Wearable Metrics Menu | Reputable Health

Device platforms

Continuous signals

Metric families

80+

Study-ready endpoints

Unified data model

Every provider is different. The five platforms use different sensors, algorithms, field names and units, and the same metric name can mean different things on different devices. That is why device selection is a design decision made before enrollment opens, and why one device is then used for the entire cohort. Reputable normalizes every platform into a single schema, so endpoint definitions and analysis methods stay consistent from study to study. Where a metric is not delivered directly by a device, we state whether we can derive it and how.

How to read this menu

Each metric below is scored against all five platforms. One symbol per device tells you whether that metric is available, and if so how it reaches your dataset. Symbols are colored by device.

Native

The device measures and reports this directly. No processing on our side.

Derived

Available, but Reputable calculates or unit-converts it from the fields the device does provide.

Not available

This platform does not report the metric. A study needing it should select a different device.

n/r

Not reviewed

Outside the scope of this review round; availability not yet confirmed either way.

88 / 88 metrics

Sleep

The richest family of metrics across all five platforms and the most common primary endpoint in supplement, device and wellness studies. Every provider reports sleep stages; they differ most in whether they expose the underlying timeline and the overnight cardiovascular signals.


Total sleep time Minutes actually asleep across the night
Time in bed Total time between lights-out and final wake
Sleep stages Minutes in light, deep, REM and awake
Sleep efficiency Share of time in bed spent asleep
Sleep latency How long it took to fall asleep
Time awake after falling asleep Total minutes awake mid-night (WASO)
Awakenings / disturbances Count of interruptions during the night
Sleep cycles completed Number of full sleep cycles
Overall sleep score Device's own 0–100 nightly quality score
Lowest heart rate overnight Resting HR proxy, a key recovery marker
Average heart rate overnight Mean HR across the sleep period
Overnight HRV Autonomic recovery marker; see the HRV design note
Respiratory rate Average breaths per minute while asleep
Skin temperature change Nightly deviation from personal baseline
Full stage timeline (hypnogram) Stage-by-stage record across the night
Overnight HR & HRV timeline Readings every 5 minutes through the night
Nap detection Daytime sleep captured separately from the night
30-day stage baselines Participant's own rolling average per stage

Daily activity & movement

Day-level totals used for adherence monitoring, covariate adjustment and energy-balance endpoints. Note that WHOOP does not report steps or distance at all. If either is an endpoint, WHOOP is not the right device to build the study on.


Daily step count Total steps taken across the day
Distance traveled Meters walked and run over the day
Active calories burned Energy burned above resting metabolism
Total calories burned Active plus resting energy expenditure
Resting metabolic burn (BMR) Baseline calories burned at rest
Floors climbed Elevation gain expressed as flights
Minutes at low / moderate / high intensity Time split by activity intensity band		n/r
Sedentary time Minutes spent inactive while awake		n/r
Device non-wear time Minutes off the body; our primary wear-compliance signal		n/r
Daily activity score Device's own 0–100 daily movement score		n/r
Training strain (0–21) WHOOP's cardiovascular load score for the day		n/r
MET-minutes by intensity Metabolic-equivalent workload, four bands		n/r
Daily heart rate summary Min, mean and max HR across the day		n/r
Time in heart rate zones Minutes in fat-burn, cardio and peak zones		n/r
Inactivity alerts Count of prolonged sitting prompts		n/r
Goal attainment Progress against the participant's own targets		n/r

Workouts & training sessions

Session-level data for exercise, performance and sports-nutrition studies. Every platform records session timing and distance; they diverge sharply on heart rate detail and on how reliably the sport type is labelled.


Session start, end & duration When the workout happened and how long it lasted
Sport / activity type What kind of workout it was		n/r
Distance covered Meters traveled during the session
Calories burned in session Energy expenditure for the workout		n/r
Average heart rate Mean HR across the session		n/r
Peak heart rate Highest HR reached during the workout		n/r
Time in heart rate zones Seconds spent in each training zone		n/r
Training strain (0–21) Cardiovascular load for the session		n/r
Perceived intensity label Easy / moderate / hard classification		n/r
Elevation gain Meters climbed during the session		n/r
Steps & average speed Step count and pace within the workout		n/r
Swim stroke count Strokes recorded in pool sessions
Auto-detected vs manually logged Whether the participant entered it themselves		n/r
Recording completeness Share of the session with valid sensor data		n/r
GPS-tracked flag Whether location data backed the session		n/r

Continuous physiological signals

Beyond daily summaries, we can pull time-stamped readings throughout the day. Forty-seven distinct signal types are supported in total, though coverage varies enormously by platform. Apple Health Kit reaches into clinical territory that no other consumer wearable does.


Heart rate Continuous beat-to-beat readings
Heart rate variability Beat-interval variation, in milliseconds
Respiratory rate Breaths per minute through the day
Blood oxygen (SpO₂) Peripheral oxygen saturation
Stress level Device-computed physiological stress score
Body temperature Absolute reading and deviation from baseline
VO₂ max Estimated cardiorespiratory fitness
One-minute heart rate recovery How fast HR drops after peak exertion
Weight & body composition Weight, body fat %, BMI, lean mass
Daylight exposure Time-weighted light exposure; useful for circadian endpoints
Standing time & stand hours Time upright and hours with movement
Water intake Logged hydration volume
Sleep breathing disturbance Breathing irregularity events per hour asleep
Blood pressure Systolic and diastolic from a paired cuff
Blood glucose Readings from a paired CGM or meter
ECG voltage & AFib burden Single-lead ECG trace and rhythm burden
Lung function FEV1, forced vital capacity, peak flow, inhaler use
Nutrition logging Carbohydrates and caffeine intake
Mindfulness minutes Logged meditation and breathing sessions
Safety events Falls, irregular-rhythm and sleep-apnea alerts

Recovery & readiness

Day-level composite scores that roll sleep, HRV and resting heart rate into a single number. Useful for tracking day-to-day fluctuation within a participant, but the scores are proprietary and not directly comparable across platforms — a WHOOP recovery of 62 and an Oura readiness of 62 mean different things.


Recovery score (0–100) Device's overnight recovery composite
Readiness score Aggregated readiness-to-perform indicator
Body battery Garmin's daily energy reserve, 0–100
Daytime stress score Continuous physiological stress estimate
Strain vs recovery balance Whether the day's load exceeded recovery capacity
Sleep need vs actual Recommended sleep minus what was achieved
HRV baseline deviation Today's HRV against a personal rolling baseline
Daily resting heart rate Lowest resting HR recorded for the day
Recovery time recommendation Hours advised before the next hard effort
Autonomic balance Sympathetic vs parasympathetic dominance

Women's health & cycle

Cycle and reproductive-health tracking. Apple, Fitbit and Garmin offer dedicated cycle logging; Oura infers cycle phase from temperature trends. WHOOP surfaces no cycle data at all, which rules it out for any study with a cycle-related endpoint.


Menstrual cycle logging Period start, end and flow level
Cycle phase prediction Current phase of the menstrual cycle
Ovulation prediction Estimated ovulation date
Fertility window Days of peak fertility in the cycle
Cycle regularity Variability in cycle length over time
Basal body temperature Morning waking temperature
Symptom & mood logging Cramps, mood and sex-drive entries
Pregnancy tracking Weeks gestation and milestones
Contraceptive logging Method and adherence tracking

Coverage

Platform coverage at a glance

Share of the 47 continuous signal types each platform supports. Breadth is not the same as fitness for purpose. Oura supports fewer signal types than Fitbit but delivers the most detailed overnight data of any device on this list.

Apple Health Kit45 / 47

Garmin22 / 47

Fitbit21 / 47

Oura Ring16 / 47

WHOOP9 / 47

Availability also depends on device model and on the permissions each participant grants at enrollment. We confirm both during onboarding and report actual capture rates in every study report.

Design

What this means for your study design

Selecting the device for your endpoints

As every participant in a Reputable study wears the same device, device selection is an endpoint decision, made before enrollment opens and fixed for the duration of the study. We work backward from the claim: list the endpoints the study has to support, identify which platforms deliver all of them, then choose among those on participant burden, cost and expected compliance. A metric that only one platform reports can absolutely carry a primary analysis, provided that platform is the one the study is built on.

HRV is not one measurement

Apple Health Kit reports HRV as SDNN; Garmin, Oura and WHOOP report RMSSD. These are different calculations on the same underlying signal and produce different absolute values. Running one device per study removes this problem inside the study, but it still shapes what the numbers can be compared against. Published reference ranges, competitor benchmarks and results from any prior study only apply if the HRV method matches. We confirm which method the selected device uses and pre-specify it in the protocol.

Wear compliance is measured, not assumed

Missing data in a wearable study is rarely random. We track device non-wear and per-metric capture rates for every participant, pre-specify the minimum wear threshold for a valid day, and report attrition transparently in the study report.

Quality control on incoming data

Consumer wearable feeds contain occasional implausible values: out-of-range heart rates, corrupted duration fields, unit mismatches. Every incoming record passes automated range and consistency checks before it enters a locked analysis dataset, and any field that fails validation is excluded with the exclusion documented in the study report rather than silently repaired.

Next step

Building a study around these metrics?

Tell us the claim you want to support and we will map it back to the endpoints, device and sample size that can actually support it, then design, run and report the study end to end.