Abstract P051: Association of Subclinical Hypoglycemia with Incident Cardiovascular Disease and All-cause Mortality: The Multi-Ethnic Study of Atherosclerosis
Trials of intensive glucose control have not improved cardiovascular disease (CVD) risk in type 2 diabetes; however data are inconsistent about the effects of maintaining below normal glucose levels in the general population. Further, it has been suggested that fasting glucose and HbA1c in the lower ranges have a different relationship with CVD and mortality.
In 5992 participants from the Multi-Ethnic Study of Atherosclerosis, free of CVD at baseline, we used logistic regression to investigate the associations of low fasting glucose (<80mg/dL) and HbA1c (<5.0%), from baseline and averaged across follow-up over 13 years, with incident CVD and mortality. We used the normal range (80 to <100mg/dL and 5.0 to <5.7%) as reference. We excluded participants with diabetes at baseline from the low, normal, and impaired groups at baseline, and excluded participants with diabetes at any visit from those groups for averaged values. We matched covariates to the visit of glycemic data collection.
Participants with low fasting glucose were more likely to be younger, female, Asian, use glucose lowering medication, have lower BMI and be normotensive. Those with low HbA1c were more likely to be younger, Asian or Hispanic, and have higher income at baseline. Adjusted for age, sex, race/ethnicity, education, and income, glucose and HbA1c in the impaired (100 to <126mg/dL and 5.7 to <6.5%) and diabetic ranges (≥126mg/dL and ≥6.5%) at baseline were significantly associated with increased CVD and mortality. Low baseline glucose and low baseline HbA1c were positively, but not significantly, associated with mortality; while low average glucose and low average HbA1c were both strongly and significantly associated with increased risk of mortality (Figure).
The relationships between fasting glucose and HbA1c with mortality are J-shaped at baseline, but strongly and significantly U-shaped when average levels are used. Consistently low fasting glucose or HbA1c may be better markers for risk from subclinical hypoglycemia than a single low measurement.
Author Disclosures: M. Mongraw-Chaffin: None. A.G. Bertoni: None. S. Golden: None. N. Mathioudakis: None. D.D. Sears: None. M. Szklo: None. C.A.M. Anderson: None.
- © 2017 by American Heart Association, Inc.