Using Heart Rate Variability to Guide Training

This article was made possible through the generous support of our voluntary subscribers.  If you value coverage like this, please support FasterSkier with a voluntary subscription.  

In my coaching and consulting work, I get a lot of questions about using heart rate variability (HRV) to guide training. This is a follow-up to a previous article, where I want to move from theory to practice and show how HRV has been one of the most valuable tools I’ve used to assess recovery and adaptation in endurance athletes.

There is no shortage of research documenting the value of HRV in endurance training. That’s well established. What matters to me is how it works in coaching practice. In my experience, HRV is especially useful for helping athletes avoid the most damaging downsides of poorly implemented training. Most of the big performance issues I have seen over the years could have been avoided with the use of HRV.

Over the years, I’ve collected long-term HRV data on many athletes, some spanning close to ten years. The example I’m using here covers nearly four years and serves as an excellent illustration of how HRV reflects training implementation, adaptation, and the consequences of training changes over time. This athlete was 50 years old at the start of the data set, but I’ve seen the same pattern in juniors, elites, and masters’ athletes alike. The trend is the same; only the absolute HRV values differ.

Most of the athletes I’ve worked with over the past 15–20 years arrived with either inconsistent training or worse, chronic over-training. In this example, training lacked consistency and relied too heavily on endurance work performed at intensities that were simply too high. To be clear, many of the long-term problems I’ve seen, performance stagnation, recurring illness, persistent fatigue, would have been obvious much earlier had HRV been used properly.

The chart below shows an overview of the data, collected using a Firstbeat sensor and analyzed through the Firstbeat Sports platform. With improved training structure and better implementation, this athlete’s RMSSD baseline steadily increased:

• 2022: ~45 ms
• 2023: ~55 ms
• 2024: ~65 ms
• 2025: ~75 ms

 

For anyone familiar with HRV, these are substantial improvements. From a health and wellness perspective, they’re meaningful. From a performance perspective, they’re even more telling. Population data suggests that an RMSSD of around 75 ms is very good for someone in this age group. Performance metrics tracked alongside this, ski erg tests and six-minute roller ski tests, showed parallel and consistent improvements over the same period.

Some people will point to what looks like large day-to-day variation in the data. Yes, the coefficient of variation can appear wide at first glance. In athletes with low or suppressed HRV, I often see variability in the 20–30% range. As fitness and autonomic stability improve, that typically narrows to the 10–20% range. So, while the spread may look dramatic, it’s often exactly what we expect. Larger variability usually reflects poor training implementation or an athlete drifting toward an overtrained state.

I’ve seen this same pattern repeatedly in underperforming and overtrained athletes, at both junior and elite levels. HRV is often much lower than expected. When training and recovery are corrected, athletes frequently feel better before their HRV confirms they are ready to resume normal training. This is a critical point. Returning too early, before HRV plateaus, is a common mistake. When training is reintroduced with proper control, HRV not only stabilizes but often continues to rise beyond previous levels.

Population-level HRV data can provide useful context. There will always be individual differences, but these ranges help determine whether an athlete is operating in a reasonable zone for their age. In practice, well-trained endurance athletes often sit above the “high” end of population norms. I’ve coached athletes in their 20s with RMSSD values consistently in the 110–120 ms range. The same 50-year-old athlete referenced here now sits closer to 75–80 ms.

Normal Resting RMSSD Ranges by Age

(Morning, seated or standing, ECG-based measurement)

Age	Low / Suppressed	Typical / Normal	High / Very Good
<20	<50 ms	60–90 ms	>90 ms
20–29	<40 ms	50–80 ms	>80 ms
30–39	<35 ms	45–75 ms	>75 ms
40–49	<30 ms	40–65 ms	>65 ms
50–59	<25 ms	35–60 ms	>60 ms
60–69	<20 ms	30–50 ms	>50 ms
70+	<15 ms	25–45 ms	>45 ms

To make HRV actionable, the data must be analyzed correctly. I typically use a seven-day acute average of RMSSD and establish a normal range using a standard deviation of ±0.5 over a two- to three-month period. When training is well implemented, we expect to see a gradual upward drift in HRV. Eventually, HRV will reach a physiological ceiling, and further increases won’t occur unless the athlete is recovering from illness, injury, or prior overtraining.

In a healthy training state, the acute HRV average should remain within that normal range. If it drops below, the first step is always context: life stress, sleep, travel, illness, or recent training load. Sometimes it’s a conversation. Other times, it’s an immediate adjustment to the plan.

Daily fluctuations are normal. Hard sessions or unusually long workouts often cause short-term drops in HRV, and that’s not a problem provided values don’t fall well below baseline. What matters is the seven-day trend and whether that trend begins drifting downward.

Chronic suppression tells us several things:

• the autonomic nervous system is not recovering
• training stress exceeds adaptive capacity
• intensity distribution is likely wrong, even if volume looks reasonable

This is where HRV is most powerful. It catches problems early, before performance falls apart. Large single-day drops can also flag oncoming illness or signal the need to review the previous few days of training.

HRV often exposes poor intensity distribution long before performance declines. Athletes can hold things together for surprisingly long periods while autonomic stress accumulates underneath. HRV shows that the effects  early, particularly in athletes doing too much work in the higher intensity zones.

This is especially relevant for junior and masters’ athletes, where recovery margins are smaller and the consequences of mistakes are greater.

I will not increase training load, weekly or monthly unless the acute HRV is at baseline or higher (the midpoint of the error bars). This is one of the most common issues I see when consulting with athletes and coaches: loads are increased according to the plan, not according to how the athlete is recovering and adapting.

Both experience and research are clear here. Increasing load or adding high-intensity work when HRV is suppressed does not lead to adaptation. I’ve worked with athletes whose acute RMSSD dropped 20–30 ms below normal for weeks or months, with the expectation that a short taper would fix it. At that point, the system is deeply suppressed. It often takes weeks or months for the nervous system to normalize and for performance to follow.

RMSSD is a valid marker of systemic stress. When HRV is low, the body is not positioned to recover from or adapt to training especially high-intensity work and racing. The daily process to collect and review this data takes less than 15 minutes.

The protocol is simple. The athlete wakes up, uses the bathroom if needed, then puts on a chest strap and sensor. In a seated or standing position, they remain still for a minute to allow heart rate to stabilize, then complete a three-minute test. I discourage optical sensors for HRV; they do not provide true ECG-based data. Sitting or standing introduces a mild orthostatic stress, which reduces variability and improves readiness assessment. We used supine testing for years and found far greater noise.

Use reliable software, Firstbeat, Kubios, or one of Marco Altini’s phone-based applications to analyze the data. I export everything to Excel to dig deeper into respiration, coefficient of variation, and resting heart rate.

If the goal is to prepare athletes optimally, manage recovery, and understand when they are truly positioned to perform, HRV is not optional. It’s inexpensive, easy to use, and one of the most valuable data streams we can collect.

HRV doesn’t make training smarter.
It is exposed when we aren’t smart already.

 

Help Support Our Olympic Coverage!

FasterSkier will have our own Nat Herz on the ground at the Olympics, which, to be honest, is a very big financial lift for us. If you value the thoughtful work and access FasterSkier provides at championships like this one, please consider becoming a Voluntary Subscriber. Your support directly fuels the work we do to cover the people, places, and moments that make our sport special.

Join the FasterSkier community!