Effects of a backward running training on backward peak velocity running, V̇O2max, vVO2max and 3 km forward running performance in male adults: a pilot study

Aims: The study aimed to evaluate the effects of a backward running (BR) training program prescribed by the peak backward running velocity (Vpeak_BR) on physiological variables and a 3 km forward running (FR) performance. Methods: Eight untrained running male adults in running took place in the study. All the participants underwent five weeks of BR training prescribed based on Vpeak_BR. They performed a maximal incremental test on the treadmill to determine the maximal oxygen uptake (V̇O2max) and the velocity associated with V̇O2max (vV̇O2max). The participants were also tested on the track field to determine the Vpeak_BR and undertook a 3 km FR performance. All initial assessments were also performed after the training period. Results: The results showed statistically significant improvements in 3 km FR performance (14.2 ± 1.2 min vs. 13.5 ± 1.0 min) and Vpeak_BR (8.0 ± 0.8 km·h vs. 8.5 ± 0.5 km·h) after the training period. Conclusion: BR training effectively improved 3 km FR performance and Vpeak_BR, demonstrating that Vpeak_BR determined according to the protocol proposed in this study can be used for the prescription of BR training. Further, BR training represents an effective training method that can be inserted into an FR running training program.


Introduction
Backward running (BR) has been presented in the literature as a form of locomotion retrograde to forward running (FR), which presents a unique energetic, biomechanical profile and cardiopulmonary response 1,4 . It is characterized as a low-impact exercise that promotes improvements in neuromuscular performance, reducing the risk of joint damage 1,5 . In this context, BR has been used as a method of maintaining cardiovascular fitness as well as a means of rehabilitating athletes and physical exercise practitioners suffering from joint injuries 4,6,7 .
Moreover, BR has been used in sports training programs to increase muscle pre-activation. It is also deployed as a method of promoting improvements in physiological and performance variables [e.g., maximal oxygen uptake (V̇O 2max ), anaerobic power, lower limb strength parameters, and agility] 1,2,7 . However, BR has not yet been investigated as a training method for improving FR performance.
Studies on BR training prescriptions have used intensities based on the ability of participants to perform sprints or have been based on self-selected intensities 1,2 . For example, Terblanche et al. 1 found a significant reduction in body fat percentage (2.4%) and demonstrated a decrease in the sum of skinfolds. The authors also found a considerable improvement in the predicted V̇O 2max values in the forward 20 m shuttle-run test (5.2%) performed after BR training. Similarly, a previous study reported significant improvement in the 10 and 20 m sprint performances [effect size (ES) = 20.47 and 20.26, respectively] and countermovement jump (CMJ) height (ES = 0.51) of the BR group compared to the FR group. The authors indicated that young male athletes could auto-regulate the BR and FR strategy to achieve desired running intensities 40 to 55% (slow), 60 to 75% (moderate), and > 90% (fast) of the maximal 2 .
Despite these outcomes, the studies mentioned above-applied training protocols based on intensities that were self-selected by the runners. Such proposed inten-sities may be underestimated or overestimated and can cause inaccurate results of this intensity.
However, it is important to use variables that individualize the BR prescription, as the peak velocity (V peak ) that is considered a good predictor of endurance FR performance 8,11 .
The outcomes of the current investigation can help improve our understanding of BR training and provide evidence that supports BR as a useful method to improve athlete performance.
Therefore, this study aimed to evaluate the effects of a BR training program prescribed by the peak backward running velocity (V peak_BR ) on physiological variables and a 3 km FR performance.

Participants
Eight male adults participated in this study (24.2 ± 4.5 years, 69.3 ± 3.4 kg, 171.0 ± 0.02 cm, 13.1 ± 4.6 % fat). As an inclusion criterion, the participants did not have any cardiovascular or respiratory dysfunction, as well as they did not have muscle or joint injuries in the lower limbs. In addition, they could not be included in any running training program. The exclusion criteria were being injured during the study, not performing the pre-and or post-training tests, and not meeting 90% of the proposed training protocol. The participants in this study provided written informed consent. The procedures performed in this research followed the regulations required by the Declaration of Helsinki and was approved by the Human Research Ethics Committee (# 1.262.502/2015)

Procedures
Participants performed before and after 5-weeks of BR training a FR performance on a 400 m track field and two incremental tests: for V peak_BR determination and other to obtain V̇O 2max and velocity associated with V̇O 2max (vV̇O 2max ). Evaluations were carried out with an interval of 48 h between them. In all tests heart rate (HR), lactate concentrations [Lac], and the rating of perceived exertion (RPE) were monitored. All tests were performed at the same time of the day (5:00-9:00 pm hours). During the period of the study, in the tests and training on the track, the ambient temperature was around 18-29°C, and the relative humidity between 56%-72%, among the days. Under standard laboratory conditions, the temperature has been around 20-22°C and relative humidity between 50-60%.

V̇O 2max and vV̇O 2max determination
The V̇O 2max and vV̇O 2max incremental exercise tests were performed on a motorized treadmill (Inbrasport Super ATL ® , Porto Alegre, Brazil) with a gradient set at 1% 12 . After a warm-up that consisted of walking at 6 km·h −1 for 3 min, the protocol started with an initial speed of 8 km·h −1 , followed by an increase of 1 km·h −1 every 3 min between each successive stage until volitional exhaustion (i.e., the participant was unable to continue running) 9 . Gas exchange was collected to determine the V̇O 2max using a portable gas analyzer K4b2 Cosmed ® , Rome, Italy), and the V̇O 2max was considered as the maximal value obtained during the test, measured at an average of 15 s intervals. The vV̇O 2max was the minimum velocity at which the participants were running when V̇O 2max occurred 13 .

V peak_BR determination
The V peak_BR test consisted of BR in a 20 m course at progressively increasing speeds controlled by audio. The protocol adapted from Machado et al. 9 , consisted of 3 min warm-up backwards walking at 4 km/h, followed by BR at 5 km/h and an increase of 1 km/h every 3 min until volitional exhaustion or until the participant failed twice in a row to overtake with one foot in the cone line. The V peak_BR was the maximal running speed reached during the incremental test and if the last stage was not completed the V peak_BR was calculated from the equation proposed by Kuipers et al 14 : where V complete is the running velocity of the last complete stage, Inc the speed increment (i.e., 1 km/h), t the number of seconds sustained during the incomplete stage and T is the number of seconds required to complete a stage (i.e., 180 s).

km FR performance
A track field test was performed to determine the time to complete the 3 km FR performance. The participants had a self-determined 10 min warm-up.
[Lac], HR, and RPE determination During the tests, the peak blood lactate concentration ([Lac peak ]), maximal heart rate (HR max ), and maximal rating of perceived exertion (RPE max ), variables were monitored following the protocols below 8,10 .
Earlobe capillary blood samples (25 μL) were collected into a capillary tube at the start and end of the tests (time zero of recovery) and at the third and fifth minutes of passive recovery, while participants were sitting in a comfortable chair. From these samples, [Lac] was subsequently determined by electroenzymatic methods using an automated analyzer (YSI 2300 STAT ® , Yellow Springs, Ohio, USA). [Lac peak ] was defined for each participant as the highest post-exercise [Lac] value. HR was monitored during all tests, in the 3 km test the HR was monitored every 400 m, while in the V̇O 2max and V peak_BR tests the HR was monitored at the end of each stage (Polar ® RS800sd; Kempele, Finland) and HR max was defined as the highest HR value recorded during the test. RPE was also monitored during all tests by using a 6-20 Borg scale 15 , and the highest RPE value was adopted as the RPE max .

BR training program
The 5 weeks BR training program was performed twice a week with two types of BR sessions: continuous moderate-intensity and high-intensity interval training. All training sessions were held on a 400 m outdoor track field, between 5:00 and 9:00 pm, under the supervision and guidance of the researchers. Continuous BR training was performed out around the official track field and interval training for BR was performed in a straight line, with a distance of 50 m demarcated by two cones. The total duration of the training session and the intensities are described in Table 1. Sessions were preceded by 10 min free warm-up, and after the main part of the session, it was a free cool-down of self-selected low-intensity running and stretching. A stopwatch and track measurements were used to control the running velocity of each participant.

Statistical analyses
Data are presented as mean ± standard deviation (SD) and were analyzed using the Statistical Package for the Social Sciences Software (SPSS ® version 22.0). The normality was verified using Shapiro-Wilk test and lead to parametric statistics. The student's t-test for dependent samples was used to compare variables between pre-and post-training. The association between 3 km FR performance and V peak_BR was performed by Pearson correlation (r). Percentage changes and effect size (ES) 16 were calculated to assess the magnitude of changes between pre-and post-training. The ES were classified as: < 0.2 (trivial), 0.2-0.6 (small), 0.6-1.2 (moderate), ≥ 1.2 (large) 17 . A significance level of P < 0.05 was adopted. Table 2 demonstrated that there were significant differences between pre-and post-training for V peak_BR and test duration (P = 0.001). The ES values were moderate, except for the HR max that showed small ES. Table 3 showed that there were no significant differences between pre-and post-training for all variables obtained during the V̇O 2max test. ES values were small and trivial for all the comparisons, except for the [Lac peak ] that showed a moderate ES. Figure 1 demonstrate that the time to complete the 3 km FR performance was significant different between pre-and post-training [14.2 ± 1.2 vs. 13.5 ± 1.0 min, respectively; P = 0.005; ES = −0.64 (moderate); %change = −4.6 ± 3.9]. Significant difference was also observed for the [Lac peak ] between pre-and post-training (8.1 ± 1.8 vs. 11.3 ± 1.8 mmol·L −1 , P = 0.046). HR max and RPE max were not different between moments (P = 0.38; P = 0.12, respectively).

Results
The correlation between V peak_BR and 3 km FR performance at post-training was high and negative (r = −0.77; P = 0.03).

Discussion
This study aimed to evaluate the effects of a BR training program prescribed by the V peak_BR on physiological variables and a 3 km FR performance. The main findings were that there was a significant improvement in the V peak_BR and the 3 km FR performance after the train- Table 1 -Continuous and interval training of BR.

Weeks 2, 3, 4 and 5 -Training Protocol
Continuous training 20 ± 2.5 min (5 min increase each week) at 80 ± 4% of V peak_BR Interval training X # sprints 50 m at 165 ± 2% of V peak_BR with effort ratio and break (1:1) # The number of series performed by each participant was adjusted so that the total duration of interval training session corresponded to 15 ± 2.5 min. ing period. A high correlation was found between the V peak_BR and the 3 km FR performance. The significant improvement found in the V peak_BR after 5 weeks of BR training is like other studies that used the V peak for the prescription of FR 8,18 . It is important to mention that V peak is a performance variable in FR and that it is sensitive to changes stemming from training; further, it may be used to evaluate, prescribe, and monitor the training of runners.
These results are important because an appropriate training prescription requires the use of variables that can control and monitor effort intensities and possible physiological adaptations resulting from the practice 8,9,11 .
In contrast to the V peak_BR , the pre-and post-training V̇O 2max was not significantly different. This outcome is also aligned with studies that observed the positive effects of FR endurance training on other physiological and performance variables, but not in the V̇O 2max 8,18 . It is suggested that V̇O 2max is not a sensitive variable capable of detecting the effects of training; thus, the use of other variables such as V peak may be necessary 18 .
The improvement observed in the 3 km FR performance in the present study conformed to the reports of previous studies that verified the effects of FR training prescribed for endurance running performance according to the V peak 8,18 . Finally, the significantly high and negative correlation between the V peak_BR and the 3 km FR performance at the post-training testing resulted from the efficient application of the choice of the V peak_BR for the prescription of BR training. This correlation is similar to the findings of studies that verified the correlation between the V peak and the FR endurance performance at different distances 8,9 .
Despite the important findings, this study must acknowledge certain limitations, such as the low number of participants and the absence of a control group comprising solely of FR training. It is suggested that future studies should investigate the insertion of BR based on the V peak_BR into an FR training program to verify whether BR training can enhance FR performance.

Conclusions
The BR training program undertaken for this study effectively enhanced the V peak_BR and 3 km FR performance of the participants. Therefore, coaches and runners aiming to optimize athletic performance should consider the following advantages of implementing BR training for FR athletes. First, BR training is recommended as an effective training method that could be included in FR training programs to improve the 3 km FR running performance of runners. Second, the V peak_BR determined according to the 3 km protocol used in this study may effectively be used for the prescription of BR training. Finally, BR training can be implemented in the FR training program, to provide different stimuli to avoid training monotony and improve performance.