Abstract Body (Do not enter title and authors here): Background: Cardiac arrhythmias, particularly atrial fibrillation (AF), are highly prevalent in diabetes due, in part, to cardiac autonomic neuropathy. More frequent AF may contribute to negative cardiovascular outcomes such as stroke and heart failure. It is unknown if episodes of higher glucose are associated with paroxysmal AF occurrences.

Hypothesis: Episodes of higher glucose levels are associated with higher risks of paroxysmal AF onset.

Methods: We invited Atherosclerosis Risk in Communities participants with diabetes at visits 9 and 10 (2021-2023) to undergo 14 days of concurrent continuous glucose monitoring (CGM) and “patch” electrocardiographic (ECG) monitoring regardless of arrhythmia history. Among them, we modelled the odds of episodic AF onset, defined as the transition from sinus rhythm to an AF episode (>30 seconds of irregularly irregular rhythm without P-waves), based on same-hour glucose using multilevel logistic regression, and 4-hour CGM data preceding AF onset using generalized scalar-on-function regression.

Results: Of 211 CGM-ECG study participants, AF was detected in 18. Ten individuals with at least one episodic AF onset (mean age 81 years, 5 females, 3 Black adults) formed the analytic sample where a total of 129 days of ECG data and 11,453 glucose measurements were analyzed. Among individual monitoring periods, the number of AF episodes ranged from 1 to 159, and the median CGM time below (<70 mg/dl), in (70-180 mg/dl), and above (>180 mg/dl) range was 2%, 65%, and 32%, respectively. Using multilevel logistic regression dividing data into distinct one-hour intervals and accounting for diurnal patterns and participant random effects, we found a 4% higher odds of AF onset per 10 mg/dl higher same-hour mean glucose (odds ratio [OR] 1.04, 95% confidence interval [CI] 1.00-1.09, p=0.04; Figure 1A). Generalized scalar-on-function regression with participant random effects and adjustment for time of day found a potentiating association between AF onset and glucose elevations accumulated over 4 preceding hours (per 10 mg/dl consistent elevation over 4 hours, integrated OR 1.17, 95% CI 1.11-1.24, p<0.001; Figure 1B).

Conclusion: Among older, US community-dwelling adults with diabetes, the risk of a paroxysmal AF episode was preceded by higher glucose levels for up to 4 hours as measured by CGM. This finding highlights the importance of continuous monitoring with wearable devices and timely detection of hyperglycemia as a marker of AF risk.