Methods to Madness

Reading Greg Nuckols and thinking about what it means to get stronger by science.

In the noisy chaos of science-based wellness on social media, there are a few reasonable voices. One of my favorites¹ is Greg Nuckols, a champion powerlifter who has been blogging about the science of strength training for well over a decade. Over the weekend, I read his 2015 books The Art of Lifting and The Science of Lifting, and they hammer home how a simple topic can be made overly complicated when we decide to write scientific papers about it.

Nuckols’ two-volume series is less than two hundred pages long and written in an engaging, bloggy voice. Though the books link out to systematic reviews and randomized trial reports, he doesn’t get bogged down in that material. Instead, he focuses on one simple modeling principle and its consequences: adaptation.

Though the human body is impossibly complex, its many interconnected systems deal with stressors in a surprisingly uniform way. Stressor here means all sorts of things: wounds, infections, temperature changes, environmental changes, activity changes. Something out of the ordinary. The model of homeostasis posits that bodies actively work to stay the same as much as possible. Most of our daily activities don’t stress the body, and we remain in a pleasant, steady state. However, the body changes and adapts if it gets pushed far enough out of its comfort zone. If you apply enough stress, the body initiates a response to prevent harm or death. There are limits to this response, and too much stress will cause injury or death. But in the right Goldilocks range, the body will reconfigure itself to resist the stress next time it sees it.

Exercise can be modeled as a stressor. It’s certainly stressful. It increases muscle tension and body temperature. It increases demand for oxygen and nutrients. It lights up your sympathetic nervous system and triggers the release of adrenaline. Your body responds to exercise similarly to how it fights off other stressors like injuries and infections. It panics and then marshals resources to make sure it sucks less the next time you go to the gym.

There are many models for the temporal behavior of the stress reaction. The most common mathematical model is the fitness fatigue model developed by Tom Calvert, Eric Banister, and collaborators in the 1970s. Notably, Calvert was an electrical engineer who introduced his physiology colleagues to cybernetics and impulse responses. The fitness-fatigue model is a parametric model of adaptation: stressors introduce bad effects (fatigue) from which the body works hard to recover quickly. It also introduces positive training effects, the adaptations you were chasing. The fitness-fatigue model posits that the positive effects decay more slowly than the negative ones. For the dynamical systems nerds out there: they model this with a simple second-order linear system. Hence, if you repeat this process of enough but not too much stress over time, you nudge your body to handle more and more stress.

This feels science-y, no? The general adaptation syndrome is a well-tested, clean model that certainly helps guide how to train athletes and implement physical therapy. This model hasn’t changed since 1980. It prescribes working hard and continuing to do so over time to increase adaptation. It predicts that training harder than last time is beneficial. These two together form the principle of progressive overload, a very simple concept that influencers try to make absurdly complicated. Moreover, this model of adaptation tells you that you need to recover to hasten the elimination of fatigue. It even suggests implementing tapering periods before competition, a common practice.

OK great, now let’s try to apply this model to practice. What is the right shape of the actual stress to give you the results you want? If you want to get stronger, which exercises should you do? How frequently should you do them? What weights should you use? How many repetitions should you do for each exercise?

Unlike most of the charlatans in the science-based fitness space, Nuckols is brutally honest about this: the answer to all of these questions is that we don’t really know. Or at least, we at best only have partial answers. You probably need to go above some nominal effort to induce adaptation. Because there is so much intersubject variability, the exact perfect amount is hard to pin down. But if you work hard and make sure to rest and eat, you’ll get stronger over time.

Nuckols thinks the broader scientific research on sports training tells us more than I think it does. From my reading of too many studies in this space, the ns are always really small, the data are always poor, and the assessments are always ill defined. Do we need thousands of studies to tell us that we should “keep it simple, stupid?” People obviously can go to the gym and get stronger.

But what’s interesting to me about how Nuckols writes is that he doesn’t consider “science” to be some combination of biomolecular pathways and randomized trials. Instead, his scientific view is one of simple predictive models and their limitations. For an athlete like Nuckols, the scientific view helps systematize what matters in training. Working hard matters. Working harder over time matters. But resting and recovery are also critical to good performance. The right balance might not be easy to quantify, but the qualitative arc and predictions based on principles of adaptation are helpful. The general adaptation model for training tells us that training isn’t that complicated if you think carefully about what you are going to measure, and how you are going to track progress. This cybernetic view of biomedical science—one closer to what I would call engineering—feels like a promising path for many other aspects of medicine.

1. Maybe for kicks one day, I’ll write about the rest of my favorites.