Intramuscular hypertrophy is not homogenous along the length of the muscle (10,20,24,25). Although previous studies had found high-force knee extensions elicit greater distal-thigh than mid-thigh hypertrophy occurs in regions that provide the greatest biomechanical advantage for the specific exercise performed during training (10). Because region-specific hypertrophy is believed to have implications for joint function (5,10), health (15), and sports performance (1) and is highly dependent upon the training stimuli, new means of predicting regional hypertrophy are necessary.
Neuromuscular activation is often viewed as an important stimulus for hypertrophy (20) because greater neurological activation of a muscle fiber will result in greater internal forces and metabolic demands of that tissue (2,11). In a recent study by Wakahara et al. (24), muscle activity of the triceps brachii was measured using T2-weighted magnetic resonance imaging (T2-MRI), after doing a chest press exercise and then compared with regional hypertrophy after 12 weeks of training with the same exercise. In this study, the regional muscle activity observed during the exercise displayed a similar pattern to the regional hypertrophy observed post-training. Although these results substantiate the relationship between regional muscle activity and hypertrophy, because of the high costs and difficulty of access that is often associated with the use of T2-MRI, alternative methods of measuring regional muscle activity may hold great clinical significance.
Surface electromyography (sEMG) is a relatively inexpensive noninvasive tool often used by physical therapists and exercise scientists to record muscle activity between 2 electrodes (15). Because it is well established that sEMG amplitude is highly correlated with muscle contraction intensity of the tissue it directly measures (2), peak and mean sEMG during a given motion are often used to estimate the maximal and relative neurological activation of this tissue (6,8,9). Because sEMG signals selectively measure tissue located between the electrodes, and muscle fibers of the vastus lateralis (VL) do not run the length of the muscle, the placement of electrodes will influence which fibers and motor units are being recorded (12,14).
Although sEMG may be a viable tool in estimating muscle activation, unlike T2-MRI, measurements are taken during the exercise from a fixed perspective on the skin and thus are influenced by changes in joint position (8,12,16), muscle geometry (11,22), and movement of the muscle relative to the skin (11,12). Furthermore, because sEMG signals are influenced by the location of the muscle fibers being measured relative to the electrodes, superficial muscle fibers may be overrepresented in the signal (12,22). In the VL, because a greater proportion of type II fibers are located superficially, such fibers will likely predominate the recorded signal (18). This may limit sEMG's ability to compare motions of different speeds or intensities (17,21,27).
To our knowledge, only one study has attempted to link regional sEMG to previously observed regional hypertrophy (19). In this study, subjects performed 4 sets of 8 knee extensions with a load of 80% of their maximum, whereas the root mean square of regional sEMG was measured and compared between the mid region and distal region of the thigh (19). Although previous studies had found high-force knee extensions elicit greater distal-than mid-thigh hypertrophy (20), no differences were observed in sEMG; thus, the authors concluded that regional muscle activity was not responsible for regional hypertrophy from knee extensor training. In this study, mid-thigh and distal-thigh muscle activity were directly compared within a single movement in which the thigh was fixed and thus regional hypertrophy would not alter joint biomechanics. Thus, it cannot be determined if regional muscle activity is greater in areas of the thigh that are most biomechanically beneficial to the exercise used in training and that have been linked to regional hypertrophy for this reason.
The purpose of the present study was to determine if regional muscle activity of the VL, as measured by sEMG, would reflect the patterns of regional hypertrophy previously seen during parallel-depth heavy squat (HS) and unloaded jump squat (JS) training. We hypothesize that greater peak and mean sEMG would be present in the proximal thigh during HS and distal thigh during JS matching previously observed patterns of hypertrophy.
Experimental Approach to the Problem
All subjects were familiarized to the experimental protocols before taking part in any testing. Subjects reported for testing on 2 separate days. Before each day of testing, subjects refrained from any lower body resistance exercise for at least 72 hours, any form of exercise or exertional activity for at least 48 hours, and caffeine and alcohol 12 hours before testing.
On the first day of testing, subjects underwent strength testing during which their 1 repetition maximum (1RM) was determined for the parallel-depth back squat exercise. On the second day of testing, subjects performed parallel-depth back squats (HS) with an external load of 80% of their 1RM and unloaded parallel-depth jump squats (JS) in a randomized order, during which surface electromyography (sEMG) of the subject's VL was recorded in the proximal-thigh, midthigh, and distal-thigh regions. Muscle activity was then compared within subjects, between exercises and muscular regions.
Ten healthy, physically active men (n = 6) and women (n = 4) (age: 20.7 ± 1.2 years, age range: 19–22 years, height: 172.1 ± 11.5 cm, mass: 75.6 ± 11.5 kg) participated in this study. Subjects had previous experience with the exercises performed in the study and were excluded from the study if they had any history of lower body musculoskeletal injury that would interfere with testing. This study was reviewed and approved by the Human Research Ethics Committee of the University and all subjects were informed of the benefits and risks of the investigation before signing an institutionally approved informed consent document to participate in the study.
After completing a warm-up consisting of 5 minutes of low-intensity cycling and a series of bodyweight and submaximal load (50, 75, and 90% of estimated 1RM) parallel-depth squats, subjects had their free-weight parallel-depth squat 1RM determined using the protocols outlined by Bachele and Earle (3). Subjects were permitted up to 6 attempts to find their 1RM with 3 minutes passive recovery between each attempt. The 1RM was defined as the greatest load that could be successfully lowered to and raised unassisted from parallel-depth while maintaining correct technique. Parallel-depth was defined as the line between greater trochanter of the femur and the intercondyloid space of the knee being parallel to the ground and assessed using real-time 2D motion capture.
On the second day of testing, subjects were outfitted with electrodes placed over their right VL at 33% (prox-), 50% (mid-), and 67% (dist-) of the distance between their anterior superior iliac spine and the lateral aspect of their patella (14) and used to determine regional muscle activity during HS and JS. At each location, the skin was shaved, abraded, and cleaned using a disposable alcohol wipe after which 2 self-adhesive surface electrodes with 0.5 cm diameter (PerformancePlus; Vermed, Bellows Fall, VT, USA) were placed over the VL muscle belly in a bipolar configuration. Interelectrode distance was set at 2.5 cm, and all signals were checked to ensure interelectrode impedance was <5 kΩ and the signal was free of exogenous noise during movement. Raw sEMG signals were recorded using a biopotential recorded (IX-BIO8, iWorx, Dover, NH, USA) to associated software (LabScribe 2.3; iWorx) at an analog-to-digital conversion rate of 2,000 Hz at 16-bit resolution after being amplified (×1,000). Recorded signals were filtered using a zero-lag dual-pass, sixth-order 10–500 Hz band-pass Butterworth filter. After which sEMG signals were full-wave rectified and relative muscle activity was determined by creating a linear envelope using a low-pass, fourth-order Butterworth filter with a cut-off frequency of 6 Hz (9).
Maximal Voluntary Isometric Squat (MVIS)
After being outfitted with electrodes, subjects repeated the warm-up previously performed before strength testing and then performed a ramped maximal voluntary isometric squat (MVIS) that was used to normalize sEMG signals collected throughout testing. To perform the MVIS, a rigid barbell was fixed to the squat rack so that the subject would be in a ¼ squat position (knee angle ∼160°) with the subject's feet positioned directly beneath the bar. The subject was then instructed to perform a ramped maximal intensity isometric squat with a 2-second buildup, 3-second maximal contraction, and 2-second build-down. During the MVIS prox-sEMG, mid-sEMG and dist-sEMG of the VL was recorded. Collected signals were filtered using a dual-pass, sixth-order 10–500 Hz band-pass Butterworth filter and maximal sustained muscle activity for each region was determined using a 500-milliseconds moving window RMS.
Heavy Squats and Jump Squats
After performing the MVIS, subjects performed HS with an external load of 80% 1RM and unloaded JS during which sEMG was recorded. Exercise order was randomized and counterbalanced between subjects. Both HS and JS were performed to a parallel-depth as previously defined. To ensure consistent depth and knee range of motion, reflective markers were placed over the greater trochanter of the femur, lateral intercondylar space of the knee, and lateral malleoli of the left limb and sagittal motions were recorded using 2D motion capture. Knee angle and bar depth were measured immediately after each repetition in video software (DartFish Express; DartFish, Alpharette, GA, USA) and feedback was given to subjects after each repetition. Only repetitions in which the subject's knee angle was within ±2.5° of parallel-depth were accepted for analysis.
Subjects performed 2 repetitions of each movement each separated by 1 minute of passive rest and were given 3 minutes of passive rest between exercises. Additional repetitions were performed if the subject did not obtain accepted depth or knee angle after 1 minute of additional rest.
Regional sEMG signals collected during the HS and JS were normalized to MVIS and reported as a percentage of MVIS. During each movement, the maximum and average sEMG values during the movement in each region were reported and used for comparison. The initiation of the movement was defined as when muscle activity of the linear envelope reached 5% of its maximum during that movement and did not drop below this value for at least 50 milliseconds. The end of the motion was defined as when muscle activity dropped below 5% after the movement was completed or in the case of a sustained submaximal contraction when muscle activity decay ended.
Because the purpose of the present study was to determine if muscle activity in the prox-thigh, mid-thigh or dist-thigh differed between HS and JS, a 3 × 2 (location and exercise) repeated measures ANOVA with Bonferroni post hoc tests was used to compare peak and mean muscle activity (p ≤ 0.05). Results are reported as means and standard deviations. Normality of data was confirmed via a Mauchly's test of sphericity. An a priori power analysis was used to calculate anticipated adequate sample size for a significance level of 0.05 and power of 0.7 from a one-tailed dependence comparison of means. Because the null hypothesis was proven a 2-tailed post hoc power analysis was used to confirm results, which found actual statistical power of 0.61, repeated measurements of regional sEMG were found to be reliable (ICC = 0.862–0.949). All statistical analyses were conducted using PASW 18.0.1 (IBM, New York, NY, USA) with the exception of power analysis and Cohen's D effect sizes, which were calculated using GPower 188.8.131.52 (University of Dusseldorf, Dusseldorf, NRW, Germany).
Results from the present study do not support the hypothesis that regional muscle activity as recorded by sEMG would match previously reported patterns of regional hypertrophy during HS and JS exercises. Specifically, we hypothesized that sEMG would be greater in the prox-region during HS and dist-region during JS.
Observed regional-specific muscle activity during the HS and JS is reported in Table 1 and depicted in Figure 1. Of particular interest, peak prox-sEMG was significantly greater in JS (411.3 ± 216.6%) than HS (287.5 ± 146.9%), proving the null-hypothesis (p = 0.033, ES = 0.794). In contrast, no significant differences were present in the mid-thigh or dist-thigh regions (p = 0.521 and 0.594, respectively). Mean muscle activity was not significantly different between exercises in the prox-region (p = 0.150, ES = 0.498), mid-region (p = 0.710, ES = −0.125), or dist-region (p = 0.979, ES = 0.009).
The 3 × 2 ANOVA also revealed a significant main effect of region on peak and mean sEMG (see Figure 2). Specifically: greater peak sEMG was observed in prox- (349.4 ± 191.0%) than mid-region (232.3 ± 112.0%) and dist-region (218.5 ± 103.8%) (p = 0.006 and 0.003, respectively) and greater mean sEMG was observed in prox-region (128.0 ± 75.6%) than mid-region (86.3 ± 42.0%) and dist-region (82.1 ± 38.1%) (p = 0.003 and 0.001, respectively). However, no significant differences were present between mid-region and dist-region in peak or mean sEMG (p = 1.000 and 1.000, respectively).
A significant main effect for exercise was also revealed for peak sEMG in that JS (292.7 ± 175.9%) was significantly greater than HS (240.8 ± 118.2%: p = 0.028); however, no such relationship was observed in mean sEMG (HS: 95.2 ± 50.0 and JS: 102.4 ± 64.9%: p = 0.296), see Figure 3.
Results from the present study demonstrated that peak and mean regional sEMG during HS and JS do not reflect previously observed patterns of regional hypertrophy. In support of this, peak prox-sEMG was greater in JS than HS (p = 0.033, ES = 0.794) and mean prox-sEMG was similar between exercises (p = 0.150), despite previous observations greater proximal hypertrophy in response to HS than JS training (10). Furthermore, in the Dist-region where previous research found only JS training was able to elicit significant hypertrophy, there were no significant differences between mean or peak sEMG during the HS and JS exercises (p = 0.979 and 0.594, respectively). Based on these observations, amplitude-derived measures of sEMG do not appear to be a viable tool in predicting selective regional hypertrophy during dynamic exercises with matched movement patterns but differing training outcomes.
The present results are in contrast to Wakahara et al. (24,25) who observed greater regional activity of the triceps brachii, measured using T2-weighted magnetic resonance imaging (T2-MRI), in regions of the muscle in which greater regional hypertrophy occurred after a training intervention with the same exercise. However, it should be noted that T2-MRI, unlike sEMG, is not biased by tissue depth because sEMG overly represents superficial fibers that are closer to the surface electrodes (12,17,22). Lexell et al. (18) observed that fiber distribution of the VL is characterized by greater abundance of type II fibers closer to the skin; therefore, it is possible that sEMG signals overly express this fiber type over type I fibers (17). This may be a limitation when comparing HS and JS because it has previously been postulated that the contribution of type II fibers to powerful, high-speed movements like the JS is greater than longer duration contractions like the HS because of shorter movement duration (27) or possible selective recruitment (21). For these reasons, although T2-MRI and sEMG are strongly correlated and considered valid methods of estimating muscle activation (2), T2-MRI may be a more appropriate measure of muscle activation when comparing exercises with expected differential contribution of type II fibers. However, more research into T2-MRI is necessary to investigate such a possibility.
It should also be noted that T2-MRI is a postexercise measure used to estimate longitudinal activity and thus is sensitive to cumulative activation of tissue throughout an exercise and across multiple sets and repetitions preformed (2). Because of this, Adams et al. (2) validated T2-MRI to time-dependent integrated sEMG signals instead of mean or peak sEMG. However, because integrated sEMG represents the time integrate of mean sEMG and JS was of shorter movement duration than HS, this measure would be inappropriate for the purposes of our comparison and would demonstrate a similar regional pattern to the reported means.
The results from the present study were similar to those previously observed by Miyamoton et al. (19) who observed that sEMG did not reflect previously observed regional hypertrophy in response to knee extensor training. Of interest to note is that Miyamoton et al. (19) did however find that tissue oxygenation as measured by near infrared spectroscopy (NIRS) did follow similar patterns to previously observed hypertrophy (19,20), suggesting either a technical limitation of sEMG in measuring true muscle activation, or that NIRS measured muscle oxygenation is more representative to muscle hypertrophy than sEMG. If the latter is true, previously observed velocity-dependent differences in intramuscular blood flow may be one possible explanation (7). However, because it is unknown if regional intramuscular pressures differ between high-force and high-speed contractions, more research is needed in this area.
In the VL, standard electrode placement is represented by the Dist-location used in the current study (14). Because no differences in muscle activity were recorded at this location between exercises but differences in hypertrophy have been observed, traditional electrode placement would be unable to predict anticipated future training adaptations. It should be noted that placement of electrodes in the mid-region and prox-region may have been influenced by the location of neuromuscular innervation zones (23). However, because of the within subject design and matched movement patterns of HS and JS, it is unlikely that these directly affected our results. Furthermore, even if sEMG signals were affected by such innervation zones, it does not dispute the current findings that sEMG is not an effective tool to predict regional hypertrophy in dynamic motions.
Because sEMG measures electrical currents relative to the electrode alignment and placement of the skin, signals are independently influenced by fiber alignment (11), fascicle angle (11,12), and movement of the muscle underneath the skin (8,13,16), all of which will change as joint angle changes. To control for this, disparity movement range of motion was matched between HS and JS. However, as peak muscle activity may have occurred at different joint positions, it cannot be concluded that this did not affect our results. One possible method to rectify this limitation would be to normalize sEMG signals to specific joint position using either a series of isometric contractions (4,26) or an isokinetic contraction (8,16).
In summary, although sEMG is a relatively inexpensive and effective tool to measure relative muscle activity during a given exercise, regional peak and mean sEMG observed during training are not predictive to anticipated regional hypertrophy and thus not a valid tool to be used for these purposes. This limitation of sEMG may be related to its sensitivity to anatomical features of the muscle or overrepresentation of type II fibers in the signal. The use of alternative estimates of muscle activation such as T2-MRI or measurement of muscle oxygenation using NIRS may be more representative to preferential regional hypertrophy. However, more research is necessary to explore these technologies in context to the HS and JS exercises.
Muscle activation during exercise is innately linked to muscle hypertrophy from chronic training with that exercise. Although intramuscular muscle activity as measured by NIRS and T2-MRI has been linked to patterns of intramuscular hypertrophy, sEMG displays dissimilar patterns and thus cannot be used to predict patterns of regional hypertrophy. Such differences are likely due to technological limitations of sEMG such as sampling bias and the influence of heterogeneous muscle composition and structure on the collected signals.
1. Abe T, Kumagai K, Brechue WF. Fascicle length of leg muscles is greater in sprinters than distance runners. Med Sci Sport Exerc 32: 1125–1129, 2000.
2. Adams GR, Duvoisin MR, Dudley GA. Magnetic resonance imaging and electromyography as indexes of muscle function. J Appl Physiol 73: 1578–1583, 1992.
3. Baechle TR, Earle RW. Essentials of Strength Training and Conditioning (3rd ed.). Champaign, IL: Human Kinetics, 2008.
4. Bandy WD, Hanten WP. Changes in torque and electromyographic activity of the quadriceps femoris muscles following isometric training. J Orthop Sports Phys Ther 73: 455–465, 1993.
5. Blazevich AJ, Gill ND, Zhou S. Intra- and intermuscular variation in human quadriceps femoris architecture assessed in vivo. J Anat 209: 189–210, 2006.
6. Cormie P, McBride JM, McCaulley GO. Power-time, force-time, and velocity-time curve analysis of the countermovement jump: Impact of training. J Strength Cond Res 23: 177–186, 2009.
7. Dennis R, Wilkinson J, De Vito G. Influence of angular velocity on vastus lateralis and rectus femoris oxygenation dynamics during knee extension exercises. Clin Physiol Funct Imaging (31): 352–357, 2011.
8. Earp JE, Newton RU, Cormie P, Blazevich AJ. Knee angle-specific EMG
normalization: The use of polynomial based EMG
-angle relationships. J Electromyogr Kinesiol 23: 238–244, 2013.
9. Earp JE, Newton RU, Cormie P, Blazevich AJ. The influence of loading intensity on muscle-tendon unit behavior during maximal knee extensor stretch shortening cycle exercise. Eur J Appl Physiol 114: 59–69, 2014.
10. Earp JE, Newton RU, Cormie P, Blazevich AJ. Inhomogeneous quadriceps femoris hypertrophy
in response to strength and power training. Med Sci Sports Exerc 47(11): 2389–2397, 2015.
11. Farina D, Cescon C, Merletti R. Influence of anatomical, physical, and detection-system parameters on surface EMG
. Biol Cybern 86: 445–456, 2002.
12. Farina D, Merletti R, Nazzaro M, Caruso I. Effect of joint angle on EMG
variables in leg and thigh muscles. IEEE Eng Med Biol Mag 10: 62–71, 2001.
13. Hahn D. Lower extremity extension force and EMG
properties as a function of knee angle and their relation to joint torque. J Strength Cond Res 56: 1622–1631, 2011.
14. Hermens JH, Freriks B. The State of the Art on Sensors and Sensor Placement Procedures for Surface Electromyography: A Proposal for Sensor Placement Procedures. Netherlands: Roessingh R&D, 1996.
15. Irish SE, Millward AJ, Wride J, Haas BM, Shum GL. The effect of closed-kinetic chain exercises and open-kinetic chain exercise on the muscle activity of vastus medialis oblique and vastus lateralis. J Strength Cond Res 24: 1256–1262, 2010.
16. Kellis E, Balzopoulos V. Agonist and antagonist moment and EMG
-angle relationship during isokinetic eccentric and concentric exercise. Isokinetics Exerc Sci 6: 79–87, 1996.
17. Knight CA, Kamen G. Superficial motor units are larger than deeper motor units in human vastus lateralis muscle. Muscle Nerve 31: 475–480, 2005.
18. Lexell J, Henriksson-Larsen K, Sjostrom M. Distribution of different fibre types in human skeletal muscles: A study of cross-sections of whole m. vastus lateralis. Acta Physiol Scand 117: 115–122, 1983.
19. Miyamoto N, Wakahara T, Ema R, Kawakami Y. Non-uniform muscle oxygenation despite uniform neuromuscular activity within the vastus lateralis during fatiguing heavy resistance exercise. Clin Physiol Funct Imaging 33: 463–469, 2013.
20. Narici MV, Hoppeler H, Kayser B, Landoni L, Claassen H, Gavardi C, Conti M, Cerretelli P. Human quadriceps cross-sectional area, torque and neural activation
during 6 months strength training. Acta Physiol Scand 157: 175–186, 1996.
21. Nardone A, Schieppati M. Shift of activity from slow to fast muscle during voluntary lengthening contractions of the triceps surae muscles in humans. J Physiol 409: 363–381, 1988.
22. Rainoldi A, Nazzaro M, Merletti R, Farina D, Caruso I, Gaudenti S. Geometrical factors in surface EMG
of the vastus medialis and lateralis muscles. J Electromyogr Kinesiol 10: 327–336, 2000.
23. Simmons DG. Do endplate noise and spikes arise from normal motor endplates?. Am J Phys Med Rehabil 80: 134–140, 2001.
24. Wakahara T, Fukutani A, Kawakami Y, Yanai T. Nonuniform muscle hypertrophy
: Its relationship to muscle activation
in training session. Med Sci Sports Exerc 45: 2158–2165, 2013.
25. Wakahara T, Miyamoto N, Sugisaki N, Murata K, Kanehisa H, Kawakami Y, Fukunaga T, Yanai T. Association between regional differences in muscle activation
in one session of resistance exercise and in muscle hypertrophy
after resistance training. Eur J Appl Physiol (112): 1569–1576, 2012.
26. Watanabe K, Akima H. Effect of knee joint angle on neuromuscular activation
of the vastus intermedius muscle. Scand J Med Sci Sports 21: 412–420, 2011.
27. Zehr EP, Sale DG. Ballistic movement: Muscle activation
and neuromuscular adaptation. Can J Appl Physiol 19: 363–378, 1994.