Elevated cortisol, not glucose, independently increased hospital mortality risk in acute coronary syndrome patients (OR 1.073, p<0.0001 vs. OR 1.0, p=0.9).
Do elevated cortisol and glucose levels independently increase hospital mortality in patients with acute coronary syndromes?
In patients with acute coronary syndromes, elevated cortisol, rather than hyperglycemia per se, appears to drive increased hospital mortality, suggesting cortisol acts as a marker of stress response that confounds the glucose-mortality relationship.
Abstract Background Hyperglycemia is associated with an increased risk for death in acute coronary syndromes (ACS). Elevated blood glucose could be related to underlying glucose metabolism abnormalities or be caused by a counter-regulatory stress response. An important part of this response involves cortisol, which in turn can cause hyperglycemia through various physiological mechanisms and has also shown to be associated with increased mortality. To determine the causal relationships underlying the effects of glucose, cortisol and diabetes over mortality in ACS, we propose using a causal inference method approach. Methods Single-center, prospective, observational study. Patients admitted to the coronary care unit with a diagnosis of ACS were included. On admission, blood samples were obtained to measure serum glucose and cortisol. Clinical variables were obtained from electronic medical records. A directed acyclic graph (DAG) relating clinical variables, inflammation markers, sympathetic nervous system activity, glucose and cortisol with hospital mortality was constructed with DAGitty v3.1. The obtained adjustment sets were used to build a logistic regression model. Statistical analysis was performed using R version 4.3.3, and significance was set at p 0.05. Results The DAG can be observed in figure 1. The minimal sufficient adjustment sets for estimating the total effect of cortisol on hospital mortality were: age, C-reactive protein, diabetes, glucose, Killip class, ST segment deviation, and troponin elevation. We included 290 patients. The mean age was 68.5 ± 12.3 years, and 62% were male. 31% had STEMI, 67% hypertension, and 27% diabetes. Hospital mortality was 7.2%. Fatal cases demonstrated significantly higher cortisol levels compared to survivors (32.36 ± 19.74 mg/dl vs. 15.48 ± 11.41 mg/dl, p = 0.0009). Similarly, glucose levels were also elevated in fatal cases (206.3 ± 142.7 mg/dl vs. 131.8 ± 60.8 mg/dl, p = 0.03). Correspondingly, univariate analysis revealed a significant association between elevated cortisol and mortality (OR = 1.073, 95% CI: 1.044 - 1.104, p 0.0001), as well as between high glucose levels and mortality (OR = 1.009, 95% CI: 1.005 - 1.014, p = 0.0001). The regression model results can be observed in the table. Cortisol, as well as age, diabetes, Killip D class and ST segment deviation were significantly associated with a greater hospital mortality. Troponin elevation was numerically associated with increased mortality, with OR = 9.56 and p = 0.093. Glucose levels were not associated with increased mortality, with an OR of 1.0 and p = 0.9, nor was C-reactive protein. Conclusion This causal inference analysis suggests that hyperglycemia per se is not directly associated with worse outcomes in acute coronary syndromes. Instead, cortisol, as a marker of the stress response, may drive increased mortality and confound the observed relationship between glucose levels and adverse events.Figure:directed acyclic graph Table:Model results
Costa et al. (2025) studied this question. Elevated cortisol, not glucose, independently increased hospital mortality risk in acute coronary syndrome patients (OR 1.073, p<0.0001 vs. OR 1.0, p=0.9).