Aerobic Threshold estimation using HRV based method DFA-⍺1

Polarized training has drawn a lot of attention lately, and there’s quite a bit of interesting publications on the subject, particularly from Dr Stephen Seiler.

A quick Google search will give you something to read. Here’s a long and detailed post about it (Complete Polaraized Training for Cyclist Guide)

Even TrainerRoad, that has been long behind the Sweet Spot approach has finally started to consider it.

Polarized training has 3 fundamental zones, delimited by 2 thresholds: the higher threshold (between Z2 and Z3) is fairly easy to determine with classic FTP tests, but the lower one, the Aerobic Threshold (AeT) has typically been harder to pin down.

The idea is to ride a certain amount of high intensity (HIT) or Z3 and the rest in Z1 (LIT), religiously avoiding Z2 (also known as Sweet Spot, hence the archenemy of Polarized Training)

This is where this HRV comes to play. This metric is a good way to determine the level of recovery, but it can be useful also during active training.

There are obviously other ways to determine AeT, but mostly require a visit to a sport lab, or poking your earlobes to draw some blood, so for now we’ll try with a cell phone and a HR strap.

What is DFA-⍺1

Ok, I was already confused by HRV, what is this DFA-alpha1 anyway…?

Not going to try and explain, save my and your time and head over to this explainer.

For our practical purposes here, it is a non linear method of processing RR samples (the intervals between each heartbeat) that is correlated to where our AeT is at.

The important bit is that for this to work, pretty much every single heart beat needs to be accounted for. If you miss one or you have an ectopic beats you have so called artifacts, which basically significantly affect measurement.

A testing protcol is necessary in order to standardize results and being able to compare and measure progress.


I use HRV logger from the excellent Marco Altini, available for iOS devices and Android.

The recommended HR strap is Polar H10. Other straps may work too, but the polar seem to have the most accurate and less noisy measurement around. Anecdotal reports show ANT+ being not as robust as BLE. Strongly recommend to pair via BLE (on Android devices at least, not an option on iOS anyway)

I recommend wearing the strap as tight as possible without getting uncomfortable,, wetting the contacts and your chest, quite a bit. You’ll start sweating if it’s not too cold, but it’s a good idea to improve contact.

There are HR monitor gels that are pretty cheap and do a good job of improving contact if sweat is not happening.

The test protocol

The protocol consist of a 15min warm up at low intensity, followed by a gradual ramp from 60% to 90% FTP and then back down. The resistance is increased by 2% every minute, so it will take 15min to go up and the same 15min to come back down.

The reason for the double sloped ramp is to cross the Aerobic Threshold from both directions, going up or down.

AeT Testing Protocol

Make sure that strap is snug and the contacts are wet. There are solutions avaiable, but just making sure the strap is wet generally works well. Sweat will do the rest.

It is also important to minimize movement and avoid changing position during the ramps. Find something that feels comfortable (tops, hoods) and stick with it.

At the end of the 15min warmup, start recording on the HRV Logger.

Ideally you’ll want to synchronize the start on the minute, so that the averaging will be done for 1min at constant power.

In the HRV logger Settings select 60 sec for the computation window and “Workout” for the RR-intervals correction.

Once the ramp is completed (up and down) you can stop the logger and export the csv file to Dropbox. You can use your favorite editor for post processing data: I wrote a Python script that generates a plot with the results on it.

AeT extrapolation

Since we are talking about cycling, figuring out where this thresholds is in terms of power can be pretty useful: luckily it is easy to check that on, in my case is at 214W, which happens to be at around 72% of my FTP.

AeT in Watts

Once we have this number down, we can use it to guide most of the volume training to be at or around this threshold.

Another useful representation of the same data for the ramp up and down:


If you’re interested in checking Polarized Training, it’s important to get the thresholds right.

In particular the Z1/Z2 threshold is critical. A simple protocol based on a standard ramp test and using HRV is described to determine AeT.

Happy Training!


A new app called FatMaxxer, currently only for Android has been published to estimate AeT based on this method.

You can find it in the Play Store here.

2 thoughts on “Aerobic Threshold estimation using HRV based method DFA-⍺1

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s