Valid estimates of energy expenditure are based on relatively longer duration, steady rate, easy to moderately intense aerobic exercise where concomitant measurements of steady-state oxygen (O2) uptake are likewise collected. Steady-state O2 uptake measurements are considered the gold standard in the estimation of exercise-energy expenditure. Lifting a weight and resistance training are however often brief and intense conditions that preclude the use of steady-state O2 uptake to accurately estimate energy expenditure. Moreover, at moderate lifting intensities, O2 delivery to working skeletal muscle can be compromised because muscle contraction is intense enough to effectively “pinch-off” arterial and venous flow (12), further suggesting that the sole measurement of exercise O2 uptake cannot be used to accurately estimate resistance training energy expenditure. Indeed, lifting a weight and resistance training are generally thought to be largely “fueled” by anaerobic metabolism, invoking an anaerobic component to energy expenditure (13). Because physiological conditions differ between aerobic (e.g., running) and anaerobic (e.g., resistance training) exercise, nonsteady state methods to estimate energy expenditure for intense strength, speed, and power-related activities have been proposed (9-11).
One nonsteady state approach to the estimation of energy expenditure is to collect not only a measure of O2 uptake during exercise but also a measure of blood lactate after exercise to estimate anaerobic energy expenditure, along with a modified measurement of recovery O2 uptake to estimate the energy demands of the recovery from 1-set of resistance training (9-11). We previously found that blood-lactate levels after free resistance training (at 60 and 80% of a 1 repetition maximum [1RM]) can be high enough to significantly contribute to exercise and recovery aerobic energy expenditure, especially when lifting occurs to fatigue (9). A follow-up investigation using the Smith machine (at 50% of 1RM) where work could be more carefully recorded, demonstrated the reliability of our non-steady-state methodology (with longer data collection periods-e.g., more reps-having greater reliability) (11). Using steady-state O2 uptake methodology and free resistance training, Robergs et al. estimated energy expenditure for the bench-press exercise in which subjects lifted at 5-23% of a 1RM over a period of 5 minutes (8). A limited comparison of the methods of Robergs et al. (8) and Scott et al.'s nonsteady state methodology (11) reveals greater than previously recorded estimates of energy expenditure estimate of energy expenditure for a single bout of bench press work (11).
In the present study, our intent was to again use nonsteady state measurements to estimate the aerobic, anaerobic, and recovery energy expenditure components of an intense bout of resistance training over a much broader range of lifting intensities. Six workloads were separately carried out for 1-set of the bench-press exercise, ranging from 37 to 90% of a 1RM, with all sets being performed to muscular fatigue (failure). These work bouts cover lifting intensities associated with both muscular endurance and strength training. It was hypothesized that anaerobic and recovery energy expenditure would make a significantly larger contribution to the energy expenditure of a single bout of resistance training as compared with exercise O2 uptake alone.
Experimental Approach to the Problem
Most investigations into the estimation of energy expenditure for resistance training only use O2 uptake measurements during the actual lift. A justification therefore exists to quantify all energy expenditure components-anaerobic and aerobic, exercise and recovery-to determine their potential significance. Because the data collection methods used here are rather new, our investigation takes the initial step of addressing only the estimation of energy expenditure for a single set of lifting. Future studies will need to determine if nonsteady state estimates of energy expenditure can, with validity, be used to examine a “typical” resistance training protocol that employs several sets.
This investigation was approved by the human subject Institutional Review Board at the University of Southern Maine. Thirteen male volunteers were informed of the experimental risks and signed an informed consent document before the investigation began: age (years) 23.8 ± 2.1, height (cm) 178.7 ± 6.6, and weight (kg) 85.9 ± 11.3, 1RM (kg) 102.5 ± 20.8. All subjects had a history of resistance training (i.e., resistance training 3 times per week for at least 3 months). Weight lifting characteristics are displayed in Table 1.
Subjects reported to the laboratory for 7 separate visits, only one lift to fatigue was performed on a given day. Subjects were asked to not exercise on the day of testing and to have fasted for at least 4 hours before testing. On the first visit, a 1RM for the bench press was recorded on a Smith machine consisting of a horizontal bar that slides on vertical tracks where weight can only be lifted in the vertical plane (York Barbell Co., York, PA, USA). Weight was gradually increased until a single repetition could not be completed. Subjects warmed up with a light weight of their choice before attempting the 1RM. During each attempt, good form was stressed and 5-point contact was maintained with the bench and floor. The tester chose the weight increase for each lift, and adequate rest (3-5 minutes) between attempts was given. Each 1RM was attained within 3-4 lifts. To minimize fluctuations in power output, subjects also practiced lifting and lowering the bar at a cadence, set by a metronome, of 1.5 seconds up and 1.5 seconds down (this was however a study limitation in that such a cadence cannot be maintained to fatigue). A small fly wheel attached to a microprocessor was connected to a moving cable on the Smith machine that recorded the distance the bar traveled (see ). Work (J) was recorded as the product of weight lifted and distance the bar traveled (6). During the following 6 visits to the laboratory, subjects were randomly assigned to bench pressing 1 of 6 lifts; 36.4 kg (80 lb), 45.5 kg (100 lb), 54.5 kg (120 lb), and 70, 80, and 90% of their 1RM. All lifts were performed until muscular fatigue (failure), when the subject could not raise the bar back to the starting position without a spotter's assistance. Each day of lifting consisted of the following measurements: 5 minutes, resting, supine energy expenditure (liters of O2 uptake per minute); resting blood lactate (mmol); exercise O2 uptake (kJ); peak recovery blood lactate (mmol); recovery O2 uptake (excess postexercise O2 consumption, EPOC, kJ) and work (J).
Oxygen uptake was measured using a standard metabolic cart (MMS-2400, PavoMedics, Sandy, UT, USA). The metabolic cart was calibrated a minimum of 2 times immediately before all testing, using room air and calibration gas (16% O2, 4% CO2); ventilation was calibrated using a 3-L syringe. Oxygen uptake was measured in 15-second sampling periods (milliliters per minute O2 uptake measures were converted into 15-sec liters of O2 uptake measures, then summed to cover the exercise and EPOC time periods). Before each lift, resting O2 uptake was averaged over a 5-minute period with each subject lying supine with their back on the bench (feet on floor). At the end of the 5-minute rest, each subject began lifting at the required cadence while O2 uptake continued to be measured throughout the exercise period. Aerobic exercise energy expenditure was estimated at 1 L O2 = 21.1 kJ. After the lifts were completed and the weight was racked, each subject had their feet elevated on a stool parallel to the height of the bench; EPOC was recorded until 2 subsequent measurements fell below 5.0 ml·kg−1·min−1 (a typical standing, resting O2 uptake). Excess postexercise O2 consumption was converted to energy expenditure as 1 L of O2 = 19.6 kJ to dismiss any glycolytic component from the O2 uptake measurement (9-11). Each subject's resting O2 uptake was subtracted during the course of exercise and EPOC.
All blood-lactate measurements were recorded in triplicate using 3 handheld lactate analyzers (Lactate Pro, Arkray, Inc., Kyoto, Japan). For data analysis, the 2 closest lactate measurements were averaged (if no outlier was present, all 3 measures were averaged). Resting blood lactate was collected with subjects lying supine, before the resting O2 uptake measurement. Blood lactate for the exercise periods was recorded in the supine position at 2 and 4 minutes postexercise. Peak blood lactate was taken as the highest blood-lactate concentration recorded at either 2 or 4 minutes postexercise. Anaerobic exercise energy expenditure was calculated as the difference between resting and peak lactate values multiplied by body weight (kg), then by 3.0 ml of O2 (4). This O2 equivalent estimate was converted to Joules as 1 L O2 = 21.1 kJ. Total energy expenditure was recorded as the sum of aerobic and anaerobic exercise energy expenditures and EPOC (10).
Data analysis was examined for differences using standard t-test (single comparisons) and ANOVA (group comparisons) with the appropriate post hoc test (SigmaStat 3.5; Point Richmond, CA, USA). Pearson correlations and regression also were performed. Alpha level was set at p ≤ 0.05. Statistical power was calculated with a group size of 3, subject number of 13, and an alpha level of 0.05.
The total energy expenditure (TEE: aerobic and anaerobic exercise + recovery-energy expenditures) to work relationship (regression) is provided in Figure 1 where r = 0.87 (p = 0.001) and TEE = 19.751 + (0.0893 × Work). Correlation between overall anaerobic energy expenditure and work was r = 0.79 (p < 0.0001), between overall aerobic energy expenditure and work r = 0.87 (p < 0.0001). The muscular endurance lifts resulted in larger total energy expenditure (60.2 ± 14.5 kJ) as compared with the strength lifts (43.2 ± 12.5 kJ, p = 0.001); work also was greater for muscular endurance (462 ± 131 J) as opposed to strength (253 ± 93 J, p = 0.001).
Lifting characteristics are displayed in Table 1. As a general statement, the lighter the weight (as a percentage of 1RM), the more repetitions were completed until fatigue and the more work that was completed. Correlations for TEE and each workload were significant for all but the highest load (90% 1RM) that had the lowest number of repetitions (4.5 ± 1.7, p = 0.09).
Comparisons within the strength-type exercises (70, 80, and 90% of 1RM, lifts to muscular fatigue; Table 2): statistical analysis (ANOVA) of energy expenditure for anaerobic (Δ lactate), aerobic (exercise O2), and EPOC revealed significant differences between anaerobic and EPOC energy expenditure as compared with aerobic (exercise) energy expenditure (p < 0.001). Anaerobic energy expenditure was larger for 70 and 80% as compared with 90% of 1RM (p < 0.006). No differences were found among the 3 strength-type lifts for aerobic and EPOC energy expenditures.
Comparisons within the muscular endurance exercises (37, 46, and 55% of 1RM, lifts to muscular fatigue; Table 2): ANOVA revealed that anaerobic energy expenditure was significantly greater than EPOC (p < 0.001), with EPOC and anaerobic energy expenditure significantly greater than aerobic energy expenditure (p < 0.001). No differences were found among the 3 muscular endurance lifts for anaerobic, aerobic, and EPOC energy expenditures.
Comparisons between strength and muscular endurance (Table 3) indicated that the anaerobic energy expenditure of all muscular endurance exercises were significantly larger than the strength-type lifts at 80 and 90% of 1RM (p < 0.05). Aerobic energy expenditure was larger for all muscular endurance lifts when compared with 90% of 1RM and, for the lightest muscular endurance lift (36.4 kg) when compared with 70 and 80% of 1 RM (p < 0.05). No differences between strength EPOC and muscular endurance EPOC were found.
Our data reveal the significant contributions of anaerobic and recovery energy expenditure (i.e., EPOC) over a series of single-set bench press lifts to fatigue. Statistical analysis indicated that the aerobic energy expenditure (O2 uptake) during lifting was always the lowest whether strength or muscular endurance routines were scrutinized (Table 2). There were no differences between anaerobic and EPOC energy expenditure among the strength training protocols (70, 80, and 90% of 1RM). However, with the muscular endurance protocols (37, 46, and 56% of 1RM), the anaerobic energy expenditure contribution was significantly larger than that of EPOC. These data emphasize that the sole measure of nonsteady state O2 uptake during resistance training likely underestimates the true energy expenditure of a single set of lifting to fatigue.
When a single set of resistance training is performed until muscular fatigue, the lesser weight lifted should result in the greatest number of repetitions, with heavier weights allowing for the least number of repetitions. This trend is indicated in Table 1. Along similar reasoning, the largest energy expenditure may be expected to result from the greatest amount of absolute work performed, with the least amount of work being associated with the lowest energy expenditure. This finding also was apparent in the current investigation albeit only generally so in that correlations between total energy expenditure and the work completed for each lifting intensity ranged from 0.49 to 0.85, perhaps reflective of an earlier finding revealing lower measurement reliability with brief sets (11) (at 90% of 1RM, work did not achieve statistical significance in a correlation with total energy expenditure [Table 1]). Figure 1 reveals a good overall relationship between total energy expenditure and work for all lifts (r = 0.87, p < 0.001). Correlations between overall work and anaerobic exercise energy expenditure (r = 0.79) and aerobic exercise energy expenditure (r = 0.87) also were apparent (p < 0.0001).
When data from the separate protocols were compiled and examined, lifting to muscular fatigue resulted in greater energy expenditure for 1-set of muscular endurance-type lifting (60.2 ± 14.5 kJ) as compared with 1-set of the strength-type lifts (43.2 ± 12.5 kJ, p = 0.001). Work also was greater for muscular endurance (462 ± 131 J) as opposed to strength (253 ± 93 J, p = 0.001). In application, the use of exercise in the promotion of weight loss would be one where the greatest amount of energy is expended. Our results suggest that the best resistance training program to promote energy expenditure might be lifting at a lighter percentage of a 1RM to fatigue. Whether resistance training protocols that employ 20-40 repetitions to muscular failure will be accepted by those who want or need to lose weight remains to be seen. Moreover, multiple (large) muscle group exercises as opposed to isolation exercises with smaller muscles would likely need to be chosen when designing a weight loss program that focuses on resistance training (e.g., bench press vs. triceps extension; leg press vs. leg extension).
Decades ago, it was incorrectly assumed that blood lactate concentrations were related to the amount of O2 uptake consumed in the recovery from aerobic exercise (i.e., the O2 debt hypothesis) (5). Is this also true for anaerobic-type exercise? Our estimates of anaerobic energy expenditure were based exclusively on blood-lactate measures where the anaerobic energy expenditure contributions differed among strength and muscular endurance protocols. Yet surprisingly, EPOC energy expenditure did not differ among muscular endurance and strength formats. It is therefore apparent that EPOC after a single bout of resistance training is not dependent on lactate concentrations as aerobic studies have previously concluded (5). In fact, correlation between anaerobic energy expenditure and EPOC for our study was not only poor it also was nonsignificant (r = 0.22; p = 0.06).
With aerobic exercise training, EPOC volume (and by inference recovery energy expenditure) appears to be linearly related to duration and exponentially related to intensity (3) (importantly, at relatively low to moderate aerobic exercise intensities as compared with higher intensities, fatigue may not be an end point). In this regard, our EPOC data for resistance training do not appear to follow an aerobic format (where fatigue was the end result, regardless of percentage of 1RM). More research has been called for concerning anaerobic-type exercise and EPOC (7) and our current findings, where EPOC was similar among all trials, support this call. Indeed, after a single bout of resistance training to fatigue, EPOC appeared to reach a limit or a plateau regardless of the previous amount of prior work. We found a poor relationship when the amount of work and EPOC energy expenditure were compared (r = 0.35, p = 0.002). A poor relationship also was found with EPOC and aerobic energy expenditure (r = 0.27, p = 0.02). We are aware of a few comparative studies where mice sprinting over brief periods at varying durations and intensities revealed similar EPOCs after several various trials (1,2). Our EPOC data appear to mimic these comparative findings but exactly why this is so remains unclear. Clearly, more research is needed to investigate this phenomenon.
In summary, anaerobic and EPOC energy expenditures exceed that of aerobic energy expenditure during 1-set of strength-type and endurance-type lifts to fatigue. Our data indicate that a single set of muscular endurance-type lifting to fatigue expends more total energy than a single set of strength-type lifting to fatigue; this apparently is related to the amount of work performed within the set. Excess postexercise O2 consumption energy expenditure had poor relationships with aerobic and anaerobic exercise energy expenditure and work.
When weight loss concerns dominate the need for an exercise prescription, the results of the current study help to answer to the question: What type of lifting “burns” the most calories, muscular endurance, or strength-type training? The answer appears to be muscular endurance-type exercise (when a greater volume of work is performed).
1. Baker, EJ and Gleeson, TT. EPOC and the energetic of brief locomoter activity in Mus domesticus
. J Exp Zool
280: 114-120, 1998.
2. Baker, EJ and Gleeson, TT. The effects of intensity on the energetic of brief locomoter activity. J Exp Biol
202: 3081-3087, 1999.
3. Borsheim, E and Bahr, V. Effect of exercise intensity, duration and mode on post-exercise oxygen consumption. Acta Phys Scand
33: 1037-1060, 2003.
4. di Prampero, PE and Ferretti, G. The energetic of anaerobic muscle metabolism: A reappraisal of older and more recent concepts. Respir Physiol
118: 855-866, 1999.
5. Gaesser, GA and Brooks, GA. Metabolic bases of excess post-exercise oxygen consumption: A review. Med Sci Sport Exerc
16: 29-43, 1984.
6. McBride, JM, McCaulley, GO, Cormie, P, Nuzzo, JL, Cavill, MJ, and Triplett, NT. Comparison of methods to quantify volume during resistance exercise. J Strength Cond Res
23: 106-110, 2009.
7. Meirelles, C, de-M, and Gomes, PSC. Acute effects of resistance exercise on energy expenditure: Revisiting the impact of the training variables. Rev Bras Med Esporte
10: 131-138, 2004.
8. Robergs, RA, Gordon, T, Reynolds, J, and Walker, TB. Energy expenditure during the bench press and squat exercises. J Strength Cond Res
21: 123-130, 2007.
9. Scott, CB. Contributions of blood lactate to the energy expenditure of resistance training. J Strength Cond Res
20: 404-411, 2006.
10. Scott, CB. A Primer for the Exercise and Nutrition Sciences: Thermodynamics, Bioenergetics, Metabolism
. Totowa, NJ: Humana Press, 2008.
11. Scott, CB, Croteau, A, and Ravlo, T. Energy expenditure before, during and after the bench press. J Strength Cond Res
23: 611-618, 2009.
12. Tamaki, T, Uchiyama, S, Tamura, T, and Nakano, S. Changes in muscle oxygenation during resistance training exercise. Eur J Appl Physiol
68: 465-469, 1994.
13. Tesch, PA, Colliander, EB, and Kaiser, P. Muscle metabolism during intense, heavy resistance exercise. Eur J Appl Physiol
55: 362-366, 1986.