Associations of blood glucose and systolic blood pressure variability with mortality in patients with acute cerebrovascular disease in intensive care units: a retrospective cohort study

Acute cerebrovascular disease (ACD) carries a high mortality rate despite advances in intensive care management. Glycemic variability (GV) and systolic blood pressure variability (SBPV) have been identified as prognostic indicators, but their combined and mediating effects on mortality remain unclear. This retrospective cohort study analyzed patients with cerebral infarction, intracerebral hemorrhage, or subarachnoid hemorrhage from the MIMIC-IV database. GV and SBPV were calculated using the coefficient of variation and stratified into tertiles. Logistic and Cox regression models were used to assess associations with 30-day, 1-year, and in-hospital mortality. Mediation analysis using the PROCESS macro with bootstrapping evaluated the interrelationship between GV and SBPV. Among 7,286 patients with ACD, both higher GV and SBPV were independently associated with increased 30-day, 1-year, and in-hospital mortality (all p < 0.001). A dose-response relationship was observed, with the combined elevation of GV and SBPV conferring the highest risk. For in-hospital mortality, SBPV showed the strongest predictive performance for in-hospital mortality (AUC = 0.672). Further exploratory analysis suggested a directional pathway, wherein the detrimental effect of GV on in-hospital mortality appeared to be partially dependent on the level of SBPV (proportion mediated = 18.0%, p < 0.001). Elevated GV and SBPV are independently associated with 30‑day, 1‑year, and in‑hospital mortality. Furthermore, higher combined variability scores are associated with a progressive increase in mortality risk. These findings suggest that concurrent monitoring of metabolic and hemodynamic fluctuations reflects physiological instability and may serve as supplementary markers alongside the assessment of absolute physiological levels.