Is the Current Definition for Diabetes Relevant to Mortality Risk From All Causes and Cardiovascular and Noncardiovascular Diseases?

  1. The DECODE Study Group and
  2. on behalf of the European Diabetes Epidemiology Group

    Abstract

    OBJECTIVE—To assess the relation between fasting plasma glucose (FPG) or 2-h plasma glucose (2hPG) and mortality from all causes, cardiovascular disease (CVD), and non-CVD and to determine whether the relationship is graded or threshold.

    RESEARCH DESIGN AND METHODS—Diabetes Epidemiology: Collaborative Analysis Of Diagnostic Criteria in Europe (DECODE) is a collaborative prospective study of 22 cohorts in Europe with baseline glucose measurements for 29,714 subjects aged 30–89 years who were followed-up for 11 years (329,050 person-years). Hazard ratio (HR) for death was estimated using Cox regression analysis.

    RESULTS—High glucose concentrations as well as very low glucose levels were associated with increased risk of death. Compared with an FPG of 4.50–6.09 mmol/l, the multivariate-adjusted HR (95% CI) for FPG <4.50 mmol/l was 1.2 (1.0–1.4) for all-cause, 1.3 (1.0–1.8) for CVD, and 1.1 (0.9–1.4) for non-CVD mortality; the corresponding HRs for diabetes (FPG ≥7.0 mmol/l) were 1.6 (1.4–1.8), 1.6 (1.3–1.9), and 1.6 (1.4–1.9), respectively. For a 2hPG of 3.01–4.50 mmol/l, as compared with a 2hPG of 4.51–5.50 mmol/l, the HRs were 1.1 (1.0–1.2), 1.1 (0.9–1.3), and 1.1 (1.0–1.3), respectively; the corresponding HRs for diabetes (2hPG ≥11.1 mmol/l) were 2.0 (1.7–2.3), 1.9 (1.5–2.4), and 2.1 (1.7–2.5), respectively. The HR for previously undetected diabetes defined by 2hPG was not significantly different from that for known diabetes, which was significantly higher than that for undetected diabetes based on FPG. Subjects with a 2hPG of 10.01–11.09 mmol/l had mortality risks similar to those diabetic subjects defined by an FPG ≥7.0 mmol/l.

    CONCLUSIONS—The relation between mortality and glucose was J shaped rather than showing threshold effect at high glucose levels, except for CVD mortality and 2hPG, where the relation was graded and increasing.

    Footnotes

    • Address correspondence and reprint requests to Dr. Qing Qiao, Diabetes and Genetic Epidemiology Unit, Department of Epidemiology and Health Promotion, National Public Health Institute, Mannerheimintie 166, KTTL, 6 krs, FIN-00300 Helsinki, Finland. E-mail: qing.qiao{at}ktl.fi.

      Received for publication 17 June 2002 and accepted in revised form 26 November 2002.

      Members of the DECODE Study Group are listed in the appendix.

      Additional information for this article can be found in an online appendix at http://care.diabetesjournals.org.

      A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.

    | Table of Contents