Most of my writing revolves around my belief that human health and performance is human-centred. This might strike you as being obvious. However, it’s not.
“What are you talking about, Shane? I choose my health and performance goals!”. “I entered that marathon”. “I joined that gym”.
I hear you. However, as Rafiki once advised a juvenile Simba, look harder. People everywhere are telling you what you should do. More importantly what to focus on, where to spend your money, and where to spend your time. Luckily, you subscribe to The Active Edge, and we’re passionate about helping you navigate that swamp of bullshit.
We’re not alone. In 2022, researchers highlighted that “there is a disconnect between the complexity underpinning human health and performance”. So when those researchers found that health and performance subjective monitoring measures, like our perception of effort, outperformed objective measures, like % of VO2 max or heart rate, naturally, I gravitated towards my biases (*ahem, I mean passion). Thus, that led to writing one of my most-read articles “The Data I See In The Context Of Me”.
Of course, I recommend reading it, but to briefly summarise and set the scene for my next few paragraphs, the conclusion was that AI and objective data monitoring still isn’t as useful as our innate human ability to rapidly integrate insane amounts of personal and environmental information. Or to take an excerpt: Our fancy Apps don’t account for how incredible human beings can be.
And thus fast-forward to the present day, reading a recent research paper on using machine learning to predict recovery in endurance athletes. I should make it clear that I am all for utilising AI and big data into health decisions and coaching processes. Moreover, AI should, and will replace lazy plan-only orientated people who call themselves coaches. Apps with financial backing, such as Runna and Centr, will eventually improve and offer cost-effective options for people wanting to work towards health and performance goals. However, AI won’t replace true human-centred care for people, which is what actual coaching is.
So this was actually a really interesting study aiming to see if machine learning could predict how recovered we feel in the morning - subjective - and changes in Heart-Rate Variability (HRV) - objective. Don’t worry, you can keep reading - this isn’t a science review. We’ll skip right to the takeaways.
Firstly, a complex machine learning model predicted perceived recovery well, when interpreting all of the data (looking at everyone as a collective). Furthermore, a simpler model (using only soreness, sleep quality, perceived life stress, and recovery scores from the previous two days), stacked up just as well as the complex one. The message? We don’t need to use loads of complex variables and data to understand our readiness to train!
However, these models didn’t do so well when used at the individual level. Interesting… So each individual had their data analysed and they each had different variables best predicting their recovery, and thus their readiness to train.
The conclusion? THE DATA I SEE IN THE CONTEXT OF ME: VOL 2!
A coaching point I cover regularly was highlighted as a key takeaway in this paper: “start by monitoring a wide array of variables and reduce the number based on feedback from different models”. Of course, they refer to feedback from machine learning models, but unless you’re an expert coder, perhaps try using training perception, long-term adaptation, and overall enjoyment before you sign up to a course at MIT.
YOU still matter. How YOU feel is still important. Just because Disco Dave down the road says HRV is the best recovery metric, doesn’t mean it will be for you. Yet it could.
It’s not rocket science. Athletes and coaches have been refining this since the inception of the modern olympics. I’m lucky enough to have been training for 18 years. Mostly, my key variables that predict my physical readiness to train: sleep index (how long I slept combined with how fresh I feel in the morning) and HRV. Ronny’s is muscle soreness and resting heart rate.
A lot of what we get told comes from the limit of knowledge of who’s telling you. People mostly tell you based on their own experience. And, whilst they mean well, now we’re in a knowledge dominated service industry, it’s not good enough to advise based on experience alone.
My key takeaway (apart from Dominos) I hope you get from this newsletter is that we should still listen to our bodies. I hope it helps you better navigate the swamp of bullshit. It’s down to us, and our support network, to work out what variables are reliable to predict our readiness to train. Readiness to train is important for adaptation (future email), and adaptation to training is important for our fitness goals, and ultimately, our health and performance.
Safe reading.
Shane.
PS - Next week, Nicole writes about arousal regulation. If you can’t wait until then, and you’re looking for performance insights right now, you can head over to our substack here to read some of our previous editions.
If you found this of value, then we value your support - Buy us a coffee: https://ko-fi.com/activeedge
Website, Videos and more: https://linktr.ee/theactiveedge
Email - info@activeedgesportstherapy.co.uk
Active Edge Recommendations
Integrative Proposals of Sports Monitoring: Subjective Outperforms Objective Monitoring. This still remains one of my favourite academic research papers. It’s also not written in overly complicated sciency language, the authors aren’t trying to show how clever they are, and you get a genuine feel that they care about improving research for human health and performance!
Predicting daily recovery during long-term endurance training using machine learning analysis. This is more sciency. And, not the easiest to digest. Amazing research by some proper clever cloggses. But, if you want to dissect more of the paper, read Alex Hutchinson’s article on Outside. I won’t lie, I had to use some of Alex’s piece to better understand the paper!