Selvin · The New England journal of medicine 2010 · prospective cohort study · n=11,092

Glycated hemoglobin, diabetes, and cardiovascular risk in nondiabetic adults.

Cited 1572 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective cohort study

PubMed 20200384 · doi:10.1056/NEJMoa0908359 · record verified 2026-08-30

What was done

Glycated hemoglobin was measured at baseline in 11,092 Black or White adults without a history of diabetes or cardiovascular disease participating in the Atherosclerosis Risk in Communities (ARIC) study (visit 2, 1990–1992). The authors evaluated and compared the prognostic value of glycated hemoglobin and fasting glucose for predicting incident diabetes, coronary heart disease (CHD), stroke, and all-cause mortality using multivariable-adjusted hazard models.

What was found

Compared with baseline glycated hemoglobin of 5.0% to <5.5% (reference, HR 1.00), multivariable-adjusted hazard ratios (95% CI) for diagnosed diabetes were 0.52 (0.40 to 0.69) for <5.0%, 1.86 (1.67 to 2.08) for 5.5% to <6.0%, 4.48 (3.92 to 5.13) for 6.0% to <6.5%, and 16.47 (14.22 to 19.08) for >=6.5%. For CHD, corresponding HRs were 0.96 (0.74 to 1.24), 1.00, 1.23 (1.07 to 1.41), 1.78 (1.48 to 2.15), and 1.95 (1.53 to 2.48), with similar HR patterns observed for stroke. All-cause mortality showed a J-shaped association with glycated hemoglobin. These associations remained significant after adjusting for baseline fasting glucose, whereas fasting glucose was not significantly associated with cardiovascular disease or all-cause mortality after adjusting for glycated hemoglobin and covariates. Adding glycated hemoglobin to models containing fasting glucose significantly improved risk discrimination for CHD.

Why it matters

These findings support using glycated hemoglobin to screen for diabetes and demonstrate that it provides superior prognostic information for cardiovascular disease and all-cause mortality compared to fasting glucose in nondiabetic adults.

Limits

As an observational cohort study, residual confounding cannot be ruled out. The study was conducted in US Black and White adults, so results may not generalize to other racial or ethnic groups. Follow-up length, exact event counts, and specific numerical risk estimates for stroke and mortality were not reported in the abstract.

Cited by