A comparison of the stress hyperglycemia ratio, glycemic gap, and glucose to assess the impact of stress-induced hyperglycemia on ischemic stroke outcome

Background The aim of this study is to compare metrics specific for stress-induced hyperglycemia (SIH) with glucose for predicting ischemic stroke outcome. Methods This observational retrospective study (n = 300) included patients acutely hospitalized for ischemic stroke over a 3.8-year period. We assessed the association between acute ischemic stroke outcome with the stress hyperglycemia ratio (SHR, relative increase in glycemia) and glycemic gap (GG, absolute increase in glycemia) using admission values and 5-day maximum values, along with incidence of poor outcome above recognized clinical thresholds of glucose 10 mmol/L, SHR 1.14, and GG 2.5 mmol/L. Results At admission, only SHR was associated with outcome after adjustment for clinical covariates (odds ratio [OR] = 2.88; 95% CI: 1.05-7.91; P = .041), while glucose or GG were not. Admission SHR >= 1.14 was also an indicator of poor outcome (39.1% vs 23.4%, P = .016), but not glucose >= 10 mmol/L or GG >= 2.5 mmol/L. All 5-day maximum glucose metrics were associated with outcome, as was any SHR >= 1.14 (40.9% vs 20.1%, P < .001) or GG >= 2.5 mmol/L (42.9% vs 23.4%, P = .011), but not glucose >= 10 mmol/L. Increased comorbidity was strongly associated with worse outcome (P < .001) in all models. Conclusions SHR provided the best prognostic insight at admission to assess the relationship between SIH and ischemic stroke outcome. Absolute glucose levels failed to account for natural interpatient variation in background glycemia and provided little prognostic insight. To assess the impact of SIH, future interventional studies need to be designed using designated markers of SIH such as SHR in preference to absolute glucose.