This post is an excerpt from a project I started working on close to two years ago and which may never see the light of the internet. But there are some decent sections in it which may be interesting. This piece is taken from a chapter called “The Overtraining Myth”.
The Math of Recovery
The stability-via-change dance of allostasis leaves supercompensation theory in an awkward position. No depleted biomolecules to trigger the supercompensation effect. No exhaustion of stress hormones to leave you overtrained.
Most everything the textbooks tell us to think about recovery has to be viewed through new eyes. Supercompensation isn’t exactly wrong, but it’s not quite right-enough either.
The question can be put simply: What is recovery, exactly? Are you recovered when your muscles have their energy stores replenished? When proteins have rebuilt themselves? When the muscle stops showing signs of tissue damage? When you don’t feel so sore anymore?
What about the tendons and ligaments and other connective tissues? What about the immune system? The autonomic nervous system? Your mood and emotional state? Do all those systems supercompensate? Do they all peak and return to their set points on the same schedule?
Does it even make sense to talk about “recovery” as a concept? Tempting as it is to focus on our muscles, they aren’t our number one priority. Tissue repair processes carry on their normal business unless you’re in an extreme state of starvation or trauma (trauma like massive burns or infection).
Even the words we use for recovery take for granted a “hydraulic model”, in which our bodies have a mysterious energy kept in a fuel tank somewhere. These recovery energies, which we draw upon to heal ourselves, aren’t different from the vital force or the vitreous humours once thought to animate living organisms. When that reserve is topped up, by good food and sleep and relaxed mood, we’re “recovered”. When we drain that reserve, we’re exhausted and sick.
That’s the essence of reductionist, linear, mechanical thinking. There’s even an element of pseudo-science quackery hiding in there. As a metaphor, the notion of “adaptive energies” works. As a reality, it fails.
If “recovery” is to be useful at all, there has to be more to it than waiting on your biceps to recover for seven days. No single measurement can describe you as “recovered” or “not-recovered”.
Later in his career, Selye wrote of stress as having both positive and negative qualities. Positive eustress promotes growth and vitality, while negative distress leads through the resistance and exhaustion stages. Exercise represents a disruption to our bodies, which can be good in moderate amounts or harmful if taken to excess.
In the early 1980s, sports scientist Eric Banister realized this and developed a model accounting for the positive and negative effects of exercise. Banister’s two-factor model treats workouts as having both fitness-building benefits and negative stress-induced effects (henceforth called fatigue).
The key insight here is that the positive and negative factors operate simultaneously, with the tissue-repairing recovery things and the bad stress-response things going on at the same time. You can feel absolutely smashed the day after a hard session while still experiencing the positive muscle- and strength-building effects of training. The trick is that you don’t always notice these fitness gains right away, not because you’re “not recovered”, but because the accumulated fatigue effects drown them out. How well you can realize your fitness level at any moment depends on the balance of positive and negative after-effects, otherwise known as preparedness.
“Recoveredness” isn’t about muscles or the CNS or any of the pieces. It’s about the accumulated disorder across your whole body relative to your current fitness level.
While I enjoy some good navel-gazing over biology, we also have to ask if the two-factor (or “multi-factor”) model makes any real difference in what we do. If your body’s so stressed out that it’s not working right and our mood and gym performance reflect that, does it really matter that we’re not technically exhausted?
Diminishing Returns — Or Positive Feedback?
The notion of a limited pool of “recovery energy”, that our capacity to recover from punishment has limits, is not entirely unfounded. Any extreme physical trauma or illness will readily demonstrate the limits of the human body, and it isn’t terribly difficult for an athlete, recreational or otherwise, to find that point through exertion.
However, we also assume that our powers of recovery never change throughout our life. A lifter reaches some “normal” level of recovery-powers, able to handle squatting twice or maybe three times a week, but no more. You’re riding a fine line, and doing more is simply impossible. Don’t even think about it.
We are of course back at the ideas of the HIT and Hardgainer schools, who believe that more training leads to diminishing returns. Physical exercise and recovery from it exist in an “either-or” dance of negative feedback. You can either do lots of training, or you can be well-rested. Never both.
You might find some progressive thinkers who grudgingly agree that training more does lead to larger absolute results, but these gains are only marginal when compared to the efficiency of fewer workouts. According to this thinking, the majority of your gains come from a baseline of two or three weekly workouts. This creates perhaps 80-90% of your possible gains. Further sessions might add to your progress, but only in proportionally smaller increments. A fourth session might only add 5%, a fifth only 2%, and so on. Instead of milking that 90% and calling it a day, you’d be adding two extra days for a piddling 7% on top of your gains. After a point it becomes so time-inefficient that you’d be a fool to bother.
There is research in favor of this view. Meta-analyses done independently by Matthew Rhea of Arizona State University and Mathias Wernbom of Sweden’s Göteborg University analyzed the body of published strength-training studies and identified inverted-U “sweet spots” for both frequency and volume. The inverted-U is like a camel’s hump on a graph. Toward the zero-point, stimulus is low and the effect is low; you don’t do much and you don’t get much back. At the high end, stimulus is high and the effect is still low, suggesting that more isn’t always better if you overwhelm yourself in the process. The effect is maximized near the middle, when a moderate stimulus generates the largest the effect.
Rhea and Wernbom independently reached similar conclusions on the effects of strength-training based on the findings of published research. Frequency maxes out around two or three sessions per week per exercise, and number of work sets starts at around four for a beginner and expands to around eight in more advanced athletes. This all fits with, if not a perspective of Doing Less, at least an eye towards moderation. That’s a position I find reasonable — not aiming to do the least possible amount of training, but an ideal amount of training to net you the best-possible gains for the work that you do.
But there is a caveat. The papers used in these meta-analyses were largely limited to the untrained or lightly-trained subjects so often used in academic exercise science, with a paucity of data on more advanced athletes. These numbers were only ever intended to illustrate what works as a theoretical starting point, and for that purpose they work wonderfully. Let’s keep them there, as guidelines rather than the final answer.
What these papers do indicate clearly is a trend towards “more” as athletes become more trained over time. Workloads used in training expand and unfold as you improve, trending toward more sets and more workouts as strength improves over a career. Your body becomes more resistant to overload as you get stronger, and as a consequence of that resistance you need “more” in order to encourage adaptation.
Supercompensation and the hitpoints model of recovery are more philosophical views on exercise — dare I say dogma — than hard facts of science. The scope of evidence drawn from outside exercise physiology suggests that our bodies just don’t work like that. Banister’s fitness-fatigue model, as one conspicuous example, explains the situation just as well, and in some instances — like the unpacking of workloads in more advanced lifters — it is superior.
The marginal returns view, while probably close to the mark for any brief snapshot of training time, doesn’t account for findings like the Frequency Project.
We need to think long-term. Stress has positive effects, remember, and these can add up over time. Your body doesn’t retreat to a point of stable equilibrium; it can adopt many states depending on what you do and what you’ve done previously.
The trick is to expose yourself to stress in such a way that it doesn’t overwhelm you. Hanging out in stress-mode all the time is bad, but spending a lot of time cruising along with occasional bursts of all-in effort is actually good for you. This pattern of under- and over-stimulation is crucial (more on this later).
By gradually coaxing your body into a more robust condition, you can handle and thrive on more training. This is the crucial difference in the two models. Supercompensation treats the body as a finite system with a cap on its capacity to grow and adapt — and it low-balls the cap at that. Drawing on models influenced by complexity and allostasis — of which Banister’s two-factor model is an example — we can instead treat the body as a “growth system”, one which does have limits but is nevertheless far more adaptable than the simple model would allow. We are less machines with limited resources and more ecosystems which can adapt and grow with the right encouragement. Diminishing returns can be, to some degree, replaced by increasing returns.
Russian sports science, while still using the imagery of “adaptive energy”, knew that a critical goal of physical training was improving the athlete’s resistance to high workloads (and they knew that there were no real energies, but rather underlying biological processes that restored you to a state of normalcy). In Supertraining Mel Siff wrote of these adaptive reserves, noting that “the capacity of these reserves is not fixed, but alters in response to the demands placed on them by stresses such as training.”
Recovery itself, in as much as the word has a meaning at all, can be trained with practice. Conditioning yourself for lots of training means you can handle lots of training. The more you expose yourself to stressful events, within reason, the less stressful they become.
On one level this is obvious. A workout that absolutely floors a beginner can be a light day for an advanced lifter. You don’t stop training in your first few months of lifting weights because you got really sore after those workouts. We take it for granted that we’ll become more robust with experience, at least in that dimension.
Instead of diminishing returns, we can think start thinking about positive feedback — that doing more can create more. Training and recovery are more of a “both-and” situation, rather than the “either-or” of supercompensation. Your body’s condition, its ability to handle stress and to recover from it, depends at least in part on being exposed to stress in the first place.
Want to train the same exercise five days in a row? You can, and you can benefit from all of those workouts. You eventually have to pay down your fatigue debt, of course, and in order to see your gains realized — to see that new muscle growth, to lift a new PR weight — you may have to let the fatigue bleed off.
This is not about recharging your hitpoints. Instead, we’re letting the stress-response stabilize and settle down. It doesn’t matter if you train “legs” or “arms” two or three or five days a week as long as you allow for adaptation to occur at some point in time.
Once you stop looking at recovery as hitpoints, a whole world of new possibilities opens up. Training is a continual process of biological growth and change which is balanced and checked by accumulated wear-and-tear. You’re always walking on a knife’s blade, at “the edge of chaos” so to speak, but as your fitness level grows, so does your ability to tolerate a thrashing. You can handle more intensity as weights get heavier, and more volume and more workouts as work capacity improves. The result is a system which grows to match what it’s been trained to do.
It would be unfair to say that workouts based on supercompensation “can’t work”, because obviously they can and do. Minimalist-style workouts, ranging from straight-up HIT to meat-and-potatoes powerlifting workouts work, and work just fine. For some people.
I don’t think supercompensation is wrong as much as it is incomplete, much as Newton’s laws of gravity can be treated as a special case of Einsteinian relativity. It’s in the incompleteness that we find possibilities.