In this study, we investigated the time-varying association between BP and MBE risk after LHI. We found that the influence of BP on MBE risk varied by time epochs: In LHI patients, 24-hour mean SBP demonstrated a U-shaped association, with no significant effect during 1–12 h but a positive linear correlation in 13–24 h after onset. DBP exhibited a threshold effect, primarily influencing MBE development during first 12 h post-onset. To our knowledge, this is the first study specifically addressing BP management in LHI and its association with MBE. These findings underscored the importance of considering temporal BP dynamics in optimizing MBE prevention strategies for LHI patients.

In our study, SBP declined in non-MBE group but remained elevated in MBE group during 13–24 h after onset, with a positive association identified between mean SBP and MBE risk during this period. No significant association was observed for SBP during 1–12 h epoch. This pattern aligned with previous studies. Shin et al. [10] and VISTA collaboration [15] reported a linear relationship between SBP and poor clinical outcomes during the 8-24-hour window. Prasad et al. [16] found that neurological deterioration was associated with greater SBP variability within first three days after thrombectomy, primarily during 12–24 h. This effect may result from progressive impairment of cerebral autoregulation, where elevated SBP exacerbates brain edema and intracranial pressure, increasing MBE risk. Alternatively, the progression of brain tissue edema and raised intracranial pressure during the late acute phase could trigger the Cushing reflex, further elevating SBP. In contrast, no significant association was observed between SBP during 1–12 h after onset and MBE, consistent with Shin et al. [10] and the VISTA collaboration [15], who reported an insignificant relationship between SBP and functional outcome or early neurological deterioration within first 8 h post-stroke. Cerebral autoregulation may remain partially intact during the initial post-stroke hours, potentially mitigating the impact of SBP on MBE risk [15, 17]. Our findings emphasized the critical importance of SBP management especially during the 13–24 h phase after onset to reduce MBE risk, while also highlighting the complexity of SBP’s role during acute phase and the need for further research to clarify its underlying mechanisms.

A U-shaped association was found between mean SBP and MBE risk during 24-hour period, with lowest risk at 120 mmHg. Previous studies linked higher admission SBP to MBE risk [9] and elevated 24-hour mean SBP to radiological evidence of brain edema [18]. However, they did not specifically focus on LHI, nor use the full MBE definition, which includes both clinical and radiological criteria. Our findings suggested that the optimal mean SBP range for LHI within 24 h after onset was around 120 mmHg, consistent with prior trials. Post-hoc analysis demonstrated that SBP targets below 140 mmHg were associated with better outcome compared to those below 180 mmHg [19], while more aggressive reductions below 120 mmHg after thrombectomy were linked to worse outcomes [20]. Elevated BP could disrupt the blood-brain barrier, causing brain edema [21], while excessively low BP could reduce cerebral blood flow, increasing ischemic injury [22].

We identified a threshold effect between DBP and MBE risk. Changes in 24-hour mean DBP below 70 mmHg were not linked to MBE development, while elevation in DBP above this threshold significantly raised MBE risk. A similar threshold of 75 mmHg was observed during 1–12 h epoch. The role of DBP in cardiovascular or cerebrovascular outcomes remains controversial, with some studies reporting a J-shaped relationship and optimal DBP ranges of 70–80 mmHg [23,24,25,26], while others showed a linear association [27, 28]. Research on DBP and MBE, particularly in LHI patients, was limited. We are the first to demonstrate this threshold effect, suggesting a broader safe range for DBP management compared to SBP. The time-dependent association likely reflected the importance of maintaining brain perfusion during the first post-onset 12 h, where DBP playing a greater role than SBP, as indicated by the formula: mean arterial pressure = 1/3 × (SBP + 2 × DBP). Notably, the association between DBP and MBE was significant only in patients without thrombectomy or successful reperfusion. Consistent with our findings, Hong et al. [29] reported that higher baseline BP correlated with better functional outcomes in patients with successful reperfusion but poorer outcomes in those without. Patients without thrombectomy treatment or successful reperfusion often presented later, had larger infarct areas, or experienced more severe symptoms, necessitating more precise blood pressure management. In such cases, maintain brain perfusion becomes more essential, with DBP playing a pivotal role.

In our study, the median interval from symptom onset to MBE occurrence was 28.97 h, consistent with prior research [30, 31]. Among the 117 MBE cases, 81.2% (95/117) occurred > 24 h after onset, and all BP measurements of these patients were included for analysis. For the remaining 22 patients developing MBE within 24 h, only pre-MBE BP data were kept to ensure correct temporal sequence between studied parameters and outcome, which ensured that the BP data analyzed were obtained strictly before MBE, making it more probable that BP variations influenced MBE development rather than the reverse. Nevertheless, our findings still indicate association rather than causation. Progressive cerebral edema, even before meeting MBE criteria, might influence BP through mechanisms like Cushing reflex [32, 33], suggesting a potential bidirectional or cyclical relationship. Further studies, including animal experiments and randomized controlled trials (RCTs), are needed to clarify causality.

Our study had several strengths. We are the first to focus on BP management in among LHI population. Currently, no specific recommendations for BP management in LHI are available. Our findings could serve as a guide for personalized BP management in this critically ill population and provide preliminary data for future RCTs. Second, we offered a detailed analysis of temporal BP dynamics and their link to MBE, taking time-varying effect of BP into account. Third, we identified key BP thresholds relevant to MBE after LHI, providing insights for targeted management. Lastly, we identified that while SBP affected MBE consistently among subgroups, DBP had a greater impact on non-reperfusion patients, highlighting the need for personalized management based on treatment and reperfusion status.

Our study had some limitations. First, not all patients had BP measured for 24 times, which was inevitable in studies analyzing dynamic indicators at multiple time points. However, 90% (357/414) had at least three measurements in both 1–12 h and 13–24 h epochs, and 95% (385/414) had at least six measurements during 24 h. Previous studies on hemodynamic parameters within 12–72 h post-stroke included patients with 1–10 measurements, which was comparable to our study [6, 16, 34]. Second, pre-stroke BP, which influences post-stroke BP elevation [35], was unavailable due to patients’ limited habit of regular monitoring. However, we substituted baseline BP and hypertension history, which partially reflect pre-stroke BP level. Third, we excluded patients developing space-occupying hemorrhagic transformation before MBE occurrence, which may introduce selection bias. However, other reasons of mass effect interfere with MBE assessment, and previous studies also excluded such patients [13, 36]. Fourth, some potentially influential factors, such as laboratory parameters, were not available and should be included in future studies. Fifth, as an observational study, our findings demonstrated associations rather than causation. Although we excluded post-MBE BP data and most MBE cases occurred > 24 h after BP measurement, reverse causality (e.g., via Cushing reflex) cannot be fully excluded. Further RCTs and animal studies are needed to clarify causality.