Immediate E ﬀ ects of an Inverted Body Position on Energy Expenditure and Blood Lactate Removal after Intense Running

: We compared the immediate e ﬀ ects of a cool-down strategy including an inverted body position (IBP: continuous 30-s alternations of supine and IBP) after a short period of an intense treadmill run with active (walking) and passive (seated) methods. Fifteen healthy subjects (22 years, 172 cm, 67 kg) completed three cool-down conditions (in a counterbalanced order) followed by a 5-min static stretch on three separate days. Heart rate, energy expenditure, blood lactate concentration, fatigue perception, and circumference of thighs and calves were recorded at pre- and post-run at 0, 5, 10, 20, and 30 min. At 5 min post-run, subjects performing the IBP condition showed (1) a 22% slower heart rate ( p < 0.0001, ES = 2.52) and 14% lower energy expenditure ( p = 0.01, ES = 0.48) than in the active condition, and (2) a 23% lower blood lactate than in the passive condition ( p = 0.001, ES = 0.82). Fatigue perception and circumferences of thighs and calves did not di ﬀ er between the conditions point (F 10,238 < 0.96, p < 0.99 for all tests). IBP appears to produce an e ﬀ ect similar to that of an active cool-down in blood lactate removal with less energy expenditure. This cool-down strategy is recommended for tournament sporting events with short breaks between matches, such as Taekwondo, Judo, and wrestling. was performed. After completion of the static the post-run 10-, 20-, and 30-min measurements were recorded.


Introduction
Progressive overload is an essential principle in developing physical fitness, and it often requires fatigue-inducing training [1]. Fatigue is defined as a physiological (and psychological) constraint that results in failure to maintain a certain amount or intensity of voluntary muscle contractions [2]. By overloading exercise-induced fatigue, athletes may experience a temporary (e.g., a few days) performance reduction, referred to as functional overreaching [3]. Nonfunctional overreaching is also considered as the normal process of training adaptation, although it may take a longer time (e.g., a few weeks) to restore function [4]. Overtraining results in performance decrease with persistent physical and/or psychological fatigue for an extended period of time (e.g., a few months) [5]. While athletes and coaches seek optimal training adaptations, inadequate physiological (and psychological) recovery between training sessions can lead to overtraining [6]. Therefore, cool-down strategies to balance training stimulation and physical recovery are just as important as the training program.

Experimental Design
We used a three (condition) × six (time) single-blinded cross-over design with repeated measures of time. To objectively increase exercise-induced blood lactate, a treadmill run was implemented. In this approach, we assumed that heart rate increase by exercise correlates linearly with the amount of blood lactate. To test the condition effect, dependent measurements among the conditions were compared at each time point. Each condition had a 10-min cool-down activity that consisted of 5 min of IBP, walking on the treadmill or sitting on a chair, and another 5 min of static stretching (same activity for each condition). Therefore, the post-run 5-min time point reflected the acute effect of the IBP whereas the post-run 20-and 30-min time points corresponded to the combined effect of IBP and stretching. The temperature and relative humidity in the laboratory were maintained at 25 • C and 50% during the data collection period. One researcher guided the treadmill run and conducted the measurements while being blinded to the cool-down condition by leaving the laboratory while another researcher conducted the cool-down protocol. The study was approved by the Institutional Review Board and was conducted in accordance with the Declaration of Helsinki.

Subjects
Fifteen healthy active adults (5 females: 20.4 ± 0.6 years, 161.4 ± 3.3 cm, 57.8 ± 4.2 kg; 10 males: 22.9 ± 0.9 years, 176.5 ± 1.9 cm, 72.1 ± 0.7 kg) who were participating in club sports (>240 min/week) were recruited for this study. Subjects had no history of lower-extremity surgery and were free from musculoskeletal injury within the last six months. Subjects with current lower-extremity Upon arrival at the laboratory for the first time, subjects read and signed the informed consent form. At each visit, subjects were fitted with a heart rate monitor, and asked to lie down on a treatment table for a 10-min rest to achieve cardiovascular stability before measuring their resting heart rate.
Pre-run measurements were recorded in the order of heart rate, energy expenditure, fatigue perception, circumference of the thighs and calves, and blood lactate (measurements were recorded in the same order at all other time points). Subjects then performed an exercise protocol on the treadmill (Jog Forma, Technogym S.p.A, Gambettola, Italy) where the run began at 4 km/h without treadmill inclination and the speed and inclination increased by 1 km/h and 0.5% every minute. The exercise protocol was ended when the subjects reached the target heart rate (90% of the maximal heart rate calculated using the Karvonen formula) [24]. The post-run 0-min measurements were recorded immediately. After a set of measurements, the researcher left the laboratory and a different researcher came into the laboratory to guide the cool-down conditions. Subjects executed the three cool-down conditions (IBP, active, and passive) in a counterbalanced order. For the IBP condition, subjects performed five sets of continuous alternating positions on the inversion table. Each set was defined as a position change from A to B for 30 s (Figure 2A to Figure  2B) followed by a position change from B to A ( Figure 2B to Figure 2A). The duration of inversion was determined to prevent discomfort due to increased intracranial pressure [25]. To change positions, the researcher asked a subject to press a button on the inversion table. For the active condition, subjects walked on the treadmill at speeds of 5 and 6 km/h for females and males, respectively, for 5 min. The walking treadmill speed was determined in our pilot study where we wanted it to be neither too fast (to avoid additional fatigue) nor too slow (no effect on active recovery). For the passive condition, subjects remained seated on a chair for 5 min. After the next set of measurements (post-run 5 min), a 5-min static stretch of the hip adductors, extensors, abductors, and flexors, and knee extensors with a 30-s hold on each side (the same protocol was applied for each condition) was performed. After completion of the static stretch, the post-run 10-, 20-, and 30-min measurements were recorded. Upon arrival at the laboratory for the first time, subjects read and signed the informed consent form. At each visit, subjects were fitted with a heart rate monitor, and asked to lie down on a treatment table for a 10-min rest to achieve cardiovascular stability before measuring their resting heart rate.
Pre-run measurements were recorded in the order of heart rate, energy expenditure, fatigue perception, circumference of the thighs and calves, and blood lactate (measurements were recorded in the same order at all other time points). Subjects then performed an exercise protocol on the treadmill (Jog Forma, Technogym S.p.A, Gambettola, Italy) where the run began at 4 km/h without treadmill inclination and the speed and inclination increased by 1 km/h and 0.5% every minute. The exercise protocol was ended when the subjects reached the target heart rate (90% of the maximal heart rate calculated using the Karvonen formula) [24]. The post-run 0-min measurements were recorded immediately. After a set of measurements, the researcher left the laboratory and a different researcher came into the laboratory to guide the cool-down conditions. Subjects executed the three cool-down conditions (IBP, active, and passive) in a counterbalanced order. For the IBP condition, subjects performed five sets of continuous alternating positions on the inversion table. Each set was defined as a position change from A to B for 30 s (Figure 2A,B) followed by a position change from B to A (Figure 2A,B). The duration of inversion was determined to prevent discomfort due to increased intracranial pressure [25]. To change positions, the researcher asked a subject to press a button on the inversion table. For the active condition, subjects walked on the treadmill at speeds of 5 and 6 km/h for females and males, respectively, for 5 min. The walking treadmill speed was determined in our pilot study where we wanted it to be neither too fast (to avoid additional fatigue) nor too slow (no effect on active recovery). For the passive condition, subjects remained seated on a chair for 5 min. After the next set of measurements (post-run 5 min), a 5-min static stretch of the hip adductors, extensors, abductors, and flexors, and knee extensors with a 30-s hold on each side (the same protocol was applied for each condition) was performed. After completion of the static stretch, the post-run 10-, 20-, and 30-min measurements were recorded.

Measurements
The total running time, average treadmill speed, and inclination were recorded for each subject during each treadmill run.
For the blood lactate measurements, a finger-pricking lancet needle (26 G, Lancets, Moa, Korea) was administered to collect 0.7 μL blood samples on lactate test strips (Lactate Plus Lactate Test Strips, Nova Biomedical, Waltham, MA, USA). The strips were then inserted into a lactate meter (Lactate Plus, Nova Biomedical, USA) for analysis. The blood lactate analysis using this device was previously validated [26].
To measure heart rate and energy expenditure, the subjects wore a heart rate monitor consisting of a Polar M400 GPS watch on their wrist and a Polar H7 strap on their chests throughout the experiment (Polar Electro Oy, Kempele, Finland). After the device was wirelessly connected to the mobile application (Polar Beat), the subject's age, sex, height, and mass were entered. All data recorded throughout the experiment were downloaded from a website (http://flow.polar.com). The perception of subjective fatigue level was quantified using a modified 10-cm visual analogue scale [27,28] with the terms "unfatigued" and "fatigued" at each end. Subjects were asked to mark their fatigue level at each measurement time.
Using a tape measure, the circumferences of the thighs and calves were recorded (to the nearest 0.1 cm) at the midpoint between the anterior superior iliac spine and the superior pole of the patella and the thickest circumference on the triceps surae, respectively. Both sides were recorded each time (left side first).

Statistical Analyses
Our sample size was calculated based on a pilot study conducted in our laboratory. We expected a mean difference in blood lactate concentration of 3 mmol/L with a standard deviation of 3.5 mmol/L (an effect size of 0.86). These calculations estimated that 14 individuals would be necessary in each condition (an alpha of 0.05 and a beta of 0.2). This calculation was supported by a similar previous study (n = 14) that reported a difference of 3.17 mmol/L [10].
Descriptive data (means and 95% confidence intervals) were calculated and checked for

Measurements
The total running time, average treadmill speed, and inclination were recorded for each subject during each treadmill run.
For the blood lactate measurements, a finger-pricking lancet needle (26 G, Lancets, Moa, Korea) was administered to collect 0.7 µL blood samples on lactate test strips (Lactate Plus Lactate Test Strips, Nova Biomedical, Waltham, MA, USA). The strips were then inserted into a lactate meter (Lactate Plus, Nova Biomedical, USA) for analysis. The blood lactate analysis using this device was previously validated [26].
To measure heart rate and energy expenditure, the subjects wore a heart rate monitor consisting of a Polar M400 GPS watch on their wrist and a Polar H7 strap on their chests throughout the experiment (Polar Electro Oy, Kempele, Finland). After the device was wirelessly connected to the mobile application (Polar Beat), the subject's age, sex, height, and mass were entered. All data recorded throughout the experiment were downloaded from a website (http://flow.polar.com). The perception of subjective fatigue level was quantified using a modified 10-cm visual analogue scale [27,28] with the terms "unfatigued" and "fatigued" at each end. Subjects were asked to mark their fatigue level at each measurement time.
Using a tape measure, the circumferences of the thighs and calves were recorded (to the nearest 0.1 cm) at the midpoint between the anterior superior iliac spine and the superior pole of the patella and the thickest circumference on the triceps surae, respectively. Both sides were recorded each time (left side first).

Statistical Analyses
Our sample size was calculated based on a pilot study conducted in our laboratory. We expected a mean difference in blood lactate concentration of 3 mmol/L with a standard deviation of 3.5 mmol/L (an effect size of 0.86). These calculations estimated that 14 individuals would be necessary in each Appl. Sci. 2020, 10, 6645 5 of 12 condition (an alpha of 0.05 and a beta of 0.2). This calculation was supported by a similar previous study (n = 14) that reported a difference of 3.17 mmol/L [10].
Descriptive data (means and 95% confidence intervals) were calculated and checked for normality using the Shapiro−Wilk test. To test condition effects over time, a mixed model analysis of variance (random variable: subjects; fixed variables: condition and time) and Tukey's test were performed on each dependent measurement. To determine practical significance, Cohen's d effect size [29] was also calculated when statistical significance was found. To test for a correlation between the blood lactate concentration and fatigue perception, Pearson's correlation coefficient was calculated. SAS software (SAS 9.3, Institute Inc., Cary, NC, USA) was used for all tests (p < 0.05).

Results
The mean ± 95% confidence interval total running time was 7 min and 39 s ± 46 s at a treadmill speed of 7.3 ± 0.4 km/h and an inclination of 1.7 ± 0.2%. The mean treadmill speed and inclination at the end of each run were 11.1 ± 0.8 km/h and 3.6 ± 0.4%, respectively.
The blood lactate concentration was different among the three conditions at post-run 5 and 10 min (condition × time interaction: F 10,238 = 3.04, p = 0.001; condition effect: F 2,238 = 12.38, p < 0.0001; Table 1). Specifically, the IBP condition (6.7 mmol/L, 23%, p = 0.001, ES = 0.82) and the active condition performed on each dependent measurement. To determine practical significance, Cohen's d effect size [29] was also calculated when statistical significance was found. To test for a correlation between the blood lactate concentration and fatigue perception, Pearson's correlation coefficient was calculated. SAS software (SAS 9.3, Institute Inc., Cary, NC, USA) was used for all tests (p < 0.05).

Results
The mean ± 95% confidence interval total running time was 7 min and 39 s ± 46 s at a treadmill speed of 7.3 ± 0.4 km/h and an inclination of 1.7 ± 0.2%. The mean treadmill speed and inclination at the end of each run were 11.1 ± 0.8 km/h and 3.6 ± 0.4%, respectively.

Discussion
The purpose of this study was to examine how the IBP condition consisting of continuous alternation between 30-s IBP and a supine position for 5 min would help physiological and psychological recovery after intense running relative to active (walking) and passive (sitting) recovery methods. While walking is a commonly recommended method [10,31] and sitting is considered as a control, we were interested in comparing these existing and frequently practiced methods to a recovery condition containing IBP. During the initial 5 min of the recovery period, the IBP was effective in reducing the blood lactate and heart rate to levels similar to the active and passive conditions. In other words, the IBP condition produced an effect similar to that of the active condition in blood lactate with less energy expenditure. This observation suggests that the recovery condition including IBP could be an alternative recovery strategy after a short bout of intense exercise. Since blood lactate removal is dependent on the amount of blood flow to working muscles, and the process is facilitated by muscle contraction [17], performing IBP could be beneficial for individuals who need a fast recovery while minimizing energy consumption.
The IBP and active (walking at a speed of 5 to 6 km/h) strategies in this study reduced blood lactate by 32% (9.8 to 6.7 mmol/L) and 27% (9.4 to 6.9 mmol/L), respectively, at 5 min postexercise (post-run 0 min). The effects of recovery strategies with active movements [10,31,32], electrotherapy [33], and massage [31] were previously reported. Comparable corresponding values at similar time points in other studies included a 19% (14.9 to 12.0 mmol/L) reduction with a 5-min treadmill run (at 50% of maximal speed) [32] and a 49% (3.9 to 2.0 mmol/L) reduction with 6 min of motor level electrotherapy (at 9, 8, and 7 Hz for 2 min each via 8 electrodes) [33]. At 10 min postexercise (after subjects had finished a 5-min static stretch in our study), blood lactate was reduced by 45% (9.8 to 5.4 mmol/L) and 53% (9.4 to 4.4 mmol/L) in the IBP and active conditions, respectively. Other studies reported reductions of 40% (14.9 to 8.9 mmol/L) with a 10-min treadmill run (at 50% of the maximal speed) [32], 51% (11.7 to 5.7 mmol/L) with a 10-min free style swim (at 65% of each individual's best 200-m front crawl swim velocity) [31], and 39% (11.6 to 7.1 mmol/L) with a 10-min massage (using the effleurage, petrissage, tapotement, and compression techniques) [31]. As indicated by these indirect comparisons at 5 and 10 min postexercise, our IBP condition (a combination of the 5-min IBP and static stretch) appears to be as effective as other recovery activities. Although not directly measured, we believe that the effect of the IBP could have been attributed to hemodynamic changes similar to those of using compression garments [34,35]. According to men's and women's Taekwondo Kyorugi schedules at the Olympic Games in Rio in 2016, the rests between matches ranged from 360 min to less than 60 min. In the Asian Games in Jakarta−Palembang in 2018, the rest times in the men's fencing épée ranged from 80 min to 175 min. During such short time periods, a recovery activity enhancing a rapid blood lactate removal with minimal energy expenditure would contribute to the performance in the subsequent match. Because all of the comparable recovery strategies reported require an electrical device [33], massage personnel [31], or to expend more energy [31,32], our method with IBP could be the best option. Although the blood lactate values in the three conditions in our study became similar at 20 min postexercise, IBP is still beneficial because it resulted in a longer time to rest with a low blood lactate level.
Fatigue perception did not differ among the three recovery conditions at each time point. Previously, electrotherapy (motor level stimulation via 8 electrodes) and the passive (sitting on a chair) recovery method reduced perceived exertion by 46% and 38% compared with active recovery, respectively [33]. Another study [32] reported 18% less perceived exertion with passive (sitting on a chair) recovery than with active (treadmill run at 50% maximal speed) recovery [32]. Although the results of the Borg scales (0 to 10 points: [33]; 6-20 points: [32]) are comparable to our fatigue perception because both are measures of subjective physical condition, the differences of the measurement characteristics could explain why there was no difference across the conditions in our study. The Borg scale matches a numeric scale with descriptions of exertion, asking subjects to indicate their exertion level, whereas our measurement simply asked subjects to place themselves between "unfatigued" and "fatigued" on a visual analogue scale. Measurements were also taken during exercise using the Borg scale and after exercise in our study. Since subjects in previous studies felt less exertion in the passive condition than those in the active condition [32,33], providing different descriptions (e.g., fresh or recovered) or using a more detailed numeric scale may have produced a different result. Interestingly, blood lactate and fatigue perception values had a strong correlation in our study (r = 0.75, r 2 = 0.57). Thus, at least 57% of an individual's perceived fatigue can be explained by blood lactate and vice versa. However, fatigue perception returned to the pre-run level at 20 min postexercise (post-run 20 min) whereas blood lactate levels did not. This observation has several practical implications. First, people may feel complete recovery although their bodies are still in the process of physiological recovery. Second, a complete physiological recovery may be unnecessary for psychological (perceptual) recovery. Third, in addition to blood lactate, there could be other factors which contribute to the level of fatigue perception. A future study should examine the relationships between psychological, physiological, and athletic performance.
Similarly, there were no changes in the circumferences of lower extremities at any time point. Given that a tape measure was used to obtain circumferences in many previous studies [34][35][36][37], the measurement reproducibility in our study was strong (ICC value of 0.99 for between-session reliability in all measurements), measurement technique or errors were unlikely to contribute to the results. In fact, we did not even observe a circumference change at the post-run measurements, which could be explained by the exercise duration and mode (specific energy system). Previously reported postexercise extremity circumference was induced by resistance [34,35] or endurance exercise [36,37]. Therefore, an approximately 7-min all-out treadmill run may not have been sufficient to increase leg swelling. Exercise-induced transient hypertrophy, also known as "the pump", is thought to be caused by the accumulation of intracellular fluid from reactive hyperemia [38] or osmolytic activity [39]. In endurance activities, the expansion of extracellular volume [37] due to increased plasma volume and sodium retention [40] is thought to be responsible for oedematous swelling. While we assumed that our intense treadmill run mostly used the combined energy system of anaerobic glycolysis, it produced neither of the physiological mechanisms, resulting in no circumference change. Because body fluid removal is an important facet of physical recovery, a future investigation should examine how the IBP strategy affects the increased circumference caused by either resistance or prolonged endurance exercise. Appendicular body composition analysis along with skinfold thickness would provide better accuracy in the assessment of exercise-induced anthropometric changes.
There are several limitations and assumptions in our study that may benefit future research. Although good criterion-related validity and high reliability in the estimation of energy expenditure using the same heart rate monitoring device were established [41], the major contributing factors such as the physical activity and sedentary levels [42], and percentage of body fat [43] were not reflected. Additionally, a lower heart rate recorded during the IBP might have underestimated the whole-body metabolism and energy consumption. Therefore, we acknowledge the level of accuracy in estimating energy expenditure as a limitation. While accepted, our hypothesis was based on increased blood flow for a body position with 15 • head-down tilt [21][22][23] and the lack of direct measures on blood flow parameters limit us to explain and interpret the observed benefits. As an elevated blood pressure during and after exercise is typical, performing IBP immediately after exercise may be harmful by applying additional pressure to blood vessels. When comparing the inversion duration (90 • for 2 min) in previous reports [44,45] to ours (90 • for 13 s: Figure 2), we assume that the possibility of additional blood pressure was minimal. Additionally, our subjects' heart rate at the last minute of the treadmill run and the beginning of IBP were 183 and 121 bpm, respectively. A 2-min interval for the post-run measurements resulted in a 34% reduction in heart rate (183 to 121 bpm). Considering that both heart rate and blood pressure are linearly related to exercise intensity [46], a reduced heart rate value indirectly supports the idea that our subjects performing IBP were not in a deleterious situation. However, future studies should attempt to compare the change of blood flow volume and pressure during the recovery period to resolve our limitations. Lastly, the contraindications of IBP should be addressed since the IBP may be unsafe for pregnant individuals or those with abnormal blood pressure or other cardiovascular diseases, eye or ear injuries, or herniated discs.
According to our data, a 10-min recovery strategy consisting of 5 min of IBP and a 5-min static stretch was the most effective of the tested protocols in removing blood lactate and minimizing energy expenditure during the acute recovery period (the first 30 min, postexercise). Therefore, the recovery condition including our IBP protocol is an appropriate strategy to enhance physical and psychological recovery. We particularly recommend this IBP protocol for athletes who complete several matches (events) within a few hours on a single day (e.g., tournaments for Taekwondo, Judo, weightlifting, fencing, wrestling, sprint cycling, and swimming). In the final stage of matches (e.g., after quarterfinals), the fatigue level would be high. In this situation, physical fitness and stamina can play bigger roles than technical and tactical strategies. Therefore, we recommend practicing our IBP method between matches. If an IBP device is unavailable, peer-assisted IBP or a supine position with legs elevated against a wall could be used instead.