Abstract
INTRODUCTION: Severe hypoglycemic events (SHEs) are associated with significant morbidity, mortality and costs. However, the more common non-severe hypoglycemic events (NSHEs) are less well explored. We investigated the association between reported frequency of NSHEs and SHEs among patients with type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) in the PREDICTIVE study. METHODS: PREDICTIVE was a global, prospective, observational study. Patients with T1DM (n = 7,420) or T2DM (n = 12,981), starting treatment with insulin detemir, reported the number of NSHEs and SHEs experienced during the 4 weeks prior to baseline and follow-up visits (mean 14.4 weeks). Logistic regression was used to determine the odds ratio (OR) of experiencing ≥1 SHE, in patients having 1-4 or ≥5 NSHEs, versus those having 0 NSHEs, while controlling for baseline covariates. RESULTS: Hypoglycemia rates were lower at follow-up than baseline. At baseline 59.2% (T1DM) and 18.8% (T2DM) reported any hypoglycemia and at follow-up 39.5% (T1DM) and 8.6% (T2DM). There was a significant (P < 0.0001) increase in the odds of ≥1 SHEs with increasing frequency of NSHEs in T1DM and T2DM, for both crude and adjusted estimates. At baseline, in T1DM, ORs for ≥1 SHE were 1.92 and 2.13 for 1-4 and ≥5 NSHEs, respectively; the corresponding ORs in T2DM were 10.83 and 15.36, respectively. At follow-up, the ORs for ≥1 SHE were 2.01 and 3.20 (T1DM) and 18.99 and 24.29 (T2DM) for 1-4 and ≥5 NSHEs, respectively. CONCLUSION: A statistically significant association between NSHE and SHE frequency was found in T1DM and T2DM. These data provide a clear rationale for the reduction of hypoglycemic events, regardless of severity, while striving for optimal glycemic control.