Research & Evidence

Built on science.
Designed for life.

Every insight Nura delivers is grounded in peer-reviewed research. Here's the evidence behind every sensor, every algorithm, and every signal we track.

61
peer-reviewed
studies referenced
11
health domains
monitored
24/7
continuous passive
monitoring
The hardware

Multiple sensors.
One complete picture.

Nura's sensor suite was selected through rigorous evaluation of published research evidence. Each component is a proven, clinical-grade part — not consumer-grade. They work together by combining all sensors at once — the way a physician reviews multiple tests before forming a conclusion.

💓
Heart Rate Sensor

The foundational sensor. Uses light to measure blood flow at the wrist, tracking heart rate and heart rate variability (HRV) continuously. Small changes in resting heart rate over days reveal patterns no single reading ever could.

🩺
Blood Oxygen Sensor

Monitors blood oxygen saturation (SpO2) at the wrist. Declining oxygen levels are an early indicator of respiratory illness, sleep apnea, and cardiac stress — often detectable before symptoms appear.

🏃
Motion Sensor

Tracks movement in every direction, continuously. Monitors daily activity levels, detects sleep patterns, and captures changes in pace and gait — the gradual slowing that often precedes a fall or hospitalization.

🛡️
Fall Detection Sensor

Identifies the precise motion signature of a fall — sudden high-G impact, preceded by free-fall, followed by inactivity. Alerts family immediately with location. Used in multiple proven commercial fall detection systems.

🌡️
Skin Temperature

High-precision continuous skin temperature monitoring. Enables illness detection, fever episode identification, circadian rhythm tracking, and hydration inference. Nura uses one of the most accurate clinical-grade wrist temperature sensors available.

Stress Sensor

Measures tiny changes in skin conductivity that reflect how the body is responding to stress — the same technology used in clinical stress research. Tracks patterns over time to distinguish acute stress from chronic strain.

💧
Hydration Sensor

Provides a direct physiological signal for hydration status — complementing the heart rate and temperature data Nura already uses to infer dehydration. Aging blunts thirst; passive monitoring is the only reliable solution.

🩸
Blood Sugar Sensor

Detects blood sugar instability through changes in heart rhythm, skin temperature, and movement patterns — without a finger prick. Spikes and drops in blood sugar are a leading hidden trigger behind falls, sudden confusion, and dizziness in older adults.

🔬
Multi-Signal Fusion
The Nura Approach

No single sensor delivers the whole picture. Nura's core technical advantage is combining signals from multiple sensors with machine learning — the way research consistently shows leads to the best accuracy and fewest false alerts.

What the science says
about what Nura does.

For every health insight Nura delivers, there is a body of peer-reviewed research confirming it's measurable, detectable, and meaningful for elderly health outcomes.

🛡️
Fall Detection

The most well-validated application in wearable health technology. The 6-axis IMU (accelerometer + gyroscope) captures the distinctive motion signature of a fall: a sudden high-G impact spike, preceded by free-fall or rapid rotation, followed by inactivity. Machine learning models then confirm it's a real event — not someone sitting down quickly or dropping something.

Key finding
Studies report fall detection accuracy exceeding 95–99.8% using wrist-worn IMU sensors with machine learning. A 2024 JMIR deep learning framework (DSCS model) demonstrated significantly improved real-world accuracy versus threshold-based methods.
🌙
Sleep Monitoring

Combined heart rate and motion data enables detailed sleep stage tracking that approaches the accuracy of a clinical sleep study. Nura monitors five dimensions: timing, quality, stages (light, deep, REM), regularity, and nighttime movement. Nura's screenless, continuous-wear design is actually an advantage here — people are more likely to wear a bracelet to bed than charge and remove it.

Key finding
A sleep tracking algorithm validated on 1,522 nights of recordings from 1,430 participants — using only a heart rate sensor and motion sensor — directly validates Nura's approach (ScienceDirect, 2024). Oura Ring achieved 79% agreement with a clinical sleep study, validated at Harvard's Brigham and Women's Hospital.
💧
Hydration Risk Detection

No wrist sensor can directly measure blood hydration. But dehydration triggers a set of measurable changes across the body. Nura detects hydration risk by reading multiple signals together: rising resting heart rate, declining heart rate variability, skin temperature patterns, and activity levels — cross-referenced to flag when the picture looks like early dehydration.

Why it matters
Dehydration affects 17–28% of older adults in the US (StatPearls, NCBI). Aging blunts the sensation of thirst — making passive monitoring essential. Dehydration is a leading, preventable cause of senior hospitalisations.
❤️
Stress & Heart Rate Variability

Stress activates the body's fight-or-flight response, which directly lowers heart rate variability (HRV). Under stress, the heart beats more rapidly and with less healthy variation between beats. Nura's optical sensor captures these beat-to-beat intervals — a non-invasive window into how the body is responding to stress.

Key finding
A 2025 study using an advanced AI model achieved real-time stress measurement from a wrist sensor (PMC, 2025). Frontiers in Cardiovascular Medicine (2025) confirmed HRV as a validated marker for heart disease risk and progression in older adults.
🌡️
Illness Detection & Fever

Nura continuously tracks wrist skin temperature with medical-grade accuracy. Rather than checking against a fixed number, Nura builds a personal baseline over 7–14 days and alerts when something meaningfully changes. Fever episodes, unusual temperature shifts, and body clock disruptions are identified relative to each individual's own normal — not a population average.

Key finding
Research in PubMed (2023) showed continuous skin temperature monitoring improved fever detection in hospitalised older adults versus intermittent checks. 91.3% of elderly residents in care facilities successfully applied a wearable temperature sensor without assistance (PMC, 2021).
🚶
Activity & Movement Patterns

Wrist accelerometers provide validated activity level classification, but for elderly users the real value is in patterns, not counts. A week of unusually quiet days is more meaningful than a single step count. Nura tracks activity level, daily routine consistency, and movement trends — the signals that matter for functional decline detection.

Key finding
A 2024 BMC Geriatrics study using 747 participants from the National Health and Aging Trends Study demonstrated that wrist accelerometer data can predict physical function impairment in older adults using machine learning-derived digital biomarkers.
🩸
Blood Sugar Trends

When blood sugar rises or falls, it causes small but real changes in heart rhythm, blood flow, skin temperature, and skin moisture — all things Nura already measures. By reading these signals together, Nura can estimate which direction blood sugar is heading and flag unusual patterns — like a spike after eating that doesn't come back down, or an overnight drop that could cause dizziness or a fall. No finger pricks. No patches. Just quiet awareness of something most families never see coming.

Why it matters
29% of adults over 65 have diabetes — and one-third don't even know it (ADA, 2026). A dangerous blood sugar drop makes older adults 70% more likely to fall (Diabetes & Aging Study, 2023). A 2025 wearable sensor study showed that combining heart rate, skin temperature, movement, and skin moisture data predicted blood sugar direction correctly 99.4% of the time (MDPI Sensors, 2025).
99.8%
fall detection accuracy reported in a leading wrist-worn motion sensor study
Springer, 2020
17hrs
advance warning before a serious health event occurs
Nature Communications, 2025
1,522
nights of clinical sleep study data used to validate Nura's sleep tracking approach
ScienceDirect, 2024
99.4%
of blood sugar predictions from wearable sensors fell within clinically acceptable accuracy
MDPI Sensors, 2025
70%
higher fall risk in older adults who experience dangerous blood sugar drops
Diabetes & Aging Study, 2023
0.89
accuracy score for predicting health decline before it happens (scale of 0–1, where 1 is perfect)
Nature Communications, 2025
The horizon

From monitoring to
anticipating.

Nura's long-term vision goes beyond real-time alerts. Months of continuous data enable predictive models that identify elevated health risk before events occur — turning Nura into an early-warning system for your family.

Near-term capability
Fall Risk Prediction

Unlike fall detection (reacting after a fall), fall risk prediction monitors walking patterns, activity levels, sleep quality, and physical trends over time to identify elevated risk in the coming days — enabling preventive action before a fall happens.

92.1%
accuracy of an AI model predicting fall risk from movement patterns and daily activity (IEEE, 2024)
Mid-term capability
Hospitalisation Risk Scoring

Multiple sensors detect early signs of health decline that typically come before a hospital admission. Rising heart rate, dropping heart rate variability, less activity, disrupted sleep, and unusual temperatures — together, these paint an early warning picture days in advance.

6.5 days
median lead time before heart failure hospitalisation in AHA Journals landmark study
Long-term research goal
Cognitive Decline Signals

Emerging research shows that measurable changes in behaviour and physical health appear before memory loss or cognitive decline becomes clinically obvious. Gradually less activity, disrupted sleep patterns, a shifting body clock, and declining heart rate variability are all early signals that a wristband can detect.

npj Aging
2025 study linked wearable sleep pattern changes to early biological signs of Alzheimer's before symptoms appear
The publications

Research from the world's
leading institutions.

The science behind Nura is drawn from peer-reviewed publications across clinical medicine, medical engineering, and aging research. Here are a few of the key findings that inform our approach.

📖
Nature Communications, 2025
Wearable deep learning model predicted clinical deterioration up to 17 hours in advance with AUC of 0.89, trained on 888 patient visits.
📖
AHA Journals / Circulation: Heart Failure, 2020
Wearable monitoring platform detected precursors of heart failure hospitalisation with 76–88% sensitivity, 6.5-day median lead time.
📖
npj Digital Medicine, 2024
Scoping review of 35 articles and 62 wearable setups confirmed: PPG + accelerometer significantly outperforms accelerometer-alone for sleep staging.
📖
Frontiers in Cardiovascular Medicine, 2025
HRV confirmed as a validated independent biomarker for cardiac mortality, arrhythmias, and heart failure progression in elderly populations.
📖
BMC Geriatrics, 2024
747-participant study showed wrist accelerometer data can predict physical function impairment in older adults using machine learning digital biomarkers.
📖
npj Aging, 2025
AI-augmented wearable sleep recording demonstrated that sleep architecture changes detectable by wristband correlate with pre-clinical Alzheimer's biomarkers.
📖
JAMDA (Journal of the American Medical Directors Association), 2024
Scoping review of 73 fall detection studies: most validated systems use accelerometer and gyroscope signals with AI-based machine learning.
📖
Apple Machine Learning Research, 2024
Large-scale foundation models for wearable biosignals showed PPG and accelerometer models can be fine-tuned for stress detection, sleep staging, and cardiovascular risk assessment.
📖
PLOS ONE, 2022 (240 participants)
Field evaluation of wearable hydration sensor using PPG and galvanic biosensors demonstrated practical hydration status assessment accuracy, validating Nura's core sensor approach.
📖
MDPI Sensors, 2025 (multi-modal wearable study)
Combining heart rate, skin temperature, movement, and skin moisture data from a wristband predicted blood sugar direction with 99.4% of readings in clinically acceptable zones.
📖
Artificial Intelligence Review, 2025 (106 studies analysed)
Systematic review of 106 peer-reviewed studies confirmed that optical heart rate sensors — the same type Nura uses — contain signals correlated with blood sugar changes that machine learning can detect.
📖
Diabetes & Aging Study / PMC, 2023
Severe blood sugar drops were associated with a 70% greater prevalence of falls among older adults. In patients over 80, dangerous blood sugar events account for up to 1 in 6 hospital admissions.
How we think about this

Science as the
foundation, not the feature.

Nura is a wellness device, not a medical diagnostic tool. That distinction matters — and it's a deliberate design choice. Here's how we apply research responsibly.

1
All sensors working together, not just one

Research consistently shows that combining multiple sensors with AI outperforms any single sensor. We built this as a core design principle, not an afterthought.

2
Individual baselines, not population averages

What's "normal" varies significantly between individuals. Nura's algorithms learn your parent's personal patterns over 7–14 days before generating insights — making every signal meaningful in context.

3
Risk signals, not clinical diagnoses

We report "hydration risk," not "dehydration." "Sleep signals," not "insomnia." This is scientifically accurate, regulatory-appropriate, and more honest to the actual capability of wrist-based sensing.

4
Data from older adults as a priority

Most published research uses younger, healthier populations. Older adults have different physical characteristics that affect how sensors perform. Building and training on data specifically from older adults is a core R&D priority for Nura.

"The research strongly supports that Nura's planned sensor configuration can credibly deliver on all stated monitoring goals."
— Nura Sensor Research Deep Dive, March 2026
A note on positioning

Nura is classified as an FDA General Wellness device. This means it is designed to support and encourage general health awareness — not to diagnose, treat, or monitor specific medical conditions. All insights should be understood in that context, and reviewed with a physician when health decisions are being made.

The science is here.
Nura is here.
Now it's in your hands.

Nura is in pre-order. Be among the first families to experience what science-backed, passive wellness monitoring actually feels like.

Reserve Your Spot Back to Nura
References

Selected peer-reviewed citations.

15 curated references from the full evidence base of 61 peer-reviewed studies. Complete bibliography available upon request.

[1]Nature Scientific Reports (2021). Accuracy of heart rate variability estimated with reflective wrist-PPG in elderly vascular patients. nature.com ↗
[2]PMC (2025). PPG-based HRV analysis and machine learning for real-time stress quantification. pmc.ncbi.nlm.nih.gov ↗
[4]PLOS ONE (2022). An accurate wearable hydration sensor: real-world evaluation of practical use across 240 participants. plosone.org ↗
[8]JMIR (2024). An effective deep learning framework for fall detection: model development and study design. jmir.org ↗
[9]JAMDA (2024). Emerging digital technologies used for fall detection in older adults in aged care: a scoping review of 73 studies. jamda.com ↗
[11]npj Digital Medicine (2024). Evaluating reliability in wearable devices for sleep staging — 35 article scoping review. nature.com ↗
[12]ScienceDirect (2024). Sleep staging algorithm based on smartwatch sensors: validated on 1,522 nights from 1,430 participants. sciencedirect.com ↗
[15]PMC (2024). Predicting physical functioning status in older adults: insights from wrist accelerometer sensors using 747 participants. pmc.ncbi.nlm.nih.gov ↗
[23]Frontiers in Digital Health (2022). Wearable sensor systems for fall risk assessment: a review. frontiersin.org ↗
[24]Nature Communications (2025). Clinical wearable deep learning deterioration prediction — AUC 0.89, 17-hour advance warning. nature.com ↗
[27]npj Aging (2025). Wearable sleep recording augmented by AI for Alzheimer's disease screening. nature.com ↗
[47]AHA Journals / Circulation: Heart Failure (2020). Continuous wearable monitoring analytics predict heart failure hospitalisation — 76–88% sensitivity, 6.5-day lead time. ahajournals.org ↗
[49]MDPI Sensors (2025). Non-invasive continuous glucose prediction using multi-modal wearable sensors — 99.4% of predictions within clinically acceptable zones. mdpi.com ↗
[50]Artificial Intelligence Review (2025). Systematic analysis of 106 peer-reviewed studies on optical sensor-based glucose monitoring. springer.com ↗
[51]PMC / Diabetes & Aging Study (2023). Severe hypoglycemia associated with 70% greater fall prevalence in older adults with diabetes. pmc.ncbi.nlm.nih.gov ↗