The Science Behind Wearables: Health Science Newsletter

The Gap Between the Metric and the Biology

A large portion of health tech gets built on metrics that product teams treat as black boxes. HRV comes out of the API. Sleep stages get rendered in a chart. Readiness scores appear in a dashboard. What those numbers actually represent, physiologically, is treated as someone else's problem.

That assumption creates real consequences. When you're building products where data informs health decisions, the space between "the API returns a number" and "here's what that number measures and what the peer-reviewed literature says about its validity" affects how features get designed and how confident you can be that your product is telling users something accurate.

The existing content around health tracking rarely fills that gap. Academic papers require domain expertise to parse. Consumer wellness content strips out the mechanism to get to the tip. Popular wearable coverage tends to treat HRV as a single "stress score" and sleep staging as a solved problem. Neither framing is wrong exactly, but both omit most of what makes the underlying science useful.

Momentum builds healthcare software and wearables infrastructure, working with companies across the US and Europe. As part of that work, we wanted to contribute more than tooling. The science behind what wearables measure deserves rigorous, sourced, practitioner-level coverage. That's why we launched "The Science Behind Wearables" on Substack.

About the Newsletter

"The Science Behind Wearables" is authored by Anna Zych, Health Science Lead at Momentum. Anna's background spans neuroscience research at Max Planck Institute for Biological Intelligence and Princeton University. She was recognized in Forbes 25 Under 25 in Poland.

At Momentum, she works where health science and product development meet: translating what the research actually shows into context that's useful for both health enthusiasts and the teams building the products they use. The newsletter reflects that orientation.

Each post covers one health metric or physiological concept that wearables track. The structure is consistent across posts: what the metric actually is, how it gets measured, what the literature says about that measurement method, and where the most common misconceptions originate. RMSSD and SDNN are not interchangeable. Sleep staging from a consumer wrist device is not equivalent to polysomnography. PPG signal quality depends on contact pressure in ways that matter for how you interpret the output.

The audience spans two groups: health enthusiasts who want to understand their data at more than surface level, and product builders who need to understand what they're building on top of. The posts assume scientific curiosity rather than scientific credentials. They cite the literature, define the terminology, name the measurement tradeoffs, and treat the reader as someone capable of handling the actual mechanism.

The framing matters. Most health content either talks down to the reader or assumes graduate-level familiarity with physiology. The newsletter is written for people who want the real explanation and are willing to engage with it, without requiring a biology degree as the price of entry. That's a narrower target than general wellness content, but it's a more useful one.

Some posts will be exclusive to Substack subscribers. The platform gives us space to publish depth pieces that warrant a dedicated readership rather than general blog traffic, and it gives subscribers a reason to follow directly rather than wait for a summary.

What We’ve Published

What Is HRV? The Science Behind Heart Rate Variability

The first post covers heart rate variability from first principles. HRV measures the variation in time between successive heartbeats. It’s shaped by the autonomic nervous system: the parasympathetic branch slows the heart and increases variability, while the sympathetic branch does the opposite. Most popular health content treats HRV as a single readiness score, but that framing collapses most of what makes the metric scientifically useful.

The post covers how HRV is calculated, the differences between RMSSD and SDNN as quantification methods, and what each one actually captures about autonomic function. It also addresses which lifestyle factors move HRV in practice, covering sleep quality, alcohol, training load, and acute stress responses, grounding each in the literature rather than wellness advice. The goal is to give readers a working understanding of the metric they're reading on their device, not a repackaged summary of what their device already tells them.

HRV and HR Accuracy in Wrist-Worn Wearables

The second post goes one layer deeper: how wrist-worn devices actually convert light into a heartbeat signal. Photoplethysmography (PPG) works by shining LED light into the skin and detecting changes in light absorption caused by blood flow. The accuracy of that signal depends heavily on contact pressure between sensor and skin, which is why the same device produces different quality data across users and conditions.

The post covers the full signal chain from optical sensor to derived metric, and explains why the distinction between SDNN and RMSSD matters when interpreting wearable HRV data. Using the wrong metric doesn't give you a slightly inaccurate number on the same scale. It gives you a number from a different measurement entirely, one that captures different physiological information. That's the kind of distinction that matters if you're building features around HRV data or interpreting your own readings seriously.

Decoding Sleep: What Your Body Does at Night

The third post covers the science of sleep. Not sleep hygiene advice, but the actual brain architecture that drives it. The post covers the adenosine-based sleep pressure system, circadian regulation, and the structure of NREM and REM cycles across a full night, including how those cycles shift in proportion as the night progresses and what that pattern means for recovery.

The second half addresses measurement. It covers the expert consensus on what constitutes quality sleep data, where consumer wearables perform well relative to clinical polysomnography, and where the gap between them is large enough to matter. Sleep staging from an accelerometer-based wrist device is not the same measurement as EEG-based staging in a sleep lab. The post explains what consumer devices can’t reliably detect, and why sleep quality functions as a signal of systemic health rather than a measure of rest in isolation.

For more on building software that handles health data responsibly, including the architectural decisions that affect data quality, see our guide to healthcare software development.

{{lead-magnet}}

Early Response

The newsletter launched without a large pre-existing audience. After two posts, it reached 256 subscribers, up from 65 at the start of that period, a 293.8% increase. Views over the same 14 days came in at nearly 1,900, up from 525, an increase of 262.3%.

More informative than the aggregate numbers is who subscribed. The CEO of Oura is among the early readers which signals that the audience isn’t limited to general health enthusiasts. People building health technology products are reading it, which is exactly the overlap the newsletter is designed to reach.

The growth suggests there’s demand for health science content that takes the underlying biology seriously rather than packaging it for maximum shareability. When rigorous, sourced coverage of wearable science gets written for practitioners rather than for general audiences, it finds them.

The 14-day window captured by these numbers is also the period when the newsletter had no brand recognition and no distribution network beyond a cold start. The shape of the subscriber curve reflects organic sharing rather than a coordinated launch push. That’s a meaningful signal about the quality of the content itself.

What’s Coming and How to Subscribe

The next posts will continue covering the metrics wearables track. Each will follow the same approach: the physiology, the measurement method, the literature, and the real tradeoffs for people who care about data quality.

Some upcoming posts will be exclusive to subscribers. Subscribing is free and gives you access to the full depth pieces as they come out.

If you’re building health tech products or working with wearables data, subscribe at thesciencebehindwearables.substack.com. If your goal is to understand what the devices you wear are actually measuring, the same applies.

Momentum also maintains Open Wearables, an open-source platform that connects wearable devices to health applications. If you’re building on wearables data and want an infrastructure layer that handles provider integrations, health scores, and the API surface, Open Wearables is designed for that. You can read about the latest release in our Open Wearables 0.4.3 release notes.