Maximizing Muscle Growth

Preface

At The Gym

• Untrained people should work out 3 or more days per week. Trained people can get away with 2. dose response for strength development. Training twice per week is better than training once per week, but not after controlling for weekly volume Effects of resistance training frequency on measures of muscle hypertrophy: a systematic review and meta-analysis.Training 2 or 3 days per week is better than training once per week, but not after controlling for volume Grgic.
• You should perform about 4 sets for each exercise dose response for strength development. Multiple sets causes 40% more growth than a single set, and this gain seems almost entirely concentrated in the second and third sets Single vs. multiple sets. Multiple sets yield greater gains than a single set Bågenhammar Durall.
• Resting 3 minutes between sets promotes more growth than shorter durations rest periods.
• As long as you spend less than 10 seconds, it doesn't matter how long each rep takes repetition duration.
• There's no reason to train to failure repetition failure.
• Novices should lift as many reps as they can while lifting 60% of their max, while trained lifters should lift 80% dose response for strength development. According to the Brzycki formula, this corresponds to about 23 and 8 reps, respectively.
• Trained lifters should lift as many reps as they can while lifting 85% of their max Maximizing strength development in athletes. According to the Brzycki formula, this corresponds to about 5 reps.
• The free-weight bench press activates more muscles than the machine bench press, weakly suggesting that free-weights might prove better at stimulating muscle growth than machine exercises McCaw.
• Full-body workouts appear as good as split-body workouts Ratamess Benton.
• Working out once per week may be enough to maintain muscle mass Trappe. A third the original is enough to maintain strength Bickel
• When comparing dumbbells, barbells, and machines, it's complicated... Farias.

In the Kitchen

• Protein up to 1.6 grams per kg of body weight per day improve muscle gains protein supplementation - that's about double the 0.8 recommended for the general population. A literature review recommends higher levels for people losing weight: between 2.3 and 3.1 dietary protein during caloric restriction g/kg/d.
• Protein supplements work Protein supplementation, especially whey advantage whey,
• Consuming more than 30g of protein in a meal yields no benefits, so you have to split your protein consumption throughout the day moderate serving of high-quality protein.
• Consuming protein around the time of your workout has no effect on muscle growth except insofar as it may boost your total protein consumption protein timing.
• You should definitely take creatine supplements, which have been found to increase muscle mass by something like a sixth of a pound per week among older adults. The only side-effects found were minor gastrointestinal issues creatine supplementation.
• Caffeine doesn't increase strength Astorino.

General Health Recommendations

Literature Reviews

[TODO: Integrate this section into the "At The Gym" section]

One literature review Manipulating resistance training program variables to optimize maximum strength in men: a review found

Another literature review is (TODO): Bird

A third literature review (TODO) Wernbom concludes

Creatine Dosage

I couldn't find any meta-analyses on the optimal creatine dosage so I decided to take a look at the data on my own. I looked at the creatine meta-analysis creatine supplementation and assumed the average American weight of ~182lb to convert between studies using creatine in g vs. g/kg. Then I constructed a scatter plot of the effect size on lean body mass (relative to the control group) after adding a little bit of noise for illustrative purposes. Unfortunately, the chart isn't very illuminating:

The charts representing the effect of creatine on chest press and leg press are similarly unhelpful.

Next, I tried dividing the studies into low-creatine (< 5.8 g/d) and high-creatine (>8.2 g/d). The average effect size in the high-creatine was 0.23 higher than in the low-creatine group. However, the p-value was fairly high (p=0.26) and this difference represents just a 16% increase in effect size, suggesting that the first ~5 g/d of creatine represent most (if not all) of the gains.

This analysis has two major problems, however. The first is high error bars due to small sample sizes and noisy muscle growth. The second is that there just aren't that many studies on the effect of very high creatine consumption on muscle growth (just 2 where more than 20 g/d was consumed).

Both these problems are mitigated by examining chest press strength gains instead:

Here, we actually find a negative difference between low and high creatine consumption, but p=0.4, so really the finding is just "no effect".

In short, while there's clear evidence that non-zero creatine consumption promotes muscle growth, we don't have a clear idea on how much creatine is ideal. There isn't good evidence to suggest that consuming more than 5 g/d helps much.

Suspicious Studies

Now I want to move on to the controversial part of this page where I disagree with the academics with decades of experience. I think several of these meta-analyses suffer from significant problems.

Take dose response for strength development, for example. The authors found that trained athletes saw smaller returns from exercising 3 times per week rather than 2.

However, I'm suspicious of this study. It didn't look at a bunch of studies that investigated this question. Instead, it looked at a bunch of studies that investigated weight-lifting generally and compared how much strength increased relative to the group's standard deviation.

This choice of analysis doesn't make much sense to me. Suppose the meta-analysis had two studies:

1. A set of 10-year-old apathetic girls do 2 sets of bench press 3 times per week. Each lift is 3 inches. Another set is the control group.
2. A set of driven 20-year-old men do 1 sets of bench press 2 times per week and are given 90g of protein as supplements. Each lift is 2 feet. Another set is the control group.

It's likely the second study would find a larger effect size due to the demographic and procedural differences. The analysis in dose response for strength development would then conclude that 1 set is better than 2, and 2-days-per-week is better than 3. This is absurd reasoning.

They mention that they also did an analysis based on gender and age, but (1) they don't present this analysis and (2) this still ignores other factors such as protein intake.

The analyses Effects of resistance training frequency on measures of muscle hypertrophy: a systematic review and meta-analysis, Maximizing strength development in athletes, and Grgic also suffer from the same problems.

Instead, I think we're better off finding studies that keep the same sets and reps and just change the number of days worked per week.

I should note that some of the studies don't fall victim to these issues. For instance, Single vs. multiple sets and Bågenhammar both look at studies with groups that differ only in one variable and look at the differences in muscle growth between the groups.

At some point I should come back to this and re-examine the other meta-analyses (TODO).

These short-comings are a big problem since these were the studies we were relying on for optimal reps and optimal days-per-week estimates. For this reason, I think it's worth diving into these two topics a bit more.

Optimal Number of Reps (Bench Press)

This table lists studies where participants performed 3 sets of the bench press 3 times per week. "X > Y" means that performing X reps led to greater gains than Y reps. Some of the results are not statistically significant.

It looks like 5-9 reps is probably around optimal.

Days per week

TODO