This study was designed to examine the association of heart rate variability (HRV) with blood glucose levels in a large community-based population. Previous reports have shown HRV to be reduced in diabetics, suggesting the presence of abnormalities in neural regulatory mechanisms. There is scant information about HRV across the spectrum of blood glucose levels in a population-based cohort. One thousand nine hundred nineteen men and women from the Framingham Offspring Study, who underwent ambulatory electrocardiographic recordings at a routine examination, were eligible. HRV variables included the SD of normal RR intervals (SDNN), high-frequency (HF, 0.15 to 0.40 Hz) and low-frequency (LF, 0.04 to 0.15 Hz) power, and LF/HF ratio. Fasting plasma glucose levels were used to classify subjects as normal (<110 mg/dl; n = 1, 779), as having impaired fasting glucose levels (110 to 125 mg/dl; n = 56), and as having diabetes mellitus (DM >/=126 mg/dl or receiving therapy; n = 84). SDNN, LF and HF power, and LF/HF ratio were inversely related to plasma glucose levels (p <0.0001). SDNN and LF and HF powers were reduced in DM subjects (4.28 +/- 0.03, 6.03 +/- 0. 08, and 4.95 +/- 0.09) and in subjects with impaired fasting glucose levels (4.37 +/- 0.04, 6.26 +/- 0.10, and 5.06 +/- 0.11) compared with those with normal fasting glucose (4.51 +/- 0.01, 6.77 +/- 0.02, and 5.55 +/- 0.02, all p <0.005), respectively. After adjusting for covariates (age, sex, heart rate, body mass index, antihypertensive and cardiac medications, systolic and diastolic blood pressures, smoking, and alcohol and coffee consumption), LF power and LF/HF ratio were lower in DM subjects than in those with normal fasting glucose (p <0.005). HRV is inversely associated with plasma glucose levels and is reduced in diabetics as well as in subjects with impaired fasting glucose levels. Additional research is needed to determine if low HRV contributes to the increased cardiovascular morbidity and mortality described in subjects with hyperglycemia.