Over the past few years, the data indicating that low intensity training, when taken to the point of concentric failure, produces comparable hypertrophy to high load training has literally piled up. There’s a growing collection of studies in both trained and untrained individuals, considering whole muscle growth and that of specific fiber types, all of which demonstrate the comparable effectiveness of various training intensities to promote muscle growth (1-6).
Despite the growing collection of studies on our outcome of interest (hypertrophy), there’s still resistance to the idea that low intensity training (i.e. 30%-1RM) can be as effective as high intensity training for muscle mass (>65%-1RM). While there are studies that do not support our new-found understanding (7,8), much of the resistance stems from the idea that light weights do not recruit as much muscle as is required to move heavy ones (9).
If the idea that a muscle fiber must be recruited in order to adapt to training is true, then any stimulus that fails to stimulate a good proportion of the motor unit pool within a muscle would likely produce less growth than one that does. Score one for heavy weights, right?
Not so fast.
A brief overview of surface electromyography (EMG)
The intent of this article is not to give an in-depth description of the methodology and technical aspects of the use of EMG in experimental research. For those looking for more detail, I’d recommend reviewing Chris Beardsley’s EMG entry in his encyclopedia of Strength and Conditioning Research.
At the simplest level, electromyography involves the recording of electrical activity within the muscle (10). With sport and exercise science, this technique is often used to determine whether muscles contribute, or are active, during a given movement/exercise, how active a muscle is relative to the activity during a maximum contraction, and the relationship of the activity between muscles during human movement (10).
In response to stimulation by an innervating alpha motor-neuron, depolarization along the muscle fiber membrane ultimately simulates muscle contraction. Surface EMG electrodes, when applied to the skin above the muscle, are able to detect alterations in electrical potential, and are indicative of the electrical activity within the muscle (11). The “activity” in these electrodes is limited only to a certain number of fibers adjacent to the electrode, termed the receptive field, yet is often thought to accurately represent the collective activity within the entire muscle.
There are two recent studies that have misapplied EMG to imply that heavier training to failure recruits more motor units than lighter training to failure, but first, let’s look at a simple analogy to understand what using surface EMG amplitude actually tells you in the context of estimating motor unit recruitment.
Imagine you are outside a party listening to everyone talking through a door. I don’t tell you how many people are in the room, but I have them all yell as loud as they can at the same time, the equivalent of the maximum voluntary contraction in an EMG experiment.
After giving you an idea of the maximum volume, I have you listen at the door to a given number of people talking, but again I don’t tell you how many. Could you tell me how loud it was relative to the maximum conversation volume? It’s possible. You may not be able to tell me that 50/100 people were talking, but it’s possible you could tell that it was 50% of the volume of the maximum condition.
Now I have you listen for an extended period of time as some people talk, others start and stop, and new people chime in. Could you tell me how many people talked relative to the maximum? It’s possible that you could again tell me what the peak volume was relative to our initial maximum.
What if I brought you into the room and asked you if you could tell me exactly who talked out of everyone at the party? Or if you could tell me how many of them had talked at some point in the conversation? You probably couldn’t do that.
This is the equivalent to EMG amplitude for assessing motor unit recruitment. You can describe the EMG amplitude during a task relative to the maximum, as is commonly done and with more precision than our party analogy. While there are more complicated methods to determine with greater specificity motor unit contributions during tasks, ultimately, surface EMG amplitude cannot tell us which motor units were used, or how many were used at some point during the measurement (the equivalent of identifying who was talking at the party).
Comparing EMG amplitude during low- and high-intensity training
Two recent studies have used EMG amplitude to make conclusions regarding the effectiveness of low intensity strength training. In the first, Looney et al (9) had 10 resistance-trained men perform Smith machine squats to parallel depth while measuring surface EMG of the vastus lateralis and medialis muscles and participant ratings of perceived exertion (RPE). Participants completed two experimental protocols, one involving two sets of submaximal repetitions at 50 and 70%-1RM followed by a dropset to fatigue (90%, 70%, 50%-1RM), and the other involving two submaximal sets at 50 and 70%-1RM and one to failure at 50%-1RM.
In general, submaximal sets had lower EMG amplitudes than maximal sets at an equivalent intensity in both muscles. For training intensity, EMG amplitude was greater at higher training intensities in both the VL and the VM. RPE was comparable across training intensities when taken to failure and was only lower when training short of failure at 50%-1RM or following the 50%-1RM dropset.
These results lead Looney et al (9) to conclude:
The results of this investigation indicate that using higher external resistance is a more effective means of maximizing muscle activation than increasing the number of repetitions performed. Accordingly, previous recommendations for the use of heavier loads during resistance training programs emphasizing strength and hypertrophy are further supported.” (emphasis mine)
In the second study, Jenkins et al (12) had 18 participants complete two training conditions in which they completed three sets of leg extension to failure with either 80 or 30% of their one-repetition maximum (1RM). Surface EMG was collected through the trials, and ultrasound was used to detect changes in muscle swelling (cross sectional area, echo intensity) after each loading condition.
Not surprisingly, volume and work as well as time under tension were greater when training at 30%-1RM than 80%-1RM. EMG amplitude was consistent across the three sets at 80%-1RM. At 30%-1RM, EMG amplitude increased over the course of the three sets, but was still lower than the sets at 80%-1RM. Muscle cross-sectional area increased to a greater extent when training at 30%-1RM, with no difference detected in echo intensity, which is often used as a marker of exercise-induced swelling (edema).
Jenkins et al (12) are much more conservative in their message, choosing to acknowledge the existing literature showing comparable growth between the conditions. As they also measured other variables with a differential response between conditions, the authors suggest that low- and high-intensity training may arrive at comparable growth, but in response to differing cellular stimuli, stating:
Our results demonstrated that muscle activation was 38–62 % lower for the superficial quadriceps femoris muscles during resistance exercise at 30% than at 80% 1RM, despite fatigue- induced increases in EMG AMP (amplitude) at both 80 and 30% 1RM. However, the cumulative volume and muscle activation, time under load, and increases in mCSA from pre- to post- exercise (i.e., muscle swelling) were greater, while the decreases in EMG MPF were more pronounced for 30% 1RM. These fundamental differences in fatigue manifestations may help explain the unexpected chronic adaptations in hypertrophy vs. strength observed in previous studies.”
The difference in wording between the two groups is important: The first suggests that heavier is better (9), whereas the latter (12) attempts to integrate their results with the fact that the present literature does not support a differential hypertrophic response between low- and high-intensity training (1). Differences in EMG may exist, but it’s possible that other mechanisms contribute to the hypertrophic “potential” of low-intensity training.
Putting these results in context – or why we shouldn’t jump to any conclusions
I was contacted a few months back to assist Andrew Vigotsky – an active, upcoming researcher in the health and fitness industry – to address both of these studies in a letter to the editor. This is an important part of the scientific process, where researchers are allowed to criticize (constructively) each other’s work in the hopes of fostering a greater understanding of the issue. Unfortunately, these letters are often published months after the original study and rarely receive the attention they deserve. For those interested, the letters to Looney et al (9) can be found here (13), and Jenkins et al (12) here (14). Andrew has also written more on this on Bret Contreras’s site here.
We had some concerns with how these studies are used to support the argument that high loads are required to maximize muscular hypertrophy, or that high loads are inherently more effective at promoting hypertrophy than lower loads. These concerns can be broken down into the following questions:
Is surface EMG amplitude an appropriate measure of motor unit recruitment?
The normalization and interpretation of EMG signals is a complicated area, and one that even researchers get wrong (15). These issues lead Carlo De Luca (16), a prominent researcher in the field of neuromuscular physiology, to state:
To its detriment, electromyography is too easy to use and consequently too easy to abuse.”
The surface EMG signal is inclusive of multiple components which limit our ability to determine what happens at the motor unit level. The EMG signal is representative of both neural and peripheral (muscular) components, all of which may be subject to change during dynamic, fatiguing contractions (17) like those used by both Jenkins et al (12) and Looney et al (9). As discussed by Vigtosky et al (13,14), changes in the characteristics of intracellular action potentials within muscle during fatiguing actions may alter surface EMG signals (17), independent of any changes in motor unit recruitment.
There are more rigorous EMG methodologies that tell us more about motor unit recruitment, yet these methods are rarely used in exercise physiology. Suffice it to say, that’s why most researchers play it safe, and use the general term “activation” as opposed to recruitment. Yet the connotation is that greater “activation” implies greater recruitment, and to some, a greater hypertrophic stimulus.
Should EMG amplitude be equivalent with fatigue between high and low-load training?
The relationship of motor unit recruitment to fatigue is complex, and our understanding in this area is constantly evolving. The comparable hypertrophy between low and high-intensity training is thought to be a consequence of motor-unit cycling (18,19). Motor units have different recruitment thresholds that decrease under fatiguing conditions. While it is likely that high-intensity training requires greater simultaneous activation of motor units due to the higher force requirements, low-intensity training may recruit just as many motor units, just at different times. During a fatiguing contraction, some motor units will remain active, others will stop working, and new ones will start. With the prolonged time-under-tension of low-load training, it’s possible that motor unit cycling allows a comparable population of motor units to be recruited during the task, just with fewer being recruited at the same time.
In the high-intensity condition, motor unit recruitment is likely higher initially to meet the increased force demands of the task relative to low intensity, otherwise you’d fail to lift the weight. It is entirely plausible that a greater number of motor units are recruited at a single point in time, which produce at least a sufficient amount of force to complete the task. This greater simultaneous recruitment could explain the elevated initial EMG amplitude as compared with a lower force task observed by both Jenkins et al and Looney et al (9,12), which would need less simultaneously active units given the lower force requirements at any point in time.
So if high-intensity loads required greater simultaneous recruitment, and low-intensity loads rely on an asynchronous recruitment strategy, it could be plausible that there would be at least some level of difference in EMG amplitude between the two conditions, which would decrease with increasing fatigue. Despite these differences, it’s possible that the exact same number of motor units were recruited between the high and low-load conditions; it’s just that the high-load condition recruited more of them at the same time.
Should data from an acute experimental design be given greater weight than measurements of the direct outcome of interest (muscle growth) in longer-term training studies?
This is perhaps the most important argument, as it is a common problem in the health and fitness world. We often use proxies of our actual primary outcomes to determine the effectiveness of a given training program, and in doing so, end up ignoring data from well-designed studies. If we have data from multiple training studies, directly measuring any form of muscle growth (CSA, thickness) or even just lean-body mass, this should be given greater weighting in our decisions than acutely measured secondary variables.
Attempting to use any other measure to predict muscle growth requires a substantial burden of evidence to establish its predictive value. I’m not saying we should ignore acute protein synthesis, EMG, or signaling protein data, but if it doesn’t reconcile with data on our primary, relevant outcome measures (hypertrophy), the decision you make should be an easy one.
In this case, we have two studies demonstrating a difference in EMG amplitude (9,12), that some have used to suggest that motor unit recruitment is greater with high-intensity training, and therefore a superior hypertrophic stimulus to low intensity training. Yet when the training studies have been performed, and muscle growth directly measured, we know this isn’t the case (1-6).
If you find yourself questioning the hypertrophic effectiveness of low-intensity training, I’d suggest you’re asking the wrong question. Rather, we should be questioning what acute studies assessing surface EMG amplitude are actually telling us about the hypertrophic adaptations to training. I’m not questioning the differences between the two conditions observed by both Jenkins et al (12) and Looney et al (9), but given the data demonstrating comparable whole muscle and fiber-type specific growth, I’d suggest it isn’t telling us much, at least as far as muscle growth is concerned.
- Surface EMG amplitude is often used as a measure of motor unit recruitment, but it cannot provide any specifics regarding the recruitment of specific populations of motor units. This doesn’t mean we shouldn’t use EMG, but should be aware of its limitations when interpreting it.
- Two recent studies (9,12) have demonstrated reduced EMG amplitude with low- as compared to high-intensity training, and this is often used to suggest low-intensity training would result in less muscle growth than high-intensity training.
- Despite acute differences in EMG amplitude, numerous studies have demonstrated comparable muscle growth between high and low-intensity training, in trained and untrained individuals, considering both whole muscle and fiber-type specific growth. Data on our primary outcome trumps indirect physiological arguments every day of the week.
- There is no evidence that the simultaneous activation of a large number of motor units is superior for hypertrophy as compared to a more time-intensive, gradual recruitment strategy. When designing a hypertrophy training program there is no need to pick specific program parameter combinations (intensity, tempo, etc…) outside of exercise selection to maximize surface EMG amplitude.
- Mitchell CJ, Churchward-Venne TA, West DWD, Burd NA, Breen L, Baker SK, et al. Resistance exercise load does not determine training-mediated hypertrophic gains in young men. J Appl Physiol. 2012 Jul;113(1):71–7.
- Ogasawara R, Loenneke JP, Thiebaud RS, Abe T. Low-load bench press training to fatigue results in muscle hypertrophy similar to high-load bench press training. International Journal of Clinical Medicine. 2013 Feb;4:114–21.
- Schoenfeld BJ, Peterson MD, Ogborn D, Contreras B, Sonmez GT. Effects of Low- Versus High-Load Resistance Training on Muscle Strength and Hypertrophy in Well-Trained Men. J Strength Cond Res. 2015 Apr 3;:1.
- Lamon S, Wallace MA, Léger B, Russell AP. Regulation of STARS and its downstream targets suggest a novel pathway involved in human skeletal muscle hypertrophy and atrophy. J Physiol (Lond). 2009 Apr 15;587(Pt 8):1795–803.
- Léger B, Cartoni R, Praz M, Lamon S, Dériaz O, Crettenand A, et al. Akt signalling through GSK-3beta, mTOR and Foxo1 is involved in human skeletal muscle hypertrophy and atrophy. J Physiol (Lond). 2006 Nov 1;576(Pt 3):923–33.
- Schoenfeld BJ, Ratamess NA, Peterson MD, Contreras B, Sonmez GT, Alvar BA. Effects of different volume-equated resistance training loading strategies on muscular adaptations in well-trained men. J Strength Cond Res. 2014 Oct;28(10):2909–18.
- Campos GER, Luecke TJ, Wendeln HK, Toma K, Hagerman FC, Murray TF, et al. Muscular adaptations in response to three different resistance-training regimens: specificity of repetition maximum training zones. 2002 Nov;88(1-2):50–60.
- Holm L, Reitelseder S, Pedersen TG, Doessing S, Petersen SG, Flyvbjerg A, et al. Changes in muscle size and MHC composition in response to resistance exercise with heavy and light loading intensity. J Appl Physiol. 2008 Nov;105(5):1454–61.
- Looney DP, Kraemer WJ, Joseph MF, Comstock BA, Denegar CR, Flanagan SD, et al. Electromyographical and Perceptual Responses to Different Resistance Intensities in a Squat Protocol: Does Performing Sets to Failure With Light Loads Recruit More Motor Units? J Strength Cond Res. 2015 Aug 10.
- Massó N, Rey F, Romero D, Gual G, Costa L. Surface electromyography applications in the sport. Apunts Med Esport. 2010;45(165):121–30.
- Reaz MBI, Hussain MS, Mohd-Yasin F. Techniques of EMG signal analysis: detection, processing, classification and applications (Correction). Biol Proced Online. Springer-Verlag; 2006;8(1):163–3.
- Jenkins NDM, Jenkins NDM, Housh TJ, Bergstrom HC, Bergstrom HC, Cochrane KC, et al. Muscle activation during three sets to failure at 80 vs. 30 % 1RM resistance exercise. Eur J Appl Physiol. 2015 Jul 10.
- Vigotsky AD, Beardsley C, Contreras B, Steele J, Ogborn D, Phillips SM. Greater electromyographic responses do not imply greater motor unit recruitment and “hypertrophic potential” cannot be inferred. J Strength Cond Res. 2015 Dec;:1–6.
- Vigotsky AD, Ogborn D, Phillips SM. Motor unit recruitment cannot be inferred from surface EMG amplitude and basic reporting standards must be adhered to. Eur J Appl Physiol. 2015 Dec 24.
- Enoka RM. Inappropriate interpretation of surface EMG signals and muscle fiber characteristics impedes progress on understanding the control of neuromuscular function. J Appl Physiol. American Physiological Society; 2015 Jul 9;119(12):jap.00280.2015–1518.
- De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech. 1997.
- Dimitrova NA, Dimitrov GV. Interpretation of EMG changes with fatigue: facts, pitfalls, and fallacies. J Electromyogr Kinesiol. 2003 Feb;13(1):13–36.
- Adam A, De Luca CJ. Recruitment order of motor units in human vastus lateralis muscle is maintained during fatiguing contractions. J Neurophysiol. 2003 Nov;90(5):2919–27.
- Burd NA, Mitchell CJ, Churchward-Venne TA, Phillips SM. Bigger weights may not beget bigger muscles: evidence from acute muscle protein synthetic responses after resistance exercise. Appl Physiol Nutr Metab. 2012 Jun;37(3):551–4.