Real-World Risk Indicators of Severe Hypoglycemia in T2D—Results of the InHypo-DM Study

• Published in: Diabetes (New York, N.Y.) Vol. 67; no. Supplement_1
• Main Authors: Ratzki-Leewing, Alexandria, Harris, Stewart, Mequanint, Selam, Au, Natalie H, Black, Jason E, Reichert, Sonja M, Brown, Judith B, Ryan, Bridget L
• Format: Journal Article
• Language: English
• Published: New York American Diabetes Association 01.07.2018
• Subjects: Complications; Diabetes; Diabetes mellitus; Diabetes mellitus (non-insulin dependent); Disease prevention; Hypoglycemia; Insulin; Population studies; Risk factors
• ISSN: 0012-1797; 1939-327X
• DOI: 10.2337/db18-377-P

Insight into the self-reported frequency of third-party assisted severe hypoglycemia (SH) is critical for the therapeutic optimization of type 2 diabetes (T2D). This study presents real-world epidemiological evidence on the self-reported incidence and related risk indicators of SH. A validated questionnaire (InHypo-DMPQ) was administered to a population-based panel of adult Canadians with T2D treated with insulin and/or secretagogues. Questions pertained to respondents’ past hypoglycemia events as well as socio-demographic and clinical characteristics, including hypoglycemia unawareness. Univariable analyses (p≤0.20) followed by a multivariable zero-inflated negative binomial (ZINB) analysis were performed to explore the influence of potential risk indicators on the occurrence of any and repeated SH events. The current evaluation is based on a cohort of 436 respondents (mean age: 53 years; male: 54%). Among these individuals, 37.8% (95% CI: 33.3%-42.4%) self-reported at least one SH event in the past year; the incidence rate was 2.4 events (95% CI: 2.26-2.55) per person-year. Based on the ZINB analysis, only non-severe (NH) hypoglycemia was independently associated with SH, holding other indicators constant. Among respondents who reported ≥1 (vs. zero) NH events, the odds of never experiencing an SH event decreased by 82% (95% CI: 39.3%-94.7%, p=0.006). Likewise, a history of NH, among those who had a chance of experiencing SH, increased the expected rate of SH by a factor of 4.29 (95% CI: 2.25-8.17, P<0.001). Real-world estimates of SH in T2D are often limited by sub-optimal reporting and gaps in clinical practice. To help clarify the true frequency of SH, this large, population-based study leveraged the value and clinical relevance of self-reported hypoglycemia data. The results of this study suggest that the burden of third-party assisted SH is substantial in people with T2D. Clinical strategies to prevent NH may reduce the risk of any/repeated SH events and resulting sequelae.