Postprandial, but not fasting, glucose predicts CV risk in type 2 diabetes

Postprandial blood glucose levels, but not fasting blood glucose, predict cardiovascular events and all-cause mortality in people with type 2 diabetes, according to 14 years of follow-up in the San Luigi Gonzaga Diabetes Study.

HbA_1c was also a predictor of both outcomes over the long term, according to Dr Franco Cavalot (University of Turin, Italy), who presented long-term follow-up data here at the European Association for the Study of Diabetes (EASD) 2011 Meeting.

Franco and colleagues reported five-year findings from this study back in 2006: at that time, postprandial blood glucose was predictive of CVD events after adjustment for HbA_1c and other cardiovascular risk factors.

The study included 505 patients with type 2 diabetes reporting for regular follow-up at the diabetes clinic. Patients underwent blood glucose tests at four time points: fasting, two hours after breakfast, two hours after lunch, and before dinner. Three separate Cox regression models were applied to confirm the results, with one of the models adjusting for age, sex, known diabetes duration, smoking, body-mass index (BMI), systolic and diastolic blood pressure, cholesterol levels, triglycerides, creatinine, and albumin excretion rate.

Over the 14 years of follow-up, 147 patients died and 172 cardiovascular events occurred. Across all of the analyses conducted, HbA_1c and blood glucose levels after lunch predicted both future cardiovascular events and all-cause deaths. Further analyses suggested that the information provided by postprandial blood glucose added to the predictive power of HbA_1c alone for both cardiovascular events and all-cause mortality. For all-cause mortality, the predictive ability of postprandial test was similar to that of HbA_1c.

"The 14-year follow-up of the San Luigi Gonzaga Diabetes Study supports the conclusion that postprandial blood glucose is not only an essential determinant of HbA_1c, but also per se a relevant tool in cardiovascular events and mortality risk stratification in type 2 diabetes mellitus," Cavalot concluded.

Cavalot noted that the lack of an association between future events and blood tests taken after breakfast is not surprising, given that, in Italy, breakfast tends to be a light meal. Asked why the study had not measured blood glucose after supper, he explained that all tests were taken during clinic hours; he speculated that these would likely be similar to the postlunch findings.