Effects of HRV-Guided vs. Predetermined Block Training on Performance, HRV and Serum Hormones

Abstract

The aim of this study was to compare heart rate variability -guided (HRVG) and predetermined (PD) block periodization of high intensity aerobic training (HIT). Endurance performance, neuromuscular performance, heart rate variability (HRV) and serum hormone concentrations were measured before, in the middle and after the 8-week training period in 24 endurance trained males. Both groups improved significantly maximal treadmill velocity (V_max) (p<0.001) and 3000 m running performance (HRVG; p<0.001 and PD; p=0.001). The relative changes in V_max and countermovement jump were significantly greater in HRVG (p<0.05). Nocturnal heart rate decreased in both groups (p<0.01), but HRV (RMSSD, LF and TP) increased significantly only in HRVG (p<0.05). The significant increase in serum testosterone concentration was observed from mid to post in HRVG (p<0.05). Significant correlations were found between individual V_max changes and absolute serum testosterone levels. Individual baseline level of HF correlated significantly with V_max changes in PD. Block periodization of HIT seems to be an effective way to improve endurance and running performance in already endurance trained males. Based on training induced increases in endurance and neuromuscular performance combined with significant changes in HRV and serum testosterone levels observed in HRVG, individually HRV -guided block training may be more optimal compared to predetermined training.

Key word

running performance - block periodization - VO2max - interval training - autonomic nervous system - heart rate variability

Introduction

The questions that endurance coaches and athletes ask themselves daily are for how long, how hard and how often. Studies by Tonnessen et al. [41] [42] found that elite athletes divide their endurance training quite uniformly into 80% low-intensity and 20% high-intensity training. Recent research further suggests that division should be done using a polarized model [24] [40]. It has been shown that both low and high-intensity endurance training are needed to gain favourable peripheral adaptations in the muscle and central adaptations in the circulatory system [39]. Despite consensus on training intensity distribution, optimal periodization during shorter microcycles and over the whole year allows much more room for different interpretations.

Block periodization of high-intensity intervals has been shown to be an effective way to improve endurance performance [32]. There have been shorter shock microcycles lasting 1–2 weeks with almost daily HIT sessions [2] [4] and longer periods of 4–12 weeks of alternating LIT and HIT blocks [32] [34]. Both models have improved VO_2max, time to exhaustion and submaximal endurance measured as speed or power at lactate threshold. It has been speculated that a high number of highly concentrated workloads may allow greater improvements than more concurrent kinds of training. The idea behind block training is to train different target abilities in series rather than concurrently [18]. However, the actual mechanisms behind block periodization and its effects on heart rate variability (HRV), serum hormone concentrations and neuromuscular performance have remained mostly unsolved.

During intensive training periods that may lead to overreaching, the role of monitoring performance and fatigue becomes more important to ensure sufficient recovery [29]. Monitoring can be divided into external and internal methods. The external methods include performance tests like countermovement jumps [5] or submaximal performance tests [47]. The internal methods include markers such as hormone concentrations of testosterone [6] or testosterone/cortisol ratio [13] and heart rate or HRV measurements [19] [45]. As a non-invasive method to evaluate the autonomic nervous system function, HRV is a potential tool to analyse the current recovery status during intensive training periods. It has been shown that HRV decreases after heavy and moderate endurance sessions [16]. Both the intensity [25] [38] and the duration of the work performed [25] may have an effect on the magnitude of delay observed in the recovery of the autonomic nervous system.

HRV-guided training has been studied, for example, by Kiviniemi et al. [19] and Vesterinen et al. [45]. The idea behind HRV-guided training is to adjust the training load or intensity based on the autonomic nervous system status. It is assumed that the decrease in HRV indicates lowered cardiac parasympathetic modulation, which in turn may be related to the reduced level of the recovery status [16] [38]. In studies by Vesterinen et al. [45] and Kiviniemi et al. [19], HRV has been monitored daily, and the intensity of daily endurance sessions has been defined by the result of the HRV test after having been compared to an individually scaled reference or control value. For the definition of changes in HRV, it has been recommended to use assessment of averages of longer periods instead of individual daily values due to the natural day-to-day variation in HRV [30].

Due to high demands of HIT blocks, also observed as changes in autonomic modulation, one may speculate that HRV guidance of these blocks may allow a more optimal outcome compared to predetermined programming. To the best of our knowledge, no research using HRV guidance of HIT blocks has been published.

The purpose of this study was to compare predetermined and HRV-guided block periodization of HIT and its effects on endurance and neuromuscular performance, HRV and serum hormone concentrations. We hypothesized that HRV-guided training provides greater adaptations compared to predetermined training.

Materials and Methods

Subjects

Thirty-two recreationally trained males were recruited for this study. Subjects were 19–37 years old and used to regular endurance training. Resting ECG was checked by a cardiologist before inclusion in the study. During the intervention there were five drop-outs due to illness (n=1), injuries (n=2) and personal reasons (n=3). Three subjects were excluded due to poor adherence to training (less than 90% of sessions). Finally, twenty-four subjects were included in the study analyses. After the control tests, subjects were divided into pairs based on their age, training background, 3000 m performance and resting HRV. After that, subjects were randomly assigned to the HRV-guided group (HRVG, n=13, age: 29±4 years, height: 180±7 cm, weight: 76.4±9.4 kg) and predetermined group (PD, n=11, age: 31±5 years, height: 176±5 cm, weight: 74.0±5.7 kg). The study was approved by the Ethics Committee of the University of Jyväskylä and it meets the ethical standards of the journal [10].

Experimental design

The study consisted of the 3-week control period and the 8-week training period. During the control period subjects maintained their regular amount of endurance training. However, they were instructed to plan their training so that they were fully recovered at the beginning of the training period. In addition, one interval session (3×10×30 s) and one strength session was preprogrammed to familiarise subjects with these sessions before the training intervention. The control period started from the control tests and ended before the pre-tests. After the pre-tests, the next three days were preprogrammed in both groups, but after that the groups utilized their own training program. After four weeks of training, the mid-tests were performed and the training program started from the beginning in both groups. The post-tests were performed after the 8-week training period.

Anthropometrics, neuromuscular measurements and 3000 m running tests were performed at the beginning of the control period, the beginning of the training period, at the middle and after the training period. The tests were performed during one day so that subjects arrived in a fasted state to give blood samples and for anthropometric measurements. After these measurements, a light breakfast was eaten. Thereafter, the maximum running velocity test, countermovement jump (CMJ) and 1RM dynamic leg press were performed. In the afternoon, the 3000 m running test was performed. The incremental treadmill test was performed before and after the training period. All the tests were performed individually at the same time of the day (± 2 h). Before each test, at least three days of low-intensity training was performed.

Training

Endurance training consisted of low-intensity training (LIT) and high-intensity training (HIT). All sessions were specified to be performed on a flat, solid surface and individually at the same time of the day. Subjects wore Garmin XT920 heart rate monitors (Garmin Ltd, Schaffhausen, Switzerland) in each training session. They also kept a training diary and wrote down the training mode, session length, heart rate and their own comments. GPS and heart rate data from each training session were sent to the research group to be checked manually. Every week at least one voluntary supervised session was held. From the training data, a weekly training frequency, amount of endurance, other and total training, and intensity distribution with the time-in-zone method on a three-zone scale (1<82% HRmax, 2=82–87% HRmax and 3>87% HRmax) were analyzed. In addition, the weekly training distribution based on the session goal of the training mode (HIT, LIT, strength) was analyzed.

LIT sessions were performed under the individual aerobic threshold. Subjects were instructed to maintain their typical length of LIT sessions, but at least 30 min and at the maximum of 90 min. One longer LIT session (over 60 min) was performed every other week. Sessions included mainly running, but alternative forms were also allowed to avoid overuse injuries [45]. Two types of interval sessions were performed during the training period, which were 4×4 min intervals at the intensity corresponding to 90–95% of maximal heart rate, individually over the anaerobic threshold with 3 min of active recovery between intervals [12] and 3×10×30 s at a running velocity equal to 95% of V_max with 15 s of active recovery between intervals and 3 min between sets [33].

A strength session was performed five times during the training period. The session was a mixture of maximum and explosive strength training. Leg press, knee flexion and two upper body exercises were performed at loads of 70–85% of 1RM with 2–3 sets and 5–10 repetitions. Bench step (body weight) and half squat (30/40/60%/1RM) were performed as explosively as possible using 3 sets of 5–6 repetitions. In addition, two core exercises were performed with three sets of 20 repetitions.

The training at the PD group consisted of HIT block weeks (4–5 HIT sessions) and recovery weeks (1 HIT session) which were performed in turns through the training period [34]. One rest day was included in each week. The same 4-week training program model was used in weeks 1–4 and 5–8. ([Fig. 1]).

Fig. 1 Description of the training program in PD and HRVG. In PD the program was divided to HIT block weeks and recovery weeks. HRVG had the same training modes, but the program was divided to six blocks (B1–B6). Moving from one block to another was based on the quick recovery test result. Only LIT sessions were performed in HRVG as long as the test result was below the individual reference values. Strength sessions were performed as LIT sessions and they were placed by research group. Both groups started the similar training program from the beginning after four weeks of training.

The HRVG group had the same training modes as the PD group, but the way those modes were programmed differed. In HRV, the training program was divided into six blocks ([Fig. 1]). Each block was preprogrammed, but moving from one block to another was based on the quick recovery test result (Firstbeat Technologies Ltd, Jyväskylä, Finland) performed each morning. The control period allowed the software to adapt to the individual range of heart rate and HRV. During the training period, a three-day running average of the quick recovery test results was used for training guidance due to day-to-day variation in HRV [30]. The individual average of all the quick recovery test results from the control period was used as a reference value for starting the next block. After finishing the predetermined block, LIT was performed until the three-day running average of the QRT score was higher than the individual reference value. After the mid-tests or finishing block 6, subjects started the training program from block 1. The strength sessions were programmed by the research group between HIT blocks, so that they were performed as an LIT session and at the same time as in the PD group.

Subjects performed the quick recovery test (Firstbeat Technologies Ltd, Jyväskylä, Finland) in a supine position every morning during the control and the training period. The test required 3 min of RR interval data collection. The software performed artefact correction and data filtering [35]. The RR interval data was collected with a Garmin 920XT heart rate monitor (Garmin Ltd, Schaffhausen, Switzerland). The quick recovery test score was derived by the Firstbeat SPORTS Monitor v. 2.0 (Firstbeat Technologies Ltd, Jyväskylä, Finland) using heart rate and RMSSD parameters for describing vagal activity. The results were adaptively scaled based on the average and standard deviation of the user’s personal measurement history. The results were presented from 0 to 100% (0–30% poor, 30–70% moderate, 70–90% good, and 90–100% excellent recovery).

3000-meter running and incremental treadmill tests

The 3000 m running test was performed on the 200 m indoor running track. Before the test, a 15 min warm-up was performed. The running tests were performed in groups of four subjects on average. After the test lactate samples were taken from the fingertip immediately after and 4 min after the end of the test (Biosen S_line Lab+ lactate analyzer, EKF Diagnostic, Magdeburg, Germany). In addition, the maximum heart rate was analyzed.

The incremental treadmill test was performed in the laboratory of Biology of Physical Activity at the University of Jyväskylä (Telineyhtymä, Kotka, Finland). The test started at a velocity of 8 km/h or 10 km/h based on the fitness level of each subject. The same starting velocity was used in both tests. The 8 km/h velocity was increased to 10 km/h and by 1 km/h at each 3 min stage thereafter. After each stage the treadmill was stopped for fingertip blood samples (20 s). Lactate samples were analyzed with Biosen S_line Lab+ lactate analyzer (EKF Diagnostics, Magdeburg, Germany). During the test, the heart rate was recorded with a Garmin XT920 heart rate monitor (Garmin Ltd, Schaffhausen, Switzerland). The incline was kept at 0.5 degrees through the test. Oxygen consumption was measured using the breath-by-breath test (OxyconPro, Jaeger, Hoechberg, Germany). Before each test the gas analyzer was manually calibrated.

VO_2max was defined as the highest 60 s average of oxygen consumption. V_max was defined as the highest speed finished, or if the stage was not finished, as the speed of the last completed stage (km/h) +(running time (s) of the unfinished stage – 30 seconds) / (180 – 30 seconds)*1 km/h. Aerobic (LT1) and anaerobic (LT2) thresholds were determined based on lactate values during the test. The aerobic threshold was set at 0.3 mmol/l above the lowest lactate value and the anaerobic threshold at the intersection point between 1) a linear model between LT1 and the next lactate point, and 2) a linear model for the lactate points with an La increase of (at least) 0.8 mmol/l [46].

Anthropometrics and neuromuscular measurements

Body weight and fat percentage were analyzed after 12 h of fasting using the InBody720 analyzer (Biospace Co. Ltd, Seoul, South Korea).

The maximum running velocity test was performed in the indoor track. Before the test, subjects performed 10 min warm-up, dynamic stretching and three accelerations of 40–50 meters. The maximum velocity (m/s) was calculated from the 10 m running distance between the photocells after the 25 m acceleration. Subjects had three attempts unless more than 5% improvement was observed between the second and third attempts. Between the attempts a recovery of 2 min was allowed.

Countermovement jumps were performed on a force plate (Department of Biology of Physical Activity, Jyväskylä, Finland). During the jumps, subjects held their hands on their hips. Knee angle was instructed to be at 90 degrees. Three jumps were performed with 1 min recovery between jumps, unless more than 5% improvement was observed between the second and third attempts. Jumping height was analyzed from the force impulse. The analysis was done with the Signal 4.10 program (Cambridge Electronic Design Ltd, Cambridge, UK).

The dynamic leg press action was performed concentrically using the David 210 dynamometer (David Sports Ltd., Helsinki, Finland). The starting knee angle was individually set to 60 degrees. The warm-up protocol consisted of five repetitions with the loads at 70% of 1RM, three repetitions at 80% of 1RM and two repetitions at 90% of 1RM. Between the sets, a one-minute recovery was allowed. After the warm-up, one repetition at a time was performed until the subject could not finish the increased load. Between the repetitions a 1.5 min recovery was allowed.

Heart rate variability

Heart rate and HRV were recorded every morning and every other night throughout the study period. The morning measurements were done with Garmin 920XT heart rate monitors (Garmin Ltd, Schaffhausen, Switzerland). Subjects were instructed to perform the measurement right after awakening and emptying the urinary bladder. The measurement was 3 min long and performed in a supine position. Before starting data collection, subjects were instructed to wait until their heart rate became steady. Subjects sent HR data to the research group and data was analyzed with the Firstbeat Sports software. The weekly average of the morning heart rate and RMSSD was analyzed. Nocturnal measurements were done with the Firstbeat Bodyguard device (Firstbeat Technologies Ltd, Jyväskylä, Finland). Subjects were instructed to put the device on when going to sleep and to release the device immediately after awakening. From the nocturnal measurements, the 4-h period starting 30 min after going to sleep was analyzed. Recorded RR intervals were edited by an artefact detection filter in the Firstbeat Sports software, which excluded all falsely detected, missed, and premature heartbeats. If the error percentage representing the number of corrected interbeat intervals shown by the software was higher than 33%, recordings were excluded from the analysis in line with the suggestion by Vesterinen et al. [44]. Heart rate, RMSSD, low frequency (LF), high frequency (HF), and total power (TP) were analyzed from the whole control period and during training weeks 4 and 8.

Serum hormone concentrations

Serum hormone concentrations were measured at the same time of the day (8:00–9:00) after 12 h of fasting. Blood samples were taken from the antecubital vein into serum tubes (Vacuette, Greiner Bio One International GmbH, Kremsmünster, Austria) using standard laboratory procedures. Whole blood was centrifuged at 3500 rpm (Megafuge 1.0R, Heraeus, Hanau, Germany) for 10 min. After that serum was removed and refrigerated at −80 degrees until the final analysis. Serum testosterone and cortisol were analyzed with chemical luminescence techniques (Immulite 2000 XPi, Siemens, NY, USA) and hormone-specific immunoassay kits (Siemens, NY, USA). The sensitivity of testosterone and cortisol assays was 0.5 nmol/l and 5.5 nmol/l, respectively. The intra-assay coefficients of variation for testosterone and cortisol were 7.3% and 8.3%, respectively.

Statistical analysis

All the values are presented as mean±standard deviation. Normality of the data was assessed with the Shapiro-Wilk test. To test for differences between the groups at baseline and within groups between the control and pre-tests, unpaired two-tailed t-tests and paired two-tailed t-tests were used. Within-group differences on the incremental treadmill test were compared using a paired two-tailed t-test (VO_2max, V_max, LT1, and LT2). Neuromuscular performance, 3000 m test, serum hormone concentrations and HRV were analyzed using repeated measures ANOVA. If the ANOVA reached significance, a Fisher’s LSD test was performed for post hoc analysis. VO_2max l/min and LF (ms²) values of the PD group were not normally distributed, so the data was analyzed using a nonparametric Wilcoxon signed-rank test. To test for differences in relative changes from the pre-intervention to post-intervention between the groups, unpaired Students t-tests were performed. In addition, effect size (ES) of between-group differences in the relative changes of key performance and physiological variables was calculated as Cohen’s d. The magnitude of changes was stated as<0.2 trivial, 0.2–0.5 small, 0.5–0.8 moderate and >0.8 large. The correlation analysis was done using Pearson moment product method. Significance was set at p≤0.05*, p<0.01**  and p<0.001** *. Results were analyzed with Microsoft Excel 2010 (Microsoft Corporation, WA, USA) and the IBM SPSS Statistics v.24 program (SPSS Inc., IL, USA).

Results

There were no significant between-group differences in the baseline levels of endurance, neuromuscular, HRV or serum hormone concentration variables. No significant changes were found during the control period, except for maximal running velocity in HRVG.

Anthropometrics

There were no significant changes from the pre- to post-training period in body weight (HRVG 76.5±9.0 kg vs. 76.4±9.4 kg; PD 74.0±5.5 kg vs. 74.0±5.7 kg) or fat percent (HRVG 12.6±4.2% vs. 12.6±4.4%; PD 12.6±2.7% vs. 12.2±3.2%).

Training

No significant differences were observed between the groups in the amount of training or training intensity distribution during the control and training periods. In both groups, significant increases were found from the control period to the training period in training frequency (HRVG 5.3±2.1 vs. 6.3±1.4, p=0.007; PD 5.0±1.1 vs. 6.1±0.4, p=0.001) and amount of Zone 2 training (HRVG 10±7% vs. 15±6%, p<0.005; PD 6±4% vs. 12±5%, p=0.008). In the PD group a significant increase was found in the amount of endurance (4.7±1.7 h vs. 5.3±1.8 h p<0.001) and total training (5.2±1.8 h vs. 6.0±1.9 h p<0.001). Training characteristics of both groups during the training period are presented in [Table 1].

Table 1 Training characteristics of both groups during the 8-week training period.

	HRVG (n=13)	PD (n=11)
	Weeks 1–8	Weeks 1–8
Training frequency (sessions/week)	6.3±1.4	6.1±0.4
Endurance training (h/week)	5.1±2.1	5.3±1.8
HRzone1 (%)	82 ±8	84±7
HRzone2 (%)	15±6	12±5
HRzone3 (%)	3±3	4±3
Other training (h/week)	0.6±0.2	0.7±0.2
Total training (h/week)	5.7±2.1	6.0±1.9

Zone 1<82%/HR_max, zone 2=82–87%/HR_max and zone 3>87%/HR_max.

No significant differences were found between the groups in the number of HIT sessions during the training period. The total number of HIT sessions in PD and HRVG were, on average, 21.8±0.6 vs. 19.8±4.1, respectively. The PD group performed, on average, 10.9±0.3 HIT sessions during the first and last four weeks of the training period, whereas the HRVG group performed 10.3±2.7 HIT sessions during the first four weeks and 9.5±2.8 HIT sessions during the last four weeks. No significant correlation was found between the number of HIT sessions and endurance performance changes in the HRVG group.

In the weekly training distribution significant differences were found between the groups. During weeks 2 (p=0.008) and 7 (p=0.030) the relative amount of weekly HIT sessions was significantly different between groups. During weeks 1, 3, 4, 6, and 8, between-group differences in the relative amount of HIT sessions approached the significance level (p=0.054–0.075) ([Fig. 2]).

Fig. 2 Weekly average distribution of different training sessions in both groups. LIT=low-intensity training, HIT=high-intensity training, strength=strength training. Statistical significances: # p<0.05, between-group difference.

Endurance performance

Both groups increased their V_max significantly (HRVG p<0.001; PD p<0.001). A significant difference and large effect size (ES=0.95) between the groups was found in the relative increase of V_max (p=0.033) ([Fig. 3]). VO_2max relative to body weight and in absolute values increased in HRVG (p=0.001, p=0.011) and PD (p=0.005, p=0.036). Moderate effect size (ES=0.52) was found between the groups in the relative increase of absolute VO_2max. Anaerobic threshold (LT2) increased in both groups significantly (HRVG p<0.001; PD p=0.050). Significant increases were found in the aerobic threshold (LT1) in HRVG (p=0.021) and PD (p=0.027) ([Table 2]).

Fig. 3 Relative mean changes of the groups and individual changes of each subject in V_max. Black line presents the mean of the whole group (4.0%). Statistical significances: # p<0.05 between-group difference.

Table 2 Incremental treadmill test results and between-group effect sizes.

	HRVG (n=13)		PD (n=11)		ES (pre-post)
	Pre	Post	Pre	Post	HRVG vs. PD
VO 2max (ml/kg/min)	53.6±4.2	56.7±3.4**	54.2±4.1	56.4±4.7**	0.42 (small)
VO 2max (l/min)	4.1±0.3	4.3±0.4*	4.0±0.3	4.1±0.3*	0.52 (moderate)
V max (km/h)	17.6±1.3	18.5± 1.2***	18.0±1.1	18.5±1.2***	0.95 (large)
LT1 (km/h)	11.0±1.5	11.8±1.1*	11.6±1.2	12.2±1.2*	0.32 (small)
LT2 (km/h)	14.1±1.0	15.0±1.1***	14.7±0.9	15.3±1.2*	0.37 (small)

Statistical significances within group changes: * p<0.05, **  p<0.01, *** p<0.001

Both groups improved their performance in the 3000 m test from pre- to post-training (HRVG –5.2±2.4%, p<0.001; PD –5.2±3.1%, p=0.001), pre- to mid-training (HRVG –3.1±1.3%, p<0.001; PD –3.5±2.6%, p=0.002) and mid- to post-training (HRVG –2.2±1.5%, p<0.001; PD –1.5±1.1%, p=0.001). Maximum lactate values increased significantly in the HRVG group from mid- to post-training (12.8±18.4%, p=0.039). In the PD group, a similar trend was observed from pre- to post-training (16.0±23.5%, p=0.064) ([Table 3]).

Table 3 Running performance, serum hormone concentrations and heart rate variability (HRV) at pre-, mid- and post-training. Effect size of between-group differences was analysed from pre- to post-training. Hormones were measured on the same day as the running test was performed. HRV was analysed as an average of the control period (pre), week 4 (mid) and week 8 (post).

	HRVG (n=13)			PD (n=11)			ES (pre-post)
	Pre	Mid	Post	Pre	Mid	Post	HRVG vs. PD
Running
3000 m (min:s)	11:13±:50	10:52±:49***	10:38±:52***ab	11:07±:47	10:42±:42**	10:32±:41**  ab	0.00 (trivial)
MaxLa (mmol/l)	14.2±3.0	14.3±3.4	15.8±2.9*b	12.4±2.6	13.6±3.0	13.9±1.7
MaxHR (bpm/min)	187±9	186±7	187±7	193±6	192±7	192±7
Hormones
Testosterone (nmol/l)	19.0±5.3	17.7±4.8	20.6±4.8*b	18.3±5.6	16.5±4.3*	17.2±4.3	0.59 (moderate)
Cortisol (nmol/l)	446±70	464±91	488±109	458±79	474±50	482±65	0.03 (trivial)
Testosterone/cortisol	0.43±0.11	0.39±0.13	0.45±0.14	0.42±0.12	0.35±0.10	0.36±0.09	0.24 (small)
Heart rate variability
HR (bpm/min)	50.9±5.6	48.9±5.5	46.5±5.0** a	52.2±5.4	49.9±5.8*	48.6 ±5.5** a	0.20 (small)
RMSSD (ms)	76±25	80±22	89±22*a	67±12	72±19	81±29	0.10 (trivial)
LF (ms2)	4898±1415	5438±1532	6232±2090*a	5165±1904	5830±2301	5768±2204	0.65 (moderate)
HF (ms2)	4055±2313	4324±2177	4865±2085	3542±1210	3876±1552	4163±1923	0.36 (small)
TP (ms2)	8952±3265	9762±3208	11097±3814*a	8707±2727	9706±3342	9931±3353	0.43 (small)

Statistical significance within group changes: * p<0.05, ** p<0.01. ^a=Pre-post, ^b=Mid-post

Neuromuscular performance

No significant changes were found within groups in the CMJ during the training period. From pre- to mid-training, jumping height in PD tended to decrease (29.0±3.8 cm vs. 28.4±3.7 cm, p=0.073), while increasing trend was seen in HRVG from pre- to post-training (31.4±4.8 cm vs. 32.1±5.2 cm). A significant difference between the groups was found in the relative change of CMJ from pre- to post-training (p=0.048) with a large effect size (ES=0.88) ([Fig. 4]).

Fig. 4 Relative changes in countermovement jump and maximal running velocity from pre-mid, mid-post and pre-post. Statistical significances:  ** p<0.01 within groups, # p<0.05 between groups.

Maximal running velocity increased in HRVG significantly from the control to the pre-test (8.14±0.31 m/s vs. 8.20±0.32 m/s, p=0.008). From pre- to mid-test, maximal running velocity in PD decreased significantly (7.95±0.40 m/s vs. 7.87±0.36 m/s, p=0.008) and tended to decrease in HRVG (–0.5±1.0%, p=0.054). No other significant differences were found between or within groups ([Fig. 4]).

1RM increased from pre- to post-training significantly in both groups (HRVG; 206±29 kg vs. 229±32 kg, p=0.001; PD 202±32 kg vs. 225±33 kg, p<0.001). From pre- to mid-training, only PD increased leg press significantly (1.9±3.0% p=0.024). From mid- to post-training, HRVG (8.7±6.6%, p<0.001) and PD (9.5±6.3%, p=0.001) increased their 1RM. No significant differences were found between the groups in the relative change of 1RM.

Heart rate variability

Nocturnal heart rate decreased significantly from pre- to post-training in both groups (HRVG; p=0.004 and PD; p=0.008) ([Table 3]). The decrease in HR from pre- to mid-training was significant only in PD (p<0.042). In HRVG, a significant increase was observed from pre- to post-training in RMSSD (p=0.028), LF (p=0.024), and TP p=0.046), whereas no significant changes were found in PD. In the morning measurements, no significant differences were observed in the relative change of RMSSD from the control period to any week during the training period ([Fig. 5]). Heart rate was significantly lower compared to the control period during weeks 4–8 (p<0.05) in HRVG and during weeks 3, 7 and 8 in PD (p<0.05). Effect size showed a moderate between-group effect in heart rate change from the control period to week 8 (ES=0.65) and a small effect in RMSSD (ES=0.42)

Fig. 5 Weekly average of morning RMSSD and heart rate. Statistical significance of within-group changes from the control period: *p<0.05, **p<0.01.

Serum hormone concentrations

Serum testosterone concentration decreased significantly in PD from pre- to mid-training (p<0.037), whereas no significant change was observed in HRVG ([Table 3]). From mid- to post-training, testosterone increased significantly in HRVG (p<0.029). Effect size showed a moderate between-group effect in testosterone change from pre- to post-training. No significant within- or between-group differences were found in the serum concentration of cortisol or the testosterone/cortisol ratio, although the testosterone/cortisol ratio increase approached significance from mid- to post-training in HRVG (p=0.051).

Correlations

A significant correlation was found between individual baseline HF and individual changes in V_max in PD (r=0.656, p=0.028) ([Fig. 6]), while no such a correlation was observed in HRVG. Individual resting HR changes from pre- to post-training correlated with 3000 m changes (r=–0.630, p=0.01), as well as individual HF (r=0.488, p=0.018) and TP (r=0.467, p=0.025) changes from pre- to post-training in the total group of subjects. In the morning measurements, individual RMSSD changes from the control period to weeks 5–8 correlated significantly with V_max changes (r=0.499, p<0.015) in all subjects. In the nocturnal measurements, individual changes from mid- to post-training in HF (r=0.414, p=0.049) and TP (r=0.485, p=0.019) correlated with V_max changes in the total group of subjects. Individual TP (r=0.462, p=0.026) and HF (r=0.503, p=0.014) changes from pre- to post-training correlated positively with absolute average serum testosterone concentrations and average testosterone/cortisol ratio (r=0.479. p=0.021 and r=0.465, p=0.025, respectively) in all subjects. In HRVG a significant correlation was found between absolute morning RMSSD values and the number of HIT sessions during the last four weeks (r=0.592, p=0.042).

Fig. 6 a Correlation between the absolute HF values during the control period and relative changes of V_max in the PD group. b Correlation between average serum testosterone concentration at the pre, mid and post measurements and relative changes of V_max in all subjects.

Individual average serum testosterone concentrations (pre-, mid- and post-training) correlated significantly with individual changes in V_max (r=0.510, p=0.01) ([Fig. 6]) and in 3000 m (r=0.570, p<0.01) in all subjects. A significant correlation was also found between average serum testosterone/cortisol ratio and changes in V_max (r=0.457, p=0.025) and 3000 m (r=0.510, p=0.011) in the total group of subjects. Individual changes in testosterone concentrations from mid- to post-training correlated positively with changes in V_max (r=0.527, p=0.008) in all subjects. Individual changes in CMJ from mid- to post-training correlated positively with changes in V_max (R=0.469, p=0.021) in the total group of subjects.

Discussion

Both groups significantly improved their 3000 m running result and endurance performance in the incremental treadmill test. However, the main finding of the current study was the significantly greater increase in V_max and countermovement jump after HRV-guided compared to predetermined training after 8-weeks of high-intensity block training. Significant increases in HRV and serum testosterone concentration were observed in HRVG, but not in PD. This study suggests that individually HRV-guided programming of HIT blocks contributes to greater positive adaptations compared to predetermined training.

Training

Training intensity distribution based on time in the zone was almost identical between the groups. When analyzing weekly distribution using the session goal approach, significant differences were observed. In HRVG, blocks tended to be performed in a more even way through the training period. In previous studies on HRV-guided training, significant differences between predetermined and HRV-guided groups were observed in the amount of HIT training [19] [45]. Although no significant differences were observed in the present study, much larger interindividual variation in the number of HIT sessions was found in HRVG (SD=4.1) compared to PD (SD=0.6). It seems that some individuals were able to recover and benefit from a greater number of HIT sessions than others. During short training periods of 1–2 weeks, HIT sessions have been performed almost daily but not in longer interventions [2] [4]. No significant correlation was found between the individual number of HIT sessions and endurance performance adaptations in HRVG, suggesting that similar adaptations can be gained with differently periodised HIT training.

Endurance performance

Both groups improved endurance performance in the incremental treadmill test and in 3000 m running. Also VO_2max and velocity at the thresholds increased significantly in both groups. The magnitude of improvement was in line with previous studies on block periodization [32] and HRV guided training [19] [45]. All subjects improved their V_max on the treadmill as well as 3000 m running time. Nonetheless, a significant difference and large between-group effect size was found in the relative change of V_max. In addition, a moderate between-group effect size was noted in the change of absolute VO_2max. These findings were interesting due to the almost identical improvement of 3000 m in both groups. They were also somewhat different compared to the study by Vesterinen et al. [45] where the HRV-guided group performed better only in the 3000 m run but not on the treadmill. This difference might be explained by the different kinds of periodization of the training in the predetermined group, because in the current study it was more similar to the HRV-guided group. Neuromuscular performance and fatigue may at least partly explain the observed group difference in V_max, because the significant between-group difference was also found in the CMJ change. No correlation was found between individual pre-to-post changes in CMJ, but individual mid-to-post changes in CMJ correlated significantly with individual relative changes of V_max. The negative trend in CMJ may indicate neuromuscular fatigue caused by too many or badly timed HIT blocks. V_max speed was on average 10% higher than average speed during the 3000 m test, indicating that neuromuscular demand at V_max may be higher. As stated by Paavolainen et al. [28], a so-called muscle power factor may be an important determinant of maximal running performance.

Neuromuscular performance

To the best of our knowledge, no previous study has investigated the effects of block periodization of HIT sessions on neuromuscular performance. In the current study, the significant between-group difference with large effect size was found in the CMJ change. While CMJ tended to increase in HRVG, it tended to decrease in PD. This finding is in line with the Vesterinen study [46], in which only the HRV-guided group improved their reactivity jump during the training period. A too high amount of endurance training may disturb neuromuscular adaptations and performance, especially rapid force development [17] [23], which was also observed as decreased maximal running velocity during the first four weeks of the present training intervention. In previous running interventions, both aerobic and supramaximal intervals improved maximal running velocity and endurance performance [3]. It seems that block periodization of HIT may be more challenging from the perspective of rapid force development. Gomez et al. [7] found decreased force production in isokinetic knee flexion and CMJ after the 10-kilometre race even after 48 h. Demands of the interval sessions used in the present study may not be as high, but a similar type of trend might be possible. During HIT blocks of the current study such a long recovery time was not allowed, and it can be speculated that heavy HIT microcycles with insufficient recovery may lead to cumulated fatigue and a decrease in neuromuscular performance. HRV guidance seems to allow more optimal recovery between blocks. Despite between-group differences observed in CMJ, both groups significantly increased the 1RM leg press. This improvement can be partly explained via learning effects, but it seems that even a very low number of strength sessions, such as used in this study, combined with HIT training may be enough to gain some measurable adaptations in maximum force, at least in non-strength-trained subjects.

According to neuromuscular performance, running biomechanics also need to be considered. Whereas cycling, for example, includes mostly concentric muscle action, running involves stretch-shortening cycles with an eccentric component [20]. Microdamage in muscle may decrease endurance performance and thus lead to a decreased training stimulus during intensive periods [1] [22]. Earlier block periodization studies did not include runners as subjects [2] [4] [32] [34]. This aspect is important to consider when planning block training for runners.

Heart rate variability

Nocturnal HRV increased in the HRVG group, whereas only the positive trend was observed in PD. Earlier investigations have found no significant change [43], an increase [27] and an acute decrease [31] in HRV following intensive training. In the current study, both the absolute values and the increase of parasympathetic HRV markers (nocturnal HF and TP, morning RMSSD) were associated with positive adaptations to endurance training, which strengthens the use of HRV in monitoring endurance training load and adaptations. In addition to HRV, the significant decrease in nocturnal heart rate was observed, similar to previous studies [36] [44]. In the current study, the magnitude of decrease was quite great considering the short amount of time and subjects with already low baseline values. This finding may suggest that specific cardiac adaptations, such as increased stroke volume, were caused by block periodization.

Although increased HRV was associated with improved performance in the current study, this was not the case in all other studies. For example, in the study by Le Meur et al. [21], significant parasympathetic hyperactivation was found followed by overreaching. Overreaching was achieved with an increase in training volume, so it can be speculated that reactions may differ between training interventions with increased training volume or intensity. Schmitt et al. [37] also underlined the individuality of HRV reactions followed by intensive training. They found four different kinds of fatigue shifts of HRV patterns. More research is still needed to examine different types of individual HRV reactions and how they are possibly related to the type of training performed.

An interesting but not a novel finding was the association between baseline HF and V_max change in PD [11] [44]. In the Vesterinen et al. [44] study, a similar correlation was found between the individual baseline HF and V_max adaptations to HIT training. The correlation was negative with LIT training. Based on the association found in PD but not in HRVG, the timing and amount of HIT training may be more relevant than training intensity or volume. In the present study, a significant correlation was found between the individual morning RMSSD during the last four weeks of the training period and the number of HIT sessions in HRVG. This may also be related to the link between absolute HRV and ability to cope with high amounts of high-intensity training. Based on the correlations found between the individual HRV mid-to-post changes and individual changes in performance, these associations may become more critical as the length of the intensive training period increases.

As in previous studies, morning and nocturnal measurements of HRV showed slightly different trends. The morning measurements were used for periodization in the current study for practical reasons. Although nocturnal measurements are often stated to be a more standardized method, in the Hynynen et al. [15] study no changes in nocturnal HRV markers were observed in the overtrained athletes, whereas significant decreases were found in the morning measurements. The authors speculated that waking always causes a kind of stress reaction that may lead to different results compared to the night measurement. In the present study, there were no significant differences between the groups according to nocturnal heart rate or HRV changes during the training period. In the morning measurements, a small between-group effect size in RMSSD and moderate in heart rate were observed when the relative change from the control period was compared to the week-8 values.

Due to different results obtained from different kinds of measurements, it is important to always use the same kind of protocol. In addition, using averages instead of individual values is highly recommended as stated by Plews et al. [30]. In the current study, the 3-day rolling average of the quick recovery test was used. Previous studies have used 7-day averages [45] and daily values [19]. Averaging results for a longer period may decrease the risk of false results due to high day-to-day variation in HRV [30], but at the same time averages of a very long period may make it difficult to react quickly to changes in the autonomic nervous system. An average of three to four days may be a good compromise, because it decreases the value of an individual result but still makes it possible to react quickly to observed trends. The reference value also plays an important role in regulating the start of HIT blocks. In the current study, the average value of the control period was used, which seemed to allow good recovery in most subjects. A few individuals had trouble obtaining a test result over the reference value, probably due to stress outside training. Despite a low number of HIT blocks, they still improved their performance. The quick recovery test scaled the result based on individual measurement history. The reference value was therefore continuously updated as more HRV data was collected. Updating the reference or control value during longer training periods may be recommended because in the current study significant increases in HRV markers were observed following training.

Serum hormone concentrations

It has been found that endurance athletes tend to have lower testosterone concentrations compared to controls [8]. However, adaptations observed after the endurance training period has varied from a decrease [13], an increase [6] to no change [43]. Training mode may affect on hormonal response, because a greater acute free testosterone response has been found after a high-intensity interval session compared to a lower intensity steady-state session [9]. In the current study, a significant increase in serum testosterone concentration and a tendency for the testosterone/cortisol ratio to increase were observed from mid- to post-training in the HRVG group. The significant decrease was noted in testosterone in PD from pre- to mid-training. No significant changes were observed in concentrations of other hormones examined.

Individual basal serum testosterone concentration as well as the testosterone/cortisol ratio correlated with changes in V_max and 3000 m. Hoogeveen and Zonderland [13] found no correlation between the improvement of cycling performance and changes in testosterone or cortisol during a training period. However, Mäestu et al. [26] concluded that the first sign of decreased adaptivity in athletes is a decreased resting level of free testosterone and a lower maximal exercise-induced acute increase in free testosterone concentration. Most studies have focused on typical high-volume endurance training, so it can be speculated that high-intensity training may induce different adaptations. Zinner et al. [48] found that after 2 weeks of HIT training, a positive correlation between the improvement in endurance performance and an increase in basal testosterone concentration was observed.

An interesting relationship was also found between both individual absolute testosterone concentrations and testosterone/cortisol ratios and individual changes in HRV during the present training period. Similarly, Huovinen et al. [14] found a significant correlation between the testosterone/cortisol ratios and the changes in HF during the stressful first week of military service. According to intensive block training, in addition to positive changes in testosterone and the testosterone/cortisol ratio, higher absolute serum testosterone concentrations may also be beneficial.

Conclusions

The present results suggest that block periodization of HIT is an effective way to improve endurance and running performance in a short amount of time in already endurance-trained males. Individually HRV-guided timing and the number of HIT blocks seem to provide greater endurance and neuromuscular adaptations compared to predetermined training. Individually guided training may reduce the risk of overtraining observed as positive changes in HRV and serum testosterone concentrations. Both baseline heart rate variability and testosterone levels may be associated with the capacity of an individual to adapt to intensive block training.

Conflict of Interest

The authors have no conflict of interest to declare.

Acknowledgements

Partial funding for this study has been provided by Firstbeat Technologies Ltd. The results of the current study do not constitute endorsement of the product by the authors or the journal.

• References

• 1 Black CD, Gonglach AR, Hight RE, Renfroe JB. Time-course of recovery of peak oxygen uptake after exercise-induced muscle damage. Respir Physiol Neurobiol 2015; 216: 70-77
• 2 Breil FA, Weber SN, Koller S, Hoppeler H, Vogt M. Block training periodization in alpine skiing: effects of 11-day HIT on VO2max and performance. Eur J Appl Physiol 2010; 109: 1077-1086
• 3 Cicioni-Kolsky D, Lorenzen C, Williams MD, Kemp JG. Endurance and sprint benefits of high-intensity and supramaximal interval training. Eur J Sport Sci 2013; 13: 304-311
• 4 Clark B, Costa VP, O'Brien BJ, Guglielmo LG, Paton CD. Effects of a seven day overload-period of high-intensity training on performance and physiology of competitive cyclists. PLoS One 2014; 9: e115308
• 5 Gathercole R, Sporer B, Stellingwerff T, Sleivert G. Alternative countermovement-jump analysis to quantify acute neuromuscular fatigue. Int J Sports Physiol Perform 2015; 10: 84-92
• 6 Grandys M, Majerczak J, Duda K, Zapart-bukowska J, Kulpa J, Zoladz JA. Endurance training of moderate intensity increases testosterone concentration in young, healthy men. Int J Sports Med 2009; 30: 489-495
• 7 Gómez AL, Radzwich RJ, Denegar CR, Volek JS, Rubin MR, Bush JA, Doan BK, Wickham RB, Mazzetti SA, Newton RU, French DN, Häkkinen K, Ratamess NA, Kraemer WJ. The effects of a 10-kilometer run on muscle strength and power. J Strength Cond Res 2002; 16: 184-191
• 8 Hackney AC, Szczepanowska E, Viru AM. Basal testicular testosterone production in endurance-trained men is suppressed. Eur J Appl Physiol 2003; 89: 198-201
• 9 Hackney AC, Hosick KP, Myer A, Rubin DA, Battaglini CL. Testosterone responses to intensive interval versus steady-state endurance exercise. J Endocrinol Invest 2012; 35: 947-950
• 10 Harriss DJ, Atkinson G. Ethical standards in sport and exercise science research: 2016 update. Int J Sports Med 2015; 36: 1121-1124
• 11 Hautala AJ, Mäkikallio TH, Kiviniemi A, Laukkanen RT, Nissilä S, Huikuri HV, Tulppo MP. Cardiovascular autonomic function correlates with the response to aerobic training in healthy sedentary subjects. Am J Physiol 2003; 285: 1747-752
• 12 Helgerud J, Høydal K, Wang E, Karlsen T, Berg P, Bjerkaas M, Simonsen T, Helgesen C, Hjorth N, Bach R, Hoff J. Aerobic high-intensity intervals improve VO2max more than moderate training. Med Sci Sports Exerc 2007; 39: 665-671
• 13 Hoogeveen AR, Zonderland ML. Relationships between testosterone, cortisol and performance in professional cyclists. Int J Sports Med 1996; 17: 423-428
• 14 Huovinen J, Tulppo M, Nissilä J, Linnamo V, Häkkinen K, Kyrolainen H. Relationship between heart rate variability and the serum testosterone-to-cortisol ratio during military service. Eur J Sport Sci 2009; 9: 277-284
• 15 Hynynen E, Uusitalo A, Konttinen N, Rusko H. Heart rate variability during night sleep and after awakening in overtrained athletes. Med Sci Sports Exerc 2006; 38: 313-317
• 16 Hynynen E, Vesterinen V, Rusko H, Nummela A. Effects of moderate and heavy endurance exercise on nocturnal HRV. Int J Sports Med 2010; 31: 428-432
• 17 Häkkinen K, Alen M, Kraemer WJ, Gorostiaga E, Izquierdo M, Rusko H, Mikkola J, Häkkinen A, Valkeinen H, Kaarakainen E, Romu S, Erola V, Ahtiainen J, Paavolainen L. Neuromuscular adaptations during concurrent strength and endurance training versus strength training. Eur J Appl Physiol 2003; 89: 42-52
• 18 Issurin V. Block periodization versus traditional training theory: A review. J Sports Med Phys Fitness 2008; 48: 65-75
• 19 Kiviniemi AM, Hautala AJ, Kinnunen H, Tulppo MP. Endurance training guided individually by daily heart rate variability measurements. Eur J Appl Physiol 2007; 101: 743-51
• 20 Komi PV. Stretch-shortening cycle: A powerful model to study normal and fatigued muscle. J Biomech 2000; 33: 1197-1206
• 21 Le meur Y, Pichon A, Schaal K, Schmitt L, Louis J, Gueneron J, Vidal P, Hausswirth C. Evidence of parasympathetic hyperactivity in functionally overreached athletes. Med Sci Sports Exerc 2013; 45: 2061-2071
• 22 Marcora SM, Bosio A. Effect of exercise-induced muscle damage on endurance running performance in humans. Scand J Med Sci Sports 2007; 17: 662-71
• 23 Mikkola J, Rusko H, Izquierdo M, Gorostiaga EM, Häkkinen K. Neuromuscular and cardiovascular adaptations during concurrent strength and endurance training in untrained men. Int J Sports Med 2012; 33: 702-710
• 24 Muñoz I, Seiler S, Bautista J, España J, Larumbe E, Esteve-lanao J. Does polarized training improve performance in recreational runners?. Int J Sports Physiol Perform 2014; 9: 265-72
• 25 Myllymäki T, Rusko H, Syväoja H, Juuti T, Kinnunen ML, Kyröläinen H. Effects of exercise intensity and duration on nocturnal heart rate variability and sleep quality. Eur J Appl Physiol 2012; 112: 801-809
• 26 Mäestu J, Jürimäe J, Jürimäe T. Monitoring of performance and training in rowing. Sports Med 2005; 35: 597-617
• 27 Nummela A, Hynynen E, Kaikkonen P, Rusko H. High-intensity endurance training increases nocturnal heart rate variability in sedentary participants. Biol Sport 2016; 33: 7-13
• 28 Paavolainen L, Häkkinen K, Hämäläinen I, Nummela A, Rusko H. Explosive-strength training improves 5-km running time by improving running economy and muscle power. J Appl Physiol 1999; 86: 1527-1533
• 29 Pichot V, Busso T, Roche F, Garet M, Costes F, Duverney D, Lacour JR, Barthelemy JC. Autonomic adaptations to intensive and overload training periods: a laboratory study. Med Sci Sports Exerc 2002; 34: 1660-1666
• 30 Plews DJ, Laursen PB, Le meur Y, Hausswirth C, Kilding AE, Buchheit M. Monitoring training with heart-rate variability: How much compliance is needed for valid assessment?. Int J Sports Physiol Perform 2014; 9: 783-790
• 31 Plews DJ, Laursen PB, Buchheit M. Day-to-day heart rate variability (HRV) recordings in world champion rowers: Appreciating unique athlete characteristics. Int J Sports Physiol Perform 2016; 1-19 (epub ahead of print)
• 32 Rønnestad BR, Ellefsen S, Nygaard H, Zacharoff EE, Vikmoen O, Hansen J, Hallén J. Effects of 12 weeks of block periodization on performance and performance indices in well-trained cyclists. Scand J Med Sci Sports 2014; 24: 327-335
• 33 Rønnestad BR, Hansen J, Vegge G, Tønnessen E, Slettaløkken G. Short intervals induce superior training adaptations compared with long intervals in cyclists – an effort-matched approach. Scand J Med Sci Sports 2015; 25: 143-151
• 34 Rønnestad BR, Hansen J, Thyli V, Bakken TA, Sandbakk Ø. 5-week block periodization increases aerobic power in elite cross-country skiers. Scand J Med Sci Sports 2016; 26: 140-146
• 35 Saalasti S, Seppänen M, Kuusela A. Artefact correction for heart beat interval data. Advanced methods for processing bioelectrical signals. In: Proceedings of the ProBisi Meeting. Jyväskylä, Finland 2004; 1–10
• 36 Scharhag-Rosenberger F, Meyer T, Walitzek S, Kindermann W. Time course of changes in endurance capacity: a 1-yr training study. Med Sci Sports Exerc 2009; 41: 1130-1137
• 37 Schmitt L, Regnard J, Parmentier AL, Mauny F, Mourot L, Coulmy N, Millet GP. Typology of "fatigue" by heart rate variability analysis in elite Nordic-skiers. Int J Sports Med 2015; 36: 999-1007
• 38 Seiler S, Haugen O, Kuffel E. Autonomic recovery after exercise in trained athletes: intensity and duration effects. Med Sci Sports Exerc 2007; 39: 1366-1373
• 39 Seiler S. What is best practice for training intensity and duration distribution in endurance athletes?. Int J Sports Physiol Perform 2010; 5: 276-291
• 40 Stöggl T, Sperlich B. Polarized training has greater impact on key endurance variables than threshold, high intensity, or high volume training. Front Physiol 2014; 5: 33
• 41 Tønnessen E, Sylta Ø, Haugen TA, Hem E, Svendsen IS, Seiler S. The road to gold: training and peaking characteristics in the year prior to a gold medal endurance performance. PLoS ONE 2014; 9: e101796
• 42 Tønnessen E, Svendsen IS, Rønnestad BR, Hisdal J, Haugen TA, Seiler S. The annual training periodization of 8 world champions in orienteering. Int J Sports Physiol Perform 2015; 10: 29-38
• 43 Vesterinen V, Häkkinen K, Hynynen E, Mikkola J, Hokka L, Nummela A. Heart rate variability in prediction of individual adaptation to endurance training in recreational endurance runners. Scand J Med Sci Sports 2013; 23: 171-180
• 44 Vesterinen V, Häkkinen K, Laine T, Hynynen E, Mikkola J, Nummela A. Predictors of individual adaptation to high-volume or high-intensity endurance training in recreational endurance runners. Scand J Med Sci Sports 2016; 26: 885-893
• 45 Vesterinen V, Nummela A, Heikura I, Laine T, Hynynen E, Botella J, Häkkinen K. Individual endurance training prescription with heart rate variability. Med Sci Sports Exerc 2016; 48: 1347-1354
• 46 Vesterinen V. Predicting and monitoring individual endurance training adaptation and individualizing training prescription with endurance performance, cardiac autonomic regulation and neuromuscular performance [Dissertation]. Jyväskylä, Finland: University of Jyväskylä; 2016
• 47 Vesterinen V, Nummela A, Laine T, Hynynen E, Mikkola J, Häkkinen K. A submaximal running test with postexercise cardiac autonomic and neuromuscular runction in monitoring endurance training adaptation. J Strength Cond Res 2017; 31: 233-243
• 48 Zinner C, Wahl P, Achtzehn S, Reed JL, Mester J. Acute hormonal responses before and after 2 weeks of HIT in well trained junior triathletes. Int J Sports Med 2014; 35: 316-322