Abstract Body (Do not enter title and authors here): Background: Genomic data suggest atrial fibrillation (AF) susceptibility and progression may be associated with decreased transcriptional responses to metabolic stress, as might be exerted by obesity and aging. Metformin, nutrient deprivation and exercise are reported to have effects on AMP kinase, a master regulator of metabolic stress.

Hypothesis & Purpose: To determine if upstream targeting of metabolic pathways with metformin and/or weight loss and exercise can reduce AF burden and progression.

Population: Patients with pacemaker (PM) or implantable cardioverter-defibrillator (ICD) devices capable of recording daily AF burden, and at least 5 minutes of AF recorded over a 3-month period.

Methods: In a prospective, randomized, open label, blinded endpoint study, 150 patients were randomized in a 2x2 factorial design to Standard of Care (SOC, including written diet and exercise advice), metformin extended release up to 750 mg twice daily, lifestyle and risk factor modification (LRFM), or both metformin and LRFM. The primary endpoint was a composite of change from baseline to 1 year of average daily AF burden% after a 3-month blanking period and survival at 1 year.

Results: Among 149 eligible randomized patients (39 SOC, 37 Metformin, 35 LRFM, 38 LRFM+Metformin), mean age was 73.7±9.3 years, 58 (39%) were female, and median baseline AF burden was 4.3%. AF burden decreased in the SOC and LRFM arms over time, most significantly in the SOC arm. The primary composite endpoint was no different between SOC and LRFM and LRFM+Metformin arms, but tended to be worse (p=0.033) in the Metformin arm, which had a high rate of discontinuation due to side effects. LRFM+Metformin appeared synergistic compared to each intervention alone (interaction p value 0.011). AF symptom scores improved in the LRFM arms, but did not significantly change in the SOC or Metformin arms.

Conclusions: Metformin to reduce AF burden tended to be worse compared to SOC but was limited by side effects. The decrease in AF burden over time was not significantly different between SOC, which included written materials, and the more contact-intensive LRFM arms, though LRFM was associated with more improved AF symptom scores compared to SOC.