Peak velocity and its time limit are as good as the velocity associated with VO 2max for training prescription in runners.

This study compared the effects of 4 weeks of training prescribed by peak velocity (V peak ) or velocity associated with maximum oxygen uptake (vVO 2max ) in moderately trained endurance runners. Study participants were 14 runners (18-35 years) randomized into 2 groups, named group VO 2 (GVO 2 ) and group V peak (GVP). The GVO 2 had training prescribed by vVO 2max and its time limit (t lim ), whereas the GVP had training prescribed by V peak and its t lim . Four tests were performed on a treadmill: 2 maximum incremental for V peak and vVO 2max and 2 for their t lim . Performance (10 km) was evaluated on a 400 m track. Evaluations were repeated after 4 weeks of endurance training. The results showed a significant effect of training on V peak [GVP (16.7±1.2 - 17.6±1.5 km . h -1 ), GVO 2 (17.1±1.9-17.7±1.6 km·h -1 )]; vVO 2max [GVP (16.4±1.4-17.0±1.3 km·h -1 ), GVO 2 (17.2±1.7-17.5±1.9 km·h -1 )]; and 10 km performance [GVP (41.3±2.4-39.9±2.7 min), GVO 2 (40.1±3.4-39.2±2.9 min)]. The V peak highly correlated with performance in both pre- and post-training in GVP (-0.97;-0.86) and GVO 2 (-0.95;-0.94), as well as with vVO 2max in GVP (-0.82;-0.88) and GVO 2 (-0.99; -0.98). It is concluded that training prescribed by V peak promoted similar improvements compared to training prescribed by vVO 2max . The use of V peak is recommended due to its practical application and the low cost of determination.

RCT Entities:   Population Interventions Outcomes

Introduction

Success in endurance racing depends on an elaborate training prescription utilizing appropriate loads and recovery periods. Such prescriptions should be planned according to the needs of the individual athlete for achieving the highest level of adaptation possible prior to the competition 16 22 29 . For proper training prescription, it is necessary to use variables that control and monitor the intensity of effort and possible physiological adaptations resulting from this practice and, most importantly, that show a correlation with performance 7 .

Currently, the velocity associated with the occurrence of maximum oxygen uptake (vVO 2max ) is considered a good variable to predict performance and to monitor and prescribe endurance running training 2 9 27 . In addition, the application of its time limit (t lim ) may improve the prescription of the most adequate set duration for high-intensity interval workouts 2 . Previous studies show that training prescribed by vVO 2max and its respective t lim promoted improvements in performance of 3, 5 and 10 km 9 13 34 . In addition, the training prescribed by these variables can promote improvements in VO 2max , running speed at the lactate threshold, and parameters related to heart rate (HR) among others 9 13 34 . However, the vVO 2max determination requires the use and handling of expensive and delicate equipment, as well as the interpretation of data, limiting its use to only a few research laboratories, coaches, and athletes. Moreover, the vVO 2max refers to estimating the minimum speed required to achieve VO 2max , as a result of calculating vVO 2max based on VO 2max determination, whereas the peak velocity (V peak ) is the maximum speed directly measured and associated with VO 2max 26 .

Thus, V peak is an attractive variable that has been gaining attention among researchers, trainers, and endurance runners due to its practicality and financial accessibility. Despite the fact that V peak is associated with vVO 2max and is a great predictor of endurance performance in tests 3–90 km 25 26 30 , it is necessary to test its applicability to endurance training prescription as well as the applicability of its t lim to determine duration of high-intensity interval sets. Although the intra-individual differences between V peak and vVO 2max might be very small, the t lim differences may be large, which would meaningfully change the duration of high-intensity interval sets.

Given that, as far as it is known, V peak based training prescription for moderate intensity continuous training and high-intensity interval training has not been tested yet, the aim of this study was to evaluate the effect of 4 weeks of training prescribed by V peak , vVO 2max, and their respective t lim in moderately trained endurance runners. Our hypothesis was that both training models would improve aerobic, anaerobic, and performance parameters of moderately trained runners in a similar manner. We also hypothesized that V peak would demonstrate a higher correlation with the 10 km performance than the vVO 2max before and after training, given that V peak is the ‘measured’ speed associated with VO 2max , and vVO 2max is the ‘estimated’ speed associated with VO 2max 11 26 . Should this be shown, it would demonstrate that the V peak was a more sensitive variable to the effects of training for moderately trained runners.

Methods

Participants

Fourteen moderately trained endurance runners were recruited for participation in this study and showed average speed (AS) between 14 and 16 km·h −1 (≅ 62–71% of the world record). They performed at least 5 training sessions per week. Their average training distance during the study was 40.9±4.5 km ∙ week −1 , which was similar to their training distance before the study. Subjects had the following characteristics: (mean±SD, age 29.2±5.3 years, weight 71.9±11.0 kg, height 175.1±4.3 cm) with a minimum of 1 year of experience in competitive long distance races. Before the study, the subjects were informed about the testing and training and possible risks involved and provided written informed consent. This study was approved by the University’s Human Research Ethics Committee (#1.022.468). All research was conducted ethically according to international standards and as required by the International Journal of Sports Medicine 15 .

Experimental design

Runners were randomized into 2 groups using random numbers. One group was trained by V peak (GVP; n=8) and the other group by vVO 2max (GVO 2 ; n=6). The experiment involved the implementation of 2 different endurance running training programs (GVP vs. GVO 2 ) using the prescribed external workload (%V peak or %vVO 2max ) for 5 sessions per week over a 4 week period, for a total of 20 sessions. Before and after the training intervention, in a counterbalanced order, the subjects were evaluated using 2 incremental tests on a treadmill to measure VO 2max and V peak and 2 to determine their t lim . Performance (10 km) was evaluated on an official running track (400 m). In addition, variables such as heart rate (HR), blood lactate concentration [LA], and rating of perceived exertion (RPE) were also evaluated during the tests. The tests were performed over 2 weeks, with a period of at least 48 h separating each of them.

Determination of V peak and its t lim

The V peak was assessed on a motorized treadmill (Super ATL; Inbrasport, Porto Alegre, Brazil) (with the gradient set at 1% 21 . After a 3 min warm-up walking at 8 km·h −1 , the protocol started with an initial velocity of 10 km·h −1 , followed by an increase of 1 km·h −1 every 3 min until volitional exhaustion (i. e., participant was unable to continue running). If the last stage was not completed, the V peak was calculated on the partial time remaining in the last stage using the equation proposed by Kuipers et al. 23 : V+(t/180×1.0), where V was the last completed velocity (km·h −1 ) and t, the time (s) of the uncompleted step (180 s). The t lim at V peak was assessed after a 15 min warm-up at 60% V peak , when velocity was increased to V peak . The subjects were verbally encouraged to run to volitional exhaustion 4 .

Determination of VO 2max and its t lim

The protocol used for determining the VO 2max was the same as that used for the determination of V peak ; additionally, exhaled gas was collected to determine the VO 2max using a portable gas analyzer (k4b 2, Cosmed, Roma, Italy). The VO 2max was regarded as the maximum value obtained during the test, measured at an average of 15-s intervals, and when at last 2 of the following criteria were met: (1) LA peak ≥8 mmol·L −1 , (2) HR max ≥100% of endurance-trained age-predicted HR max using the age-based “206–0.7×age” equation 37 and (3) RPE max ≥18 in the 6–20 Borg scale 6 . The vVO 2max was the minimal velocity at which the athlete was running when VO 2max occurred 2 4 . To determine t lim at vVO 2max, the same protocol was applied as that used for determining the t lim at V peak using the values of vVO 2max as parameters.

Time trials of 10 km

Participants undertook 10 km time trials on a 400 m outdoor running track at 6:00 pm. The trial was preceded by a self-selected pace warm-up of 10 min duration. A hydration station was set up on the track with natural water. The participants were encouraged to achieve their best performance. Split times were registered at each 400 m and the average velocity of each section was calculated.

Determination [LA], HR, and RPE

Earlobe capillary blood samples (25 μl) were collected into a capillary tube at the end of the tests (time zero of recovery) and at the third, fifth, and seventh minutes of passive recovery with participants seated in a comfortable chair. From these samples, [LA] was subsequently determined by electroenzymatic methods using an automated analyzer (YSI 2300 STAT, Yellow Springs, Ohio, USA). Peak [LA] (LA peak ) was defined for each participant as the highest post-exercise [LA] value. RPE was also monitored during all tests by using a 6–20 Borg scale 6 , and the highest RPE value was adopted as the peak RPE (RPE peak ). HR was monitored during all tests (Polar RS800sd; Kempele, Finland) and HR max was defined as the highest HR value recorded during the test.

Training programs

All training sessions were held on a 400 m outdoor running track, between 5:00 and 9:00 pm hours due to the availability of participants and the fact that their performance would be better in the evening 10 . The training protocol consisted of 2 types of running training: continuous moderate-intensity and high-intensity interval training (short interval and long interval). The running intensity was prescribed based on the V peak and t lim for the GVP group, and the vVO 2max and t lim for the GVO 2 group ( Table 1 ).

Table 1 Continuous and interval training prescribed for GVP and GVO 2 groups.

GVP and GVO 2
Continuous training	45*±2.5 min at 75±4% of V peak or vVO 2max . (weeks 1 & 2) 60±2.5 min at 75±4% of V peak or vVO 2max . (weeks 3 & 4)
Short interval training	X* # sets at 120±2% of V peak or vVO 2max with duration 10% their respective t lim and intervals (passive) with duration 30% of t lim at V peak or vVO 2max .
Long interval training	X* # sets at 100±2% of V peak or vVO 2max with duration 60% their respective t lim and intervals (passive) with duration 60% of t lim at V peak or vVO 2max .

# The number of series performed by each participant was adjusted so that the total duration of interval training session corresponded to 30±2.5 min

* The intensity and duration of training was the same for both groups with differences only in the prescription variable: the GVO 2 had the training prescribed by vVO 2max and its respective t lim and GVP had training prescribed by V peak and its respective t lim

Training was based on studies by Buchheit et al. 9 ; Esfarjani and Laursen 13 ; Smith; Coombes, and Geraghty, 34 ; Billat et al. 2

The GVO 2 and GVP training sessions were preceded by a 15 min warm-up consisting of 5 min of low intensity running at a self-selected velocity, 5 min of stretching, and 5 min of running at 60% of V peak or vVO 2max 35 . After the warm-up, the main training session (continuous or interval training) was conducted, followed by a cool-down comprised of self-selected low-intensity running and stretching.

The training participants of both groups were trained 5 times per week for 4 weeks. They performed 10 sessions of continuous training and 10 of interval training. During the odd weeks, participants performed 3 sessions of continuous training and 2 sessions of interval training; and the reverse during even weeks. The training sessions of the groups were differentiated by the prescription method (V peak and their respective t lim to GVP and vVO 2max and their respective t lim to GVO 2 ). The intensity and volume of training were maintained throughout the protocol, except for continuous training in weeks 3 and 4 when the duration was increased from 45 to 60 min for both groups.

Statistical analyses

All statistical analyses were performed using the SPSS software (v.20, SPSS Inc., Chicago, IL, USA). The variables are presented as mean±standard deviation (SD). Data normality was verified by the Shapiro-Wilk test. The comparison between the pre- and post-training for the 2 groups was made by mixed ANOVA for repeated measures. Correlations between aerobic and anaerobic parameters with 10 km running performance were performed using the Pearson correlation coefficient. The differences (i. e., effect size [ES]) were considered small when ES≤0.2, moderate when ES≤0.5 and large when ES>0.8. Furthermore, magnitude-based inferences were applied to estimate the chances of a true observed effect being positive, trivial or negative, considering the smallest worthwhile change per Hopkins et al. 18 . The probability of a positive/trivial/negative effect of the training programs was interpreted following the recommendations of Hopkins et al. 18 ; effect: <1% almost certainly not; 1–5% very unlikely; 5–25% unlikely; 25–75% possibly; 75–95% likely; 95–99% very likely;>99% almost certainly. When the chance of having positive or negative effects in an outcome were both above 10%, the qualitative inference result was considered as unclear.

Results

The results show V peak improvement in both groups after the 4 week training period: GVP=0.9 [0.4–1.4] km·h −1 ( p =0.01) and GVO 2 =0.6 [0.2–1.0] km·h −1 ( p =0.03) ( Table 2 ). A significant increase in the total duration of the incremental test was observed in both groups: GVP=2.8 [1.5–4.1] min ( p =0.01) and GVO 2 =2.2 [0.4–3.9] min ( p =0.06) ( Table 2 ).

Table 2 Mean±standard deviation (SD) difference (90% CI), magnitude of inference, and significance level for group×time interaction ( P ) for the variables: V peak , (km  .  h −1 ) Total time of the incremental test (min), HR max (bpm) RPE max (AU), LA peak (mmol  .  L −1 ) and t lim at V peak (min) obtained from the experimental protocol for determining the V peak .

	GVP (n=8)				GVO 2 (n=6)
Variable	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Group×time interaction ( P )
V peak (km·h −1 )	16.7±1.2	17.6±1.5*	0.9 [0.4–1.4]	Very likely 98/2/0	17.1±1.9	17.7±1.6*	0.6 [0.2–1.0]	Possible 72/28/0	0.352
Duration (min)	23.0±3.7	25.8±4.4*	2.8 [1.5–4.1]	Very likely 99/1/0	24.3±5.7	26.4±4.7*	2.2 [0.4–3.9]	Likely 81/19/0	0.566
HR max (bpm)	189±5.0	191±6.0	1.6 [−0.6–3.8]	Possible 66/32/2	183±10.0	184±12.0	1.8 [−4.1–7.7]	Unclear 42/48/10	0.943
RPE max (AU)	19.9±0,4	19.9±0.4	−0.1 [−0.6–0.3]	Unclear 20/23/57	19.7±0.5	19.8±0.4	0.2 [−0.5–0.8]	Unclear 55/25/20	0.449
LA peak (mmol·L −1 )	9.3±0.6	10.3±0.8	0,9 [−0.2–2.1]	Likely 81/15/3	8.0±0.6	9.0±1.0	1.1 [−1.0–3.1]	Unclear 74/15/11	0.914
t lim (min)	6.8±1.6	6.7±1.3	−0.1 [−0.6–0.4]	Unlikely 7/72/21	7.7±1.8	6.8±2.3	−0.9 [−1.7- −0.1]	Likely 1/14/85	0.130

* P <0.05 in relation to pre moment to the same group. Dif=Difference; (P/T/N)=Positive/Trivial/Negative

No significant differences were observed in either group between pre- and post-training for HR max , RPE max , t lim at V peak , t lim at vVO 2max , and LA peak .

After 4 weeks of training, we observed a significant improvement in vVO 2max only in the GVP group: 0.6 [–2.2–1.8] km·h −1 ; ( p =0.01). In relation to the total duration of the test, a significant increase was observed in both groups: GVP=1.7 [0.4–3.0] min ( p =0.036) and GVO 2 =1.2 [0.2–2.2] min ( p =0.047) ( Table 3 ).

Table 3 Mean±standard deviation (SD) difference (90% CI), magnitude of inference, and significance level for group×time interaction ( P ) for the variables: VO 2max (ml·kg −1 ·min −1 ), vVO 2max (km·h −1 ), total duration of incremental test (min) HR max (bpm) RPE max (AU), LA peak (mmol·L −1 ) and t lim at vVO 2max (min) obtained from the determination of the protocol vVO 2max

	GVP (n=8)				GVO 2 (n=6)
Variable	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Group×time interaction ( P )
VO 2max (ml . kg −1 ·min −1 )	50.2±3.5	50.0±2.3	−0.2 [−2.2–1.8]	Unclear 19/52/29	49.0±6.9	48.9±6.1	−0.1 [−1.7–1.6]	Unlikely 4/90/6	0.957
vVO 2max (km·h −1 )	16.4±1,4	17.0±1.3*	0.6 [0.3–1.0]	Likely 93/7/0	17.2±1.7	17.5±1.9	0.3 [−0.1–0.8]	Possible 37/62/1	0.317
Duration (min)	21.6±4.8	23.3±4.2*	1.7 [0.4–3.0]	Likely 81/19/0	23.7±5.9	24.9±5.2*	1.2 [0.2–2.2]	Possible 36/64/0	0.601
HR max (bpm)	193±11.0	190±6.0	−2.9 [−9.0–3.3]	Possible 7/37/54	183±8.0	182±7.0	−0.8 [−4.2–2.5]	Possible 8/65/27	0.623
RPE max (AU)	18.8±2.1	19.5±1.1	0.8 [0.0–1.5]	Possible 75/24/1	19.0±1.7	19.3±1.3	0.3 [−0.3–1.0]	Possible 43/53/4	0.470
LA peak (mmol·L −1 )	9.1±1.9	8.8±1.3	0.5 [−0.5–1.5]	Possible 67/24/9	8.4±1.1	8.0±2.5	−0.8 [−1.6–0.1]	Possible 1/28/71	0.911
t lim (min)	7.5±1.7	6.7±1.1	−0.8 [−2.3–0.6]	Possible 8/21/72	6.3±1.4	6.1±2.1	0.5 [−0.8–1.7]	Unclear 57/30/13	0.225

* P <0.05 in relation to pre moment to the same group. Dif=Difference; (P/T/N)=Positive/Trivial/Negative

Table 4 shows the values of the variables both pre- and post-training obtained in the 10 km performance. In both groups, there was a significant reduction in the time it took to run a 10 km distance after the training program (GVP (– 1.4 [−2.5 to −0.3] min; p =0.04) and GVO 2 (– 0.9 [–1.6–0.2] min; p =0.048)). Furthermore, there was a significant increase in the AS after 4 weeks of training (0.6 [0.1–1.0] km·h −1 for GVP (p=0.04) and 0.4 [0.1–0.6] km·h −1 for GVO 2 ( p =0.036)). The runners’ AS was between 14 and 16 km·h −1 (≅62–71% of the world record).

Table 4 Mean±standard deviation (SD), difference (90% CI), magnitude of inference, and significance level for group×time interaction ( P ) for the variables in the time trial of 10 km (min), average speed (AS) 10 km (km  .  h −1 ) HR max (bpm), RPE max (AU) and LA peak (mmol·L −1 ), obtained from the 10 km track performance.

	GVP (n=8)				GVO 2 (n=6)
Variable	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Pre	Post	Dif. (90% CI)	Inference (P/T/N)	Group×time interaction ( P )
Time (min)	41.3±2.4	39.9±2.7*	−1.4 [−2.5- −0.3]	Likely 1/8/91	40.1±3.4	39.2±2.9*	−0.9 [−1.6–0.2]	Possible 0/37/63	0.517
AS 10-km	14.6±0.9	15.1±1.1*	0.6 [0.1–1.0]	Likely 92/7/1	15.1±1.3	15.4±1.2*	0.4 [0.1–0.6]	Possible 60/40/0	0.478
HR max (bpm)	179±8.0	179±5.0	2.9 [−1.5–7.3]	Possible 58/39/3	171±10.0	173±8.0	6.5 [−1.0–14.0]	Likely 86/11/3	0.404
RPE max (AU)	18.8±1.9	18.8±1.9	−0.1 [−0.8–0.5]	Unclear 15/49/35	18±2.8	17.0±2.6	0.0 [−0.5–0.5]	Unclear 15/70/15	0.792
LA peak (mmol·L −1 )	7.8±2.0	7.7±1.7	−0.1 [−1.7–1.5]	Unclear 27/39/34	6.7±0.6	7.4±0.8	0.7 [−0.2–1.6]	Likely 85/9/6	0.486

* P <0.05 in relation to pre moment to the same group. Dif=Difference; (P/T/N)=Positive/Trivial/Negative

The effect size for the comparison between GVP and GVO 2 for the percentage variation after the 4 week running training period revealed a small effect for V peak and 10 km time and a moderate effect for vVO 2max , all favorable to GVP ( Fig. 1 ).

Fig. 1

Effect sizes of the comparison between GVP and GVO 2 for the variation (%) of vVO 2max (km·h −1 ) V peak (km  .  h −1 ) and the 10 km time after the 4 week running training period.

The V peak and vVO 2max were significantly correlated with the 10 km performance in both pre- and post-training time in both groups ( Table 5 ). The VO 2max , however, did not correlate with the 10 km performance at any time ( Table 5 ).

Table 5 Correlation between the performances of 10 km before and after 4 weeks of training with the variables: V peak (km  .  h −1 ), VO 2max (ml·kg −1 ·min −1 ), vVO 2max (km·h −1 ).

	GVP (n=8)		GVO 2 (n=6)
Variable (Pre and Post)	Performance Pre	Performance Post	Performance Pre	Performance Post
V peak (km·h −1 )	−0.97*	−0.86*	−0.95*	−0.94*
VO 2max (ml·kg −1 min −1 )	−0.35	0.03	−0.64	−0.70
vVO 2max (km·h −1 )	−0.82*	−0.88*	−0.99*	−0.98*

* P <0.05

Discussion

The aim of the study was to evaluate the effect of 4 weeks of training prescribed by V peak , vVO 2max , and their respective t lim in moderately trained endurance runners.

The main finding of the study was that the training prescribed by V peak or by vVO 2max promoted similar improvements for moderately trained endurance runners, which confirmed a previous hypothesis. Effect size analysis showed slightly favorable changes for GVP. A significant correlation was observed between the 10 km performance and the V peak and vVO 2max , but our hypothesis was disproven because only in the pre-training time, the GVP showed a higher correlation of the V peak with the 10 km performance compared with the vVO 2max .

For proper training prescription, it is necessary to use variables that can control and monitor the intensity of effort and possible physiological adaptations resulting from this practice and, most importantly, show a correlation with performance 7 .

In our review, we found no previous studies that had used V peak in the prescription of individualized endurance training. The GVP showed improvement in 10 km performance after 4 weeks of training, suggesting that V peak is an effective variable for prescribing training and is able to promote improvements in performance after a period of training. The improvements found in performance caused by the training prescribed by V peak were similar to those described by studies that used vVO 2max for training prescription 13 35 . As for GVO 2 , the improvements in the 10 km performance were similar to those observed for GVP after 4 weeks of training. This improvement in performance is in line with previous studies that used the same variable for training prescription 13 35 . Esfarjani and Laursen 13 observed improvements in 3 000 m performance after applying a 10 week training in 17 moderately trained runners whose training sessions were prescribed by vVO 2max and their respective t lim . Similar improvements observed in the 10 km performance by both prescription variables (V peak and vVO 2max ) can be explained by the fact that both variables are highly interrelated, as well as related to endurance performance 12 25 32 . This similarity is of great interest to coaches, athletes, and researchers, because currently vVO 2max is widely known as a variable to predict performance, monitoring, and training prescription 9 24 27 . However because it requires the use of expensive equipment, its use is limited to only a few research laboratories, coaches, and athletes. Thus, the V peak is an attractive alternative variable because of its practicality and low financial cost.

Both the V peak and vVO 2max groups showed improvement after the training program. This improvement is mainly associated with the prescription model used in the study for interval training sessions. The intensity of the V peak and vVO 2max were related to VO 2max , which is considered the ideal intensity to utilize the maximum aerobic production system energy and maintain it as long as possible 31 . Moreover, the stimuli had a duration of 60% of t lim at V peak and vVO 2max , which is considered the time required to achieve and maintain the VO 2max , resulting in an improvement in the prescription variable 5 31 . No evidence was provided, however, about the existence of a limit to the improvement in prescription variables with training, or if they might be bettered by improving the performance test.

The improved V peak demonstrates the sensitivity of this variable in that it is capable of accurately monitoring the changes caused by this type of training, which is one of the main requirements for an athletic training prescription variable 7 . Regarding vVO 2max , improvement was observed for post-training GVP, but no difference was found in the GVO 2 after the 4 weeks of training. It is noteworthy that even without a statistical difference in vVO 2max , there was significant improvement in the total duration of the incremental test when we observe the pre- and post-training duration (23.7±5.9 vs. 24.9±5.2 min, respectively). The improvement in test time and the absence of improvement in vVO 2max may be related to the methodology for its determination, which is to record the minimum intensity at which the occurrence of VO 2max was observed 2 4 . In addition to being dependent on VO 2max , this estimation is not considered the total period of the test; therefore, even with the improvement in test duration, the occurrence of VO 2max can be observed at similar intensities between the pre- and post-training, with no change in vVO 2max. This does not occur with the V peak when the Kuipers et al. 23 adjustment (which takes into account the precise length of the incomplete stage) is applied. This result shows that vVO 2max determined by this protocol is a less accurate alternative variable for monitoring training when possible adaptations are small. It also supports the use of V peak as a variable for monitoring and training prescription because it is sensitive to small changes caused by training. This sensitivity is of great interest since the more highly trained the athletes, the smaller the improvements will be. Even detection of these small gains would warrant a new training protocol.

As for t lim at V peak and vVO 2max , no difference was found for these variables after the 4 week training program. This result deserves further consideration, however, because after the training program the participants have managed to remain at t lim the same amount of time while exercising at higher intensities. These results were similar to those of Billat et al. 2 , who also found no difference at t lim after a 4 week training protocol. The t lim seems to be a variable that does not follow the changes caused by training 24 . Despite that, the application of t lim for prescribing interval training favors greater individualization of the duration of each high-intensity effort, given the large variation between subjects at t lim , even if V peak or vVO 2max do not show major differences between the subjects.

No improvements were seen at VO 2max in either group after the training program. Results from previous studies observed the effect of a training program on VO 2max in trained endurance runners with similar training prescriptions to those used in our study 2 31 35 . Even without changes in VO 2max , these studies have in common a significant improvement in performance, demonstrating that VO 2max seems to be a less sensitive training variable, which in turn suggests that the use of other variables for monitoring adaptations may be warranted 8 20 26 .

No changes were observed in variables HR max ·LA peak , or RPE max, either in the treadmill test or in track performance. The absence of change to these variables after training was expected because they are routinely used for the identification of physiological responses generated by the effort 17 . They serve as a parameter for identifying the maximum effort during the incremental test 14 . Thus, for already moderately trained runners such as our participants, the 4 week training period is a short time to promote changes in the said variables, especially in HR.

The correlation among V peak , vVO 2max , and performance in the present study was also observed in previous studies 3 26 . In the present study, the GVO 2 presented higher correlation of the performance with vVO 2max than with V peak . The ability to predict performance by vVO 2max is related to the fact that it is a variable that shows the interaction between VO 2max and running economy (RE) 3 12 26 , which are important variables for predicting performance. However, they are not able to predict the performance as isolated variables 19 , especially in individuals with similar VO 2max and/or who have a high level of training 28 . Unlike the GVO 2 , the GVP group showed a higher correlation between V peak and 10 km performance in the pre-training time. Previous studies have also shown high correlations between V peak and performance 11 36 . Noakes et al. 30 , in a study on expert runners over long distances (20 marathoners and 23 ultra-marathoners) with different performances, found that V peak determined on a treadmill and lactate threshold (LT) were the 2 best performance predictors from 10- to 90 km running performances, concluding that V peak is a great predictor of performance. Even in groups presenting different correlations of each variable (V peak and vVO 2max ) with performance, it was observed that both were able to predict performance, justified by the fact the 2 variables are highly interrelated 26 .

Although studies show that VO 2max has a great capacity for performance prediction in races ranging from 3 km through ultramarathons 1 26 28 , in this study no correlation was found between VO 2max and 10 km performance in either the pre-training time or post-training time in either group. The fact that the runners present a similar VO 2max may indicate that the VO 2max is not as efficient a variable to predict the performance when individuals have similar VO 2max 12 . The results demonstrated in this study have important practical implications for teams, coaches, and athletes in obtaining information about the adaptations induced by training, especially its effects on performance, given that the V peak is a variable of great practicality and low financial cost because it does not require expensive equipment (gas analyzer).

Based on the results of this study, it was concluded that the training prescribed by V peak promoted improvements similar to the training prescribed by vVO 2max in moderately trained endurance runners. Therefore, we recommend the additional use of V peak associated with its time limit for endurance training prescription in recreational runners with a similar training level to that of the study participants.