Skip to main content
NCBI home page
European Journal of Sport Science logoLink to European Journal of Sport Science
. 2024 Sep 19;24(11):1539–1551. doi: 10.1002/ejsc.12175

Correlation properties of heart rate variability for exercise prescription during prolonged running at constant speeds: A randomized cross‐over trial

Thomas Gronwald 1,2,✉, Leonie Horn 3, Marcelle Schaffarczyk 1, Olaf Hoos 3
PMCID: PMC11534628  PMID: 39300759

Abstract

The study explores the validity of the nonlinear index alpha 1 of detrended fluctuation analysis (DFAa1) of heart rate (HR) variability for exercise prescription in prolonged constant load running bouts of different intensities. 21 trained endurance athletes (9 w and 12 m) performed a ramp test for ventilatory threshold (vVT1 and vVT2) and DFAa1‐based (vDFAa1‐1 at 0.75 and vDFAa1‐2 at 0.5) running speed detection as well as two 20‐min running bouts at vDFAa1‐1 and vDFAa1‐2 (20‐vDFAa1‐1 and 20‐vDFAa1‐2), in which HR, oxygen consumption (VO2), respiratory frequency (RF), DFAa1, and blood lactate concentration [La‐] were assessed. 20‐vDFAa1‐2 could not be finished by all participants (finisher group (FG), n = 15 versus exhaustion group (EG), n = 6). Despite similar mean external loads of vDFAa1‐1 (10.6 ± 1.9 km/h) and vDFAa1‐2 (13.1 ± 2.4 km/h) for all participants compared to vVT1 (10.8 ± 1.7 km/h) and vVT2 (13.2 ± 1.9 km/h), considerable differences were present for 20‐vDFAa1‐2 in EG (15.2 ± 2.4 km/h). 20‐vDFAa1‐1 and 20‐DFAa1‐2 yielded significant differences in FG for HR (76.2 ± 5.7 vs. 86.4 ± 5.9 %HRPEAK), VO2 (62.1 ± 5.0 vs. 77.5 ± 8.6 %VO2PEAK), RF (40.6 ± 11.3 vs. 46.1 ± 9.8 bpm), DFA‐a1 (0.86 ± 0.23 vs. 0.60 ± 0.15), and [La‐] (1.41 ± 0.45 vs. 3.34 ± 2.24 mmol/L). Regarding alterations during 20‐vDFAa1‐1, all parameters showed small changes for all participants, while during 20‐vDFAa1‐2 RF and DFAa1 showed substantial alterations in FG (RF: 15.6% and DFAa1: −12.8%) and more pronounced in EG (RF: 20.1% and DFAa1: −35.9%). DFAa1‐based exercise prescription from incremental testing could be useful for most participants in prolonged running bouts, at least in the moderate to heavy intensity domain. In addition, an individually different increased risk of overloading may occur in the heavy to severe exercise domains and should be further elucidated in the light of durability and decoupling assessment.

Keywords: decoupling, DFAa1, endurance sports, HRV, intensity distribution

Highlights

  • DFAa1‐based exercise prescription from incremental testing shows potential for prolonged constant load exercise, at least in the moderate to heavy intensity domains for most trained runners.

  • Caution is advised for the heavy to severe exercise domains as individual overload may occur.

  • The relationship of DFAa1 and vDFAa1 seems to be highly individual as well as perspectives for the durability and decoupling assessment and should be further elucidated during longer exercise bouts.

1. INTRODUCTION

Analyses of the nonlinear characteristics of heart rate (HR) variability (HRV) show that the short‐term scaling exponent alpha1 of detrended fluctuation analysis (DFAa1) may be a sensitive marker for assessing global organismic demands during acute endurance exercise (Gronwald, Törpel, et al., 2020; Gronwald & Hoos, 2020). DFAa1 was shown to provide a wide dynamic range encompassing the moderate, heavy, and severe exercise intensity domains (3‐zone‐model) during exercise compared to linear metrics of HRV (Rogers & Gronwald, 2022).

In general, the exponent quantifies the fractal scale and correlation properties of HR time series in cardiac beat‐to‐beat intervals and represents a rather qualitative marker of the autonomic nervous system (ANS) regulation, which means that under resting conditions, DFAa1 values around 1.0 mirror the homeodynamic behavior of control systems to dynamically self‐organize in between order (persistence) and disorder (de Godoy, 2016; Goldberger et al., 2002; Kauffman, 1995). During exercise, DFAa1 shows strongly correlated patterns (values well above 1.0) at low‐intensity exercise in the moderate domain, transitions to fractal patterns (value at around and below 1.0) at moderate to heavy exercise intensities, and drops to uncorrelated and anticorrelated patterns at the highest intensities with values around and below 0.5, which indicates a loss of fractal dynamics and a change toward random and/or anticorrelated behavior (Hautala et al., 2003). Easily accessible HRV data acquisition with chest belt sensors allows for laboratory and in‐field use and also opens up opportunities to provide real‐time feedback on exercise intensity (Gronwald, Berk, et al., 2021; Rogers & Gronwald, 2022).

Given these properties and based on signal‐theory background, applying this metric may be used as a biomarker for exercise intensity domain delineation. It has been shown that discrete numerical values of this biomarker may demarcate the transition from moderate to heavy intensity exercise around the aerobic threshold (DFAa1 of 0.75) and from heavy to severe intensities around the anaerobic threshold (DFAa1 of 0.5), respectively, corresponding to traditional threshold markers based on different organismic subsystem measures, such as blood lactate concentration [La‐] or gas exchange data, taking into account the potential limitations and deviations on an individual level (Mateo‐March et al., 2023; Rogers, Giles, Draper, Hoos, et al., 2021; Rogers, Giles, Draper, Mourot, et al., 2021; Schaffarczyk et al., 2023; van Hooren, Mennen, et al., 2023). Further, DFAa1 has been shown to be useful as a marker of acute fatigue in terms of a systemic perturbation (Rogers, Mourot, et al., 2021; Schaffarczyk et al., 2022; van Hooren, Bongers, et al., 2023; van Hooren, Mennen, et al., 2023) or as an internal load measure of fatigue resistance in studies with prolonged exercise (Gronwald et al., 2018, 2019; Gronwald, Rogers, et al., 2021). Therefore, expanding these findings to future approaches of real‐time monitoring of prolonged exercise seems to be promising, as the DFAa1 marker might bear the potential to mirror decoupling mechanisms as alterations of external‐to‐internal load relationships or “durability” aspects of endurance performance that were recently described as “the time of onset and magnitude of deterioration in physiological‐profiling characteristics over time during prolonged exercise” (Maunder et al., 2021; Smyth et al., 2022). Jones (2023) recently introduced this construct in terms of physiological resilience constituting an independent, fourth dimension of endurance exercise performance in addition to the classical ones of maximal oxygen uptake (VO2MAX), economy or efficiency, and fractional utilization of VO2MAX (e.g., Joyner, 1991; Joyner & Coyle, 2008). In that regard, recent studies show prior moderate‐intensity exercise could decrease power output at threshold estimates between moderate to heavy exercise intensities and time trial performance (Gallo et al., 2024; Hamilton et al., 2024; Stevenson et al., 2024). This alteration of internal‐to‐external load relationship could be also shown from an internal load perspective with fixed external load in the moderate intensity domain according to HR metrics and DFAa1 (Rogers, Mourot, et al., 2021; van Hooren, Bongers, et al., 2023; van Hooren, Mennen, et al., 2023). Here, it should be still differentiated between pre and post comparisons (e.g., Hamilton et al., 2024; Stevenson et al., 2024) and alterations during continuous prolonged exercise bouts (e.g., Gronwald et al., 2018; Gronwald, Rogers, et al., 2021; Maunder et al., 2021; Smyth et al., 2022).

However, validation data of DFAa1 during prescribed prolonged exercise bouts are still scarce and the true significance for exercise prescription remains to be fully elucidated. In addition, exercise prescription based on a percentage of maximum HR, oxygen consumption, or various approaches of fixed [La‐] and individual gas exchange utilization most often lead to an inaccurate and inconsistent representation of the relationship between external and internal load during prolonged exercise with large interindividual variability of the internal load situation (Brownstein et al., 2022; Fleckenstein et al., 2023; Jamnick et al., 2020; Mann et al., 2013). Therefore, the aim of the present pilot study was to evaluate the construct validity of DFAa1 to prescribe and monitor exercise intensity during continuous running bouts at the transition of moderate to heavy and heavy to severe exercise domains.

2. METHODS AND MATERIALS

2.1. Participants

21 trained (McKay et al., 2022) endurance athletes (9 w and 12 m; age: 25.9 ± 3.6 years, height: 178.4 ± 9.9 cm, body weight: 70.8 ± 8.7 kg, body fat: 12.2 ± 4.4%, peak heart rate – HRPEAK: 198.4 ± 7.9 bpm, and peak oxygen consumption – VO2PEAK: 59.0 ± 8.3 mL/kg/min) voluntarily participated in this study. All subjects were informed about risks and benefits of the procedures and signed an informed consent form. The ethics committee of the MSH Medical School Hamburg (reference no.: MSH‐2022/172) approved all tests performed in the study. The study was also carried out in accordance with the principles set forth in the most recent revision of the Declaration of Helsinki.

2.2. Study design

All participants visited the laboratory on three separate days. On the first day, they were informed about risks and benefits of the study and were accustomed to treadmill, face mask, and blood lactate concentration measurement procedures. Body fat percentage (BF%) was measured using bio‐impedance analysis (InnerScanV). Afterward, participants completed an incremental running test on a treadmill (Woodway DESMO Pro XL, Woodway Europe, Weil am Rhein) with an increment of 1 km/h per minute starting at 7 km/h for women and 8 km/h for men until volitional exhaustion. Prior to the test, athletes completed a 10‐min warm‐up on the treadmill at the initial test speed. The running speeds (v) at the first and second ventilatory threshold (vVT1 and vVT2) and at DFAa1 of 0.75 and 0.5 (vDFAa1‐1 and vDFAa1‐2) were determined during the running test. Afterwards, two 20‐min continuous running bouts at vDFAa1‐1 and vDFAa1‐2 (20‐vDFAa1‐1, 20‐vDFAa1‐2) were conducted in randomized and counterbalanced order within one week. Prior to 20‐vDFAa1‐1 and 20‐vDFAa1‐2, all participants warmed‐up on the treadmill at a speed corresponding to 80% of their vDFAa1‐1 speed for 10 min (see Figure 1). All tests were conducted at similar times of the day with at least a 72 h' time‐lag to the next or previous test or intense exercise session. The participants were instructed to obtain enough sleep the night before testing (at least 7 h), to refrain from alcohol intake for 24 h, and to have a last light meal 2–3 h before the investigation.

FIGURE 1.

FIGURE 1

Open in a new tab

Schematic view of heart rate (HR, grey) and DFAa1 (blue) of one 20‐min continuous running bout at vDFAa1‐1 (here: 11.7 km/h); warm‐up at 80% of vDFA1‐1 (here: 9.4 km/h); and passive recovery data included (Kubios HRV Scientific, version 4.1). Data of HR, HR variability (HRV), oxygen consumption (VO2), and respiratory frequency (RF) were continuously recorded, blood lactate concentration [La‐] before (Pre) and at the end (End) of the running bouts.

2.3. Data measurement

On all three laboratory days, HR and HRV were continuously measured during all test conditions using Polar H10 (transmitter sensor as chest belt recording) and V800 (receiver as wrist‐worn watch) HR devices (Polar Electro Oy, Kempele, Finland; Schaffarczyk et al., 2022). [La‐] was collected from the ear lobe before and at the end of the continuous running bouts using an enzymatic‐amperometric method (Biosen C‐line, EKF‐Diagnostics, Eppendorf, Germany). Further, oxygen consumption (VO2 in ml/kg/min) and respiratory frequency (RF in breaths per minute, bpm) were continuously measured during all three exercise sessions using a portable breath‐by‐breath metabolic cart (Metamax 3B, Cortex Biophysik GmbH, Leipzig, Germany; van Hooren et al., 2024).

2.4. HRV analysis and threshold determination

To analyze RR‐intervals (in ms) and HR (in beats per minute, bpm), data were exported from the Polar Flow web service (Polar Electro Oy, Kempele, Finland) and processed in Kubios HRV Scientific (version 4.1, Biosignal Analysis and Medical Imaging Group, Department of Physics, University of Kuopio, Kuopio, Finland). Preprocessing settings were set to the default values, including the RR detrending method, which was kept at “smoothness priors” (Lambda = 500). The RR‐interval series were then corrected using the Kubios HRV “automatic correction” method. To calculate DFAa1, the root mean square (RMS) fluctuations of the integrated and detrended RR‐intervals were analyzed in observation windows of different sizes and then further processed as the slope between the RMS‐fluctuation data in relation to the different window sizes on a log‐log scale (Peng et al., 1995). Window size was set to 4 ≤ n ≤ 16 beats in the software preferences. Data were also scanned visually for artefacts (e.g., spikes in RR‐interval series which were not corrected by the artefact correction algorithm), marked as “noise” and removed manually by an expert with experience in HRV data analysis. Datasets with >5% artefacts were excluded from HRV analysis. The time varying DFAa1 kinetic was then calculated over a 120 s window width with grid intervals of 5 s for the threshold estimation during incremental testing and 10 s for the alignment with ventilatory data during prolonged exercise. HRV thresholds were determined based on fixed values of DFAa1. Linear regression was performed on the subset of data consisting of the rapid, near straight‐line drop from DFAa1 values close to 1.0 to approximately 0.5 or below if the values continued in a non‐deviating fashion. The running speeds where DFAa1 reached either 0.75 (vDFAa1‐1) and 0.5 (vDFAa1‐2) were calculated based on the regression equation from that linear section (Rogers, Giles, Draper, Hoos, et al., 2021; Rogers, Giles, Draper, Mourot, et al., 2021) or based on established multiphasic dose‐response modeling (Di Veroli et al., 2015), if its goodness‐of‐fit exceeded the one of standard linear regression.

2.5. Ventilatory threshold determination

To determine VT1, the V‐slope method was used (Beaver et al., 1986). In case of inconclusive VCO2 and VO2 trajectories, additionally an increase in end tidal O2 pressure versus time and a rise in the O2 equivalent without a simultaneous increase in CO2 equivalent was used as criteria for VT1. VT2 was defined as the point of a disproportional rise in minute volume (VE) versus VCO2. Additional criteria were a decrease in end expiratory CO2 pressure versus time and an increase in the CO2 equivalent versus time (Meyer et al., 2005). VT1 and VT2 were both determined independently by two researchers. In cases where both researchers did not agree, the mean value was used for VT1 and VT2 estimation.

2.6. Efficiency factor

For the analysis of internal‐to‐external load relationship and a possible decoupling mechanism, an efficiency factor (EF) was defined. This internal‐to‐external workload ratio was calculated for the start and end of the continuous running bouts using the ratio of internal load indicators (HR, VO2, RF, and DFAa1) and running pace (in km/h). For those participants who exhausted before the end of 20‐vDFAa1‐2, minutes 4/5 versus the last 2 minutes of exercise were used. The difference of the EF from the start and end was calculated and divided by the EF from the start multiplied by 100 to get a percentage of alteration (%). Thus, a value of 10% indicates that internal‐to‐external ratio was 10% greater at the end segment compared to that observed in the start segment (Maunder et al., 2021; Smyth et al., 2022).

3. STATISTICAL ANALYSES

Statistics were conducted using the SPSS 27.0 (IBM), and Microsoft Excel (Microsoft Corp, Redmond, USA). Prior to all tests, normality of distribution was tested using the Shapiro–Wilk testing. To analyze the effects of the exercise bouts on dependent variables, a two‐way analysis of variance with repeated measures (ANOVA; intensity x time) was applied and both the main effects and the interaction (intensity × time) were reported. In addition, post‐hoc testing and comparison of different approaches of external load assessment at estimated exercise intensity thresholds as well as mean differences between 20‐vDFAa1‐1 and 20‐vDFAa1‐2 were conducted via paired t‐tests. Additionally, partial η2 was used to denote main effects and Hedge's g for the effect size estimation of t‐test results (difference between mean values divided by corrected standard deviation; Hedges, 1981), with no effect (d < 0.2), small effect size (d < 0.5), moderate effect size (d ≥ 0.5), and large effect size (d > 0.8) (Cohen, 1988). In addition to mean values for the complete continuous running bouts, the values at start and end at minutes 4/5 versus 19/20 were used for comparisons (see Figure 1). Statistical tests were deemed to be significant at p ≤ 0.05. All results are reported as means ± standard deviation (SD).

4. RESULTS

The analyzed data exhibited a normal distribution. External loads of vDFAa1‐1 (10.6 ± 1.9 km/h) and vDFAa1‐2 (13.1 ± 2.4 km/h) for all participants were comparable to vVT1 (10.8 ± 1.7 km/h, p = 0.418, and g = 0.13) and vVT2 (13.2 ± 1.9 km/h, p = 0.661, and g = 0.07), but considerable differences between both methods were present for the comparison of vVT2 versus vDFAa1‐2 in EG (see also descriptive analysis and differentiation for FG and EG in Table 1). 20‐vDFAa1‐1 was successfully performed by all participants. In this regard, 20‐vDFAa1‐2 was completed by only 15 participants (finisher group, FG), while six participants had to stop ahead of time at 11:47 ± 03:13 min: s due to exhaustion (exhaustion group, EG), as indicated by [La‐] mean values of 9.98 ± 2.41 mmol/L (see Table 1). Due to artefact rates >5% in RR‐interval raw data, two participants had to be excluded from the analysis of 20‐vDFAa1‐2 (for DFA a1) in the FG.

TABLE 1.

Comparison of peak running speeds (vPEAK) and total duration from the incremental test, vDFAa1‐1, vDFAa1‐2, vVT1, vVT2, and blood lactate concentration [La‐] before (Pre) and at the end (End) of both running bouts for all participants, the finisher group and the exhaustion group.

All [n = 21] FG [n = 15] EG [n = 6]
vPEAK [km/h] 17.5 ± 1.7 17.5 ± 1.7 17.6 ± 1.8
Test duration [min] 10.9 ± 1.4 10.8 ± 1.4 11.3 ± 1.4
vDFAa1‐1 [km/h] (%vPEAK) 10.6 ± 1.9 (60.2 ± 8.0) 10.0 ± 1.5 (57.4 ± 5.9) 12.0 ± 2.5 (67.4 ± 8.4)
vVT1 [km/h] (%vPEAK) 10.8 ± 1.7 (61.7 ± 6.8) 10.6 ± 1.8 (60.4 ± 7.3) 11.5 ± 1.4 (65.0 ± 4.3)
vDFAa1‐2 [km/h] (%vPEAK) 13.1 ± 2.4 (74.5 ± 10.0) 12.3 ± 1.9 (69.9 ± 6.8) 15.2 ± 2.4 (85.8 ± 7.6)
vVT2 [km/h] (%vPEAK) 13.2 ± 1.9 (75.5 ± 6.3) 13.1 ± 2.0 (74.8 ± 7.0) 13.6 ± 1.5 (77.2 ± 3.8)
[La‐] Pre 20‐vDFAa1‐1 [mmol/L] 0.95 ± 0.29 0.92 ± 0.27 1.01 ± 0.37
[La‐] End 20‐vDFAa1‐1 [mmol/L] 1.56 ± 0.59 1.41 ± 0.45 1.94 ± 0.78
[La‐] Pre 20‐vDFAa1‐2 [mmol/L] 1.08 ± 0.22 1.09 ± 0.24 1.07 ± 0.15
[La‐] End 20‐vDFAa1‐2 [mmol/L] 5.24 ± 3.79 3.34 ± 2.24 9.98 ± 2.41
Open in a new tab

Note: FG: group of 15 participants who finished both prolonged exercise tests; EG: group of 6 participants who did not finish 20‐vDFAa1‐2.

ANOVA results indicate a significant main effect for intensity in comparison of 20‐vDFAa1‐1 and 20‐DFAa1‐2 for all analyzed parameters in FG. In addition, a significant main effect of time could be determined in HR, VO2, and RF but not for DFAa1. An effect of interaction (intensity × time) could only be found for HR (intensity: F = 43.3, p < 0.001, η 2 = 0.756; time: F = 139.9, p < 0.001, η 2 = 0.909; and interaction: F = 18.2, p = 0.001, η 2 = 0.565) and VO2 (intensity: F = 34.6, p < 0.001, η 2 = 0.712; time: F = 25.7, p < 0.001, η 2 = 0.647; and interaction: F = 15.3, p = 0.002, η 2 = 0.522), but not for RF (intensity: F = 21.9, p < 0.001, η 2 = 0.610; time: F = 17.5, p = 0.001, η 2 = 0.556; and interaction: F = 4.3, p = 0.056, η 2 = 0.236), and DFAa1 (intensity: F = 12.3, p = 0.004, η 2 = 0.505; time: F = 1.1, p = 0.320, η 2 = 0.082; and interaction: F = 0.4, p = 0.526, η 2 = 0.034), respectively.

20‐vDFAa1‐1 and 20‐DFAa1‐2 yielded substantial mean differences in FG for HR (150.4 ± 10.6 vs. 170.3 ± 9.8 bpm, p < 0.001, g = 1.90; 76.2 ± 5.7 vs. 86.4 ± 5.9 %HRPEAK, p < 0.001, g = 1.71), VO2 (36.7 ± 5.2 vs. 46.0 ± 8.7 mL/kg/min, p < 0.001, g = 1.26; 62.1 ± 5.0 vs. 77.5 ± 8.6 %VO2PEAK, p < 0.001, g = 2.13), RF (40.6 ± 11.3 vs. 46.1 ± 9.8 bpm, p < 0.001, g = 0.50), DFA‐a1 (0.86 ± 0.23 vs. 0.60 ± 0.15, p = 0.004, g = −1.30), and [La‐] at the end of the running bouts (1.41 ± 0.45 vs. 3.34 ± 2.24 mmol/L; p < 0.001, g = 1.16, see Figures 2, 3, 4, 5, and descriptive analysis and differentiation data for EG in Tables 1 and 2).

FIGURE 2.

FIGURE 2

Open in a new tab

Comparison of heart rate (%HRPEAK, mean, and SD) kinetics during 20‐vDFAa1‐1 and 20‐vDFAa1‐2 of the finisher group.

FIGURE 3.

FIGURE 3

Open in a new tab

Comparison of oxygen consumption (%VO2PEAK, mean, and SD) kinetics during 20‐vDFAa1‐1 and 20‐vDFAa1‐2 of the finisher group.

FIGURE 4.

FIGURE 4

Open in a new tab

Comparison of respiratory frequency (RF, mean, and SD) kinetics during 20‐vDFAa1‐1 and 20‐vDFAa1‐2 of the finisher group.

FIGURE 5.

FIGURE 5

Open in a new tab

Comparison of DFAa1 (mean and SD) kinetics during 20‐vDFAa1‐1 and 20‐vDFAa1‐2 of the finisher group.

TABLE 2.

Summary of all physiological mean data (± standard deviation) from 20‐vDFAa1‐1 and 20‐vDFAa1‐2 for the finisher group and the exhaustion group.

FG [n = 15, n = 13 for DFAa1 in 20‐vDFAa1‐2] EG [n = 6]
20‐vDFAa1‐1 20‐vDFAa1‐2 Statistics 20‐vDFAa1‐1 20‐vDFAa1‐2
HR [bpm] (%HRPEAK) 150.4 ± 10.6 (76.2 ± 5.7) 170.3 ± 9.8 (86.4 ± 5.9) p < 0.001, g = 1.90 165.1 ± 10.5 (82.3 ± 5.3) 189.3 ± 5.5 (94.3 ± 2.1)
VO2 [ml/kg/min] (%VO2PEAK) 36.7 ± 5.2 (62.1 ± 5.0) 46.0 ± 8.7 (77.5 ± 8.6) p < 0.001, g = 1.26 41.5 ± 7.2 (71.7 ± 10.7) 53.9 ± 7.3 (92.9 ± 2.2)
RF [bpm] 40.6 ± 11.3 46.1 ± 9.8 p < 0.001, g = 0.50 40.6 ± 6.8 51.1 ± 3.4
DFAa1 0.86 ± 0.23 0.60 ± 0.15 p = 0.004, g = −1.30 0.76 ± 0.27 0.48 ± 0.17
Open in a new tab

Note: T‐test statistics and effect sizes are stated for the comparison of 20‐vDFAa1‐1 versus vDFAa1‐2 in FG. FG: group of 15 participants who finished both prolonged exercise tests; EG: group of 6 participants who did not finish 20‐vDFAa1‐2.

In comparison of the start and end of 20‐vDFAa1‐1 in FG HR and RF increased moderately from 148.4 ± 11.0 to 153.9 ± 11.0 bpm (p < 0.001 and g = 0.48) and 38.7 ± 10.7 to 41.3 ± 11.8 bpm (p = 0.023 and g = 0.22), while VO2 and DFAa1 remained rather stable with values of 36.6 ± 5.0 versus 37.1 ± 5.6 mL/kg/min (p = 0.108 and g = 0.09) and 0.86 ± 0.28 versus 0.84 ± 0.19 (p = 0.666 and g = −0.08), respectively (see Table 3). Regarding the alteration of the calculated EF, all parameters showed small changes for FG (HR: 3.7%, VO2: 1.3%, RF: 6.0%, and DFAa1: −2.5%). In addition, EG also showed small alterations of EF <10% for 20‐vDFAa1‐1 (HF: 4.8%, VO2: 1.4%, RF: 7.8%, and DFAa1: −8.3%).

TABLE 3.

Summary of all physiological mean data (± standard deviation) from the start and end of 20‐vDFAa1‐1 and 20‐vDFAa1‐2 for the finisher group and the exhaustion group.

20‐vDFAa1‐1 20‐vDFAa1‐2
Start End Statistics Start End Statistics
FG [n = 15, n = 13 for DFAa1 in 20‐vDFAa1‐2]
HR [bpm] (%HRPEAK) 148.4 ± 11.0 (75.2 ± 6.0) 153.9 ± 11.0 (78.0 ± 5.8) p < 0.001, g = 0.48 166.6 ± 9.9 (84.5 ± 6.1) 175.7 ± 10.5 (89.1 ± 6.2) p < 0.001, g = 0.87
VO2 [ml/kg/min] (%VO2PEAK) 36.6 ± 5.0 (61.9 ± 5.1) 37.1 ± 5.6 (62.7 ± 5.7) p = 0.108, g = 0.09 44.2 ± 7.7 (74.5 ± 7.3) 46.8 ± 9.0 (78.9 ± 9.2) p < 0.001, g = 0.30
RF [bpm] 38.7 ± 10.7 41.3 ± 11.8 p = 0.023, g = 0.22 42.0 ± 11.2 48.4 ± 9.5 p = 0.002, g = 0.59
DFAa1 0.86 ± 0.28 0.84 ± 0.19 p = 0.666, g = −0.08 0.65 ± 0.21 0.57 ± 0.17 p = 0.262, g = −0.41
EG [n = 6]
HR [bpm] (%HRPEAK) 161.8 ± 10.1 (80.7 ± 4.7) 169.7 ± 11.2 (84.6 ± 5.9) ‐ 188.5 ± 4.7 (94.0 ± 2.7) 194.3 ± 5.9 (96.8 ± 2.5) ‐
VO2 [ml/kg/min] (%VO2PEAK) 41.2 ± 6.4 (71.1 ± 8.4) 41.8 ± 7.4 (72.3 ± 12.3) ‐ 51.9 ± 6.8 (89.4 ± 2.4) 55.5 ± 8.3 (95.4 ± 2.0) ‐
RF [bpm] 38.8 ± 7.6 42.0 ± 7.3 ‐ 46.3 ± 5.4 55.7 ± 3.2 ‐
DFAa1 0.81 ± 0.36 0.62 ± 0.26 ‐ 0.55 ± 0.25 0.34 ± 0.13 ‐
Open in a new tab

Note: T‐test statistics and effect sizes are stated for the comparison of “Start” versus “End” in 20‐vDFAa1‐1 and vDFAa1‐2 in FG. FG: group of 15 participants who finished both prolonged exercise tests; EG: group of 6 participants who did not finish 20‐vDFAa1‐2.

During the 20‐vDFAa1‐2 running bout of FG HR und RF rose substantially from 166.6 ± 9.9 to 175.7 ± 10.5 bpm (p < 0.001 and g = 0.87), and 42.0 ± 11.2 versus 48.4 ± 9.5 bpm (p = 0.002 and g = 0.59); while, VO2 increased moderately with 44.2 ± 7.7 versus 46.8 ± 9.0 mL/kg/min (p < 0.001 and g = 0.30) and DFA‐a1 remained rather stable with 0.65 ± 0.21 versus 0.57 ± 0.17 (p = 0.262 and g = −0.41) (see Table 3). The calculated EF showed small changes for HR and VO2 (HR: 5.4% and VO2: 5.6%), while RF and DFAa1 showed more substantial alterations above 10% (RF: 15.6% and DFAa1: −12.8%). In addition, in EG, small to moderate alterations of EF indices during 20‐vDFAa1‐2 were present for HR (2.9%) and VO2 (6.7%), while changes in RF (20.1%) and DFAa1 (−35.9%) were substantially more pronounced.

5. DISCUSSION

The aim of this pilot study was to evaluate the ability of DFAa1 to prescribe and monitor exercise intensity during continuous running at the boundaries of moderate to heavy and heavy to severe exercise intensities, respectively. Even though it was not the primary goal to directly compare different approaches of exercise intensity domain demarcation (see e.g., Galán–Rioja et al., 2020; Kaufmann et al., 2023), our data indicate that external loads of vDFAa1‐1 and vDFAa1‐2 were comparable to vVT1 for all participants and to vVT2 for most of the participants (FG), respectively. In addition, the overall comparison of mean values from 20‐vDFAa1‐1 and 20‐vDFAa1‐2 yielded a clear distinction between the two exercise intensities in all physiological parameters and indicate a reasonable demarcation of a 3‐zone‐model of intensity distribution for moderate, heavy, and severe exercise domains (Jamnick et al., 2020; Haugen et al., 2022). Therefore, as already suggested based on signal‐theory and findings from prior studies using incremental exercise tests (Gronwald, Törpel, et al., 2020; Gronwald & Hoos, 2020; Rogers & Gronwald, 2022; Sempere‐Ruiz et al., 2024), exercise prescription using running speeds at DFAa1 values of 0.75 and 0.5 (vDFAa1‐1, vDFAa1‐2) separate reasonable intensity domains for prolonged constant exercise bouts for most participants. In our approach, these different domains do not directly rely on metabolic markers but rather are based on the complex changes in autonomic modulation due to parasympathetic withdrawal, sympathetic activation, altered nonneural factors, and the potential loss of interaction between the two branches of the ANS with increased exercise intensity (Persson, 1996; White & Raven, 2014). However, substantial interindividual fluctuations in internal load occur for both prolonged running bouts that are related to general problems of exercise prescription for prolonged exercise when intensity zone markers are derived from incremental exercise tests (Iannetta, de Almeida Azevedo, et al., 2019, 2020; Jamnick et al., 2020; Zuccarelli et al., 2018). Further, our data also support the notion that the magnitude and practical relevance of these interindividual differences depend on the addressed intensity domain, as substantial and practically relevant differences were mainly found in vDFAa1‐2, and these differences lead to premature exhaustion in the high intensity running bout for more than 25% of the participants (EG, n = 6).

5.1. Prolonged constant load exercise prescription from incremental testing

Prolonged constant load exercises derived from external load prescriptions of incremental tests bear the general problem that the physiological response may vary considerably between individuals both at the beginning and throughout the constant load exercise (e.g., Iannetta, de Almeida Azevedo, et al., 2019; Jamnick et al., 2020). This could be seen in all our metabolic and cardiorespiratory markers, including DFAa1, which was used for prescription of vDFAa1‐1 (DFAa1 = 0.75) and vDFAa1‐2 (DFAa1 = 0.5). In addition, recent findings even indicate an unclear assignment of an intensity domain during constant load exercise of prolonged duration when using the highly individual acute responses of %HRPEAK as a benchmark (Iannetta, de Almeida Azevedo, et al., 2019). In that regard, for prolonged exercise prescription, it must be considered that laboratory testing with incremental design (step and ramp) needs to account for specific response kinetics of the corresponding physiological markers, and the magnitude of the interindividual variability depends on the interaction of the used biomarker, its response kinetics, and the incremental exercise protocol (Zuccarelli et al., 2018; Iannetta, Murias, & Keir, 2019). External load prescription assumes that physiological responses are rather static (Jamnick et al., 2020; Maunder et al., 2021) and neglect the influence of internal and external factors leading to heterogeneity in exercise tolerance and physiological responses over time (e.g., personal or environmental factors, Gronwald, Törpel, et al., 2020; Meyler et al., 2023). The large variation of internal load responses in our data seem to be also present when prescriptions based on fixed submaximal threshold approaches (e.g., [La‐]) or individual submaximal threshold approaches are made (e.g., GET) (Fleckenstein et al., 2023; Brownstein et al., 2022). In addition, pre‐exhaustion due to incremental testing may also lead to an altered internal‐to‐external load relationship at an intensity domain transition compared to the beginning of prolonged exercise (with or without standardized warm‐up procedures) revealing the need to address exercise duration as an important independent prescription factor (Hofmann & Tschakert, 2017; Tschakert et al., 2022). This also applies to other external load indicators such as movement frequency (e.g., cadence in cycling exercise; Beneke & Leithäuser, 2017; Gronwald et al., 2018). In this regard, recent findings from DFA of HRV during running show that repeated incremental running tests shift the agreement between gas exchange thresholds and DFAa1‐derived boundaries for intensity demarcation, leading to the assumption of fatigue‐related inter‐ and intraindividual physiological perturbations depicted in DFAa1 kinetics (van Hooren, Bongers, et al., 2023; van Hooren, Mennen, et al., 2023). Therefore, it seems reasonable to address these issues using an internal‐load‐based approach of exercise prescription for prolonged constant load exercise to consider the individual and daily responses to a prescribed external load accounting for personal and environmental influences on the most prominent underlying trend in HR data, the cardiovascular drift (Ekelund, 1967). Recently, a HR‐based exercise intensity control approach using a HR clamp that kept HR constant over time by adjusting speed/power (Zuccarelli et al., 2018; Li et al., 2023) showed substantial differences for standard HRV metrics during exercise when compared to constant load exercise with corresponding HR drifts (Brockmann & Hunt, 2023; Hernando et al., 2018; Hunt & Saengsuwan, 2018). However, in this context, further questions need to be clarified about suitability of different subsystem parameters of internal load and an “optimal” and feasible real‐time monitoring approach for the control of exercise intensity (e.g., HR drift and the potential underestimation of rating of perceived exertion (RPE); Cartón–Llorente et al., 2022). Here, a dimensionless, global, and systemic internal load indicator, such as DFAa1 (in addition to RPE as an easily accessible subjective marker), could provide the potential for further investigation in prolonged exercise regimes (Gronwald, Berk, et al., 2021; Rogers & Gronwald, 2022). However, it seems mandatory to evaluate mid‐ and long‐term training outcomes when controlling exercise intensity based on internal load markers, as this may lead to significant and interindividual varying reduction in exercise and training stimuli (Zuccarelli et al., 2018).

5.2. Intensity domain dependency for prolonged exercise prescription

Besides the already discussed general problems of exercise prescription for prolonged constant load exercise from incremental testing, our data also points toward an intensity specific aspect of these issues that needs to be further addressed. On the one hand, our data show that the intensity prescription for 20‐vDFAa1‐1 elicits a [La‐] below 2 mmoL/L, and magnitudes of %HRPEAK and %VO2PEAK for FG and EG (Tables 1, 2, 3) that clearly mirror medium to upper levels of a moderate intensity domain in established 3‐zone‐models (Jamnick et al., 2020; Haugen et al., 2022). Besides the mean values the interindividual differences in %HRPEAK and %VO2PEAK are in line with the magnitudes shown in a recent study based on [La‐]‐derived LT prescription (Fleckenstein et al., 2023). In addition, when considering recently proposed ratios for the assessment of internal‐to‐external load relationships and decoupling mechanisms (Maunder et al., 2021; Smyth et al., 2022) in terms of EF of HR, VO2, RF, and DFAa1 in relation to running speed, a comparison of start and end of the running bouts lead to rather small alterations of EF <10% in all parameters in both FG and EG. These values of a small HR drift without a slow component in VO2 support the notion of a DFAa1‐derived separation of the moderate and heavy intensity domain.

On the other hand, for 20‐vDFAa1‐2 derived from DFAa1 values of 0.5 during incremental testing, [La‐]‐values of approx. 3.5 mmol/l as well as %HRPEAK and %VO2PEAK in FG (n = 15) can be matched within the lower to medium range of the heavy intensity domain, while the corresponding values for EG (n = 6) that could not sustain the 20‐min exercise duration clearly exceed the boundary toward the severe intensity domain (Jamnick et al., 2020; Haugen et al., 2022). Further, 20‐vDFAa1‐2 lead to substantial alterations of EF with a magnitude of >10% for RF and DFAa1 in FG and >20% for EG, which shows the potential of these two internal load parameters regarding further decoupling analysis of internal‐to‐external relationship. In this context, RF was recently mentioned as a promising internal load marker in exercise physiology and offers new possibilities for wearable analyses in research and practical settings (Nicolo et al., 2017; Nicolo & Sacchetti, 2023).

The probable overestimation of running velocity in the subsample of EG is further in line with previous studies that already discussed potential overestimations for a relevant number of participants using the present approach of nonlinear HRV analysis to delineate the heavy from the severe intensity domain (Rogers, Giles, Draper, Mourot, et al., 2021; Mateo‐March et al., 2023). One factor might be that (linear) HRV metrics are both intensity and time dependent and may reach close to minimum values with low signal‐to‐noise ratio rapidly (Brockmann & Hunt, 2023). However, to what extent this is true for nonlinear measures, such as DFAa1, needs to be further elucidated in, for example, HR clamp exercise. Further, as stated before (Gronwald & Hoos, 2020; Kaufmann et al., 2023; Rogers & Gronwald, 2022), despite the need for comparison with established intensity domain threshold concepts, it should be kept in mind that the present approach is based on the theoretical framework of a self‐organized dynamic regulation of the central autonomic network (CAN, Benarroch, 1993) that is reflected in the correlation properties of HR dynamics. Therefore, it is rather complementary to and does not necessarily coincide with classical metabolic threshold concepts based on metabolic and/or respiratory biomarkers. As mentioned above, the definition of intensity domain boundaries might involve different approaches based on performance indicators of external load (e.g., critical speed or power, CS, and CP), subsystem indicators of internal load such as [La‐] (e.g., lactate threshold and maximal lactate steady state) and/or gas exchange data (e.g., gas exchange threshold, GET), that interact with their corresponding different testing protocols. This may produce inconsistent results leading to substantially different delineations of the boundaries in a 3‐zone‐model of intensity zones and is reflected in the still ongoing debate about gold standard approaches to delineate moderate from heavy, and especially heavy from severe intensity domains (Chicharro et al., 1997; Galán‐Rioja et al., 2020; Hopker et al., 2011; Iannetta, de Almeida Azevedo, et al., 2019; Jamnick et al., 2018; Pallarés et al., 2016; Poole et al., 2021). Therefore, misleading comparisons between protocols as well as undesired training outcomes in athletes attempting to emulate a proposed method are also present when other approaches are used, and this seems to be most problematic for the boundary of heavy to severe exercise intensity (Galán‐Rioja et al., 2020; Iannetta, de Almeida Azevedo, et al., 2019; Jamnick et al., 2018; Poole et al., 2021).

Taken together based on the present results and the available data of previous studies using DFAa1 as an complementary exercise prescriptor (Mateo‐March et al., 2023; Rogers, Giles, Draper, Hoos, et al., 2021; Rogers, Giles, Draper, Mourot, et al., 2021; Schaffarczyk et al., 2023; van Hooren, Mennen, et al., 2023), it must be noted that for some individuals the present approach does not lead to an adequate specification of exercise intensity at the boundary of the heavy to severe exercise domain. Therefore, further investigations should be dedicated to the considerable differences of DFAa1‐derived threshold determination for 20‐vDFAa1‐2 in EG, leading to overestimation of running velocity in this subsample. Here, for example, the influence of individuality in time and intensity dependent changes in the scaling behavior of DFA (Molkkari et al., 2020; Kanniainen et al., 2023) as well as the model fitting and the model type (e.g., linear, polynomial, and sigmoidal) of DFAa1‐derived threshold determination could be subject to further investigations. In prior studies, we observed inappropriate suppression of correlation properties of HR time series in some individuals losing dynamic range of DFAa1 despite good ECG waveform and little artefact (van Hooren, Mennen, et al., 2023). In addition, future studies should enlighten the relevance of standardization of methodological aspects (e.g., quality of data acquisition, preprocessing, and artifact correction methods depending on the type/mode of exercise and/or laboratory vs. field conditions) on DFAa1‐derived exercise prescription and evaluate more thoroughly the significance of primary internal (e.g., breathing) and external influencing factors (e.g., environmental conditions).

6. LIMITATIONS

The prolonged exercise bouts were too short to provide evidence whether our approach of DFAa1‐derived exercise prescription may be useful for typical duration of running training (e.g., 30–60 min), as these longer durations may further complicate potential “decoupling mechanisms”. However, our data as well as findings from prior studies at least indicated that the magnitude of duration‐related influences and the potential for fatigue resistance assessment during prolonged exercise regimes could be further evaluated using a DFAa1 approach (Gronwald et al., 2018, 2019; Gronwald, Berk, et al., 2021). Longer exercise bouts are also needed to increase sensitivity for the evaluated EF. In addition, since external load was maintained constantly for the prolonged exercise bouts, the analysis of internal‐to‐external load relationship and decoupling mechanism would be more appropriate to use within the application in‐field conditions and/or self‐paced scenarios. Whether other theoretically appealing approaches using ratios of individually designed external‐to‐internal load markers, such as, maximal or submaximal external load markers like CP or CS in combination with DFAa1 and/or %HR recovery or %HRMAX might be helpful also remains an open question.

7. CONCLUSION

For most participants, DFAa1 shows great potential as a dimensionless and systemic index for internal load‐based exercise prescription with a clear demarcation perspective for a 3‐zone training intensity distribution model. However, for some individuals, the present approach does not lead to an adequate separation of exercise intensities, especially not for the heavy to severe exercise domains. Therefore, further investigations are recommended to account for interindividual differences and to better understand the relationship of DFAa1 and vDFAa1 and its relevance for the time evolution of fatigue during prolonged constant load exercise. In this regard, the potential for internal load‐based real‐time monitoring and intraindividual internal‐to‐external load analysis as a regular biological calibration procedure accounting for personal and environmental factors might be strengthened by further exercise specific methodological refinements of DFA. In addition, future studies should elucidate possible decoupling mechanisms of DFAa1 and other internal load measures in relation to external load (and other influencing factors, e.g., exercise mode, environmental conditions, and pre‐exhaustion) during even longer exercise bouts that correspond to typical exercise durations of real‐world running training (>30 min).

AUTHOR CONTRIBUTIONS

Olaf Hoos and Thomas Gronwald designed the research. Leonie Horn and Olaf Hoos conducted the experiments and data processing. Thomas Gronwald and Olaf Hoos conducted data analysis and interpretation. Thomas Gronwald drafted the raw manuscript. All authors provided critical comments on the manuscript, read, and approved the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflicts of interest.

ACKNOWLEDGMENTS

The authors would like to thank all participants for their time for taking part in this study. The authors would also like to thank Sebastian Kaufmann, Moritz Wagenhaeuser, and Lukas Berberich for their help in data acquisition. None of the authors received funding for this work from any organization other than salary support for the authors from their respective institutions.

Open Access funding enabled and organized by Projekt DEAL.

DATA AVAILABILITY STATEMENT

Data are available from the corresponding author on reasonable request.

REFERENCES

  1. Beaver, W. L. , Wasserman K., and Whipp B. J.. 1986. “A New Method for Detecting Anaerobic Threshold by Gas Exchange.” Journal of Applied Physiology 60(6): 2020–2027. 10.1152/jappl.1986.60.6.2020. [DOI] [PubMed] [Google Scholar]
  2. Benarroch, Eduardo E. 1993. “The Central Autonomic Network: Functional Organization, Dysfunction, and Perspecti‐Ve.” Mayo Clinic Proceedings 68(10): 988–1001. 10.1016/s0025-6196(12)62272-1. [DOI] [PubMed] [Google Scholar]
  3. Beneke, Ralph , and Leithäuser Renate M.. 2017. “Maximal Lactate Steady State's Dependence on Cycling Cadence.” International Journal of Sports Physiology and Performance 12(3): 304–309. 10.1123/ijspp.2015-0573. [DOI] [PubMed] [Google Scholar]
  4. Brockmann, Lars , and Hunt Kenneth J.. 2023. “Heart Rate Variability Changes with Respect to Time and Exercise Intensity During Heart‐Rate‐Controlled Steady‐State Treadmill Running.” Scientific Reports 13(1): 8515. 10.1038/s41598-023-35717-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brownstein, Callum G. , Pastor Frederic Sabater, Mira José, Murias Juan M., and Millet Guillaume Y.. 2022. “Power Output Manipulation from Below to Above the Gas Exchange Threshold Results in Exacerbated Performance Fatigability.” Medicine & Science in Sports & Exercise 54(11): 1947–1960. 10.1249/mss.0000000000002976. [DOI] [PubMed] [Google Scholar]
  6. Cartón‐Llorente, Antonio , Roche‐Seruendo Luis E., Mainer‐Pardos Elena, Nobari Hadi, Rubio‐Peirotén Alberto, Jaén‐Carrillo Diego, and García‐Pinillos Felipe. 2022. “Acute Effects of a 60‐min Time Trial on Power‐Related Parame‐Ters in Trained Endurance Runners.” BMC Sports Science, Medicine & Rehabilitation 14(1): 142. 10.1186/s13102-022-00538-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chicharro, J. L. , Pérez M., Vaquero A. F., Lucía A., and Legido J. C.. 1997. “Lactic Threshold vs Ventilatory Threshold During a Ramp Test on a Cycle Ergometer.” The Journal of Sports Medicine and Physical Fitness 37(2): 117–121. [PubMed] [Google Scholar]
  8. Cohen, J. 1988. Statistical Power Analysis for the Behavioural Sciences. Hillsdale, MI, USA: Erlbaum. [Google Scholar]
  9. Di Veroli, Giovanni Y. , Fornari Chiara, Goldlust Ian, Mills Graham, Koh Siang Boon, Bramhall Jo L., Richards Frances M., and Jodrell Duncan I.. 2015. “An Automated Fitting Procedure and Software for Dose‐Response Curves with Multiphasic Features.” Scientific Reports 5(1): 14701. 10.1038/srep14701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ekelund, L. G. 1967. “Circulatory and Respiratory Adaptation During Prolonged Exercise of Moderate Intensity in the Sitting Position.” Acta Physiologica Scandinavica 69(4): 327–340. 10.1111/j.1748-1716.1967.tb03529.x. [DOI] [PubMed] [Google Scholar]
  11. Fernandes de Godoy, Moacir . 2016. “Nonlinear Analysis of Heart Rate Variability: A Comprehensive Review.” Journal of Cardiology: Therapy 3: 528–533. 10.17554/j.issn.2309-6861.2016.03.101-4. [DOI] [Google Scholar]
  12. Fleckenstein, Daniel , Seelhöfer Jannik, Walter Nico, and Ueberschär Olaf. 2023. “From Incremental Test to Continuous Running at Fixed Lactate Thresholds: Individual Responses on %VO2max, %HRmax, Lactate Accumulation, and RPE.” Sports (Basel). 11(10): 198. 10.3390/sports11100198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Galán‐Rioja, Miguel Ángel , González‐Mohíno Fernando, Poole David C., and González‐Ravé José M. ª.. 2020. “Relative Proximity of Critical Power and Metabolic/Ventilatory Thresholds: Systematic Review and Meta‐Analysis.” Sports Medicine 50(10): 1771–1783. 10.1007/s40279-020-01314-8. [DOI] [PubMed] [Google Scholar]
  14. Gallo, Gabriele , Faelli Emanuela Luisa, Ruggeri Piero, Filipas Luca, Codella Roberto, Plews Daniel J., and Maunder Ed. 2024. “Power Output at the Moderate‐To‐Heavy Intensity Transition Decreases in a Non‐linear Fashion During Prolonged Exercise.” European Journal of Applied Physiology: Advance online publication. 10.1007/s00421-024-05440-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Goldberger, Ary L. , Amaral Luis A. N., Hausdorff Jeffrey M., Ivanov Plamen Ch, Peng C. K., and Stanley H. Eugene. 2002. “Fractal Dynamics in Physiology: Alterations with Disease and Aging.” PNAS 99(suppl_1): 2466–2472. 10.1073/pnas.012579499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gronwald, Thomas , Berk Sander, Altini Marco, Mourot Laurent, Hoos Olaf, and Rogers Bruce. 2021. “Real‐Time Estimation of Aerobic Threshold and Exercise Intensity Distribution Using Fractal Correlation Properties of Heart Rate Variability: A Single‐Case Field Application in a Former Olympic Triathlete.” Frontiers in Sports and Active Living 3: 668812. 10.3389/fspor.2021.668812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gronwald, Thomas , and Hoos Olaf. 2020. “Correlation Properties of Heart Rate Variability During Endurance Exercise: A Systematic Review.” Annals of Noninvasive Electrocardiology 25(1): e12697. 10.1111/anec.12697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gronwald, Thomas , Hoos Olaf, and Hottenrott Kuno. 2019. “Effects of Acute Normobaric Hypoxia on Non‐linear Dynamics of Cardiac Autonomic Activity during Constant Workload Cycling Exercise.” Frontiers in Physiology 10: 999. 10.3389/fphys.2019.00999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gronwald, Thomas , Ludyga Sebastian, Hoos Olaf, and Hottenrott Kuno. 2018. “Non‐linear Dynamics of Cardiac Autonomic Activity During Cycling Exercise with Varied Cadence.” Human Movement Science 60: 225–233. 10.1016/j.humov.2018.06.013. [DOI] [PubMed] [Google Scholar]
  20. Gronwald, Thomas , Rogers Bruce, and Hoos Olaf. 2020. “Fractal Correlation Properties of Heart Rate Variability: A New Biomarker for Intensity Distribution in Endurance Exercise and Training Prescription?” Frontiers in Physiology 11: 550572. 10.3389/fphys.2020.550572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gronwald, Thomas , Rogers Bruce, Hottenrott Laura, Hoos Olaf, and Hottenrott Kuno. 2021. “Correlation Properties of Heart Rate Variability During a Marathon Race in Recreational Runners: Potential Biomarker of Complex Regulation During Endurance Exercise.” Journal of Sports Science and Medicine 20(4): 557–563. 10.52082/jssm.2021.557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gronwald, Thomas , Törpel Alexander, Herold Fabian, and Budde Henning. 2020. “Perspective of Dose and Response for Individualized Physical Exercise and Training Prescription.” Journal of Functional Morphology and Kinesiology 5(3): 48. 10.3390/jfmk5030048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hamilton, Kate , Kilding Andrew E., Plews Daniel J., Mildenhall Mathew J., Waldron Mark, Charoensap Thanchanok, Cox Tobias H., Brick Matthew J., Leigh Warren B., and Maunder Ed. 2024. “Durability of the Moderate‐To‐Heavy‐Intensity Transition Is Related to the Effects of Prolonged Exercise on Severe‐Intensity Performance.” European Journal of Applied Physiology: Advance online publication. 10.1007/s00421-024-05459-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Haugen, Thomas , Sandbakk Øyvind, Seiler Stephen, and Tønnessen Espen. 2022. “The Training Characteristics of World‐Class Distance Runners: An Integration of Scientific Literature and Results‐Proven Practice.” Sports Medicine ‐ Open 8(1): 46. 10.1186/s40798-022-00438-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hautala, Arto J. , Mäkikallio Timo H., Seppänen Tapio, Huikuri Heikki V., and Tulppo Mikko P.. 2003. “Short‐term Correlation Properties of R‐R Interval Dynamics at Different Exercise Intensity Levels.” Clinical Physiology and Functional Imaging 23(4): 215–223. 10.1046/j.1475-097x.2003.00499.x. [DOI] [PubMed] [Google Scholar]
  26. Hedges, Larry V. 1981. “Distribution Theory for Glass’s Estimator of Effect Size and Related Estimators.” Journal of Educational Statistics 6(2): 107–128. 10.3102/10769986006002107. [DOI] [Google Scholar]
  27. Hernando, David , Hernando Alberto, Casajús Jose A., Laguna Pablo, Garatachea Nuria, and Bailón Raquel. 2018. “Methodological Framework for Heart Rate Variability Analysis During Exercise: Application to Running and Cycling Stress Testing.” Medical, & Biological Engineering & Computing 56(5): 781–794. 10.1007/s11517-017-1724-9. [DOI] [PubMed] [Google Scholar]
  28. Hofmann, Peter , and Tschakert Gerhard. 2017. “Intensity‐ and Duration‐Based Options to Regulate Endurance Training.” Frontiers in Physiology 8: 337. 10.3389/fphys.2017.00337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hopker, J. G. , Jobson S. A., and Pandit J. J.. 2011. “Controversies in the Physiological Basis of the 'Anaerobic Threshold' and Their Implications for Clinical Cardiopulmonary Exercise Testing.” Anaesthesia 66(2): 111–123. 10.1111/j.1365-2044.2010.06604.x. [DOI] [PubMed] [Google Scholar]
  30. Hunt, Kenneth J. , and Saengsuwan Jittima. 2018. “Changes in Heart Rate Variability with Respect to Exercise Intensity and Time During Treadmill Running.” BioMedical Engineering Online 17(1): 128. 10.1186/s12938-018-0561-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Iannetta, Danilo , de Almeida Azevedo Rafael, Keir Daniel A., and Murias Juan M.. 2019. “Establishing the V̇o2 VSersus Constant‐Work‐Rate Relationship from Ramp‐Incremental Exercise: Simple Strategies for an Unsolved Problem.” Journal of Applied Physiology 127(6): 1519–1527. 10.1152/japplphysiol.00508.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Iannetta, Danilo , Inglis Erin Calaine, Mattu Anmol T., Fontana Federico Y., Pogliaghi Silvia, Keir Daniel A., and Murias Juan M.. 2020. “A Critical Evaluation of Current Methods for Exercise Prescription in Women and Men.” Medicine & Science in Sports & Exercise 52(2): 466–473. 10.1249/mss.0000000000002147. [DOI] [PubMed] [Google Scholar]
  33. Iannetta, Danilo , Murias Juan M., and Keir Daniel A.. 2019. “A Simple Method to Quantify the V˙O2 Mean Response Time of Ramp‐Incremental Exercise.” Medicine & Science in Sports & Exercise 51(5): 1080–1086. 10.1249/mss.0000000000001880. [DOI] [PubMed] [Google Scholar]
  34. Jamnick, Nicholas A. , Botella Javier, Pyne David B., and Bishop David J.. 2018. “Manipulating Graded Exercise Test Variables Affects the Validity of the Lactate Threshold and VO2peak.” PLoS One 13(7): e0199794. 10.1371/journal.pone.0199794. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Jamnick, Nicholas A. , Pettitt Robert W., Granata Cesare, Pyne David B., and Bishop David J.. 2020. “An Examination and Critique of Current Methods to Determine Exercise Intensity.” Sports Medicine 50(10): 1729–1756. 10.1007/s40279-020-01322-8. [DOI] [PubMed] [Google Scholar]
  36. Jones, Andrew M. 2023. “The Fourth Dimension: Physiological Resilience as an Independent Determinant of Endurance Exercise Performance.” The Journal of Physiology: Advance online publication. 10.1113/JP284205. [DOI] [PubMed] [Google Scholar]
  37. Joyner, M. J. 1991. “Modeling: Optimal Marathon Performance on the Basis of Physiological Factors.” Journal of Applied Physiology 70(2): 683–687. 10.1152/jappl.1991.70.2.683. [DOI] [PubMed] [Google Scholar]
  38. Joyner, Michael J. , and Coyle Edward F.. 2008. “Endurance Exercise Performance: The Physiology of Champions.” The Journal of physiology 586(1): 35–44. 10.1113/jphysiol.2007.143834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Kanniainen, Matias , Pukkila Teemu, Kuisma Joonas, Molkkari Matti, Lajunen Kimmo, and Räsänen Esa. 2023. “Estimation of Physiological Exercise Thresholds Based on Dynamical Correlation Properties of Heart Rate Variability.” Frontiers in Physiology 14: 1299104. 10.3389/fphys.2023.1299104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Kauffman, S. A. 1995. At Home in the Universe: The Search for Laws of Self‐Organization and Complexity. Oxford, UK: Oxford University Press. [Google Scholar]
  41. Kaufmann, Sebastian , Gronwald Thomas, Herold Fabian, and Hoos Olaf. 2023. “Heart Rate Variability‐Derived Thresholds for Exercise Intensity Prescription in Endurance Sports: A Systematic Review of Interrelations and Agreement with Different Ventilatory and Blood Lactate Thresholds.” Sports Medicine ‐ Open 9(1): 59. 10.1186/s40798-023-00607-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Li, Siu Nam , Peeling Peter, Scott Brendan R., Peiffer Jeremiah J., Shaykevich Alex, and Girard Olivier. 2023. “Automatic Heart Rate Clamp: A Practical Tool to Control Internal and External Training Loads during Aerobic Exercise.” Frontiers in Physiology 14: 1170105. 10.3389/fphys.2023.1170105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Mann, Theresa , Lamberts Robert Patrick, and Lambert Michael Ian. 2013. “Methods of Prescribing Relative Exercise Intensity: Physiological and Practical Considerations.” Sports Medicine 43(7): 613–625. 10.1007/s40279-013-0045-x. [DOI] [PubMed] [Google Scholar]
  44. Mateo‐March, Manuel , Moya‐Ramón Manuel, Javaloyes Alejandro, Sánchez‐Muñoz Cristóbal, and Clemente‐Suárez Vicente J.. 2023. “Validity of Detrended Fluctuation Analysis of Heart Rate Variability to Determine Intensity Thresholds in Elite Cyclists.” European Journal of Sport Science 23(4): 580–587. 10.1080/17461391.2022.2047228. [DOI] [PubMed] [Google Scholar]
  45. Maunder, Ed , Seiler Stephen, Mildenhall Mathew J., Kilding Andrew E., and Plews Daniel J.. 2021. “The Importance of 'Durability' in the Physiological Profiling of Endurance Athletes.” Sports Medicine 51(8): 1619–1628. 10.1007/s40279-021-01459-0. [DOI] [PubMed] [Google Scholar]
  46. McKay, Alannah K. A. , Stellingwerff Trent, Smith Ella S., Martin David T., Mujika Iñigo, Goosey‐Tolfrey Vicky L., Sheppard Jeremy, and Burke Louise M.. 2022. “Defining Training and Performance Caliber: A Participant Classification Framework.” International Journal of Sports Physiology and Performance 17(2): 317–331. 10.1123/ijspp.2021-0451. [DOI] [PubMed] [Google Scholar]
  47. Meyer, T. , Lucía A., Earnest C. P., and Kindermann W.. 2005. “A Conceptual Framework for Performance Diagnosis and Training Prescription from Submaximal Gas Exchange Parameters‐‐Theory and Application.” International Journal of Sports Medicine 26(Suppl 1): S38–S48. 10.1055/s-2004-830514. [DOI] [PubMed] [Google Scholar]
  48. Meyler, Samuel , Bottoms Lindsay, Wellsted David, and Muniz‐Pumares Daniel. 2023. “Variability in Exercise Tolerance and Physiological Responses to Exercise Prescribed Relative to Physiological Thresholds and to Maximum Oxygen Uptake.” Experimental Physiology 108(4): 581–594. 10.1113/ep090878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Molkkari, Matti , Angelotti Giorgio, Emig Thorsten, and Räsänen Esa. 2020. “Dynamical Heart Beat Correlations during Running.” Scientific Reports 10(1): 13627. 10.1038/s41598-020-70358-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Nicolò, Andrea , Massaroni Carlo, and Passfield Louis. 2017. “Respiratory Frequency During Exercise: The Neglected Physio‐Logical Measure.” Frontiers in Physiology 8: 922. 10.3389/fphys.2017.00922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Nicolò, Andrea , and Sacchetti Massimo. 2023. “Differential Control of Respiratory Frequency and Tidal Volume during Exer‐Cise.” European Journal of Applied Physiology 123(2): 215–242. 10.1007/s00421-022-05077-0. [DOI] [PubMed] [Google Scholar]
  52. Pallarés, Jesús G. , Morán‐Navarro Ricardo, Ortega Juan Fernando, Fernández‐Elías Valentín Emilio, and Mora‐Rodriguez Ricardo. 2016. “Validity and Reliability of Ventilatory and Blood Lactate Thresholds in Well‐Trained Cyclists.” PLoS One 11(9): e0163389. 10.1371/journal.pone.0163389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Peng, C. K. , Havlin Shlomo, Stanley H. Eugene, and Goldberger Ary L.. 1995. “Quantification of Scaling Exponents and Crossover Phenomena in Nonstationary Heartbeat Time Series.” Chaos 5(1): 82–87. 10.1063/1.166141. [DOI] [PubMed] [Google Scholar]
  54. Persson, P. B. 1996. “Modulation of Cardiovascular Control Mechanisms and Their Interaction.” Physiological Reviews 76(1): 193–244. 10.1152/physrev.1996.76.1.193. [DOI] [PubMed] [Google Scholar]
  55. Poole, David C. , Rossiter Harry B., Brooks George A., and Gladden L. Bruce. 2021. “The Anaerobic Threshold: 50+ Years of Controversy.” The Journal of Physiology 599(3): 737–767. 10.1113/jp279963. [DOI] [PubMed] [Google Scholar]
  56. Rogers, Bruce , Giles David, Draper Nick, Mourot Laurent, and Gronwald Thomas. 2021. “Detection of the Anaerobic Threshold in Endurance Sports: Validation of a New Method Using Correlation Properties of Heart Rate Variability.” Journal of Functional Morphology and Kinesiology 6(2): 38. 10.3390/jfmk6020038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Rogers, Bruce , Giles David, Draper Nick, Hoos Olaf, and Gronwald Thomas. 2021. “A New Detection Method Defining the Aerobic Threshold for Endurance Exercise and Training Prescription Based on Fractal Correlation Properties of Heart Rate Variability.” Frontiers in Physiology 11: 596567. 10.3389/fphys.2020.596567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Rogers, Bruce , and Gronwald Thomas. 2022. “Fractal Correlation Properties of Heart Rate Variability as a Biomarker for Intensity Distribution and Training Prescription in Endurance Exercise: An Update.” Frontiers in Physiology 13: 879071. 10.3389/fphys.2022.879071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Rogers, Bruce , Mourot Laurent, Doucende Gregory, and Gronwald Thomas. 2021. “Fractal Correlation Properties of Heart Rate Variability as a Biomarker of Endurance Exercise Fatigue in Ultramarathon Runners.” Physiological Reports 9(14): e14956. 10.14814/phy2.14956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Schaffarczyk, Marcelle , Rogers Bruce, Reer Rüdiger, and Gronwald Thomas. 2022. “Fractal Correlation Properties of HRV as a Noninvasive Biomarker to Assess the Physiological Status of Triathletes During Simulated Warm‐Up Sessions at Low Exercise Intensity: A Pilot Study.” BMC Sports Science, Medicine & Rehabilitation 14(1): 203. 10.1186/s13102-022-00596-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Schaffarczyk, Marcelle , Rogers Bruce, Reer Rüdiger, and Gronwald Thomas. 2022. “Validity of the Polar H10 Sensor for Heart Rate Variability Analysis During Resting State and Incremental Exercise in Recreational Men and Women.” Sensors 22(17): 6536. 10.3390/s22176536. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Schaffarczyk, Marcelle , Rogers Bruce, Reer Rüdiger, and Gronwald Thomas. 2023. “Validation of a Non‐linear Index of Heart Rate Variability to Determine Aerobic and Anaerobic Thresholds During Incremental Cycling Exercise in Women.” European Journal of Applied Physiology 123(2): 299–309. 10.1007/s00421-022-05050-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Sempere‐Ruiz, Noemí , Sarabia José Manuel, Baladzhaeva Sabina, and Moya‐Ramón Manuel. 2024. “Reliability and Validity of a Non‐linear Index of Heart Rate Variability to Determine Intensity Thresholds.” Frontiers in Physiology 15: 1329360. 10.3389/fphys.2024.1329360. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Smyth, Barry , Maunder Ed, Meyler Samuel, Hunter Ben, and Muniz‐Pumares Daniel. 2022. “Decoupling of Internal and External Workload During a Marathon: An Analysis of Durability in 82,303 Recreational Runners.” Sports Medicine 52(9): 2283–2295. 10.1007/s40279-022-01680-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Stevenson, Julian D. , Kilding Andrew E., Plews Daniel J., and Maunder Ed. 2024. “Prolonged Exercise Shifts Ventilatory Para‐Meters at the Moderate‐To‐Heavy Intensity Transition.” European Journal of Applied Physiology 124(1): 309–315. 10.1007/s00421-023-05285-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Tschakert, Gerhard , Handl Tanja, Weiner Lena, Birnbaumer Philipp, Mueller Alexander, Groeschl Werner, and Hofmann Peter. 2022. “Exercise Duration: Independent Effects on Acute Physiologic Responses and the Need for an Individualized Prescription.” Physiological Reports 10(3): e15168. 10.14814/phy2.15168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Van Hooren, Bas , Bongers Bart C., Rogers Bruce, and Gronwald Thomas. 2023. “The Between‐Day Reliability of Correlation Properties of Heart Rate Variability during Running.” Applied Psychophysiology and Biofeedback 48(4): 453–460. 10.1007/s10484-023-09599-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Van Hooren, Bas , Mennen Bram, Gronwald Thomas, Bongers Bart C., and Rogers Bruce. 2023. “Correlation Properties of Heart Rate Variability to Assess the First Ventilatory Threshold and Fatigue in Runners.” Journal of Sports Sciences: 1–10: Advance online publication. 10.1080/02640414.2023.2277034. [DOI] [PubMed] [Google Scholar]
  69. Van Hooren, Bas , Souren Tjeu, and Bongers Bart C.. 2024. “Accuracy of Respiratory Gas Variables, Substrate, and Energy Use from 15 CPET Systems during Simulated and Human Exercise.” Scandinavian Journal of Medicine & Science in Sports 34(1): e14490. 10.1111/sms.14490. [DOI] [PubMed] [Google Scholar]
  70. White, Daniel W. , and Raven Peter B.. 2014. “Autonomic Neural Control of Heart Rate during Dynamic Exercise: Revisi‐Ted.” The Journal of Physiology 592(12): 2491–2500. 10.1113/jphysiol.2014.271858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Zuccarelli, Lucrezia , Porcelli Simone, Rasica Letizia, Marzorati Mauro, and Grassi Bruno. 2018. “Comparison between Slow Components of HR and V˙O2 Kinetics: Functional Significance.” Medicine & Science in Sports & Exercise 50(8): 1649–1657. 10.1249/mss.0000000000001612. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Data are available from the corresponding author on reasonable request.

Articles from European Journal of Sport Science are provided here courtesy of Wiley

RESOURCES