The weird science of recovery

You'll learn: What actually helps people recover from exercise and life, and whether recovery scores from activity trackers can help.

Feb 19, 2024

My favorite recovery practice is a long walk in the desert with Stockton.

Housekeeping

This post and its audio version in the bar above—like all Monday posts—is free to all subscribers of 2%. Enjoy.
If you want full access to Wednesday and Friday 2% posts and their audio versions, become a Member below.
Our promise to you: You’ll save more time and money in the long run than the cost.

This post summarized in four sentences:

Many new fitness trackers give users daily recovery scores, which attempt to quantify how ready the user is for the day ahead.
These recovery scores are useful for some groups, but research shows the scores often don’t correlate with physiological measurements of recovery (or how people feel).
The best measurement of recovery is how you feel—that’s because your brain is more sophisticated than a fitness tracker.
Sleeping and eating enough are the two practices science consistently shows help people recover.

Now onto the post …

Last Monday, we released a lengthy piece on new fitness trackers that are using made-up metrics like Strain to score everything that happens to you in a day—from your workouts to your everyday activity to your life stress.

These grand metrics have upsides and downsides downsides.

Read or listen to the piece here to learn more.

Today, we’re looking at a different made-up score from fitness trackers: recovery.

Recovery scores take many measurements and then score how ready you are for the day ahead. As one company put it, they “quantify how your body is feeling.”

Oura uses Readiness Score.
WHOOP uses Recovery score.
Garmin uses Training Readiness score.
Fitbit uses Stress Management and Daily Readiness scores.

You get the point.

Today, we’ll cover the following:

What recovery is.
How fitness tracker recovery scores are calculated.
Whether recovery scores are accurate.
Some downsides of recovery scores.
The best way to “measure” recovery.
How to recover.

Let’s roll …

What is recovery?

In short

Recovery is basically your ability to do one thing after doing another thing.

The details

Definitions matter. What do we mean when we say “recovery” in the context of exercise?

I opened up an exercise textbook to see how its authors defined recovery. And … the word isn’t in the book. The closest they got was:

A section on avoiding destructive behaviors.
A section on sleep.
A section on stress.

Otherwise, nothing.

I read a bunch of studies. Many used the word but didn’t define it.

The best definition I found came from WHOOP. They said, “Recovery reflects how well prepared your body is to take on (physical stress/exercise), and is a measure of your body's ‘return to baseline’ after a stressor.”

It’s basically your ability to do something at the skill level you want after doing another thing.

It happens within workouts. For example:

Let’s say you lift a weight 10 times and then rest for one minute.
Then you repeat the lift but can only lift it 8 times
This means you haven’t fully recovered.

It also happens between workouts. For example:

Let’s say you run a 7-minute mile on Wednesday.
Then you run a 7-minute mile again the next day.
Congratulations, you were recovered.

Recovery scores from trackers measure your recovery across the days and between workouts.

The point of exercise is to incite positive changes in your body. That’s what improves your strength, cardio, etc.

But you don’t improve when you actually do the workout. You improve afterward as your body rebuilds and increases your fitness in preparation for the next round of exercise.

How are fitness tracker recovery scores calculated?

In short

Recovery scores are calculated through physiological measurements the tracker takes—like heart rate variability, resting heart rate, etc—and assumptions about your behavior (for example, less sleep or more exercise usually leads to a lower recovery score).

The details

Wearable companies factor in as many as 20 measurements to determine your recovery score.

For example, WHOOP’s website states:

WHOOP calculates your recovery on a scale of 0 to 100% during your sleep, looking at your heart rate variability (HRV), resting heart rate, respiratory rate, SpO2, sleep performance, and skin temperature to see how your body is adapting to physiological and psychological stress. The biggest influence is by far your HRV but it also considers your health, behaviors, stress levels, hydration, and more.

Oura says the rings:

(T)akes into consideration over 20 different body signals — including temperature, heart rate, HRV, and sleep — to tell you how ready you are for the day ahead.

All those data points flow into proprietary algorithms invented by the companies.

When you wake up, you get a recovery score.

For example, if you have a low recovery score, you could have a lower-than-usual HRV or higher-than-usual resting heart rate. Or your sleep could have been shorter than usual. Or your skin temperature could be higher. Or … (Insert all sorts of factors).

If you slept 7 hours instead of your usual 8, the band might tell you you’re only 45% recovered.
If your sleep was normal but HRV was lower, you also might have a lower score.

Recall that these bands also give you a daily exercise goal in the form of Strain (WHOOP), Active Calorie Burn (Oura), etc.

Your recovery score then changes your daily exercise goal.

In the case of WHOOP, for example:

If your recovery score was 45%, your Strain goal might be 8.
If your recovery score was 90%, your Strain goal might be 16.

The idea is pairing the recovery score and exercise goal pushes you to exercise harder when your body is most ready to take on exercise and adapt. That theoretically allows you to improve more.

Are recovery scores accurate?

In short

Recovery scores are best at picking up common sense behaviors and don’t always jive with reality.

But they can help people, especially those new to wellness, identify and fix bad behaviors.

They can be potentially negative for Type-A people who already practice healthy behaviors and understand how their body works.

The details

Recall the expert Marco Altini from last Monday’s post. He has a PhD in applied machine learning, a master’s in computer science engineering, and a master’s in human movement science and high-performance coaching—the ideal mix to understand wearables.

He explained that the raw physiological measurements from wearables—like HRV and RHR—can be useful.

But recovery scores bundle in behaviors, not just raw physiological measurements (like heart rate or HRV).

So, for example, your recovery score would likely be lower if you slept less the night before or exercised harder the day before.

But your body might actually be fine if we assume raw HRV is a more important gauge of your body’s ability to take on and adapt to stress.

Here’s Marco:

I think wearables are useful to capture the raw physiological data, but make them a disservice when trying to interpret it providing readiness or recovery scores. As athletes or coaches, you really want to see how the body responds to (training and other) stressors, and mudding the waters mixing behavior (e.g. sleep and activity) with physiology (e.g. heart rate and HRV) makes it so that the data reflects assumptions made by a generic algorithm (e.g. less sleep or more activity requires more recovery) as opposed to what the athlete's physiology actually showed (e.g. a good response to an increased training volume).

Marco pointed me to a study where researchers asked a bunch of swimmers how stressed they were with a questionnaire.

The scientists then compared the swimmers’ responses to raw HRV data from WHOOP bands and Recovery scores from the bands.

They discovered that the swimmers who reported more stress had lower HRVs and higher resting heart rates. That’s to be expected and a great capturing of raw data from WHOOP.

Then the researchers compared the data to WHOOP’s Recovery Score. They found self-reported stress and resting heart rate correlated with HRV as you would expect.

But here’s Marco writing about what they found about Recover Score.

There was zero correlation between the Recovery Score and all other variables. This is quite something, I am not sure I’d be able to design a score that ends up being as useless as this one, even if I wanted to.

Please do yourself a favor: if you want to use a wearable, at least use it to look at the physiology and ignore made-up metrics (recovery, readiness, etc.).

Here’s the data:

If you’re a serious athlete, Marco recommends avoiding recovery scores and, if possible, taking raw HRV data from your device at the same time each morning (that way you get consistent data). Note: Not all trackers allow you to do this, but WHOOP does.

You can do this with the following devices:

Phone camera (iPhones and supported Androids).
Scosche Rhythm24 in HRV mode.
Coros Apex Pro 2 (or other Coros with ECG).
Apple Watch using the Breathe app.
Whoop in broadcasting mode.
Kyto HRM-2931 or the following versions.
Polar, Garmin chest straps.

Here’s a deep dive from Marco on how to do this and make use of the data.

This doesn’t mean recovery scores are totally useless for the average person, especially someone just starting a health journey.

Here’s Marco again:

Can it work for the average person? I think so, despite my criticisms. Most people do not have a plan, and poor sleep or more activity than they normally do will require some recovery, so the assumptions of the model are not bad, but this is not why you normally wear a sensor. You wear it to learn about your body's response, and the body's response is in the actual data (HR, HRV), not in the recovery score.

The band’s assumptions, for example, might pick up bad sleep after nights where you drink or watch too much Netflix. The poor recovery score the next morning might compel you to make different decisions.

That example seems like common sense. And it is. But some people need the bad recovery score number staring them down in order to make a change.

I asked the wise Dr. Trevor Kashey what he tells clients who ask him about the utility of WHOOP and other trackers. He said:

I say ‘if you want to pay WHOOP to tell you to quit Netflix and go to bed on time, go for it.’

Downsides of recovery scores

Some people get obsessive about their recovery and sleep tracker scores. Both The Wall Street Journal and The Atlantic recently ran stories on this. And there are plenty of Reddit threads where people discuss how their wearables stress them out.

I’ve also heard this from many friends and readers. These people tend to be fit and interested in health and wellness. They’re doing all the right things but the tracker gives them seemingly random scores—and that stresses them out.

Call it “wearable anxiety.”

Applied to sleep, it’s a condition called Orthosomnia. One paper in Nature and Science of Sleep defined it as:

The obsessive pursuit of optimal sleep metrics based on fitness tracker or mobile phone app data is termed “orthosomnia”, coined by combining “ortho” referring to straight, right or proper, and “somnia” referring to sleep.

I wrote about this in my recent book Scarcity Brain.

Trackers with made-up scores lean on The Scarcity Loop—the potent combination of opportunity, unpredictable rewards, and repeatability—that is unparalleled at getting people hooked. Read more here.

Marco explained:

I think these sensors often promote a reactive approach that is not ideal, maybe for the sake of engagement, while metrics are really more useful when we already have a clear plan, and make (very minor) adjustments to it based on the data, on rare occasions. The data should not be used to make up a plan as you go, without any context. Start with a plan, then use the data (actual physiological data, not made up scores) to make small adjustments. This is what research has shown to actually be effective in e.g. improving performance.

What’s the best way to measure recovery?

In short

HRV shows lots of promise, but how you feel seems to be most practical.

The details

There’s a reason most trackers give a lot of weight to HRV. Some strong research suggests HRV gives us the best insight into our body’s ability to take on and adapt to exercise stress.

HRV is a measurement of the amount of variability between your heart beats. Higher variability correlates strongly with your body’s ability to take on stress and adapt. For example:

Scientists in Finland discovered that people who did HIIT when their HRV was high experienced greater fitness gains.
A Spanish study found that the performance of cyclists who tailored their training to their HRV improved by up to 14% more than that of others who relied on more traditional methods.

The catch is that you need to measure HRV correctly, as we laid out above. And HRV is definitely not an end-all-be-all.

But here’s the good news: If you’re not getting paid to exercise or play a sport, there’s probably a simpler and maybe even more accurate way to figure out your recovery.

Christie Aschwanden is one of the best science journalists in the game. She’s won prestigious awards for her science writing and is a deep thinker about communicating science.

She also happens to be a great athlete. She was on the Team Rossignol Nordic ski racing squad, where she raced in the US and Europe.

In 2019, she released a book about recovery called Good to Go: What the athlete in all of us can learn from the strange science of recovery.

To report that book, she conducted hundreds of interviews and read hundreds of papers on recovery. She investigated different recovery methods—from massage to compression pants to ice baths to supplements and on and on.

About recovery scores and the best way to gauge recovery, she told The Verge this:

I was surprised by the the extent to which there are so many efforts to quantify recovery. All these companies trying to find “the magic metric,” that thing you can measure on a watch or blood test and so you’ll know whether you’ve optimized your recovery. But the very best thing we have is this subjective measure of “how do we feel?”

The subjective measures trump the objective measures every time in multiple studies and analyses. It turns out our brains are the most sophisticated instruments here in ascertaining recovery and that psychological feeling of wellness and readiness is the very best measure of recovery.

What actually helps us recover

In short

Sleeping and eating enough.

The details

After reporting the book, Aschwanden basically came down to this conclusion, as summarized by reviewers at Columbia University:

Other than getting enough sleep and eating enough calories, nobody knows for sure what helps athletes recover between events or training sessions better than just living a normal life.

That said, the research on recovery is new and imperfect. If you have a special recovery practice that you strongly believe works for you—do it.

Have fun, don’t die, sleep and eat enough, and live a normal life.

-Michael

The weird science of recovery

You'll learn: What actually helps people recover from exercise and life, and whether recovery scores from activity trackers can help.

Housekeeping

This post summarized in four sentences:

Now onto the post …

What is recovery?

In short

The details

How are fitness tracker recovery scores calculated?

In short

The details

Are recovery scores accurate?

In short

The details

Downsides of recovery scores

What’s the best way to measure recovery?

In short

The details

What actually helps us recover

In short

The details

Sponsored by GORUCK​

Sponsored by Maui Nui Venison

Sponsored by Momentous​

Discussion about this post

Sponsored by GORUCK

Sponsored by Momentous