Hyperglycaemia at hospital discharge is associated with worse cardiovascular outcomes after hospitalization for acute decompensated heart failure

Impaired glycaemic control is common in hospitalized patients, and hyperglycaemia is associated with poor clinical outcomes including increased length of stay, hospital readmission, and mortality.1 Acute hyperglycaemia is associated with adverse outcomes following acute coronary syndromes.2,3 There are limited data regarding the association between impaired glycaemic control on outcomes after acute decompensated heart failure (ADHF). We aimed to evaluate the association between discharge hyperglycaemia and adverse outcomes after ADHF. We performed a retrospective analysis of all patients admitted for ADHF from July to December 2019 on a high-acuity cardiovascular ward. Patients with acute coronary syndromes and cardiogenic shock were excluded. Patients were analyzed for clinical characteristics and discharge fasting plasma glucose (DG). We used Kaplan–Meier analysis and the log-rank test to assess the incidence of the primary outcome of ADHF readmission or cardiovascular death at 90 days. We constructed Cox hazards models to compare the primary outcome by tertile of DG, adjusting for factors previously associated with ADHF readmission. The final model adjusted for age, sex, body mass index, Charlson comorbidity index, left ventricular ejection fraction (LVEF), discharge systolic blood pressure, discharge creatinine, B-type natriuretic peptide level, and discharge loop diuretic requirement in furosemide equivalents. We also adjusted for number of HF medications (beta-blocker, renin-angiotensin-aldosterone system inhibitor, mineralocorticoid antagonist) at discharge in patients with reduced LVEF. We additionally adjusted for A1c and number of diabetes mellitus (DM)-directed medications (insulin and oral agents) at discharge in diabetics. Statistical analysis was performed with SPSS (version 27.0, IBM Corp., Armonk, NY, USA). Three hundred and twenty-six patients were included with mean DG of 132 ± 21 mg/dL (Table 1). Patients were mostly non-Hispanic White (65%), male (59%), and had a reduced LVEF <40% (53%). One hundred and thirteen (35%) had T2DM with mean haemoglobin A1c of 8.3 ± 1.1%, and diabetic patients had DG of 149 ± 24 mg/dL (P = 0.03 compared with non-diabetics). One hundred and one (31%) patients experienced the primary outcome. There was significant difference between patients in the lowest and highest tertiles of DG in rates of 90-day HF readmission (16.7 vs. 30.6%, P = 0.001) and cardiovascular death (5.5 vs. 10.2%, P = 0.03). In patients experiencing readmission (n = 77), mean DG and readmission fasting glucose were similar (139 vs. 131 mg, P = 0.1). Baseline clinical characteristics of patients stratified by highest and lowest tertile of fasting plasma glucose at discharge Continuous data are presented as mean with standard deviation and compared using Student’s t-test. In Kaplan–Meier analysis of the primary outcome, the primary outcome was significantly different at 90 days between the highest and lowest tertile of DG (log-rank P = 0.001; Figure 1). In the fully adjusted model, higher DG (highest tertile compared with lowest tertile) was significantly associated with higher rate of 90-day HF readmission or cardiovascular death (HR 1.18, 95% CI: 1.08–1.29, P = 0.009). There remained a significant association after additional adjustment for admission glucose level (HR 1.10, 95% CI: 1.02–1.17, P = 0.02). There was a significant interaction by DM status (HR 1.25 in diabetics, 95% CI: 1.07–1.44, P for interaction = 0.04), though the relationship was significant in non-diabetics alone. There was no interaction by LVEF status. Survival with freedom from heart failure readmission following acute decompensated heart failure by tertile of discharge glucose. This study investigated the association between fasting hyperglycaemia and 90-day ADHF readmission or cardiovascular death in patients admitted with ADHF. Our results demonstrate that elevated fasting DG is an independent risk factor for adverse cardiovascular events after hospital discharge. To the best of our knowledge, this is the first report of the relationship between hyperglycaemia at hospital discharge and outcomes following ADHF. The primary finding of this report is a significant association between fasting hyperglycaemia and 90-day ADHF readmission or cardiovascular death following hospitalization for ADHF. These findings are consistent with prior findings of this relationship in patients with acute coronary syndromes.2,3 Importantly, this association was still significant in patients without DM, suggesting this relationship is independent of baseline insulin resistance. The mechanisms of the detrimental effects of hyperglycaemia in ADHF are multifactorial. Hyperglycaemia during acute illness increases inflammation and endothelial dysfunction through mechanisms including oxidative stress, electrolyte disturbances, and disruption of intracellular pathways including nitric oxide signalling.4 Moreover, acute hyperglycaemia may have negative effects on myocardial mechanics.5 However, hyperglycaemia is known to be a normal physiologic stress response from sympathetic over-activation and excessive release of stress hormones.4 Therefore, discharge hyperglycaemia may represent a marker of ongoing physiologic stress related to ADHF rather than the direct mediator of poor outcomes. The results of this report should be interpreted in the context of three limitations. First, this was a single-centre report of patients on a high-acuity cardiovascular ward at a tertiary centre, and these findings may not be generalizable to all ADHF hospitalizations. Second, it is possible that discharge hyperglycaemia in diabetic patients may represent poor medication reinitiation, though the relationship persisted with adjustment for discharge antihyperglycaemic agents and admission glucose. Third, this report cannot establish whether treatment of acute hyperglycaemia improves ADHF outcomes. Therefore, these findings should be viewed as hypothesis generating. In conclusion, we identify that fasting hyperglycaemia at hospital discharge following ADHF is significantly associated with 90-day ADHF readmission or cardiovascular death. These findings suggest that discharge hyperglycaemia may be an important prognostic factor for adverse cardiovascular events after hospital discharge from ADHF. This project received no funding.