Identification of impaired fasting glucose, healthcare utilization and progression to diabetes in the UK using the Clinical Practice Research Datalink (CPRD)

Abstract

Purpose: Few studies have examined patients with prediabetes in usual, "real-world" clinical practice settings. Among patients with impaired fasting glucose (IFG), we aimed to describe the rates of progression to diabetes and to examine the long-term reduction in diabetes risk associated with regression to normoglycemia at 1 year.

Methods: The UK-based study included 120 055 non-diabetic patients in Clinical Practice Research Datalink from 2001 to 2012 aged 25+ years and with ≥1 fasting plasma glucose (FPG) test between ≥6.1 and <7.0 mmol/l indicating IFG who were followed for progression to diabetes. In a subgroup of 45 167 patients with IFG with subsequent FPG results 1 year later, we assessed the 1-year glycemic status change and estimated the relative hazard of diabetes comparing patients with regression to normoglycemia (IFG-normoglycemia) to those who remained in IFG (IFG-IFG) using a multivariable Cox model.

Results: Among patients with IFG with over 414 649 person-years of follow-up, 52% received a subsequent FPG test, and 10% developed diabetes within 1 year after recognition of IFG. The incidence rate of diabetes was 5.86 (95% CI: 5.78 to 5.93) per 100 person-years. In the subgroup analysis, 31% of these patients remained in IFG, while 53% and 16% converted to normoglycemia or diabetes, respectively. The adjusted hazard ratio of developing diabetes was 0.33 (95% CI: 0.31 to 0.35) comparing IFG-normoglycemia to IFG-IFG.

Conclusions: IFG is a high-risk state for diabetes. Regression to normoglycemia from IFG strongly reduces the long-term risk of developing diabetes. Our study also shows the feasibility of identifying patients with IFG in the Clinical Practice Research Datalink. Copyright © 2016 John Wiley & Sons, Ltd.

Keywords: Clinical Practice Research Datalink; diabetes; impaired fasting glucose; pharmacoepidemiology; prediabetes.