Association of physical activity and heart rate variability in people with overweight and obesity: A systematic review

Background: Obesity is a major public health issue globally which is intrinsically linked to reduced heart rate variability (HRV). Physical inactivity and reduced resting HRV are linked to an increased risk of coronary heart disease, while athletes have a greater HRV. However, the exact correlation between physical activity and HRV remains uncertain. This systematic review aims to collect, report, and critically assess the current scientific literature about the association between physical activity and HRV in individuals with higher weight and obesity. Methods: A systematic search was carried out in electronic databases (Medline/PubMed, SCOPUS and CINAHL Plus) to retrieve studies that evaluated the relationship between physical activity and HRV in individuals with higher weight and obesity. Case-control, longitudinal/cohort, cross-sectional and observational studies were included. Using a critical narrative approach, information about the HRV, and physical activity was extracted and synthesized. The study was registered in PROSPERO: CRD42020208018 on October 9 2020. Results: After removing duplicates, 980 title/abstract records were checked for eligibility, and 12 papers were finally included in the narrative synthesis. The included studies contained physical activity as well as HRV in adults with higher weight or obesity with or without comorbidities. A negative relationship between moderate to vigorous physical activity and HRV indices had been found in two studies. There was also a negative relationship between sedentary time and HF (p = 0.049) and LF/HF (p = 0.036), as well as a positive relationship between sedentary time and LF (p = 0.014). Also dose-response association was found between vigorous exercise and higher SDNN, LF power, and HF power in one of the studies. Conclusions: This systematic review revealed a wide range of responses to physical activity and HRV; however, the current evidence uses a variety of approaches to objectively assess physical activity and measure HRV with different equipment.


Introduction
Having higher weight or obesity is characterized by abnormal or excessive fat accumulation that poses a health concern leading to chronic medical issues and a greater risk of disability. 1,2 Obesity affects about 40% of young adults aged 20-39 years, 44.8% of middle-aged adults (aged 40-59 years), and 42.8 % of people aged ≥ 60 years globally. 2 The World Health Organization estimated the global economic impact of obesity to be around 2.8% of the world's gross domestic product. However, despite its widespread prevalence, and the accompanying cost, disease burden, and complications, obesity is often ignored as a disease. 3 In the low-and-middle-income nations, uncontrolled urbanization and a shift in eating patterns from traditional to western-style diets are leading to an alarming rise in the incidence of obesity. It is reported that a curvilinear relationship between relative weight and mortality begins after the age of 18 years. 1 Several diseases like diabetes, hypertension, cancer, stroke, osteoarthritis, liver, renal disease, sleep apnea, and depression are linked to obesity, and the association increases with age and the presence of comorbidities. [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21] Another notable dysfunction associated with obesity is the disruption of cardiac autonomic function leading to alterations in the normal parasympathetic and sympathetic regulation, which can be detected using heart rate variability (HRV). [22][23][24][25] HRV is a non-invasive tool for evaluating autonomic function by measuring beat-to-beat differences in R-R intervals. 26 It is known that low HRV is an independent predictor of cardiovascular mortality and sudden cardiac death, is related to higher skinfold thickness, higher body mass index (BMI), higher body fat percentages, and lower levels of physical activity. [27][28][29] Furthermore, reduced physical activity and cardiorespiratory fitness are also potential risk factors contributing to cardiovascular diseases 30 which are worsened by higher associated adiposity indices. 31 Research shows that increased epicardial fat thickness is related to a lack of physical activity, impacting the autonomic nervous system. [32][33][34] Despite these facts, the relationship between physical activity and the autonomic cardiac function as measured by HRV has not been explored well. Accordingly, this systematic review aimed to collate, review and critically assess current scientific and clinical knowledge on the relationship between physical activity and HRV in adults with either higher weight or obesity.

Methods
The study was registered in PROSPERO: CRD42020208018 on October 9 2020.

Search strategy
Search strategies were developed related to physical activity and HRV with the use of boolean operators "AND," and "OR" to allow logical interconnections between concepts. Synonyms or Medical Topic Heading (MeSH) keywords and the title and abstract text for each of the terms were searched. Multiple databases, Medline, SCOPUS, Cumulative Index to Nursing, and Allied Health Library (CINAHL Plus) [Ebscohost], were searched from their earliest record to January 2021. Before adaptation for the other databases, the search strategy was initially built in Medline. For any additional studies, the reference list entries of the included studies and systematic reviews were inspected. We used the following keywords in the quest-exercise, physical activity, incidental physical activity, aerobic fitness, fitness, habitual physical activity, obese, overweight, healthy people, older adults, cardiac autonomic function, autonomic nervous system, heart rate variability, autonomic function, sympathetic function, and parasympathetic function. Table 1 presents the search strategy in detail.

Eligibility criteria
Only case-control, longitudinal/cohort, cross-sectional, and observational studies about physical activity (occupational, transportation, and leisure) and HRV in people with higher weight and obesity with or without any comorbidity were included in this review. A reported mean or median BMI value of ≥ 25 kg/m 2 was also considered for inclusion. The physical activity assessment in the studies could be either subjectively or objectively evaluated. The HRV considered for inclusion was either normalized unit or log-transformed data of time domain or in frequency domain measures; all time-domain measures were reported in ms and frequency domain measures in ms 2 . Studies that failed to examine the results of interest by using only healthy subjects or not including the overweight and obese group were excluded from the analysis.

Study selection
MKS screened all study titles after removing duplicates. Then, MKS and VK independently assessed all relevant titles and abstracts for eligibility, and the full texts of all potentially qualifying papers for independent review were retrieved. Any disputes were addressed by dialogue with the third reviewer, AGM. Articles considered suitable were further processed for data extraction and quality assessment.

Data extraction
The following data were extracted from each qualifying article: study design, study setting, funding source, ethical approval, study population characteristics (age, gender, BMI, body fat percentage), inclusion criteria, obesity grading, sample size, objectives, outcome measures, and compiled in a data-recording table. Initially, MKS and VK piloted the data extraction and qualitative evaluation methods on three papers. MKS conducted data extraction and it was checked by VK. Any conflicts were resolved by contacting the third reviewer, AGM, SK, SU and RS.

Data analysis
A narrative synthesis of the studies was performed while accounting for the differences in the individual investigations regarding their design, findings, and objectives. A meta-analytical approach was not feasible in this analysis due to variations in the measurement of physical activity and HRV and the different study designs.

Results
Literature identified through the search strategy We retrieved 1,902 published papers following a database search based on the keywords described above. After removing the duplicates, 980 records of titles/abstracts were screened for eligibility, of which, 12 studies were finally included in the narrative synthesis of this systematic review. A Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram for the inclusion process is presented in Figure 1. A summary of the 12 studies is presented in Table 2. 23,[35][36][37][38][39][40][41][42][43][44][45] Quality assessment of included studies The quality of the studies was determined by the quality index score. 46 For the present review, the score for the included studies ranged from 16 to 23; however, a few of the points were not scored because they were specific to the intervention study (Table 3). Furthermore, the majority of the included studies did not contain adequate information to understand the role of potential confounders in their relationship with HRV.

Participant characteristics
The participants were aged between 18 and 83 years and all studies included both men and women, except Tonello et al. 37 who included young non-menopausal women with higherweight but in good health. Three of the included articles focused on only people with obesity, 23,35,38 one study had recruited only people with higher weight (Mean BMI = 26.3 AE 4.1 kg/m 2 ) 37 , and another included people with higher weight and obesity 36

Identification of studies via databases
Identification Screening Included Figure 1. PRISMA Flow diagram. participants were with higher weight or obesity, but their mean BMI was >25 kg/m 2 (40-45 kg/m 2 ), and one study included people with a BMI both less than and more than 25 kg/m 2 . 39 Physical activity status of the participants is depicted in Table 4.
The included studies' levels of physical activity were presented in the following, moderate, vigorous, and moderate to vigorous. Figures 2, 3, and 4 provide specifics of the meta-analysis.
HRV and physical activity analysis A brief overview of the studies assessing relationships between HRV and physical activity is presented in Table 5.
Two studies had reported a negative association between moderate to vigorous physical activity and rMSSD. 23,40 One of these also described a negative correlation between sedentary time and HF (p = 0.049) and LF/HF (p = 0.036), as well as a positive association between sedentary time and LF (p = 0.014). 23 Likewise, Föhr et al. 41 observed that both moderate and vigorous physical activity is linked to rMSSD and the LF/HF ratio.
The GCPS, a method to assess physical activity, was reported to be moderately correlated with ln SDNN and ln HF, with correlation coefficients values of 0.61 (p = 0.001) and 0.59 (p = 0.001), respectively. 35 In other three studies, significant correlations were observed between vigorous physical activity (VPA) and rMSSD. 36 One of the studies reported a dose-response relationship between vigorous exercise and higher SDNN, LF power, and HF power (p = 0.05, p = 0.01, and p = 0.01, respectively). 39 Leisure-time behavior was cross-sectionally linked to higher SDNN (p trend =0.001) and higher ULF (p trend <0.0001) according to a study by Soares-Miranda et al. 44 On the other hand, Kluttig et al. found no consistent or statistically significant connection between physical activity and HRV variables. 45

Discussion
This systematic review elucidated the link between physical activity and HRV-derived cardiac autonomic function. We found that rMSSD is negatively correlated to moderate to vigorous physical activity, and sedentary time has a positive    The studies reported that regular physical activity had a positive effect on autonomic control of the heart in adults by decreasing the resting heart rate and an increase in HRV. Further, regular physical activity and aerobic fitness improve a range of health outcomes and lower mortality from all causes. Nevertheless, the use of objective tools to assess physical activity and thorough body composition measurements has been less frequent.

HRV measurement
The included studies used a variety of equipment to record HRV-V800 Polar, ANS watch monitor, GE MC1200 for resting electrocardiography (ECG); RS800CX Polar heart rate monitor, Polar R-R recorder, 12-lead electrocardiogram, Firstbeat Bodyguard device for ambulatory HRV recording; ECG Holter, BIOPAC MP-36 system, 24-hour Holter and Modular ECG Analysis System. However, the use of different equipment to measure HRV in the included studies limits the scope of comparison of results between the studies. Many confounding factors for HRV were identified which included the level of physical activity, recording time, room temperature, individual stress or nutrition, gender, and several hormones. 23,[35][36][37][38][39][40][41][42][43][44][45] The reporting may also be influenced by the HRV data or log transformation of the data.

Physical activity and HRV
Oliveira et al. 23 stated that physical activity and HRV-derived cardiac autonomic activity, rMSSD, have a negative relationship with moderate to VPA and higher levels of MVPA were associated with reduced LF; however, their participants were with severe obesity with only minimal time spent on MVPA (98.92 AE 41.00 min/week). Another important finding was the association between WC and the deterioration of vagal tone. 23 In the same article, sedentary time was linked to all of the frequency domain indices, with a negative correlation with the parasympathetic component and sympathovagal HRV balance, and a positive association with cardiac sympathetic modulation. 23 This suggests that PA may protect the heart, as opposed to increased sympathetic tone. Higher levels of PA are also associated with a favourable modulation of the cardiac sympathetic nervous system in severely obese individuals. As a result, PA reduces the risk of cardiovascular disease associated with obesity through enhancing autonomic function. 23 Similarly, De Liao et al. 32 reported higher cardiac frequency and a decrease in the HF index in elderly persons with higher weight, and a lower HRV in the poor physical capacity group compared to the high physical capacity group.  On analyzing further, Kiviniemi et al. 40 found a negative relationship between rMSSD and MVPA where physical activity was monitored continuously over two weeks, which can be considered representative of the overall current physical activity level. However, the authors stated that cardiorespiratory fitness was a better predictor of cardiac autonomic function than MVPA.
Pope et al. 36 found independent associations between accelerometer-estimated VPA, MPA, and LPA and time-domain HRV measures. Their findings suggested that LPA participation is affordable to a large portion of the population regardless of disease status or age, arguing for more LPA participation to replace sedentary time, highlighting the benefits of LPA. When compared to sedentary subjects, Buchheit et al. 42 found that those with the greatest self-reported MVPA had greater SDNN and rMSSD values as well as higher HF power. According to Soares L et al. 44 prospective study, walking and leisure activities are connected to improved HRV. Additionally, older people who increased their walking distance or pace over a five-year period of follow-up had a more favourable HRV than those who decreased it. Kleiger RE et al. 47 also observed the benefits of physical activity and significant correlations between physical activity and higher SDNN and ULF. 47 Likewise, Kluttig et al. 45 observed that physical activity was inconsistently related with HRV and HRV were closely linked to biomedical risk factors, resulting in a high predictive capacity for potential cardiovascular events like diabetes and obesity.

Conclusion
This review found that physical activity and obesity indices were both independently associated with changes in the cardiac autonomic modulation of obese individuals. Improved cardiac autonomic function has been shown to have positive effects on health outcome, it is crucial to promote it because PA greatly impacts adipocity indices. In this systematic review, we observed a wide range of responses to physical activity and HRV (time and frequency domain variables); however, the existing literature contains a variety of methodologies and equipment for quantifying physical activity and measuring HRV. Furthermore, the evidence for a definite association between different levels of physical activity and HRV is insufficient. Future studies should aim to delve deeper into the influence of intensity and type of physical activity on cardiovascular parameters accounting for all potential regulated confounders. Also, the variability in the equipment used to determine HRV makes it difficult to draw objective conclusions, thereby limiting generalizability.

Ethical approval
Formal ethical approval is not required as this article is a review paper.

Data availability
Underlying data All of the data that underlies the results are included in the article and no additional source data are required.