Stroke ranks as the second-leading cause of mortality worldwide, frequently accompanied by severe complications during both its acute and recovery stages (1). Post-stroke epilepsy (PSE) affects approximately 5–9% of patients after stroke (2). It is widely recognized that PSE contributes to extended hospitalization durations and escalates mortality rates in cerebrovascular disease patients. Ischemic stroke has been identified as the predominant cause of PSE, accounting for 70–85% of cases associated with stroke-related epilepsy (3). The reported frequency of post-stroke ictal and interictal epileptiform activity after stroke is 7 and 8%, respectively (4). Addressing the profound influence of PSE on patient prognosis necessitates a thorough investigation of its risk factors, which is vital for enabling precise risk prediction and for generating hypotheses that may inform future preventive and therapeutic research (5).

Recent studies have highlighted serum uric acid (SUA) as a biomarker of growing importance in stroke and PSE research (5, 6). SUA, a metabolic byproduct of purine degradation, exhibits contrasting roles within the nervous system. While moderate levels exert antioxidant properties that counter neuronal damage caused by oxidative stress, elevated SUA concentrations may act as pro-oxidants, intensifying inflammatory responses and amplifying neural excitability (7). Previous research has revealed a “U-shaped” dose–response correlation between SUA levels and ischemic stroke risk (8), emphasizing that both exceedingly high and low SUA levels markedly elevate stroke risk (9, 10). This dualistic behavior suggests SUA’s complex involvement in stroke pathophysiology, wherein it provides protective effects at optimal levels but contributes to adverse outcomes at extremes (11). Nonetheless, the association between SUA levels and epilepsy risk remains contentious. Certain studies propose a “L-shaped” relationship, arguing that lower SUA levels heighten epilepsy risk, whereas higher levels seemingly plateau without further impact (12). Supporters of this view underscore the significance of SUA’s antioxidant activity, asserting that insufficient levels heighten vulnerability to oxidative damage, thereby fostering epileptogenesis (13). Contrarily, alternative findings link elevated SUA to an increased likelihood of late-onset seizures post-ischemic stroke, implicating oxidative stress and inflammatory cascades as underlying mechanisms (14, 15). These divergent results imply that SUA’s role in PSE may vary depending on disease stage and individual patient conditions.

In the specific context of ischemic stroke, the interaction between SUA and PSE emerges as particularly intricate and remains insufficiently understood. For instance, low SUA levels may undermine antioxidant defenses, whereas excessive levels could aggravate inflammation and neuronal injury, collectively elevating epilepsy risks (16). This dual role accentuates the need for elucidating SUA’s biological mechanisms in PSE. Accordingly, the present study endeavors to systematically explore the association between SUA levels and PSE in ischemic stroke patients. By analyzing these relationships comprehensively, this study aims to provide foundational evidence supporting the prediction and clinical prevention of PSE.

The present study provides compelling evidence of a U-shaped relationship between SUA levels and the risk of PSE in patients with acute ischemic stroke. Using data from a large cohort, we demonstrated that both low (Q1: <305.3 μmol/L) and high (Q4: >373.9 μmol/L) SUA levels were associated with significantly increased risks of PSE compared to the moderate levels (Q2-Q3: 305.3–373.9 μmol/L). After adjusting for multiple confounders, including demographic factors, stroke severity (NIHSS score), comorbidities, and laboratory markers, the adjusted ORs for the Q1 and Q4 groups were 2.32 (95% CI: 1.79–2.99) and 1.60 (95% CI: 1.26–2.02), respectively. This pattern was confirmed through RCS analyses, which identified an inflection point at 335.3 μmol/L. Moreover, the robustness of these findings was reinforced by PSM and sensitivity analyses. After PSM, the non-linear association remained statistically significant. Notably, a significant interaction between SUA quartiles and sex was observed (P for interaction < 2.2 × 10⁻¹⁶), suggesting that the overall association reflects an average effect and is more pronounced in men (Supplementary Tables S1, S2). Given the close link between SUA and renal function, we adjusted for urea, creatinine, and uremia (Model 3) and performed PSM; however, slight post-matching imbalance in renal biomarkers remained (Supplementary Figure S1), and residual confounding cannot be excluded. Collectively, our results underscore the dual role of SUA as both a potential protective factor and a risk factor for PSE, depending on its concentration.

This non-linear relationship can be mechanistically explained by the dual biological properties of uric acid (16, 25). On one hand, SUA acts as a potent endogenous antioxidant, capable of scavenging reactive oxygen species (ROS) and inhibiting lipid peroxidation (26, 27). This antioxidant capacity is particularly relevant in the context of ischemic stroke, where oxidative stress contributes significantly to neuronal damage and subsequent epileptogenesis (28). Moderate SUA levels may thus confer neuroprotection by mitigating oxidative injury, thereby reducing the likelihood of PSE. On the other hand, at high concentrations, SUA can crystallize and trigger inflammatory pathways, such as the NLRP3 inflammasome, leading to the release of pro-inflammatory cytokines (e.g., IL-1β, IL-18) (29). Chronic inflammation is a known driver of epileptogenesis, as it promotes neuronal hyperexcitability and synaptic reorganization (30–32). Additionally, elevated SUA is often linked to metabolic disorders (e.g., hyperuricemia, gout) and renal impairment (15, 33), which may independently exacerbate stroke outcomes and increase susceptibility to seizures.

Conversely, very low SUA levels may reflect a state of compromised antioxidant defense (34). Inadequate ROS scavenging could leave neurons vulnerable to oxidative damage, facilitating excitotoxic neuronal death and aberrant synaptic plasticity—key processes in epilepsy development (35, 36). The identified inflection point (335.3 μmol/L) is consistent with this dual-role hypothesis, suggesting that the balance between antioxidant and pro-inflammatory milieus may vary across SUA levels.

Clinically, these findings suggest that SUA could serve as a valuable biomarker for PSE risk stratification. Patients with SUA levels outside the range associated with the lowest predicted odds in our data may warrant further investigation in future studies. For instance, low and high SUA levels may reflect different underlying biological profiles (e.g., oxidative stress–related vs. inflammatory/metabolic states). These mechanistic hypotheses require validation in experimental studies and prospective cohorts.

Our results both corroborate and extend the existing literature on SUA and neurological outcomes (37). Several studies have highlighted the neuroprotective role of SUA in stroke recovery. For example, Chamorro et al. (38) reported that higher SUA levels were associated with improved functional outcomes after acute ischemic stroke, attributing this to SUA’s antioxidant properties. Similarly, Dhanesha et al. (39) found that SUA administration reduced infarct size and improved neurological recovery in animal models of stroke. These studies support our observation that moderate SUA levels (Q2-Q3) are protective against PSE.

However, our discovery of increased PSE risk at high SUA levels appears to contradict some prior reports. This discrepancy may stem from differences in outcome measures: while earlier studies focused on short-term functional recovery, our work specifically examined PSE-a long-term complication influenced by chronic inflammation and neuronal hyperexcitability. Importantly, recent evidence has begun to acknowledge the “paradoxical” role of SUA. Zhang et al. (40) reported that high SUA levels were associated with increased risk of post-stroke complications, including epilepsy. Similarly, Weir et al. (41) identified SUA as a predictor of poor outcomes in stroke patients, particularly among those with pre-existing comorbidities. Wang et al. demonstrated that aberrant serum uric acid levels, whether excessively elevated or markedly reduced, are associated with an increased risk of epilepsy secondary to cerebral infarction, supporting SUA as a clinically relevant biomarker associated with PSE risk (14). These studies highlight the complex interplay between SUA, inflammation, and oxidative stress in the post-stroke brain.

Unique to our study is the characterization of this non-linear association using flexible modeling (RCS) and robustness checks (PSM and sensitivity analyses), which strengthens confidence that the pattern is not solely driven by measured confounding. In contrast to prior research that primarily focused on overall stroke outcomes, our study specifically investigates PSE, a clinically significant but underexplored complication of stroke.

While our study provides important insights, several limitations must be acknowledged. First, the observational design limits our ability to infer causality between SUA levels and PSE risk. Second, despite PSM and multivariable adjustments, slight post-matching imbalance in renal biomarkers persisted (Supplementary Figure S1), and confounding by renal function and other unmeasured factors cannot be fully ruled out. Third, excluding patients who died in the ultra-early phase may shift the sample toward survivors with milder disease, thereby limiting the generalizability of the observed SUA–PSE association. Likewise, excluding patients with severe comorbidities (e.g., uremia, hydrocephalus, or cerebral herniation) may further limit the applicability of our findings to broader stroke populations, as these patients may exhibit distinct SUA–PSE dynamics. Additionally, SUA levels were measured at a single time point, which does not capture potential temporal fluctuations or the cumulative effects of SUA over time. Longitudinal studies are needed to better understand how changes in SUA levels influence PSE development. Moreover, the study population was derived from a single center, which may limit the external validity of our findings to other regions or healthcare settings. Finally, the original publication did not report detailed diagnostic criteria for post-stroke epilepsy (e.g., EEG confirmation, ICD coding rules, exclusion of acute symptomatic seizures) or a prespecified adjudication workflow; therefore, we used the dataset-labeled outcome as provided. This may introduce outcome misclassification, which should be considered when interpreting the estimated associations. Future research should aim to elucidate the mechanisms underlying the U-shaped relationship between SUA and PSE through experimental and longitudinal studies. Investigating the role of SUA modulation in animal models of stroke and epilepsy could provide valuable mechanistic insights. Additionally, randomized clinical trials evaluating interventions to optimize SUA levels in stroke patients are warranted. Expanding the study population to include diverse stroke subtypes and severity levels will enhance the generalizability of future findings. Finally, integrating genetic, metabolic, and dietary data could provide a more comprehensive understanding of the factors influencing SUA-PSE dynamics.

Our study builds on existing evidence and provides novel insights into the association between serum uric acid (SUA) and post-stroke epilepsy (PSE). We observed a non-linear relationship, with the lowest predicted odds of PSE at intermediate SUA levels, suggesting that SUA may reflect heterogeneous biological states relevant to post-stroke neuroinflammation and oxidative stress. These findings are hypothesis-generating and warrant validation in prospective cohorts and mechanistic studies.