App‐Based Mental Training to Reduce Atrial Fibrillation–Related Symptoms After Pulmonary Vein Isolation: MENTAL AF Trial

Abstract

Background

Even after atrial fibrillation (AF) catheter ablation, many patients still experience relevant symptom burden. The objective of the MENTAL AF trial was to determine whether app‐based mental training (MT) during the 3 months following pulmonary vein isolation reduces AF‐related symptoms.

Methods and Results

Patients scheduled for pulmonary vein isolation were enrolled and randomized 1:1 to either app‐based MT or usual care. Of 174 patients, 76 in the MT and 75 in the usual care group were included in the final analysis. The intervention was delivered by a daily 10‐minute app‐based MT. The primary outcome was the intergroup difference of the mean AF6 sum score, an AF‐specific questionnaire, during the 3‐month study period. Secondary outcomes included quality‐of‐life measures such as the AFEQT (Atrial Fibrillation Effect on Quality of Life). Mean age (SD) was 61 (8.7) years and 61 (41%) were women. The mean AF6 sum score over the study period was 8.9 (6.9) points in the MT group and 12.5 (10.1) in the usual care group (P=0.011). This referred to a reduction in the AF6 sum score compared with baseline of 75% in MT and 52% for usual care (P<0.001). The change in the AFEQT Global Score was 22.6 (16.3) and 15.7 (22.1), respectively; P=0.026.

Conclusions

MENTAL AF showed that app‐based MT as an adjunctive treatment tool following pulmonary vein isolation was feasible. App‐based MT was found to be superior to standard care in reducing AF‐related symptom burden and improving health‐related quality of life.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: NCT04067427.

Nonstandard Abbbrevations and Acronyms

AFEQT

atrial fibrillation effect on quality of life

MT

mental training

PROMIS

patient reported outcome measures

UC

usual care

Clinical Perspective.

What Is New?

• App‐based mental training (mindfulness, breathing, and relaxation techniques) was effective in improving atrial fibrillation–related symptoms as well as health‐related quality of life during the 3 months following pulmonary vein isolation as compared with usual care, likely by changing disease perception.

What Are the Clinical Implications?

• Further research investigating effective complementary treatment methods is needed.

• Because patients with atrial fibrillation experience multifaceted symptoms and distress and show substantial overlap with underlying cardiovascular conditions, implementing such complementary app‐based mind and body approaches holds promise to improve comprehensive atrial fibrillation care and bears important potential for digital interventions that are ubiquitously accessible and scalable.

Atrial fibrillation (AF) as the most common sustained arrhythmia is not only associated with various cardiovascular conditions but also with a multifaceted symptomatology, ranging from almost asymptomatic to severe palpitations and dyspnea. Additionally, about one‐third of patients with AF experience mental health impairments, notably stress and depressive and anxiety symptoms. ¹ AF recurrences may cause further distress and symptom aggravation (and vice versa). ¹ In the majority of patients, AF leads to an impaired quality of life (QoL). ² Catheter ablation for AF (pulmonary vein isolation, PVI) offers an effective treatment option. ² However, >50% of patients may still experience a relevant symptom burden, notably during the 3 months following ablation, the so‐called blanking period, which can be a particularly challenging time for the patients.

Mind and body interventions such as meditation or mental training (MT) and yoga are fundamental methods to reduce stress and stress‐related health issues by relieving physical and mental symptoms alike and improving patient engagement. ³ In the YOGA My Heart study, yoga reduced symptomatic and asymptomatic AF episodes and improved several domains of QoL. ⁴ Very recently, behavioral therapy has been shown to substantially improve AF‐related QoL ⁵ , supporting the necessity for treatment alternatives beyond common pathways.

MT describes a systematic and repeated application of MT techniques and exercises to acquire abilities in stress management and emotional regulation and positively influence interoception. ⁶ It could therefore help to ameliorate arrhythmia perception. Whether MT as a complementary approach in terms of comprehensive AF care has beneficial effects on AF‐related symptoms and QoL has not yet been investigated systematically. The primary objective of the MENTAL AF trial was to determine whether an app‐based MT intervention reduces AF‐related symptoms during the 3 months following PVI based on patient‐reported outcomes using the mean AF6 sum score.

METHODS

Study Design and Patient Population

The MENTAL AF study was an investigator‐initiated, prospective, open‐label, single‐center, randomized clinical trial conducted to assess whether an app‐based MT could improve AF‐related symptoms during the 3 months following catheter ablation compared with usual care (UC). The trial was approved by the ethics committee and conducted in accordance with the Declaration of Helsinki. All patients gave written informed consent. The study protocol is available in Data S1. The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Population

Patients were eligible for participation if they had paroxysmal or persistent, symptomatic AF (EHRA II‐IV) and were scheduled for PVI within the same hospital visit. Further inclusion criteria were the availability of a compatible smartphone, internet access, and capability to use mobile apps. Excluded from study participation were patients under 18 years of age, without appropriate smartphones, or unwillingness to perform MT at least twice per week. The primary follow‐up ended in January 2022.

Workflow and Interventions

A study‐specific website (www.mental‐af.de) had been created to provide study information and obtain patient‐reported outcomes measures. Patients were randomized 1‐to‐1 to either MT or UC with stratification for paroxysmal and persistent AF by a programmed randomization tool. PVI/catheter ablation was conducted according to the center's standard of care protocol and has been reported previously. ⁷ Antiarrhythmic therapy was generally discontinued after ablation.

Patients randomized to MT were instructed to train at least 2 times per week for 3 months. A multimodal audio‐guided MT empirically adapted for patients with AF delivered by the MINDANCE app was applied. Each session's duration was ≈10 minutes. In the first 6 weeks, the interventions used different relaxation techniques to relieve physical and mental discomfort. In the second 6 weeks, the interventions were based on mindfulness exercises aiming to refine a patient's ability to perceive physical sensations, psychological events, and sensory inputs. ⁸ , ⁹ The training protocol is provided in Data S1 and MT background in Data S1.

All patients were instructed to answer the AF6 questionnaire weekly via the study webpage (www.mental‐af.de). At 3‐month follow‐up, patients were initially seen in the outpatient clinic, but due to the pandemic were later approached by email, phone, and postal service; access to the MINDANCE app was then provided to the UC group.

Primary Outcome

The primary outcome was the intergroup difference of the mean AF6 sum scores during the 3 months following catheter ablation (mean score of the first, second, and third month) on a per protocol analysis. Therefore, drop‐out was predefined if MT was performed less than twice weekly on average and if AF6 was reported less than once per month. This 6‐item questionnaire assesses AF‐related symptoms with a recall period of 1 week ¹⁰ : breathing difficulties at rest; upon exertion; limitations in day‐to‐day life due to AF; feeling of discomfort due to AF; tiredness due to AF; and worry/anxiety due to AF. Each item ranges from 0, no symptoms, to 10, most severe symptom, reported for each item. All 6 items are added up to the AF6 sum score.

Secondary Outcomes

Secondary end points were changes of measures of health‐related QoL as assessed by the AFEQT (consisting of a Global Score and the subdomains Symptoms, Daily Activity, and Treatment Concern) and the PROMIS Global Health 10 (Physical Health and Mental Health) questionnaires before and 3 months after AF ablation. ¹¹ Three‐month follow‐up also included a 7‐day Holter (time in AF/atrial tachycardia, and heart rate).

Statistical Analysis

Continuous variables are presented as mean and SD if following parametric distribution, otherwise as median and interquartile range. AFEQT and PROMIS were also presented as median and interquartile range, as changes from baseline did not follow norm distribution. Distribution was tested by the Kolmogorov–Smirnov test. Categorical variables were presented as frequencies and percentages and compared using χ ² test. As predefined, the primary outcome was assessed in all subjects randomized, who underwent AF ablation, trained at least 2 sessions/week in the MT group and who provided at least 1 AF6 per month. The primary outcome was analyzed by Student t test. A general linear model accounting for treatment and subgroup allocation was performed to assess interaction of subgroups with treatment effects with noncategorical variables dichotomized in accordance with the 50th percentile. For intergroup comparisons, the Mann–Whitney U test was used in case of nonnormal distribution and Student t test in case of normal distribution of date. A 2‐tailed P <0.05 was considered significant. All data were analyzed using SPSS version 20 (IBM, Armonk, NY).

Sample Size Calculation

In a previous study using the AF6 questionnaire following AF ablation, ¹² a score of 17 at 6 months was observed. Based on the high AF recurrences rate during the blanking period, 50% higher symptom burden/higher AF6 score at 3 months corresponding to 26 points in the UC group was assumed. According to a 33% symptom reduction by app‐based MT, the AF6 score within the MT group was expected to be 17. The above‐mentioned previous trial using AF6 reported an interquartile range of 28, which translates into an SD of 22 (interquartile range/1.35=SD). Given a power of 0.8 and a probability of 0.05, a sample size of 74 per group was calculated. A drop‐out rate of 10% was assumed for this per‐protocol analysis, suggesting a minimum sample size of n=164. Power calculation was performed by using G*Power Version 3.1.9 (Franz Faul, University of Kiel, Germany).

RESULTS

Between August 2019 and July 2021, a total of 233 patients with symptomatic AF scheduled for catheter ablation of AF were screened, of which 31 did not fulfill eligibility criteria and 19 refused study participation, leaving 183 patients for enrollment (Figure 1). Because some patients did not receive catheter ablation, additional patients had been recruited. Of 183 patients enrolled, 174 received PVI and 151 completed study activities and were included in the final analysis (76 in the MT and 75 in the UC group; Figure 1). The drop‐out rate was 13.2%.

Figure 1. CONSORT diagram.

AF indicates atrial fibrillation; CONSORT, Consolidated Standards of Reporting Trials; and MT, mental training.

Patient Characteristics

Baseline characteristics including underlying conditions were comparable between the groups and are shown in Table 1. Mean age (SD) was 61 (8.7) years, 61 (41%) were women, 83 (55%) had paroxysmal AF, and 47 (31%) underwent prior ablation. The groups did not differ with regard to procedure time (P=0.51) and left atrial low‐voltage areas (P=0.74) as an indicator for more advanced disease (Table 1). PVI was achieved in all patients. There was 1 pericardial tamponade necessitating postprocedural pericardiocentesis in the UC group. Mean weekly MT time was 50±21 minutes, which translates to a practicing frequency of ≈5 times per week (Figure 2). No study‐related harm was reported in either group.

Table 1.

Patient Characteristics

Baseline characteristics	Mental training	Usual care
	n=76	n=75
Age, mean±SD, y	60.6±8.7	61.5±8.7
Sex, no. (%)
Female	32 (42.1)	30 (40.0)
Male	44 (57.9)	45 (60.0)
BMI,* mean±SD	28.0±4.4	28.09±5.6
Comorbidities and risk factor, no. (%)
Diabetes	4 (5.3)	9 (12)
Arterial hypertension	56 (73.7)	53 (70.7)
Dyslipidemia	17 (22.4)	19 (25.3)
History of congestive heart failure	6 (7.9)	8 (10.7)
Coronary artery disease	5 (6.6)	8 (10.6)
Vascular disease	5 (6.6)	3 (4.0)
Stroke, transient ischemic attack	4 (5.3)	4 (5.3)
Sleep apnea	2 (2.6)	6 (8.0)
CHA2DS2−Vasc score†
≤2	50 (65.8)	51 (68.0)
>2	26 (34.2)	24 (32.0)
AF characteristics, no. (%)
Paroxysmal AF	42 (55.3)	41 (54.7)
Persistent AF	34 (44.7)	34 (45.3)
Prior catheter ablation for AF	25 (32.9)	22 (29.3)
Prior direct cardioversion	37 (48.7)	45 (60.0)
β‐blocker therapy	64 (84.2)	61 (81.3)
Prior antiarrhythmic therapy†	30 (39.5)	35 (46.7)
Amiodarone at discharge	2 (1.3)	1 (2.7)
AF on admission	34 (44.7)	34 (45.3)
EHRA class	2 (2, 3)	2 (2, 2.5)
II, mild‐to‐moderate symptoms	47 (62)	56 (75)
III, severe symptoms	29 (38)	19 (25)
Cardiac magnetic resonance imaging, mean±SD
Left atrial dimension, cm2	28.1±7.0	26.2±7.8
LVEDV, mL	148.4±47.2	147.8±36.6
IVS, mm	10.6±2.1	10.9±2.7
LVEF, %	58.0±7.3	57.5±7.8
Catheter ablation procedure
Left atrium low voltage areas‡, no. (%)	19 (25.0)	17 (22.7)
Substrate ablation, no. (%)	18 (23.7)	16 (21.3)
Procedure duration, min	100.7 (43.8)	103.6 (48.4)
High power short duration, no (%)	53 (69.7)	58 (77.3)

AF indicates atrial fibrillation; BMI, body mass index; CHA₂DS₂‐Vasc, congestive heart failure, arterial hypertension, age 65 to 75 years, age 75 years (doubled), diabetes, stroke/transient ischemic attack/thromboembolism (doubled), vascular disease, sex category (female); EHRA, European Heart Rhythm Association; IVS, interventricular septum; LVEDV, left ventricular end‐diastolic volume; and LVEF, left ventricular ejection fraction.

^*BMI calculated as weight in kilograms divided by the square of the height in meters.

^† Antiarrhythmic drug therapy, flecainide, amiodarone, dronedarone.

^‡ Low voltage defined by <0.5 mV in the electroanatomical map.

Figure 2. Mental training time.

 

Primary Outcome

The average AF6 sum score during the 3 months PVI was lower in the MT compared with the UC group (ie, 8.9 [6.9] versus 12.5 [10.1] points; P=0.011; Figure 3A and Table 2). As compared with a single baseline assessment of the AF6, which was higher in the MT group (28.6 [10.4] versus 24.1 [9.8]), both groups reported lower AF6 sum scores throughout the study period. Significant differences between the groups developed after the first month of MT with lower scores in the MT group: 7.8 (7.5) versus 12.1 (10.9) for month 2 and 6.8 (7.1) versus 10.8 (10.5) for month 3, P=0.006, respectively; Figure 3B and Table 2. The effect size was estimated at 0.42 by Cohen's d. Additionally, the overall reduction of the mean AF6 sum score compared with baseline was 75% in the MT and 52% in the UC group (P <0.001). Also, AF6 sum scores ≤6 indicating no or only few symptoms were reported by 65% of MT and 48% of UC patients (P=0.041). In the intention‐to‐treat analysis including 3 patients with insufficient MT, the mean AF6 mean sum score was 12.5±10.1 for UC versus 9.1±7.0 for the MT group (P=0.018).

Figure 3. Mean AF 6 sum scores: primary outcome and time point assessment.

A, Mean AF6 score of 3 months post catheter ablation. B, Mean AF6 score at baseline and at each month following ablation. AF6 sum score ranges from 0 to 60; 0 indicating no symptoms, 60 indicating maximum symptom burden.

Table 2.

Secondary Outcomes: Atrial Fibrillation Effect on Quality of Life Global and Subdomain Scores and PROMIS Global Health 10 and Primary Outcome (AF6 Sum Score)

AFEQT, mean (SD)	Mental training (n=70)			Usual care (n=67)			Intergroup change P value
	Baseline	3 mo	Change from baseline	Baseline	3 mo	Change from baseline
Global score‡	49.6±15.5	72.2±15.5	22.6±16.3*	52.5±18.4	68.3±20.6	15.7±22.1*	0.026
Symptoms‡	52.8±20.1	75.8±21.0	23.0±25.1*	56.5±22.4	75.2±23.5	18.7±16.7*	0.348
Daily activity‡	51.2±21.2	76.2±19.6	25.0±24.3*	51.9±24.7	68.5±28.1	16.6±29.9*	0.030
Treatment concern‡	53.7±20.6	74.6±17.5	20.9±18.7*	58.0±19.8	71.3±21.6	14.3±22.9*	0.066

PROMIS Global Health 10, mean (SD)	Mental training (n=71)			Usual care (n=66)
Physical health§	42.6±6.1	47.3±6.5	4.6±5.6*	44.3±8.0	45.2±7.5	0.8±7.3†	<0.001
Mental health§	44.5 (6.5)	47.4 (6.7)	2.9 (5.7)*	45.8 (7.3)	46.6 (8.2)	1.0 (7.0)†	0.043

AF6 sum score, mean (SD)	Mental training (n=76)	Usual care (n=75)	P value
Mo 1	12.0±8.7	14.6±11.7	0.13
Mo 2	7.8±7.4	12.1±10.9	0.006
Mo 3	6.8±7.1	10.8±10.5	0.006
Average mo 1–3	8.9±6.9	12.5±10.1	0.011

AFEQT indicates Atrial Fibrillation Effect on Quality of Life; and PROMIS, patient‐reported outcome measures.

^*Intragroup baseline to follow‐up comparison P value ≤0.0001.

^† P value >0.1.

^‡ Score range: 0 to 100; score <70 indicates severely symptomatic; 70 to 89, mildly to moderately symptomatic; ≥90 minimally symptomatic or asymptomatic.

^§ T‐score mean of general population = 50 (SD 10).

The average weekly response rate of the AF6 scores for the first to third month was 79.6%, 85.9%, and 85.9%, respectively, in the MT group and 76.0%, 78.7%, and 82.0% in the UC group (P=0.2).

Examination of subgroups based on clinical and demographic characteristics showed that the effect of MT was consistent across analyzed subgroups (Figure 4). Due to numerical differences in some baseline parameters potentially related to worse ablation outcomes in the UC group, a multivariable linear regression analysis was performed showing that group allocation to MT remained significantly associated with an improved AF6 sum score (Table S1).

Figure 4. AF6 sum score subgroup analysis.

Forest plot: the squares represent adjusted mean AF6 sum scores and the bars indicate confidence intervals. AF indicates atrial fibrillation; EHRA, European Heart Rhythm Association; and CHA₂DS₂‐Vasc, congestive heart failure, arterial hypertension, age 65 to 75 years, age 75 years (doubled), diabetes, stroke/transient ischemic attack/thromboembolism (doubled). ^aEHRA Score assesses the severity of symptoms attributable to atrial fibrillation; score range class I, no symptoms to IV, disabling symptoms. ^bCHA₂DS₂‐Vasc: On a scale of 0 to 9, with 0 indicating the lowest risk of stroke, and 9, the highest risk of stroke.

Secondary Outcomes

The AFEQT Global Score and the 3 subdomains improved significantly in both groups at 3‐month follow‐up compared with baseline (Table 2). The improvement was more pronounced in the MT compared with the UC group. The mean change in the Global Score was 22.6 (16.3) in the MT group and 15.7 (22.1) in the UC group (P=0.026). This difference was mainly driven by a more pronounced improvement in the daily activity domain. The improvement in both domains of the PROMIS was significant in the MT group only (Table 2). Physical Health improved by 4.6 (5.5) in the MT as compared with 0.9 (7.3) in the UC group; P=0.001. In Mental Health there was also a more pronounced improvement for MT (Table 2).

In 3‐month follow‐up 7‐day‐Holter, recording time was 153 (36) and 156 (33) hours in the MT and UC groups, respectively, and 10 patients (16.9%) versus 11 patients (14.3%) exhibited AF/atrial tachycardia (P=0.67) referring to a time in AF/atrial tachycardia of 479 (1863) versus 523 (1862) minutes (P=0.37). In the UC group, 35 (53.8%) and in the MT group 25 patients (35.7%) reported heart rhythm–related symptoms during the registration period (P=0.034). Beyond palpitations, symptoms included dyspnea, anxiety, uneasiness, dizziness, and weakness. In the MT group, 24% and in the UC group 31% did not show an arrhythmic correlate (P=0.57). There were no differences regarding heart rate (mean 70.4 (9.2) versus 69.4 (8.7) beats per minute; P=0.53) and heart rate variability with an SD of the normal sinus‐initiated interbeat interval of 119 (32) in the MT and 117 (40) in the UC group (P=0.79), and no difference in systolic or diastolic blood pressure at 3‐month follow‐up (Table S2).

Within the follow‐up period, 9 patients in the MT group (11.3%) and 12 (16%) in the UC group (P=0.48) received electric cardioversion due to AF recurrence or occurrence of atrial flutter. One patient in the MT group reported new‐onset acute myeloid leukemia with symptomatic anemia. One patient in the UC group had a syncope due to sinus arrest and junctional escape rhythm during Holter recording but refused pacemaker implantation.

DISCUSSION

In the first 3 months following AF catheter ablation, the blanking period, recurrences of AF and limiting symptoms are common. ¹³ Thus, this period is typically excluded from the analysis of trials but was the focus of the intervention in the MENTAL AF trial. The key findings of this study can be summarized in the following: (1) Given the overall consenting rate into the trial of 94% of approached patients, app‐based MT post‐AF ablation appeared to be well received and applicable in the majority of patients; (2) MT resulted in significant symptomatic relief in the first 3 months following AF ablation as demonstrated by a reduction of AF6 sum scores reported throughout the study period; (3) the benefits of MT were mainly consistent across subgroups; (4) MT led to a more pronounced improvement in health‐related QoL measures at 3‐month follow‐up as compared with UC.

In recent years, there has been a paradigm shift away from the perception of AF as a primary disease toward understanding AF as a process of its antecedent risk factors. ¹⁴ Analogous with the efficacy of common risk factor management preventing AF, ¹⁵ , ¹⁶ there is a role for addressing mental health–related factors contributing to the occurrence and tolerance of AF, such as severe (perceived) stress and symptom preoccupation. ¹ , ⁵ Mind and body interventions, notably MT/meditation and yoga, have been shown to reduce stress, anxiety, and depression, and to improve QoL. ⁴ , ¹⁷ , ¹⁸ Yoga, in a pre‐post study, was able to additionally reduce AF episodes. ⁴ Recently, Särnholm et al elegantly showed that behavioral therapy improved AF‐related QoL after 3 months as well as heart‐focused anxiety. ⁵ These data emphasize the importance of implementing techniques in ‘usual’ AF care, which sustainably support patients in coping with their AF symptoms.

MENTAL AF explored the role of the generalizable concept of an app‐based MT as an additional treatment tool in the context of standard of care for symptomatic AF, which in most cases is PVI. ² The primary end point of this trial was the mean AF6 sum score during the study period, which deserves some discussion. The concept of repetitive assessment of patient‐reported outcome measures was chosen to reflect the potential effect of MT throughout the intervention time. Greater MT effects were expected to develop over time, considering that MT techniques require practice, and a training effect cannot be expected to arise within just a couple of weeks. Additionally, AF‐related symptoms might be reduced at the end of the blanking period. ¹⁹ Both considerations made a single end point assessment less appropriate in this setting. Also, patients performing MT started their training with a certain delay after receiving ablation.

A central question in patient‐reported outcomes is how much change reflects clinical meaningfulness. The overall reduction in the AF6 sum score was 75% in the MT and 52% in the UC group. The effect size as suggested by Cohen's d of 0.42 is moderate. Data for comparison of the AF6 are limited; in a prior study the median AF6 sum score was 17 (2–30) 6 months after ablation, which is twice the AF6 score of the MT group. ¹² However, this was a single assessment challenging further comparison. A more intuitive assessment of clinical relevance is the comparison of patients with no or few symptoms (AF6 sum score ≤6), which applied to 65% of MT and 48% of UC patients.

MT also led to a more pronounced improvement in AF‐related QoL, which was remarkably low at baseline (mean AFEQT Global Score of 51.0 [17.0]), indicating high symptom burden and quite severe symptoms. Interestingly, this more pronounced improvement of the Global Score was driven by an improvement of the Daily Activity score. Mind and body interventions have in common that health is approached actively, conjoining the physical and mental level and engaging patients to take responsibility for their individual health. Correspondingly, the PROMIS Physical Health domain improved by MT, which corresponds to data in patients with symptomatic compared with asymptomatic arrhythmia. ²⁰ As for the AFEQT, 5 points are considered to represent a clinically meaningful difference, and 2 to 6 T‐score points for the PROMIS. ²¹ , ²² The data (Table 2) indicate that MT significantly improved the AF‐ and general health‐related QoL, even though this effect was less pronounced than after behavioral therapy, ⁵ for which a difference of 15 points in the AFEQT Global Score between the therapy and UC group was demonstrated, ⁵ an even much higher difference as in larger ablation versus drug therapy trials. ⁵ , ²³ , ²⁴

Patient‐reported outcomes are prone to different biases, notably in unblinded trials such as recall or reporting bias, ²⁵ because patients may report toward a presumably more desirable response. On the other hand, there may be increased interoceptive awareness concerning body and mind, leading to worse reported scores. Patients may experience disappointment about their allocation without sham procedure, again influencing reporting. Despite all potential confounding, patient‐reported outcomes measure symptom burden, which is what is meaningful to patients. ²⁵ It is important to keep in mind that there is a known poor correlation of symptoms with documented AF episodes, ²⁶ and patients in AF who believe they are in sinus rhythm appear less symptomatic than patients in sinus rhythm believing they are in AF. ²⁷ Additionally, AF ablation is known to reduce symptomatic AF episodes. ²⁰ Therefore, reducing time in AF/atrial tachycardia is only 1 side of the coin in improving AF health care. Change in perception of AF episodes may add to an improved QoL. ² , ²⁰ Hence, MT might explore its effects of reduced AF6 scores and improved QoL not only by reducing AF‐related symptoms but equally by improved coping with AF. Interestingly, no effect was seen in the AFEQT symptom domain. The latter includes questions that predominantly refer to palpitations/heart rhythm, whereas the AF6 adds limitations and discomfort due to AF as well as worry/anxiety and is a better reflection of coping with AF. Accordingly, in the 7‐day Holter at 3‐month follow‐up, no differences were detected regarding time in AF/atrial tachycardia but in only 10 and 11 patients could episodes be documented. On the one hand, single Holter assessment is limited and even repetitive 7‐day Holter had failed to detect differences in AF recurrences in other trials. ²⁸ , ²⁹ On the other hand, the lack of actual arrhythmia burden difference speaks for a modification of general disease perception and coping by app‐based MT. This hypothesis is supported by the fact that fewer MT patients experienced symptoms during Holter monitoring. Interestingly, about one‐third of all patients reported arrhythmia uncorrelated to general symptoms, potentially hinting toward significant symptom overlap with conditions such as heart failure with preserved ejection fraction but also anxiety, both conditions being prevalent in patients with AF. Additionally, a recent behavioral therapy study found no association between symptomatic improvement and AF burden. ⁵ A more favorable sympathetic–parasympathetic balance and stress reduction represent plausible modes of action. ³ Potentially, the response to MT notably in patients with hypertension may be in favor of this theory, though this cohort was certainly underpowered, not allowing a clear statement. The significant reduction of mean systolic and diastolic blood pressure seen in both groups can most likely be attributed to the different modalities of assessment (in‐hospital versus ambulatory), as well as better blood pressure control after hospital admission. In this postablation cohort, mean systolic and diastolic blood pressure values did not differ between the groups, but they were within normal range at 3‐month follow‐up.

Presumably, informed consent for the ablation procedure as well as for the study has added a patient educational component because the importance of lifestyle optimization on AF recurrences had been addressed, which applies for both groups.

Nevertheless, this trial is hypothesis generating and the exact mechanisms by which MT reduces symptoms remain to be investigated. How the MT effect develops in the long term and whether a maintenance training is required (which is likely to be the case) needs further evaluation. ³⁰

Implementing complementary eHealth approaches holds promise to improve patient care in AF. Larger‐scale trials ideally combining patient education, psychotherapeutic interventions, and MT/meditation programs are desirable.

Limitations

Primarily, this MENTAL AF was neither a blinded trial, nor has there been a sham intervention. As with all unblinded, nonsham trials, it cannot be ruled out that there is an enhanced placebo effect, which is equivalent to true treatment effects. There is also a reporting bias toward the supposedly desired response. It remains unknown whether patients who refused reporting of AF6 scores or who did not perform training were particularly symptomatic or asymptomatic. Baseline AF6 scores were higher in the MT group; however, according to subgroup analysis, this might have instead led to an underestimation of MT effects. Moreover, the fact that both groups received PVI might have further attenuated MT effects. In the MT group, there is a numerically lower number of patients with diabetes, which is a condition known to also cause neuropathy in the autonomic nervous system; thus interoceptive ability may be attenuated. On the other hand, these patients may be more prone to recurrences. However, the number of patient with diabetes is small, and diabetes was not associated with the primary outcome (Table 2).

Furthermore, the observed differences in the mean of AF6 scores between the treatment and control arm was significantly smaller than expected, yet given the narrower SDs observed in our cohort, the assumed effect size and the observed effect size were strikingly similar (assumed effect size 0.41, observed effect size 0.42), resulting in an actual power of the study of 81.6% (as compared with the initially calculated 80.0%).

CONCLUSIONS

MENTAL AF showed that app‐based MT as an adjunctive treatment tool following catheter ablation of AF was feasible. App‐based MT was found to be superior to standard care in reducing AF‐related symptom burden and improving health‐related QoL.

Sources of Funding

This project was funded by Deutsche Herzstiftung (F/49/19) and the Helios Health Institute.

Disclosures

L. Stenzel had a professional and financial relationship with Mindance and was Co‐Founder and Chief Scientific Officer of Mindance. The other authors have no professional or financial relationship with Mindance. All data acquisition was performed independently. The presented app is also available to other research teams on request for study purposes. The remaining authors have no disclosures to report.

Supporting information

Data S1

Tables S1–S2

References 31–35