Background and Purpose: Serial glucose could more accurately reflect glycemic status in acute ischemic stroke (AIS) than presenting glucose. We sought to investigate the clinical implications of various parameters of serial glucose on the outcomes of patients with AIS treated with intravenous thrombolysis (IVT). 
Methods: This was a single-center, prospective, observational study of stroke patients treated with IVT. Blood glucose (BG) was serially measured at 6 time points during the first 24h of IVT. The following parameters related to BG were analyzed: initial blood glucose (iBG), mean blood glucose (mBG), maximal blood glucose (max BG), standard deviation (SD), coefficient of variance (CoV), and J-index. The primary endpoint analyzed was a good outcome at 3m (modified Rankin Scale of 0–2).
Results: Among the 492 patients in the cohort (age, 70±12y; men, 57%), the overall BG level was 131±33mg/dl. At 3m, 40.4% of the patients had a good outcome. Patients with good outcomes had significantly lower mBG (121 vs 128mg/dl) and higher CoV (17% vs 14%) but no differences in the other parameters. For patients with higher mBG (every 30mg/dl), the likelihood of achieving a good outcome decreased (OR 0.81, 95% CI 0.66-0.99). For patients with higher CoV (every 10%), the likelihood of a good outcome increased (OR 1.37, 95% CI 1.11-1.69). But non-significant association was observed in SD and J-index.
Conclusions: The results showed that higher mBG was consistently associated with worse outcomes in IV thrombolyzed stroke patients, suggesting that lowering mBG might be potential therapeutic target in these patients.

Clinical Implications of Serial Glucose Measurements in Acute Ischemic Stroke Treated with Intravenous Thrombolysis