approaching a million views on medium (🔗: medium.com/@altini_marco )

in this thread, an overview of my main blogs by topic:

• basics of heart rate variability (HRV)
• more advanced topics related to HRV
• wearables
• training
• biofeedback

Thank you for reading

/1 BASICS OF HRV

• ultimate guide to HRV (4 parts): medium.com/@marco_alt/the…
• heart rate vs HRV, what's the difference (5 parts): medium.com/@altini_marco/…
• 3 steps to make use of HRV: medium.com/p/f23e7f1c743
• what's behind changes in HRV: medium.com/p/ec3c02353373

/2

when looking at heart rate variability (HRV) we are analyzing the body's response to stressors

a common misconception is that hard training should cause a dip in HRV the following day

well, that's the case only if something went wrong

research and a practical example below

🧵 it is certainly true that hard training will cause a disruption in ANS activity as captured by HRV

however, this is the immediate, acute response, which will fade in time

clearly shown in @StephenSeiler's paper (pubmed.ncbi.nlm.nih.gov/17762370/): see ↓ HRV for hard training

/2

thank you Michael, glad you found these features useful

here are two papers I wrote about them:

• estimating VO2max: researchgate.net/publication/31…
• estimating performance (10 km time, what we call lactate threshold in HRV4Training): researchgate.net/publication/32…

brief summary below

/1

https://twitter.com/Nittaz/status/1572408006448664576

simply put, estimating VO2max relies on how external and internal load change across and within individuals

if you are more fit, your heart rate (internal load) at a given pace (external load) will tend to be lower, when analyzed in the long term

/2

simple workaround in case you would like to use your Oura ring to measure your HRV in the morning (and not only the night), and then interpret it with respect to your normal range in HRV4Training

/1 in the morning, from the Oura app, tap the plus icon, and start an Unguided session

/2

I've enjoyed reading "Individualized Endurance Training Based on Recovery and Training Status in Recreational Runners", by @OPNuuttila and co-authors

in this study, there were a few interesting changes with respect to previous work on HRV-guided training

👇

/1 HRV-guided training simply means that training is adjusted based on our physiological response

the typical benefits of this approach are due to how not only the type of stressor but also the timing of it, are key to improving long-term performance

medium.com/@altini_marco/…

/2

since I'm at it today... here is a poor choice used by Garmin to report resting heart rate (HR):

"HR is calculated using the lowest 30-minute average in a 24-hour period"

a simple example where acute sickness was obvious when measured by any device but Garmin HR

👇

/1 woke up fine in the morning, went for a workout, I was doing still okay in the afternoon

mid-afternoon, I started feeling sick, by 8pm I have a 38.7 C fever

around midnight I'm better, during the night the fever goes away and in the morning I have no fever, just a bit tired

/2

a touch of food poisoning to keep things interesting when it seems to go too well

life

see how for sickness, HR is typically a sensitive enough signal, as we reported in our paper: mdpi.com/1424-8220/21/2…

sitting data can capture this better than lying down

day off :)

https://twitter.com/altini_marco/status/1564214782844182528

in these situations (with a clear stressor) you can see how an orthostatic measurement (e.g. going from lying down to sitting or standing, then measuring) is a lot more sensitive:

• night HR: 55bpm
• morning lying down measurement: 55bpm
• sitting: 68bpm
• standing: 91bpm

Some tools I built to track changes in cardiorespiratory fitness and performance for runners

these can be, in my view, more effective than the "training load analysis"

• intervals analysis
• lactate threshold estimation (LT2, or critical speed)
• aerobic efficiency

1/6

https://twitter.com/bramwillemse/status/1558378744871587840

1) intervals analysis

the best predictor of performance is...performance

interval analysis in HRV4Training Pro uses Strava Laps (normally automatically logged by your watch) to summarize your sessions and give you an overview of performance changes

hrv4training.com/blog/interval-…

2/6

hot day in Amsterdam, so I cut sleep shorter, woke up early, and went out for my hard session before it gets too hot

wearables are freaking out telling me to "concentrate on recovery", or that I have "just enough to get by" (so much yellow)

what about my physiology? 👇

/1 as you can see in HRV4Training.com, HRV is at the high end of my normal range, daily and for the baseline:

today is a great day for a workout

we have two issues here:
1) wearables pretending to know your sleep quality
2) mixing sleep data with HRV to provide advice

/2

in the past year, I've measured HRV in the morning while lying down and sitting (added the latter thanks to great input from @andrew_flatt)

mostly due to a large change in my resting heart rate (see pic) I've finally come across parasympathetic saturation when lying down

🧵

/1 what's parasympathetic saturation?

a situation in which your parasympathetic activity is high, but HRV does not reflect it well

despite increased acetylcholine (the main neurotransmitter of the parasympathetic system), heart rhythm does not change

hrv4training.com/blog/parasympa…

/2

some thoughts below in terms of:

• activity tracking
• sleep tracking
• made up metrics vs actual physiology
• form factor
• do you even need one

hope this helps!

https://twitter.com/bloombergtrades/status/1553276050410393600

ACTIVITY:

in case you have any interest in tracking your activities (e.g. running, cycling, walking, other), then Garmin, ideally paired with a strap, is the way to go

all other wearables often produce inaccurate data and have no GPS, nor do they let you pair a strap

/1

when providing daily advice (color-coding and message) in HRV4Training we combine your physiology and your subjective feel (outputs)

not your behavior (input)

this differs from what you get in terms of readiness or recovery scores in most wearables

why is that?

/1 the whole point of assessing your state, either objectively via heart rate variability (HRV) or subjectively by feel, is to determine how you responded to your given circumstances

you already know the input (behavior) and are assessing the output (physiology or feel)

/2

getting the same race time estimates in HRV4Training Pro and Stryd

Stryd uses Critical Power, while our model, even though I never thought of it this way, estimates Critical Speed (we call it lactate threshold, LT2)

both use workouts data (no specific tests)

I love consistency here is our paper describing how we derive our estimate from previous workouts data

you can link HRV4Training to Strava to get these estimates in HRV4Training Pro (hrv4t.com)

researchgate.net/publication/32…

To make sense of changes in physiology (in particular, HRV and heart rate), we need to interpret them with respect to what we call your "normal range"

In the scientific literature, this is called the smallest worthwhile change (or SWC)

How we define this range, is key 🧵

/1 if you are new to the concept of the normal range or SWC, please check out this blog post first: hrv4training.com/blog/the-impor…

here I'd like to focus on a single and often overlooked aspect:

how much data should we include to determine the normal range?

/2

a common misconception is that HRV should dictate how you can perform

this is not the case

for one, delayed onset muscle soreness (DOMS) and nervous system recovery are on a different schedule

what does it mean when your HRV is still within normal after a hard workout?

/1 this can be confusing at times, but there is nothing better than a good (within normal range) HRV after a hard session

why?

a good HRV after a hard session shows that you were able to quickly bounce back

this is a sign of good fitness and an adequate training stimulus

/2

I often talk about your "normal values" as the only meaningful way to make use of Heart Rate Variability (HRV) data

But what are these normal values?

Why are they important?

A short thread with some examples

👇

1/ Simply put, your normal values are a representation of your historical data

They allow you to understand if acute (daily) and chronic (weekly) HRV responses are showing meaningful changes or just small variations that you should not worry about

They make the data actionable

/2

some data and thoughts about pre and post-exercise autonomic control

mostly stating the obvious, but it helps to fine-tune the protocol and to gather feedback, bear with me

• rationale
• tools
• heart rate and rMSSD
• alpha 1
• protocol

Go 🚀

👇

cc @StephenSeiler

/1 RATIONALE

based on Stephen Seiler's reasearch, using pre and post-exercise HRV we can better understand the impact of different workouts

measuring *immediately* before and after, we isolate the training stressor in a way that allows us to answer different questions

/2

a great tool to play with this is exphyslab.com/lactate by @felipe_mattioni

below is some data I had collected during the "pandemic treadmill" days

many methods to derive the aerobic threshold (e.g. OBLA 2.0 or LTP1) are clustered around the same speed (12.5 here)

/1

https://twitter.com/danielgross/status/1482820023953477635

which gives me a heart rate of 150 bpm, exactly what I've considered my aerobic threshold since starting with polarized training ~5 years ago (medium.com/p/e604b2f0a910)

This is "50+ km" intensity (easy!)

note that using DFA alpha 1 gives me a much lower HR rate, 135-140 bpm

/2

.. and as expected, back to normal despite training almost five hours yesterday (at very easy intensity (!): strava.com/activities/652…, 66% of max heart rate)

some thoughts about acute HRV drops: these are somewhat more difficult to interpret

why?

my two cents below

/1

https://twitter.com/altini_marco/status/1481867890861162496

we have two types of HRV suppressions:

• acute: single day dips
• chronic: a couple of days, or baseline drops

chronic drops highlight clear stressors, which explains why research on HRV-guided training uses baseline changes outside of your normal range for prescription

/2

In the past weeks, I had the opportunity to talk about heart rate variability (HRV) and HRV4Training with the authors of two of my favorite running books:

@mattfitwriter (80/20 Running) and @jasonkoop (Training Essentials for Ultrarunning)

Episodes: marcoaltini.com

/1 Matt's book, covering @StephenSeiler's research, made running a lot more fun for me (and a bit faster too, see my writeup here: medium.com/@altini_marco/…)

I've read plenty more from Matt, and I'm looking forward to his next one with @BenRosario1 (here: amazon.com/dp/B096DH95W8/…)

/2

Last paper in our special issue, great work by @OPNuuttila (here: mdpi.com/1424-8220/22/1…)

"Validity of the Wrist-Worn Polar Vantage V2 to Measure Heart Rate and HRV at Rest"

The data looks good for both heart rate and rMSSD

Short thread about night HRV data below 🧵

/1 Regardless of the accuracy of a sensor, there is a second key aspect to consider:

Is the data sampled meaningfully?

This is for example where others, that are also accurate, fail in providing useful data (e.g. the Apple Watch: medium.com/@altini_marco/…)

What about Polar?

/2