How to Interpret Your Wearable Health Data (Oura, WHOOP, Apple Watch)

Modern wearables can surface 30 or more distinct data points, but clinical utility is not evenly distributed across them. The five metrics with the strongest evidence base are heart rate variability (HRV), resting heart rate (RHR), sleep architecture (total duration plus deep and REM percentages), blood oxygen saturation (SpO2), and VO2 max estimates. Each of these reflects a distinct physiological domain — autonomic nervous system regulation, cardiovascular fitness, sleep quality, respiratory function, and aerobic capacity respectively — and all have robust bodies of research linking them to health outcomes and mortality risk.

A secondary tier of metrics provides useful context but is less directly actionable. Skin temperature deviation is sensitive to illness onset and hormonal cycles (particularly ovulation in women). Stress scores and body battery readings aggregate HRV, sleep, and activity data into composite scores that can flag high-load days. Respiratory rate during sleep is an early indicator of illness, respiratory conditions, and overtraining syndrome, with resting values typically in the 12-20 breaths-per-minute range and deviations of more than 2-3 breaths warranting attention.

Some heavily marketed metrics deserve more scepticism. Calorie burn estimates from wrist-worn PPG sensors carry a mean error of approximately 20-28% across validation studies — accurate enough for broad awareness, but not for precise dietary planning. Step count as a standalone metric lacks clinical nuance; a sedentary desk worker hitting 8,000 steps in brief bursts receives less cardiovascular benefit than one accumulating them through consistent low-intensity movement. For a comprehensive comparison of how these devices stack up across metrics, see our article on Oura Ring vs WHOOP vs Apple Watch compared.

The Oura Ring's readiness score is a composite metric ranging from 0 to 100 that aggregates six underlying contributors: HRV balance (your overnight HRV relative to your 30-day personal average), resting heart rate (overnight nadir versus your baseline), body temperature (deviation from your personal mean, typically within ±0.5°C), recovery index (time taken for heart rate to stabilise in the final hours of sleep), previous night's sleep quality, and previous day's activity level. A readiness score below 70 is Oura's signal that recovery is incomplete — this should prompt reduced training intensity, additional sleep, or deliberate stress management.

The Oura sleep score weights total sleep time most heavily, followed by sleep efficiency (percentage of time in bed actually asleep, with 85% as a healthy threshold), restfulness (movement and awakenings), REM sleep duration, deep sleep duration, sleep latency (time to fall asleep — ideally under 20 minutes), and timing relative to your chronotype. Understanding these sub-scores allows targeted intervention: a low efficiency score points to time-in-bed exceeding sleep drive, while low deep sleep may indicate alcohol consumption, elevated stress, or sub-optimal room temperature (the sweet spot for slow-wave sleep is 60-67°F / 15-19°C).

One of Oura's most underappreciated features is its nightly temperature deviation tracking, which records skin temperature fluctuations as small as 0.1°C relative to your baseline. Research conducted with Oura's data showed temperature elevation of +0.2°C or more predicted COVID-19 infection 1-3 days before symptom onset with reasonable sensitivity. For women, temperature deviation also tracks the biphasic menstrual cycle: a sustained rise of approximately 0.3-0.5°C signals ovulation and the luteal phase. This makes Oura's temperature sensor arguably more clinically versatile than any other single wearable feature beyond ECG.

WHOOP's core innovation is its explicit strain-to-recovery model, which frames training load as something that must be earned by recovery capacity rather than accumulated independently of it. Strain is measured on a 0-21 scale derived from time spent in various heart rate zones, with a daily score below 10 reflecting low exertion, 10-14 moderate, 14-18 high, and 18-21 all-out effort. WHOOP calculates recovery on a 0-100% scale from four overnight metrics: HRV (sampled in 5-minute epochs throughout the night, providing far more data points than many competitors), resting heart rate, respiratory rate, and sleep performance relative to the night's sleep need.

The strain-recovery interaction is where WHOOP provides its most differentiated value for athletes. Taking on a strain score of 16 when recovery is at 90% produces positive adaptation; the same strain on a 30% recovery day risks overtraining, immune suppression, and injury. WHOOP's coach function operationalises this by recommending a day's optimal strain ceiling based on current recovery. The system also maintains a rolling sleep debt tracker — a cumulative deficit calculated from the difference between sleep need (individualised by algorithm based on strain, age, and historical data) and actual sleep obtained. Most adults carry a chronic sleep debt of 1-2 hours that suppresses HRV, performance, and immune function.

WHOOP is particularly well-suited to endurance athletes and those whose training volume fluctuates significantly week to week, since the strain-recovery model rewards periodisation. The device's continuous 24/7 optical PPG sampling (with a green plus red plus infrared tri-wavelength sensor) also gives it an advantage for detecting subtle respiratory rate changes indicative of illness. See how AI is transforming wearable health monitoring for a deeper look at how platforms like WHOOP are integrating machine learning into recovery prediction.

Apple Watch occupies a distinct position in the wearable landscape because several of its features have received FDA clearance as medical-grade monitoring tools. The most significant is the single-lead ECG (lead I recording, enabled from Series 4 onward), which captures the electrical activity of the heart across a 30-second epoch. In the Apple Heart Study — the largest prospective study of its kind, enrolling over 400,000 participants — the irregular heart rhythm notification algorithm demonstrated sensitivity of approximately 71% and specificity of 99% for AFib during 48-hour ambulatory ECG patches, with the on-demand ECG showing even higher performance. For context, undetected AFib is responsible for approximately 15-20% of strokes annually in the US.

VO2 max estimation via Apple Watch (Apple calls it "cardio fitness") uses optical heart rate data from outdoor walks and runs, combined with GPS speed data, to estimate maximal aerobic capacity without laboratory testing. Multiple independent validations place Apple Watch VO2 max estimates within 5-10% of gold-standard metabolic cart testing for most users, with accuracy highest for regular runners who train outdoors. The metric is clinically significant: a VO2 max below the "Low" threshold for your age-sex group (typically below 31 ml/kg/min for men under 50 or below 25 for women under 50) is associated with substantially elevated all-cause mortality risk.

Additional features with meaningful health implications include the blood oxygen app (SpO2, measuring peripheral oxygen saturation via infrared LED — useful for flagging hypoxia at altitude or potential sleep apnoea patterns), crash detection and fall detection (the latter validated in populations over 65 with accelerometer algorithms), and the walking steadiness metric introduced in watchOS 8. Walking steadiness analyses gait symmetry, balance, and consistency using the accelerometer and has been validated as a fall risk predictor in older adults. One important limitation of the Apple Watch for research-grade HRV measurement: the optical wrist PPG is less accurate than finger PPG (Oura) or chest ECG (Polar H10) for HRV, particularly at rest, where skin perfusion at the wrist is lower.

Sleep architecture is divided into four stages that cycle approximately every 90 minutes through the night. NREM Stage 1 (N1) is the lightest transitional stage, comprising roughly 5% of total sleep time. NREM Stage 2 (N2), the most abundant stage at 45-50%, is where sleep spindles and K-complexes occur — critical for motor memory consolidation and immune function. NREM Stage 3, or slow-wave sleep (SWS, also called deep sleep), represents 15-20% of sleep in healthy young adults; during this stage, growth hormone is released, cellular repair occurs, metabolic waste products (including amyloid-beta) are cleared from the brain via the glymphatic system, and episodic memory is consolidated. REM sleep (20-25%) is characterised by rapid eye movements and near-complete muscle atonia, and is essential for emotional regulation, creative problem-solving, and procedural learning.

Healthy targets shift with age. Adults under 30 typically achieve 20-25% SWS; this declines to 10-15% by age 50 and 5-10% after 60, partly explaining age-related cognitive decline and reduced physical recovery capacity. For a 7.5-hour sleep, a 30-year-old should target roughly 90-100 minutes of deep sleep and 90-100 minutes of REM. Wearable data showing consistently less than 45 minutes of deep sleep or less than 60 minutes of REM (absent known age-related decline) warrants investigation of contributing factors.

Consumer sleep staging accuracy is a significant limitation to understand. Validation studies comparing wrist-worn accelerometry plus PPG against polysomnography (the clinical gold standard with EEG, EOG, and EMG electrodes) typically find epoch-by-epoch agreement of 70-80%, with the greatest error in distinguishing N1 from wakefulness and N2 from N3. This means individual-night staging data should be interpreted directionally rather than precisely. The practical approach: treat your wearable's sleep stage data as reliable for tracking multi-week trends (consistent SWS decline = meaningful signal), but not for precise nightly staging. For more on accuracy limitations and how to use the data despite them, see our article on how accurate consumer sleep trackers really are.

Signs of poor sleep quality visible in wearable data include: sleep efficiency below 85%, more than 2 awakenings per night lasting over 5 minutes, sleep latency consistently above 30 minutes, total REM below 15% of sleep time, and morning resting heart rate more than 5 bpm above your 30-day baseline. Any of these patterns sustained over two or more weeks warrants evaluation of sleep hygiene factors — blue light exposure, alcohol, caffeine timing, bedroom temperature, and consistency of sleep and wake times.

The most important conceptual shift in interpreting wearable data is recognising that absolute values matter far less than your personal baseline trends. A morning HRV of 42 ms means very little without context — for a highly trained endurance athlete, it may represent a low day, while for a sedentary 55-year-old it could be above average. Population reference ranges exist, but they aggregate across age, sex, fitness level, and measurement methodology, making them a crude benchmark. The higher-value approach is to establish a 30-day personal baseline for each key metric, then flag deviations from that individual normal. Most wearable apps now compute this automatically, but understanding the principle helps you use the data more intelligently.

A practical framework for daily decisions is a traffic light system. Green (more than 10% above your baseline HRV or recovery score): prioritise challenging training, demanding cognitive work, or important decisions. Yellow (within ±10% of baseline): maintain planned activity but monitor how you feel as it progresses. Red (more than 10% below baseline, especially if combined with elevated RHR or poor sleep scores): modify the day meaningfully — replace intense training with zone-1 movement or complete rest, reduce cognitive load where possible, and prioritise sleep that night. Research on elite athletes consistently shows that those who modulate training intensity based on daily recovery biomarkers accumulate greater long-term performance adaptations than those following fixed periodisation plans. The full physiological underpinning of HRV as a decision-making metric is covered in our guide to HRV and cardiovascular health.

Certain wearable signals should prompt medical evaluation rather than just lifestyle adjustment. Persistently low SpO2 (below 94% at sea level across multiple nights) may indicate sleep-disordered breathing or cardiac issues. A resting heart rate that trends upward by more than 10 bpm over two to three weeks without a clear explanation (increased training load, heat, caffeine, or acute illness) deserves clinical assessment. The same applies to a body temperature deviation that stays elevated for more than 72 hours without a resolved infection. Apple Watch AFib notifications, if received, should be followed up with a cardiologist for a standard 12-lead ECG to confirm — the wrist-based single-lead ECG cannot diagnose all arrhythmia subtypes reliably.

To get the most from sharing wearable data with a physician, export a 30-60 day data summary from your device's companion app (Oura, WHOOP, and Apple Health all support CSV or PDF exports). Highlight specific anomalies — nights with SpO2 below 94%, clusters of low HRV days, or sustained elevated resting heart rate — rather than presenting raw data. Physicians are increasingly familiar with interpreting this data, particularly as longitudinal studies now validate wearable metrics against clinical outcomes. Present data as a screening pattern rather than a diagnosis, and let the physician determine whether formal testing (polysomnography, Holter monitor, VO2 max lab test) is warranted.