Monday, June 10, 2013

Intra-Set Rest Periods Boost Power (+38%) & Strength Gains (+65%) Without Hampering Skeletal Muscle Hypertrophy

Up your strength /check! Up your power /check! Up your muscle gains in the long term /likely!
Intra-Set Rest Periods? No, that's not identical to "clustering" or "cluster training", although it may appear as if it was. Ok, let's briefly summarize what IRS / clustering is: Basically the idea is to perform the exact same amount of reps and thus work with a heavier weight than usual to induce greater gains (usually strength gains). Now how can you do that? Well, one possibility would be to simply add another set, so that you would now be doing 4 set of 10 reps instead of the usual 3 sets of 10 reps. Unfortunately, this is yet nothing but another instance of the notorious more helps more approach of which you should by now have realized that it is pointless and ineffective.

Now things are getting a little complicated, because contrary to a classic clustering regimen, where you would add weight to the bar, do say 8 reps, take 10 breaths and rep out the other two, the intra-set rest (IRS) period protocol in the Oliver study employed a tightly controlled and not an "ad hoc" variety of "clustering". Contrary to the control group who performed a regular 4x10 routine, the IRS group broke their 4 sets of 10 with 120s rest between sets up to eight sets of five with only 60s of rest in-between the exercises (see table 1 for more details)
Table 1: Overview over the workout program (adapted based on Oliver 2013)
It's easy to see, there is no black magic involved. Same lifts, same plan, same progression, but different set x rep scheme and inter-set rest for the STD vs. ISR groups. In the fourth unloading week, tests were conducted and the same protocol was repeated with accordingly increased weights. All sessions were supervised and the compliance was 95% over the whole 3x4 week study period.

Remember: This is not "clustering"! When you "cluster" your training volume will increase, the IRS protocol, on the other hand, is standardized in a way that the total workload will (and in in the study at hand did) remain the same.
The participants of the study had all been doing upper + lower body resistance training for at least 2 years and were thus familiar with the basics of strength training, before the study they received nutritional counseling which is probably the reason that all of them increased their protein intake significantly (ca. +50g to ~190-200g/day; no inter-group difference).

The same goes for the changes in muscle fiber and body composition. All subjects experience an increase in MHC-IIa fibers (glycolytic) and gained lean mass continuously. Unfortunately, they also gained body fat, so that the body fat % (DEXA measured) did not change at all.

The power and strength advantage

What did however react to the modified rest times were the changes in strength and power on all the major lifts. According to Oliver, et al.
Table 2: Effect size and qual. inferences on intergroup difference; (t)rivial, s - (s)mall, (m)edium., (l)arge (Oilver. 2013).
"Only subjects in ISR experienced an increase [in 1RM bench press & squats] at 4, 8 and 12 weeks. This corresponded to greater increases at 4 (ISR, 6.6±6.6 kg; STD, -1.4±6.2 kg; p= 0.012), 8 (ISR, 9.9±6.8 kg; STD, 2.9±5.8 kg, p = 0.016) and 12 (ISR, 15.1±8.3 kg; STD, 9.1±3.7 kg; p = 0.051) weeks.

[...] Again, only ISR increased at 4, 8 and 12 weeks. Greater percent increase from baseline was observed in ISR at 4 and 8 weeks (p = 0.017 and 0.034, respectively), with 12 weeks approaching significance (p = 0.082)." (Oliver. 2013)
Similar differences were observed on the power output strength, where the group difference at both 8 weeks (ISR, 151.0±74.0 W; STD, 97.5±60.9 W; p = 0.084) and 12 weeks (ISR, 282.1±104.1 W; STD, 204.9±70.2 W; p = 0.063) approached significance with subjects in ISR showing a greater increase. Interestingly, the advantage became even more obvious, when the changes for normalized for the subjects' body weight (from p = 0.084 to p = 0.016 and p = 0.063 to p = 0.038 after 8 and 12 weeks, respectively - that means what may be at best a trend on an absolute level is a statistically significant advantage if you take the muscle mass into account, as well)

Bottom line: It sure looks as if IRS would in fact be a viable training strategy for anyone training for muscle hypertrophy, strength and power. Neither the workload, nor the workout time changes and still the results improve. And I can hardly phrase it better than Oliver et al. did:
Suggested read: "Training for Size & Strength: Does Rest Matter? Study Finds 7-9% Greater Increase in Muscle Size With Decreasing Rest Periods." (read more)
"Based on these results, it could be suggested the incorporation of ISR in the hypertrophic phase of a traditional or non-traditional periodized training program would allow for greater improvements  in strength and power. 

Whether these improvements would result in greater gains in strength and power output over an entire mesocycle is unknown, but hypothetically entering the strength and power phases of a training mesocycle at higher performance ability (strength and power) would allow a continued improvement above that achieved during traditional training models." (Oliver. 2013)
What? Oh, yes... I will tell you if the researchers ever conduct this study. To be honest, I am yet pretty sure that this is not going to happen. After all, the results won't be patentable and since we are dealing with healthy people (worse athletes!) the governments of the Western Obesity Belt are more likely to burn a few additional millions to find new pharmacological methods to prolong the misery of the increasingly obese majority of their citizens - I mean, who would vote for them if they told people that we already had the solution to the obesity epidemic, but it was not available in convenient pill form?

  • Oliver JM, Jagim AR, Sanchez AC, Mardock MA, Kelly KA, Meredith HJ, Smith GL, Greenwood M, Parker JL, Riechman SE, Fluckey JD, Crouse SF, Kreider RB. Greater Gains In Strength And Power With Intra-Set Rest Intervals In Hypertrophic Training. J Strength Cond Res. 2013 Jun 3. [Epub ahead of print]


  1. This sounds conceptually very similar to the best program I have ever found ( for me anyway) for upper body shoulder centric movements. The routine is from Steve Justa's book and the Dragon Door website, and it seemed appealing in that it is the most unique approach I have found and it allows a lot more total volume and intensity without provoking over-training...and it goes like this.. It is based on doing sets of one rep very day. You start by calculating 75% or so of your 1 rep max and on day one work up to 3 sets of 1 rep each. Next day do 5 sets of one rep each, and so on adding two sets per day until you reach 15 sets of one rep. Then add 5 pounds and drop back down to 3 sets and go through the routine again. This sounds like a vacation and it will initially feel deceptively easy, however I have found it to be very effective. It also allowed me to improve my form and joint integrity quite a bit. I tweaked the basic routine so that for set six (only) I add 5 or ten pounds and for the bench press, I alternate close and wide grip on sets 8-14. I found this routine to work great for shoulder-centric moves like all forms of pressing and pull-ups, however I over-trained on this and went backwards when I tried it for squats and deadlifts...just too much work. I eventually had to back off some from every-day work once I was well into new personal record territory, but it is still producing gains...just a thought for those of you looking for something completely different...

    1. Superchunk... nice handle! My take on this is that you are chronically depleted from an adrenal perspective. If you have the inclination, find an integrative physician or a clinical nutritionist, and get an Adrenal Stress salivary panel. If it checks out, I will be surprised.

  2. I think the way the article was written was quite confusing. The title is "Intra-Set Rest Periods Boost Power..." and "intra" means within, so that would mean resting within the set. Doing 8 sets of 5 reps with 60 seconds rest between sets has nothing to do with "intra set rest," because there was no rest within the set!

    The other confusing part is defining cluster training as doing a straight set, then resting for a short time, then doing another 1-3 reps. This is not what cluster training is! Cluster training would be more like doing one heavy rep, resting 10-15 seconds, doing another rep, resting 10-15 seconds, etc. This allows you to complete the same number of reps (say 5) in a "set" but using a higher percentage of a 1RM. Doing a straight set of say 8 reps, taking a short rest, and then doing another 1-3 reps would be considered rest-pause training, or simply an extended set method, not cluster training.

    1. Could not agree more, pretty confusing article, good info though.

    2. I concur completely, needlessly confusing

    3. I agree. No offense to the author, but you do not know what either "intra-set" rest is, or what a "cluster set" is. Intra-set rest was defined by Neil Gaus above.

      True clusters are specialized rep schemes specifically geared to tap into higher threshold motor units that allow for a greater training volume at the training intensity. The best example is the 5x5 rep scheme.

      In a typical 5x5 rep scheme, most lifters will use 85% of 1RM for their work sets. A cluster methodology allows the same lifter to use 89-91% of 1RM for the same volume.

      To use an example of dead lifting, if a lifter can dead lift 500 lbs, they would generally perform 5x5 in the neighborhood of 425 lbs. In a cluster set, they would load the bar to approximately 450 lbs, and perform each set of 5 reps as follows:
      1 rep - 10 seconds - 1 rep (2) - 10 seconds - 1 rep (3) - 10 seconds - 1 rep (4) - 10 seconds - 1 rep (5) rest.

  3. On an unrelated note... Deadlift injury. Constructive advice will be greatly appreciated from anyone. So far I've decided to just not deadlift til I feel it's appropriate again and see my GP who might re-direct me to a specialist/chiropractor/physiotherapist.
    Pain localized to very lower back, mostly on the left side. It's not painful enough for me to take some NSAIDs and I have no problems bending down, although I can clearly feel something's not right.
    Also, when the injury first occurred (3 or 4 weeks ago - today was the first time I attempted to deadlift again a slight bit of pain has returned), I had extreme difficult lifting my legs to any height where I could put on pants (creates a fun image in one's mind, but is in reality not that fun). Both legs were affected, but especially the right one.
    Air squats result in a definitely painful sensation.
    Any thoughts or advice?

    1. nothing to add to Aileen's comment - without someone actually looking at it, I can't say anything.

  4. Sorry for the typos resulting in seemingly poor grammar and spelling, but there's simply edit button once the comment's published. ;)

  5. This is a very interesting result, but it brings to mind a few questions. Will it work if I am already in a low rep regime (i.e. half of 6 reps rather than 10 reps)? Should I also do something similar for all my assistance work? Finally are there any concrete results on optimal rest times between sets and if so should I half them to follow this protocol?

    1. (a) probably not. At least hypertrophy wise I would suspect it sucks

      (b) I would recommended the assitance work (for 6reps it may not be necessary, but even with 5x5 I usually suggest 3x10 for at least 1 exercise as an adjunct)

      (c) 1-2min is the best we can say based on studies with shorter = more hypertrophy / body composition and longer = more strength oriented

  6. Doesn’t this directly contradict the following study:

    Do you have any thoughts on why the outcomes were so different?

    1. the protocol is fundamentally different (fewer reps, longer rest, probably insufficient workload, just a single exercise - not surprising that did not work):

      from the FT: "The Singles group performed 6 repetitions with a rest between every repetition; each repetition commenced after a 23-second interval timer sounded. The Doubles group rested after every 2 consecutive repetitions and commenced on a 56-second interval, until 6 total repetitions had been performed. The Triples group performed 3 consecutive repetitions on a 109-second interval."

      I mean come on 3 workload matched reps with 109s rest? Or singles with 23 seconds break and a weight you would perform six reps with - even my grandma wouldn't gain on that training regimen

    2. Must’ve been wrong link, and I keep getting an error message when I try to access Pubmed. This is the study I was talking about, though:

      "Volume (sets × repetitions × %6RM) between groups was equated and both groups completed all sets in the same time period (13 minutes and 20 seconds)."

      I don’t have the full paper right now, but from what I can tell they equated load, volume and total time and still the CR group came out ahead.

    3. This comment has been removed by the author.

    4. Ok, not sure what happened because the link goes to the right study.

      You must be talking about this later study:

      Well, that is obviously a completely different protocol - but again, not the one I was talking about.

  7. Adel, This was confusing to me and I can't find anything but abstracts, which don't clarify: "eight sets of five with only 60s of rest in-between the exercises". Shouldn't that rather read: "eight sets of five with only 60s of rest in-between _each set_".

    That would seem to make sense given the table and it seems that the point of the study was to compare 4x10/120s vs 8x5/60s with the total workload, intensity and time exercising being the same for both protocols. Thanks.

    1. well, the description in the full text is not exactly perfect, either, but my understanding is the same - identical workload, identical intensity and with standardized TUT in the end even the same time span for the whole workout with e 4x10/120s vs 8x5/60s (inter-set rest)

  8. This does not strike me as a landmark study. Basically, you have two sets of loading parameters - 4x10, 120sec rest, and 8x5, 60sec rest. The discussion of whether its clusters, rest-pause, or intra-set rest, etc. is semantics. And then you have a boat load of assistive work in the 3x10, 90sec rest realm for both groups.

    What is to say that relative overtraining did not occur in the Traditional group? The volume of training is certainly high enough for it to be a possibility. That alone would make the ISR group look better by comparison with regard to hypertrophy. Additionally, who is to say that the hypertrophy in the ISR group did not result from all of the assistive work? There is no way to know, unless they used another control that was not discussed, but I am willing to bet on it, as the average workout time for Push days would have been about 66 minutes, excluding warmup, and the Leb/Back day would have been closer to 70 minutes.

    Furthermore, if you present two identical scenarios for volume, density of work, amount of rest, and intensity, and the only thing that you modulate is the rate of the work period, of course the group performing the work at a faster rate - the ISR group - is going to improve power more. You have created a scenario where the only variable that you manipulate is time.

    Also, anyone who thinks that power can be improved through training with sets of 10 is ignorant. I am not saying that the authors of the study believed this; I am saying that when you manipulate the scenario as they did, that is absolutely the only result that you can get - the ISR group has to come out on top.

    However, as Power = Force x Velocity, you can manipulate mass, acceleration (torque), velocity, or even displacement. Provided you control the other variables enough, you can still increase power, so I cannot agree that this particular brand of intra-set rest is promising as a tool to improve power and hypertrophy; it just looks like another set of loading parameters.