Resistance exercise is recommended by the American College of Sports Medicine (ACSM), the American Heart Association (AHA), and the National Strength and Conditioning Association (NSCA) (6,8,12). The beneficial effects of resistance exercise can include increased muscular strength and endurance (19), increased HDL cholesterol (15), decreased LDL cholesterol (11,15), decreased triglycerides (11), decreased blood pressure (18), improved insulin sensitivity (7), and increased bone mineral density (9). In addition, because resistance exercise increases daily energy expenditure, it may be valuable during weight loss or weight loss maintenance programs when combined with caloric restriction (20) and endurance exercise (1). Increased total daily energy expenditure after resistance exercise (33) could result from increased lean body mass (19), increased metabolic rate (30), increased exercise recovery energy expenditure (4), and spontaneous increases in daily physical activity (10).
Because of the exponential increase in obesity in the United States during the past 15 yr (24) and the subsequent effect on human health (26), it seems prudent to recommend exercise programs that expend the greatest amount of energy. The ability to accurately assess energy expenditure is critical in the design of clinical trials and is necessary to develop successful exercise prescriptions. However, the measurement of resistance exercise energy expenditure can be difficult or costly or can restrict research volunteers to exercises not used in typical resistance exercise protocols. For instance, metabolic chambers are costly and may alter exercise choice because of space or weight limitations. Portable indirect calorimeters have been used to assess resistance exercise energy expenditure; however, this technique is expensive, and the equipment must be repositioned on the basis of the exercise (e.g., prone, supine, barbell location).
Recently, the use of accelerometers to objectively measure physical activity and estimate physical activity energy expenditure has grown in popularity (31). Accelerometers assess "counts" of physical activity and eliminate the potential inaccuracy associated with self-report of physical activity (32). Counts of physical activity obtained from accelerometers can be used to estimate exercise intensity and energy expenditure. In addition, accelerometers are lightweight and small, which may allow for a more accurate and less cumbersome assessment of resistance exercise energy expenditure compared with indirect calorimetry. Accelerometers can be worn on the extremities to assess limb activity, although it is currently unknown what anatomical site provides the best estimate of overall energy expenditure during resistance exercise. In addition, accelerometers cost significantly less than portable metabolic systems, making them more cost effective.
Accelerometers are a valid means of assessing physical activity during behaviors such as walking, running, or activities of daily living (31). However, the use of accelerometers to determine resistance exercise energy expenditure has not been studied. Relative to walking and running activities, resistance exercise is unique because of the number of exercises performed seated or lying down, the rhythmic yet intermittent nature of resistance exercise, the duty cycles that vary based on training protocol, the lack of movement of some limbs during certain exercises (e.g., the legs during the bench press), and the extreme differences in training load performed at similar accelerations. It is unlikely that regression equations based on walking, jogging, or activities of daily living will accurately predict the energy expended during resistance exercise. In addition, the appropriate location to wear the accelerometer (e.g., wrist, waist, and ankle) has not been studied with resistance exercise.
We are unaware of any studies that have investigated resistance exercise energy expenditure using accelerometry. The purpose of this study was 1) to estimate resistance exercise energy expenditure using accelerometry and 2) to determine whether there are differences in counts of activity during resistance exercise on the basis of accelerometer location.
Testing was conducted in the morning after an overnight fast. Participants refrained from caffeine ingestion and exercise for 12 h before testing and refrained from alcohol ingestion for 24 h before testing. Age, height, weight, skinfold thickness, circumference measurements, limb lengths, and strength (10-repetition maximum [10RM]) were assessed on the first visit. The 10RM testing was conducted using standardized procedures (2) for the following exercises: Smith machine bench press, Smith machine shoulder press, Smith machine squat, leg extension, leg curl, lat pulldown, triceps push-down, and barbell bicep curl. Briefly, participants were instructed in the proper technique for each exercise, completed an exercise-specific warm-up of 10 repetitions with a very light weight, and completed three to five additional sets of 10 repetitions with progressive resistance (approximately 5% for upper body exercises and 10% for lower body exercises) and 2-min rest between sets. The 10RM was achieved when a load allowing only 10 repetitions was reached. During the second visit, energy expenditure during a resistance exercise protocol was assessed using a CosMed K4b2 portable metabolic system while wearing accelerometers on the wrist, waist, and ankle.
A sample of 30 (15 males, 15 females) healthy college-aged individuals ranging in age from 18 to 30 yr were recruited for this study. All volunteers were recruited from the Bloomsburg area. The study was approved by the Human Subject's Committee of the institutional review board at Bloomsburg University. Before participating in this investigation, participants read and signed an institutionally approved informed consent document. A Physical Activity Readiness Questionnaire was used to exclude participants with known cardiovascular or musculoskeletal problems. Participants engaged in recreational resistance training but were not highly trained.
Resistance exercise protocol
The resistance exercise protocol was based on the ACSM recommendations for resistance exercise to achieve good health (29). To increase muscular strength/endurance, lean body mass, and bone density, one set of 8-12 repetitions of 8-10 resistance exercises involving major muscle groups (i.e., arms, shoulders, abdomen, back, hips, and legs) on 2-3 d·wk−1 is recommended, with multiple sets potentially offering greater improvements (29).
In the current study, participants warmed-up on a cycle ergometer for 5 min at 60% of their age-predicted maximum heart rate. After the warm-up, three accelerometers were randomized to the volunteer's wrist, waist, and ankle. Once the devices were activated, participants performed two sets of 8-10 repetitions of a group of exercises at a 10RM intensity. All participants performed the following exercises: Smith machine bench press, Smith machine shoulder press, Smith machine squat, leg extension, leg curl, lat pulldown, triceps push-down, and barbell bicep curl. There was a seated 1-min rest between sets and a seated 2-min rest between exercises. Participants were instructed to remain as still as possible and to minimize movement between exercises and sets. During the workout, the investigators carried a chair to each exercise station for the volunteers to sit down. This was meant to minimize between-exercise movements. Exercise cadence was controlled with a metronome set at 60 bpm. After the exercise protocol, there was a 2-min walking cooldown period. Lifting volume for each exercise was calculated as follows: lifting volume = ((load × set × reps)1 + (load × set × reps)2). The calculated lifting volume for each exercise was then summed to calculate a total lifting volume score.
Participants were fitted with the CosMed K4b2 portable metabolic system. The CosMed K4b2 system was calibrated before each use according to the manufacturer's guidelines, and a gas calibration was conducted to ensure that O2 and CO2 readings were in normal ranges. Turbine and delay calibrations were conducted to ensure that the calorimeter was collecting data in sync with participant breathing patterns. The unit was attached to the participants' chest with a harness, and a Hans Rudolph face mask (Kansas City, MO) was attached using a head strap. Once fitted with the CosMed K4b2 portable metabolic system, participants were placed in a comfortable chair for 20 min for the estimate of resting metabolic rate.
Participants wore three uniaxial ActiGraph GT1M (Pensacola, FL) accelerometers located on the right wrist (between the ulnar and the radial styloid processes), right side of waist (on the hip), and right ankle (superior to the lateral malleolus). An accelerometer, such as the model used in the current study, records integrated acceleration information as an activity count, which can be subsequently used to estimate intensity of body movement and energy expenditure. The epoch period was set for 1 s. Data were downloaded to a computer with the ActiLife computer software program. The summation of counts was determined from the beginning of the first set to the end of the last set of the resistance exercise session using Microsoft Excel.
Anthropometry and body composition.
Height was determined using a wall-mounted stadiometer, and body mass was recorded with a calibrated balance scale. Circumference measurements were taken with a Gulick II anthropometric tape measure (Country Technologies, Inc., Gays Mills, WI) at the neck, shoulder, chest, waist, abdomen, hip, thigh, calf, ankle, arm, wrist, and forearm (13). Upper and lower body limb lengths were obtained for each participant according to procedures described by Lohman et al. (22). Participants stood with arms relaxed at the sides and palms rotated medially, and the lengths from the acromion to the dactylion and from the greater trochanter to the base of the calcaneus were recorded using a Gulick II anthropometric tape measure. Skinfold thickness was assessed at three sites (16,17) with Lange calipers (Beta Technologies, Cambridge, MA). The skinfold sites assessed were the triceps, the suprailiac, and the thigh for women and the chest, the abdomen, and the thigh for men (13). Each site was tested three times with the average of the three trials used to determine body density. Body density was converted to body composition using the Siri equation (13).
Statistical analyses were performed using SPSS (Chicago, IL) and Sigma Stat (Richmond, CA). Means and standard deviations were calculated, an independent t-test was used to assess differences in baseline data between sexes, an ANOVA was used to compare wrist, waist, and ankle accelerometer counts, and a Tukey's post hoc analysis was conducted to locate significant differences. The association between accelerometer counts and energy expenditure was assessed with Pearson product moment correlation coefficients. Forward stepwise regression analysis was used to develop an equation to predict resistance exercise energy expenditure from accelerometer counts. Statistical significance was set a priori at P ≤ 0.05.
Men were significantly taller and had a higher body mass and more fat-free mass, whereas the women had a significantly higher body fat percentage. Descriptive characteristics of the participants (mean ± SD) are presented in Table 1. Men had significantly larger neck, shoulder, chest, waist, abdomen, ankle, arm, and forearm circumferences and longer upper body limb lengths (Table 2). There was a significant difference between sexes in total lifting volume and 10RM strength for each exercise (all P values <0.001; Table 3). Resting metabolic rate, gross energy expenditure, and net energy expenditure were significantly higher in men than women (all P values < 0.001; Table 4). The duration of the resistance exercise protocol was ≈31.7 min, with ≈7.2 min of exercise time.
Accelerometry and energy expenditure.
An ANOVA indicated a significant difference between accelerometer counts by location, and a post hoc analysis revealed that counts of activity were different between the wrist (61,282 ± 8358), the ankle (26,886 ± 3998), and the waist (6565 ± 2445; Fig. 1). Net energy expenditure was significantly associated with ankle (r = 0.50; P < 0.01) and waist (r = 0.77; P < 0.001) accelerometer counts. There was a trend for an association between net energy expenditure and wrist accelerometer counts (r = 0.31; P = 0.10; Table 5). Figures 2-4 depict the relationship between net energy expenditure and accelerometer counts. Total waist accelerometer counts explained 59% of the variance (R2 = 0.59) in net energy expenditure. In addition, most anthropometric (e.g., height, body mass, fat-free mass, circumferences, limb lengths) and resistance exercise (e.g., 10RM, training volume) variables correlated with net energy expenditure.
A forward stepwise regression analysis was conducted to develop an equation to predict resistance exercise energy expenditure on the basis of accelerometer counts, anthropometric (e.g., height, body mass, fat-free mass, circumferences, limb lengths), and resistance exercise (e.g., 10RM, training volume) variables. The following variables were used in this analysis: waist, ankle, and wrist accelerometer counts; sex, height, body mass, body mass index, body fat percentage, fat-free mass, circumference of the neck, shoulder, chest, waist, abdomen, hip, thigh, ankle, calf, ankle, arm, wrist, and forearm; upper and lower body limb length; and total lifting volume. Fat-free mass explained the greatest percentage of the variance in the estimate of resistance exercise energy expenditure (R2 = 0.85). The addition of sex and waist accelerometer counts increased the final R2 for the estimate of resistance exercise energy expenditure to 0.90. The addition of other variables, including counts from ankle and wrist accelerometers, did not improve our ability to predict net energy expenditure.
The following equation was developed to calculate net energy expenditure during resistance exercise:
Note that sex indicates 1 for females and 2 for males; FFMkg is the fat-free mass in kilograms; and counts is the total counts from the accelerometer worn on the waist.
In the current study, we demonstrated that resistance exercise energy expenditure can be estimated from fat-free mass, sex, and counts of activity from a uniaxial accelerometer worn at the waist. The purpose of this study was 1) to estimate resistance exercise energy expenditure using accelerometry and 2) to determine whether there are differences in counts of activity during resistance exercise on the basis of accelerometer location. As previously reported (21,27), we found that anthropometric and training load variables correlated with net energy expenditure. Unique to this study is the finding that waist accelerometer counts were highly correlated with (r = 0.77) and improved our ability to predict resistance exercise energy expenditure. In addition, we determined that accelerometer counts varied considerably between wrist (61,282), ankle (26,886), and waist (6565) accelerometers.
Resistance exercise is recommended by the leading sports medicine and health organizations (ACSM, AHA, and NSCA) (6,8,12). The benefits of resistance extend beyond the well-known improvements in muscular (19) and bone health (9) and can include improvements in various cardiovascular risk and metabolic health biomarkers (7,11,15,18). Resistance exercise expends fewer kilocalories than continuous endurance exercise when matched for total time and relative intensity (cycling 70% V˙O2max = 441 kcal; free-weight squatting 70% 1RM = 269 kcal) (3). However, when combined with caloric restriction (20) and endurance exercise (1), resistance exercise is a valuable addition to weight loss or weight-loss maintenance programs. Resistance exercise may alter physical activity behaviors as well and is known to cause spontaneous increases in daily physical activity in older adults (10). Accelerometers are increasingly being used to estimate the energy cost of rhythmic exercises (e.g., walking, jogging) and activities of daily living (e.g., yard work, vacuuming) but not resistance exercise.
There are several strengths of the current study. First, we assessed the energy expenditure associated with a standardized resistance exercise protocol (8 exercises, 10 repetitions, 2 sets; cadence = 60 bpm; 1-min rest between sets and 2-min rest between exercises) that is recommended for good health (29). In addition, participant movement between sets and exercises was minimized. Second, we assessed multiple anthropometric (e.g., height, body mass, circumferences, body composition, limb lengths) and exercise-related (e.g., 10RM, training load/lifting volume) variables, which correlated with net energy expenditure. We also collected data from three accelerometers (wrist, ankle, and waist) during resistance exercise. However, with the exception of fat-free mass and sex, none of these measures improved our ability to predict energy expenditure. Our data demonstrate that resistance exercise energy expenditure can be estimated in young adults using waist accelerometer counts, fat-free mass, and sex. Finally, the resistance exercise energy expenditure in the current study is similar to what has been reported in previous studies (25,27).
Several regression equations to predict the energy cost of physical activity have been validated. However, these equations were not based on resistance exercise but on walking, running, and activities of daily living. Resistance exercise presents a challenge because many exercises are performed seated or lying down, some limbs are often inactive while others are moving (e.g., the legs during the bench press), and the majority of time spent during a resistance exercise session is often at rest (i.e., between sets). In addition, unlike continuous endurance exercise, resistance exercise programs may have wide variations in rest periods (zero to several minutes), training load (50%-95% of 1RM), training volume (single vs multiple sets), and repetition cadence (high-velocity Olympic lifts vs slower dynamic contractions). Thus, it is difficult to know if the results of the current study can be generalized to other resistance exercise protocols.
It cannot be known why the accelerometer worn at the waist, relative to the accelerometers worn at the ankle and wrist, was most highly correlated with energy expenditure. It could be speculated that this resulted from the nature of the workout (i.e., total body workout, three lower body exercises, five upper body exercises). Possibly, a radically different workout protocol (e.g., only upper body) would yield dissimilar results. In addition, it has been shown that differences in repetition velocity can influence resistance exercise energy expenditure (14,23), which may influence the relationship between accelerometer counts and energy expenditure. One could speculate that the high correlation between counts from an accelerometer worn at the waist would remain regardless of repetition speeds, provided the exercise protocol was balanced and included both upper and lower body exercises.
Although we were able to explain 90% of the variance in resistance exercise energy expenditure, several methodological and analytical factors should be considered in future studies. As resistance exercise takes place in several planes, the use of triaxial accelerometers may improve our ability to estimate resistance exercise energy expenditure. Although not a resistance exercise study, Pober et al. (28) analyzed accelerometer data using quadratic discriminant analysis and hidden Markov models and reported more accurate estimates of time spent in different exercise intensities relative to the traditional cut point approach. Chang et al. (5) successfully analyzed accelerometer data using naive Bayes classifiers and hidden Markov models to recognize resistance exercise type and count repetitions. We must also acknowledge the fact that although metabolic collection was continuous, accelerometers were essentially not collecting data as the subjects were sitting still between sets. Given the greater popularity of traditional resistance training that includes between-exercise/set rest periods relative to circuit training where there are no between-exercise/set rest periods, we decided to keep between-exercise/set movement minimal. Future studies will have to address this issue if resistance exercise protocols of a more continuous nature are used. Thus, although counts from a uniaxial accelerometer worn at the waist combined with fat-free mass and sex provided an accurate estimate of resistance exercise energy expenditure, it is possible that a triaxial accelerometer or other analytical techniques could improve predictability.
On the basis of the findings and within the limitations of this investigation, we reported that resistance exercise energy expenditure can be estimated using a uniaxial accelerometer worn at the waist and by the sex and fat-free mass of the participant. As there was large variability in ActiGraph counts (wrist = 48,644-84,352; waist = 3165-11,860; and ankle = 12,625-32,835) in this relatively homogeneous population, future investigations should contrast different population samples such as older and younger age groups, elite athletes and untrained individuals, and healthy versus patient populations. In addition, resistance exercise protocols that involve explosive exercises such as the Olympic lifts should be tested. Finally, the use of different analytical techniques such as naive Bayes classifiers and hidden Markov models to recognize exercise type or triaxial accelerometers to track body posture should be explored.
The authors would like to thank the dedicated volunteers for their participation. The authors have no conflicts of interest to report, and the results of the present study do not constitute endorsement by the ACSM.
1. Arciero PJ, Gentile CL, Martin-Pressman R, et al. Increased dietary protein and combined high intensity aerobic and resistance exercise improves body fat distribution and cardiovascular risk factors. Int J Sport Nutr Exerc Metab
2. Baechle TB, Earle RW. Essentials of Strength and Conditioning
. 3rd ed. Champaign (IL): Human Kinetics; 2008. p. 250-92.
3. Bloomer RJ. Energy cost of moderate-duration resistance and aerobic exercise. J Strength Cond Res
4. Burleson MA Jr, O'Bryant HS, Stone MH, Collins MA, Triplett-McBride T. Effect of weight training exercise and treadmill exercise on post-exercise oxygen consumption. Med Sci Sports Exerc
5. Chang K, Chen MY, Canny J. Tracking free-weight exercises. In: Proceedings of the UbiComp 2007: Ubiquitous Computers
. 2007, Berlin, Springer, p. 19-37.
6. Conley MS, Rozenek R. National Strength and Conditioning Association Position Statement: health aspects of resistance exercise and training. Strength Cond J
7. DiPietro L, Dziura J, Yeckel CW, Neufer PD. Exercise and improved insulin sensitivity in older women: evidence of the enduring benefits of higher intensity training. J Appl Physiol
8. Donnelly JE, Blair SN, Jakicic JM, Manore MM, Rankin JW, Smith BK. American College of Sports Medicine Position Stand. Appropriate physical activity
intervention strategies for weight loss and prevention of weight regain for adults. Med Sci Sports Exerc
9. Dornemann TM, McMurray RG, Renner JB, Anderson JJ. Effects of high-intensity resistance exercise on bone mineral density and muscle strength of 40-50-year-old women. J Sports Med Phys Fitness
10. Fiatarone MA, O'Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med
11. Goldberg L, Elliot DL, Schutz RW, Kloster FE. Changes in lipid and lipoprotein levels after weight training. JAMA
12. Haskell WL, Lee IM, Pate RR, et al. Physical activity
and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation
13. Heyward VH, Wagner DR. Applied Body Composition Assessment
. 2nd ed. Champaign (IL): Human Kinetics; 2004. p. 49-82.
14. Hunter GR, Seelhorst D, Snyder S. Comparison of metabolic and heart rate responses to super slow vs. traditional resistance training. J Strength Cond Res
15. Hurley BF, Hagberg JM, Goldberg AP, et al. Resistive training can reduce coronary risk factors without altering V˙O2max
or percent body fat. Med Sci Sports Exerc
16. Jackson AS, Pollock ML. Generalized equations for predicting body density of men. Br J Nutr
17. Jackson AS, Pollock ML, Ward A. Generalized equations for predicting body density of women. Med Sci Sports Exerc
18. Kelley GA, Kelley KS. Progressive resistance exercise and resting blood pressure: a meta-analysis of randomized controlled trials. Hypertension
19. Kraemer WJ, Adams K, Cafarelli E, et al. American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Med Sci Sports Exerc
20. Kraemer WJ, Volek JS, Clark KL, et al. Influence of exercise training on physiological and performance changes with weight loss in men. Med Sci Sports Exerc
21. Kuehl K, Elliot DL, Goldberg L. Predicting caloric expenditure during multi-station resistance exercise. J Appl Sport Sci Res
22. Lohman TG, Roche AF, Martorell R. Anthropometric Standardization Reference Manual
. Champaign (IL): Human Kinetics; 1988. p. 9-26.
23. Mazzetti S, Douglass M, Yocum A, Harber M. Effect of explosive versus slow contractions and exercise intensity on energy expenditure. Med Sci Sports Exerc
24. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA
25. Morgan B, Woodruff SJ, Tiidus PM. Aerobic energy expenditure during recreational weight training in females and males. J Sports Sciences Med
26. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. JAMA
27. Phillips WT, Ziuraitis JR. Energy cost of the ACSM single-set resistance training protocol. J Strength Cond Res
28. Pober DM, Staudenmayer J, Raphael C, Freedson PS. Development of novel techniques to classify physical activity
mode using accelerometers. Med Sci Sports Exerc
29. Pollock ML, Gaesser GA, Butcher JD, et al. American College of Sports Medicine Position Stand. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med Sci Sports Exerc
30. Pratley R, Nicklas B, Rubin M, et al. Strength training
increases resting metabolic rate and norepinephrine levels in healthy 50- to 65-yr-old men. J Appl Physiol
31. Troiano RP. Translating accelerometer
counts into energy expenditure: advancing the quest. J Appl Physiol
32. Tudor-Locke CE, Myers AM. Challenges and opportunities for measuring physical activity
in sedentary adults. Sports Med
33. Van Etten LM, Westerterp KR, Verstappen FT, Boon BJ, Saris WH. Effect of an 18-wk weight-training program on energy expenditure and physical activity
. J Appl Physiol