The Week in Blood Pressure & Wearables — September 25-October 1, 2026
By Yair Lurie, Founder and CEO, Cardiowell. Published October 2, 2026.
On October 1, Pixel Watch owners got their first monthly blood pressure trend reports from Google. Some got a "High," and some had normal readings on a home cuff. That's this week's deep dive.
Also this week: a large insurer said it plans to copy Medicare's ACCESS Model for commercial plans, Oura shelved its IPO, Sky Labs heads to the AHA Hypertension meeting, and a new study shows home blood pressure runs higher in winter.
Deep dive: Google's first blood pressure trend reports, and what a clinic should do with a "High"
What Google launched
On September 24, Google switched on "blood pressure trends" in the Google Health app for Pixel Watch 3, 4 and 5, per Google's announcement. It's free on those watches. Fitbit Air users will get it later this year with a paid Google Health Premium subscription.
Here is how it works, based on Google's own feature page. The watch uses its optical and motion sensors to record the pulse wave at the wrist, including during sleep. A Google AI model called WavesFM analyzes the pulse's shape and speed. No cuff and no calibration step.
The output is a monthly trend report that tells the user whether their trend looks persistently high. The user doesn't see a systolic or diastolic number. The report needs at least five consecutive days of wear in a week, according to 9to5Google.
Google says the model was validated in everyday conditions against 24-hour ambulatory blood pressure monitoring (ABPM), the reference test. The page I reviewed gives no participant numbers or accuracy figures.
Google is clear about the limits. The feature is "not a prescreener for hypertension," is "not intended for medical use," and "should not be relied upon for blood pressure management, medication adjustments, or treatment plans." It isn't for people who are pregnant. Google positions it as a general wellness feature, and I found no mention of FDA clearance. That fits the FDA wellness guidance from January that we covered last week.
What happened on October 1
The first reports arrived this week. Gadgets & Wearables reported that some users received "High" classifications for blood pressure, insulin resistance, or both. Some questioned a high blood pressure trend after getting more reassuring numbers from a regular cuff. Others got a high insulin resistance flag right after normal glucose or A1C results.
One Reddit user quoted in the piece pointed out that the report doesn't give "a conventional reading such as 140/90." That's by design, and it's the source of the confusion.
How hard cuffless accuracy still is
The same week, Amsterdam UMC researchers published a validation study in the European Journal of Applied Physiology on a different cuffless method. It estimates blood pressure from pulse timing measured with ECG and impedance sensors, and compares it with a cuff.
In 41 healthy young adults, the estimates were repeatable but inaccurate. Even after calibrating to each person, the average error was 10.1 to 12.9 mmHg, and only 59% to 68% of estimates fell within 10 mmHg of the cuff. The method missed blood pressure changes during mental stress. The authors concluded it isn't valid enough for clinical or stress research use.
This study didn't test Google's method, so it says nothing directly about Pixel Watch. It does show how hard the problem is, even with calibration.
A different tier: cuffless as a regulated medical device
Sky Labs' CART BP Pro sits in another category. It's a ring that uses an optical (PPG) sensor and a deep learning model for 24-hour cuffless monitoring, per Sky Labs' September 7 release. It received Korean medical device approval in 2023, and Korea's national insurance reimburses it under the existing 24-hour ABPM code. The company reports more than 300,000 prescriptions.
At ESC Congress 2026, Sky Labs presented data in patients with severe heart failure (mean ejection fraction 29%), where the ring met the European Society of Hypertension error criteria against standard 24-hour ABPM. In atrial fibrillation patients, about 80% of the data was usable, as summarized in a LinkedIn post Sky Labs' CEO shared. Sky Labs' materials I reviewed don't mention FDA clearance.
Disclosure: Cardiowell has an integration relationship with Sky Labs.
My view as a founder who has been through 510(k)s
I think Google made a responsible design choice. A monthly trend with no number is harder to misuse than a daily systolic reading on a watch face. Users will still get a "High," though, and some of them will call their doctor.
That call matters for clinics. A flag from a wellness feature should lead to a validated cuff and a proper home measurement schedule. Hypertension guidelines already call for confirming with readings taken outside the office. A J-HOP analysis published this week of 3,353 people found a one-week home monitoring schedule tracked a two-week schedule closely. One week of good home readings gives a clinic something it can act on.
Our BT107 blood pressure monitor is FDA 510(k) cleared. In general, a 510(k) requires showing that a device is substantially equivalent to one already legally on the market. For a blood pressure monitor, that typically means clinical validation against reference measurements under a recognized protocol, and labeling limited to what the evidence supports.
Home blood pressure also runs higher in winter. A Japanese study of 7,908 people, also published this week, found home systolic pressure was about 3 to 4 mmHg higher in winter than summer. Google's first reports cover September. I'd guess January reports will flag more people, and clinics should expect more calls then.
There's good news here too. Last week's Oxford study found 59% of adults with hypertension in a 1.4 million-person English cohort were undiagnosed. A passive flag on a smartwatch could reach some of those people. Some flags will likely be false alarms. Clinics need a confirmation step ready for when patients bring them in.
My suggestion: if a patient brings in a wearable flag, treat it as a reason to start a week of validated home readings. Make the decision from those.
Devices & Wearables
Oura shelves its IPO
On September 29, Oura postponed its planned IPO of up to $2.2 billion, citing uncertainty in the IPO market, TechCrunch reports. The company has 5.7 million paying members and reported $907.9 million in 2025 revenue.
In The Device Files, Blythe Karow reads it as a valuation and timing decision. She suggests investors tend to understand either consumer subscription wearables or regulated medtech, rarely both. That gap matters for every company selling health data from a consumer device.
Apple is testing a screenless fitness band
Apple has built prototypes of a display-free fabric band with a small sensor module for round-the-clock health tracking, MacRumors reports, citing Bloomberg. Apple hasn't decided whether to make it, and a launch would likely be 2028 at the earliest. The reports don't mention blood pressure.
Leonard Rinser's widely shared LinkedIn post framed it as a distribution threat to Whoop and Oura, proving people will wear a screenless tracker and pay a subscription for it. It joins a growing list of screenless trackers, from Whoop and Oura to Google's Fitbit Air.
Sky Labs lists on KOSDAQ and heads to AHA Hypertension
Sky Labs, maker of the CART ring, listed on Korea's KOSDAQ market on September 4, per its LinkedIn page. The company is exhibiting at the AHA Hypertension Scientific Sessions next week in Arlington, Virginia (Booth 9). It's also sponsoring the meeting's first Established Investigator Award for Clinical Hypertension, which went to Dr. Annet Kirabo.
Sky Labs is also hosting a "Beyond the Cuff" luncheon seminar at the Japanese Society of Hypertension meeting on October 11.
Policy & Regulation
Blue Shield of California plans an ACCESS-style model for commercial members
Blue Shield of California, with 6 million members, is one of 17 major health plans that have committed to payment models based on Medicare's ACCESS Model, STAT reports. It plans comparable commercial offerings in 2027. Chief Medical Officer Ravi Kavasery said payment levels and partners are still being decided.
ACCESS pays technology-enabled care organizations based on patient outcomes, and hypertension is part of its early cardio-kidney-metabolic track. If commercial plans follow, outcome-based payment for hypertension care could reach far more patients than Medicare alone.
Employers and health plans are tying digital health payments to outcomes
A Peterson Health Technology Institute survey of 321 buyers found 68% of employers and 55% of health plans use performance-based contracts for digital health, Fierce Healthcare reports. In 85% of those contracts, at least 25% of fees are at risk. Thirty-eight percent of buyers said they can't independently verify the outcomes.
The 38% matters. Paying for outcomes only works if both sides trust the measurement.
Rural Health Transformation grants name remote patient monitoring
CMS announced $169.6 million in Rural Health Transformation Program awards for Colorado, covering 91 grantees and about 250 projects, per CMS. The release names remote patient monitoring along with ICU telemedicine and mobile clinics.
Texas received $51 million for rural chronic disease and nutrition programs, including cardiovascular disease, per CMS. North Carolina received $20 million for telehealth and health IT, per CMS.
37 states sign on to outcome-based Medicaid quality measures
CMS launched a voluntary Medicaid quality pledge with 37 states, shifting reporting from process measures toward health outcomes, with prevention and chronic disease management among the priorities, per CMS. Blood pressure isn't named specifically yet.
Clinical & Research
Home blood pressure is 3 to 4 mmHg higher in winter
Tohoku University researchers analyzed home readings from 7,908 people synced through a personal health record app, and published the findings in Hypertension Research. Home systolic pressure peaked in December and January. Once researchers accounted for the higher winter average, day-to-day variability barely changed by season, suggesting variability is a stable trait for each person.
One week of home readings tracked two weeks closely
In 3,353 participants from Japan's J-HOP study, a home blood pressure "stability score" calculated from either week of a two-week schedule agreed closely with the full 14-day score, per Hypertension Research. The authors say this supports using a standard one-week home monitoring schedule.
Telemonitoring is moving past blood pressure alone
A mini review by Tomitani and Kario in Hypertension Research describes hypertension telemonitoring that combines home blood pressure with activity, sleep, medication adherence and environmental data. It lists open problems: data quality standards, clinical workflow, patient engagement, digital inequity, and privacy.
Blood pressure swings over time are linked to death and stroke
In 57,611 UK Biobank participants, each standard deviation increase in systolic variability was associated with an 11% higher risk of death, and diastolic variability with a 12% higher risk of stroke, News-Medical reports. People with type 2 diabetes and high systolic variability had 2.8 times the risk of death. The analysis is observational and was presented at the EASD meeting in Milan.
Wall squats once a week kept blood pressure down
In a trial of 100 adults published in BMJ Open Sport & Exercise Medicine, four weeks of wall-squat training lowered blood pressure, and one session a week was enough to keep most of the reduction, Healthline reports. Three sessions a week kept the largest drop. People who stopped returned to baseline. The maintenance phase lasted only four weeks.
A case report on wearable data before a sudden cardiac death
Stanford Medicine described a JACC: Case Reports paper on a 76-year-old man with known heart disease whose Apple Watch and Oura Ring data changed in the months before his sudden cardiac death, per Stanford. According to a summary in Physiologically Speaking, his nighttime heart rate variability fell from 180 to 63 milliseconds and his deep sleep fell by more than half. It's one retrospective case, and the authors say it can't show that wearables predict sudden death.
Also worth a glance
Virtual reality and music in the emergency room. In a 130-patient trial in Ankara, VR with classical music lowered mean arterial pressure 12% at 15 minutes in hypertensive urgency, compared with 8% in controls, News-Medical reports. The difference was gone by 60 minutes.
The thread this week
Most of this week's stories depend on what happens after a sensor flags something.
Google is sending blood pressure trend flags to Pixel Watch owners. Sky Labs is taking a regulated cuffless ring to the US hypertension community. Insurers and employers are moving money toward outcomes, and more than a third of buyers still can't verify them.
Each of these depends on a confirmation step someone can trust: a validated reading, a defined schedule, and a care team that acts on it. The sensors are getting better quickly. I'm optimistic that the workflows will catch up.
References
Our new health and safety tools are live in the Google Health app. Google.
Blood Pressure. Google Health.
Google rolling out blood pressure & insulin resistance, more on Pixel Watch 3, 4, 5. 9to5Google.
Not FDA's Whoop-ing Boy: How a Warning Letter Became a Wellness Playbook. ArentFox Schiff.
Pixel Watch Health Guardian is freaking out some users on its first reporting day. Gadgets & Wearables.
Validation study on cuffless ambulatory blood pressure monitoring method using bio-impedance, impedance cardiography and electrocardiography. European Journal of Applied Physiology.
CART Blood Pressure Monitor Demonstrates Utility in BP Monitoring for AF and Heart Failure Patients. PR Newswire.
Sky Labs Inc. Advances Cuffless 24-Hour BP Monitoring in AF and Heart Failure. Jack Shuang Hou on LinkedIn, reposted by Sky Labs CEO Jack (Byunghwan) Lee.
Agreement between seven-day and fourteen-day home systolic blood pressure stability scores: A J-HOP study analysis. Hypertension Research.
Seasonal variation in day-to-day blood pressure variability: large-scale within-individual home blood pressure data. Hypertension Research.
Two-Thirds of Adults With High Blood Pressure Do Not Know They Have It. University of Oxford.
Oura shelves its $2.2B IPO, citing uncertainty in the market. TechCrunch.
Oura Pumps the Brakes on Its IPO. The Device Files (Blythe Karow).
Apple Developing Screenless Fitness Band. MacRumors.
More than 100 million people were wearing an Apple Watch by the end of 2020. Leonard Rinser on LinkedIn.
Sky Labs on LinkedIn: KOSDAQ listing, AHA Hypertension 2026 and JSH 2026. Sky Labs on LinkedIn.
A major insurer on how it may replicate Medicare's chronic-care experiment. STAT.
Employers increasingly tie digital health payments to outcomes, PHTI survey finds. Fierce Healthcare.
Trump Administration Announces $169 Million to Expand Specialty Care, Strengthen Emergency Services, and Bring Lifesaving Blood Transfusions Closer to Colorado Patients. Centers for Medicare & Medicaid Services.
Trump Administration Announces $51 Million for Rural Texas Nutrition and Chronic Disease Prevention Programs. Centers for Medicare & Medicaid Services.
Trump Administration Announces $20 Million to Expand Telehealth Access and Upgrade Healthcare Technology. Centers for Medicare & Medicaid Services.
CMS Refocuses Medicaid Quality on Health Outcomes, Launches Innovative Partnership With 37 States. Centers for Medicare & Medicaid Services.
Telemonitoring in the era of digital hypertension: beyond remote blood pressure monitoring. Hypertension Research.
Greater variation in blood pressure associated with higher risk of death and strokes. News-Medical.
High Blood Pressure: Isometric Exercise Once a Week May Help. Healthline.
Researcher's sudden death shows wearables' potential as 'check-engine light'. Stanford Medicine.
Physiology Friday #337: Can Your Wearable Predict When (or How) You'll Die?. Physiologically Speaking.
Virtual reality and music may help lower blood pressure during emergencies. News-Medical.