EOS count trajectory independently predicts both short- and long-term mortality in critically ill AMI patients, establishing its role as a reliable marker for risk stratification and prognostic evaluation.
Key Findings
Results
Three distinct eosinophil count trajectories were identified in critically ill AMI patients using group-based trajectory modeling.
Trajectory 1 was characterized as 'stable-low'
Trajectory 2 was characterized as 'low-level steady rise'
Trajectory 3 was characterized as 'medium-level rapid rise'
The trajectories were derived from 1,493 critically ill AMI patients from the MIMIC-IV database
Group-based trajectory modeling (GBTM) was applied to identify the distinct trajectories during ICU admission
Results
Both rising eosinophil trajectories were associated with significantly reduced 28-day mortality risk compared to the stable-low trajectory.
Trajectory 2 (low-level steady rise) showed a significant reduction in 28-day mortality risk: HR = 0.68, 95% CI: 0.47–0.99
Trajectory 3 (medium-level rapid rise) showed a significant reduction in 28-day mortality risk: HR = 0.63, 95% CI: 0.50–0.79
Both comparisons were made relative to Trajectory 1 (stable-low) as the reference group
Associations were evaluated using multivariable logistic/Cox regression
Survival differences were assessed by Kaplan-Meier curves and log-rank tests
Results
Trajectory 3 (medium-level rapid rise) was associated with a 34% lower 1-year mortality risk compared to Trajectory 1 (stable-low).
HR = 0.72, 95% CI: 0.60–0.86 for 1-year mortality
The abstract specifically notes a '34% lower 1-year mortality risk' for Trajectory 3 vs. Trajectory 1
1-year mortality was one of two primary outcomes assessed
28-day mortality was the other primary outcome
Secondary outcomes included severe AKI incidence and ICU mortality
Results
Acute kidney injury (AKI) partially mediated the association between eosinophil count trajectories and 28-day mortality.
Mediation analysis was conducted to investigate the potential mediating effect of AKI on mortality
The mediation was described as 'partial,' indicating AKI accounts for only part of the trajectory-mortality relationship
Severe AKI incidence was a secondary outcome in the study
The mediation analysis specifically examined the pathway between EOS trajectories and 28-day (not 1-year) mortality
Methods
The study population consisted of 1,493 critically ill AMI patients drawn from the MIMIC-IV database.
Patients were enrolled from the MIMIC-IV database, a large publicly available ICU database
All patients were critically ill AMI patients admitted to the ICU
Eosinophil count trajectories were measured during ICU admission
Previous clinical studies had demonstrated 'conflicting evidence' regarding EOS count and adverse outcomes in AMI, motivating this trajectory-based approach
What This Means
This research suggests that tracking how eosinophil counts (a type of white blood cell) change over time during an ICU stay can help predict survival outcomes in patients who have had a heart attack (acute myocardial infarction). Using data from nearly 1,500 critically ill heart attack patients, researchers identified three distinct patterns of eosinophil count change: one group where counts stayed consistently low, one where counts rose gradually from a low level, and one where counts rose quickly from a medium level. Patients whose eosinophil counts rose — either gradually or rapidly — had meaningfully lower risks of dying within 28 days compared to those whose counts remained persistently low. The group with the most rapid rise also had a 34% lower risk of dying within one year.
The study also found that acute kidney injury (AKI), a common complication in heart attack patients, partially explained the link between eosinophil patterns and short-term death risk, suggesting that eosinophils may influence outcomes in part through their effect on kidney health. Importantly, previous research had produced conflicting results when looking at eosinophil levels at just a single point in time; this study's approach of examining how counts change over the entire ICU stay appears to provide a clearer and more reliable picture.
This research suggests that monitoring eosinophil count trajectories — rather than single measurements — could serve as a useful, non-invasive tool for identifying which heart attack patients in the ICU are at highest risk of dying or developing kidney complications. If validated in future studies, this approach could help clinicians better tailor monitoring and treatment intensity for high-risk patients.
Zhang W, Shuai W, Huang X, Dai J, Huo J, Shen M, et al.. (2026). Eosinophil count trajectories are associated with the prognosis of acute myocardial infarction patients: Insights from ICU data analysis.. PloS one. https://doi.org/10.1371/journal.pone.0349827