In several free-weight exercises (e.g., squat), there is a mismatch throughout the range of motion (ROM) between the torque created by the weights and the muscles' ability to produce torque (10,13,18). In the upper part of the movement, the torque-generating capacity of the muscle is greater than the torque caused by the lever arm of the external load. Consequently, if the movement is not performed explosively, the force output will be reduced toward the end of the concentric phase (13). Therefore, the training stimuli may not be optimal during the free-weight squat, and there could be a potential for greater performance improvements.
Variable resistance training can in theory overcome the possible weakness of constant resistance training and can be defined as resistance training where the resistive force alters during the movement (13). Variable resistance can be performed by using machines with cam-based systems (19) or by adding chains or elastic bands to the barbell (6,21). In the back squat, elastic bands will offer an external ascending resistance curve that can enhance the training stimuli and muscle activation throughout the ROM (25). However, the results are inconclusive. Some studies have reported increased muscle activation and force production throughout the ROM using variable resistance training (1,27), whereas other studies observed no differences (2,3,9). It was also recently reported that variable resistance training elicited higher acute hormonal responses than constant resistance training (27), which over time could lead to greater muscle hypertrophy and strength development.
Most intervention studies comparing variable and constant resistance training regimes have reported similar increases in dynamic strength (8,16,21,22) or isometric strength (7,19,26). However, Melo Mde et al. (22) reported a greater gain in isometric knee extensor strength at near extended knee angles in response to 12-week variable knee extensor resistance training compared with constant resistance training using the same relative intensity (e.g., 6–10 repetition maximum [RM]). As the load of the variable resistance is substantially higher compared with constant resistance at near extended knee angles, it seems plausible that there could be greater potential for strength gains with variable resistance training in the study by Melo Mde et al. (22), but more research is needed to verify this.
Several studies have investigated effects of variable vs. constant resistance training on explosive performance (4,8,14,16,24). However, only 1 study observed differences. Anderson et al. (4) reported results in favor of variable resistance training—as these authors reported greater average power during countermovement jump (CMJ) after 7 weeks of back squat training with either free weights or free weights in combination with elastic bands. To the best of our knowledge, no studies have examined the effect of variable resistance on ballistic muscle performance at different phases of the ROM. Theoretically, it seems plausible that variable resistance training could be beneficial for improving jumping height in CMJs with a shorter descending phase, as greater resistance is applied at more extended knee angles in variable resistance training.
To the best of our knowledge, only 1 study (26) has assessed effects of a variable vs. a constant resistance training program on muscle activation. Walker et al. (26) recruited elderly men to a quasirandomized controlled study where they performed resistance training for 20 weeks. No significant differences were observed between groups. However, both regimes increased the agonist activation during isometric knee extensor maximal voluntary isometric contraction (MVC) assessed at a 107° knee angle, but only the constant resistance training group increased the electromyographic (EMG) amplitude significantly during concentric knee extension from pre- to postintervention.
Previous studies suffer from various limitations. Some recruited participants with no experience with resistance training (7,22,26), others have used low training volumes (8,16,19), and some performed additional training (i.e., sprint and plyometric training), which could mask potential differences between programs (8,16). Therefore, there is uncertainty concerning the effects of variable vs. constant resistance training on maximal muscle strength, neural activation, and explosive muscle strength at different phases of the ROM for persons with resistance training experience.
The primary aim of this study was to compare the effects of a variable vs. a constant resistance training program (10 weeks) on isometric muscle strength and muscle activation at different knee angles in addition to CMJ height using different descending distances. Secondarily, we wanted to assess the effect on dynamic muscle strength and activation for 6RM back squat. The constant resistance was provided with free weights, whereas the variable resistance was free weights in combination with elastic bands. We hypothesized that the variable training regime would increase muscle activation, strength, and jump height at more extended knee angles (120°) compared with the constant training regime, as the loading by the bands is substantially higher toward the end of the concentric movement. We expected similar improvements from the 2 training regimes on tests performed at more flexed knee angles (60°) and in the dynamic squat.
Experimental Approach to the Problem
A randomized controlled within- and between-groups design was used to examine the effects of a 10-week lower extremities strength training program using variable vs. constant resistance. The participants were randomly assigned to elastic band group (EBG) or free-weight group (FWG) by drawing slips of paper with the participants' number from a hat, 16 participants in each group. The strength training program consisted of 2 exercises targeting the hip and knee extensors; back squat and split. Both groups performed the exercises with the same number of sets and repetitions, using the same relative resistance. However, the FWG performed the exercises with free weights and the EBG used elastic bands in combination with free weights. Knee extensor MVC, muscle strength, and CMJs were assessed at 60, 90, and 120°, in addition to 6RM in free-weight back squat. Muscle activity was obtained by surface EMG during MVCs and back squat.
Thirty-two healthy women (range: 20–44 years old) volunteered as participants. Inclusion criteria were at least 6 months of recreational strength training experience—involving the hip and knee extensors on a weekly basis, be able to perform full back squats with proper technique. Exclusion criteria were no injuries or pain that could impair maximal performance during the tests or training. The participants had to refrain from any additional strength training of the legs during the study, but were encouraged to continue other types of training, if they performed such activities. All participants were informed (oral and written) of the procedures and possible risks associated with the study. Informed consent was obtained from the subjects before they were included in the study. The study had approval from the Sogn og Fjordane University College Review Board, and all appropriate consent pursuant to law was obtained before the start of the study.
One participant from each group withdrew during the intervention because of reasons not related to the study; hence, 15 participants for each group were used in the analysis. Anthropometric parameters are presented in Table 1.
Seven to twelve days before the pretest, a 6RM strength test was performed for the back squat as familiarization to the pretest. The warm-up protocol was identical on all test days. The warm-up protocol consisted of 10-minute general warm-up before the participants performed a specific progressive warm-up protocol in back squat consisting of 10–15 repetitions at 30–50% of 6RM and 6 repetitions at 50–70% of 6RM. The participants self-reported 6RM was used at the familiarization session, whereas the 6RM obtained at the familiarization test or the latest work out session was used at the pre- and posttest, respectively. Two- to three-minute rest was given between the sets.
On the pre- and posttest, all tests were conducted in 1 session and in the following order; 6RM in back squat, unilateral knee extensor MVC of the dominant leg, and CMJ (Figures 1A–C). The knee extension tests and CMJs were performed with knee angles of 60, 90, and 120°, where a full knee extension was defined as 180°. The order of the different angles in both the MVC and CMJ was randomized and kept identical at pre- and posttest. Each individual performed the pre- and posttest approximately at the same time of the day.
Six Repetition Maximum in Back Squat
In the first familiarization test, the lowest position of the movement was found—defined as when trochanter major aligned with the upper part of the patella (17). A horizontal rubber band was set at the same height as the lower part of the hamstrings. The participant had to touch the rubber band before starting the concentric lifting phase (Figure 1A). The position of vertical displacement and preferred distance between the feet were measured, controlled, and the same in every session.
The lifting tempo was self-paced but had to be controlled and without sudden jerks. The repetitions were executed continuously without any rest between the repetitions. In the experimental test, the load started at the 6RM load achieved in the familiarization test. The load was either increased or decreased by 2.5 or 5 kg until 6RM was obtained. All participants reached their 6RM loads within 3 attempts. The testing was ended if the participants were not able to complete all 6 repetitions of a set, if the participant and the test leader agreed that the participant could not lift a higher weight, or if the participants were not able to maintain proper technique. A minimum of 3-minute rest was given between each attempt. A linear encoder (sampling frequency of 100 Hz; Ergotest Technology AS, Langesund, Norway) was used to measure the lifting time and to identify the beginning and end of each repetition. The measurement system synchronized the signals from the linear encoder, force cell, and EMG (MuscleLab 4020e; Ergotest Technology AS).
Knee Extensor Maximal Voluntary Isometric Contraction
The participants were sitting on a table with a 90° angle (controlled with a protractor) in the hip with back support. Straps around the thighs (upper and lower parts) were used to hold the participants in the same position during the MVC tests. The arms were held across the chest touching onto the opposite shoulder. The dominant leg was attached to a force cell with a nonelastic sling (Ergotest Technology AS) at the ankle (Figure 1B). The force cell was attached to a bolt fastened to the floor. A protractor was held along the femur and tibia before the test to identify the 60, 90, and 120° knee angles. The sling was adjusted so that knee extensor MVCs could be performed at the 3 angles. The participants were instructed to extend their leg with maximal force for 3 seconds. Three attempts were conducted on each angle with a minimum of 1-minute rest between each attempt. The mean maximal force of 2 seconds of the best attempt was used in further analyses (20). The force cell was synchronized with the EMG recording system (MuscleLab 4020e; Ergotest Technology AS).
Countermovement jumps were performed on a force platform (MuscleLab Force Plate model 2; Ergotest Technology AS) (Figure 1C). In each condition, the participants were instructed to lower themselves rapidly to a depth of 60, 90, and 120° in the knees before starting the concentric phase as explosive as possible. The angles were identified using a protractor, and a horizontal rubber band was used to mark the bottom position. If the participants did not touch or touched the band excessively (maximum 2 cm displacement), the attempt was not approved. The hands were kept at the hips throughout the jump. The same test leader visually controlled this throughout the study. Each participant had to complete 3 approved attempts in each condition and the best jump height was used in further analysis. All participants were able to complete 3 successful jumps within 3–5 attempts. A minimum of 1-minute rest was given between each attempt. The jump height was calculated by the impulse using a commercial software program (MuscleLab V8.13; Ergotest Technology AS).
In the 6RM back squat and leg extensor MVCs, EMG measurements of the vastus lateralis, vastus medialis, and rectus femoris on the dominant leg were conducted. Before placing the gel-coated self-adhesive electrodes (Dri-Stick Silver circular sEMG Electrodes AE-131; NeuroDyne Medical, Cambridge, MA, USA), the skin was shaved, abraded, and washed with alcohol according to the recommended procedures (15). The electrodes (11-mm contact diameter and a 2-cm fixed center-to-center distance) were placed along the presumed direction of the underlying muscle fiber and located according to the recommendations by SENIAM (15). In other words, different anatomic landmarks were used to strengthen the reliability of the placement between the pre- and posttest. The electrode on vastus lateralis was located two-thirds down the line between spina iliaca anterior superior and the lateral side of the patella. The electrode on vastus medialis was placed four-fifths down the line between spina iliaca anterior superior and the cavity in front of the medial collateral ligament. The electrode sampling EMG data from rectus femoris was placed halfway down the line from spina iliaca anterior superior and the upper part of the patella (15). To minimize noise from the surroundings, the raw EMG signal was amplified and filtered using a preamplifier located close to the sampling point. The preamplifier had a common mode rejection ratio of 100 dB, low cut frequency 8 Hz, and high cut frequency 600 Hz. The EMG signals were converted to root mean square (RMS) signals using a hardware circuit network (frequency response 0–600 kHz, averaging constant 100 ms, total error ±0.5%). Finally, the RMS-converted signal was sampled at 100 Hz using a 16-bit A/D converter. The mean of each of the 6 repetitions in back squat (the average of the concentric and eccentric phase) and the mean of the 2 seconds of knee extensor MVC with the highest force production were used to calculate the RMS EMG using a commercial software (MuscleLab V8.13; Ergotest Technology AS).
The 10-week intervention consisted of 2 sessions per week with a minimum of 48 hours between the sessions. The program consisted of 2 exercises, squat and split (Figure 2). A 10-minute general warm-up (jogging or cycling) and 3 specific warm-up sets were always completed before the training sessions. The specific warm-up consisted of 2 sets of back squat (10–15 repetitions at 30–50% of 6RM and 6 repetitions at 50–70% of 6RM) and one set of split (6 repetitions at 50–70% of 6RM). The training program progressed from 3 sets with 10RM (weeks 1–4); 3 sets with 8RM (weeks 5 and 6), and 4 sets with 6RM (weeks 7–10). If a participant completed more or less repetitions than the intended number, weights were added or removed before the next set or session. The set counted as one of the training sets. The different intensities (6–10RM) corresponded to a progression from 75 to 85% of 1RM throughout the intervention period (5). In the split exercise, the repetitions with the dominant and nondominant leg in front were executed consecutively and considered as one set (Figure 2B). A 2- to 3-minute rest was given between each set. All repetitions were performed in a controlled self-selected tempo. An experienced instructor supervised all sessions to ensure safety, that the load matched the intended intensity and assisted the participants with the last repetition if necessary. In the EBG, the exercises were performed using free weights in combination with 2 elastic bands, one at each side of the barbell (Figures 2A, B). Throughout the intervention, the load was increased in the EBG by adding more weights; however, the relative intensity was at all times matched between the 2 groups. To ensure that the bands provided the same force during the intervention, they were controlled every second week. Each band was stretched to a standard length of 125 cm and should then produce a force of approximately 180 ± 10 N. If not, the band was replaced. Before and after the intervention, the total resistance (free weights + elastic bands) was calculated in the squat for all participants in the EBG. This was done by measuring the height in the upper and lower position in squat. The bands provided a near perfect stretching length-force output relationship (r = 0.9), and the total resistance was then calculated by adding the free-weights load and the force provided by the bands in upper and lower position. The contribution to the total resistance from the bands in the upper position (straight knee and hip) was 58% in the beginning and 38% in the end of the intervention. In the bottom position (trochanter major aligned with the upper part of the patella), the contribution was 44% in the first and 27% in the last training session.
A mixed design (within-time, between-group) analysis of variance (ANOVA) with repeated measures was used to analyze the main and interaction effects for time and group. When differences were detected by ANOVA, paired t-tests with a Bonferroni's post hoc correction were applied to determine where the differences lay. Statistical analyses were performed with SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). All results are presented as mean ± SD values and Cohen's d effect size. An effects size was considered small, medium, and large at 0.2, 0.5, and 0.8, respectively. The level of significance was set at p ≤ 0.05. The reliability (intraclass correlation [ICC]) ranged from 0.93 to 0.99 at the different angles in the CMJs and MVCs. In the back squat, the ICC between habituation and pretest was 0.93.
Pre- and postintervention data for knee extensor MVC, CMJ height, and squat strength, as well as EMG, are presented in Table 2. There were no significant differences between the groups at baseline.
Analyzing force output in the knee extensor MVC tests revealed no time × group interactions at any angles (F = 0.60–1.77, p = 0.19–0.45) nor any main effect for the group (F = 0.02–0.24, p = 0.63–0.88). There was, however, a main effect for time at all angles (F = 6.29–26.73, p < 0.01–0.02). Post hoc analysis showed that the EBG increased their force output at 60° with 15 ± 15% (p < 0.01, ES = 0.64), but not at 90 or 120°, respectively (p = 0.96 and 0.48; Figure 3). The FWG increased their force output at 60 and 90° by 24 ± 26% and 15 ± 16%, respectively (60°: p < 0.01, ES = 1.02; 90°: p < 0.01, ES = 0.69). At 120°, there was a tendency toward improved force output (p = 0.06).
Analyzing muscle activation for knee extensor MVC revealed no significant time × group interactions (F < 0.01–2.96, p = 0.10–0.96) nor main effects (time: F < 0.01–2.94, p = 0.10–0.92; group: F < 0.01–3.12, p = 0.09–0.95) at 60, 90, nor 120°; however, there was a tendency toward a main effect for vastus lateralis at 120° (p = 0.09).
There were no time × group interactions (F = 0.06–0.92, p = 0.35–0.81) nor group main effects at any angles (F = 0.03–0.55, p = 0.47–0.87); however, a main effect for time was revealed at all angles (F = 11.88–21.94, p < 0.01). Post hoc analysis showed that the EBG had a 15 ± 14% and 10 ± 13% improvement (p < 0.01, ES = 0.54 and p = 0.02, ES = 0.46; Figure 4) in CMJ performance when ending the descending phase at knee angles of 60 and 90°. There were 16 ± 26% and 15 ± 16% improvement (p = 0.01, ES = 0.72 and p < 0.01, ES = 1.01) for the FWG. At 120°, the FWG had a 12 ± 15% (p = 0.01, ES = 0.85) improvement, whereas the EBG had no significant improvement (p = 0.20).
There was no time × group interaction (F = 0.18, p = 0.67) nor group main effect (F = 0.26, p = 0.61) in load lifted. However, a time main effect was detected (F = 328.41, p < 0.01). Post hoc tests showed a significant increase of 25 ± 8% and 23 ± 7% in 6RM load in the EBG (p < 0.01, ES = 1.7) and FWG (p < 0.01, ES = 1.7), respectively, from baseline to posttest (Figure 5).
Concerning muscle activation in the squat, there was no time × group interaction (F < 0.01–1.87, p = 0.18–0.96) nor any main effects for time (F ≤ 0.01–1.56, p = 0.22–0.98) or group (F = 0.02–0.14, p = 0.71–0.89) (Table 2).
This study showed improvements for training with FWG and EBG in combination with free weights. In the 6RM back squat, both groups increased their performance loads similarly. In the knee extensor MVC, the FWG increased their force output significantly at 60 and 90°, whereas the EBG only increased significantly at 60°. In the CMJ, the FWG increased their jump height significantly at all angles, whereas the EBG only improved at 60 and 90°. There were no significant changes in neural drive, assessed by EMG.
The FWG improved their force output at 2 angles (60 and 90°), whereas the EBG only improved at the angle with greatest knee flexion (60°). These results were in contrast to the hypotheses as it was expected that variable resistance would be more effective in improving muscle strength at more extended knee angles (i.e., 120°). One possible explanation is that the intervention consisted of relatively deep squat and splits that were rather unfamiliar to the subjects who typically performed half-squats or half-leg presses in their usual training routine and therefore could have more potential for improvement at more flexed knee angles. In contrast to the present results, Melo Mde et al. (22) found greater improvement after variable resistance training, but only at more extended knee angles than the ones assessed in this study. However, our findings of no significant differences between the training groups are in line with most previous studies (7,19,26). The lack of differences in improvement between the 2 training regimes could be caused by matched relative training intensity. Training intensity is one of the most paramount variables in resistance training and is likely to be much more important than whether the resistance is constant or variable. Second, the training regimes became increasingly similar throughout the intervention, as we for practical reasons increased the training load in the EBG by adding more plates to the bar rather than adding resistance from elastic bands. As the participants increased their training load, the percentage contribution from elastic bands to the total load (loads + elastic band resistance) decreased throughout the intervention (from 58 to 38% in the upper phase). Third, it could be that the isometric knee extension test was too nonspecific from the training regimes to be able to detect small differences between the groups.
In contrast to previous studies (7,8,26), this study compared the effect of variable and constant resistance on dynamic ballistic force production at 3 different knee angles. There were no significant differences between the groups, but the CMJs improved for both groups starting the ascending phase from 60 and 90°, and for the FWG also at 120°. This stands in contrast to our hypothesis that the variable regime should be beneficial for neuromuscular performance at the end of the ROM. Again, it could be that lack of experience with knee extensor training at high levels of knee flexion provided more room for improvement here. The results were in line with Joy et al. (16) who examined self-selected lowering depth in CMJ and found improvements in both constant and variable resistance training without any difference between the training regimes. Cronin et al. (8), however, did not find improvement in neither the variable nor the constant training group after 10 weeks of resistance training. Still, when considering the knee extension and CMJ results, although no significant differences were discovered between the groups, the differences in effect size could imply that constant resistance training is advantageous compared with variable resistance training.
The EMG data demonstrated no interaction or main effects in either the knee extensor MVC or dynamic squat. The only study comparable with the present is Walker et al. (26). In contrast to our results, they found an increased activation of vastus lateralis and medialis (presented as a mean of the 2 muscles) in both groups during concentric knee extension. In knee extensor MVC, the variable resistance training group had an improved activation after 10 weeks of training (midtest), whereas the constant group trained for 20 weeks before significantly increasing the EMG activity. The discrepant results between Walker et al. (26) and our study may be due to including older men without much experience with resistance training as participants, who may have more potential for neural adaptations than trained adults. In addition, Walker et al. had a longer intervention and used knee extension as a part of their training regime, which could increase the muscle activation because of a longer training stimuli and higher specificity between the training and the tests.
Although the FWG was more accustomed to the 6RM squat test, both groups had similar increases in 6RM squat without any differences in the effect size. These results are in agreement with our hypothesis and several comparable studies measuring dynamic strength (7,8,21,22), but in contrast to Bellar et al. (6) whom reported results favoring variable resistance training and 2 studies favoring constant resistance (4,23).
There are some limitations to this study. Only healthy women with resistance training experience were recruited and the results can therefore not necessarily be generalized to other populations. Differences in menstrual cycle, between pre- and posttest, use of contraceptive pills, or variation in hormone levels were not controlled. There was no control group included. However, the main purpose of this study was to compare the 2 different strength training regimes, and earlier studies have clearly demonstrated the effect of strength training vs. no training. Surface EMG provides only an estimate of the muscle activation, and there is a possible risk of crosstalk from neighboring muscles (12). However, electrodes with a small and fixed center-to-center distance were used to limit this source of error (28). Furthermore, there are additional methodological limitations with dynamic EMG recordings (11) and performing EMG measurements before and after an intervention (15), which could mask small effects.
In conclusion, constant and variable resistance training provided similar increases in dynamic and isometric strength, as well as ballistic muscle performance. Still, the constant resistance training group improved in more of the tests than the variable resistance training group. There was no evidence for changes in neuromuscular activity.
The use of constant resistance (e.g., free weights or conventional training machines) provides constant external force rather than constant tension throughout the ROM. The use of variable resistance (e.g., elastic bands) could in theory provide more optimal muscle stimuli across the entire ROM for squat exercises, which in turn could lead to higher training adaptations over time. However, our results suggest that athletes, professionals, and other resistance-trained adults can expect similar increases in both maximal- and explosive knee extensor strength at various knee angles from constant or variable resistance training. If we were to recommend one of the modalities, greater effect sizes imply that constant resistance training could be more beneficial. However, they provided broadly similar improvements and both modalities could be included in a periodized training program. These results are highly valid for athletes and coaches and their real-world training as the study, in contrast to earlier studies, recruited resistance-trained subjects, had a sufficient training volume with a high periodized intensity, and did not involve any other forms of training that could mask the results.
The authors thank the participants for their positivity and participation in the study. This study was conducted without any funding from companies or manufacturers or outside organizations.
1. Aboodarda SJ, Hamid MSA, Muhamed AMC, Ibrahim F, Thompson M. Resultant muscle torque and electromyographic activity during high intensity elastic resistance and free weight exercises. Eur J Sport Sci 13: 155–163, 2013.
2. Aboodarda SJ, Shariff MA, Muhamed AM, Ibrahim F, Yusof A. Electromyographic activity and applied load during high intensity elastic resistance and nautilus machine exercises. J Hum Kinet 30: 5–12, 2011.
3. Andersen LL, Andersen CH, Mortensen OS, Poulsen OM, Bjornlund IB, Zebis MK. Muscle activation and perceived loading during rehabilitation exercises: Comparison of dumbbells and elastic resistance. Phys Ther 90: 538–549, 2010.
4. Anderson CE, Sforzo GA, Sigg JA. The effects of combining elastic and free weight resistance on strength and power in athletes. J Strength Cond Res 22: 567–574, 2008.
5. Baechle TR, Earle RW. Essentials of strength training and conditioning. Champagne, IL: Human Kinetics 2008.
6. Bellar DM, Muller MD, Barkley JE, Kim CH, Ida K, Ryan EJ, Bliss MV, Glickman EL. The effects of combined elastic- and free-weight tension vs. free-weight tension on one-repetition maximum strength in the bench press. J Strength Cond Res 25: 459–463, 2011.
7. Colado JC, Garcia-Masso X, Pellicer M, Alakhdar Y, Benavent J, Cabeza-Ruiz R. A comparison of elastic tubing and isotonic resistance exercises. Int J Sports Med 31: 810–817, 2010.
8. Cronin J, McNair PJ, Marshall RN. The effects of bungy weight training on muscle function and functional performance. J Sports Sci 21: 59–71, 2003.
9. Ebben WP, Jensen RL. Electromyographic and kinetic analysis of traditional, chain, and elastic band squats. J Strength Cond Res 16: 547–550, 2002.
10. Eloranta V. Coordination of the thigh muscles in static leg extension. Electromyogr Clin Neurophysiol 29: 227–233, 1989.
11. Farina D. Interpretation of the surface electromyogram in dynamic contractions. Exerc Sport Sci Rev 34: 121–127, 2006.
12. Farina D, Merletti R, Enoka RM. The extraction of neural strategies from the surface EMG. J Appl Physiol (1985) 96: 1486–1495, 2004.
13. Frost DM, Cronin J, Newton RU. A biomechanical evaluation of resistance: Fundamental concepts for training and sports performance. Sports Med 40: 303–326, 2010.
14. Ghigiarelli JJ, Nagle EF, Gross FL, Robertson RJ, Irrgang JJ, Myslinski T. The effects of a 7-week heavy elastic band and weight chain program on upper-body strength and upper-body power in a sample of division 1-AA football players. J Strength Cond Res 23: 756–764, 2009.
15. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 10: 361–374, 2000.
16. Joy JM, Lowery RP, Oliveira de Souza E, Wilson JM. Elastic bands as a component of periodized resistance training. J Strength Cond Res 2013.
17. Kraemer WJ, Nindl BC, Ratamess NA, Gotshalk LA, Volek JS, Fleck SJ, Newton RU, Häkkinen K. Changes in muscle hypertrophy in women with periodized resistance training. Med Sci Sports Exerc 36: 697–708, 2004.
18. Kulig K, Andrews JG, Hay JG. Human strength curves. Exerc Sport Sci Rev 12: 417–466, 1984.
19. Manning RJ, Graves JE, Carpenter DM, Leggett SH, Pollock ML. Constant vs variable resistance knee extension training. Med Sci Sports Exerc 22: 397–401, 1990.
20. Marshall PW, Murphy BA. Increased deltoid and abdominal muscle activity during Swiss ball bench press. J Strength Cond Res 20: 745–750, 2006.
21. McCurdy K, Langford G, Ernest J, Jenkerson D, Doscher M. Comparison of chain- and plate-loaded bench press training on strength, joint pain, and muscle soreness in Division II baseball players. J Strength Cond Res 23: 187–195, 2009.
22. Melo Mde O, Brodt GA, Gomes LE, Candotti CT, La Torre M, Loss JF. Effects of added elastic tubes on open-chain knee extensor strength training. Sports Biomech 12: 195–203, 2013.
23. Pipes TV. Variable resistance versus constant resistance strength training in adult males. Eur J Appl Physiol 39: 27–35, 1978.
24. Rhea MR, Kenn JG, Dermody BM. Alterations in speed of squat movement and the use of accommodated resistance among college athletes training for power. J Strength Cond Res 23: 2645–2650, 2009.
25. Saeterbakken AH, Andersen V, Jansson J, Kvellestad AC, Fimland MS. Effects of BOSU ball(s) during sit-ups with body weight and added resistance on core muscle activation. J Strength Cond Res 28: 3515–3522, 2014.
26. Walker S, Peltonen H, Sautel J, Scaramella C, Kraemer WJ, Avela J, Hakkinen K. Neuromuscular adaptations to constant vs. variable resistance training in older men. Int J Sports Med 35: 69–74, 2014.
27. Walker S, Taipale RS, Nyman K, Kraemer WJ, Häkkinen K. Neuromuscular and hormonal responses to constant and variable resistance loadings. Med Sci Sports Exerc 43: 26–33, 2011.
28. Winter DA, Fuglevand AJ, Archer SE. Crosstalk in surface electromyography: Theoretical and practical estimates. J Electromyogr Kinesiol 4: 15–26, 1994.