Insulin Resistance Trends: An Industry First

The headline feature of Health Guardian is Insulin Resistance Trends, the first time a consumer wearable has attempted to estimate metabolic burden over time .

Unlike continuous glucose monitors (CGMs) that measure blood sugar directly, this feature uses passive wrist sensors — accelerometer, barometer, heart rate, and skin temperature — combined with AI to estimate monthly shifts in how hard the body works to produce energy from food . It does not provide real-time glucose readings or blood sugar levels . Instead, users receive a monthly summary in the Google Health app that flags notable changes, potentially prompting lifestyle adjustments or a doctor visit .

Clinical Validation

The feature is built on research from the WEAR-ME study, a peer-reviewed study published in Nature in March 2026 . The study enrolled 1,165 remote participants across the US who wore Fitbit or Pixel devices . By combining wearable data with routine blood biomarkers (fasting glucose, lipid panel), the deep neural network model predicted insulin resistance with an auROC of 0.80, 76% sensitivity, and 84% specificity at a HOMA-IR cut-off of 2.9 . The model performed better when using both wearable data and blood biomarkers than either alone .

This validation underpins the consumer feature, but Google emphasizes that the tool is designed for trend awareness, not diagnosis .

Blood Pressure Trends

Blood Pressure Trends is a monthly trend summary that estimates vascular stress without requiring a traditional blood pressure cuff . The watch passively observes pulse and movement patterns using optical blood flow and motion sensors, analyzed by a multimodal AI model .

Google tested the feature against 24-hour blood pressure monitors, using an advanced waveform model trained on more than 1 billion minutes of data from over 500,000 participants . Like insulin resistance, this is not a continuous measurement — it provides a single contextualized insight on the first of each month . In Japan and Mexico, the feature is referred to as "Vascular Stress / Vascular Stress Trends" .

Sleep Breathing Quality Trends

Sleep Breathing Quality Trends delivers a monthly report tracking changes in breathing patterns during sleep . Google explicitly states this is not sleep apnea detection or a diagnostic tool — it is a wellness-oriented awareness feature that surfaces shifts in respiratory patterns over time .

Breathing Emergency Detection

The most safety-critical feature in the suite is Breathing Emergency Detection, an opt-in, industry-first capability on supported Pixel Watches . It monitors for a critical, sustained drop in blood oxygen saturation (SpO2) — for instance, from drug toxicity, severe pneumonia, or choking . If the user becomes unresponsive, the watch automatically calls emergency services and pre-registered contacts .

Crucially, this feature is initially available only in EU countries where it carries a CE mark . It has not received FDA clearance in the US at launch . Google also notes the feature is not intended for people with certain pre-existing respiratory conditions and may not detect all breathing emergencies .

Availability and Rollout

Availability varies by region/country and device . Google recommends consistent wear for accurate passive tracking, as first detailed summaries arrive after about one month of use .

Key Limitations to Understand

• Not diagnostic: Google does not present any Health Guardian feature as a medical diagnostic tool. They are trend-awareness features meant to surface shifts, not confirm conditions .
• No FDA clearance for US: Breathing Emergency Detection lacks FDA clearance in the US at launch . The FDA status of the trend features (insulin resistance, blood pressure) is not detailed — they ship as wellness features, not regulated medical devices .
• Insulin resistance is inferred, not measured: The feature estimates trends using surrogate sensor signals and does not provide real-time glucose values .
• Requires consistent wear: Like all passive-tracking features, accuracy depends on regular sensor data collection .
• No additional hardware needed: All features rely on sensors already in the devices — no blood pressure cuffs or glucose sensors required .

Insulin Resistance Trends: An Industry First

The headline feature of Health Guardian is Insulin Resistance Trends, the first time a consumer wearable has attempted to estimate metabolic burden over time .

Unlike continuous glucose monitors (CGMs) that measure blood sugar directly, this feature uses passive wrist sensors — accelerometer, barometer, heart rate, and skin temperature — combined with AI to estimate monthly shifts in how hard the body works to produce energy from food . It does not provide real-time glucose readings or blood sugar levels . Instead, users receive a monthly summary in the Google Health app that flags notable changes, potentially prompting lifestyle adjustments or a doctor visit .

Clinical Validation

The feature is built on research from the WEAR-ME study, a peer-reviewed study published in Nature in March 2026 . The study enrolled 1,165 remote participants across the US who wore Fitbit or Pixel devices . By combining wearable data with routine blood biomarkers (fasting glucose, lipid panel), the deep neural network model predicted insulin resistance with an auROC of 0.80, 76% sensitivity, and 84% specificity at a HOMA-IR cut-off of 2.9 . The model performed better when using both wearable data and blood biomarkers than either alone .

This validation underpins the consumer feature, but Google emphasizes that the tool is designed for trend awareness, not diagnosis .

Blood Pressure Trends

Blood Pressure Trends is a monthly trend summary that estimates vascular stress without requiring a traditional blood pressure cuff . The watch passively observes pulse and movement patterns using optical blood flow and motion sensors, analyzed by a multimodal AI model .

Google tested the feature against 24-hour blood pressure monitors, using an advanced waveform model trained on more than 1 billion minutes of data from over 500,000 participants . Like insulin resistance, this is not a continuous measurement — it provides a single contextualized insight on the first of each month . In Japan and Mexico, the feature is referred to as "Vascular Stress / Vascular Stress Trends" .

Sleep Breathing Quality Trends

Sleep Breathing Quality Trends delivers a monthly report tracking changes in breathing patterns during sleep . Google explicitly states this is not sleep apnea detection or a diagnostic tool — it is a wellness-oriented awareness feature that surfaces shifts in respiratory patterns over time .

Breathing Emergency Detection

The most safety-critical feature in the suite is Breathing Emergency Detection, an opt-in, industry-first capability on supported Pixel Watches . It monitors for a critical, sustained drop in blood oxygen saturation (SpO2) — for instance, from drug toxicity, severe pneumonia, or choking . If the user becomes unresponsive, the watch automatically calls emergency services and pre-registered contacts .

Crucially, this feature is initially available only in EU countries where it carries a CE mark . It has not received FDA clearance in the US at launch . Google also notes the feature is not intended for people with certain pre-existing respiratory conditions and may not detect all breathing emergencies .

Availability and Rollout

Availability varies by region/country and device . Google recommends consistent wear for accurate passive tracking, as first detailed summaries arrive after about one month of use .

Key Limitations to Understand

• Not diagnostic: Google does not present any Health Guardian feature as a medical diagnostic tool. They are trend-awareness features meant to surface shifts, not confirm conditions .
• No FDA clearance for US: Breathing Emergency Detection lacks FDA clearance in the US at launch . The FDA status of the trend features (insulin resistance, blood pressure) is not detailed — they ship as wellness features, not regulated medical devices .
• Insulin resistance is inferred, not measured: The feature estimates trends using surrogate sensor signals and does not provide real-time glucose values .
• Requires consistent wear: Like all passive-tracking features, accuracy depends on regular sensor data collection .
• No additional hardware needed: All features rely on sensors already in the devices — no blood pressure cuffs or glucose sensors required .