Only have a second? Check out the takeaway below. Have 5 minutes? Check out the rest of the newsletter.

TRAINING TAKEAWAY: Training with heavier loads will likely improve moderate-term strength gains. There is some indication that this relationship may be stronger in trained lifters. Thus, trained lifters could benefit from organizing their training to include frequent exposures to heavy loads to maximize strength.

Background

It's well established that load (% of 1RM) and maximal strength gains have a strong relationship. Put simply, to improve the skill of lifting heavy weights, you need to lift heavy weights. That said, the potential diminishing returns of this relationship are often not discussed. Much of the available data on this topic compares groups training at very light versus moderate-heavy loads (e.g., 30 vs. 80% of 1RM), which doesn’t reflect how lifters train in the real world. A more interesting comparison is in the 60-100% of 1RM range. A recent meta-analysis by Lopez and colleagues (and its corrigendum) offers an opportunity to look at this relationship closer.

Study Overview

This meta-analysis demonstrates that, on average, training with heavy loads (≤8RM or ≥80% of 1RM) leads to greater strength gains compared to light (>15RM or <60% of 1RM) and moderate loads (9-15RM or 60-79% of 1RM). Specifically, both heavy and moderate loads lead to significantly better strength gains than light loads, but the comparison between moderate and heavy loads did not reach the significance threshold (p=0.145, SMD = 0.16). These results, at face value, seem to suggest diminishing returns between load and 1RM strength gains.

However, the question I am most interested in - whether there are diminishing returns in strength gains in the moderate to heavy loading range - can’t be totally answered by this analysis as it dichotomizes moderate and heavy loads. To better answer my specific question, I took all of the studies included in the relevant comparisons from Lopez et al. and performed an exploratory meta-regression. The type of analysis I performed (using within-group effect sizes as the dependent variable) can help to visualize the relationship between strength gains and load when it’s treated as a continuous rather than categorical variable. You can see the overall model below:

While this comparison is not without limitations, the relationship between load and strength gains was not strong (a pseudo R^2 = ~0%), potentially indicating diminishing returns. This ~0% value suggests that variation in strength gains was not at all explained by the load used. However, we know training status is an important variable to consider when examining resistance training research. When I included training status as a variable within the model, the pseudo R^2 improved considerably to ~70%. This indicates that training status is likely playing a role here. To be clear, the interaction between load and training status was non-significant (p > 0.05), but I speculate this could be due to the inability of the sample size to detect small differences that may be present. In this case, it's plausible that the “threshold” of load needed to improve strength gains at the maximal rate in untrained individuals is lower than that of their trained counterparts. So, I created two more models this time separating the data by training status. Here are the meta-regressions for each trained and untrained participants:

Here you can see a stronger relationship between load and strength gains in trained versus untrained individuals (pseudo R^2 values of ~18% vs. ~3% of the variance explained for the respective models). Tentatively, this may suggest that training status impacts the relationship between load and strength gains. Specifically, training with heavier loads may be more beneficial for trained individuals. That said, it's important to keep in mind this analysis further limits the pool of already sparse data (from 6 down to 3 studies) and the actual interaction was non-significant. So, take this with a grain of salt.

Application

Overall, the relationship between load and strength gains is well supported. From diving into this data, there’s some indication that the relationship is stronger in trained individuals.

Anecdotally, consistently training with heavy loads can come with a high recovery demand, so it’s essential to find the loads that elicit the best return on investment on an individual level.

Finding creative ways to organize training so that experienced lifters can tolerate frequent exposures to heavy loads while minimizing unnecessary recovery burden is likely the name of the game. This is where manipulating things like protocol type (e.g., top set + backoff sets), proximity to failure (e.g., 3 sets x 6 reps @ 80% of 1RM vs. 6 sets x 3 reps @ 80% of 1RM), amount of volume allocated towards the main lifts versus accessory movements, and exercise selection can help support heavier average and peak loads.