Background and Aims: Patients with cardiovascular disease are at high risk of type 2 diabetes (T2D), but little has been done to provide robust T2D risk estimates. We used Acarbose Cardiovascular Evaluation (ACE) trial data to develop a T2D risk calculator for Chinese people with coronary heart disease (CHD) and impaired glucose tolerance (IGT).

Methods: ACE recruited 6522 Chinese CHD patients with IGT, randomized 1:1 to acarbose or placebo, with prospective ascertainment of new-onset diabetes over median 5.0 years. Logistic regression was used to develop basic and full risk prediction models in the 6238 with the requisite baseline data. Clinically relevant variables and variables showing a univariate association (P<0.1) with incident diabetes were included.

Results: Overall 918 (15%) of 6238 participants developed diabetes. The basic model, using only clinical variables, was a poor fit with an area under the receiver-operating characteristic curve (AUC) of 0.60, whilst the full model including routinely-available laboratory results had an AUC of 0.76 with excellent goodness of fit (Table).

Conclusion: In Chinese people with IGT and CHD, lower age, male sex, obesity, prior hypertension, calcium channel blocker use, nondiabetic hyperglycaemia and dyslipidaemia were major determinants for new-onset diabetes. A model comprising these factors can be used to predict 5-year T2D risk.


S. Xu: None. C.A. Scott: None. R.L. Coleman: None. J. Tuomilehto: Advisory Panel; Self; Novo Nordisk A/S. Board Member; Self; AstraZeneca. Research Support; Self; Bayer AG. R.R. Holman: Advisory Panel; Self; Bayer AG, Novartis AG, Novo Nordisk A/S. Research Support; Self; AstraZeneca, Bayer AG, Merck Sharp & Dohme Corp. Other Relationship; Self; AstraZeneca, Bayer AG, GlaxoSmithKline plc., Janssen Pharmaceuticals, Inc.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at