An example of how NOT to interpret research

Note: Because this is not a review article, it has not been peer-reviewed.

With the rise of the “evidence-based” fitness movement (EBF) figures such as Brad Schoenfeld have become famous and adopted a guru-like status in social media. While he is certainly part of a recent wave of researchers addressing relevant questions in the skeletal muscle hypertrophy field, I’ve had my reservations on some of his opinions on specific topics (which earned me a blocked status from him) and studies (co-)authored by him. Inasmuch as I am very critical of everyone (including myself), I’ve been able to discuss and debate some other EBF figures respectfully (for instance, my exchange with his co-author James Krieger). Thus, I don’t think I have been particularly critic of him compared to others; strong debate is part of the daily life of scientists. But I disgress.

Because of their status, EBF figures should be careful of what they post on social media and should be held to the higher standards of scrutiny as, in a way, they represent the scientific community in social media.

I came across an Instagram post by Brad Schoenfeld that showed a slide of an upcoming presentation dealing with the topic of repetition tempo and hypertrophy:


In it, Brad shows two studies finding apparently different results for the effects of performing an exercise with a slow eccentric phase. His interpretation of the data:

“…it can be speculated that in a 1 second eccentric action, you’re allowing gravity to do most of the work, not the muscles. Alternatively, taking 2 seconds on the eccentric forces the muscle to lower the weight; prolonging the duration of the eccentric action thus has no further benefit.”

The two studies cited are Pereira et al., 2016 and Shibata et al., 2018. Interventions are summed up below:


As can be seen in the table, the graph shared in the Instagram post is wrong, as Shibata et al., 2018 compared 2-0-2 to 2-0-4, not 1-0-2 to 1-0-4 (concentric-transition-eccentric). But these mistakes happen, and, in theory, shouldn’t affect much the conclusions as the focus is on the eccentric part (see below). However, it suggests that the cited paper hasn’t been carefully read. But there are more methodological details that have to be considered for comparing both studies, not just the eccentric tempo.

For avid hypertrophy-related readers, it is readily apparent that we are talking about two very different populations: trained in Pereira et al., 2016 and untrained in Shibata et al., 2018. Both studies also differed on the duration of the intervention; the training protocol on Shibata et al., 2018 (6 weeks) lasted half the time of Pereira et al., 2016 (12 weeks). The length of the intervention is important in hypertrophy research, which might already had influenced the results. So, from the start, both the population being studied and the duration prevents any proper direct comparison between the two studies (especially without a caveat).

But some other nuances come into play when analyzing the results that also affect the interpretation. The first thing that came to my mind when I saw these studies was the duration of the concentric. It is a very bad idea to purposedly slow the concentric, as it would reduce the load that can be lifted. A 1 second concentric means a controlled, albeit fast concentric. A 2 second concentric at 75% 1RM means doing the movement slow on purpose since repetition number 1. A 1 second concentric doesn’t impede lifting a high load; a 2 second or slower concentric proportionally reduces the load that can be lifted. As we can produce more force eccentrically than concentrically, a slower eccentric phase means higher resistance (and hence greater mechanical tension) at a given load (during normal resistance training conditions, the eccentric phase is under-loaded). On the contrary, a slower concentric automatically reduces the load that can be lifted and thus the load that is resisted during the eccentric (thus making it more under-loaded). According to this, a 2-0-2 tempo would be better than a 2-0-4 because the latter would reduce the load that can be lifted in a controlled fashion. This would be specially apparent in multi-joint, compound exercises like the squat, in which performing a slow tempo as prescribed needs more focus on balancing the weight than on actually targeting the muscle. And this is exactly what the authors observed in the study: the load lifted for 75% of 1RM increased by 5.5kg (7.7%) in the 4-sec eccentric group, and by 9.1kg (12.7%) in the 2-sec eccentric group. Therefore, the slow concentric reduced the load that could be lifted and, coupled to the slower eccentric tempo, reduced the amount of weight used. Because of this, the 4-sec eccentric group performed less reps with less weight than the 2-sec eccentric group:


Basically, such a set up ended up having the 4-sec eccentric participants lift less weight for less reps. Nevertheless, statistically, there were no differences in muscle growth between groups, which would suggest that, despite reducing the load and reps performed, a slower eccentric still promoted a similar hypertrophic response than a faster eccentric with a higher load. Therefore, in contrast to Brad’s superficial interpretation, this study shows that a slower eccentric might be more effective than a faster one (similar hypertrophy with less overall load). This interpretation agrees with the findings of Pereira et al., 2016.

When the weakening effect of a slow concentric is eliminated by using a faster concentric (1 second), and the exercise used does not become a balancing act when using a slow eccentric (and hence, a higher load can be used without problems), then the benefit of using a slower eccentric becomes more apparent. Compared to a 1-sec eccentric, a 4-sec eccentric promoted a higher increase in CSA (4.8 cm2, 16.5%; versus 1.9 cm2, 6.25%). This difference was not statistically significant (only 6 subjects per group), but the effect size was almost three times larger for the 4-sec group (1.3 versus 0.5).



Contrary to what Brad’s Instagram post suggests, there appears to be a benefit of slower eccentrics on muscle hypertrophy (NB: taking into considerations the uncertainty regarding the relationship between muscle size measured by MRI or US and actual myofibrillar protein accrual). In trained subjects, when the absolute load is not compromised by a slow concentric, a slower eccentric appears to induce greater hypertrophy than a shorter eccentric. In untrained subjects, a slower eccentric appears to promote the same magnitude of hypertrophy despite using less load and performing less repetitions. It remains to be seen if the latter also occurs in trained subjects and at what eccentric speed a threshold is observed. Both studies show the same effect: a slower eccentric outperforms a faster eccentric.

An adequate comparison of both studies has to take into account the differences in population and design to reach a proper conclusion, instead of only looking directly at one variable (tempo) on its own. Carefully looking into methodological details and results allows proposing an hypothesis which better explains the observations. It’s easier to just be driven by the conclusions from the authors (eg. effect versus no effect), but this can lead to completely wrong interpretations, like the one in the Instagram post.

Inadequate comparison of studies is a major confounder and limitation in skeletal muscle hypertrophy research, and introduces systematic errors in the literature when doing it in meta-analyses assessing different variables. Likewise, EBF prominent figures should properly and carefully analyze research before making a bold recommendation, and their statements should be critically evaluated and not taken at face-value. In this particular case, the incorrect description of Shibata et al., 2018 further reduces the accuracy of the statement (while strengthening Schoenfeld’s position), but it is most likely a reflection of not properly reading the studies before making the post/slide (or rather doing it quickly without much thought).

Research-wise, a study directly comparing different eccentric tempos without slowing the concentric (for instance 1 vs 2 vs 4 seconds), while performing a high load resistance exercise program to failure in trained subjects should answer directly if there is a benefit of slower eccentrics.




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s