Enhancement of knee extension voluntary and electrically-evoked strength with the body tilted backward

Abstract

BACKGROUND:

While extreme changes in body position (e.g., standing vs. sitting) have been shown to affect knee extensor strength, little is known regarding how less extreme changes in body position affect knee extensor strength and activation.

OBJECTIVE:

This study evaluated the effect of four seated positions on the maximum voluntary isometric contraction (MVIC) and electrically-evoked isometric strengths of the knee extensors.

METHODS:

Forty healthy subjects participated. MVIC strength, electrically-evoked strength, and voluntary activation of the quadriceps muscles were measured under four seated positions: 1) 22.5–25 ${}^{\circ}$ forward tilt, 2) 0 ${}^{\circ}$ tilt, 3) 22.5–25 ${}^{\circ}$ backward tilt, and 4) 45–50 ${}^{\circ}$ backward tilt. The effect of seated position on each dependent variable was assessed using a one-way repeated measures ANOVA with linear contrasts.

RESULTS:

MVIC and electrically-evoked strengths increased ( $P\leqslant$ 0.037) by 5–11% as the body was moved from the 22.5–25 ${}^{\circ}$ forward-tilted to 45–50 ${}^{\circ}$ backward-tilted position. However, voluntary muscle activation and EMG amplitudes of the knee extensors during MVIC did not change.

CONCLUSIONS:

Knee extensor MVIC and electrically-evoked strengths were modestly enhanced as the body was tilted backwards. The use of a semi-reclined position may assist in obtaining a more forceful contraction of the knee extensors, thereby potentially resulting in better strength training outcomes.

Keywords

Muscle activation electromyography quadriceps interpolated twitch

1. Introduction

Strength assessment and strengthening of the knee extensor muscle group (i.e., quadriceps femoris) play an important role in research and patient care for various musculoskeletal disorders, such as knee ligament or meniscus tears, cartilage degeneration, and patellofemoral pain syndrome. However, the optimal testing position(s) and/or joint configuration(s) to evaluate knee extensors remain unknown. For example, while the optimal knee angle(s) for maximum voluntary isometric contraction (MVIC) activation and strength of the knee extensors have been studied extensively, there is no consensus on that angle. Several studies suggested an optimal angle of approximately 90 ${}^{\circ}$ and 70 ${}^{\circ}$ of knee flexion for assessing knee extensor MVIC activation and strength, respectively [1, 2, 3, 4]. Other studies, however, have indicated similar activation and/or strength across a range of knee flexion angles [5, 6].

Given that the rectus femoris muscle spans both the hip and knee joints, position of the hip joint is another factor that may affect MVIC activation and strength of the knee extensors. However, previous findings on the optimal hip flexion angle(s) for knee extensor MVIC strength remain controversial [7, 8, 9]. Opposite results have been reported regarding whether a greater knee extensor MVIC strength could be achieved when the hip was flexed at 90 ${}^{\circ}$ vs. 0 ${}^{\circ}$ [7, 9]. In addition, Hasler et al. found electromyographic (EMG) amplitude of the vastus lateralis (VL) and vastus medialis (VM) muscles during a MVIC to be highest at both 90 ${}^{\circ}$ and 0 ${}^{\circ}$ as compared to the other intermediate angles, and observed no effect of hip flexion angle on the EMG amplitude in the rectus femoris [8].

One possible factor that may have contributed to the inconsistent findings regarding the optimal knee/hip joint angle(s) for knee extensor activation and strength is the variation among studies in body position in the sagittal plane because of the wide variation in seating among different dynamometers that have been used to assess knee extensor strength. While holding hip and knee angles constant (120 ${}^{\circ}$ and 60 ${}^{\circ}$ of flexion, respectively), Thomee and colleagues observed knee extensor strength to be greater in a so-called standing position as opposed to a sitting position [10]. A more recent study by Paddock and Behm found that inversion of the body (with the subjects remained seated with hips and knees flexed at 90 ${}^{\circ}$ ) reduced knee extensor strength and VL EMG amplitude by 6% and 30%, respectively, as compared to an upright position [11]. The same laboratory also reported a similar effect of body inversion on the elbow flexors [12].

However, little is known regarding the effects of less extreme changes in a sitting testing position on muscle activation and strength, which is more relevant and practical to be implemented clinically when compared to extreme modifications in testing positions (e.g., inversion, standing). Such information is thus critical for determining the optimal sitting position for knee extensor strength testing/training and standardizing testing protocols during muscle functional evaluation. The primary objective of our studies, therefore, was to assess the effect of altering body tilting position during sitting in the sagittal plane, while keeping the knee and hip angles constant, on MVIC muscle activation and strength as well as electrically-evoked strength of the knee extensors muscles. We hypothesized that altering body position during sitting would modify the activation and strength of the knee extensor muscles.

2. Methods

2.1 Participants

Twenty young, healthy subjects [14 males and 6 females; mean ( $\pm$ SD) age: 28.0 ( $\pm$ 6.4) yr] completed Study I, and 20 additional young, healthy subjects [11 males and 9 females; mean age: 27.0 ( $\pm$ 6.7) yr] participated in Study II. Subjects reported no contraindications for participation such as musculoskeletal injuries of the lower extremities within the last year, pregnancy, cardiopulmonary disease, disorders of the nervous system, or cancer. Based on the data reported by Thomee and colleagues [10] with the assumption of equal variance and moderate correlation between body positions (i.e., $r=$ 0.5), our sample size of 20 per study would provide a statistical power greater than 90% to detect a significant difference in knee extensor MVIC strength between body positions. The research protocols were approved by the University Institutional Review Board and all subjects gave written informed consent.

2.2 Study design

This study was based on a repeated-measures design. MVIC muscle activation and strength as well as electrically-evoked strength of the knee extensors muscles were measured under 4 different seated positions.

2.3 Procedures

2.3.1 Study I: Effect of sagittal body position on strength and activation of the knee extensors

The knee extensors of the right leg were tested by having the subject seated in a knee extension/flexion machine (NT-1260, Nautilus Fitness Products; Louisville, CO, USA) with the hip and knee set at 90 ${}^{\circ}$ and 70 ${}^{\circ}$ , respectively, of flexion. The subject was positioned so that his/her knee’s axis of rotation (defined by the femoral epicondyles) coincided with that of the knee extension/flexion machine. A five-point racing harness attached to the machine was used to secure the torso. In addition, a custom seat belt along with thigh and right ankle straps were used to hold the subject securely in place (Fig. 1A). To prevent neck flexion, an elastic wrap was placed over the forehead and wrapped around the back support of the knee extension/flexion machine. To ensure consistent positioning of the subject from one test position to the next, the harness and strap length settings during the initial testing position were recorded and kept identical during the subsequent testing positions. Subjects were tested in four body positions in a random order: 1) 22.5 ${}^{\circ}$ forward tilt, 2) 0 ${}^{\circ}$ tilt (i.e., upper body parallel to the vertical), 3) 22.5 ${}^{\circ}$ backward tilt, and 4) 45 ${}^{\circ}$ backward tilt (Fig. 1A).

Figure 1.

Schematic of the experimental setup. The subject’s head, trunk, and legs were secured with straps. The four testing positions were then achieved by tilting the knee extension machine forward and backward in a random order (A). To calculate knee extension torque, a force transducer (for Study I) and strain gauge (for Study II) were attached to the metal frame of the knee extension/flexion machine (B).

To measure knee extensor torque production, a force transducer (SBO-300, Transducer Techniques; Temecula, CA, USA) was attached on one end to the back of the bar upon which the subject kicked against with his/her right leg and the other end attached to immovable bracing of the machine (Fig. 1B). The output signal of the force transducer was amplified (TMO-2, Transducer Techniques; Temecula, CA, USA) and sampled at 5000 Hz using a combined A/D and D/A board (KPCI-3116, Keithley Instrument; Solon, OH, USA), 2.2-GHz Pentium computer, and custom written programs with TestPoint software (ver. 7.0, Measurement Computing, Norton, MA, USA). The force recorded was then multiplied by the moment arm (distance from the knee axis to the center of bar that the subject kicked against) to calculate the knee extensor torque. To remove the varying effects of the gravity among testing positions on strength recording, for each testing position, the passive torque exerted by the weight of the leg was determined prior to each measurement when the subject was instructed to relax completely. The recorded value of the passive torque (average from a 1-s recording) was then used to correct the subsequent strength measurements.

An interpolated twitch procedure was used to quantify MVIC muscle activation and strength as well as electrically-evoked strength at each of the four body positions [13, 14, 15]. For stimulating the knee extensors, two 7-cm $\times$ 10-cm adhesive electrodes (UniPatch 620SS; Wabasha, MN, USA) were placed on the skin overlying the thigh, one over the distal VM muscle and the other over the proximal VL muscle near the anterior superior iliac spine of the ilium (ASIS). Impedance of the electrodes was recorded using an impedance meter (model EZM5D, Grass Technologies; Braintree, MA, USA) both before and after testing. Impedance was checked to ensure that the electrodes were not damaged and that there was adequate adhesion to the skin (i.e., $\leqslant$ 10 k $\Omega$ ). The electrodes were connected to a constant-current stimulator (model DS7AH, Digitimer; Hertfordshire, England) that was controlled using the data acquisition and control system described previously [13, 14, 15].

To determine the stimulation current needed for the interpolated twitch procedure, a series of electrically-evoked isometric contractions of the knee extensors was performed with the current being increased on succeeding stimulations. Each stimulation consisted of a paired-pulse stimulation, i.e., two 0.2-ms pulses separated by 10 ms. Stimulator current was initially set to 100 mA and increased by 20 mA on each succeeding stimulation. Progressively stronger stimulations were delivered once every 20 s until the peak contraction torque reached a plateau and then showed a decline on two successive stimulations. The current eliciting the highest peak torque on the plateau of the torque-current curve was used for the remainder of the test session. Determination of the optimal stimulation current was performed at each of the 4 body positions.

To do the interpolated twitch procedure, the subject was instructed to perform a 3-s MVIC of his/her right knee extensor muscles. Auditory cues elicited by the custom-written TestPoint software were used to signal the subject when to start and stop the contraction. At 2.5 s into the MVIC, the muscle group was stimulated with a paired-pulse stimulation, and the increase in muscle torque over the MVIC level (i.e., interpolated twitch torque, ITT) was measured. At 2 and 4 s after the end of the MVIC, the same paired-pulse stimulation was administered again to relaxed muscle to determine the peak electrically-evoked torque (EET). The average value for the two stimulations was used as the electrically-evoked strength and as the EET for the percent muscle activation calculation. The percent muscle activation during MVIC was calculated as 100% (1 – ITT/EET) [13, 14, 15]. MVIC strength was defined as the average torque between 2 and 2.5 s into the MVIC.

Muscle electrical activities of the knee extensor and flexor muscles were also assessed during the interpolated twitch procedure using preamplified surface EMG electrodes (Delsys DE-2.1, Boston, MA, USA). Overlying skin was first prepared by shaving any hair and cleaning with an alcohol wipe. Electrode placement was as per Rainoldi et al. [16]. For the VL muscle, an EMG electrode assembly was placed at 33% of the distance between the superior lateral patella and ASIS. For the VM muscle, an electrode assembly was placed over muscle at 30% of the distance between the superior medial patella and ASIS. For the biceps femoris (BF) and semitendinosus/semimembranosus (ST/SM) muscles, the subject was placed in the prone position with the knee extended. For the BF muscle, an electrode assembly was placed at 67% of the distance from the ischial tuberosity to the lateral border of the popliteal cavity. For the ST/SM muscles, an electrode assembly was placed at 67% of the distance from the ischial tuberosity to the medial border of the popliteal crease across the ST tendon. The longitudinal axis of each electrode assembly was aligned parallel to the estimated longitudinal axis of the underlying muscle fibers. An adhesive 2-cm $\times$ 2-cm reference electrode was placed on the skin overlying the fibular head. The electrodes were connected to an 8-channel EMG amplifier (Delsys Bagnoli EMG system, Boston, MA, USA) using a 20–450 Hz band-pass filter and a gain setting of 1000. The amplified output was sampled at 5000 Hz using the data acquisition and control system described above. A muscle’s EMG amplitude was calculated as the root-mean-square (RMS) over the 0.5-s interval between 2 and 2.5 s into the MVIC, i.e., coinciding with the time period used for the determination of MVIC strength.

During testing for each body position, six interpolated twitch procedures were performed, with one minute of rest between procedures. Of the six interpolated twitch procedures, the three best attempts (i.e., the three trials with the highest peak MVIC torques) were determined and their data averaged together for use in the statistical analyses.

Figure 2.

The MVIC strength (A), electrically-evoked torque/strength (B), and percent activation during MVIC (C) recorded under the four body positions. Values are shown as Mean $\pm$ SE. ${}^{*}$ Denotes a significant linear trend with increased backward body tilt.

Figure 3.

The EMG RMS values for the VL (A), VM (B), BF (C), and ST/SM (D) muscles. ${}^{*}$ Denotes a significant linear trend with increased backward body tilt. Values are shown as Mean $\pm$ SE.

Figure 4.

The un-potentiated (A) and potentiated (B) electrically-evoked isometric torque of the knee extensor muscles recorded under the four body positions. Values are shown as Mean $\pm$ SE. ${}^{*}$ Denotes a significant linear trend with increased backward body tilt.

2.3.2 Study II: Effect of sagittal body position on potentiated and unpotentiated electrically- evoked strength of the knee extensors

Because of our surprising finding in Study I of a sagittal body position effect on electrically-evoked strength, we conducted Study II to confirm and extend our Study I findings. In Study II, we used a force-measuring methodology (i.e., strain gauge-based) different from that used in Study I. This was done because the strain gauge-based system was less susceptible to noise induced by sudden impact forces on the knee extension/flexion machine. In Study II, we also focused exclusively on measuring electrically-evoked strength, specifically strength measured under unpotentiated and potentiated conditions.

The equipment and procedures used in Study II were identical to those used in Study I with the following exceptions. First, as mentioned above, the force transducer-based system for measuring torque was replaced. A strain gauge (model SGD-30/120, Omega Engineering, Inc.; Norwalk, CT, USA) was bonded to the back of the bar upon which the subject kicked against (Fig. 1B). Additionally, an Omega BCM-1 bridge completion module and DMD-465 signal amplifier were used to record the torque. Second, three of the four body positions used were changed slightly to obtain the maximal range possible in the sagittal plane with our equipment. The forward and backward tilts of 22.5 ${}^{\circ}$ were increased to 25 ${}^{\circ}$ tilts. Similarly, the 45 ${}^{\circ}$ backward tilt was increased to a 50 ${}^{\circ}$ tilt. Finally, the interpolated twitch protocol was replaced with a sequence of electrically-evoked contractions during which torque production was recorded. Prior to beginning this sequence, the subjected rested for 2 min; and the passive torque exerted by the weight of the leg was determined and used to correct the subsequent strength measurements. The subject was then stimulated using the two-pulse stimulation using optimal current as described for Study I. Peak torque produced under this condition was considered to be unpotentiated electrically-evoked strength. Ten seconds later, the subject performed a 3-s MVIC. Electrical stimulations were elicited at 2 and 4 seconds after the end of the MVIC contraction. The peak torques recorded following these two stimulations were averaged together and considered to be potentiated electrically-evoked strength. The sequence of electrically-evoked contractions was repeated two more times with two minutes in between. The values from the three trials were averaged together and used in the subsequent statistical analyses.

2.4 Statistical analysis

The dependent variables for Study I included 1) MVIC strength, 2) percent muscle activation during MVIC, 3) electrically-evoked strength, and 4) EMG RMS of the knee extensor (i.e., VM and VL) and flexor (i.e., BF and ST/SM) muscles. The dependent variables for Study II were the unpotentiated and potentiated electrically-evoked strengths of the knee extensor muscles. For each variable of interest, a one-way repeated measures ANOVA with linear contrast was used to compare the difference among the four testing positions with a significance level of 0.05 (SPSS ver. 21, IBM; Armonk, NY, USA).

3. Results

For Study I, a significant linear relationship was found between sagittal seated body position and knee extensor strength, both MVIC and electrically-evoked ( $P\leqslant$ 0.037, Fig. 2). Specifically, as the body was moved from the 22.5 ${}^{\circ}$ forward-tilted to the 45 ${}^{\circ}$ backward-tilted position, these strengths increased by 5–9%. No significant effect was found for body position on the percent muscle activation during MVIC ( $P=$ 0.39, Fig. 2) as well as the EMG RMS values in the VM, VL, and BF muscles ( $P\geqslant$ 0.12, Fig. 3). However, a significant linear relationship was found between body position and ST/SM EMG RMS ( $P=$ 0.005, Fig. 3); the EMG RMS increased from the 22.5 ${}^{\circ}$ forward-tilted to the 45 ${}^{\circ}$ backward-tilted position but the increase was small, corresponding to an increase from 12% to 14% of the EMG RMS recorded during knee flexor MVIC. For Study II, a significant linear relationship was found between sagittal body position and electrically-evoked strength, both potentiated and unpotentiated ( $P\leqslant$ 0.001, Fig. 4); as the body was moved from the 25 ${}^{\circ}$ forward-tilted to the 50 ${}^{\circ}$ backward-tilted position, these strengths increased by 9–11%.

4. Discussion

The primary purpose of this study was to determine the effects of sagittal body position on knee extensor muscle activation and strength during a MVIC as well as the effect on electrically-evoked strength. The results in part supported our hypothesis that altering body position may result in changes in the strength and activation of the knee extensor muscles. Specifically, we found in Study I that as body position was changed from the 22.5 ${}^{\circ}$ forward-tilted to the 45 ${}^{\circ}$ backward-tilted position (i.e., a total change of 67.5 ${}^{\circ}$ ), both MVIC and electrically-evoked strengths increased linearly but modestly (i.e., 5–9%). The electrically-evoked strength finding was confirmed in Study II using another force-measuring methodology (i.e., strain gauge-based vs. force transducer-based) and a slightly greater range of angles in the sagittal plane (i.e., 75 ${}^{\circ}$ vs 67.5 ${}^{\circ}$ ). On the other hand, however, neither percent muscle activation nor EMG amplitude for the VM or VL muscles measured during MVIC were affected by a change in body position. Taken together, our findings indicate that a semi-reclined position may assist in generating a more forceful contraction of the knee extensor muscle group.

While the effects of body position on knee extensor strength have been investigated [10, 11], the experimental approaches in previous studies involved extreme changes in body position (e.g., standing vs. sitting, upright vs. inverted). Our modification of a sitting position by tilting the subject forward/backward while maintaining the same body/joint configurations is more comparable to knee extensor strength testing protocols commonly used in clinical practices. The observed linear increase in MVIC and electrically-evoked strengths with increased backward tilt angle may help explain some of the discrepancies in previous studies concerning the optimal joint position(s) of assessing knee extensor strength [1, 5, 7, 9]. Based on our findings, standardization of body tilting position during sitting in addition to the knee and hip angles must be considered when assessing knee extensor muscle strength. Future studies with combinations of various body tilting and lower-extremity joint positions may help determine the optimal body configuration(s) for examining knee extensor strength.

The mechanism(s) associated with the increased MVIC strength with backward body tilt remain(s) unclear. The varying effects of the gravity among testing positions were minimized by removing the passive torque exerted by the weight of the leg prior to each strength measurement. The increased MVIC strength also cannot be explained by increased central activation or increased muscle/nerve tissue excitability because no significant changes were observed among body positions in either percent muscle activation or EMG amplitude of the VM or VL muscles. A very small increase in ST/SM EMG amplitude indicating enhanced co-contraction was observed as the body was tilted backwards but this observation is directionally wrong in terms of explaining the body position effect on MVIC strength.

Our only explanation for the MVIC strength increase with backward tilting is based on our observation of a body position effect on electrically-evoked strength. This observation suggests that as the body is tilted backwards, intrinsic contractility of the knee extensor muscle tissue is increased. Assuming that all (or at least the same number of) muscle fibers in the knee extensors are activated by electrical stimulation in the four body positions, we can only hypothesize that perhaps calcium release or actomyosin interaction during contraction is altered by sagittal plane body position. However, we are unaware of precedence for such a mechanism under a similar scenario and thus consider this to be an unlikely possibility. Intuitively, it would seem more plausible that our observed sagittal body position effect on MVIC strength would be due to a vestibular effect on central activation and/or muscle/nerve tissue excitability during MVIC. We recommend that future studies consider probing the possibility.

4.1 Limitations

Several limitations must be addressed. First, the distinct varying tilting angles were selected mainly to examine the effects of changing body positions on knee extensor muscle function. These titling angles do not represent the best/worst seating position(s) when assessing knee extensor muscles. While we had ensured that all subjects were well secured during testing, it is still possible that stabilizations of the subjects to the testing chair were not the most ideal, particularly when the subjects were tilting forward (i.e., facing downward), thereby potentially affecting their abilities to perform the knee extensor muscle testing. Lastly, the present study combined both men and women into analyses because we did not intend to compare men to women. Our use of repeated measures study design and significant results/findings of the corresponding statistical analyses would suggest comparable responses in men and women to the changes in body position. Specifically, if men and women demonstrated different responses (e.g., increased performance in men but decreased performance in women when the body was tilted forward or backward), it would likely result in non-significant results instead. However, given our uneven and small samples in women and men, we were unable to fully test this premise through proper statistical analyses that compared men to women. Future studies with sufficient sample sizes of men and women will be needed to examine whether men may respond differently than women in terms of the knee extensor strength and activations under varying seated positions.

5. Conclusions

Strength training of the knee extensors, either through voluntary contraction or electrical stimulation, has been commonly prescribed to patients with knee injuries such as knee ligament or meniscus tears, cartilage degeneration, and patellofemoral pain syndrome. Our findings suggest that placing an individual in a semi-reclined sitting position can assist in obtaining a more forceful contraction of the knee extensor muscle group. Presumably, this would result in a better strength training outcome. Finally, we are unable to provide a definitive physiological mechanism for the sagittal body position effect on MVIC and electrically-evoked strengths.

Footnotes

Conflict of interest

The authors declare no conflict of interest.

References

Pincivero

Salfetnikov

Campy

Coelho

. Angle- and gender-specific quadriceps femoris muscle recruitment and knee extensor torque. Journal of Biomechanics. 2004; 37(11): 1689-97. doi: 10.1016/j.jbiomech.2004.02.005.

Becker

Awiszus

. Physiological alterations of maximal voluntary quadriceps activation by changes of knee joint angle. Muscle & Nerve. 2001; 24(5): 667-72.

Suter

Herzog

. Extent of muscle inhibition as a function of knee angle. Journal of Electromyography and Kinesiology: Official Journal of the International Society of Electrophysiological Kinesiology. 1997; 7(2): 123-30.

Kubo

Tsunoda

Kanehisa

Fukunaga

. Activation of agonist and antagonist muscles at different joint angles during maximal isometric efforts. European Journal of Applied Physiology. 2004; 91(2-3): 349-52. doi: 10.1007/s00421-003-1025-x.

Newman

Jones

Newham

. Quadriceps voluntary activation at different joint angles measured by two stimulation techniques. European Journal of Applied Physiology. 2003; 89(5): 496-9. doi: 10.1007/s00421-003-0836-0.

Weir

McDonough

Hill

. The effects of joint angle on electromyographic indices of fatigue. European Journal of Applied Physiology and Occupational Physiology. 1996; 73(3-4): 387-92.

Fisher

Pendergast

Calkins

. Maximal isometric torque of knee extension as a function of muscle length in subjects of advancing age. Archives of Physical Medicine and Rehabilitation. 1990; 71(10): 729-34.

Hasler

Denoth

Stacoff

Herzog

. Influence of hip and knee joint angles on excitation of knee extensor muscles. Electromyography and Clinical Neurophysiology. 1994; 34(6): 355-61.

Maffiuletti

Lepers

. Quadriceps femoris torque and EMG activity in seated versus supine position. Medicine and Science in Sports and Exercise. 2003; 35(9): 1511-6. doi: 10.1249/01.MSS.0000084426.03247.93.

10.

Thomee

Grimby

Svantesson

Osterberg

. Quadriceps muscle performance in sitting and standing in young women with patellofemoral pain syndrome and young healthy women. Scandinavian Journal of Medicine & Science in Sports. 1996; 6(4): 233-41.

11.

Paddock

Behm

. The effect of an inverted body position on lower limb muscle force and activation. Applied Physiology, Nutrition, and Metabolism = Physiologie Appliquee, Nutrition et Metabolisme. 2009; 34(4): 673-80. doi: 10.1139/H09-056.

12.

Hearn

Cahill

Behm

. An inverted seated posture decreases elbow flexion force and muscle activation. European Journal of Applied Physiology. 2009; 106(1): 139-47. doi: 10.1007/s00421-009-0999-4.

13.

Park

Maresca

McKibans

Morgan

Allen

Warren

. Caffeines enhancement of maximal voluntary strength and activation in uninjured but not injured muscle. International Journal of Sport Nutrition and Exercise Metabolism. 2008; 18(6): 639-52.

14.

Cureton

Warren

Millard-Stafford

Wingo

Trilk

Buyckx

. Caffeinated sports drink: ergogenic effects and possible mechanisms. International Journal of Sport Nutrition and Exercise Metabolism. 2007; 17(1): 35-55.

15.

Collier

Hardy

Millard-Stafford

Warren

. Small beneficial effect of caffeinated energy drink ingestion on strength. J Strength Cond Res. 2016; 30(7): 1862-70. doi: 10.1519/JSC.0000000000001289.

16.

Rainoldi

Melchiorri

Caruso

. A method for positioning electrodes during surface EMG recordings in lower limb muscles. Journal of Neuroscience Methods. 2004; 134(1): 37-43. doi: 10.1016/j.jneumeth.2003.10.014.