Background and Aims: Hypoglycemia is a major threat of insulin therapy, but only severe events are often considered as valid endpoints. Surrogates are required; mild events are logical candidates. The International Hypoglycemia Study Group (IHSG) proposed thresholds (3.9 and 3.0 mmol/L) for representative mild events, but evidence is scarce.

Materials and Methods: We studied the correlation of the rate (or the incidence) of non-severe events with the rate (or the incidence) of severe events in a study-level systematic analysis of published data (1/1/00-6/1/18) in insulin-treated patients. Mild event was defined as either “any hypoglycemia,” or “severe or confirmed <3.9 mmol/L,” or “severe or confirmed <3.0,” with a tolerance of +/-0.4 mmol/L to manage the multiplicity of definitions. Nocturnal period was defined as per study design, therefore not homogeneously. Severe events were mostly defined as per IHSG.

Results and Conclusion: 127 out of 15570 studies were eventually included. Pearson R² are presented (Figure). The mild hypoglycemia were usually better correlated with incidence rather than rate of severe events. Incidence of severe or confirmed <3.0 mmol/L, especially if limited to the night, appeared to be the best surrogate for incidence of severe events.


R. Roussel: Advisory Panel; Self; Merck Sharp & Dohme Corp., Sanofi. Consultant; Self; Abbott, AstraZeneca, Eli Lilly and Company, Medtronic, Novo Nordisk A/S, Physiogenex. Research Support; Self; Janssen Pharmaceuticals, Inc. A. Moutairou: None. J. Riveline: Board Member; Self; Abbott, Lilly Diabetes, Novo Nordisk Inc. Consultant; Self; Novo Nordisk Inc. Research Support; Self; Air Liquide, Amgen Inc. Speaker's Bureau; Self; Abbott, Lilly Diabetes, Sanofi. J. Gautier: None. L. Potier: Advisory Panel; Self; Merck Sharp & Dohme Corp. Consultant; Self; Sanofi.



Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at