How WHOOP validates heart rate, sleep, and WHOOP Strap 3.0 data

Podcast episode originally published on July 16, 2019

How does WHOOP validate heart rate and sleep data, and what should you do if your strap seems off? This article pulls six practical questions out of Episode 031 of the WHOOP Podcast, where WHOOP co-founder John Capodilupo explains how wrist-based heart rate sensing works, why band placement changes signal quality, what caused early WHOOP Strap 3.0 connectivity issues, and how member feedback shaped fast software fixes. He also walks through two research efforts, one with Weill Cornell Medicine on sleep changes tied to Alzheimer’s risk, and another with the University of Arizona on insomnia treatment, that show how WHOOP data can extend beyond training into sleep science.

Note: This article covers WHOOP 3.0. For the latest hardware, see the current WHOOP hardware page.

To listen to Episode 031 of the WHOOP Podcast in full, head to the full podcast episode on Spotify.

How WHOOP validates heart rate, sleep, and WHOOP Strap 3.0 data

If you want a broader primer on what WHOOP measures, start there. This article stays focused on six narrower questions: how the sensor reads heart rate, how placement changes signal quality, why syncing issues can happen, how support logs help engineering, and what two early sleep studies suggested about using WHOOP in research settings.

How does WHOOP measure heart rate from the wrist?

WHOOP measures heart rate from the wrist with photoplethysmography, or PPG. In simple terms, the sensor shines light into the skin and analyzes how the reflected light changes as blood volume rises and falls with each heartbeat.

Capodilupo explains that the basic signal is useful, but only after heavy filtering. Light conditions change when you move from indoors to sunlight. Skin contact shifts when the band moves. Arm motion adds noise that can overwhelm the raw pulse pattern. WHOOP addresses that by combining the optical signal with other sensor inputs, then running algorithms that separate the pulse signal from motion and environmental interference.

The larger point from Capodilupo is that heart rate accuracy at the wrist is an algorithm problem as much as a hardware problem. A clean sensor helps, but the real job is recovering enough true signal from a messy real-world stream to estimate heart rate and heart rate variability continuously, all day and all night.

In the episode, Capodilupo puts the technical challenge plainly:

“The raw signal itself is incredibly noisy. We’ve developed a set of sophisticated algorithms over the last 7 years to be able to take the signal in almost any condition, filter out any of the noise sources, and what’s left is an estimation of the heart rate and also heart rate variability.”

That framing also explains why WHOOP spent years collecting reference data against chest straps and electrocardiograms. The goal was not just to prove the sensor could work in a lab. The goal was to train and test it across skin tones, body types, environments, and movement patterns that look like actual use.

What you should take away

• WHOOP uses photoplethysmography, or PPG, to estimate heart rate by measuring changes in reflected light from blood flow under the skin
• Wrist-based heart rate accuracy depends on both the sensor and the algorithms that remove motion, light, and skin-contact noise
• WHOOP was designed to estimate heart rate and heart rate variability continuously, rather than only during isolated workouts or checks.

If you want to hear Capodilupo unpack how photoplethysmography becomes a usable heart rate signal, head to the full podcast episode on Spotify.

What placement changes make WHOOP heart rate data more accurate?

Once the sensing model is clear, the next question is fit. WHOOP data usually improves when the strap sits about 1 inch above the wrist bone on the non-dominant wrist, with firm contact against the skin.

Capodilupo describes fit as the first thing to check when heart rate looks erratic. A loose band lets the sensor shift and creates avoidable motion noise. An overly tight band can also hurt the reading by compressing the tiny blood vessels the sensor needs to read. In the episode, Will Ahmed offers a practical rule of thumb: the band should be tight enough that you can just fit a pinky under it, but not so tight that the band fully pinches down on the skin.

Placement can matter just as much as tightness. Moving the strap slightly higher on the forearm can help because the wrist has tendons and structures that change shape every time you flex the hand or fingers. For some activities, especially those with constant wrist flexion or impact, moving even farther up the arm can produce a cleaner signal.

Capodilupo also addresses one common question directly. WHOOP testing did not show a statistical difference between wearing the sensor on top of the wrist versus underneath it.

“We have seen no statistically significant difference in the quality of the WHOOP data from wearing it on the top of the wrist or the other wrist.”

For sports such as squash or lifting, the better answer may be a different wear location entirely. Capodilupo says WHOOP saw better heart rate estimation in some high-motion cases when members used a bicep band or arm sleeve, because those locations reduce the wrist-specific noise source.

What you should take away

• WHOOP heart rate data is usually strongest when the strap sits about 1 inch above the wrist bone on the non-dominant wrist
• A loose strap can add motion noise, and an overly tight strap can reduce blood flow under the sensor
• Wearing WHOOP under the wrist did not show a statistical advantage in WHOOP testing
• High wrist-motion activities can produce cleaner heart rate data when WHOOP is worn higher on the arm

If you want to hear Capodilupo go deeper on strap fit, wrist placement, and bicep wear options, head to the full podcast episode on Spotify.

Why can WHOOP Strap 3.0 show data lag or connectivity issues?

After fit, the next issue is data transfer. WHOOP Strap 3.0 used Bluetooth Low Energy, and that changed how the app and strap talked to each other.

Capodilupo explains that with WHOOP Strap 3.0, the app is responsible for asking the strap for data. That means force quitting the WHOOP app can create lag, because iOS may stop the app from running the background process that requests new data from the strap. In practical terms, leaving the app open in the background gives WHOOP the best chance to stay current.

He also notes that Apple controls background behavior. Even when you do everything right, iOS can still decide to terminate a background app to free system resources. WHOOP responded with app updates that improved background performance, including version 2.0.5 at the time of the episode.

Capodilupo explains the force-quit issue in exact terms:

“If you force kill it, it won’t do anything, including asking the WHOOP Strap for data, even if the WHOOP Strap is like, ‘Hey, I’m here. I have some data. Do you want it?’”

The episode also covers a separate launch issue. WHOOP found a defect in one electronic component that caused a small percentage of straps to reboot when a certain internal command was issued. A reboot could break the connection and prevent consistent syncing. Capodilupo says WHOOP confirmed the defect with the component manufacturer and replaced affected units.

His troubleshooting sequence was simple. Double tap the strap to check battery lights. Take it off the wrist and tap repeatedly to trigger pairing mode. If the green sensor lights never appear, support should review the case. Many of those early lessons carried forward into later generations, including the product changes discussed in Introducing the WHOOP 4.0.

What you should take away

• WHOOP Strap 3.0 used Bluetooth Low Energy, so the WHOOP app had to request data from the strap
• Force quitting the WHOOP app could create data lag because iOS may block the background process WHOOP needs for syncing
• Some early WHOOP Strap 3.0 units had a component defect that caused reboots and pairing issues, and affected units were replaced
• Updating the app and firmware was part of the fix path for many early syncing problems

If you want to hear Capodilupo unpack Bluetooth Low Energy, background app behavior, and early 3.0 troubleshooting, head to the full podcast episode on Spotify.

How does WHOOP use member feedback to improve algorithms and support?

Once a signal reaches the app, the next layer is confidence. WHOOP does not just estimate heart rate. It also estimates how confident the system is in that estimate based on signal quality, motion, and related conditions.

That matters for two reasons. First, it lets WHOOP monitor edge cases across the wider member base and look for patterns tied to a specific sport, body type, or wear condition. Second, it turns support from a reactive team into a source of engineering data. Capodilupo says that if WHOOP sees the same pattern often enough, the team sets up experiments, reviews the logs, and adjusts the algorithm.

His description of that loop is one of the clearest parts of the episode:

“If we see a pattern emerge, we’ll take a look at the data, set up experiments, and then refine the algorithm.”

Capodilupo also spent time in member support himself during the launch period. He says that being close to incoming tickets helped him see how bugs affected real people and what needed to move up the roadmap. The operational lesson was just as important as the technical one. A technical issue should be sent through the WHOOP app, using Help and then Email Support, because that process attaches a diagnostic log. Those logs include app-side and strap-side information that engineers can use to identify the fault faster.

That loop between continuous data collection, member tickets, and product updates sits inside the larger WHOOP model described in The Story of WHOOP, where ongoing physiology data feeds both coaching insights and product improvement.

What you should take away

• WHOOP tracks confidence in its heart rate estimation, not just the heart rate value itself
• Repeated member issues can become algorithm experiments when WHOOP sees the same pattern across the population
• Sending a technical support request through the WHOOP app helps because the message includes a diagnostic log
• Support feedback can change both engineering priorities and the order of product fixes

What did the Weill Cornell Medicine Alzheimer’s study show with WHOOP?

The episode then shifts from troubleshooting to validation. In a study from Weill Cornell Medicine titled [Sleep Patterns and Autonomic Function in Patients at Risk for Alzheimer’s Disease May Be Used to Predict Cognitive Performance], WHOOP monitored 33 participants selected for elevated Alzheimer’s risk.

Capodilupo says the study looked at sleep architecture, meaning the breakdown of sleep into stages such as slow-wave sleep, REM sleep, light sleep, and awakenings. Researchers also measured cognitive performance and executive function, then compared those outcomes with the sleep and autonomic data.

The result, as Capodilupo describes it, was that WHOOP captured sleep-architecture changes that helped separate participants with a family history or elevated likelihood of Alzheimer’s disease from those without those markers. The episode does not present WHOOP as a diagnostic tool for Alzheimer’s disease, and the study does not justify that leap. What it does suggest is that longitudinal wearable data may help researchers identify meaningful sleep patterns in at-risk groups outside a one-night lab snapshot.

Capodilupo summarizes the WHOOP finding this way:

“WHOOP did show these changes in the architecture of their sleep, and it was able to split the population between people that had a family history of Alzheimer’s disease and people likely to develop Alzheimer’s disease from those who didn’t.”

That is one reason this episode still holds up. The discussion starts with wrist fit and battery life, but it ends with a bigger idea: if sleep architecture can be measured continuously and comfortably, it can support research questions that are hard to study with only infrequent lab visits.

What you should take away

• A Weill Cornell Medicine study used WHOOP to monitor 33 participants at elevated risk for Alzheimer’s disease
• The study focused on sleep architecture and autonomic function, alongside measures of cognitive performance and executive function
• WHOOP data showed sleep-pattern differences between higher-risk participants and comparison participants in that research setting
• The episode frames WHOOP as a research tool for sleep monitoring, not as a diagnostic test for Alzheimer’s disease

If you want to hear Capodilupo go deeper on sleep architecture and the Alzheimer’s research thread, head to the full podcast episode on Spotify.

How did the University of Arizona insomnia study use WHOOP outside a sleep lab?

From Alzheimer’s research, the conversation moves naturally to a second sleep question: can WHOOP help researchers evaluate insomnia treatment without sending every participant into a lab? Capodilupo says that was the appeal of a University of Arizona study called [Cloud-based evaluation of wearable-derived sleep data in insomnia trials].

The study paired WHOOP with cognitive behavioral therapy for insomnia. Instead of relying on polysomnography, or PSG, in a sleep lab, researchers tracked participants in a setting closer to normal life. That choice matters because insomnia is hard to study when the test itself creates an unfamiliar sleep environment.

Capodilupo says WHOOP reflected the expected improvements from the therapy. In his summary, sleep time increased, sleep efficiency improved, wake after sleep onset fell, and total actual sleep time rose. Those are exactly the kinds of outcomes researchers care about when they want to know whether an intervention changed sleep in a useful way.

He explains the logic behind the study design this way:

“They didn’t want to have these people go to the sleep lab where they’ll probably sleep worse anyways and they already have insomnia.”

That point connects back to a broader WHOOP theme. Continuous, at-home monitoring can answer questions that a single night in a lab may distort. It is one reason WHOOP has kept showing up in settings outside performance, including the healthcare-focused discussion in Using technology to improve lives of healthcare workers.

What you should take away

• A University of Arizona insomnia study used WHOOP to evaluate sleep changes during cognitive behavioral therapy for insomnia
• Researchers chose at-home wearable tracking because a sleep lab can distort sleep, especially in people who already have insomnia
• Capodilupo says WHOOP reflected improvements in sleep time, sleep efficiency, wake after sleep onset, and total actual sleep time
• WHOOP can support sleep research by extending observation beyond a single, unfamiliar lab night

The bottom line

• WHOOP measures wrist heart rate with photoplethysmography and then uses algorithms to separate pulse data from motion, light, and skin-contact noise
• WHOOP heart rate quality improves when the strap sits about 1 inch above the wrist bone, on the non-dominant wrist, with firm but not overly tight contact
• WHOOP testing discussed in this episode found no statistically significant difference between wearing the sensor on top of the wrist and underneath it
• WHOOP can produce cleaner heart rate data during high wrist-motion activities when the sensor is moved higher on the arm
• WHOOP Strap 3.0 could show data lag if the WHOOP app was force quit, because the app had to request data from the strap over Bluetooth Low Energy
• WHOOP support works best on technical issues when members send the request through the WHOOP app, because that process includes diagnostic logs for engineering review
• WHOOP was used in a Weill Cornell Medicine study of 33 participants at elevated risk for Alzheimer’s disease to track changes in sleep architecture and autonomic function
• WHOOP was also used in a University of Arizona insomnia study to reflect changes in sleep time and sleep efficiency during cognitive behavioral therapy for insomnia

Frequently asked questions about things discussed in this episode

How does WHOOP measure heart rate from the wrist?

WHOOP measures heart rate from the wrist with photoplethysmography, or PPG. The sensor shines light into the skin, reads how reflection changes with blood volume, and uses algorithms to remove noise from motion, ambient light, and shifting skin contact.

What placement does WHOOP recommend for the strongest wrist signal?

WHOOP recommends wearing the strap about 1 inch above the wrist bone on the non-dominant wrist when signal quality is the goal. WHOOP also recommends a snug fit that keeps steady skin contact without compressing the area so much that blood flow under the sensor drops.

Does WHOOP work differently on top of the wrist versus underneath it?

WHOOP testing discussed in this episode found no statistically significant difference between top-of-wrist and underside-of-wrist placement. WHOOP still allows members to experiment with both positions if one feels more stable during a specific activity.

What does WHOOP do if the app is force quit?

WHOOP Strap 3.0 can show data lag after the WHOOP app is force quit because the app has to ask the strap for data over Bluetooth Low Energy. WHOOP works best when the app is left open in the background rather than manually terminated.

What does WHOOP support need to diagnose a technical issue?

WHOOP support needs the diagnostic log attached through the WHOOP app for the fastest technical review. WHOOP includes both app-side and strap-side information in that log, which gives engineering more detail than a plain email.

What did WHOOP contribute to the Weill Cornell Medicine Alzheimer’s study?

WHOOP contributed continuous sleep and autonomic data in a Weill Cornell Medicine study of 33 participants at elevated risk for Alzheimer’s disease. WHOOP data helped researchers analyze changes in sleep architecture alongside cognitive and executive-function measures.

What does WHOOP do for insomnia research outside a sleep lab?

WHOOP can track sleep changes in normal living conditions, which makes it useful for insomnia research that would be distorted by a lab night. WHOOP was used in the University of Arizona study discussed in this episode to reflect changes during cognitive behavioral therapy for insomnia.

For this topic, WHOOP is most useful when better fit, cleaner signal, and complete diagnostic logs turn wrist data into something reliable enough to guide training decisions and inform real sleep research.