Assessment of hypoglycaemia during basal insulin therapy: Temporal distribution and risk of events using a predefined or an expanded definition of nocturnal events

Abstract Aim To describe in type 2 diabetes the 24-hour distribution of hypoglycaemia and compare the frequency of nocturnal events based on a predefined nocturnal window or an expanded interval, using illustrative data for two insulin glargine formulations. Methods Temporal distribution of hypoglycaemic events was assessed descriptively and by profile using participant-level data from three randomized trials comparing insulin glargine 300 U/mL (Gla-300) and 100 U/mL (Gla-100). Risk of hypoglycaemia and annualized event rates were compared for the predefined nocturnal interval (00:00 to 05:59h) and an expanded window (22:00h to the pre-breakfast glucose measurement). Results Confirmed (≤3.9mmol/L [≤70 mg/dL]) or severe hypoglycaemic events were reported most frequently between 06:00 and 10:00 h with both insulins. Nearly threefold more events were identified using the expanded nocturnal interval. Risk of ≥1 nocturnal event was 25% lower with Gla-300 than Gla-100 with the predefined, and 16% lower with the expanded interval; annualized event rates were 31% and 24% lower with the predefined and expanded window, respectively. The between-insulin difference in number of nocturnal events depended markedly on the chosen nocturnal interval (556 vs. 1145 fewer events with Gla-300 using the predefined vs. expanded interval). Conclusions The predefined 00:00–05:59h nocturnal interval excluded many hypoglycaemic events occurring during the actual overnight interval. While Gla-300 reduced hypoglycaemic events versus Gla-100 (regardless of the interval considered), the results obtained using the expanded window better reflect the clinical experience of people treated with basal insulin.