Glucose levels in the normal range predict incident diabetes in families with premature coronary heart disease

Stasia S. Reynolds, Lisa R. Yanek, Dhananjay Vaidya, Samia Mora, Taryn F. Moy, Christopher D. Saudek, Lewis C. Becker, Diane M. Becker

Research output: Contribution to journalArticlepeer-review

Abstract

Purpose: Little is known about excess risk of incident diabetes conferred by fasting plasma glucose (FPG) within the normal range (<5.6 mmol/l) for high risk families. Methods: Healthy 30-59 year old non-diabetic siblings (N = 542) of index cases with documented premature coronary disease were followed prospectively for type 2 diabetes. Results: During 8.7 ± 3 years of follow-up, incident diabetes was identified in 7.8%. Rates were incremental with baseline non-diabetes FPG thresholds of 5.0, 5.6, 6.1, and 6.7 mmol/l (p for trend <0.0001). FPG was the strongest predictor of incident diabetes even across levels within the normal range. The multivariable adjusted relative risk was 14.9 (95% CI = 3.4-65.2) at FPG thresholds ≥5.0 mmol/l versus FPG <5.0 mmol/l. The maximal diagnostic efficiency for FPG was 5.50 mmol/l; with sensitivity and specificity 0.782. All FPG thresholds in the normal range between 5.0 and 5.6 mmol/l showed efficiency levels >0.74. The overall area under the ROC curve predicting incident diabetes for normal and prediabetes ranges of FPG was 0.867. Conclusion: Higher FPG levels within the designated "normal" range in high risk families are a potent independent risk factor for type 2 diabetes and may serve as a sentinel to trigger primary preventive interventions.

Original languageEnglish (US)
Pages (from-to)267-273
Number of pages7
JournalDiabetes Research and Clinical Practice
Volume74
Issue number3
DOIs
StatePublished - Dec 2006

Keywords

  • Coronary heart disease
  • Families
  • Incident diabetes
  • Prevention

ASJC Scopus subject areas

  • Internal Medicine
  • Endocrinology, Diabetes and Metabolism
  • Endocrinology

Fingerprint Dive into the research topics of 'Glucose levels in the normal range predict incident diabetes in families with premature coronary heart disease'. Together they form a unique fingerprint.

Cite this