HRV for cyclists: is it worth tracking, and how to use it

Heart rate variability is the most oversold number in endurance sport and, at the same time, one of the few genuinely useful ones. Both statements are defensible, and which one applies to you depends almost entirely on whether you look at single readings or at trends.

This piece covers what HRV actually measures, how to take a reading that means something, what the training trials show, and why the number your app showed you this morning is far less informative than it looks.

What HRV is

Your heart does not beat like a metronome. Even at a steady 60 beats per minute, the interval between consecutive beats varies — 980 milliseconds, then 1,020, then 995. That beat-to-beat variation is heart rate variability.

The variation comes from the autonomic nervous system, the involuntary control layer that runs your heart, gut and stress response. It has two branches: the sympathetic (mobilising, "fight or flight") and the parasympathetic (restorative, "rest and digest"), the latter acting mainly through the vagus nerve. Vagal activity modulates each beat rapidly, so when you're recovered and relaxed, the intervals vary more. Under accumulated stress — training, illness, poor sleep, a bad week at work — that modulation is blunted and the intervals become more uniform.

So, loosely: more variability tends to mean a more restored autonomic state. That is a real relationship, but it is a noisy one, and it is not a fitness measurement.

rMSSD, and why it's the one to use

There are dozens of HRV indices. The standards document that defined most of them dates from 1996 and remains the reference point. For athletes, one index dominates in practice: rMSSD — the root mean square of successive differences between beat intervals.

Its advantages are practical. It reflects vagal activity specifically, it is comparatively stable in short recordings of a minute or two, and it is less distorted by breathing rate than the frequency-domain measures. Most apps report lnRMSSD, the natural logarithm, because raw rMSSD is skewed and the log makes day-to-day comparisons more sensible.

Measurement hygiene is most of the battle

HRV is extremely sensitive to how you measure it. Inconsistent conditions will generate a number that varies more with your routine than with your physiology — and you will then make training decisions on noise.

Why one morning reading means almost nothing

This is the crux. Day-to-day HRV in a healthy, well-trained rider bounces around a great deal for reasons that have nothing to do with training readiness — sleep position, a late meal, how full your bladder is, the time you woke. Researchers in this area have argued for years that the rolling average, typically over seven days, carries the signal, and the individual day carries mostly noise.

A useful practical refinement is to track not just the average but its stability. A rolling average that stays roughly flat with modest scatter suggests you're coping with the load. An average that falls while the scatter widens has been associated with accumulating fatigue in monitoring studies of elite endurance athletes.

Does HRV-guided training actually work?

Several trials have tested the obvious application: let the morning reading decide whether today is hard or easy. Vesterinen and colleagues found HRV-guided prescription produced favourable endurance adaptations in recreational runners, and Javaloyes and colleagues reported improvements in trained cyclists compared with a traditional predefined plan.

The picture across the whole literature is more restrained. A 2021 systematic review with meta-analysis concluded that HRV-guided training was clearly better for maintaining and improving vagally-mediated HRV itself, but that its advantage for fitness and performance — if it exists at all — is small, with confidence intervals crossing zero. The reviewers were also blunt that methodology in this field is inconsistent enough to make pooling difficult.

That is a reasonable state of evidence for a monitoring tool. It suggests HRV-guided training is unlikely to hurt, may modestly help, and is nowhere near the transformative method it's sometimes marketed as.

What it's good for, and what it isn't

Where it fits alongside training load

HRV describes your autonomic state. Training-load models describe the work you've done. They answer different questions, and they're considerably more useful together than apart — a falling HRV average during a heavy block is a different situation from a falling average during a rest week.

Moveee tracks Fitness, Fatigue and Form from your actual rides, which gives you the load side of that picture, and the plan eases itself when a genuinely hard effort shows up. HRV, if you choose to track it, is the complementary reading: not how much you've done, but how well you're absorbing it.

And if you're tracking HRV mainly because you suspect you're overcooked, the more direct question is worth asking too — see overtraining vs overreaching.