The incidence of cardiac arrhythmias during exercise stress testing: a focus on patients with severe obesity undergoing sleeve gastrectomy

Obesity is associated with a higher risk of cardiac arrhythmias. Sleeve gastrectomy (SG) is a common bariatric surgery with beneficial effects on weight loss and comorbidities. The study aimed to investigate the prevalence of arrhythmias during maximal exercise testing in patients with moderate-severe obesity and to evaluate the impact of SG on these arrhythmic events. All patients with moderate or severe obesity who were considered suitable candidates for SG between June 2015 and September 2020 were recruited. Each patient underwent three incremental, maximal, ECG-monitored cardiopulmonary exercise test 1 month before and 6 and 12 months after SG; the frequency and complexity of ventricular premature beats (VPBs) and atrial premature beats (APBs) have been evaluated during rest, exercise and recovery phases. Fifty patients with severe obesity (BMI 46.39 ± 7.89 kg/m2) were included in the study. After SG, patients presented a decreased BMI (34.15 ± 6.25 kg/m2 at 6 months post-SG and 31.87 ± 5.99 kg/m2 at 12 months post-SG). At 6 months post-SG, an increase in VPBs, mainly during the recovery phase, was observed. At 12 months post-SG, a reduction in VPBs compared with the 6 months evaluation was showed. Although in the early post-surgical phase the risk of exercise-induced arrhythmias may be higher, SG does not seem to increase the occurrence of arrhythmias in the long-term. No life-threating arrhythmias were found during post-SG evaluations.


INTRODUCTION
The worldwide increasing prevalence of obesity represents a major health burden and is associated with an increased risk for many metabolic and cardiovascular diseases and comorbidities, such as arterial hypertension, coronary artery disease, heart failure and arrhythmias [1]. Bariatric surgery, and in particular laparoscopic sleeve gastrectomy (SG), is recognized as a significant treatment option for severe obesity, achieving sustainable weight loss and providing positive effects on different metabolic and cardiovascular conditions [2][3][4][5][6][7].
The increased cardiac arrhythmic risk associated with obesity has been previously described, particularly regarding atrial fibrillation (AF) and ventricular arrhythmias [8][9][10]. Obesity has emerged as an independent risk factor for AF, which can be favorably affected by lifestyle interventions leading to weight loss [11,12]. Specifically, bariatric surgery reduces cardiovascular risk factors and results in an improvement of systo-diastolic function with a regression of ventricular hypertrophy [13,14]. Improvement in functional aspects of cardiac alterations after bariatric surgery seems to be associated with a lower incidence of AF but nothing has been described on possible effects on the presence and complexity of ventricular arrhythmias [15].
Obesity is independently associated with increased onset of ectopic ventricular arrhythmia during exercise [16], however, little is known about the incidence of arrhythmic events (AF, ventricular premature beats (VPBs) and atrial premature beats (APBs)) during exercise testing in patients with severe obesity eligible for bariatric surgery as well as the effect of surgery, and the following weight loss, on exertional arrhythmias. Thus, the purpose of this study is to investigate the prevalence of arrhythmias during maximal exercise testing in patients with severe obesity and to evaluate the impact of bariatric surgery on cardiac arrhythmic risk.

MATERIAL AND METHODS Subjects and protocol
This observational cohort study consecutively recruited all patients affected by moderate-severe obesity, who were considered suitable candidates for SG after multidisciplinary evaluation. The Sports and Exercise Medicine Division of the University Hospital of Padova conducted the study between June 2015 and September 2020. All study participants followed a regionally approved clinical pathway and underwent SG (Veneto Region, resolution n.55/ CR August 4, 2015). This study was performed in accordance with the Declaration of Helsinki and approved by the local ethics committee (128n/AO/21). After obtaining written informed consent, a complete functional evaluation including anthropometric data assessment, medical examination and cardiopulmonary exercise testing was performed 1 month before and 6 and 12 months after SG [17,18].
Patients receiving beta-blocker or other antiarrhythmic therapy (including anti-hypertensive drugs with possible antiarrhythmic effects) and/or those with a history of major arrhythmic conditions were excluded from the study. Further exclusion criteria were the rejection of surgical intervention, psychotropic substance abuse, and cardiovascular/orthopedic diseases, which contraindicated or impaired exercise testing.

Cardiopulmonary exercise testing and ECG analysis
Each patient underwent an incremental, ECG-monitored, maximal cardiopulmonary exercise test (Jaeger Masterscreen-CPX, Carefusion). Tests were performed on treadmill (T170 DE, Cosmed), according to the modified Bruce protocol, with an integrated initial 5-min constant speed interval (2.7 km/h). Criteria of exhaustion were a Borg rating of perceived exertion ≥18/20 associated with a respiratory exchange ratio ≥1.10, and/or a peak heart rate (HR) ≥ 85% of predicted maximal HR, and/or the achievement of a plateau of oxygen uptake (VO 2 ).
Continues 12-lead ECG monitoring focused on detecting APBs and VPBs during each test phase (i.e., rest, exercise and recovery). Premature beats (PBs) were then categorized according to number, frequency and complexity (isolated or repetitive). QTc was calculated using the Bazett's formula.

Statistical analysis
Statistical analyses were performed with Statistical Package for Social Science (SPSS Inc., Version 20, IBM Corporation, Armank, NY). Continuous variables are expressed as mean ± standard deviation or median and interquartile range; comparisons between the same population before and after SG (1 month pre-SG, 6-and 12-months post-SG) were performed with the one-way analysis of variance test with Bonferroni correction or Friedman test. Categorical variables are expressed as percentages and comparison between the same population before and after SG was performed with McNemar test. All reported probability values are twotailed and a value of p < 0.05 was considered statistically significant.

RESULTS
A total of 50 patients with severe obesity (23 men; 27 women, BMI 46.39 ± 7.89 kg/m 2 ) were included in the study, with a mean age of 46.78 ± 12.42 years. At baseline, 13 patients (26%) were affected by type 2 diabetes mellitus and medically treated with oral hypoglycemics, 15 patients (30%) were treated for arterial hypertension and 7 patients (14%) were affected by sleep apnea syndrome. Anthropometric and baseline ECG characteristics of patients at clinical evaluations are shown in Table 1.
Six months post-SG Six months post-SG, patients lost averagely 35.24 ± 11.19 kg, i.e., 25.90 ± 5.34% of body weight, leading to a mean BMI of 34.15 ± 6.25 kg/m 2 . Post-SG, patients presented lower resting HR and QTc with increased exercise test duration (all p < 0.001).
Two patients (4%) presented repetitive APBs and two patients VPBs during pre-surgical evaluation (two patients presented a supraventricular couplet and two patient a ventricular couplet). Three patients (6%) presented repetitive APBs (two patients presented a supraventricular couplet and one patient presented six supraventricular couplets) and seven patients (14%) presented repetitive VPBs (three patients presented a ventricular couplet, one patient presented seven ventricular couplets, two patients presented a ventricular triplet and one patient presented 11 ventricular couplets and a short tract of non-sustained ventricular tachycardia of 7 beats) during post-SG evaluation. No patients presented AF during pre-surgical evaluation while one patient presented AF post-SG. No patients had sustained ventricular tachycardia.
The qualitative and quantitative analysis of the arrhythmic events during the complete test showed significant difference between the evaluations pre-SG and 6 months post-SG. The prevalence of VPBs was significantly increased during post-SG evaluation, in particular during the recovery phase, both by number of patients with at least one arrhythmic event and by the frequency of arrhythmic events per minute (Table 2 and Fig. 1).
Twelve months post-SG At 12 months post-SG evaluation, patients showed further weight loss (31.87 ± 5.99 kg/m 2 ; p < 0.001). No further reductions in HR, QTc, or significant increases in exercise time were evident. Two patients (4%) presented repetitive APBs (two patients presented a supraventricular couplet) and six patients (14%) presented repetitive VPBs (three patients presented a ventricular couplet, two patients presented two ventricular couplets, one patient presented four ventricular couplets). No patients presented AF or ventricular tachycardia. At 12 months post-SG, a reduction in the number of patients who presented at least one arrhythmic event is evident, in particular VPBs during the recovery phase (p = 0.003; Table 1). At 12 months post-SG, linear mixed models revealed a Continuous variables are expressed as mean ± standard deviation or median and interquartile range.
APBs atrial premature beats, BMI body mass index, HR heart rate, VPBs ventricular premature beats. significant decrease in VPBs frequency when compared to the 6 months post-SG evaluation, mainly during the recovery phase (Fig. 1).

DISCUSSION
We investigated the prevalence of cardiac arrhythmic events in patients with moderate-severe obesity during cardiopulmonary exercise testing before and after SG. To the best of our knowledge, no previous study analyzed the association between patients affected by severe obesity undergoing SG and exercise-induced cardiac arrhythmias. The main outcomes of this study are: • Six months after SG, patients presented lower HR, QTc interval and higher exercise tolerance. These changes persist 12 months after bariatric surgery.  Fig. 1 Arrhythmic events during maximal exercise testing in patients who received functional evaluations 1 month before, 6 and 12 months after sleeve gastrectomy (SG). At 6 months post-SG there is a global increase in ventricular premature beats (VPBs)/min, particularly in the recovery phase. At 12 months after SG VPBs/min appear reduced when compared to the 6 months follow-up evaluation. No differences in atrial premature beats (APBs)/min were detected. The red box plots refer to the complete test while the blue box plots refer to the different exercise test phases. * = Significant difference between 1 month pre-SG and 12 months post-SG vs. 6 months post-SG (both p < 0.001).
• Six months after SG, patients presented an increased occurrence of VPBs, mainly during the recovery phase.
• Twelve months after SG, patients showed a reduction in the frequency of VPBs compared to 6 months post-SG, mainly during recovery phase, similar to the pre-surgical evaluation.

SG and ECG parameters
Patients affected by obesity present usually higher resting HR compared with subjects with normal weight, probably due to a reduction in parasympathetic activity and relative predominance of sympathetic activity [19]. Weight loss after bariatric surgery has been shown to decrease resting HR and affect HR variability [20]. Moreover, several studies showed that patients with obesity have a prolonged QT interval, highlighting the effects of bariatric surgery on reducing QTc [21][22][23]. In our study, resting HR and QTc interval were found significantly decreased at 6 months post-SG and further changes are not evident 1 year after bariatric surgery. The relationship between obesity and prolonged ventricular repolarization is not clearly understood and various hypotheses have been proposed. Typical obesity related comorbidities such as insulin resistance, arterial hypertension and obstructive sleep apnea syndrome could be involved in shortening ventricular repolarization parameters [24][25][26]. All these comorbidities are represented in our study population before SG and their reduction after bariatric surgery could influence the ECG repolarization pattern. Moreover, patients post-SG showed a longer exercise phase when compared to their own pre-SG testing. This data is in accordance with previous study results that showed how patients with severe obesity have a better exercise tolerance 6 months after bariatric surgery [27]. SG has a positive impact on body composition, physical functioning and metabolic parameters, allowing to perform the same workload with less energy expenditure, also due to improved cardiac reserve [28].

SG and arrhythmic risk during exercise testing
Our analysis focused on the comparison of arrhythmias during maximal exercise testing before and after SG. It is known that obesity is associated with an increased risk of cardiac arrythmia including AF [29,30], PBs [10,16] and sudden cardiac death [31,32]. Moreover, several studies showed the importance of weight reduction in patients with obesity for the treatment of AF, while others report little or no effectiveness [11,12,15,33,34]. In our study only one patient presented a short-run AF at 6 months post-SG and none presented an exercise-induced AF, therefore, it was not possible to analyze the effect of SG on this form of arrhythmia. However, P-wave dispersion and QTc dispersion showed to be attenuated after bariatric surgery, indirectly suggesting a reduction in risk of AF, ventricular arrhythmias, and sudden cardiac death [21-23, 35, 36].
At 6 months post-SG, despite a major reduction in BMI by all patients, an increase in ventricular arrhythmias, mainly during the recovery phases, was showed. The same phenomenon does not seem to be confirmed for the rest and exercise phases. This behavior seems to contradict some of the evidence on arrhythmias and obesity since an independent association between BMI and exercise-induced ventricular arrhythmias has already been described in a large cohort of patients [16]. Indeed, several hypotheses (anatomical, immunological, hormonal and multifactorial) have attempted to explain why patients with obesity are more prone to arrhythmic onset. Different studies agree that atrial and ventricular ectopy activation could be due to the structural and electrical remodeling that increase the risk of AF and VPBs [10,37]. Moreover, the epicardial fat presents a pro-inflammatory profile, promoting chronic inflammation of the adjacent cardiac tissue, which may potentially increase the risk of ventricular arrhythmias [32,38]. A further hypothesis involves the possible role of leptin in increasing sympathetic activity and reducing vagal tone [39]. This hormone has been demonstrated to correlate with BMI and it has been associated with an increase in mean arterial blood pressure and HR in animal models, due to alterations in the autonomic nervous system [40,41].
At 12 months post-SG evaluation, an overall reduction in number of patients presenting VPBs and in VPBs/min, still primarily during the recovery phase, was observed. The prevalence of arrhythmias seems similar to 1 month pre-SG evaluation. A possible explanation for this increase and subsequent decrease in arrhythmias during the recovery phase could be due to the imbalance between the circulating catecholamines produced during exercise and the raising of vagal tone, responsible for lowering HR after exercise. An attenuated vagal reactivation might be associated with a minor suppression of ventricular activity and more frequent ventricular arrhythmias during recovery. This was also described as an important and independent predictor of increased mortality risk [42]. A previous study conducted by our research group showed that during the recovery phase HR decreases more rapidly in patients at 6 months after SG, when compared to the pre-surgical evaluation [27]. In fact, in these patients there is no difference in the maximum HR reached during exercise testing between evaluations pre and post-SG, while the resting HR appears lower at 6 months after surgery; therefore, a greater vagal effort is required for the complete recovery after exercise [27]. Results of the current study are in line with those already described and provide an implementation of data evaluating also cardiovascular adaptations 1 year after bariatric surgery. Indeed, at 12 months post-SG, the reduction of arrhythmic events during recovery phase might suggest a restoration of sympathetic-vagal balance related to weight loss maintenance and metabolic equilibrium.
Moreover, a further hypothesis could be related to the nutritional alterations that these patients have to undergo in the first period after SG [43]. In fact, bariatric surgery patients are at risk for deficiency of different nutrients, especially of trace elements, despite taking supplements [44,45]. The transition from strong catabolism during the first period after bariatric surgery to progressive anabolic recovery can cause nutritional alterations such as micronutrient deficiency (e.g., thiamine), intracellular and extracellular electrolyte disturbances (hypophosphatemia, hypokalaemia, hypomagnesemia), fluid imbalance and salt retention even in the cardiac tissue [46]. Thus, these subjects might be exposed to an increased risk of cardiac arrhythmias, in a similar way to those documented in the "refeeding syndrome" [47]. It is possible that restoring a regular diet and maintaining a stable weight will help to reduce the arrhythmic susceptibility during the first year after surgery.

Limitations and perspectives
This study focused on the first 12 months post-SG, a period in which it is observed a larger weight loss and patients have dietary restrictions and limitations on their lifestyle, especially in the first months after bariatric surgery. Thus, future studies should provide a standardized long-term follow-up postsurgery to a wider sample size, to evaluate the arrhythmic burden after more sustained and stable weight loss. Moreover, although body composition analysis has not been performed in this study because of methodological issues, future trials may investigate whether the arrhythmic burden is also associated with changes in body composition after SG. Given the lack of studies regarding weight loss and the risk of cardiac arrhythmias, further research is still needed, also with longer registration methods. By way of example, to investigate if the different treatment options for obesity, such as bariatric surgery, pharmaceutical drugs or dietary interventions, have a different influence on the incidence of arrhythmic events.
Finally, also the impact of timing, rapidness of metabolic adaptations and treatment phases should be considered by future trials.

CONCLUSION
This study provides further information about the relationship between severe obesity and arrhythmic risk during maximal exercise testing. SG seems to slightly increase the global occurrence of ventricular arrhythmias at 6 months after surgery, mainly during the recovery phase. However, at 12 months post-SG the arrhythmogenic susceptibility improves with more stable metabolic conditions returning comparable to the pre-surgical evaluation. Finally, no life-threating arrhythmias were found post-SG, suggesting that bariatric surgery does not represent a risk factor for the onset of relevant arrhythmic events in these patients. Further studies with longer follow-up after surgery considering also the different effect of pharmaceutical drugs or dietary interventions on the incidence of arrhythmic events are needed.

DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.