Single set vs Multiple sets - new research

[Richard99]

Single set vs Multiple Sets – new research

There has been an ongoing debate in the strength training and bodybuilding communities for about 40 years as to whether a single set of an exercise is superior for building size and strength than training with multiple sets. The idea that a single set of an exercise might be more effective than traditional multiple set training was first popularized in the 1970s by Arthur Jones, inventor of Nautilus strength training equipment. Based on his training observations, Jones believed that a single set per exercise taken to the point of failure- a training method that is commonly known as high intensity training (HIT)- was the most effective type of training for improving both strength and size. In a series of published bulletins widely circulated throughout the strength and bodybuilding communities, Jones made his case for the superiority of single set training, sparking heated debate on the issue that continues to this day.

The debate drew the attention of exercise physiologists around the world, resulting in a growing body of research data examining the issue. However, despite an abundance of research studies physiologists were not able to resolve the issue. The main problem was that the research was equivocal; some studies supported the idea that a single set was more effective than multiple sets, other studies found multiple sets produced greater increases in strength and size, but most studies found no statistical difference in results between the two training methods. In short, there was no consensus in the research.

In general most research indicates that multiple sets tend to produce somewhat larger increases in strength and size. However, the issue is that the difference in results between the two has not been large enough to definitively say that multiple sets are superior. On average multiple sets produce a few percentage points greater increase in strength and size, say 2-5%, but this difference has not been large enough to be statistically significant (statistical significance is important to show that the results are not just a matter of chance).

With research unable to declare a clear winner the debate continued unabated. Despite the lack of consensus the physiological community generally accepted multiple sets to be superior to a single set, which drew some very vocal criticism from a few researchers.

In response to these critics a number of “meta-analyses” have been conducted by researchers in recent years see if the conflict could be resolved. A meta-analysis is essentially a study of studies. It is a way of analyzing the results of multiple studies on the same research hypothesis to see what can be learned by looking at the entire body of research data as a whole versus the examining the results from individual studies. A meta-analysis can often more powerfully estimate the “effect size”, the true difference in results, in comparison to the smaller “effect size” of a single study. Measuring “statistical significance” is different than measuring “effect size”. The advantage of measuring effect size via a meta-analysis is that it may reveal actual differences that were missed by examining the statistical significance of the results of the individual studies comprising the meta-analysis.

Let’s have a look at these meta-analyses and see if they have finally put to rest the whole single set versus multiple set debate.

Strength Studies

The first meta-analysis was conducted by Rhea et al (4) in 2002. Examining 16 studies Rhea reported that 3-set training produced superior results to 1-set training. In 2003 Rhea et al (5) conducted another meta-analysis, this time of 140 published studies, and concluded that 4-sets produced maximum strength gains in both trained and untrained subjects. Both of these studies received some criticism due to the criteria Rhea used for study inclusion and also for his statistical analysis methods.

A third meta-analysis conducted in 2004 by Wolfe et al (6) of 16 studies found multiple sets to be superior to a single set in trained subjects and in programs lasting 17 to 40 weeks. As in both Rhea’s meta-analyses, Wolfe’s study received some criticism for his statistical analysis methods.

Aware of the criticism of the previous three analyses, Kreiger (3) conducted a fourth meta-analysis in 2009 specifically designed to improve upon the limitations of the previous studies. He examined 14 studies with 92 effect sizes measured across 30 groups of subjects comparing 1-set, 2-3 sets, and 4-6 sets. He found that 2-3 sets produced 46% greater increases in strength than 1 set in both trained and untrained subjects. Interestingly, he also found no difference in results between 2-3 sets and 4-6 sets. Performing more than 3 sets did not produce a greater increase in strength. Kreiger’s study strengthens the findings of both of Rhea’s previous studies. There were some differences between Wolfe’s findings and Kreiger’s findings in terms of the effect of volume of training but Kreiger’s study also strengthened Wolfe’s finding that multiple sets produce superior results to a single set. Finally, a 2010 meta-analysis of 72 studies by Frohlich et al (1) found single set training to be the equal of multiple set training for short training periods but multi-set training to be superior over longer periods of training.

In summary, there is now a consensus in the research literature supporting the idea that multiple sets are superior to single set training for increasing muscular strength.

Size Analysis

All of the meta-analyses cited above examined differences in strength gains; none examined the issue as to whether single or multiple-set training elicited greater muscle size gains. Increases in strength are caused by both neural and hypertrophic changes and it is possible that the superiority of multiple sets for increasing strength might be due to a greater neural effect and not hypertrophy. It is possible that multiple sets might be superior for increasing strength but not size so this issue needed to be resolved also.

In 2010 Kreiger (2) addressed this topic with another meta-analysis designed to determine if multiple set training elicited greater muscle hypertrophy compared to single set training. Examining 55 effect sizes across 19 groups in 8 studies he found that multiple sets produced 40% higher increases in muscle hypertrophy regardless of the training status of the subjects or the length of the training program. Kreiger also concluded that the 46% greater increase in strength from multiple sets revealed in his earlier meta-analysis was largely due to greater hypertrophy and not neural factors.

Interestingly, while he found no significant difference in hypertrophy from 2-3 sets or 4-6 sets he did find a trend for greater hypertrophy with 4 or more sets. One weakness of his analysis was a limited number of studies that utilized 4 or more sets so he stated that no definitive conclusion could be reached as to whether 4 or more sets was superior to 2-3 sets for inducing muscle growth.

Summary
The debate as to the superiority of single versus multiple set training has been on-going for around 40 years. High intensity training (HIT), originally popularized by Arthur Jones in the 1970s, promotes the idea that single set training is superior to traditional multi-set training for improving both strength and size. Until now research on this topic has been equivocal and unable to resolve the dispute. However, six recent meta-analyses have confirmed that multiple set training produces greater increases in both strength and size than single set training in both trained and untrained subjects.


References:

1. Frohlich M, Emrich E, Shmidtbleicher D., Outcome effects of single-set versus multiple-set training- an advanced replication study. Res Sports Med. 2010 Jul;18(3): 157-75

2. Kreiger JW., Single vs. multiple sets of resistance exercise for muscle hypertrophy: a meta-analysis. J Strength Cond Res. 2010 Apr; 24(4): 1150-9

3. Kreiger JW., Single versus multiple sets of resistance exercise: a meta-regression. J Strength Cond Res. 2009 Sep; 23(6): 1890-901.

4. Rhea, MR, Alvar, BA, and Burkett, LN. Single versus multiple sets for strength: a meta-analysis to address the controversy. Res Q Exerc Sport 73: 485–488, 2002.

5. Rhea, MR, Alvar, BA, Burkett, LN, and Ball, SD. A meta-analysis to determine the dose response for strength development. Med Sci Sports Exerc 35: 456–464, 2003.

6. Wolfe, BL, Lemura, LM, and Cole, PJ. Quantitative analysis of single- vs. multiple set programs in resistance training. J Strength Cond Res 18: 35–47, 2004.

[mbinda]

I never bought into the single set theory not because it can't be a productive way to train but because a lot of the HIT advocates perform 2-3 warm up sets with poundages/percentages that most people would consider pretty decent work sets. Then they want to say that anything but the 1 true 'work set' was just a warm up.

[Retardo-pex]

I never bought into the single set theory not because it can't be a productive way to train but because a lot of the HIT advocates perform 2-3 warm up sets with poundages/percentages that most people would consider pretty decent work sets. Then they want to say that anything but the 1 true 'work set' was just a warm up.

Yeah when you look at a 5x5 program its not that different, you usually only have one "work set" as in the heaviest weight for the same reps but it doesn't always mean the lighter sets are easy.

[Orlando1234977]

The studies are somewhat interesting but always leaves you with two unsolved issues.

1st, Similar to what mbinda brought up. In the testing how are the subjects performing the multiple sets? All to failure? Last to failure? None to failure? And how is the 1 set trainee performing their set, warmups included? no warmups? Are they truly reaching failure? Are they skilled to do so effectively?

I never bought into the single set theory not because it can't be a productive way to train but because a lot of the HIT advocates perform 2-3 warm up sets with poundages/percentages that most people would consider pretty decent work sets. Then they want to say that anything but the 1 true 'work set' was just a warm up.

2nd, Are both classes of subjects given the same frequency of workouts?

ie:
What if the multiple sets group gained 5% better results, but the 1 set group recovered 8% faster due to the lighter workload and therefore could have trained more frequently and possibly recovered or even surpassed the difference?

[Richard99]

1st, Similar to what mbinda brought up. In the testing how are the subjects performing the multiple sets? All to failure? Last to failure? None to failure? And how is the 1 set trainee performing their set, warmups included? no warmups? Are they truly reaching failure? Are they skilled to do so effectively?


2nd, Are both classes of subjects given the same frequency of workouts?

ie:
What if the multiple sets group gained 5% better results, but the 1 set group recovered 8% faster due to the lighter workload and therefore could have trained more frequently and possibly recovered or even surpassed the difference?

When doing a research study all variables have to be controlled for with only 1 variable different between the groups being tested. If there is more than 1 uncontrolled variable then it makes it very difficult to impossible to determine which variable was responsible for the results.

What this means is that in each individual study the researchers equated all training variables except # of sets. Training volume, intensity, frequency, exercises conducted, warm up, etc would all be the same, with # of sets being the 1 difference. That is the only way to determine the effect varying numbers of sets produce.

This is not to say that every singe study used the same volume, intensity, frequency, etc as every other study. The variables would be equated within a study, not across multiple studies.

[delineator]

Who wrote this? You?

It seems strange to read an article & have full details of the authors of the references in the article when you have no idea who wrote the article.

What exactly is the 'new research'? The first 3 references?

[NaturaLPumP]

I only do one work set for the most part. Works for me.

[Richard99]

Who wrote this? You?

Yes, I wrote this article. I have access to a research library and my undergraduate degree is in exercise physiology so I like this stuff and try to keep up with research.

What exactly is the 'new research'? The first 3 references?

The 3 most recent meta-analyses are the new research. I call them new because they were published so recently and I'm assuming that most readers are not aware of and/or have not read these meta-analyses.

[Richard99]

Another point, this one about selecting subjects for a research study.

All the subjects for an individual study will have a similar profile. For example, all the subjects for a study might consist of only Div 1 NCAA football players. Or all the subjects might be males with at least 3 years of bodybuilding experience. Or all the subjects might be young, healthy but sedentary males. Or all might be healthy, sedentary males over the age of 65. And so on.

Not all research will study the same subjects, but in any particular study all the subjects will have the same characteristics. No study examining the effect of # of sets would mix subjects, say 1 group of Div 1 NCAA football plaers and 1 group of males over the age of 65. Each group would have to consist of the same type of subject, with the subjects randomly assigned to a study group.

[bertostare]

i dont see it ...

its simple. if it works for you, then its good, right?

no need to make studies imo

[gomez26]

The debate drew the attention of exercise physiologists around the world, resulting in a growing body of research data examining the issue. However, despite an abundance of research studies physiologists were not able to resolve the issue. The main problem was that the research was equivocal; some studies supported the idea that a single set was more effective than multiple sets, other studies found multiple sets produced greater increases in strength and size, but most studies found no statistical difference in results between the two training methods. In short, there was no consensus in the research.
.

hit typically stagnates after 6-8 weeks. most research doesnt last much longer than this.

i disagree that there has been such a close battle for decades with not much difference that its hard to pick the clear winner. any1 with long term training experience themselves or training others knows that the debate is between delusional whacko's on 1 side & vast examples of real world results on the other.

[Richard99]

hit typically stagnates after 6-8 weeks.

I'm not aware of any research supporting this claim.

The research I've seen is quite clear that single set training (i.e. HIT) works well. Wolfe's meta-analysis of studies ranging from 17 to 40 weeks revealed that single set training was quite effective at increasing strength.

The research shows multi-set training to be more effective but that doesn't mean HIT is ineffective or produces stagnation after a short period of time.

[perpendicularis]

I'm not aware of any research supporting this claim.

[Richard99]

I guess I should have included - if anyone is aware of research showing single set training to be ineffective please cite it. I would be most interested in reviewing it.

[Orlando1234977]

When doing a research study all variables have to be controlled for with only 1 variable different between the groups being tested. If there is more than 1 uncontrolled variable then it makes it very difficult to impossible to determine which variable was responsible for the results.

Yes Richard, I understand that completely. However, since recover will be different regarding somebody doing 1 set, 3 sets or 15 sets, keeping the other variables constant will naturally line up better (optimal) with one group more than another.

ie
Maybe the higher volume would come out on top if the controlled frequency was stretched to a week for all between bodyparts. Or, Maybe the 1 set group would come out on top if the frequenced was 3 week etc. The study, while showing what may be optimal for an average controlled frequency, doesn't necessarily declare a winner. Moderate sets will always trump low or high in meta, for obvious reasons, lower and higher volume programs are set closer to the moderate parameters of recovery/frequency, diluting their actual potential.

[Richard99]

ie
Maybe the higher volume would come out on top if the controlled frequency was stretched to a week for all between bodyparts. Or, Maybe the 1 set group would come out on top if the frequenced was 3 week etc. The study, while showing what may be optimal for an average controlled frequency, doesn't necessarily declare a winner.

Just a point of clarification- none of the cited meta-analysis consisted of a single study. All were meta-analysis of multiple studies. For example one of Rhea's meta-analysis included 140 studies. Wolfe conducted his meta-analysis on 16 studies. Kreiger's two meta-analysis consisted of 14 and 8 studies.

With so many studies being included in these analyses there was a range of every training variable, including training frequency. There was no 1 training frequency per week for all the individual studies analyized. I haven't read all the individual studies but I suspect the training frequency range was 2-4 days per week across all the individual studies.

[Orlando1234977]

Just a point of clarification- none of the cited meta-analysis consisted of a single study. All were meta-analysis of multiple studies. For example one of Rhea's meta-analysis included 140 studies. Wolfe conducted his meta-analysis on 16 studies. Kreiger's two meta-analysis consisted of 14 and 8 studies.

With so many studies being included in these analyses there was a range of every training variable, including training frequency. There was no 1 training frequency per week for all the individual studies analyized. I haven't read all the individual studies but I suspect the training frequency range was 2-4 days per week across all the individual studies.

Yes I understand. It's a combination of many studies. Which means... there's an average frequency across the board. (Like you pointed out in your last sentence) Whichever volume is mosted suited to the average frequency across the board, will probably come out ahead.

If the average frequency was say 6.2 days between bodypart training, could the 1 set group compete with the 15 set group? Doubt it.

If the average frequency was say 2.2 days between bodypart training, could the 15 set group compete with the 1 set group? Doubt it.

So the moderate sets always wins in a typically meta-analysis that averages a moderate rest between training.

[all pro]

That's what I told you two years ago. Three sets, same weight, same reps > one set.
BUT One set low reps {3-5} followed by one set of medium reps {8-12} > three sets, same weight, same reps.

[Richard99]

So the moderate sets always wins in a typically meta-analysis that averages a moderate rest between training.

I understand your point now.

For the training frequencies that most training program employ (2-4 workouts per week) the research indicates multi-set training is superior. Kreiger included some studies examining 1 workout per week in his meta-analysis and found multi-set training to produce greater increases in these studies. Based on this it appears that for someone training 1-4 times per week (or muscle groups 1-3 times per week) multi-set is likely to produce better results than a single set per exercise. For someone training more frequently than 4x per week (say 7 workouts per week or 4 or more workouts per muscle group per week) the results might very well be different.

[Orlando1234977]

I understand your point now.

For the training frequencies that most training program employ (2-4 workouts per week) the research indicates multi-set training is superior. Kreiger included some studies examining 1 workout per week in his meta-analysis and found multi-set training to produce greater increases in these studies. Based on this it appears that for someone training 1-4 times per week (or muscle groups 1-3 times per week) multi-set is likely to produce better results than a single set per exercise. For someone training more frequently than 4x per week (say 7 workouts per week or 4 or more workouts per muscle group per week) the results might very well be different.

Yes that's exactly my point. Though the control systems are valid to observe, the only problem I have is when one study or meta-analysis 'conclude' one is superior, because it will steer towards the middle.

What would be further interesting is to have 9 (or more) distinct meta analysis'..

IE:
low frequency (per body) :
A low volume B moderate volume C high volume

moderate frequency:
D low volume E moderate volume F high volume

high frequency:
G low volume H moderate volume I high volume

For example;
If the results above were:
2, 4, 6
4, 5, 4
6, 4, 2

We'd have moderate winning the overall analysis at a total of 13 (total of middle column), as anticipated since it's better hedged against the extremes in both directions.

However, that 13 doesn't guarantee or prove the best results, it simply shows it's hedged better against the extremes.
With this Hypothetical it explains how it's possible to see high frequency low volume and low frequency high volume both showing more impressive results (6>5), while losing an overall meta-analysis. The conclusion would then be inaccurate to state moderate sets the clear winner.

[liftlightly]

I'm new to this stuff. What is single and multiple sets?

[Orlando1234977]

I'm new to this stuff. What is single and multiple sets?

Excellent question.

A set is a collection of numbers, objects, etc, that is treated as an entity. So a grouping of repetitions without rest, would be an example of one set.

Multiple sets would then be performing the grouping of repetitions (as described above), or variables of different repetitions, more than one time.

Be careful to distinquish and identify the context, as some will reference sets in terms of bodyparts, and others will reference sets in terms of specific exercise.

[perpendicularis]

That's what I told you two years ago. Three sets, same weight, same reps > one set.
BUT One set low reps {3-5} followed by one set of medium reps {8-12} > three sets, same weight, same reps.

You better back that up with a lab-coat's signature otherwise it can't possibly be valid.

[pumplikecuming]

This is new. People who are not retarded found this out ages ago. But one thing with these studies that get me are the total volume, workset repetitions, and all the nooks and crannies.

So for example if a study does 1 exercise, and then thats the only exercise the entire group does, and based off that they did single set vs multiple set. Of course the multiple set group would have greater hypertrophy cause the colume is so low on 1 exercise.

How about them doing one on 5 exercises all 1 set to failure vs 1 exercise 5 sets, last to failure. Or 3 totally different rep ranges all done at equal volume, but set number will varry to accomodate volume.

What I am saying is, its not the sets you do PER EXERCISE, its the OVERALL VOLUME per session A good read is the meta analysis for volume and hypertrophy. So all in all single set training is dumb. But single set training, with multiple exercises (4-6) is not.

[feltmann]

Anyone who needs studies to prove that one set is inferior is not worth listening to on any training matter.

[SuffolkPunch]

Hypertrophy = tension overload first, fatigue second. You can't get both from one set.

[adazona]

For example a big fluffy gown at an informal beachside gathering, a luxurious beaded satin slip with a delicately sheer cathedral length veil for a very formal church ceremony, and a heavily adorned crystal gown dresses for a simple afternoon cocktail affair. What allowed these brides to carry it off so well was their ability to select a gown that suited them.

[gomez26]

only a very small number of ppl have actually had long-term success on 1 set hit style routines.

the hitters will tell u its because the others lack the guts to train hard enough. they claim their methods are based on science when in reality its just a philosophy with zero scientific understanding of variation between people.

the real science will tell u that personal volume preference, like rep range is largely genetic.
http://imbodybuilding.com/articles/hit-vs-volume/?p=1
http://www.ncbi.nlm.nih.gov/pubmed/16003539

[all pro]

only a very small number of ppl have actually had long-term success on 1 set hit style routines.

the hitters will tell u its because the others lack the guts to train hard enough. they claim their methods are based on science when in reality its just a philosophy with zero scientific understanding of variation between people.

the real science will tell u that personal volume preference, like rep range is largely genetic.
http://imbodybuilding.com/articles/hit-vs-volume/?p=1
http://www.ncbi.nlm.nih.gov/pubmed/16003539

http://www.ofz.uni-erlangen.de/pubs/...20training.pdf

[wessej40]

I've been doing a good amount of reading on the primary literature and there are some issues that many people fail to see when reading studies. Studies make claims and results should not be taken at face value until an adequate analysis of their methodology has been completed, otherwise the results mean Jack **** since they could be skewed/biased.

Google A CRITICAL ANALYSIS OF THE SINGLE VERSUS MULTIPLE SET DEBATE
-It won't let me post links yet..

Overall from what I can see from my readings in JPhys is that single sets produce similar results and do not run the risk of overtraining. Additionally, since the time invested is significantly less, it becomes appealing to favor that single set method to failure. Many of the research supporting multiple sets over single sets have biases that can potentially skew results. At the end of the day give single sets and honest effort and see if it's worth it/ works for you. Obviously some people prefer/do better with multiple sets, but you cannot snub it without trying it since SOUND unbiased scientific data is lacking.

-Jeff.