The most comprehensive meta-analysis to date supports a relationship between volume and both hypertrophy and strength. However, you get diminishing returns for both as volume increases, particularly with strength. Higher frequencies may be better for strength but with diminishing returns. Frequency appears not to have much of an impact on hypertrophy when volume is equated.
Overview
- What did they test? The authors performed a meta-regression of 67 studies comprising of 2,058 subjects, looking at the impacts of volume and frequency on size and strength.
- What did they find? Both hypertrophy and strength increased as volume increased, but the diminishing returns were much stronger for strength than for size. Frequency had a negligible relationship to hypertrophy. However, strength gains increased with increasing frequency, but with diminishing returns.
- What does it mean for you? Volume is an important progressive overload tool for both strength and size. If you’ve hit a plateau, try increasing your volume, either by adding more training days in the week, or by increasing the number of sets.
What’s the problem?
Purpose
Volume and frequency are one of the two biggest tools in your toolbox that you can manipulate to maximize your gains. However, they are also sources of confusion. “How many sets should I do?” or “How many times a week should I lift?” have been frequent questions since the dawn of weight training. Research can give us some guidance, but it has limitations. Studies can seem like they’re all over the place as to the impact of volume and frequency on size and strength. Part of this is due to the small sample sizes in resistance training studies. It would be ideal to do a study of 500-1000 subjects, but that is not possible in the real world. There’s generally not a lot of money to do resistance training studies (it takes money to do research), and the studies involve volunteers who are willing to devote 3-6 months of their time to a study. Since they are often done at universities, the time frames of studies are usually within a typical quarter or semester (6-15 weeks). Thus, we end up with a lot of studies with groups of only 8-30 subjects or so that last around 2-4 months.
Changes in muscle size are very small. Improvements in muscle thickness, as measured by ultrasound, are in millimeters. There’s also a lot of variability from one person to the next in the response to training. Thus, if you’re doing a study comparing two different types of training programs, it can be difficult to see differences between groups, even if a difference exists. This is especially true with the small sample sizes and short training program durations. Thus, it can be helpful to look at an overall body of research to identify trends. One tool used by scientists is called a meta-analysis. This is where you pool studies together and statistically analyze them. Meta-analyses can be useful when you have many studies with small sample sizes.
I was the first to publish a meta-analysis on set volume and hypertrophy in 2010 1. I found there was a tendency for increasing hypertrophy as you went from one set per exercise to 4-6 sets per exercise. However, I only had eight studies to work with at the time, and I only looked at sets per exercise rather than per muscle group. In 2017 I worked with Brad Schoenfeld and Dan Ogborn to publish an updated meta-analysis with 15 studies 2. We found evidence of a dose-response relationship, with increasing hypertrophy up to 10+ weekly sets per muscle group. We had insufficient data to draw any conclusions at much higher volumes than that. Also, only two of the studies we included in our analysis involved subjects with weight training experience.
Meta-analyses on set volume and strength have also been published, showing better strength gains with higher volumes in terms of either sets per exercise per session 3 or weekly sets per exercise 4. In terms of training frequency, past meta-analyses have shown that frequency has a negligible effect on either strength 5 or hypertrophy 6 when weekly volume is equated.
All of these meta-analyses have limitations. For example, they often use categorical variables to classify volume, such as “low” (<5 weekly sets per muscle group), “medium” (5-9 weekly sets per muscle group), and “high” (10+ weekly sets per muscle group) 2. The problem with this approach is that it’s arbitrary, and may not capture meaningful differences within a category. For example, perhaps there’s a difference between 10 weekly sets and 15 weekly sets, but both are grouped into the 10+ category. Thus, it can be difficult to establish a true dose-response curve, and it may be better to treat set volume as a continuous variable rather than categorical variable.
Another limitation is that these analyses quantify all sets for a given muscle group as equal, whether it’s a primary force generating muscle (the pecs in a bench press) or a synergist muscle (the triceps in a bench press). However, some evidence suggests that synergists may not grow as much with indirect work as with direct work 7. Thus, it may be more appropriate to quantify sets involving synergist muscles as something different, like a half-set.
Finally, some past meta-analyses have used linear regression to analyze the data. This assumes a linear relationship between set volume and hypertrophy or strength. However, it’s likely that the relationship is non-linear, such as reaching a plateau at higher volumes. Thus, analyses are needed to explore these potential relationships.
Thus, the purpose of this meta-analysis was to explore the nature of the continuous dose-response relationships between:
- Weekly set volume and hypertrophy
- Weekly set volume and strength
- Weekly frequency and hypertrophy
- Weekly frequency and strength
What Did They Test and How?
The authors searched PubMed/Medline and Google Scholar databased for studies published up until April 2023. They used a variety of search terms related to hypertrophy or strength to find studies. Also, if the authors became aware of any studies between April 2023 and June 2024, they were included.
The inclusion criteria for studies were as follows:
- A study, pre-print, thesis, or abstract (if the researchers were able to correspond with and get data from the authors)
- Published or pre-printed before June 2024
- Available in English
- Resistance training lasting a minimum of 4 weeks, utilizing a randomized experimental design (either within- or between-group)
- Healthy participants
- Did not involve participants 70 years old
- Compared at least two groups featuring differences in set volume and/or frequency while controlling for the load (±5% of 1RM or ±2 RM) and proximity to failure (failure, non-failure, or mixed)
- Included pre- and post-measurements of muscle hypertrophy with a direct, site-specific measurement (ultrasound, computed tomography, MRI, and/or muscle biopsies) or included pre- and post-measurements of dynamic (up to a 10-RM), isometric, or isokinetic maximal strength
- Had not been retracted or called into question by other researchers
The quality of included studies was assessed using the TESTEX scale, which is designed specifically for exercise training studies. Data were extracted from the studies and included variables regarding study design, measurements, subject characteristics, training protocols, and outcomes. All sets were classified as direct or indirect, allowing for three classifications (“total”, “fractional”, and “direct”). “Total” was the sum of direct and indirect sets, while “fractional” counted indirect sets as a half set (indirect x 0.5 + direct). “Direct” did not account for indirect sets. For hypertrophy, direct sets were those in which the measured muscles were likely to be the primary force generator in the exercise. Indirect sets were those in which the measured muscle was likely to be meaningfully trained but not the primary force generator. For example, if a study had 5 sets of bicep curls in one training session and 5 sets of rows in the other session during a week, the weekly “total” sets would be 10, the “fractional” sets would be 7.5, and the “direct” sets would be 5. For frequency, “total” would be 2 days per week, “fractional” would be 1.5, and “direct” would be 1.
For strength, direct sets were those that trained the exact exercise used for strength assessment. Indirect sets were any that were likely to meaningfully train the muscle(s) involved in strength assessment. For example, let’s say you have a study that measures back squat 1-RM strength. It involves 5 sets of squats in one session, 5 sets of squats in a second session, and 5 sets of leg presses in the third session. The weekly “total” volume would be 15 sets, the “fractional” volume would be 12.5, and the “direct” volume would be 10. For frequency, “total” would be 3 days per week, “fractional” would be 2.5, and “direct” would be 2.
The researchers analyzed the data using multilevel meta-regression. This allows you to model the variation between muscles or exercise, between groups, and between studies. Here’s an example of how a multi-level meta-regression model looks from my first meta-analysis on strength 3 (Figure 1):
Figure 1
