Systemic Inflammation: The Link Between Atrial Fibrillation and Stroke Outcomes

Introduction Atrial fibrillation (AF) is one of the most prevalent cardiac arrhythmias worldwide and a well-established risk factor for ischemic stroke.1,2 Patients with AF tend to experience more severe neurological deficits, larger infarct volumes, and higher risks of mortality and long-term disability compared with those without AF.3 However, the mechanisms underlying the association between AF and adverse functional outcomes after stroke remain incompletely understood.4 Systemic inflammation is increasingly recognized as a potential mechanistic bridge between AF pathogenesis and post-stroke injury.5 Experimental and clinical studies show that inflammatory mediators such as tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), interleukin-2 (IL-2), interleukin-8 (IL-8), and C-reactive protein (CRP) contribute to atrial structural remodeling, endothelial dysfunction, and thrombogenesis.6–9 Recent multi-omic evidence further highlights immune-cell infiltration and fibroblast activation in AF-related atrial tissue.10 These inflammatory processes extend into the cerebrovascular system, where dysregulated innate and adaptive immunity—particularly neutrophil activation, lymphocyte depletion, oxidative stress, and blood–brain barrier breakdown—worsen cerebral injury and impair recovery after ischemic stroke.11–13 Composite inflammation-based indices, such as neutrophil-to-lymphocyte ratio (NLR), the systemic inflammation response index (SIRI) and the systemic immune-inflammation index (SII) capture multiple dimensions of immunity and have shown strong associations with cardiovascular and cerebrovascular outcomes.11,14–16 These markers reflect the balance between neutrophil-driven innate immunity and lymphocyte-mediated regulation, both of which play critical roles in stroke-related secondary injury.17–19 Emerging inflammation–lipid composite indices, such as the neutrophil-to-total cholesterol ratio (NTR) and the high-density lipoprotein-to-lymphocyte ratio (HLR) have also shown early potential, although their roles are less established.20,21 Despite growing evidence linking AF, inflammation, and stroke prognosis, it remains unclear whether systemic inflammation mediates the relationship between AF and post-stroke functional outcomes. Prior studies have primarily examined inflammatory biomarkers as predictors rather than mechanistic intermediates, and few studies have formally quantified indirect pathways using mediation analysis.22 Causal mediation analysis, combined with modern variable-selection approaches such as the least absolute shrinkage and selection operator (LASSO), offers a robust framework to identify candidate mediators and quantify indirect effects in multivariable contexts.23 Sensitivity tools such as E-values further enable the assessment of robustness to potential unmeasured confounding.24 Therefore, this study aimed to evaluate whether systemic inflammatory markers mediate the association between AF and 90-day functional outcomes after acute ischemic stroke (AIS). We focused on established inflammatory indices such as NLR and SIRI, while also exploring emerging inflammation–lipid markers. By applying LASSO-guided mediator selection and formal causal mediation analysis, this study seeks to quantify the contribution of systemic inflammation to AF-related functional impairment and identify clinically informative biomarkers in AF-associated ischemic stroke. Methods Study Design and Participants This retrospective cohort study included consecutive adult patients admitted with AIS to the Stroke Center of the First Affiliated Hospital of Soochow University between September 2023 and April 2025. All patients received standard-of-care acute stroke management, including reperfusion therapies (intravenous thrombolysis or mechanical thrombectomy) when clinically indicated, according to national stroke guidelines. The distribution of reperfusion therapies was similar between AF and non-AF groups, and for patients who underwent thrombolysis or thrombectomy, blood samples were obtained before the procedure. Blood samples for inflammatory markers were collected under fasting conditions in the early morning of the second hospital day and analyzed in a single certified hospital laboratory. AF, including both previously diagnosed AF and AF newly detected during hospitalization, was confirmed by 12-lead electrocardiography or ≥24-hour continuous cardiac monitoring. Functional outcomes were assessed at 90 days using the modified Rankin Scale (mRS), with poor outcome defined as mRS >2.25 Among the screened patients, those with pre-stroke disability (mRS >2), incomplete clinical records, active malignancy, autoimmune disorders, or documented acute infection at admission were excluded. Among 500 screened patients, 37 met the exclusion criteria (23 with incomplete data, 9 with acute systemic inflammation, and 5 with malignancy or other systemic diseases), resulting in 463 patients included in the final analysis. Clinical and laboratory data were extracted from electronic medical records. Inflammatory Marker Assessment The inflammatory indices were calculated using standard hematologic and biochemical parameters as follows: Additional inflammatory indices were calculated using conventional formulas and are summarized in Supplementary Table S8. Statistical Analysis Continuous variables were examined for distributional properties, and skewed variables were log-transformed and Z-standardized before analysis. Missing data were handled using multiple imputation with chained equations. Candidate variables for further analysis were selected based on two criteria: (1) differences between AF and non-AF groups (p < 0.1), and (2) significant association with poor 90-day functional outcome in univariable logistic regression (p < 0.05). Variables meeting these thresholds were categorized a priori into four groups: must-have confounders, the core variable of interest (AF), candidate mediators, and other associated variables. To identify appropriate mediators while minimizing multicollinearity and overfitting, a two-layer LASSO regression procedure was implemented. In the first layer, only candidate mediators were penalized while confounders were forced into the model. In the second layer, other associated variables were introduced to obtain a more comprehensive model structure. Cross-validated λ was chosen to minimize prediction error and variables with non-zero coefficients were retained for further modeling. Predictive performance of the confounder-only model and full model (must-have confounders + AF + LASSO-selected variables) was compared using receiver operating characteristic (ROC) curves, area under the receiver operating characteristic (AUC) curve, calibration metrics, and decision curve analysis. Single-mediator causal mediation analyses were performed using a traditional, frequency-based framework. For each mediator, a linear regression model was fitted with AF and confounders as predictors. The outcome model was specified for the mediator (continuous), and a logistic regression model was used for the outcome (poor 90-day mRS). Mediation effects, including the average causal mediation effect (ACME), average direct effect (ADE), total effect, and proportion mediated, were estimated using 5000 bootstrap simulations. Significance was defined as a 95% confidence interval not crossing zero. To evaluate robustness to potential confounding by baseline stroke severity, a sensitivity analysis excluding the NIHSS score from both mediator and outcome models was conducted. The E-values were calculated to assess the strength of confounding required to explain away the observed effects. Model assumptions, including linearity, normality, and homoscedasticity, were examined using residual diagnostics and Q-Q plots. In the mediation models, the mediator was modeled using linear regression and the outcome using logistic regression with a logit link function, each adjusted for AF and the prespecified confounders. These model structures were consistent across all single-mediator analyses. No correction for multiple testing was applied because the mediators were defined a priori based on biological plausibility. Software and Packages All statistical analyses were conducted in R (version 4.3.2). Causal mediation analyses were conducted using the mediation package, and data manipulation and visualization were performed with dplyr, openxlsx, and ggplot2, along with other relevant packages. Results A total of 463 patients with AIS were included, of whom 96 (20.7%) had AF. Compared with patients without AF, those with AF were older, more often female, and presented with more severe neurological deficits, as reflected by higher baseline NIHSS scores (Table 1 and Supplementary Table S1). Univariable logistic regression identified several variables associated with poor 90-day outcome (Supplementary Table S2), which were subsequently grouped into six must-have confounders, one core variable (AF), sixteen candidate mediators, and forty-seven other associated variables (Supplementary Table S3). The two-layer LASSO regression procedure was applied to identified candidate mediators and additional predictors (Supplementary Tables S4, S5 and Figure 1). In the first layer, high-sensitivity C-reactive protein (hsCRP), NLR, SIRI, and myoglobin were retained, while NTR, HLR, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were not selected. In the second layer, additional variables—including direct bilirubin (DBIL), fasting blood glucose (FBG), and high-density lipoprotein cholesterol (HDL-C) were additionally selected, and AF remained selected across models. The final multivariable model (AF + six must-have confounders + LASSO-selected variables) outperformed the confounder-only model (AUC 0.829 vs 0.757), with acceptable calibration and superior net benefit on decision curve analysis (Figure 2a–c and Supplementary Table S6). We next assessed whether the selected inflammatory markers mediated the relationship between AF and poor 90-day functional outcome (Table 2 and Figure 3). In single-mediator causal mediation analyses with 5000 bootstrap simulations, NLR and SIRI demonstrated statistically significant indirect effects, indicating a partial but meaningful inflammatory contribution: NLR: ACME = 0.031 (95% CI 0.007–0.064), proportion mediated 22.3% SIRI: ACME = 0.024 (95% CI 0.002–0.056), proportion mediated 17.0% In contrast, hsCRP and myoglobin did not show statistically significant indirect effects. ADE E-values indicated that direct effects were relatively robust to potential unmeasured confounding (hsCRP 1.520; NLR 1.466; SIRI 1.499; myoglobin 1.538). A sensitivity analysis excluding baseline NIHSS yielded results consistent with the primary mediation analysis (Supplementary Table S7). NLR and SIRI continued to show significant indirect effects, whereas hsCRP and myoglobin showed smaller and statistically non-significant effects. ADE E-values indicated that the direct effects remained robust to potential unmeasured confounding. Model diagnostics supported the assumptions of the causal mediation framework. The residuals-versus-fitted values plot showed no systematic deviations, and Q–Q plots indicated approximate normality of residuals, supporting the validity of the linear mediator models (Supplementary Figure S1). Discussion In this exploratory analysis, we examined whether systemic inflammation mediates the association between AF and poor functional outcomes in AIS. Our findings indicate that both NLR and SIRI partially mediate this relationship, accounting for 22.3% and 17.0% of the total effect, respectively. These findings underscore the contribution of inflammatory activation to AF-related post-stroke disability.7,14,26,27 The validity of these mediation effects was supported by diagnostic evaluations confirming that the assumptions of the linear mediation models were adequately met. The biological mechanisms underlying these associations are increasingly understood. AF can amplify systemic inflammatory responses, contributing to endothelial dysfunction, hypercoagulability, and immune dysregulation, all of which may worsen ischemic injury.28,29 NLR reflects the balance between neutrophil-driven pro-inflammatory activity and lymphocyte-mediated regulatory processes. Elevated neutrophils exacerbate ischemic injury by releasing reactive oxygen species, proteases, and neutrophil extracellular traps, leading to blood-brain barrier disruption and microvascular damage.11,30 Reduced lymphocyte counts further diminish anti-inflammatory and reparative responses, increasing vulnerability to secondary brain injury.29,31 SIRI extends this concept by integrating neutrophil, lymphocyte, and monocyte counts, capturing a broader systemic inflammatory burden that appears to influence stroke recovery in patients with AF.14 These findings provide a plausible mechanistic explanation for our observed mediation effects. While previous studies have reported associations between elevated NLR or SIRI and adverse outcomes in AIS, they did not quantify the extent to which inflammatory dysregulation contributes to the excess risk attributable to AF. Our mediation analysis extends the existing evidence by demonstrating that more than one-fifth of the detrimental impact of AF on 90-day functional outcome is mediated through NLR, highlighting a potentially modifiable inflammatory pathway linking AF and stroke severity. Interestingly, other inflammatory markers, including hsCRP and myoglobin, did not show significant mediation effects, despite their known associations with stroke and inflammation.32,33 This may reflect differences in specificity and timing. hsCRP often rises later than NLR or SIRI and is influenced by diverse systemic inflammatory sources, whereas myoglobin lacks central nervous system specificity.34–36 Consequently, these markers may be less sensitive for capturing the acute inflammatory processes that mediate AF-related ischemic injury. Our results align with previous work demonstrating that inflammation contributes to adverse outcomes after stroke,37,38 but extend these findings by quantifying the mediating role of specific markers in AF-related stroke. Prior studies have shown that elevated NLR or SIRI is associated with early neurological deterioration or mortality in AIS,7,26 yet the mediating pathway linking AF, inflammation, and functional outcomes had not been directly assessed. By demonstrating that NLR and SIRI capture inflammation-related risk, our study suggests that these markers could aid early prognostic assessment and guide individualized management strategies. In this context, NLR and SIRI could also be incorporated into existing stroke prognostic models and potentially serve as early triggers for anti-inflammatory therapeutic strategies in patients with AF-related ischemic stroke. For example, patients with elevated NLR or SIRI might benefit from more intensive monitoring, early rehabilitation, or potentially targeted anti-inflammatory interventions, though such strategies require validation in prospective trials.39–41 While our study focused on NLR and SIRI due to their clinical availability, reproducibility, and established associations with outcomes, emerging markers such as NTR and HLR may offer additional prognostic information in AIS. Future studies should evaluate whether incorporating these novel markers can further refine risk stratification. Several limitations should be acknowledged. First, the retrospective single-center design limits causal inference and generalizability, and reliance on electronic medical records may introduce selection bias. Second, although a cross-validated penalty was applied during LASSO selection to reduce overfitting, the modest sample size relative to the number of candidate mediators may still result in residual overfitting, potentially affecting the accuracy of estimated mediation effects. This concern is consistent with reports that LASSO can yield overly sparse solutions or retain noise predictors when applied to high-dimensional data with limited sample sizes.42 Third, the selection of biomarkers for analysis was based on predefined indices, but residual bias cannot be excluded, and other unmeasured inflammatory markers may contribute to the association between AF and outcomes. Fourth, missing data were handled using multiple imputation, which assumes that data are missing at random; deviations from this assumption could bias results. Fifth, the absence of external validation limits the generalizability of our findings. Nonetheless, sensitivity analyses—including E-value assessment and alternative model specifications—support the robustness of the observed mediation effects. Given these limitations, prospective multicenter studies with standardized inflammatory profiling and temporally precise mediator measurements are essential to confirm the causal pathways suggested by our mediation analysis.43,44 Looking forward, future research could combine detailed inflammatory profiling—including cytokines such as IL-6 and TNF-α—with neuroimaging and functional assessments to clarify the pathways linking AF, systemic inflammation, and stroke recovery. Randomized trials testing interventions that modulate NLR or SIRI may provide insight into whether reducing systemic inflammation can improve outcomes in high-risk ischemic stroke patients with AF.45,46 In summary, our study suggests that systemic inflammation, reflected by NLR and SIRI, partially mediates the association between AF and poor functional outcomes after ischemic stroke. These widely available inflammatory markers may help identify high-risk patients and inform strategies to improve stroke recovery. Conclusion In conclusion, this exploratory study suggests that NLR and SIRI may partially mediate the association between AF and poor functional outcomes after ischemic stroke. Systemic inflammation may therefore represent a partial pathway linking AF and poor prognosis rather than a definitive causal mechanism. These readily accessible inflammatory markers may assist in identifying high-risk patients and informing individualized management strategies. Nevertheless, the findings should be interpreted with caution, and prospective, multicenter studies are needed to validate these mediation effects and determine whether targeting systemic inflammation can improve outcomes in AF-related ischemic stroke. Data Sharing Statement Data are available in anonymous form upon reasonable request after ethics clearance and approval from the corresponding author. De-identified data supporting the findings of this study are available from the corresponding author upon reasonable request and after obtaining ethics clearance. Ethics Approval and Informed Consent This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of the First Affiliated Hospital of Soochow University (Approval No. 2025838). The requirement for informed consent was waived because the study used de-identified retrospective data. Author Contributions Xinyi He: Data curation, Formal analysis, Visualization, Writing – original draft. Haixing Xiao: Formal analysis, Writing – original draft. Qi Fang: Methodology, Validation, Writing – review & editing, Supervision, Funding acquisition. Xiang Tang: Methodology, Funding acquisition, Writing – review & editing. All authors took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Funding This work was supported by Suzhou Basic Research Pilot Project (no. SSD2024058, Xiang Tang), the National Natural Science Foundation of China (no. 82001125, Xiang Tang). Disclosure The authors report no conflicts of interest in this work. References 1. Rydén L, Sacuiu S, Wetterberg H, et al. Atrial fibrillation, stroke, and silent cerebrovascular disease: a population-based MRI study. Neurology. 2021;97(16):e1608–e1619. doi:10.1212/wnl.0000000000012675 2. Vinding NE, Kristensen SL, Rørth R, et al. Ischemic stroke severity and mortality in patients with and without atrial fibrillation. J Am Heart Assoc. 2022;11(4):e022638. doi:10.1161/jaha.121.022638 3. Elsheikh S, Hill A, Irving G, Lip GYH, Abdul-Rahim AH. Atrial fibrillation and stroke: state-of-the-art and future directions. Curr Probl Cardiol. 2024;49(1 Pt C):102181. doi:10.1016/j.cpcardiol.2023.102181 4. Lip GYH, Proietti M, Potpara T, et al. Atrial fibrillation and stroke prevention: 25 years of research at EP Europace journal. Europace. 2023;25(9):euad226. doi:10.1093/europace/euad226 5. Yang X, Zhao S, Wang S, et al. Systemic inflammation indicators and risk of incident arrhythmias in 478,524 individuals: evidence from the UK Biobank cohort. BMC Med. 2023;21(1):76. doi:10.1186/s12916-023-02770-5 6. Hijazi Z, Oldgren J, Siegbahn A, Granger CB, Wallentin L. Biomarkers in atrial fibrillation: a clinical review. Eur Heart J. 2013;34(20):1475–1480. doi:10.1093/eurheartj/eht024 7. Guo Y, Lip GY, Apostolakis S. Inflammation in atrial fibrillation. J Am Coll Cardiol. 2012;60(22):2263–2270. doi:10.1016/j.jacc.2012.04.063 8. Hu YF, Chen YJ, Lin YJ, Chen SA. Inflammation and the pathogenesis of atrial fibrillation. Nat Rev Cardiol. 2015;12(4):230–243. doi:10.1038/nrcardio.2015.2 9. Hu J, Wang JJ, Wang L, Cao YH, Wu JC. Combined predictive value of neutrophil-to-lymphocyte ratio and CHA2DS2-VASc score for cardiogenic cerebral embolism in NVAF patients. Int J Gen Med. 2025;18:3489–3499. doi:10.2147/ijgm.S523480 10. Sheng Y, Wang YY, Chang Y, et al. Deciphering mechanisms of cardiomyocytes and non-cardiomyocyte transformation in myocardial remodeling of permanent atrial fibrillation. J Adv Res. 2024;61:101–117. doi:10.1016/j.jare.2023.09.012 11. Gong P, Liu Y, Gong Y, et al. The association of neutrophil to lymphocyte ratio, platelet to lymphocyte ratio, and lymphocyte to monocyte ratio with post-thrombolysis early neurological outcomes in patients with acute ischemic stroke. J Neuroinflammation. 2021;18(1):51. doi:10.1186/s12974-021-02090-6 12. Anrather J, Iadecola C. Inflammation and Stroke: an Overview. Neurotherapeutics. 2016;13(4):661–670. doi:10.1007/s13311-016-0483-x 13. Alsbrook DL, Di Napoli M, Bhatia K, et al. Neuroinflammation in acute ischemic and hemorrhagic stroke. Curr Neurol Neurosci Rep. 2023;23(8):407–431. doi:10.1007/s11910-023-01282-2 14. Wang Y, Zhang Z, Hang X, Wang W. Associations of inflammatory markers with neurological dysfunction and prognosis in patients with progressive stroke. Eur J Neurol. 2025;32(2):e70080. doi:10.1111/ene.70080 15. Eyiol A, Ertekin B. Association of red blood cell distribution width to albumin ratio with prognosis in stroke patients. Biomarker Med. 2024;18(7):311–320. doi:10.2217/bmm-2023-0460 16. Eyiol A, Ertekin B. The relationship between hemoglobin-to-red cell distribution width (RDW) ratio (HRR) and mortality in stroke patients. Eur Rev Med Pharmacol Sci. 2024;28(4):1504–1512. doi:10.26355/eurrev_202402_35480 17. Li LM, Cai H. Neutrophil-related immune-inflammatory biomarkers influence the early progression of medial medullary infarction. Front Neurol. 2025;16:1528560. doi:10.3389/fneur.2025.1528560 18. Graça SC, Mosca T, Gagliardi VDB, Forte WCN, Gagliardi RJ. Prognostic impact of neutrophil-to-lymphocyte ratio in ischemic stroke. J Pers Med. 2024;14(12):1149. doi:10.3390/jpm14121149 19. Tian T, Wang L, Xu J, et al. Prediction of early neurological deterioration in acute ischemic stroke patients treated with intravenous thrombolysis. J Cereb Blood Flow Metab. 2023;43(12):2049–2059. doi:10.1177/0271678x231200117 20. Liu SL, Feng BY, Song QR, Zhang YM, Wu SL, Cai J. Neutrophil to high-density lipoprotein cholesterol ratio predicts adverse cardiovascular outcomes in subjects with pre-diabetes: a large cohort study from China. Lipids Health Dis. 2022;21(1):86. doi:10.1186/s12944-022-01695-x 21. Xu Q, Wu Q, Chen L, et al. Monocyte to high-density lipoprotein ratio predicts clinical outcomes after acute ischemic stroke or transient ischemic attack. Cns Neurosci Ther. 2023;29(7):1953–1964. doi:10.1111/cns.14152 22. Jin A, Wang S, Li J, et al. Mediation of systemic inflammation on insulin resistance and prognosis of nondiabetic patients with ischemic stroke. Stroke. 2023;54(3):759–769. doi:10.1161/strokeaha.122.039542 23. Wu P, Wang Y, Ye Y, et al. Liver biomarkers, lipid metabolites, and risk of gestational diabetes mellitus in a prospective study among Chinese pregnant women. BMC Med. 2023;21(1):150. doi:10.1186/s12916-023-02818-6 24. VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the E-value. Ann Intern Med. 2017;167(4):268–274. doi:10.7326/m16-2607 25. Banks JL, Marotta CA. Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke. 2007;38(3):1091–1096. doi:10.1161/01.STR.0000258355.23810.c6 26. Zhang A, Zhu Y, Liao J, et al. The association of systemic inflammatory response index and neutrophil-to-high-density lipoprotein ratio mediated by fasting blood glucose with 90-day prognosis in acute ischemic stroke patients. Neuroepidemiology. 2025;59(1):31–42. doi:10.1159/000539132 27. Qin P, Ho FK, Celis-Morales CA, Pell JP. Association between systemic inflammation biomarkers and incident cardiovascular disease in 423,701 individuals: evidence from the UK biobank cohort. Cardiovasc diabetol. 2025;24(1):162. doi:10.1186/s12933-025-02721-9 28. Rivard L, Friberg L, Conen D, et al. Atrial fibrillation and dementia: a report from the AF-SCREEN International Collaboration. Circulation. 2022;145(5):392–409. doi:10.1161/circulationaha.121.055018 29. Luo J, Thomassen JQ, Nordestgaard BG, Tybjærg-Hansen A, Frikke-Schmidt R. Neutrophil counts and cardiovascular disease. Eur Heart J. 2023;44(47):4953–4964. doi:10.1093/eurheartj/ehad649 30. Guan X, Zhu S, Song J, et al. Microglial CMPK2 promotes neuroinflammation and brain injury after ischemic stroke. Cell Rep Med. 2024;5(5):101522. doi:10.1016/j.xcrm.2024.101522 31. Zhao Y, Li Q, Niu J, et al. Neutrophil membrane-camouflaged polyprodrug nanomedicine for inflammation suppression in ischemic stroke therapy. Adv Mater. 2024;36(21):e2311803. doi:10.1002/adma.202311803 32. Chang A, Ricci B, Grory BM, et al. Cardiac biomarkers predict large vessel occlusion in patients with ischemic stroke. J Stroke Cerebrovasc Dis. 2019;28(6):1726–1731. doi:10.1016/j.jstrokecerebrovasdis.2019.02.013 33. Cui C, Liu L, Qi Y, et al. Joint association of TyG index and high sensitivity C-reactive protein with cardiovascular disease: a national cohort study. Cardiovasc diabetol. 2024;23(1):156. doi:10.1186/s12933-024-02244-9 34. Tsai NW, Lee LH, Huang CR, et al. Statin therapy reduces oxidized low density lipoprotein level, a risk factor for stroke outcome. Crit Care. 2014;18(1):R16. doi:10.1186/cc13695 35. Donohue MM, Cain K, Zierath D, Shibata D, Tanzi PM, Becker KJ. Higher plasma fractalkine is associated with better 6-month outcome from ischemic stroke. Stroke. 2012;43(9):2300–2306. doi:10.1161/strokeaha.112.657411 36. Chen J, Wu Q, Liu H, et al. Predictive value of remnant cholesterol inflammatory index for stroke risk: evidence from the China health and Retirement Longitudinal study. J Adv Res. 2025;76:543–552. doi:10.1016/j.jare.2024.12.015 37. Wang X, Pan Y, Zhang R, et al. Inflammation and adverse outcomes in patients with acute ischemic stroke with and without chronic kidney disease. J Am Heart Assoc. 2024;13(9):e033450. doi:10.1161/jaha.123.033450 38. Ma F, Li L, Xu L, et al. The relationship between systemic inflammation index, systemic immune-inflammatory index, and inflammatory prognostic index and 90-day outcomes in acute ischemic stroke patients treated with intravenous thrombolysis. J Neuroinflammation. 2023;20(1):220. doi:10.1186/s12974-023-02890-y 39. Weiss A, Ding Y. Inflammatory markers and stroke outcomes: lessons from neutrophil counts and antiplatelet therapy. J Am Heart Assoc. 2025;14(14):e043158. doi:10.1161/jaha.125.043158 40. Gurol ME, Wright CB, Janis S, et al. Stroke prevention in atrial fibrillation: our current failures and required research. Stroke. 2024;55(1):214–225. doi:10.1161/strokeaha.123.040447 41. Eyiol A. The relationship of pan-immune-inflammation value (PIV) and HALP score with prognosis in patients with atrial fibrillation. Medicine. 2024;103(36):e39643. doi:10.1097/md.0000000000039643 42. Van lissa CJ, van Erp S, Clapper EB. Selecting relevant moderators with Bayesian regularized meta-regression. Res Synth Methods. 2023;14(2):301–322. doi:10.1002/jrsm.1628 43. Cheng H, Song S, Tang Y, et al. Does ICU admission dysphagia independently contribute to delirium risk in ischemic stroke patients? Results from a cohort study. BMC Psychiatry. 2024;24(1):65. doi:10.1186/s12888-024-05520-w 44. Lindmark A, Norrving B, Eriksson M. Socioeconomic status and survival after stroke - using mediation and sensitivity analyses to assess the effect of stroke severity and unmeasured confounding. BMC Public Health. 2020;20(1):554. doi:10.1186/s12889-020-08629-1 45. Huang X, Li C, Zeng P, et al. Non-traditional lipid-inflammatory parameters estimate the risk of stroke in middle-aged and older Chinese adults: a nationwide prospective cohort study. J Adv Res. 2025. doi:10.1016/j.jare.2025.11.029 46. Li J, Lin J, Pan Y, et al. Interleukin-6 and YKL-40 predicted recurrent stroke after ischemic stroke or TIA: analysis of 6 inflammation biomarkers in a prospective cohort study. J Neuroinflammation. 2022;19(1):131. doi:10.1186/s12974-022-02467-1