Hb1Ac associated with subclinical atherosclerosis in individuals without diabetes

Introduction and methods

Strategies for early detection of subclinical atherosclerosis (SA) as tools to refine risk prediction in individuals with low or intermediate CV risk are getting more interest, since asymptomatic individuals can have underlying SA despite their low or moderate CV risk [1]. Moreover, more absolute CV deaths occur among individuals at low to moderate risk, since they represent the largest group [2].

Prediabetes, defined by ADA guidelines as HbA1c level 5.7% to 6.4% [3], is extremely prevalent and this offers an important window of opportunity for implementation of preventive interventions [4].

People with T2DM are automatically recognized as high-risk individuals in the SCORE model [5]. Also, ASCVD risk models recommended by the US guidelines recognize diabetes as a predictor of CV risk [6]. However, neither models have a predictor incorporated that indicates long-term glycemic exposure in individuals with prediabetes or without diabetes. This study assessed whether HbA1c levels were associated with presence and multiterritorial extent of SA in asymptomatic individuals without diabetes and with a low to moderate CV risk according to the SCORE algorithm. The added value of HbA1c as adjunct risk factor for prediction of the presence and extent of SA was also assessed.

This study used data from PESA (Progression of Early Subclinical Atherosclerosis), an observational study that prospectively included participants (40-54 years) without CVD or CKD. Participants without diabetes (n=3,973) were stratified to HbA1c strata. 10-Year risk of CV death prediction was estimated using the SCORE algorithm and participants were classified to low (<1%) or moderate (1-5%) risk categories. Atherosclerotic plaques in carotids, infrarenal abdominal aorta, and iliofemoral arteries were detected by 2-dimensional vascular ultrasound (2DVUS). Cardiac computed tomography was used to estimate coronary artery calcium (CAC) scoring. Extent of SA was defined by presence of 2DVUS-detected plaques and CAC scores ≥1. Multiterritorial extent of SA was defined by the number of vascular territories with atherosclerotic plaques. Individuals were classified as disease-free (0 vascular sites affected), or as having focal (1 site), intermediate (2 or 3 sites), or generalized atherosclerosis (4 to 6 sites).

Main results

• There was a significant association between HbA1c levels and prevalence of SA as well as multiterritorial extent of SA in all prediabetic groups and in the group below the prediabetic threshold (HbA1c category 5.5%-5.6%) (Ptrend<0.001). Elevated levels of HbA1c were associated with more SA in carotid arteries, infrarenal abdominal aorta, or iliofemoral arteries (Ptrend<0.001 for each vascular territory).
• The association between HbA1c and number of atherosclerotic plaques, number of noncoronary territories with plaque, CAC scores, and multiterritorial SA extent remained significant after multivariate adjustments (P<0.001 for each of the 4 outcomes).

• Increasing HbA1c levels were independently associated with a higher risk for SA in participants with a low SCORE risk (P<0.001). This association was absent in individuals with a moderate SCORE risk (P=0.335).

• Subgroup analysis assessing interaction between SCORE risk predictors and HbA1c categories for the clinical outcomes disease-free or focal SA vs. intermediate or generalized SA demonstrated that the effect of HbA1c levels was homogenous across SCORE predictors, except for total cholesterol.

• Adding HbA1c as a continuous parameter to the SCORE model significantly improved the multiterriorial SA risk (c-statistics SCORE 0.732 [0.717-0.748] vs. c-statistics SCORE+HbA1c 0.751 [0.735-0.766], P<0.001).

Conclusion

References

Find this article online at J Am Coll Cardiol.