Exercise Enhances the Effect of Bariatric Surgery in Markers of Cardiac Autonomic Function

Abstract

Background

Bariatric surgery improves cardiovascular health, which might be partly ascribed to beneficial alterations in the autonomic nervous system. However, it is currently unknown whether benefits from surgery on cardiac autonomic regulation in post-bariatric patients can be further improved by adjuvant therapies, namely exercise. We investigated the effects of a 6-month exercise training program on cardiac autonomic responses in women undergoing bariatric surgery.

Methods

Sixty-two women eligible for bariatric surgery were randomly allocated to either standard of care (control) or an exercise training intervention. At baseline (PRE) and 3 (POST3) and 9 (POST9) months after surgery, we assessed chronotropic response to exercise (CR%; i.e., percentage change in heart rate from rest to peak exercise) and heart rate recovery (HRR30s, HRR60s, and HRR120s; i.e., decay of heart rate at 30, 60, and 120 s post exercise) after a maximal exercise test.

Results

Between-group absolute changes revealed higher CR% (Δ = 8.56%, CI95% 0.22−19.90, P = 0.04), HRR30s (Δ = 12.98 beat/min, CI95% 4.29−21.67, P = 0.01), HRR60s (Δ = 22.95 beat/min, CI95% 11.72−34.18, P = 0.01), and HRR120s (Δ = 34.54 beat/min, CI95% 19.91−49.17, P < 0.01) in the exercised vs. non-exercised group.

Conclusions

Our findings demonstrate that exercise training enhanced the benefits of bariatric surgery on cardiac autonomic regulation. These results highlight the relevance of exercise training as a treatment for post-bariatric patients, ensuring optimal cardiovascular outcomes.

Introduction

Individuals with obesity may show cardiovascular autonomic dysfunction, characterized by increased sympathetic activity to the heart [1] and decreased cardiac parasympathetic activity [2]. These maladaptive responses are associated with metabolic dysfunction [3], functional (i.e., increases in heart rate and blood pressure), and structural (i.e., cardiac and vascular remodeling) abnormalities in the cardiovascular system [3–5]. The autonomic disfunction is not only evident in resting conditions but also in response to physiological stressors, such as physical effort. For instance, chronotropic response (CR%) to exercise and post-exercise heart rate recovery (HRR) are blunted in individuals with obesity [6], both being markers of impaired parasympathetic/sympathetic reactivity and independent predictors of cardiovascular events [7, 8].

Bariatric surgery improves heart rate reactivity to exercise, which is indicative of improved cardiac autonomic function [9]. However, surgical intervention alone might not be able to fully restore autonomic regulation. Additionally, the sustainability of the beneficial effects of surgery is likely dependent on the adoption of a healthy lifestyle. Importantly, we showed a reversal of the improvements in metabolic health 9 months following bariatric surgery [10], which might negatively affect CR% and HRR.

In this scenario, exercise emerges as a potential non-pharmacological strategy to sustain or even enhance the benefits of bariatric surgery on cardiac autonomic function. Indeed, exercise training has been shown to increase cardiac autonomic regulation in non-bariatric obese subjects [11], but it is still unknown whether these effects hold true in post-bariatric patients. Here we investigated the effects of exercise training on cardiac autonomic responses to maximal exercise in women undergoing bariatric surgery. Our hypothesis was that exercise training following bariatric surgery would elicit greater benefits in cardiovascular autonomic function, as measured by CR% and HRR, when compared with standard of care.

Patients and Methods

Experimental Design and Participants

Data reported herein are derived from a large randomized controlled trial that examined the effects of exercise training after surgery on overall cardiometabolic risk factors in individuals with obesity who had undergone bariatric surgery (NCT02441361). This dataset was collected between March 2015 and June 2018. All procedures performed were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards and all patients provided written informed consent.

Patients were recruited from the Unit of Metabolic and Bariatric Surgery of the Clinics Hospital of the University of Sao Paulo. Inclusion criteria were the following: women eligible for bariatric surgery (body mass index (BMI) > 40 kg/m² or ≥ 35 kg/m² with associated co-morbidities), 18–60 years of age, and not engaged in an exercise training program for at least 1 year prior to the study. Exclusion criteria involved cancer in the past 5 years, and any cardiovascular, neurological, or musculoskeletal disorders that would contraindicate exercise practice.

Before surgery, patients were randomly assigned (1:1) into either standard of care (RYGB) or exercise (RYGB + ET) group using a computer-generated randomization code. A 6-month, supervised exercise training program started 3 months after surgery. Before (PRE), 3 (POST3), and 9 (POST9) months after surgery, patients performed a maximal exercise test on a treadmill.

Maximal Exercise Test on a Treadmill

The maximal exercise test was performed on a treadmill (Centurion200, Micromed, Brazil), with increments in velocity and/or grade at every minute until volitional exhaustion. Oxygen uptake (VO₂) and carbon dioxide output (VCO₂) were obtained through breath-by-breath sampling and expressed as a 30-s average using a calorimetric system (Cortex-Metalyzer IIIB, Leipzig, Germany). The test was considered maximal when one of the following criteria was met: VO_2plateau (i.e.,< 150 ml/min increase between two consecutive stages), respiratory exchange ratio value above 1.10, heart rate no less than 10 beats below age-predicted maximal heart rate, and rate perceived exertion ≥ 17. Importantly, all patients met at least the RER > 1.1 criterion and the majority of patients met ≥ 2 criteria.

Heart Rate Response During Exercise and Recovery

Heart rate (HR) was continuously recorded at rest (HR_rest), during exercise and recovery using a 12-lead ECG (ErgoPC Elite, Micromed, Brazil). Resting HR (HR_rest) was measured immediately before the test, with the subjects standing on the treadmill. HR (HR_reserve) was calculated as the difference between maximum and resting HR (HR_peak and HR_rest, respectively). CR% was calculated through the following formula: ([HR_peak−HR_rest/220−age−HR_rest]×100). HR recovery was calculated as the difference in HR after 30, 60, and 120 s of recovery in relation to HR_peak (HRR_30s, HRR_60s, and HRR_120s, respectively). A blunted HRR after exercise was defined as a HRR_60s ≤ 12 bpm [7].

Exercise Intervention

RYGB + ET underwent a 6-month, three-times-a-week, supervised exercise training program, whereas RYGB received standard post-surgery care. Exercise training sessions consisted of 5-min light warm-up followed by strengthening exercises for the major muscle groups (leg-press 45°, leg extension, half-squat, bench-press, lat pulldown, seated row and calf raise) and aerobic exercise on a treadmill. Resistance exercise training comprised 3 sets of 8–12 repetitions maximum with a 60-s rest interval between sets and exercises. Load progression (5%) was employed as soon as patients were able to perform two or more repetitions than previously determined. Aerobic training consisted of 30–60 min (10-min progression every 4 weeks) of treadmill walking at an intensity corresponding to 50% of difference between the ventilatory threshold and respiratory compensation point. HR was monitored throughout every session to ensure proper exercise intensity. Any adverse events or signs and symptoms were documented, and a record of attendance was kept ensuring adherence to the protocol.

Statistical Analyses

Data were primarily analyzed using the intention-to-treat-principle. Dependent variables were tested by mixed-model with repeated measures using SAS® version 9.3. Tukey post hoc test was used for multiple comparison testing. Additionally, independent t tests were performed to test possible between-group differences in POST9 to PRE absolute changes (Δ) for all dependent variables. Chi-square test was used to compare the proportion of blunted HRR between groups at POST9. Data were presented as mean ± SD; whenever between-group differences were found, we also reported the estimated mean difference (EMD) between groups and 95%CI. The significance level was set at P ≤ 0.05.

Results

Two-hundred and eleven RYGB-eligible patients were screened for participation; 62 patients met the inclusion criteria (age = 40.6 ± 8.1 years; BMI = 48.3 ± 7.7 kg/m²; VO_2peak = 15.6 ± 2.6 ml/kg/min⁻¹) and were randomly assigned to either a standard of care non-exercise (RYGB: n = 31) or exercise group (RYGB + ET: n = 31). Ten patients from the RYGB group (did not perform surgery: n = 1; personal reasons: n = 9 (7 declined to continue during baseline assessments, 2 non-protocol related health conditions)) and nine from the RYGB + ET (personal reasons: n = 9 (1 lack of interest and 8 participants declined to continue during baseline assessments)) withdrew from the study due to non-protocol-related reasons. Adherence to the exercise program was 81.5 ± 13.1%. No adverse events were reported. Use of β-Blockers was limited to 6.4% for each group (RYGB: n = 2/31; RYGB + ET: n = 2/31). To examine a possible influence of the use of β-blockers on the outcomes, we analyzed the data including or excluding patients taking β-blockers, and data interpretation was not different. Thus, we opted to present the complete data.

HR_peak and CR% were significantly improved after surgery at POST3 (main effect of time: both P < 0.05) and continued to improve by the end of the intervention (i.e., POST9) (main effect of time: P = 0.0001 and P < 0.0001, respectively) (Fig. 1c and g). HR_rest and HR_reserve were not influenced by surgery (at POST3), but HR_rest decreased and HR_reserve increased at POST9 (main effect of time: both P < 0.0001) (Fig. 1a and e). Importantly, between-group absolute changes revealed significantly higher HR_peak (Δ = 11.47 beat/min, CI95% 2.97 to 19.97, P = 0.0098), HR_reserve (Δ = 12.13 beat/min, CI95% 3.24 to 21.02, P = 0.0091), and CR% (Δ = 8.56%, CI95% 0.22 to 19.90, P < 0.05) in the exercised when compared with the non-exercised group (Fig. 1d, f, and h, respectively).

Fig. 1.

Resting heart rate (HR_rest), peak heart rate (HR_rest), reserve heart rate (HR_reserve), and chronotropic response along study. Panels a, c, e, and g show HR_rest, HR_rest, HR_reserve, and chronotropic response, respectively, before (PRE), 3 (POST3), and 9 (POST9) months after bariatric surgery. Panels b, d, f, and h show delta changes scores (POST9-PRE) in HR_rest, HR_rest, HR_reserve, and chronotropic response, respectively. RYGB: non-exercise group; RYGB + ET: exercise group; asterisks denote significant main time effects

HRR_120s were significantly improved after surgery at POST3 (main effect of time: P < 0.0029) (Fig. 2e), but HRR_30s and HRR_60s were not influenced by surgery (at POST3) (main effect of time: P = 0.7768, P = 0.0661, respectively). HRR_30s, HRR_60s, and HRR_120s improved at POST9 when compared with PRE (main effect of time: P < 0.0277, P < 0.0001, and P < 0.0001, respectively). Mixed-model analysis revealed that HRR_60s and HRR_120s were significantly higher in the exercised than in the non-exercised group at POST9 (EMD = 16.48 beat/min, CI95% − 29.06 to − 3.89, P = 0.0036 and EMD = − 23.57 beat/min, CI95% − 38.94 to − 8.20, P = 0.0004, respectively) (Fig. 2 c and e). In addition, between-group absolute changes revealed higher HRR_30s (Δ = 12.98 beat/min, CI95% 4.29 to 21.67, P = 0.0004), HRR_60s (Δ = 22.95 beat/min, CI95% 11.72 to 34.18, P = 0.0003), and HRR_120s (Δ = 34.54 beat/min, CI95% 19.91 to 49.17, P < 0.0001) in the exercised group when compared with their non-exercised counterparts (Fig. 2 b, d, and f, respectively). Then, the proportion of participants showing blunted HRR decreased at POST3 and POST9 in both groups, indicating an effect of surgery. Importantly, the exercised group showed a significantly lower proportion of blunted HRR in comparison with the non-exercised group at POST9, indicating the enhancing effect of exercise (5% vs. 31%, P = 0.0214, respectively).

Fig. 2.

Heart rate recovery in 30 (HRR_30s), 60 (HRR_60s), and 120 (HRR_120s) s after maximal exercise along study. Panels a, c, and e show HRR_30s, 60 HRR_60s, and HRR_120s, respectively, before (PRE), 3 (POST3), and 9 (POST9) months after bariatric surgery. Panels b, d, and f show delta changes scores (POST9-PRE) in HRR_30s, 60 HRR_60s, and HRR_120s, respectively. Panel g shows the percentage of participants with blunted HRR in each group across time. RYGB: non-exercise group; RYGB + ET: exercise group; a: p < 0.05 between POST3 vs POST9. b: p < 0.05 between PRE vs POST9. c: p < 0.05 for between-group comparisons at POST9. Asterisks denote significant main time effects

Discussion and Conclusion

Our results suggest an additive, beneficial effect of exercise training to bariatric surgery on CR% and HRR, which is indicative of improvements in cardiac autonomic regulation in women after bariatric surgery. It is worth mentioning the high prevalence of abnormal HRR prior to the surgery (RYGB: 72%; RYGB + ET: 80%, respectively), which corroborates the cardiac autonomic impairment and increased cardiovascular risks in morbid obesity. These abnormalities were improved by bariatric surgery reinforcing the benefits of this intervention on the cardiac autonomic function. However, our most striking findings were that exercise training greatly enhanced the surgery-mediated improvements in CR% and HRR, as evidenced by the greater increases in CR% and HRR experienced by the exercised patients, as compared with those not engaged in exercise.

Increases in chronotropic response suggest improvements in cardiac beta-adrenergic sensitivity [12] and have been linked with improved autonomic responsivity and reduced sympathetic activity [13], whereas increases in HRR are associated with faster parasympathetic reactivation and/or reduced sympathetic activation following exercise [14]. Importantly, both responses have been associated with improved cardiovascular health and higher survival rates after major cardiovascular events [15]. It is note-worthy that exercise benefits on CR% and HRR took place despite similar group changes in body weight, body fat, lipid profile, or arterial blood pressure, as previously reported [10], strengthening the notion that exercise improves cardiovascular health beyond traditional risk factors [16].

This study is not without limitations. The study duration was relatively short, hampering deeper conclusions regarding the efficacy and feasibility of the intervention in the long run. Additionally, the results reported herein are confined to our patients’ characteristics (e.g., middle-aged women, severe obesity, generally low education level, and economic status). Notably, both HRR and CR% are indirect measures of the cardiovascular autonomic regulation, and future studies targeting direct or mechanistic measures of the autonomic nervous system (e.g., muscle sympathetic nerve activity, pharmacological approaches to assess cardiac and sympathetic tone, or baroreflex sensitivity) are warranted. One may argue that the increase in HR_peak across exercise tests (PRE to POST9) might be an unexpected response and could suggest that the previous tests failed to reach maximal effort; however, it is imperative to note that criteria for maximal test were observed in every testing session, and no differences in RER_peak were observed between groups. Therefore, the increase in HR_peak likely reflects an overall improvement in cardiovascular regulation to exercise triggered by the bariatric surgery and further ameliorated by the exercise training. Finally, HR_rest was measured immediately before the exercise test in the standing position, which makes both HR_rest and HR_reserve potentially affected by exercise expectations. While this arousal response might decrease along the study time points, they should not influence the differences between groups. Despite the completion rate herein being comparable to similar interventions, and the precautionary measures adopted for data analysis, it is not possible to completely rule out any potential bias caused by attrition.

In conclusion, a 6-month, supervised, exercise training program superimposed on the benefits of bariatric surgery on cardiac autonomic regulation. In light of these and other emerging findings [10, 17, 18], practitioners should consider recommending exercise as part of post-bariatric care to ensure optimal cardiovascular outcomes.