Blood glucose measures offered little benefit for CVD risk prediction
Assessment of HbA1c values in the context of CVD risk assessment may provide little incremental benefit for the prediction of CVD risk in people without known CVD or diabetes, new data suggest.
A 2010 American College of Cardiology Foundation/American Heart Association guideline for the assessment of CV risk in asymptomatic adults states that measurement of HbA1c may be reasonable for CVD risk assessment for adults not diagnosed with CVD or diabetes, according to the study background.
Emanuele Di Angelantonio, MD, from the University of Cambridge, and colleagues analyzed whether adding information on HbA1c levels to prognostic models consisting of conventional CV risk factors would improve the prediction of first-onset CVD outcomes in middle-aged and older adults without diabetes. They also assessed how measurement of HbA1c levels compared with other glycemia measures.
The analysis included individual participant data from 73 prospective studies of 294,998 individual (mean age, 58 years; 49% women) without CVD or diabetes at baseline. The researchers developed CVD risk prediction models with and without a glycemia measure. According to the researchers, this enabled them to measure risk discrimination for CVD outcomes and to determine whether adding a glycemia measure would reclassify participants across predicted 10-year risk categories.
The 10-year risk categories were low (<5% chance of CVD outcome within 10 years), intermediate (5% to <7.5% chance) and high (≥7.5% chance).
The researchers found 20,840 incident CVD outcomes (13,237 CHD; 7,603 stroke) during the median follow-up period of 9.9 years.
Adjustment of conventional CVD risk factors revealed a J-shaped association between HbA1c levels and CVD risk; this was also true for other glycemia measures. The association between HbA1c levels and CVD risk changed slightly after adjustment for total cholesterol and triglyceride concentrations or estimated glomerular filtration rate, but was attenuated after adjustment for concentrations of HDL and C-reactive protein.
The C-index for a CVD risk-prediction model consisting of conventional CVD risk factors alone was 0.7434, and the addition of HbA1c levels to the model was associated with a change of only 0.0018 and a net reclassification improvement of 0.42 for the predicted 10-year risk categories, according to the study.
The improvement in CVD risk prediction after adding HbA1c levels to the model was equal to or better than the improvement after adding fasting, random or postload plasma glucose levels to the model, according to the researchers.
The results “should encourage further studies to test whether very low glycemia levels are markers of ill health, such as that caused by hepatic dysfunction or other comorbidities,” Di Angelantonio and colleagues wrote.
Disclosure: See the full study for a list of the researchers’ relevant financial disclosures.