Draft “ The performance and aerobic endurance effects of high-intensity versus moderate-intensity continuous running

Journal: Applied Physiology, Nutrition, and Metabolism Manuscript ID apnm-2018-0575.R3 Manuscript Type: Article Date Submitted by the Author: 28-Jan-2019 Complete List of Authors: Jarstad, Even; Norwegian Sport Medicine Clinic, Pb 3843, Ullevaal Stadion Mamen, Asgeir; Norwegian School of Health Sciences, Kristiania University College, Pb 1155 Keyword: Recreational athletes, Time to exhaustion, Maximal oxygen consumption, Lactate threshold, Substrate oxidation, Running economy Is the invited manuscript for consideration in a Special Issue? : Not applicable (regular submission)

To the best of our knowledge, HICR has not been compared to lower-intensity running exercise of longer duration.The aim of the present study was, therefore, to D r a f t 4 investigate the performance and aerobic endurance effects of HICR versus MICR in recreational athletes.We chose a practical approach to the issue, by not matching the two training modalities (HICR and MICR) for total work.

Subjects
Twenty-seven subjects, who participated regularly in aerobic exercise two to three times a week, were enrolled for the present study.The subjects were randomly assigned to a HICR, MICR or no-intervention control (C) group.Three participants dropped out during the study due to lack of time, injury or self-inflicted logistical problems, whereas four subjects were excluded from analysis because of too few intervention sessions, illness or participation in another study.Ultimately, 20 subjects (seven women and 13 men), aged 28 ± 5 years, completed the present study.
Before the study commenced, the subjects gave written informed consent and the project was approved by the South-Eastern Norway Regional Committee for Medical and Health Research Ethics.

Exercise test procedures
Prior to the study, all the subjects performed a complete initial habituation test procedure; pre-and post-test were then conducted within 10 weeks.The same technician conducted the pre-and post-tests at approximately the same time of the day, in about the same climatic conditions (temperature and relative humidity) in the laboratory.To avoid communication problems and disturbances in the interaction between the technician and participants during testing, the participants were not allowed to listen to music or have any spectators in the laboratory.All tests started with measurement of body mass, followed by two treadmill protocols to determine running performance and aerobic endurance.
For determination of running economy, substrate oxidation and lactate threshold (LT), a stepwise test protocol was used.The stepwise test protocol started with an initial 10-min workload which corresponded to 55 ± 4% of the individual pre-training VO 2max , after which the velocity was increased by 1.0 km•h -1 each fifth min, for four to six workloads (modified from Borch et al. (1993)).The inclination was constant during the test and set to either 1.7% (1°) or 5.3% (3°) according to individual preferences.VO 2 , heart rate (HR) and pulmonary respiratory gas-exchange ratio (RER) were recorded in steady-state phase, between the third and fifth min at each workload (Helgerud et al. 1990;Barstow 1994), whereas blood lactate concentration (BLC) was measured during a standardized 60-sec break between the workloads D r a f t 6 (modified from Helgerud et al. (1990)).The individual velocity corresponding to ~70% pre-training VO 2max was used to indicate running economy (Støren et al. 2008), whereas RER at three individual, submaximal workloads up to ~75-80% pre-training VO 2max was used to calculate substrate oxidation (Jeukendrup and Wallis 2005).For determination of the individual LT, the following equation was used: individual BLC during warm-up + 1.5 mmol•L -1 , which has been shown to be a satisfactory approach to the "gold standard", maximal lactate steady state (Helgerud et al. 1990).
After 10 min of active rest, an exhausting ramp test protocol was used to measure VO 2max and time to exhaustion.The constant inclination during the ramp test was set to 5.3% (3°), and the test started with an initial velocity which corresponded to 91 ± 9% of the individual pre-training VO 2max .The running speed was then increased each min by 1.0 km•h -1 to a supramaximal workload (124 ± 6% of the individual pretraining VO 2max ) and volitional exhaustion within 4-8 min (modified from Tønnessen et al. (2014) and Ingjer (1992)).VO 2max was defined as the mean of the two highest consecutive 30-sec VO 2 measurements (Sandvei et al. 2012;Tønnessen et al. 2014).
The main criterion for attainment of VO 2max was a levelling off of VO 2 , despite increased workload (Taylor et al. 1955).If a levelling off of VO 2 was not present, at least two of the following three additional criteria had to be achieved: a RER ≥1. 10 (Franch et al. 1998;Sandvei et al. 2012), BLC of ≥6 mmol•L -1 two min after completed test (Sandvei et al. 2012) and visible exhaustion of the subject (Esfarjani and Laursen 2007).Retest was carried out if the criteria were not met.For estimation of vVO 2max , a mathematical regression line was drawn through three to four VO 2 values below LT (steady-state VO 2 values from the stepwise test procedure), and then extrapolated to the individual VO 2max recorded (modified from Franch et al. (1998)).
As time to exhaustion was used as a measure of running performance, each subject D r a f t performed identical workloads (inclination and velocities) during the ramp test preand post-training.The participants were not able to view the time on the running clock during the ramp test, or informed about their time to exhaustion until the project was finished.In accordance with Ingjer (1991a), HR max was calculated as peak HR (HR peak ) + 5 beats•min -1 ; this differs from Franch et al. (1998) and others (e.g.Helgerud et al. 2007;Sandvei et al. 2012), who seem to have used HR peak as HR max .
Therefore, to make this study comparable to others, we also indicate the exercise intensities in percentage of both HR peak and VO 2max in the HICR and MICR groups.

Instruments
All tests were carried out on a treadmill (Woodway ELG 2, Woodway GmbH, Weil am Rhein, Germany).VO 2 , carbon dioxide production (VCO 2 ), RER and ventilation were measured with an Oxycon Champion with mixing chamber (Vyaire Medical, Höchberg, Germany).During the exercise tests, the participants breathed through a rubber mouthpiece connected to a two-way non-rebreathing valve (2700 series, Hans Rudolph Inc., Kansas, USA), and wore a nose clip to avoid nasal breathing.The expired air was led through a ~2 m long tube into the mixing chamber for subsequent analyses of oxygen (O 2 ), carbon dioxide (CO 2 ) and ventilation.As far as the authors know, no scientific studies have validated the Oxycon Champion against the "gold standard", the Douglas bag method.However, the successor to Oxycon Champion, the Oxycon Pro, which relies on the same gas (O 2 and CO 2 ) and ventilation measurement principles as the Oxycon Champion (service manuals for Oxycon Champion and Pro (Vyaire Medical, Höchberg, Germany)), has been shown to be an accurate and reliable system for gas (O 2 , CO 2 and thus RER) and ventilation measures over a wide range of ventilations (~15-210 L•min -1 ), when compared to the Douglas bag method (Foss and Hallén 2005).Furthermore, unpublished data has shown that the Oxycon Before exercise testing, Oxycon Champion was calibrated against room air and calibration gases with known content (5% CO 2 and 95% N), whereas volume calibration was conducted manually with a 3 L syringe.BLC of haemolysed blood was measured with a YSI 1500 Sport Analyzer (YSI Inc., Ohio, USA).HR was measured by an HR monitor (Polar Electro OY, Kempele, Finland) and time measurements were carried out using a digital stopwatch (Hanhart Prisma 200,Hanhart,Germany).

Training intervention
According to individual preferences, the training intervention sessions were carried out on a motor-driven treadmill and/or outdoors.For physiological feedback, all the participants used a HR monitor (Polar Electro OY, Kempele, Finland) during every intervention session.After a 10-min warm-up, the HICR group performed a 20-min strenuous, almost exhausting, run above LT at ~83% VO 2max (~88% HR max /~91% HR peak ), whereas the MICR group performed a 40-min run at ~72% VO 2max (~80% HR max /~82% HR peak ).The intervention sessions were completed with a 10-min jog corresponding to 60-70% HR max .Exercise duration in the HICR and MICR groups was chosen based on feasibility and to enable the MICR group to achieve a substantially higher energy expenditure than the HICR group.Calculation of the energy output, by multiplying 20 kJ per litre VO 2 per min (Åstrand et al. 2003, p. 238), indicated that the HICR and MICR groups converted ~1320 and ~2230 kJ per session, respectively, excluding warm-up and recovery.The training diary showed that the HICR group performed 2.6 ± 0.2 and 1.8 ± 0.6 intervention and regular aerobic exercise sessions a week, respectively.In the MICR group, the equivalent figures were 2.5 ± 0.1 and 1.9 ± 0.9.There were no significant differences between the HICR or MICR groups in the number of training intervention or regular aerobic exercise sessions completed during the study.

Statistical analysis
Results are presented as mean ± SD.The significance level was 0.05.Differences between groups pre-training, and within groups pre-to post-training, were analysed by a two-way repeated measures ANOVA, whereas differences in the changes pre-to post-training between groups were analysed by the one-way ANOVA, Holm-Sidak method (SigmaPlot 14, Systat Software Inc., San Jose, California, USA).
Additionally, the effect sizes in the two training groups (HICR and MICR) versus the C group were calculated (corrected for bias) in cases where changes in the dependent variables might have been of practical significance, although not being statistically significant.This was conducted on a downloaded Excel spreadsheet (Centre for Evaluation and Monitoring, Durham University, UK (www.cem.org/effect-sizecalculator)).The confidence interval was 95%, and an effect size of minimum 0.50 D r a f t was considered as acceptable; two of the six effect size categories suggested by Sawilowsky (2009), "medium" (0.50-0.79) and "large" (0.80-1.19), were, therefore, relevant in the present study.

Anthropometry
Body mass and BMI decreased significantly by 1.4% and 1.3%, respectively, in the MICR group, while no such changes were observed in the HICR or C groups (Table 1).

Performance
Time to exhaustion, during the ~4-8 min ramp test procedure, increased significantly by 23% and 24% in the HICR and MICR groups, respectively, whereas no such change was found in the C group (Table 2).
No VO 2max related changes were found in the C group (Table 2).

D r a f t 11
The HICR group increased vLT significantly by 6.8%, while no such effect was observed in the MICR or C groups (Table 2).

Running economy
In the HICR group, the energy cost of running decreased numerically, but not statistically significantly, by 4.2%, with a medium effect size corresponding to -0.54 (-1.7 to +0.6) (Table 2).The energy cost of running increased numerically, but not statistically significantly, by 3.2%, with a large effect size of 1.19 (0.0 to +2.4), in the MICR group (Table 2).

Supplementary physiological variables below LT
The HICR group decreased BLC significantly at Workloads I, II and III by 20%, 23% and 28%, respectively (Table 3).Moreover, significant declines of %VO 2max at Workloads I, II and III, corresponding to 7.8%, 8.6% and 7.7%, respectively, were observed in the HICR group (Table 3).No such changes were observed in the MICR or C groups (Table 3).
In the MICR group, RER decreased numerically, but not statistically significantly, by 2.2% and 3.3% at Workload I and II, respectively, with a medium effect size corresponding to -0.52 (-1.6 to +0.6) and a large effect size of -0.80 (-1.9 to +0.3), respectively (Table 3).Furthermore, RER was significantly decreased by 4.2% at Workload III in the MICR group (Table 3), whereas no such effects were observed in the HICR or C groups (Table 3).

Differences in changes pre-to post-training between groups
The significant increase in time to exhaustion in the HICR and MICR groups was numerically, but not statistically significantly, greater than the change in time to exhaustion in the C group (HICR versus C: P = 0.09, MICR versus C: P = 0.08) preto post-training (Table 2).Furthermore, there was no significant difference in the D r a f t 12 change in time to exhaustion pre-to post-training between the HICR and MICR groups (P = 1.00) (Table 2).The only significant inter-group difference in changes pre-to post-training was the decreased %VO 2max at workload III observed in the HICR versus C group (Table 3).

Discussion
The main finding in the present study was that HICR increased time to exhaustion similarly to MICR, non-matched for total work.Other notable findings were that HICR increased absolute VO 2max and generated a tendency towards enhanced running economy, whereas MICR increased relative VO 2max and seems to have enhanced fat oxidation.These overall findings indicate that HICR and MICR can produce similar performance improvements, despite various physiological alterations.

HICR
The performance (time to exhaustion), maximal (VO 2max and vVO 2max ) and submaximal effects (running economy, and BLC and %VO 2max at three submaximal velocities) in the HICR group of the present study support the results in the HICR group in the study by Franch et al. (1998).Moreover, the performance and aerobic endurance improvements in the HICR group are also in accordance with training effects reported following HIIR in recreational athletes (Thomas et al. 1984;Olsen et al. 1988;Franch et al. 1998;Esfarjani and Laursen 2007;Helgerud et al. 2007).
However, no changes of RER were found following HICR, indicating no alteration in fat oxidation.This differs from the finding of Ulloa et al. (2014), who reported significantly increased maximal fat oxidation rate in recreational runners following eight weeks of HIIR (5-6 x 1 km/3 min pauses, three times a week) performed at similar exercise intensity (~83% VO 2max ) as that used in the HICR group of the D r a f t 13 present study.These discrepant findings can be explained by the use of indirect calorimetry and its limitations, as well as the use of different equations for calculating substrate oxidation (Jeukendrup and Wallis 2005).It can also be speculated whether the participants in the HICR group of the present study had already adapted their fat metabolism system through their regular running exercise activities prior to the study.
On the other hand, the MICR group, which had similar aerobic fitness pre-training to the HICR group in the present study, appeared to have increased fat oxidation (decreased RER).Therefore, further investigations may be necessary to produce more scientific evidence about the fat oxidation effects of HICR versus MICR in recreational athletes.

MICR
The improvement of performance and/or relative VO 2max (mL•kg -1 •min -1 ) in the MICR group of the present study supports findings following MICR in previous studies of recreational athletes (Thomas et al. 1984;Rowan et al. 2012;Sandvei et al. 2012).Moreover, the decrease in RER (0.02-0.04 units) at three submaximal velocities (Workloads I-III) below LT in the MICR group, which indicates enhanced fat oxidation, is in accordance with the results by Ulloa et al. ( 2014), who found enhanced maximal fat oxidation rate following eight weeks of MICR (5-6 km, three times a week, ~62% VO 2max ) in recreational runners.The small, non-significant, increase (0.4-1.8%) in absolute VO 2 (L•min -1 ) at the same three submaximal velocities (Workloads I-III) in the MICR group may support the decreased RER and a possible increase in fat oxidation in this (MICR) group, as decreases in respiratory quotient (RQ) from 0.95 to 0.91, 0.92 to 0.89 and 0.91 to 0.89 (i.e. an average increase in fat oxidation from ~25% to ~35% of the total energy expenditure) at a D r a f t 14 given workload (for a given energy yield) increase the VO 2 demand, theoretically, by ~0.5-1.0%(McArdle et al. 2015, p. 188).
In contrast to the results in the MICR groups (~24 min at 85% HR max , or 45 min at 70% HR max ) in the study by Helgerud et al. (2007), who improved their running economy, the MICR group in the present study showed a tendency towards increased energy cost of running.However, improved running economy is to be expected in previously non-running-trained subjects suddenly performing a relatively large amount of running exercise (Helgerud et al. 2007).Therefore, one explanation for the contradictory findings regarding the energy cost of running could be the fact that the participants in the present study were used to performing running exercise regularly, unlike those in the study by Helgerud et al. (2007).Furthermore, the possible increase in fat oxidation, and thereby the slightly higher absolute VO 2 submaximally, observed in the MICR group in the present study may also partially explain the discrepant results regarding the energy cost of running in the MICR group in the present study versus the MICR groups in the study by Helgerud et al. (2007).

High-intensity versus moderate-intensity running exercise
The impact of different exercise intensities in the development of the aerobic endurance "key factor", VO 2max , has been studied for decades.In this context, Helgerud et al. (2007) and others (Thomas et al. 1984;Wenger and Bell 1986) have reported that intensity of training cannot be compensated by longer duration.The findings in the present study point in the same direction as the results in the abovementioned studies (Thomas et al. 1984;Wenger and Bell 1986;Helgerud et al. 2007), as the HICR group increased absolute VO 2max significantly, whereas the MICR group did not (Table 1).Furthermore, the tendency towards increased vVO 2max observed in the HICR group, which was substantial and superior to that of the MICR group, is in D r a f t 15 accordance with the results in the study by Enoksen et al. (2011), who found that "high-intensity low-volume" training (82-92% HR max ) improved vVO 2max more than "high-volume low-intensity" training (65-82% HR max ) in middle-distance runners.
This indicates that intensity of training may also be crucial in the development of

HIIR versus HICR
In most studies aimed at investigating the training effects of high-intensity aerobic running exercise, HIIR has been the preferred modality (e.g.Thomas et al. 1984;Olsen et al. 1988;Esfarjani and Laursen 2007;Helgerud et al. 2007;Ulloa et al. 2014).Also, in practice, HIIR seems to be a favoured modality among recreational and competitive athletes, due to its intermittent character (which gives the athlete breaks "to look forward to"), time-efficiency and rapid improvement of performance and aerobic endurance.HICR, on the other hand, allows the athlete to exercise at high training intensity without spending time on breaks.Thus, within a given total exercise time, including identical warm-up and recovery, HICR enables the athlete to achieve a larger amount of effective training (i.e.longer effective running distance) at high HR max ), than the performance improvement (67%) following HIIR (~5664 m, 4-6 x 4 min/2 min pauses, ~16.6 km•h -1 /~94% HR max ) observed in that particular study.Therefore, if the purpose is to improve running performance as much as possible, within a given total exercise time available (including warm-up and recovery), HICR may be recommended instead of HIIR.However, as only one study (Franch et al. 1998) appears to have addressed this issue, further research is needed to produce more scientific evidence about the performance and physiological effects of HIIR versus HICR.

Women versus men
In the present study, women and men were pooled in the training (HICR and MICR) and C groups, which may be debatable since women have systematically lower VO 2max than men (Edvardsen et al. 2013), and perform at a lower level than men in endurance sport events (Åstrand et al. 2003, p. 269).However, women and men with the same exercise status pre-training seem to respond similarly physiologically to a given aerobic training stimulus (Thomas et al. 1984;Chandler et al. 1996).Therefore, as the women and men in the present study had approximately similar exercise status and relative VO 2max (ml•kg -1 •min -1 ) compared to normative data for age group and gender, pre-training, the use of pooled groups of women and men should not have been a disadvantage in this case.

Limitations
In the present study, the number of subjects was relatively low in each group, which limited parts of the statistical analysis performed.For example, with a higher number D r a f t 17 of participants, the significant increases in time to exhaustion within the HICR and MICR groups would most likely also have been significantly greater than the change in time to exhaustion in the C group pre-to post-training.Moreover, with larger groups, it is not unlikely that the changes in the main physiological variables, VO 2max , LT and/or running economy, as well as RER submaximally, would have differed significantly between the groups pre-to post-training.
The participants' usual aerobic exercise pattern and intensity prior to the present study was not logged.Therefore, although the C group did not show any performance or physiological changes, it cannot be completely excluded that possible alterations in habitual exercise routines during the study might have influenced the training effects in the HICR and/or MICR groups.
In the present study, time to exhaustion during the ramp test pre-and post-training was used to evaluate changes in running performance.To the best of our knowledge, no studies have investigated the reliability of time to exhaustion in this ramp test procedure.However, it has been shown that exhaustive treadmill protocols, lasting approximately 6 to 7 min, running at vVO 2max , and 1500 m time-trial time, may have a coefficient of variation (CV) of ~13-17% (Currell and Jeukendrup 2008).Shorter time-to-exhaustion protocols (~1-2.5 min) performed at supramaximal workloads (120-125% VO 2max /150% W max ) have been shown to have a CV of ~2-10% (Currell and Jeukendrup 2008).Thus, as the ramp test in the present study lasted ~4-8 min, and was completed at a supramaximal velocity (~125% pre-training VO 2max ), it is not unreasonable to expect a time-to-exhaustion CV of ~10% in this ramp test.However, this issue remains to be investigated.Therefore, although time to exhaustion was improved by >20% in the HICR and MICR groups, with no changes in the C group, D r a f t 18 the performance effects in the two training groups (HICR and MICR) must be interpreted with caution.
There may also be some limitations to be considered with respect to the interpretation of substrate utilization.Although indirect calorimetry has been shown to accurately determine substrate oxidation at exercise intensities up to ~75-85% VO 2max (Jeukendrup and Wallis 2005), it is important to be aware that overbreathing, where CO 2 "blows off" from the lungs, or additional CO 2 from increasing reliance on lactic anaerobic metabolism above ~60-65% VO 2max , generates a rise in RER that does not reflect RQ from aerobic metabolism (McArdle et al. 2015, pp. 189, 290-291).
Furthermore, the type, amount and timing of nutrient intake before exercise are also factors that may influence substrate oxidation (Achten and Jeukendrup 2004).
However, the nutrient intake was not assessed prior to the exercise tests in the present study.Therefore, the results of the present study regarding RER must also be interpreted with caution.

Practical implications
Previous studies have shown that ~25-45 min of moderate-intensity continuous (70-85% HR peak /HR max ) and/or 4-6 x 4 min high-intensity interval training (85-95% HR peak /HR max ), two to three times a week, for 6-12 weeks, can improve performance and/or aerobic endurance significantly in both cardiac (Wisløff et al. 2007;Helgerud et al. 2011) and pulmonary patients (Bjørgen et al. 2009), in moderately trained individuals (Franch et al. 1998;Helgerud et al. 2007) and in soccer players at a relatively high competitive level (Helgerud et al. 2001).It is, therefore, not unreasonable to believe that the two training modalities (HICR and MICR) successfully examined in moderately trained recreational athletes in the present study may produce performance and physiological gains also in different patients D r a f t 19 undergoing rehabilitation, as well as healthy untrained individuals and athletes competing at higher levels in various endurance and team sports.In this context, it is important to emphasize that the exercise duration and training frequency must be customized individually, based on the initial fitness level of the subject(s).This probably implies shorter training duration during HICR (<20 min) and MICR (<40 min) in patients and untrained individuals, initially, whereas longer exercise duration (HICR: >20 min, MICR: >40 min), higher training intensity (HICR: >88% HR max , MICR: >80% HR max ) and/or more frequent exercise frequency (>3 times per week), than used in the present study, may be necessary in high-level endurance and team sport athletes to gain performance and aerobic endurance effects.However, when it comes to HICR, this remains to be further investigated, as the scientific evidence about the training effects of this modality (HICR) appears to be non-existent in patients, healthy untrained individuals and high-level endurance and team sport athletes.
MICR groups were set to perform three training intervention sessions a week, and had to complete at least 80% of the intervention sessions within 10 weeks.In accordance with Franch et al. (1998), both training groups also had to maintain at least ~60% of their habitual aerobic exercise during the study.This was considered appropriate for a satisfactory assessment of training effects among initially trained subjects.All participants in the two training groups kept a diary log of exercise duration, training intensity and type of exercise during the intervention and habitual aerobic training sessions.
vVO 2max .On the other hand, among professional and some recreational runners, the main purpose of aerobic training is to improve running performance.In such a context, the findings of an equal increase in time to exhaustion in the HICR and MICR groups of the present study indicates that intensity of training can be compensated by longer duration.Therefore, based on the findings in the present study, both high-and moderate-intensity running exercise can be recommended in training regimes aimed at improving running performance and aerobic endurance.However, among recreational athletes with limited amount of time to spend on physical training, high-intensity aerobic running exercise (HIIR and/or HICR) can be a time-saving alternative.