We’ve written about the effects of range of motion on muscle growth quite a few times in MASS Research Review, tentatively concluding that the beneficial hypertrophy effects of training through a full range of motion are primarily driven by the impact of training at longer muscle lengths. In other words, training through a full range of motion is generally more effective than employing a partial range of motion where the prime movers are only trained at short muscle lengths, but employing a partial range of motion where the prime movers are trained at long muscle lengths seems to be just as effective as training through a full range of motion (2).
If training at longer muscle lengths is the critical factor, that opens the door to a new question: are exercise variations that allow you to train at longer muscle lengths inherently better for hypertrophy than exercise variations that force you to train at shorter muscle lengths? A previous study by Maeo and colleagues (3) seemed to answer this question with a resounding, “yes.” It compared the effects of seated versus lying hamstring curls. Since there are multiple biarticular muscles crossing the knee and hip, manipulating hip angles changes the muscle lengths of several knee flexors, without affecting knee flexion range of motion. Namely, seated hamstring curls train three of the four heads of the hamstrings at longer muscle lengths than lying hamstring curls. In keeping with the hypothesis that training at longer muscle lengths is better for hypertrophy, all three biarticular heads of the hamstrings (the long head of the biceps femoris, the semitendinosus, and the semimembranosus) experienced more hypertrophy following seated hamstrings curls than lying hamstring curls, whereas the sole monoarticular head of the hamstrings (the short head of the biceps femoris) experienced roughly the same amount of hypertrophy following both seated and lying hamstrings curls.
A recent study by the same group of researchers aimed to see if this same principle would generalize to another muscle group: the triceps (1). Employing a within-subject unilateral design, 21 healthy but untrained subjects completed 12 weeks of triceps training – one arm performed overhead triceps extensions on a cable machine, and one arm performed cable pushdowns. Both arms trained through 90 degrees of elbow flexion, and subjects performed 5 sets of 10 reps for each exercise with a controlled cadence (2-second eccentrics and 2-second concentrics), twice per week. Training loads increased by 5% when a subject could complete all five sets of 10 reps with a particular load. 1RM strength was assessed pre- and post-training, and triceps muscle volume was assessed pre- and post-training via MRI. The researchers hypothesized that overhead triceps extensions would result in more growth of the long head of the triceps since overhead triceps extensions train the long head of the triceps at longer muscle lengths (Figure 1). Furthermore, they hypothesized that both triceps exercises would produce similar growth for the monoarticular heads of the triceps (the lateral and middle heads), since shoulder position shouldn’t influence the muscle lengths of the two monoarticular heads.
Relative increases in training loads (Figure 2) and 1RM strength were similar between arms, though the arms performing pushdowns trained with heavier absolute loads throughout the study.
In keeping with the researchers’ hypothesis, overhead triceps extensions led to larger increases in muscle volume for the long head of the triceps (+28.5% versus +19.6%; p <0.001). However, in conflict with the researchers’ hypothesis, overhead triceps extensions also led to larger increases in muscle volume for the lateral and middle heads of the triceps (+14.6% versus 10.5%; p = 0.002). You can see these results in Figure 3.
Before digging into the results of the present study, it’s worth mentioning a prior study investigating the effects of pushdowns versus overhead triceps extensions. A 2018 study by Stasinaki and colleagues assessed hypertrophy in only the long head of the triceps, also employing a within-subject design (4). One arm performed pushdowns with an elbow range of motion spanning from 10 to 90 degrees of elbow flexion, and the other arm performed overhead triceps extensions with an elbow range of motion spanning from 150 to 70 degrees of elbow flexion. In other words, the arms performing pushdowns trained at muscle lengths that were comparable to the pushdown arms in the present study, but the arms performing overhead triceps extensions trained at even longer muscle lengths (for the long head of the triceps) than the arms performing overhead triceps extensions in the present study. Unlike the present study, the study by Stasinaki and colleagues found that overall increases in cross-sectional area for the long head of the triceps was similar following both training interventions. Overhead triceps extensions were non-significantly better for growing the distal region of the long head of the triceps (closer to the elbow), while pushdowns were non-significantly better for growing the proximal region (closer to the shoulder). I think there are three factors that could explain why the results differed between these two studies.
First, there may have just been some issues with measurement error in the study by Stasinaki and colleagues (4). The present study by Maeo and colleagues (1) measured changes in triceps size using MRI, which produces crisp, clear images. The Stasinaki study used B-mode ultrasound, which certainly can produce clear images, but image quality varies between ultrasound devices. Figure 1 in the Stasinaki study shows a representative scan with the ultrasound device used in the study, and it’s pretty clear that some judgment calls would need to be made for determining the boundaries of the long head of the triceps. Second, in the Stasinaki study, subjects always completed all sets of pushdowns before they performed overhead triceps extensions. In the present study by Maeo and colleagues, subjects alternated which arm was trained first. So, it’s possible that subjects in the Stasinaki study simply trained a bit harder when performing pushdowns, since they always trained pushdowns when they were fresh. Third, and most importantly, it’s possible that the overhead triceps extensions in the Stasinaki study were performed at too long of muscle lengths. When muscle fibers are stretched to more than about 150% of resting length, there are too few actin-myosin crossbridges to create much active tension. The present study by Maeo and colleagues presents some muscle modeling results, suggesting that the long head of the triceps is effectively “tapped out” once you reach 90 degrees of elbow flexion in an overhead triceps extension (Figure 1). Since subjects trained overhead triceps extensions from 150 to 70 degrees of elbow flexion in the Stasinaki study, it’s likely that the long head of the triceps was barely producing any active force throughout most of the range of motion being trained. In other words, instead of increasing the training stress on the long head of the triceps, the overhead triceps extensions performed in the Stasinaki study may have reduced the training stress on the long head of the triceps, by putting it in an over-stretched position.
Turning our attention to the study at hand, I’m not going to say much more about the finding that overhead triceps extensions were more effective for promoting hypertrophy in the long head of the triceps. That comports with prior research, finding that training at longer muscle lengths is generally superior for hypertrophy. However, I do want to discuss the finding that overhead triceps extensions also led to more muscle growth in the lateral and middle heads of the triceps, because it’s both a very strong and a very surprising finding.
First, it’s worth explaining why it’s a strong finding. Since exercise selection didn’t affect the muscle lengths of the lateral and middle head of the triceps, a lot of people (including the authors) expected that overhead triceps extensions and pushdowns would be similarly effective for growing the monoarticular heads of the triceps. So, I’ve seen quite a few people on social media writing this finding off as a fluke that we shouldn’t put much stock in. However, purely from a scientific and statistical perspective, this finding has a lot going for it. First, the study design itself is great – with a within-subject unilateral design, each subject can serve as their own control, so your results won’t be impacted by things like a failure of your randomization protocol. Basically, both of your “groups” are guaranteed to have the same lifestyles, genetics, nutrition, etc. (i.e., your right arm doesn’t sleep more or eat better than your left arm), which isn’t necessarily guaranteed with a parallel-groups design. Second, hypertrophy was assessed via MRI, which is the gold standard for assessing changes in whole muscle size in vivo. Third, the p-value for the comparison of monoarticular triceps hypertrophy was very low (p = 0.002), implying that overhead triceps extensions didn’t just lead to more growth of the middle and lateral heads of the triceps on average – overhead triceps extensions predictably led to more growth for the vast majority of individuals.
In absolute terms, the volume of the middle and lateral heads of the triceps increased by 42.1 ± 33.4cm3 in the arms doing overhead triceps extensions, and by 30.4 ± 26.9cm3 in the arms doing pushdowns. If this study employed a parallel-groups design, that difference between groups wouldn’t be statistically significant with 21 subjects per group. If you run an unpaired t-test on those change scores, you’ll come up with a p-value of 0.22. However, since the p-value for this comparison was very low (p = 0.002), that means that overhead triceps extensions consistently produced superior results in this study. For more on why consistency matters when you’re dealing with correlated, paired data, see the “criticisms and statistical musings” section of this prior MASS article. But, in short, both the magnitude and consistency of a finding matter when evaluating how durable the finding is likely to be. If an intervention produces 40% more muscle growth, on average, but it only produces more muscle growth for 60% of individuals, you might not be dealing with a particularly reliable and generalizable finding. However, if another intervention produces 40% more muscle growth, on average, but it produces more muscle growth for 90% of individuals, there’s a very good chance that it is truly a generalizably superior intervention.
Now, let’s turn our attention to potential explanations for this finding. The middle and lateral heads of the triceps were trained at the same muscle lengths by both exercises, and the general resistance curves would be been similar for both exercises (hardest at the start of the concentric when the arm is parallel to the floor, and easiest at lockout), but overhead triceps extensions reliably produced more muscle growth in the monoarticular heads of the triceps. What could explain these results?
The authors of the study put forth two potential explanations, and I’d like to add a third (which, admittedly, may be a bit of a stretch).
Their first explanation is purely mechanical: overhead triceps extensions put the long head of the triceps in a very lengthened position, where it isn’t capable of producing much active force. Therefore, the middle and lateral heads of the triceps would have needed to generate more force (especially at the start of the concentric) during overhead triceps extensions to “make up for” the diminished contributions of the long head. This explanation initially makes intuitive sense, but the more I’ve thought about it, the less plausible it seems. Quite simply, the nervous system doesn’t generally have issues recruiting monoarticular muscles for single-joint exercises. The subjects were training to failure or near failure on all of their sets, so the monoarticular heads of the triceps would have already been producing as much force as they were capable of (given the per-set rep targets) with both exercises. Since the long head of the triceps couldn’t produce as much force during overhead triceps extensions, the subjects just performed overhead triceps extensions with a lower load (Figure 2). Basically, reducing the force output of the long head of the triceps didn’t make the monoarticular heads of the triceps produce even more force; it just reduced total force output.
The authors’ second explanation is that overhead triceps extensions may have increased hypoxic stress for all heads of the triceps. Way back in Volume 1 of MASS, I wrote about a study by Goto and colleagues which suggested that “constant tension” training may lead to greater muscle growth by increasing hypoxic stress during training (5). There’s also some evidence suggesting that training in low-oxygen environments may lead to greater muscle growth due to a similar mechanism (6). The precise ways in which hypoxia increases muscle hypertrophy isn’t fully elucidated, but I do think this hypothesis has some legs. When your arms are overhead, they receive less arterial blood flow (because the flow of blood is being counteracted by gravity; when your arms are to your side, gravity instead aids in arterial blood flow), which could lead to greater hypoxia. So, via this potential mechanism, overhead triceps extensions may lead to greater muscle growth of the monoarticular heads of the triceps for a reason that’s completely unrelated to training at longer versus shorter muscle lengths.
My third tentative explanation is that changing the length of the long head of the triceps may have actually affected tension in the monoarticular heads of the triceps, since all three heads of the triceps share the same distal tendon. Instead of individually inserting directly at the elbow, all three heads of the triceps insert on a wide, flat tendon (an aponeurosis), which then inserts at the elbow. The long head of the triceps inserts on the medial side of this tendon, so putting a stretch on the long head of the triceps might medially displace the aponeurosis slightly, which would effectively place the fibers of the lateral heads of the triceps in a position where fiber lengths would be slightly longer with the same degree of elbow flexion. Alternatively, if passive force on the triceps tendon increases, activation of the monoarticular heads of the triceps may increase slightly due to tendon-associated reflex arcs. With that said, I do think the hypoxia explanation is a far more plausible primary explanation for these findings.
Ultimately, I think this study is a useful lesson in not getting too seduced by reductionism and single-factor thinking. The skepticism I’ve seen toward this study’s results essentially boils down to “nothing was done to put more tension on the monoarticular heads of the triceps, and the monoarticular heads of the triceps were trained at the same muscle lengths with both exercises, so the results must be wrong.” In other words, if you assume that only one or two factors could possibly influence muscle growth, and a study result conflicts with those assumptions, the result shouldn’t be trusted. However, I think it’s far more helpful to instead assume that your assumptions may have been faulty – maybe there are simply other factors that influence hypertrophy. Only time will tell, but for now, if you have lagging triceps, I think it’s probably worth giving overhead triceps extensions a shot (if they’re not already in your program).
Note: This article was published in partnership with MASS Research Review. Full versions of Research Spotlight breakdowns are originally published in MASS Research Review. Subscribe to MASS to get a monthly publication with breakdowns of recent exercise and nutrition studies.
References
- Maeo S, Wu Y, Huang M, Sakurai H, Kusagawa Y, Sugiyama T, Kanehisa H, Isaka T. Triceps brachii hypertrophy is substantially greater after elbow extension training performed in the overhead versus neutral arm position. Eur J Sport Sci. 2022 Aug 11:1-11. doi: 10.1080/17461391.2022.2100279. Epub ahead of print. PMID: 35819335.
- Pedrosa GF, Lima FV, Schoenfeld BJ, Lacerda LT, Simões MG, Pereira MR, Diniz RCR, Chagas MH. Partial range of motion training elicits favorable improvements in muscular adaptations when carried out at long muscle lengths. Eur J Sport Sci. 2022 Aug;22(8):1250-1260. doi: 10.1080/17461391.2021.1927199. Epub 2021 May 23. PMID: 33977835.
- Maeo S, Huang M, Wu Y, Sakurai H, Kusagawa Y, Sugiyama T, Kanehisa H, Isaka T. Greater Hamstrings Muscle Hypertrophy but Similar Damage Protection after Training at Long versus Short Muscle Lengths. Med Sci Sports Exerc. 2021 Apr 1;53(4):825-837. doi: 10.1249/MSS.0000000000002523. PMID: 33009197; PMCID: PMC7969179.
- Stasinaki A-N, Zaras N, Methenitis S, Tsitkanou S, Krase A, Kavvoura A, Terzis G. Triceps Brachii Muscle Strength and Architectural Adaptations with Resistance Training Exercises at Short or Long Fascicle Length. Journal of Functional Morphology and Kinesiology. 2018; 3(2):28. https://doi.org/10.3390/jfmk3020028
- Goto M, Maeda C, Hirayama T, Terada S, Nirengi S, Kurosawa Y, Nagano A, Hamaoka T. Partial Range of Motion Exercise Is Effective for Facilitating Muscle Hypertrophy and Function Through Sustained Intramuscular Hypoxia in Young Trained Men. J Strength Cond Res. 2019 May;33(5):1286-1294. doi: 10.1519/JSC.0000000000002051. PMID: 31034463.
- Ramos-Campo DJ, Scott BR, Alcaraz PE, Rubio-Arias JA. The efficacy of resistance training in hypoxia to enhance strength and muscle growth: A systematic review and meta-analysis. Eur J Sport Sci. 2018 Feb;18(1):92-103. doi: 10.1080/17461391.2017.1388850. Epub 2017 Oct 18. PMID: 29045191.
