Automatic Submaximal Fatigue Testing

A submaximal fatigue test is a steady effort, repeated regularly at the same power or pace. Because the effort is the same each time, changes in heart rate during the test and in how quickly your heart rate falls afterwards say something about your fitness and how fatigued you are, without the cost of a maximal test. Several top cycling teams use SFT tests.

Intervals.icu finds the test in your activity automatically, compares it with your previous tests and your wellness data, and tells you whether you look fitter, fatigued, or about the same, and why.

Configure the test parameters separately for each sport on the /settings page. This is a premium feature for supporters (thanks!).

Fix the FTP or threshold pace value and the target percentage for the sport at the start of the season. Do not change it as you improve.

Re-analyse existing activities to look for tests.

Doing the test

The effort. After your usual warmup, do 3 minutes steady at your test power or pace, then ease right off. By default that is 110% of FTP on the bike, or 105% of threshold pace running. You can change the target, the duration and how strict the detection is in the SFT section of your sport settings. Once you have picked a target, keep it the same: tests are only compared with others at about the same power or pace (within 5%).

Right after it. Stop or soft-pedal and keep as still as you can for a minute. How fast your heart rate drops in that minute (HR recovery) is part of the analysis, and it is only measured if you ease off.

Afterwards. Enter your RPE for the test (how hard it felt, 1 to 10) on the activity. RPE is what tells fatigue apart from fitness, and a test without one does not get a result.

Make the tests comparable. HR is easily changed by things other than fitness, so keep everything you can the same:

  • Same bike or trainer and power meter, and a chest strap rather than a wrist sensor for HR
  • Same warmup, same time of day, same pre-test meal and caffeine (or none), same room temperature and fan, same cadence
  • Same position in your training week, when fresh, so the test reflects your training block and not yesterday’s session
  • ERG mode is fine for this test: holding the power steady helps
  • Test about once a week

Getting started. The first result needs at least 3 previous tests from the last 90 days.

How to perform an SFT with power (cycling): https://www.sciencetosport.com/how-to-perform-a-sub-maximal-fatigue-test/

Firstbeat’s implementation: Firstbeat Sports Launches New Feature to Simplify Fitness Testing for Teams - Firstbeat

These tests are based on a more complicated protocol: The Lamberts and Lambert Submaximal Cycling Test (LSCT)

Activities with automatically detected sub-max fatigue tests are given the #SFT tag so you can search for them easily.

Hover over the test summary to highlight where it took place on the timeline. Click to show details and capture RPE.

What it looks at

  • HR for the same output: your average HR over the last minute of the test, shown as an efficiency factor (power or pace per heartbeat). Lower HR for the same power usually means fitter. But fatigue can also lower it, see below.
  • How fast your HR comes up: seconds for HR to reach its final level. Slower than usual is a classic sign of fatigue.
  • RPE: how hard the test felt. Harder at the same power points to fatigue. Easier points to fitness.
  • HR recovery: how far HR drops in the minute after. Faster with fitness, but also when acutely overloaded. Slower when fatigue has lasted.
  • Wellness: your sleep, stress, fatigue and soreness, morning resting HR and HRV. Context for borderline results.

Each one is compared with your own previous tests, not with population norms, and with how much your own results normally vary from test to test. A change counts as clear when it is about twice your normal test to test variation, and possible when it is bigger than that variation but not clearly so. Anything smaller is noise.

This framework was developed with help from CRC (LSCT literature review and analysis protocol), thanks!

Why lower HR is not always good news

This is the most important thing to understand about the test. When you are fit, the same power costs fewer heartbeats. But when you are acutely fatigued, for example in the middle of a training camp, your heart rate response is blunted: it rises slowly and less far, so the same power also gives a lower HR [1, 2]. On HR alone, the two look the same.

What separates them is how it feels and how your heart behaves:

  • Fitter: lower HR, feels the same or easier, HR comes up normally.
  • Fatigued: lower HR, but it feels harder and HR is slow to come up. That is not a reason to add load.

That is why your RPE and how fast your HR comes up are checked before anything else.

The results

  • Green: improved fitness. Continue or progress the plan. Confirm with a maximal test before changing your zones
  • Within noise: no real change. Continue, and judge the trend over several weeks
  • Amber: acute fatigue or functional overreaching, or a result that is unclear or not confirmed. 2-3 easy days or check the context, then retest within 48-72 h
  • Red: loss of fitness, illness or prolonged fatigue. Rest. If you have not recovered within about 2 weeks, review your training and health
  • Not enough data: not enough comparable tests yet, or the test has no RPE, no HR for its last minute, or is ignored. Keep testing and enter your RPE.

You can click the bars to navigate to different activities with tests.

Every result comes with the reasons behind it, each marked as good, neutral or bad, so you can see what drove it. It also includes the previous tests it was compared with, and your resting HR, HRV and sleep around the test.

Some things to know about how it decides:

  • Missing information never confirms a result. If your RPE could not be compared or your HR recovery was not measured, a worse test is “not confirmed” (amber) rather than red. Enter your RPE and keep still after the effort to get a clearer answer.
  • Feeling easier can confirm a possible improvement. A small drop in HR together with an easier RPE, and a normal HR response and recovery, is green. Feeling easier on its own is not enough.
  • Unclear results are decided by warning signs. When the test itself does not show a clear pattern, six warning signs are counted: RPE up, HR slow to come up, HR recovery changed, wellness worse than usual, morning resting HR up, and HRV low. 0-1 is within noise, 2-3 amber and 4 or more red.
  • A short-term drop is normal in hard training. A drop that recovers within about 2 weeks is functional overreaching and can come before an improvement. One that lasts longer is not, and illness should always be ruled out [3].

Wellness: what helps

Wellness data is not needed, but it helps a lot with borderline results. Research shows how you feel is often more sensitive to training stress than objective measures [4]. What is used:

  • Sleep: how long you slept over the 3 nights before the test compared with your usual. Each hour short is a point. Sleep scores from devices are not used, as they have not been validated, but sleep duration has [5].
  • Stress, fatigue and soreness on the day of the test, compared with your usual. Enter them in your wellness for the day.
  • Morning resting HR: up 5 bpm or more on 2 of the last 3 mornings.
  • HRV (rMSSD): your 7 day average compared with your usual range [6].

Only things that are worse than usual count. A long sleep does not cancel out a stressful day.

Limits

  • The thresholds are a proposal. The way results are decided follows a framework proposed from the research on the Lamberts Submaximal Cycle Test [7, 8], adapted to a test at a fixed power or pace. The cut-offs have not been validated.
  • HR is easily disturbed. Heat, dehydration, caffeine, poor sleep, stress and the start of an illness all change it. If a result is surprising, look for one of those first.
  • It does not replace a maximal test for setting training zones

The analysis protocol is very much under development. We will continue to work with the Intervals.icu community to improve it.

Further reading

  1. Hammes D, Skorski S, Schwindling S, et al. Can the Lamberts and Lambert Submaximal Cycle Test indicate fatigue and recovery in trained cyclists? Int J Sports Physiol Perform. 2016;11(3):328-336.
  2. Decroix L, Lamberts RP, Meeusen R. Can the Lamberts and Lambert Submaximal Cycle Test reflect overreaching in professional cyclists? Int J Sports Physiol Perform. 2018;13(1):23-28.
  3. Meeusen R, Duclos M, Foster C, et al. Prevention, diagnosis, and treatment of the overtraining syndrome: joint consensus statement of the ECSS and the ACSM. Med Sci Sports Exerc. 2013;45(1):186-205.
  4. Saw AE, Main LC, Gastin PB. Monitoring the athlete training response: subjective self-reported measures trump commonly used objective measures: a systematic review. Br J Sports Med. 2016;50(5):281-291.
  5. Chinoy ED, Cuellar JA, Huwa KE, et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep. 2021;44(5):zsaa291.
  6. Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Med. 2013;43(9):773-781.
  7. Lamberts RP, Swart J, Noakes TD, Lambert MI. A novel submaximal cycle test to monitor fatigue and predict cycling performance. Br J Sports Med. 2011;45(10):797-804.
  8. Capostagno B, Lambert MI, Lamberts RP. A systematic review of submaximal cycle tests to predict, monitor, and optimize cycling performance. Int J Sports Physiol Perform. 2016;11(6):707-714.

Other Notes

The zoom buttons zoom the timeline to the test and the optional HR recovery part.

The SFT analysis stores the target power/pace, the max CV & and tolerance % with the test. If the activity is re-analysed then these values and the existing type and duration are used, not what is currently in /settings for the sport. This preserves history.

If you want to re-do the analysis on an existing activity with the latest settings, use the activity list view:

27 Likes

Thanks David, that´s an very interesting feature. While I`m figuring out which parameters work best for me I noticed, that if I saved a threshold pace and want to delete this and save it blank, it requires me to enter any pace, but isn`t saveble without a number.

Tx. We have fixed this for the next deploy.

1 Like

Thanks very much for implementing this new feature and linking the background information!
It looks very interesting. Will have to read how i could use that for my running training.
I guess it will be released for cycling first and other sports may follow?

It’s there for each sports category. You can choose Power or Pace as target.

2 Likes

What should I be entering for Test FTP value? eFTP? Something else?

2 Likes

Thats exactly where I’m still very unsure. I will try with an empty field , which will result in using the individual FTP per Workout. Since I regularly adjustmy FTP value to the eFTP-suggestion it then corresponds to the current eFTP-value.

I will run some test at 110% FTP and will look how it works.

1 Like

You can enter your current FTP value, but it can also be something else in this ball park.
But keep it the same value for the season.

1 Like

You can AutoMate that process here.

What if I leave it blank? Would that be useful?

I guess I am not sure how this value impacts the test?

Does it matter how hard I set the value at the beginning of the season?

Like, should it be in the tempo- / threshold or vo2Max-Area? Because the heartrate recovery in these three Zones isnt’t equal, is it?

I think it’s best if it’s threshold or above. So that you still could estimate the time to exhaustion (probably difficult to estimate if you can hold the tempo power 90min , 2 hours or longer).

This is the starting value in absolute watts for the test. If left blank, the activity’s FTP value is used. However, if the FTP value changes over the course of the season, the test value would also change, which actually contradicts the whole point of the test.

For example, if you set it to 250W and 110% now, you’ll be doing a 3-minute test at 275W. And you can repeat this test as often as you like, even if your FTP might rise to 275W at some point during the season.

3 Likes

Thank you, appreciate the explanation!

We have added all the SFT fields to activity fields for the list view and activity panels:

And matching fields for /fitness so you can plot them with the subjective wellness fields:

The fitness fields use the first non-ignored test found on a given day.

We added the target range for power or pace with info on how to perform the test in /settings:

5 Likes

Hi David,
Are you using the Lamberts and Lambert Submaximal Cycling Test or a modified version?
The LSCT protocol uses 6m @ 60% maxHR, 6m @ 80% maxHR, 3m @ 90% maxHR and then 1m stationary (then take drop in HR).

The increase in power indicates better performance, and vice versa for the lower power.
Controlling how “very easy” is can render different results.

I would be better if the test was a defined protocol, like the actual LSCT protocol.

2 Likes

This is the simplified one ScienceToSport, Firstbeat (and at least 2 WT teams) use. It is very easy to add this test to the first 20 minutes of any “fresh” ride. It doesn’t require any particular warmup steps and can easily be done outdoors. Polar also have a very similar running test built into their watches.

6 Likes

We add a chart to the details dialog:

6 Likes

I have a standard running workout of 3 x 10’. I set this up to recognize an SFT successfully, but it grabs the first interval that meets the parameters. What if I want to use the 3rd interval? Thanks.

There is no way to do that currently. The test is supposed to be done in fresh state after some warmup and using the 3rd interval isn’t that.

2 Likes

Got it.