15-2.49), HLA-C*03:02 (1.75, 1.19-2.58) and HLA-DRB1*03:01 (1.51, 1.01-2.27). Class-specific analyses showed that HLA-B*13:01 was associated with Class I/II LN; HLA-B*58:01 and HLA-C*03:02 with Class III ± V LN; and HLA-B*13:01, HLA-B*58:01, and HLA-C*03:02 with Class IV ± V LN. No allele remained independently associated with Class V LN after full adjustment. Within the LN cohort, HLA-B*13:01 was the only allele associated with reduced renal function (eGFR < 60 mL/min/1.73m²; adjusted OR 2.24, 95% CI 1.30-3.85). Overall, the strongest associations were observed for HLA class I alleles, particularly HLA-B*13:01 in this study. In Taiwanese patients with SLE, a narrower set of predominantly HLA class I alleles is associated with LN susceptibility and proliferative LN phenotypes. HLA-B*13:01 showed the most consistent association across overall LN susceptibility, histologic severity and reduced renal function.

Rheumatoid arthritis (RA) is a chronic, systemic inflammatory disease that impacts quality of life and functionality. Its management involves pharmacological and nonpharmacological strategies, including physical exercise and health education. Although patient education is essential, there is no consensus regarding the specific knowledge needs of individuals with RA. This study aimed to assess the health education needs of Brazilian individuals with RA using a mixed cross-sectional observational design and an online questionnaire. Adults over 18 years with a self-reported RA diagnosis were recruited via social media. The questionnaire collected sociodemographic data and information on educational needs. Quantitative data were analyzed descriptively using Microsoft Excel. A subgroup analysis explored the association between years since diagnosis and educational needs through Spearman's correlation and simple linear regression in SPSS, after confirming non-normal distribution (p < 0.05), with a significance level of 5%. Qualitative responses were analyzed using thematic content analysis based on Bardin's method, supported by Braun and Clarke's inductive approach. Results indicate that individuals with RA consider it important to understand basic disease information, biopsychosocial aspects, and topics such as diet, treatment options, patient rights, pain, and physical exercise. A very weak inverse correlation was observed between years since diagnosis and educational needs (ρ = -0.160; p = 0.027), while regression analysis showed no significant predictive effect (β = -0.108; p = 0.137), with low explanatory power (R² = 0.012). The study highlights the actual educational needs of this population and identifies often-overlooked topics, providing a foundation for developing tailored educational materials.

Increasing physical activity (PA) is an important recommendation for the management of hip and knee osteoarthritis (OA), under which components of therapeutic exercise can be embedded. However, maintaining PA behaviour (≥ 6-months) following the cessation of structured exercise interventions remains challenging and under-explored. This mixed-methods systematic literature review aimed to identify the barriers to and facilitators of maintaining PA behaviour for those living with hip and/or knee OA. Electronic databases were searched for peer-reviewed studies examining PA behaviour maintenance (defined as at least six-months post-intervention) for those with hip and/or knee OA. Data were extracted, appraised using the Mixed-methods Appraisal Tool (MMAT), and synthesised narratively. Sixteen studies (n = 1,486; 73.7% female) were included: six quantitative, five qualitative and five mixed-methods. The most frequently cited TDF domains were beliefs about capabilities (e.g. confidence reinforced by intervention skills), environmental context and resources (e.g. competing priorities), and memory, attention and decisional processes (e.g. memory recall - forgetting). Quantitative evidence supported these domains highlighting that correlates such as self-efficacy, energy levels, social support and resource availability strongly influenced PA behaviour maintenance. Integrated findings informed the Physical Activity Maintenance for osteoArthritis (PAMA) conceptual map, illustrating the dynamic interactions between individual, social and environmental factors. This review revealed that maintaining PA behaviour is multifactorial and influenced by individual, social and environmental factors. Further research should consider these interactions in the design of future interventions and in healthcare practice to foster and support the maintenance of PA behaviour for those living with hip and knee OA. PROSPERO registration: CRD42024591820 .

Referral delays and errors in inflammatory rheumatic diseases (IRDs) are common. RhePort is a German online self-referral tool that estimates IRD probability based on a structured questionnaire and expert-derived algorithm. This study evaluated whether machine learning could improve IRD discrimination of RhePort. In this retrospective internal validation study, 1,333 unique RhePort questionnaires linked to rheumatologist-confirmed diagnoses were analyzed. A shared stratified five-fold cross-validation design was used for Logistic Regression, Neural Network, XGBoost, and LightGBM models. For each algorithm, feature selection and Bayesian hyperparameter optimization were performed, followed by evaluation of ensemble strategies. Discrimination, calibration, clinically relevant operating points, and SHAP-based interpretability were assessed against the original RhePort score. IRD prevalence was 35.6% (475/1,333). LightGBM achieved the strongest individual performance, with an area under the receiver operating characteristic curve of 0.791 using 30 features. A weighted ensemble of LightGBM and Logistic Regression further improved performance to 0.815 and achieved the lowest Brier score (0.166). At 95% sensitivity, specificity increased from 9% for RhePort to 32%; at 90% sensitivity, specificity increased from 15% to 48%. The most influential predictors included C-reactive protein, painful small joints, dactylitis pattern, sex, and erythrocyte sedimentation rate. Ensemble AUC-ROC was 0.828 among 746 patients with CRP and/or ESR information and 0.786 among 587 patients without these laboratory data. Machine learning improved IRD classification over the current expert-derived RhePort score and identified questionnaire items that may be omitted without loss of performance, demonstrating the potential of machine learning to enhance rheumatology referral.

Exercise perceptions influence exercise behavior in patients with rheumatoid arthritis (RA). This study aimed to investigate factors associated with exercise perception in women with RA. A total of 104 patients (mean age:57 years; DAS-28 score:2.7 ± 0.8) were included. Exercise perception was assessed using the Exercise Benefits and Barriers Scale (EBBS), evaluating perceived exercise benefits (PEBE) and barriers (PEBA). Physical activity (PA) was measured with the International Physical Activity Questionnaire. Exercise self-efficacy, fatigue, body awareness, social support, quality of life, and beliefs about medicines were assessed using the Exercise Self-Efficacy Scale, Bristol Rheumatoid Arthritis Fatigue Multidimensional Questionnaire, Body Awareness Questionnaire, MOS-Social Support Survey, Health Assessment Questionnaire, and Beliefs about Medicines Questionnaire. Correlation analysis with adjustments and multiple linear regression analysis were performed. Most participants were married (85.6%), had a low education level (71.1%), and were in RA remission (67.3%). PEBE and general exercise beliefs were negatively associated with fatigue (r=-0.396 and - 0.425). PEBA was negatively associated with PA level, exercise self-efficacy, and BA (r=-0.448,-0.486, and - 0.561, respectively). Additionally, higher PEBE scores was independently associated with greater PA levels and exercise self-efficacy, whereas higher PEBA was associated with lower body awareness and greater fatigue in women with RA. Higher fatigue was associated with poorer exercise benefit perceptions and greater perceived exercise barriers. Women with higher perceived barriers had lower PA levels, lower exercise self-efficacy, and poorer body awareness. Evaluating exercise perceptions may help guide rehabilitation planning for women with RA.

Assessing disease activity in psoriatic arthritis (PsA) remains challenging, and additional biomarkers that can complement conventional inflammatory markers are still needed. The CRP-albumin-lymphocyte (CALLY) index integrates inflammatory, immune, and nutritional components into a single measure. This study examined the relationship between the CALLY index and disease activity in patients with PsA. This retrospective longitudinal cohort study included 150 patients with PsA and 50 age- and sex-matched healthy controls. Blood-derived inflammatory indices, including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), platelet-to-lymphocyte ratio (PLR), systemic immune-inflammation index (SII), pan-immune inflammation value (PIV), and the CALLY index, were calculated using routine laboratory data. Disease activity was evaluated with the clinical Disease Activity Index for Psoriatic Arthritis (cDAPSA). Correlation analyses, receiver operating characteristic (ROC) analyses, subgroup analyses, and multivariable logistic regression models were performed. Patients with moderate-to-high disease activity had markedly lower CALLY index values than those with remission or low disease activity. The CALLY index showed a moderate inverse correlation with cDAPSA scores (r = - 0.529, p < 0.001). During follow-up, CALLY values increased significantly in both csDMARD-treated and biologic-treated patients (both p < 0.001), whereas no significant differences were observed between treatment groups. In multivariable analysis, lower log-transformed CALLY index values remained independently associated with moderate-to-high disease activity (OR 0.39, 95% CI 0.23-0.66). The discriminative performance of the CALLY index was similar to that of CRP (AUC 0.724 vs. 0.738; p = 0.42). Lower CALLY index values were linked to greater disease activity in PsA and improved alongside reductions in inflammatory burden during follow-up. Although its performance was comparable to CRP rather than superior, the CALLY index may represent a useful complementary biomarker for disease activity assessment in PsA. Prospective studies are needed to further clarify its clinical utility.

Artificial intelligence (AI) is increasingly used in clinical medicine. Rheumatology is well-suited to AI applications due to diagnostic complexity of rheumatic diseases, variable disease presentations, and multisystem involvement. This cross-sectional study examines global AI research in rheumatology through bibliometric analysis of Scopus data. The Scopus database was searched on May 5, 2026, using the terms "artificial intelligence" AND "rheum*" in the title, abstract, and keyword fields. No temporal restrictions were applied, and all English documents were analyzed. Bibliometric data, including publication year, country, institution, author, journal, and keywords, were extracted. Retracted articles were identified using Scopus tags and were manually verified through related journal editorial notices. Disease-specific publication counts were determined through independent searches for rheumatic diseases. Survey-based studies were identified through manual review. Linear regression analysis was conducted to assess temporal trends. A total of 1,057 publications were identified. Annual output remained below 10 until 2018, then rose sharply to a peak of 282 in 2025. Linear regression confirmed a significant upward trend (p < 0.001). The United States produced the most publications (n = 249), followed by the United Kingdom (n = 141) and India (n = 138). Harvard Medical School was the leading institution (n = 33), and Knitza, J., was the most prolific author (n = 17). Rheumatology International published the most articles (n = 27). Osteoarthritis (n = 1,017) and rheumatoid arthritis (n = 646) dominated the field, while systemic vasculitis (n = 40), Behçet disease (n = 29), and familial Mediterranean fever (n = 16) were underrepresented. Eleven survey-based studies were identified, mostly published between 2024 and 2025. Two articles were retracted due to peer review irregularities. AI research in rheumatology has increased significantly, with research outputs stemming from a limited number of countries, institutions, and disease categories. Osteoarthritis and rheumatoid arthritis are the most frequently explored diseases in the field. The increasing number of survey-based studies indicates heightened attention to clinician and patient perspectives. Future research should focus on expanding international collaboration, addressing gaps in AI research on underrepresented diseases, and strengthening ethical and methodological standards to facilitate broader AI adoption in rheumatology.

Chronic relapsing aseptic meningoencephalitis is a rare but serious central nervous system (CNS) manifestation of Sjögren's disease (SjD), particularly when refractory to conventional immunosuppression. Cryoglobulinemia occurs in a subset of SjD patients and is associated with systemic vasculitis, but its role in driving refractory CNS inflammation remains poorly defined. A 54‑year‑old woman with SjD, rheumatoid arthritis, and persistent cryoglobulinemia presented with a 5‑year history of recurrent headaches resistant to prednisone, methotrexate, mycophenolate mofetil, and cyclophosphamide. CSF analysis showed protein elevation (peak 2,906mg/L), mild pleocytosis (26 nucleated cells/µL), and elevated opening pressures (150-230 mmH₂O). MRI revealed diffuse white matter hyperintensities, cerebral atrophy, and ventriculomegaly. She received eight intrathecal dexamethasone (IT‑DEX) injections (10mg each), each producing rapid headache relief (Numeric Pain Rating Scale 7‑8 → 1‑2)and progressive CSF protein decline (from 2,906 to 696mg/L). At last follow‑up (March 2026), she remained in remission on prednisone 10mg/day and intravenous cyclophosphamide 0.4g every 4 weeks. A systematic review of 23 reported SjD‑associated recurrent aseptic meningitis cases identified universal CSF pleocytosis and elevated protein, but none reported concurrent cryoglobulinemia or progressive ventriculomegaly. This case expands the clinical spectrum of CNS‑SjD to include cryoglobulinemia‑driven refractory meningoencephalitis with hydrocephalus ex vacuo. IT‑DEX appears to be an effective rescue therapy for compartmentalized CNS inflammation when systemic immunosuppression fails.