Autoregulation has become a mainstay programming term for those looking to maximize strength or physique development, often being used as shorthand for “use lighter weights when you’re under-recovered.” Two primary issues arise from this: 1) while this can be a great method of autoregulation, it isn’t always the best adjustment depending on the circumstances; and 2) while autoregulation has permeated evidence-based fitness, the scientific evidence surrounding its utility isn’t quite as straightforward as you might think, which leads to some interesting considerations about its implementation in practice.
Autoregulation: A Bidirectional Input-Output System
There are a few key aspects of conceptualizing autoregulation. First and foremost, we can define this idea as an adjustment to programming variables based upon some training- or non-training-related signal. Crucially, this definition encompasses not only adjusting variables down when under-recovered, but also adjusting variables up when well-recovered (this idea will be important for us later). Autoregulation is bidirectional: we are merely responding according to the signal. Which leads to the next point, that the signal should be chosen specifically based upon the goal of the set. I prefer to think of this as an “input” from which a decision can be made regarding the “output” (whether to change a programming variable), which also should be specific to the goal of the set.
The Logic Checks Out, But Does the Research?
Autoregulation lies at a very interesting crossroads between research and practical application. There have been a handful of studies on the topic now, with results that require a bit of thought to make sense out of.
Strength:
Two studies examining the effects of autoregulation on strength have led to similar findings, but to truly contextualize the results, we need to look a little deeper than the surface. Both Helms et al. (2018a) and Graham and Cleather (2021) compared differences in 1RM (one repetition maximum) strength gains between experienced lifters training with either a predetermined progression scheme using a percentage of 1RM or an autoregulated approach where they were allowed to select their own weights to reach certain progressive RPE (rating of perceived exertion) targets, controlling how close to failure they trained. In this way, both groups utilized progressive overload with heavier weights as time went on (which we know is a great recipe for an effective strength program), but one group was allowed to autoregulate the load on the bar while the other could not deviate from the fixed program.
Let’s take a quick detour back to our conceptualization of autoregulation: in these studies, the input the autoregulating group used was training performance at the target repetition range, which is a logical choice specific to strength training. The output was an adjustment to the load on the bar; again, a logical choice specific to strength training, being arguably the most important variable for strength development.
After 8 weeks of training in Helms et al. (2018a) and 12 weeks in Graham and Cleather (2021), significantly larger strength increases were reported in both instances for the autoregulating group compared to the fixed-training group. The differences were relatively small but meaningful, indicating that adjusting the weight on the bar due to differences in perceived performance capabilities led to slightly better outcomes than following a rigid structure.
When examining the findings of any research study, it can be helpful to ask yourself why the findings are what they are. In the case of these two studies, it’s pretty clear: the autoregulating groups trained with higher %1RM on average than the fixed-training groups. And, like we said before, training with heavier loads is generally a great way to get stronger. This is a major reason why I emphasized early on that autoregulation is bidirectional; these findings suggest the same thing we see in the gym all the time: if you’re recovering really well and are able to increase the load on the bar a little bit faster without overshooting your RPE targets, then you’re likely to make faster strength progress because you are keeping pace with your adaptations.
Hypertrophy:
Oddly, there is more evidence regarding hypertrophy outcomes with or without autoregulation, but less clarity. I have a hypothesis as to why this is, which I’ll circle back to at the end of this section.
For starters, the same Helms et al. (2018a) study I mentioned earlier also measured changes in muscle thickness of the pecs and two different sites on the vastus lateralis, and found no differences in pre-to-post changes between groups. Again, let’s ask ourselves why… we know that hypertrophy, unlike strength, is not particularly dependent upon the weight being lifted. Therefore, it makes sense that an input and output specific to strength development might not apply the same way to hypertrophy-focused training. We need to find a different route.
In my view, the most viable option for an output that is specific to hypertrophy is volume (number of sets performed per muscle). Not only has meta-analysis established a clear, positive dose-response relationship between hypertrophy and set volume, but there are also a few potential inputs that could make sense to modulate set volume.
Gomes et al. (2022) and Bartolomei et al. (2024) both conducted studies using subjective inputs to adjust set volume in their autoregulation groups, comparing differences in muscle growth to a fixed training protocol. In these studies, the set volume in a given session was either decreased, maintained, or increased (more instances of bidirectionality) based upon the session-RPE of the previous session (in Gomes et al.) or perceived recovery status at the time of training (in Bartolomei et al.). Most recently, for my doctoral dissertation, we examined set volume autoregulation using an objective metric of set-to-set performance. The reason we used that input is because it is the only one with prior evidence to back it up, with an acute study from Helms et al. (2018b) demonstrating that when using performance as a guide on how many sets to do, set volume did in fact differ. Regardless, none of these studies identified a significant difference in hypertrophy between the autoregulating and fixed-training groups. Unlike strength, there has not been a demonstrable benefit to utilizing autoregulation when training for hypertrophy, despite trialing multiple different inputs and outputs.
But let us return once more to why that lack of difference may have occurred. In all three of these studies it’s possible to estimate the average sets performed per week (or per session) in each group. And in all three of these studies, the differences are negligible: only 1-2 additional sets per week in the autoregulating groups. It makes sense then why hypertrophy was very similar, because volume was very similar.
There are two more important points to make about these results. First, a lack of superiority for autoregulation does not indicate inferiority. Rather, these results suggest lifters can autoregulate their training volume without sacrificing results, which is ultimately what most of us are really after anyway: adjusting our plan when life gets crazy without worrying about missing out on gains. Second, unlike in the strength studies where the autoregulating groups trained with a consistently different crucial variable for their outcome, here we are not seeing consistently different volumes. Perhaps if a study included a fixed condition that was relatively easy, the autoregulating group could take advantage of accruing more volume in comparison. Or, looked at another way, perhaps neither group was under enough fatigue to make autoregulation actually worth doing. If both groups were exposed to consistently high fatigue but one was allowed to autoregulate while the other had to train through it anyway, it’s possible we see different results. In that case, even if autoregulation leads to consistently lower volume (which is likely) and therefore less hypertrophy in the short term (as I would hypothesize), it’s plausible that autoregulation may lead to a difference in adherence and sustainability which would lead to greater hypertrophy over the longer term.
Where Does That Leave Us?
The research on autoregulation is still very new and evolving. From a scientist’s perspective, I think it’s safe to say autoregulation has, at worst, a neutral effect on strength and hypertrophy. But from a practitioner’s perspective, I think most lifters would benefit from utilizing autoregulation in some capacity. The number of times a client has told me they’re going into their highest volume leg day or heaviest deadlift session and are having some external life variable (poor sleep, work/family stress, time constraints, etc.) getting in the way is impossible to quantify. The same goes for the number of times I’ve seen people rigidly sticking to programs that are too easy and holding their progress back. So, maybe the best thing to leave you with is my best approximation of how to autoregulate our primary training variables, based upon a combination of the available research and my coaching/training practice.
I want to acknowledge also that despite the very wide practical utility of autoregulation, I was careful with my word choice above when I said “most lifters” benefit from its use. There are those that for various reasons a strict program is the best call, actually. Some individuals have a harder time with things like introspection or self-regulation, or “being in tune with their body” - if you find that you struggle to judge proximity to failure accurately, manage your emotions or expectations around training and nutrition, eat mindfully towards body composition goals without strict macronutrient targets, and so on, you may find some aspects of autoregulation don’t suit you. Also, even when you don’t feel particularly ready to perform, you might be surprised how well you can still show up when it comes down to it. Oddly, some of my and my clients’ best training sessions have come after poor sleep or feeling a little under the weather. For that reason, I think autoregulation is best used reactively, not proactively. Go in with a plan to perform as normal, gather your input(s) specific to your training goal, and adjust (or don't adjust) your specific output accordingly.
Find more from Jake on…
Instagram: @jake.remmert.phd