Are low reps or high reps better for muscle growth in trained lifters?
September 30 2014
If you are a trained lifter, are high reps (8-12) reps better for muscle growth ? Or should you do low reps ( 4-6 reps) to increase a muscle growth? This was the exact question raised in a recent study.
What was the purpose of the study?
Though 8-12 repetitions are universally regarded as the optimal range for muscle growth, there is very little evidence that show any advantage of high reps (8-12) over low reps (4-6 reps). The few studies that did compare low reps to high reps suffered from 2 major problems:
Unequal Volume: The volume wasn’t same for both groups The 8-12 rep groups always did more volume than the low rep group. For example, one of the studies quoted in the paper showed an increase in muscle growth with high reps. To my knowledge, this is the only study that showed an advantage of high reps over low reps. The bodybuilding group did 9 sets of 8 - 15 reps (40-80% of 1RM). Now guess what the powerlifting group did? 5 sets of 3 reps (90% of !RM). So the volume of the bodybuilding group did almost 5 times more volume than the powerlifting group. The second study that showed identical results as reported in the paper is the exact same study, but reporting other variables. What about studies which did equate volume? The studies which did equate volume found no difference between high rep and low reps in muscle growth.
Not Trained: All the studies above used untrained subjects. In untrained subjects, everything works. In trained subjects, the muscles are less responsive.
Hence it is a great question to ask whether lower reps (or heavier load) performed at the same volume as high reps (or lower weight) could induce greater muscle growth in trained folks. And this is exactly the question the researchers asked in the current study. as mentioned in the paper, it is just a comparison of a powerlifting routine to a bodybuilding routine with similar volumes . To put it another way, is it the pump (high reps) that matters or is it the high load (low reps) that matters in muscle growth? Let’s see.
What was the study design?
Study Design: 17 well-trained subjects were randomly assigned to two groups. After 8 weeks, there were 9 in the bodybuilding type or high rep group, and 8 in the powerlifting group/low reps groups. The subjects’ experience ranged from 1 to 10 years in weight training.
Low Rep group: Hit one muscle 3 times per week
Sets * Reps : 7 sets of 3 reps
Rest between sets: 3 min
All sets to failure
High Rep group: Hit one muscle 1 time per week
Sets * Reps: 3 sets of 10
Rest between sets: 1.30 min
All sets to failure
As you can see, the volume was equalized by using Load*Reps*Sets. So for example, for the chest muscles, the low rep group did 63 reps while the high rep group did 90 reps.
How did they run the statistical analysis?
If you don’t care about the stats, you can skip this part. In my understanding, there are two common ways to analyze the results of a study:
Post Test: You can just simply look at the difference between post test results in both the groups. That being said, if the groups were different at the beginning of the study, the final results will be influenced by these differences. Even if you conduct the prefect randomization with a large sample size, chances are that the groups will be slightly imbalanced in the beginning. The group which had a higher average will tend to have a lower score, and the group which started lower will tend to have higher scores by the end of the study. This statistical phenomenon is called regression to the mean.
Change Score: Or you can take the difference from pre-test to post-test in both groups and calculate the difference to obtain the gain score or change score. Here, since we are including measurements twice -pre test and post-test, you are adding up errors unless the test are highly reliable. In addition, the regression to the mean will have an influence here too.
So what is the solution? Basically computing just one value for the pre-test for both groups. This would look like both groups started out with the same score in the beginning. This is usually referred as baseline adjusted scores or adjusted scores.. So now we can just compare the post test results without worrying about the nasty regression to the mean. This type of analysis is not very common. However, it is frequently used in drug trials and generally considered to be the best method to maximize statistical power in a two group pre-post design. So the researchers did an excellent job with statistical analysis.
What were the results?
Muscle growth: No differences were seen between the groups for biceps thickness. Both groups improved by around 13%.
Strength: For bench press, the low rep group had a 13% increase, while the high rep group had a 9% increase. For squats, the low rep group improved 26%, whereas the high rep group improved 22%. This is after the adjustment as I wrote in the analysis. The numbers shown in the table are the non-adjusted values and hence the percentages will be different if you calculated using those numbers
Conclusion: Though there was little difference in muscle growth, considering the time (90 min vs 17 min), high fatigue and injuries (2 injuries) in the high load group, the high reps or bodybuilding type of workout appears to be more suited for muscle growth.
Low reps vs high reps: The topic of low reps and high reps has always fascinated me. In fact, this was the exact question for my Master’s thesis in 2005. I used beginners, but the concept was pretty similar of equalizing time under tension in both the groups. Also, my first ever online article in 2006 was about repetition range and hypertrophy. My hypothesis was that low reps would have a greater hypertrophy than high reps if the volume was equalized. However, after all these years of reading the literature, the less and less I am convinced that there is a great difference between low reps and high reps to talk about. The last article I wrote on high reps and low reps in 2010 was based on a study which showed that there is a limit to protein synthesis, and there is no difference in protein synthesis with high reps and low reps. Apparently, the conclusions I wrote in that article were almost identical to the current study. That being said, that study was done in untrained lifters. I even had a graph to show how this may turn out in trained lifters. So I have been waiting for all these years for a study on trained lifters comparing repetition ranges and looking at actual muscle growth.
8-12 reps: So if there is no direct study to show that 8-12 reps are optimal for muscle growth, how come bodybuilders always used 8-12 reps? I think this is a good example of how we tend to confuse cause and effect in anecdotal observations. In my opinion, anecdotally, 8-12 reps are considered “optimal” for muscle growth not because it is optimal for muscle growth per se, but for reasons that have very little to do with actual muscle growth and more to do with feasibility, recovery and injury concerns.
We all know low reps cause greater fatigue and hence can impair recovery. Also, as the individual gets stronger, the heavier weight that comes with low reps can be very hard on joints often leading to joints aches and injury concerns. Now what about the practicality of low reps? What do you think is really practical in your gym? Doing 7 sets of 5 of bench press with 3- 5 minute rest, or 3 sets of 12 reps with 1 -2 min rest (35 reps vs 36 reps)? All the above concerns become very real when you are in the gym almost every day for years like a typical bodybuilder. Consequently, bodybuilders naturally gravitated towards high reps. What about powerlifters? Power lifters have no choice but to perform low reps because their sport demands low reps. And hence they have to extensively rely on load cycling, deloading, not going to failure and other strategies to minimize fatigue and stay injury-free. So 8-12 reps for muscle growth was born more out of necessity than choice, and for reasons not directly related to actual muscle growth as most people think. So there is no magic in high reps, I think.
This is the large reason why there is efficacy and effectiveness trials. Most of the studies in exercise physiology are efficacy trials: Trials which are done under ideal or controlled conditions. All these problems outlined above will not be evident or will be missed in a 8 or 10 week controlled lab study, which is primarily designed to look at muscle growth. On the other hand, effectiveness trials are trials trying to simulate the real world. In effectiveness trials, one of the primary questions is whether the protocol or workout program is actually feasible in real world conditions. So even if efficacy trials show muscle growth is more with a certain rep range, effectiveness trials may show something totally different.
Volume: At least from my perspective, the simple question is if you do 1 sets of 12 reps, can you get greater hypertrophy with 2 sets of 6 reps? Here, the time under tension for both the groups is the exact same: 12 reps. So the comparison here is just between load, which is almost double in low reps, and metabolic fatigue, which is very high for high reps and very low for high reps . I am guessing the choice of 3 reps than 5 or 6 reps was to avoid any sort of metabolic fatigue in the low rep group. In the current study, equated for volume, the TUT for low reps was 63 reps and high reps was 90 reps. Will equalizing the TUT make a difference? I don’t know. Is that practical? Maybe not.
Sample size: Sample size is obviously small. There are two schools of thought when it comes to sample size: One school of thought is that even if it is a small sample size, you should conduct the study and publish it. And in the future, a meta-analysis can pool all the small studies to come up with a large sample size. The other school of thought is that you shouldn’t embark on a study when you clearly know you won’t be able to recruit enough subjects to find a statistically significant difference. I favor the former. In fact, the study which I recently conducted had 9 and 8 subjects in each group. But the criteria for recruitment was very narrow and hence very hard to get subjects. In fields like exercise physiology, it is not practical to conduct large scale multi-center trials or hope to receive NIH grants that give you a couple of million dollars every few years.
Failure: This is one area I felt the researchers could have been done a little bit differently. For example, the low rep group performed 7 sets of 3 to failure for a body part 3 days a week. This is very taxing on your nervous system and muscles, and is evident from the subjective reports and injuries sustained in the low rep group. If ‘going to failure’ is an integral part of bodybuilding, ‘not going to failure’ is an integral part of power lifting. I would have been very happy if they had at least instructed the low rep group to stop one rep shy of failure. And this change would have made the routine more comparable to a typical powerlifting routine. I talked to Brad about it. Could the heavy load taken to failure affected recovery and thereby muscle gains? I don’t know. As is typical of Brad, he calmly acknowledged it and said it probably won’t make a big difference since they maintained strength levels and even gained some. I agree with Brad to a certain extent.
Strength: As expected, the low reps group gained more strength than high rep, especially in the squat. But strength is traditionally defined as the maximum amount of weight you can lift for one repetition or 1 RM. I would have loved to know what the 10 RM for each group was. My guess is that the high rep group would improve the 10 RM more than the low reps. As dictated by the specificity theory, the neural adaptations inherent to low reps are inherent to low reps and is less unlikely to transfer to high reps. And if it is true, I don’t see a logic for bodybuilding routines to include low reps to increase muscle strength and thus muscle growth. For a bodybuilder, what you need to worry about is the maximum amount of weight that you can do for 8 to 12 reps and not 1 repetition.
Now how to better analyze the results?
Typically, almost all studies are analyzed based on the p-value. Here is how to do it using confidence intervals.
The horizontal line says that we can be 95% confident that the true mean difference between the groups will fall somewhere on that line. For example, as you can see in the figure, the difference between the groups (mean difference) for bench press in the current study is 11 kg (10.8) in favor of the strength training group. And if we repeat the study a number of times, we can be 95% confident that it could fall somewhere between 10kg in favor of the high rep group, or 31 kg in favor of the low rep group as shown by the upper and lower limit. This interval is called the 95% confidence interval (CI). Now you will admit that it is a bit too wide of an interval to make any definitive conclusions about whether to use high reps or low reps to increase bench press. So how the can you make the line or the interval shorter?
Now here is where our sample size comes in. The greater the sample size, the shorter the horizontal line of our interval, and thus greater our confidence or certainty or precision. Here, even though the confidence intervals are pretty wide for bench press, we clearly see that there could be a large benefit as the upper limit possible for bench press approaches 31 kg and 10 kg. If I am coach or a lifter, these are huge numbers. So we might very well embark on a future study with a larger sample size to see which group is really better. Now if the upper limit was pretty small, for example 3-5 kg for bench press, I wouldn’t bother conducting a future study to see if high reps or low reps are better.
Now if we used p values, let’s see our how our interpretation goes: Since the horizontal lines cross the vertical line of no difference (P > 0.05), there is no statistically significant difference between the groups. Hence we conclude that there is no benefit of using low reps or high reps for bench press. There is no further interpretation or discussion of the study results. Apparently, both the p-value and confidence intervals are calculated using the same method , but the interpretations and conclusions we reach are very different.
There is more to it, but I hope this gives an introduction to the benefit of using CI to interpret and report results.
Note: I used the non-adjusted values since the baseline adjusted values were not shown the paper. After the baseline adjustments in the current study, the squats horizontal line favored low reps (did not cross the vertical line of no difference)
So what does 13% increase in bicep thickness mean?
Now, we all know what a 5lb or a 10lb increase in bench press or squat means. We all know what a 1/4 in or half inch increase in our biceps means. But what does a 13% increase in biceps thickness using ultrasound means?. Even if you showed a 13% increase in muscle thickness, you may not notice any difference looking in the mirror or using a tape measure. It maybe very meaningful if you are a bodybuilder on stage at 5% body fat under a row of bright lights. Finally, what counts is whether your muscles ‘look’ big. The same goes for measuring muscle size using MRI. This is usually called the practical or clinical significance. It would be good to know how this percentage translates to a tape measure reading.
Blah blah.. so should I do high reps or low reps?
In my opinion, this is one of those questions which will be debated for years. And this has more to do with our human nature and cognitive biases and less to do with the science. Things will never be black and white, but as humans we favor certainty and tend to be very uncomfortable with ambiguity. If you notice, this is one reason why a lot of quacks thrive; they always make absolute claims and people ‘subconsciously’ tend to love anyone who can take complex issues and simply portray it as black and white. Mind you, research can never eliminate uncertainty, it can only minimize it.
Now instead of worrying about doing low OR high reps, why not just do both:
- Start the cycle with high reps and culminate with low reps such as Week 1: 10-12, Week 3-4: 8-10, Week 5-6: 6-8: Week 7-8: 1-3 reps
- Perform a strength program with low reps for the big lifts such as bench press, squats,deadlfitsand shoulder press. And then follow with high reps for assistance exercises like dumbell press, rows and so forth.
- Or just do a pyramid, starting with high reps such as, set 1: 10, Set 2: 8, Set 3: 6 Set 4: 4.
There are many other options, but you get the point.
Brad Schoenfeld: Brad is the lead author of the study. Brad is one of those who is single-handedly changing the field of hypertrophy. This study is a perfect example of how he investigates questions that people have been asking for years, but no one ever bothered to conduct a study on. It is certainly a mark of a good researcher. What I really like about Brad is his passion for the field of hypertrophy and towards a scientific approach. Every time I talk to Brad, in person or on phone, we just wallow in science and theories with the same enthusiasm and amusement shown by two little kids playing outside in the rain. I am in school now and teach a bit, but I am yet to see a student that shows that hunger or desire to learn more and excel in their field. And this is exactly what is common to all those big names in the fitness field, like Lyle Mcdonald, Bret Contreras, Nick Tummenello, Alan Aragon to name a few (all science-based too).
- From a purely muscle growth or academic standpoint, we might have to wait for a few more studies to definitely conclude that high reps are not superior to low reps or high reps and low reps are equivalent for muscle growth. As of the current study, the likelihood of high reps being superior to low reps is low.
- From a practical/safety/volume perspective, it is better to rely on high reps more often than low reps. You can also mix repetition ranges, if strength is a concern.
I think that’s all I have. I know it is a bit long, but I hope you learned a little bit about research and muscle growth after reading it. Thank you for reading.
Reference: Effects of different volume-equated resistance training loading strategies on muscular adaptations in well-trained men
If you like it, please share it:
| Wed October 01, 2014
Excellent as usual, Anoop.
However, I can’t agree to definition of 8-12 reps as “high”
I would suggest to treat bellow 5’s as low, 5-15(12?) as intermediate, and more 15 as “high”
The question that fascinates me the most is as follow:
“May progress in hypertrophy be achieved by using exclusively high rep range”?
| Wed October 01, 2014
Great Article Anoop!
| Wed October 01, 2014
Anatoly I agree with your question “May progress in hypertrophy be achieved by using exclusively high rep range”?.... or as I would state it slightly differently “can (optimal) hypertrophy be obtained with exclusively high rep ranges”? I do think Schoenfeld’s novel review on mechanisms of hypertrophy was more than likely correct and I hope future research breaks this down further on trained subjects. He basically suggested training in intermediate ranges may be both easier to maintain health and just as effective while the strength ranges act to increase strength output which increases the amount of work performed when switching back to high rep growth ranges. Also, the high rep ranges increase metabolic stress and metabolites; perhaps cell stretching to some degree via hypoxia. So fascinating. Great question thought. And Anoop…bro… You have to post more. I love reading your work
Anoop | Wed October 01, 2014
Thanks Anatoly! I think some of the recent studies have looked at the question if we can have muscle growth with higher reps. In fact Brad studies this recently with 30 reps in trained lifters. And it showed some pretty interesting results. It would be better if it comes from him than me.
Thanks Michael! I will post more. As I said in one comment on facebook: This study again shows why anecdotal observations, no matter how compelling they are, should be tested before we jump to conclusions. In fact, the first studies which did look at hypertrophy started out asking “why are 8-12 reps optimal” instead of asking : Are they optimal”?.
| Thu October 02, 2014
Under the paragraph titled, ‘What was the study design?’ aren’t the workout durations the wrong way around? The higher rep group’s workouts took less time than the lower rep group’s workouts. This could be another advantage of doing higher reps: less time in the gym.
As always, great article.
Anoop | Thu October 02, 2014
Thanks for the comment. I corrected it. Good catch!
I hope people understood the stats part, especially the CI intervals.
| Tue December 02, 2014
This was a very informative post! I’ve always with high reps because it’s what I’ve been most comfortable with and seen the most progress.
Whenever I’ve gone with low reps or I see other people do it, we always tend to use heavier weight and most of the time we can’t fully control that weight. For example, when people do low reps while benching, the weight is so much that all they’re doing is bouncing the bar off their chest which is useless for muscle growth because your muscles aren’t working as hard as they should be
| Sat July 04, 2015
“For example, when people do low reps while benching, the weight is so much that all they’re doing is bouncing the bar off their chest which is useless for muscle growth because your muscles aren’t working as hard as they should be”
I agree, that’s why I do paused bench press stopping the bar about an inch over my chest. That way the reps(3-6 at the moment, strength being the main goal) are much cleaner they could ever be doing high reps. Sets of 10-12 would be pretty impossible or impractical to do paused anyway, and my technique and focus tends to deteriorate towards the end of the set when doing high reps.
| Mon October 24, 2016
Tension and fatigue are the key factors to stimulating hypertrophy.
This can be done with low or high reps depending on the application with a wide variety/combinations of sets, reps, and exercises based on the individual’s unique characteristics.
To me, it makes sense to use the least amount of load necessary for optimal results so to allow less wear and tear on the joints. My days of using max loads for low reps are over, but YMMV.