Device comparisons
Fitbit vs Apple Watch: Heart Rate and Sleep Accuracy from Published Studies
Independent studies favor Apple Watch for exercise heart rate (concordance with ECG 0.92 vs 0.67 for Fitbit Blaze in Gillinov 2017). For sleep staging, Fitbit Sense (κ 0.42) and Charge 5 (κ 0.41) came close to Apple Watch Series 8 (κ 0.53) in Schyvens 2025. Both brands ran large AFib studies, but with different definitions, so their headline numbers are not directly comparable.
Published study results
| Metric | Fitbit (Sense 2 / Charge 6) | Apple Watch Series 10 |
|---|---|---|
| Exercise heart rateConcordance vs ECGGillinov et al. 2017, n=50 | 0.67Fitbit Blaze | 0.92Apple WatchCloser in this study |
| Sleep stagesCohen's κ vs polysomnographySchyvens et al. 2025, n=62 | 0.42 / 0.41Sense / Charge 5 | 0.53Series 8Closer in this study |
Each row is one statistic from a study already cited on this page. The link names the sample size when this page already states it. This is not a new measurement.
Devices
Fitbit Charge 6 with a coral band: product photograph from the Google Store. Apple Watch Series 10, jet black aluminum with a black sport band: product photograph from Apple Newsroom, September 2024.
Quick answer
Apple Watch has the stronger evidence. During exercise its heart rate reached 0.92 concordance with ECG versus 0.67 for the Fitbit Blaze. For sleep stages against polysomnography, Apple Watch Series 8 scored κ 0.53 versus 0.42 for Fitbit Sense and 0.41 for Charge 5. Both tested older models.
Sources: Gillinov et al. 2017, Med Sci Sports Exerc; Schyvens et al. 2025, Sleep Advances
At a glance
Gillinov et al. 2017 measured exercise heart rate in 50 adults. Schyvens et al. 2025 measured sleep stages in 62 adults. The Fitbit devices were Blaze, Sense and Charge 5; the Apple Watch sleep result is Series 8.
| Specification | Fitbit (Sense 2 / Charge 6) | Apple Watch Series 10 |
|---|---|---|
| Exercise heart rateConcordance vs ECG | 0.67Fitbit Blaze | 0.92Apple WatchCloser in this study |
| Sleep stagesCohen's κ vs polysomnography | 0.42 / 0.41Sense / Charge 5 | 0.53Series 8Closer in this study |
Primary rows only. The study tables further down list more results from the accuracy database.
Hardware specs
| Specification | Fitbit (Sense 2 / Charge 6) | Apple Watch Series 10 |
|---|---|---|
| Battery LifeDiffers | 6+ days (Sense 2); up to 7 days (Charge 6) | Up to 18 hours; 36 hours Low Power Mode |
| Water ResistanceDiffers | 50 m | 50 m (ISO 22810:2010) |
| WeightDiffers | 37.64 g (Sense 2); 30 g (Charge 6) | 36.4 g (46 mm aluminum, GPS) |
| ECGDiffers | ECG app (not in all countries) | ECG app; irregular rhythm notifications |
| SpO2Differs | Red and infrared sensors (Sense 2) | Blood Oxygen app (not for medical use) |
Hardware only. Study results are not repeated in this table. Differs marks values that are not the same. An em dash means this page did not publish that figure.
Specs verified on October 17, 2026: Gillinov et al. 2017, Med Sci Sports Exerc (exercise HR vs ECG, n=50), Schyvens et al. 2025, Sleep Advances (six devices vs PSG, n=62), Perez et al. 2019, NEJM (Apple Heart Study), Lubitz et al. 2022, Circulation (Fitbit Heart Study), Google Store: Fitbit Sense 2 tech specs, Google Store: Fitbit Charge 6 tech specs, Apple Watch Series 10 technical specifications, FDA De Novo DEN180042: Apple irregular rhythm notification.
Pros and cons
Fitbit (Sense 2 / Charge 6)
Pros
- Battery 6+ days (Sense 2), up to 7 days (Charge 6)
- ECG app; Sense 2 adds cEDA skin-conductance sensing
- Fitbit Heart Study: 98.2% PPV for concurrent AF (Lubitz 2022)
Cons
- Fitbit Blaze exercise HR concordance 0.67 vs ECG (Gillinov 2017)
- Sleep-stage κ 0.42/0.41 (Sense/Charge 5) vs 0.53 for Apple Watch S8 (Schyvens 2025)
Apple Watch Series 10
Pros
- Exercise HR concordance 0.92 with ECG (Gillinov 2017)
- Highest sleep-stage κ (0.53, Series 8) of six devices in Schyvens 2025
- ECG app and FDA-authorized irregular rhythm notifications
Cons
- Up to 18 hours battery (36 hours Low Power Mode)
- Requires an iPhone
What published studies show
One table per metric. These rows come from the accuracy database and are not repeated as a second card list. From the wearable accuracy database (v1.0.0).
Heart rate at rest4 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Charge 4 | 7.8 bpm (seated rest)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
| Fitbit Charge 4 | ICC 0.80 (COPD) / 0.79 (controls)Intraclass correlation | ECG chest strap (Polar H7/ | n = 55, 5 days free-living | Hermans et al. (2024)Funding not reported |
| Apple Watch Series 9 / Ultra 2 | 5.91% (MAE 3.73 bpm)Mean absolute percentage error | Polar H10 + Kubios HRV | n = 39, Serial morning measurements over 14 days, 316 readings | O'Grady et al. (2024)Independent |
| Apple Watch Series 6R-R intervals and BPM at rest | 1.15%Mean absolute percentage error | ECG | n = 78, Lab: rest, talking, movie, before/ | Bonneval et al. (2025)Independent |
Heart rate during exercise11 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Charge 6mixed rest/stress/walking | 4.5 bpmMean absolute error | ECG chest strap (Zephyr BioHarness / | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/ | Gielen et al. (2026)Funding not reported |
| Fitbit Inspire 3mixed rest/stress/walking | 9–14 bpm (group range)Mean absolute error | ECG chest strap (Zephyr BioHarness / | n = 45, Rest, cognitive stress, steady and intermittent walking in neutral/ | Gielen et al. (2026)Funding not reported |
| Fitbit Charge 4squats | 15.7 bpm (squats)Mean absolute error | Holter ECG | n = 23, Rest, sedentary, low- and high-intensity phases | Nissen et al. (2022)Independent |
| Fitbit Versa | 4.1%–19.1% (range by intensity)Coefficient of variation | ECG chest strap (Polar H7/ | n = 25, Sitting, walking, running, sprints | Düking et al. (2020)Independent |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Sense | 4.14–10.76%Coefficient of variation | ECG chest strap (Polar H7/ | n = 60, Sitting, walking, running, resistance, cycling | Hajj-Boutros et al. (2023)Funding not reported |
| Fitbit Blazeall exercise conditions | rc = 0.67Lin's concordance correlation | ECG | n = 50, Gillinov: treadmill, bike, elliptical | Gillinov et al. (2017)Funding not reported |
9 more results in the database. Heart rate accuracy across devices →
Overnight heart rate1 study
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 6 | ICC 0.96Intraclass correlation | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
| Apple Watch Series 6 | 1.5 bpm (absolute bias)Mean absolute error | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
HRV (RMSSD)2 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 9 / Ultra 2 | 28.88% (MAE 20.46 ms)Mean absolute percentage error | Polar H10 + Kubios HRV | n = 39, Serial morning measurements over 14 days, 316 readings | O'Grady et al. (2024)Independent |
| Apple Watch Series 6 | ICC 0.67Intraclass correlation | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
| Apple Watch Series 6 | 22.5 ms (absolute bias)Mean absolute error (absolute bias) | ECG | n = 53, Overnight, sleep laboratory | Miller et al. (2022)Other maker funded |
Sleep stages (4-stage)4 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Sense | κ = 0.42Cohen's kappa | Polysomnography (PSG) | n = 37, Sleep laboratory, 1 night | Schyvens et al. (2025)Funding not reported |
| Fitbit Charge 5 | κ = 0.41Cohen's kappa | Polysomnography (PSG) | n = 39, Sleep laboratory, 1 night | Schyvens et al. (2025)Funding not reported |
| Fitbit Sense 2 | 61.7–78.0% (range across stages)Sensitivity | Polysomnography (PSG) | n = 35, Inpatient, 1 night | Robbins et al. (2024)Other maker funded |
| Fitbit Sense 2deep sleep duration | Deep sleep −15 minMean bias (signed) | Polysomnography (PSG) | n = 35, Inpatient, 1 night | Robbins et al. (2024)Other maker funded |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Sense 2 | κ = 0.42Cohen's kappa | Polysomnography (PSG) | n = 26, Two Korean sleep centres, in-lab PSG | Lee et al. (2023)Independent |
| Fitbit Sense 2 | 64.6%Epoch-by-epoch agreement | Polysomnography (PSG) | n = 26, Two Korean sleep centres, in-lab PSG | Lee et al. (2023)Independent |
7 more results in the database. Sleep accuracy across devices →
Sleep vs wake (2-stage)4 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Charge 5 | > 90% (all devices)Sensitivity | Polysomnography (PSG) | n = 62, Sleep laboratory, 1 night | Schyvens et al. (2025)Funding not reported |
| Fitbit Sense | > 90% (all devices)Sensitivity | Polysomnography (PSG) | n = 62, Sleep laboratory, 1 night | Schyvens et al. (2025)Funding not reported |
| Fitbit Sense 2 | ≥ 95%Sensitivity | Polysomnography (PSG) | n = 35, Inpatient, 1 night | Robbins et al. (2024)Other maker funded |
| Fitbit Alta HR | As good as or better than actigraphyStudy finding (no single number) | Polysomnography (PSG) | n = 34, Sleep laboratory, 3 nights incl. disrupted night | Chinoy et al. (2021)Funding not reported |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 8 | > 90% (all devices)Sensitivity | Polysomnography (PSG) | n = 62, Sleep laboratory, 1 night | Schyvens et al. (2025)Funding not reported |
| Apple Watch Series 8 | ≥ 95%Sensitivity | Polysomnography (PSG) | n = 35, Inpatient, 1 night | Robbins et al. (2024)Other maker funded |
2 more results in the database. Sleep accuracy across devices →
Total sleep time1 study
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Sense 2 | −74.5 min (older adults)Mean bias (signed) | Polysomnography (PSG) | n = 19, 1 night | Searles et al. (2026)Funding not reported |
Blood oxygen (SpO2)9 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Charge 4 | ICC 0.32 (COPD, overnight)Intraclass correlation | Medical-grade pulse oximeter | n = 30, 4 nights | Hermans et al. (2024)Funding not reported |
| Apple Watch Series 7 | 3.5% (rest)Root-mean-square error | Medical-grade pulse oximeter | n = 36, Rest, COPD outpatients | Støve et al. (2023)Funding not reported |
| Apple Watch Series 6 | r = 0.89Pearson correlation | Medical-grade pulse oximeter | n = 265, Spot readings | Spaccarotella et al. (2022)Independent |
| Apple Watch Series 6 | −0.23 (LoA −3.49 to 3.04)Mean bias (signed) | Medical-grade pulse oximeter | n = 265, Spot readings | Spaccarotella et al. (2022)Independent |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 7 | 2.2% (RMSE 2.9%)Mean absolute error | Medical-grade pulse oximeter | n = 49, Spot readings, with and without chronic pulmonary disease | Jiang et al. (2023)Independent |
| Apple Watch Series 7 | 2.9%Root-mean-square error | Medical-grade pulse oximeter | n = 49, Spot readings, with and without chronic pulmonary disease | Jiang et al. (2023)Independent |
5 more results in the database. SpO2 accuracy across devices →
Atrial fibrillation detection6 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Sensesmartwatch ECG app, inconclusive tracings counted | 66% (spec 79%)Sensitivity | 12-lead ECG | n = 201, Tertiary cardiology service, single-lead ECG algorithms | Mannhart et al. (2023)Funding not reported |
| Fitbit Sensesmartwatch ECG app, inconclusive tracings counted | 79%Specificity | 12-lead ECG | n = 201, Tertiary cardiology service, single-lead ECG algorithms | Mannhart et al. (2023)Funding not reported |
| Fitbit (PPG irregular rhythm algorithm)PPG irregular rhythm detection | PPV 98.2% (95% CI 95.5–99.5%)Positive predictive value | ECG patch | n = 455699, Remote trial; 1-week ECG patch after detection | Lubitz et al. (2022)Maker funded |
| Fitbit Ionic + FibriCheck algorithmthird-party PPG algorithm (FibriCheck), not Fitbit's own | Sens 97% / spec 100%Sensitivity | 12-lead ECG | n = 73, Around cardioversion, 1-min PPG | Selder et al. (2023)Independent |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 6smartwatch ECG app, inconclusive tracings counted | 85% (spec 75%)Sensitivity | 12-lead ECG | n = 201, Tertiary cardiology service, single-lead ECG algorithms | Mannhart et al. (2023)Funding not reported |
| Apple Watch Series 6smartwatch ECG app, inconclusive tracings counted | 75%Specificity | 12-lead ECG | n = 201, Tertiary cardiology service, single-lead ECG algorithms | Mannhart et al. (2023)Funding not reported |
3 more results in the database. AFib accuracy across devices →
VO2max estimate3 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 10 | 13.2% (MAE 6.79 mL/kg/min)Mean absolute percentage error | Indirect calorimetry | n = 40, Maximal treadmill test (modified Åstrand) | Lambe et al. (2026)Independent |
| Apple Watch | 13.31% (MAE 6.92 mL/kg/min)Mean absolute percentage error | Indirect calorimetry | n = 30, Maximal treadmill test (modified Åstrand) | Lambe et al. (2025)Independent |
| Apple Watch Series 7 | 15.79% (ICC 0.47)Mean absolute percentage error | Indirect calorimetry | n = 19, Graded cycle ergometer test | Caserman et al. (2024)Independent |
Energy expenditure (calories)6 studies
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Versa | 8.0%–32.3% (range by intensity)Coefficient of variation | Indirect calorimetry | n = 25, Sitting, walking, running, sprints | Düking et al. (2020)Independent |
| Fitbit Sense | 13.44–29.66% (poor in all activities)Coefficient of variation | Indirect calorimetry | n = 60, Sitting, walking, running, resistance, cycling | Hajj-Boutros et al. (2023)Funding not reported |
| Fitbit Sense 2physical activity energy expenditure, 10 min | +3.14 kcalMean bias (signed) | Indirect calorimetry | n = 44, 10-min recumbent cycling intervals | Kostrna et al. (2026)Independent |
| Fitbit Surge | No device achieved EE error below 20%Study finding (no single number) | Indirect calorimetry | n = 60, Sitting, walking, running, cycling | Shcherbina et al. (2017)Independent |
More results (2)
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Apple Watch Series 9 | Inconsistent across all groupsStudy finding (no single number) | Indirect calorimetry | n = 30, 10-min treadmill | Chase et al. (2024)Independent |
| Apple Watch Series 4 | 13.5%–27.1% (range by intensity)Coefficient of variation | Indirect calorimetry | n = 25, Sitting, walking, running, sprints | Düking et al. (2020)Independent |
4 more results in the database. Calories accuracy across devices →
Respiratory rate1 study
| Device tested | Result | Reference | Sample | Study |
|---|---|---|---|---|
| Fitbit Charge 4 | ICC 0.84 (COPD) / 0.77 (controls)Intraclass correlation | ECG chest strap (Polar H7/ | n = 55, 5 days free-living | Hermans et al. (2024)Funding not reported |
Verdict
Apple Watch has the stronger independent evidence for exercise heart rate and was slightly ahead for sleep staging; Fitbit offers multi-day battery and works with Android and iPhone. Both have published large-scale AFib studies.
Frequently asked questions
Is Fitbit or Apple Watch more accurate for heart rate?
In Gillinov et al. (2017), Apple Watch reached 0.92 concordance with ECG during exercise versus 0.67 for the Fitbit Blaze. Newer models have not been compared head to head.
Can Fitbit and Apple Watch detect AFib?
Both have irregular rhythm features backed by large studies. The Apple Heart Study reported a PPV of 0.84 (Perez 2019) and the Fitbit Heart Study 98.2% (Lubitz 2022), but the studies used different definitions, so the numbers should not be ranked.
Which tracks sleep better: Fitbit or Apple Watch?
In Schyvens et al. (2025), sleep-stage agreement with polysomnography was κ 0.53 for Apple Watch Series 8, 0.42 for Fitbit Sense and 0.41 for Fitbit Charge 5, all fair to moderate.
Device research records
Datasets, papers and validation results for the models named on this page.