IntroductionLong-term changes of fasting blood glucose (FBG) in relation to lower-extremity peripheral artery disease (lower-extremity PAD) in people without diabetes has barely been reported. Our study aimed to investigate the association between FBG variability and the incidence of lower-extremity PAD in people without diabetes.Research design and methodsWe included 7699 participants without prior lower-extremity PAD and diabetes from the Atherosclerosis Risk in Communities study in the final analysis. At least two measurements of FBG were required during follow-up. Variability of FBG was identified using SD, coefficient of variation (CV), variability independent of the mean (VIM) and average real variability. Lower-extremity PAD was defined as an ankle brachial index <0.9, or hospitalization with a lower-extremity PAD diagnosis. Cox regression model was used to calculate HR for incidence of lower-extremity PAD and FBG variability.ResultsDuring a median follow-up of 19.5 years, 504 (6.5 %) lower-extremity PAD events were observed, 54.4% (n=274) were male, and 17.5% (n=88) were African-American. FBG variability was positively associated with incident lower-extremity PAD, with a linear relationship. HRs for CV and VIM were 1.015 (95% CI: 1.001 to 1.03; p=0.023), and 1.032 (95% CI: 1.004 to 1.06; p=0.022) for lower-extremity PAD, respectively. Participants in the lowest quartile of CV were at lower lower-extremity PAD risk compared with the highest ones (HR: 1.499, 95% CI: 1.16 to 1.938; p=0.002).ConclusionsHigher FBG variability was independently associated with increased prevalence of lower-extremity PAD in people without diabetes.Trial registration numberNCT00005131.