I-STOP-AFib (Individualized Studies of Triggers of Paroxysmal Atrial Fibrillation) Results

Alcohol, but not caffeine, triggered afib, but managing triggers decreased afib episodes by 40%

By Mellanie True Hills

November 18, 2021

  • Summary:  The I-STOP-AFib Triggers Study found that alcohol, but not caffeine, triggered afib and that managing triggers decreased afib episodes by 40%.
  • Reading time: 4–6 minutes

The results of the I-STOP-AFib (Individualized Studies of Triggers of Paroxysmal Atrial Fibrillation) Study were presented at the just-completed American Heart Association Scientific Sessions 2021. This novel study tested triggers among those with symptomatic paroxysmal afib. This patient-initiated research featured a Patient Lead Investigator and a Patient Advisory Board.

Before we talk about the results, let’s talk about how the study came about.

How the Study Came About

You may recall that in 2014, we invited afib community members to participate in the Health eHeart Alliance Patient-Powered Research Summit. Some of you joined us there.

At the summit, we self-selected into interest groups, and there was an afib interest group. In addition, researchers interested in afib joined us to find out what research afib patients felt was needed.

The afib patients identified triggers as an important but unresearched topic. We suggested that understanding afib triggers and whether or not they really triggered afib was an essential question to be answered.

Triggers were considered controversial at the time. Some patients identify things that trigger their afib, and others are unable to do so. In addition, many doctors discount the concept of triggers.

Since testing afib triggers had never been done in a scientifically rigorous fashion, it made sense to try it. Thus, the researchers helped formulate research questions and how to conduct such a trial. At the summit, we had heard about a novel concept, an “n-of-1” study using a smartphone. With an n-of-1 study, each patient could become their own control and assess individual results. Thus, an n-of-1 study was perfect for trigger testing.

The researchers did the heavy lifting of defining the I-STOP-Afib Study and getting it funded. It was funded by PCORI (the Patient-Centered Outcomes Research Institute) and the National Institutes of Health/National Institute of Biomedical Imaging and Bioengineering.

Another novel aspect was the amount of patient involvement. The study included a Patient Lead Investigator (Kathi Sigona) and a Patient Advisory Board (Debbe McCall, Kathleen Sciarappa, Kathi Sigona, Mellanie True Hills, and other patients). Not only did patients help define the study concept, but they also advised, recruited, and helped with the manuscript, becoming study co-authors.

About the Study

When the study started, we asked for your help to fill out a survey to identify potential triggers to test as part of the study. Survey participants came from the Health eHeart Study (UCSF), StopAfib.org (Mellanie True Hills), and the Atrial Fibrillation Support Forum on Facebook (Debbe McCall). As a result, 1295 people with paroxysmal afib responded to the survey. Three-fourths reported triggers, with the most commonly reported triggers being alcohol (35%), caffeine (28%), exercise (23%), and lack of sleep (21%). Those most commonly experiencing triggers tended to be younger, female, or have a family history of afib. Most triggers were modifiable and thus could be tested as part of the study. The trigger survey results were published in Patient-reported triggers of paroxysmal atrial fibrillation, with the four afib patients as co-authors with the researchers.

The I-STOP-AFib Study goal was to improve afib patient quality of life by testing potential afib triggers to help patients predict and prevent afib. We reached out to you again to volunteer to participate in the I-STOP-AFib Study to Identify Your Atrial Fibrillation Triggers. Study participants were again recruited by the Health eHeart Study, StopAfib.org, and the Atrial Fibrillation Support Forum on Facebook.

The study was for those with symptomatic paroxysmal afib. Those in the n-of-1 group used a smartphone app to test triggers they suspected to cause their afib. Controls simply monitored their afib with the smartphone app.

For the first six weeks of each 10 week trial period, the n-of-1 participants were notified weekly to either partake of or abstain from their chosen trigger. They took daily ECG rhythm readings with an AliveCor KardiaMobile device and tracked the frequency and severity of afib episodes. After 6 weeks, n-of-1 participants were shown the results and the likelihood that their trigger influenced their afib. Over the next 4 weeks, they could make any appropriate lifestyle changes and track their afib. At the end of 10 weeks, both groups filled out the Atrial Fibrillation Effect on Quality-of-Life (AFEQT) assessment.

They could then continue tracking other triggers for up to three more successive 10-week periods or stop their study participation. The controls could then track their triggers, too.

Triggers selected for testing across the entire study were: caffeine (n = 53), alcohol (n = 43), reduced sleep (n = 31), exercise (n = 30), lying on the left side (n = 17), dehydration (n = 10), large meals (n = 7), cold food or drink (n = 5), specific diets (n = 6), and other write-in triggers (n = 4).

Study Results

The study results were presented on November 14, 2021, at the American Heart Association Scientific Sessions. In addition, Individualized Studies of Triggers of Paroxysmal Atrial FibrillationThe I-STOP-AFib Randomized Clinical Trial was published the same day (online ahead of print) in JAMA (Journal of the American Medical Society) Cardiology.

This trial of 466 patients found:

  • Alcohol increased the likelihood of an afib episode, but caffeine did not, even though it was the most commonly tested trigger (and could even be protective from arrhythmias)
  • None of the other triggers increased the likelihood of an afib episode
  • The study did not show improved afib-related quality of life as there were no significant improvements in AFEQT scores
  • Those testing triggers had a 40% reduction in the number of afib episodes based on what they learned; this was primarily driven by those who tried caffeine, dehydration, and exercise

This study is important for afib patients. Patient Advisory Board member Kathleen Sciarappa, who helped define the study, spoke for many afib patients:

“What patients really want is a sense of control. Investigating one’s own triggers is a powerful path to learning how to diminish AFib incidences and lead to a better quality of life. Living through the frightening experience of one’s heart racing up to 190 beats per minute, fainting, chest pain, shortness of breath, and the like, barely equals the terror of a five-fold increase in chances of stroke. Learning which triggers set off the chain reaction is paramount for patients wishing to avoid episodes. Patient-driven research could benefit patients with innumerable conditions. We applaud Dr. Marcus for his respectful approach to this condition.”

Unfortunately, the study didn’t find a statistically significant improvement in patient quality of life. However, afib patients know that controlling triggers can improve their quality of life. Some of the reasons the study didn’t show this could be:

  • Those who volunteered for the study, regardless of which group they ended up in, were motivated patients looking for solutions; that could have introduced biases
  • The two groups could have somehow been different in how they experienced afib
  • If some in the n-of-1 group avoided a trigger to avoid having afib or some in the control group were on their best behavior to avoid afib, that could have introduced biases
  • The survey at the end of the 10 weeks could have been subject to recall bias, with specific details from earlier in the 10 weeks not being remembered
  • Patients might have wanted to do well on the surveys, which could have affected their answers

Additionally, the study featured an array of options to allow participants to customize their experiences. However, that may have been too complex; a simpler study may be appropriate for future trigger testing.

Finally, I wonder if some afib may have been missed. For example, suppose some participants have a mix of symptomatic and asymptomatic afib. In that case, it might not have been captured in the study as there was no continuous monitoring. This is a challenge with paroxysmal afib.

Future Implications for Afib Patients

Some things that would have been too complicated to add to this study could be tested in future studies. Or patients can try them on their own. One example might be the effect of dehydration in testing alcohol as a trigger. Could it have been the alcohol itself, or the dehydration caused by alcohol, that triggered the afib? For those willing to test their own triggers, drinking extra water alongside alcohol or decreasing water intake could shed some light on your individual triggers.

Additionally, we have reported patient workarounds for consuming coffee. For example, some patients have shared that they can often drink organic coffee but not non-organic. Thus, it could be pesticide-related. Since this is commonly known in the patient community, was the coffee consumed in this study sometimes organic? (We need to consider gathering this kind of information in future trigger studies to determine factors that could have influenced results.)

Triggers are very complicated, so afib patients will likely continue to test for their own individual triggers. However, additional trigger studies are needed, building on what was learned in this study.


I am grateful to all who made this study possible.

Thank you to the researchers:

  • University of California at San Francisco: Gregory Marcus, MD; Jeffrey Olgin, MD; Mark Pletcher, MD; Ida Sim, MD, PhD; Madelaine Faulkner Modrow, MPH; Shiffen Gettabecha, MPH; Gregory Nah, MA; Sean Joyce, BS; Kelsey Ogomori, BS; Vivian Yang, BA; and Xochitl Butcher, BSA.
  • Brown University: Christopher Schmid, PhD; Jiabei Yang, MS; and Tzu-Chun Chu, MPH.

Thank you to the patients who identified the need and helped with the study: Patient Lead Investigator Kathi Sigona, MA, and Patient Advisory Board members Debbe McCall, MBA, Kathleen Sciarappa, EdD, and other patients (including me).

Thank you, most of all, to you, our community, for participating in the triggers survey and stepping up and volunteering for the study itself. Without you, this study could not have happened.

Links to Selected Media Coverage

Here are some of the media articles published to date about the study. When you review them, please remember that writers generally are not aware of the nuances of living with afib. Thus, they may have different conclusions from you.