Mobile health for walking on the tightrope of optimal physical activity to reduce the risk of atrial fibrillation

This editorial refers to ‘Accelerometer-derived physical activity and risk of atrial fibrillation’, by S. Khurshid et al., doi:10.1093/eurheartj/ehab250.

Risk factor and lifestyle optimization has public health relevance for the prevention of atrial fibrillation (AF) due to the high number of affected individuals and the high morbidity and mortality of the disease. In this context, the relationship between physical activity and AF has been of interest for a long time, but the potential benefit and harm derived from different levels of regular physical activity are still largely unknown. Extreme physical activity increases the risk for AF in endurance athletes. Among long-distance cross-country skiers who completed the Vasaloppet 90 km competition, those who participated repeatedly had the highest long-term AF incidence.¹ Several adaptions of the athlete’s heart observed in men and animal models have been reported in relation to the risk of AF, such as larger hearts with increased volume load and stretch, in particular of the atria, enhanced vagal tone, sinus node impairment, bradycardia, atrial fibrosis, and inflammation.²

Moderate physical activity, in contrast, appears to be associated with a lower risk of AF and is recommended for the prevention of AF.^{3  ,  4} The effects of low to moderate regular exercise in the population are difficult to differentiate from the potential effects of lifestyle and other correlated cardiovascular risk factors such as obesity. Immobility reflects comorbidities such as heart failure and sleep apnoea, as well as frailty, and may thus be a marker of increased risk of developing AF. However, there are direct effects, too. In patients with a body mass index ≥27 kg/m² and manifest AF, an increase in cardiorespiratory fitness ≥2 metabolic equivalents (METs) was related to a lower recurrence of AF after 5 years in a risk factor management programme.⁵

Understanding the U-shaped relationship between physical activity and AF risk has been a challenge. A hurdle was, until now, the correct quantification of physical activity in populations. Self-reported activity estimates are known to be unreliable. A recent meta-analysis that normalized physical activity to the MET of tasks reported that guideline-recommended levels are associated with a significantly lower AF risk, with remaining uncertainty with regard to the effect of very high physical activity.⁶ Frequency of physical activity, e.g. number of days per week, duration of exercise periods, and intensity certainly play a role. In addition, temporal and intensity patterns of physical activity are highly likely to alter their cardiovascular effects.

Quantifying physical activity objectively

In this issue of the European Heart Journal, Khurshid and colleagues finally overcome the imprecision of prior studies: They identified >90 000 individuals in the UK Biobank with analysable data from wrist accelerometers worn over 1 week. ⁷ The study provides a step-change in our understanding of the relationship between physical activity and AF. It also illustrates the power of open access to data pursued by the UK Biobank.

The analysis by Khurshid et al. confirms that physical activity approximating the threshold of moderate to vigorous physical activity, recommended by cardiac societies and the World Health Organization, is associated with reduced AF risk. Longer periods of more intensive physical activity were associated with incident AF and stroke over a 5-year follow-up. Not unexpectedly, the agreement between the objectively quantified physical activity and self-reported physical activity collected using the international physical activity questionnaire was weak. The self-reported physical activity was not significantly associated with either AF or stroke in the subsample with physical activity measurements, although in the >400 000 individuals from the overall UK Biobank the association reached statistical significance for AF.⁸ The current analysis advances research into physical activity and AF from the ordinal to the numeric: quantifying activity, even if only over a week, clearly provides more reliable and quantifiable information than asking participants about their activity levels. Understanding the optimal dose and pattern of physical activity will be advanced substantially by such quantification. Fortunately, quantifiable estimation of activity is now readily available via wearable devices.

The findings of the report by Khurshid and colleagues fit into the emerging U-shaped relationship between physical activity and AF risk. The extremes of little or no physical activity as well as high levels of activity are not well represented in the UK Biobank. Nonetheless, a mild increase of AF risk at the lower and upper tails can be seen in this cohort.⁹ The UK Biobank enrolled middle-aged men and women, and is thus not very likely to include many competitive athletes. Studies using objectively measured physical activity in physically active populations such as athletes are thus desirable to determine the tipping point where vigorous exercise offsets the beneficial effects on other AF risk factors. In addition, despite adjustment for AF risk factors, the mechanisms by which exercise reduces the risk of incident AF remain unknown. General effects on cardiovascular health compete with atrial-specific effects. This may in the future be addressed by research into quantifiable cardiovascular biomodulators associated with AF, e.g. natriuretic peptides: these are increased in response to acute aerobic exercise, probably reflecting atrial dilation, but decrease after chronic aerobic endurance training, possibly due to lower peripheral vascular resistance.¹⁰

From quantitative diagnostics to mobile health interventions

The population-wide use of an accelerometer provides a glimpse into future opportunities to estimate and alter health-modifying behaviour. With >100 000 mobile health apps and >400 wearable activity monitors currently available,¹¹ mobile technology offers new and affordable possibilities for prevention and patient care. Integrating information from complementary biosensors increasingly allows fine-tuning of lifestyle and risk factors in everyday life for risk optimization through semi-continuous quantification, evaluation, and instantaneous, possibly motivational, feedback. Clearly, there is an urgent need to evaluate the benefit and harm (e.g. overdiagnosis and overtreatment, or anxiety based on misinterpretation of findings) of these technologies in a cardiovascular context.

Whether mobile health applications can improve risk stratification and timing of interventions such as starting oral anticoagulation, and whether these tools can be used to inform populations about the health impacts of their behaviour needs to be demonstrated in clinical trials. An example for testing the use of mobile health monitoring is the Smart in OAC case-finding study assessing the feasibility and efficacy of a cloud-based analytical service in combination with a pulse plethysmographic wristband in detecting absolute arrhythmia (NCT04579159). The app will also be used to validate and enhance clinical and risk factor information on the participants captured during the monitoring period.

Thus, digital devices may provide simple, low-threshold access to screening technology targeting at-risk populations and patients. They can provide feedback on health behaviours and possibly help change risk factor profiles and improve outcomes through optimized, more continuous, and fast feedback loops (Graphical Abstract).

Graphical Abstract.

Overview of the potential role of mobile health applications and wearables for (big) data collection in research and improvement of lifestyle and risk factors through monitoring, quantification, and direct feedback to the user according to individualized targets. As an example, a hypothetical graph for physical activity is shown with a U-shaped association.

The study by Khurshid and the open access policy of the UK Biobank illustrate how quantitative activity monitoring can benefit research into AF. Their study broadens our thinking about mobile health, and illustrates how much well-conducted clinical research is needed to estimate the benefits and risks of mobile health-based interventions. We need to encourage controlled trials of mobile health and to continue a broad discussion about the ethical and judicial rules for using such technology in societies deeply grounded in freedom and human rights.

Conflicts of interest: P.K. is partially supported by European Union BigData@Heart (grant agreement EU IMI 116074), the British Heart Foundation (FS/13/43/30324; PG/17/30/32961 and PG/20/22/35093; AA/18/2/34218), the German Centre for Cardiovascular Research supported by the German Ministry of Education and Research (DZHK), and the Leducq Foundation. He receives research support for basic, translational, and clinical research projects from European Union, British Heart Foundation, Leducq Foundation, Medical Research Council (UK), and German Centre for Cardiovascular Research, from several drug and device companies active in atrial fibrillation, and has received honoraria from several such companies in the past, but not in the last 3 years. He is listed as inventor on two patents held by the University of Birmingham (Atrial Fibrillation Therapy WO 2015140571, Markers for Atrial Fibrillation WO 2016012783). The other authors have no conflicts to declare.

 

The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.