^{1}

^{*}

^{1}

^{1}

^{2}

^{1}

^{2}

Edited by: Jerzy Sacha, Regional Medical Center, Poland

Reviewed by: Jerzy Sacha, Regional Medical Center, Poland; George E. Billman, The Ohio State University, USA

*Correspondence: Catharina C. Grant, Section Sports Medicine, University of Pretoria, P.O. Box 37897, Faerie Glen 0043, Gauteng, Pretoria, South Africa e-mail:

This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Quantification of cardiac autonomic activity and control via heart rate (HR) and heart rate variability (HRV) is known to provide prognostic information in clinical populations. Issues with regard to standardization and interpretation of HRV data make the use of the more easily accessible HR on its own as an indicator of autonomic cardiac control very appealing. The aim of this study was to investigate the strength of associations between an important cardio vascular health metric such as VO_{2}max and the following: HR, HRV indicators, and HR normalized HRV indicators. A cross sectional descriptive study was done including 145 healthy volunteers aged between 18 and 22 years. HRV was quantified by time domain, frequency domain and Poincaré plot analysis. Indirect VO_{2}max was determined using the Multistage Coopers test. The Pearson correlation coefficient was calculated to quantify the strength of the associations. Both simple linear and multiple stepwise regressions were performed to be able to discriminate between the role of the individual indicators as well as their combined association with VO_{2}max. Only HR, RR interval, and pNN50 showed significant (_{2}max. Stepwise multiple regression indicated that, when combining all HRV indicators the most important predictor of cardio vascular fitness as represented by VO_{2}max, is HR. HR explains 17% of the variation, while the inclusion of HF (high frequency HRV indicator) added only an additional 3.1% to the coefficient of determination. Results also showed when testing the normalized indicators, HR explained of the largest percentage of the changes in VO_{2}max (16.5%). Thus, HR on its own is the most important predictor of changes in an important cardiac health metric such as VO_{2}max. These results may indicate that during investigation of exercise ability (VO_{2}max) phenomena, quantification of HRV may not add significant value.

It is known that cardiovascular disease (CVD) has reached pandemic proportions worldwide. It is identified as one of the most important causes of death in several countries including USA, Pacific and Middle Eastern countries, China and Europe (Barakat et al., _{2}max, results in a reduction in CVD risk (Barakat et al.,

The link between cardiac autonomic neuropathy (increased sympathetic, decreased parasympathetic activity or a combination of the two), and CVD is well established (Lahiri et al.,

Quantification of HRV, integrating the sympathetic and parasympathetic cardiac influences, is a complex measurement of the autonomic nervous system and its responses to internal and external stimuli (Lahiri et al., _{2}max and HR, as well as between VO_{2}max and HRV indicators including RMSSD, pNN50, LF, HF, LF/HF, SD1, and SD2 (abbreviations explained in Table _{2}max than standard HRV indicators. Regression analyses were employed to determine the relative importance of HR and HRV indicators as predictors of an important cardio vascular health metric such as VO_{2}max.

Mean RR (s) | The mean of the intervals between successive QRS complexes |

RMSSD (ms) | Root mean square of the standard deviation between RR intervals |

pNN50 (%) | The percentage of successive RR interval differences over the entire measurement larger than 50ms |

LF Power (ms^{2}) |
Peak between 0.04 and 0.15Hz |

HF Power (ms^{2}) |
Peak between 0.15 and 0.5Hz |

LF/HF | LF Power (ms^{2}) divided by HF Power (ms^{2}) |

SD1 (ms) | Indicator of the standard deviation of the instantaneous RR variability |

SD2 (ms) | Indicator of the standard deviation of the continuous or long term variability of the heart rate |

The primary hypothesis was that HR and RR intervals show stronger correlations with VO_{2}max than the HRV indicators, and are on their own stronger predictors of VO_{2}max than the more complex indicators of HRV (RMSSD, pNN50, LF, HF, LF/HF, SD1, and SD2). A secondary hypothesis was that, when including HR and HR normalized indicators of variability (RMSSD/RR, LF/RR^{2}, HF/RR^{2}, LF/HF, SD1/RR, SD2/RR) regression analysis will show that HR is still the most important predictor of VO_{2}max.

The gold standard in fitness testing is generally regarded as the VO_{2}max test for measuring aerobic fitness or cardiorespiratory endurance. VO_{2}max is defined as the maximum volume of oxygen that can be utilized in 1 min during maximal exercise. It is measured in millilitres of oxygen used in 1 min per kilogram of body weight, (ml/kg/min) (Wilmore and Costill, _{2}max values and it is thought that the more oxygen the body can use during strenuous exercise, the more energy it can produce (Wilmore and Costill,

VO_{2}max can be measured directly or indirectly. The direct measurement needs to be done in a laboratory and is expensive, time consuming and requires special expertise. The test subject must reach his/her maximum work capacity in order to determine VO2 max accurately. The test is done on a treadmill or a bicycle and a strict protocol is followed. The athlete is required to exercise while the speed and intensity is gradually increased. The volume and oxygen concentration of the inhaled and exhaled air of the athlete is measured to determine how much oxygen is used. The oxygen consumption usually rises in direct relationship to the increase in exercise intensity up to a certain point. At this point the oxygen consumption reaches a plateau and does not rise further even with a further rise in intensity. This plateau marks the VO_{2}max.

VO_{2}max can also be determined indirectly, as an estimate of the true VO_{2}max. The Cooper 12 min run test was used in the current study. This assessment uses a set period of time (12 min) and scoring according to distance (in meters). The participants were asked to run or walk for 12 min as fast as possible (Cooper, _{2max} from the distance covered at the end of the 12 min period was obtained by applying the distance run to the Cooper regression equation: VO_{2}max (ml/kg/min) = 0.0268 (distance covered in meters)—11.3 (Cooper,

The sedentary person has a much lower VO_{2}max generally compared to the active fit individual, with the elite endurance athlete having the highest VO_{2}max. Genetics play a role in an individual's exercise ability but VO_{2}max can be improved with physical training. The more unfit, the more the VO_{2}max can be improved. As exercise capacity improves, skeletal muscle strength and endurance also improve (Paterson et al.,

A hypotheses driven, cross sectional, descriptive study was performed. The study protocol was submitted and approved by the Ethics Committee of the University. A total of 145 healthy participants from a group of 235 volunteers were accepted to take part in the research study. Exclusion criteria included refusal to give voluntary written informed consent; a history of cardiovascular, hepatic, respiratory, or renal impairment, as well as pulmonary, metabolic, and orthopaedic diseases requiring medical attention; lung/respiratory tract infection in the previous 2 weeks; and medication that could influence cardiovascular control and psychological disorders. None of the participants were professional athletes or high-level sport participants. All participants gave written informed consent before commencement of the intervention. Participants fasted overnight and were asked not to use any caffeine, alcohol or to smoke 24 h prior to the measurements. Measurements were taken in a temperature regulated, quiet environment, in the morning before 12H00. The Polar 810E HR monitor system was used to record supine RR intervals for a 10 min period. The 5 to 10 min period of this recording was used to quantify the HRV. Computer software from the University of Kuopio, Finland calculated the HRV indicator values by time domain analysis (RR, STDRR, RMSSD, and pNN50), frequency domain analysis (LF, HF, and LF/HF), and Poincaré plot analysis (SD1 and SD2) (Mourot et al.,

The Pearson's correlation coefficient (

To test the hypotheses, linear regression was used to determine the most important predictors of changes in cardio vascular fitness and exercise ability as represented by VO_{2}max. Indicators included were: HR, RR interval, RMSSD, pNN50, LF, HF, LF/HF, SD1, SD2. The following HR normalized HRV indicators were also included in a separate set of regression analyses: RMSSD/RR, LF/RR^{2}, HF/RR^{2}, LF/HF, SD1/RR, and SD2/RR. Both simple linear and multiple stepwise regressions were performed to be able to discriminate between the role of the individual indicators as well as their combined relationship with VO_{2}max (Field,

A summary of the participants (

_{2}max

BMI (kg/m^{2}) |
23.07 | 2.36 | 22.62 |

VO_{2}max (ml/kg/min) |
49.54 | 8.79 | 53.53 |

2.4 km run time (s) | 733.71 | 179.10 | 661.50 |

Mean RR (ms) | 894.10 | 137.03 | 884.50 | 186.25 (791.00;977.25) |

Mean HR (beats per minute) | 69.22 | 10.5 | 68.78 | 14.61 (61.83;76.44) |

RMSSD (ms) | 76.04 | 44.84 | 65.00 | 53.25 (43.40;96.65) |

pNN50 (%) | 39.08 | 22.84 | 38.30 | 39.85 (18.40;59.25) |

LF (ms^{2}) |
1148.19 | 1291.74 | 999.00 | 1336 (582.50;1918.50) |

HF (ms^{2}) |
2760.15 | 3465.31 | 1721.00 | 2627 (582.50;1918.50 |

LF/HF | 0.81 | 0.58 | 0.67 | 0.74 (0.35;1.09) |

SD1(ms) | 54.06 | 31.83 | 46.10 | 37.80 (30.95;68.75) |

SD2(ms) | 90.48 | 37.17 | 83.10 | 51.60 (64.00;115.60) |

RMSSD/RR | 0.08 | 0.04 | 0.07 | 0.05 (0.05;0.10) |

LF/RR^{2} |
0.0017 | 0.0013 | 0.0012 | 0.001 (0.001;0.002) |

HF/RR^{2} |
0.0031 | 0.0034 | 0.0020 | 0.002 (0.001;0.003) |

SD1/RR | 0.060 | 0.030 | 0.050 | 0.04 (0.03;0.07) |

SD2/RR | 0.10 | 0.040 | 0.090 | 0.05 (0.07;0.12) |

The HR, RR intervals, and HRV indicators (normalized and non-normalized) were compared to the VO_{2}max of participants results, to establish any significant relationships (Table

_{2} max and HRV indicators

_{2}max |
_{2}max |
||
---|---|---|---|

Mean RR | 0.41(< 0.01)^{*} |
NA | NA |

Mean HR | −0.41 < 0.01^{*} |
NA | NA |

RMSSD | 0.09 (0.29) | RMSSD/RR | 0.02 (0.80) |

pNN50 | 0.19 (0.03)^{*} |
NA | NA |

LF | 0.10 (0.21) | LF/RR^{2} |
0.02 (0.95) |

HF | 0.001 (0.99) | HF/RR^{2} |
−0.03 (0.72) |

LF/HF | 0.02 (0.82) | NA | NA |

SD1 | 0.09 (0.29) | SD1/RR | 0.02 (0.80) |

SD2 | 0.14 (0.10) | SD2/RR | 0.02 (0.79) |

The results shown in Table _{2}max. When indicators were HR normalized by dividing by RR or RR^{2}, no significant correlations were found. All significant correlation coefficients were between 0.19 and 0.41 indicating very low to moderate associations.

Simple linear regression analyses (Table _{2}max. Stepwise multiple regression (Table _{2}max 20.1%). However, HR entered the model first and by itself explains 17% of the variation, while the inclusion of HF added only an additional 3.1% to the coefficient of determination. Stepwise multiple regression also showed that, when including HR and all normalized HRV indicators (RMSSD/RR, LF/RR^{2}, HF/RR^{2}, LF/HF, SD1/RR, SD2/RR), the statistical model consisting of a linear combination of HR and HF/RR^{2}, will be the best predictor of variation in VO_{2}max, that is 19.0%. HR on its own explained 16.5% and a combination of HR and HF/RR^{2} thus, explained only a further 2.5%.

_{2}max included |
^{2} |
||
---|---|---|---|

Simple Regression | HR | HR | 0.170 |

Simple Regression | RR | RR | 0.161 |

Simple Regression | pNN50 | pNN50 | 0.035 |

Multiple Regression | Nine indicators HR, RR, RMSSD, pNN50, LF, HF, LF/HF, SD1, SD2 | HR and HF | 0.201 |

Multiple Regression | HR and six normalized HRV indicators: HR/RR, RMSSD/RR, LF/RR^{2}, HF/RR^{2}, LF/HF, SD1/RR, SD2/RR |
HR and HF/RR^{2} |
0.190 |

The primary hypothesis was that straightforward measurements such as HR and RR intervals show stronger correlations with VO_{2}max than the HRV indicators, and is on their own stronger predictors of VO_{2}max than the more complex indicators of HRV. Study results partially confirmed the first hypotheses. Moderately strong correlations were found between HR and RR intervals vs. VO_{2}max. The only other correlation found (pNN50) was of very low strength (_{2}max, when entering the following HRV indicators: HR, RR, RMSSD, pNN50, LF, HF, LF/HF, SD1, SD2. With regards to the second hypothesis multiple ^{2} explained only 2.5% more of the variation in VO_{2}max, than HR alone (16.5%), when including all normalized indicators.

Recent publications by Sacha et al. (^{2}.

Correlations found between the HR, RR intervals, pNN50, and VO_{2}max were significant (_{2}max (_{2}max (_{2}max, and also higher vagal cardiac control (Nagai et al., _{2}max was only indirectly determined and not directly with gas analyses.

The fact that only one HRV indicator (pNN50) correlated significantly with VO_{2}max, may indicate that correlations between HRV indicators and VO_{2}max exist mainly due to the relationship between HR (independent of variability) and VO_{2}max.

Regression analysis confirmed that, when including HR, RR, RMSSD, pNN50, LF, HF, LF/HF, SD1, and SD2, the most important predictor of cardio vascular fitness and exercise ability as represented by VO_{2}max, is HR. Results also showed when testing the normalized indicators, HR explained the largest percentage of the changes in VO_{2}max. Thus, HR on its own is the most important predictor of changes in an important cardiac health metric such as VO_{2}max. Thus, the more complex concept of HRV quantification unlock only a small percentage of additional information with regards to autonomic function, compared to information obtained from only HR and RR interval measurements. These results may indicate that during investigation of exercise ability _{2}max) phenomena, quantification of HRV may not add significant more value. This is a novel idea which should be investigated further in clinical populations.

Limitations of this study include the fact that VO_{2}max was not directly determined with gas analyses, but only indirectly determined. Another drawback is that the tachogram used to determine HRV, was sampled only in a resting, supine position. Previous studies indicated that HRV measured during a stressor, such as standing upright, show more significant correlations with cardiopulmonary fitness indicators than supine HRV measurements (Grant et al.,

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.