Best Wearables for Sleep: Scientific Rankings

Best Wearables for Sleep: Scientific Rankings

Sleep Trackers: Which One is the Best?

Overview of Sleep Tracker Analysis

  • The speaker reviews various sleep trackers, including brands like Apple, Google, Samsung, Aura, Whoop, Garmin, Polar, and Fitbit.
  • A spreadsheet with statistics and references is available in the video description for further details.
  • The analysis focuses on two types of results: overall sleep detection (asleep vs awake) and tracking specific sleep stages (light sleep, deep sleep, REM sleep).

Initial Findings on Sleep Detection

  • The first focus is on whether devices can reliably detect when a person is asleep or awake. This is illustrated through a plot comparing different brands.
  • Sensitivity measures how well a device detects sleep time; specificity measures how well it detects awake moments. Perfect devices would score 100% on both axes.

Brand Comparisons: Garmin vs Polar vs Fitbit

  • Garmin devices show high sensitivity but low specificity; they often classify entire nights as sleep without detecting wake moments. This indicates a potential flaw in their tracking algorithm.
  • Polar devices have lower sensitivity than Garmin but better specificity; this suggests a trade-off between the two metrics rather than one brand being outright superior to another.
  • Fitbit's performance appears slightly better than both Garmin and Polar based on average results from studies focused on normal sleepers versus those with sleep issues.

Insights into Whoop and Aura Devices

  • Whoop devices sacrifice some sensitivity for improved specificity compared to Garmin; they also fall along the same trade-off line observed in other brands' data.
  • The older generations of Aura rings exhibit similar trade-offs as Whoop and Garmin; however, the newer Aura Ring 3 shows improvements in both sensitivity and specificity over its predecessors.

Apple Watch Performance

  • Limited studies exist for Apple Watch performance; however, findings suggest that recent models (Series 6 and Series 8) perform significantly better due to updated algorithms introduced recently by Apple itself.
  • Overall conclusions indicate that Apple and Aura are currently among the best at detecting sleep states since their algorithm upgrades while other brands display varying degrees of trade-offs between sensitivity and specificity across their models.

Evolution of Sleep Stage Tracking

  • Traditional measurements focused primarily on detecting if someone was asleep or awake until more complex methods emerged during the last decade to track specific stages of sleep defined scientifically (light, deep, REM).

Sleep Tracking Devices: Which Ones Are Reliable?

Overview of Sleep Tracking Technology

  • The acceptance of smartwatches and bands in sleep research has increased, highlighting the need to identify reliable devices among various options.
  • Reference devices like polysomnography and EEG are used to evaluate smartwatch performance, but studies often test only a few devices at a time.

Research Methodology

  • The speaker, Rob, is a postdoctoral scientist specializing in biological data analysis. He reviews literature to provide an overview of device performance.
  • A plot displays average agreement across four sleep stages; devices should ideally be positioned towards the top right for optimal performance.

Device Performance Insights

  • Garmin devices show poor performance, clustering towards the bottom left of the plot compared to other brands.
  • Fitbit devices outperform Garmin and Polar with better average agreement and less variance in worst sleep stage tracking.

Comparative Analysis of Devices

  • Whoop strap performs similarly to Fitbit but slightly worse; both are better than Garmin and Polar for sleep stage tracking.
  • Oura Ring 2 shows comparable results to older Whoop straps and Fitbits, while Oura Ring 3 demonstrates significant improvements due to algorithm updates.

Evaluating Apple Watch Performance

  • Apple Watch Series 8 performs well, slightly better than Fitbit and Whoop but may not be as reliable as Oura based on non-peer-reviewed studies.
  • Overall findings suggest that healthy sleepers can rely on Oura, Fitbit, Whoop, and Apple for accurate sleep stage tracking.

Considerations for Sleep Disorders

  • Xiaomi's Mi Band 7 is identified as the least reliable device based on personal experience with its accuracy issues.

Sleep Tracking Devices: Performance Overview

Best Performing Devices for Healthy Sleepers

  • The top devices for tracking sleep are the Apple Watch and Oura Ring, followed by Whoop Strap and Fitbit. Other brands like Polar, Garmin, and Xiaomi show poorer performance.
  • For individuals without major sleep issues, this overview can guide device selection; however, options are limited for those with severe sleep problems.

Research Limitations and Personal Testing

  • Most studies referenced older devices due to the lengthy process from research to publication in peer-reviewed journals.
  • The speaker tested various smartwatches on themselves using EEG and PSG as reference devices to compare results against scientific studies.

Comparative Analysis of Device Performance

  • Results indicate that Apple Watches outperform Oura Rings in average agreement during personal testing, while both perform well according to scientific studies.
  • Fitbit and Whoop Strap rank as second-tier devices; Garmin shows varied performance but generally underperforms compared to leading brands.

Insights on Understudied Brands

  • Limited testing on Polar devices suggests they perform similarly to Garmin; Xiaomi devices consistently rank lower in performance metrics.
  • A comparison of personal test results with scientific data reveals similar patterns, though Apple appears slightly better in personal tests than indicated by studies.

Conclusion on Device Efficacy

  • Both Apple and Oura are effective sleep stage trackers for healthy sleepers. However, many brands were excluded from scientific reviews despite being tested personally.
  • The speaker plans to share updates through a newsletter and social media regarding ongoing tests of various sleep tracking variables.

Additional Observations on Newer Brands

  • Some high-performing brands like HLE Pod (models 3 & 4), which lack prior scientific study backing, have shown promising results in personal testing.

Sleep Stage Tracking Devices: A Comparative Analysis

Overview of Sleep Stage Trackers

  • The effectiveness of sleep stage trackers varies, with Apple, Nua, and Aura leading the market. Fitbit, Google, and Whoop follow closely behind.
  • For budget-conscious consumers, Fitbit is recommended; however, there are concerns about potential subscription fees for accessing sleep tracking features.

Device Recommendations Based on Budget

  • If budget allows for more investment in sleep tracking, Aura is suggested as a top choice despite its subscription model.
  • Apple Watch offers excellent sleep stage tracking along with a high-quality heart rate sensor but has limitations such as a one-day battery life and poor data interpretation in the Apple Health app.

Additional Device Insights

  • Whoop strap provides decent sleep tracking capabilities with good battery life and superior sports functionalities compared to Aura but also requires a monthly fee.
  • For those willing to invest significantly without wearing devices, the AG Sleep Pod offers advanced features like independent temperature control and snore reduction in its ultra version.

Considerations for Sleep Trackers

  • Effective sleep stage tracking may only be beneficial for healthy sleepers; individuals with existing sleep issues might not see improvements from these devices.
  • Personal experiences can vary widely; feedback from users regarding their experiences with different devices is encouraged to help others make informed decisions.

Supporting the Channel

Video description

▬▬▬▬▬▬▬▬▬▬👇DISCOUNT ON BEST DEVICES!👇▬▬▬▬▬▬▬▬▬▬ 🔴 Whoop Strap 4.0: https://join.whoop.com/QuantifiedScientist ($30 off any new WHOOP membership) 🔴 Eight sleep 350$/200$ (€) discount*: https://eight-sleep.ioym.net/TheQuantifiedScientist + add code “UNBOXTHERAPY” or “TQS” for full discount 🔴 Oura Ring Link: I recommend you search for a discount on reddit with the "Refer-A-Friend" program! If you can’t find one: https://ouraring.sjv.io/rQ4L33 (affiliate link that supports the channel, but doesn't have a discount) 🔴 General Amazon link*: https://geni.us/thequantifiedscientist ☝️Amazon Affiliate* (paid) links. Supports the channel, doesn't cost you any more!☝️ 🟢 Instagram: https://www.instagram.com/quantified_scientist/ 🟠 Shorts: https://www.youtube.com/@smartwatchshorts 🔵 Newsletter: www.robterhorst.com 🟡 Twitter: @QuantifiedRob * Affiliate (paid) links support the channel by earning me a commission on any sale, and it doesn't cost you any more! My opinions are completely my own, and this content is not sponsored. ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 📸My Gear* (paid links, supports the channel and doesn’t cost you any more)📸 Camera body (main): https://geni.us/SonyA6600_body Camera lens (main): https://geni.us/SigmaContemporary Camera (B-roll 1): https://geni.us/Sony_ZVE1 Camera (B-roll 2): https://geni.us/CanonEosM50MarkII and https://geni.us/Sony_ZVE10 Light: https://geni.us/Godox_VL150 Softbox: https://geni.us/NiceFotoSoftbox Light stand: https://geni.us/WalimexLightstand ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ Reference devices: ZMax + Dreamento: https://www.biorxiv.org/content/10.1101/2023.08.18.553744v1 & https://github.com/dreamento/dreamento Polar H10: https://pubmed.ncbi.nlm.nih.gov/31004219/ 🔴 Polar H10: https://geni.us/hnXc Spreadsheet with all information: https://docs.google.com/spreadsheets/d/1XDeGOXzLuR05PTLnOx1eEDdtt1DALVjq2WHiWNieCic/edit?usp=sharing Note, as a starting point I used a table from a very nice paper: Birrer et al. 2024 Table 1 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10948771/) Timestamps: 00:00 Quick Intro 00:31 Sleep - Wakefulness Detection (Scientific Papers) 04:33 Sleep Stage Detection (Scientific Papers) 10:59 Most Recent Device Testing (Own Testing, Pt1) 14:09 Most Recent Device Testing (Own Testing, Pt2) 15:27 Conclusions I went through hundreds of scientific publications to find out which the best sleep stage trackers are that money can buy, depending on your budget. Is it Apple, Google, Samsung, Oura, Garmin, Polar or maybe Fitbit. Well, today you are going to find out! This is a video I’ve been wanting to make for a long time, and the results are super interesting. It also shows us some major problems with sleep tracking I have not addressed in my previous videos. A spreadsheet with all the statistics and references I base my findings on is in the description below, and of course there are timestamps on this video's timeline to make things easier for you. Detecting whether you are asleep or awake is what wearables would traditionally measure in the early 2010s, with the launch of the first Fitbit Tracker for instance. However, during that same decade more and more companies tried to do something much more complicated: tracking the sleep stages defined by scientists, albeit in a slightly simplified form, which were termed light sleep, deep sleep and REM sleep. This is actually really hard, since these are usually studied in the lab by measuring brain waves, eye movements, and muscle movements. Instead, a wearable has to do it based on just your heart beats and movements. In my testing I’ve seen that wearables are becoming much better at this, and more people than ever seem to be using them to keep track of their sleep patterns. Over the last few years I’ve also found that it’s becoming more and more accepted in the scientific community to use smartwatches and smart bands in sleep research. However, for both consumers and scientists it’s important to know which devices are good at tracking stages and which are not so reliable, since some are quite bad. Luckily, scientists are recognizing the value in actually testing the performance of wearables using polysomnography or EEG, giving us an idea which smartwatches are good, and which are not. However, they usually test only one or a few devices at a time, so similar to before I went through literature, and extracted the results from all papers I could find to give us a good overview of the performance of different devices. Intro Music from #Uppbeat (free for Creators!): https://uppbeat.io/t/moire/new-life License code: OBLD6VBBCTRDFZQP