Nearly one in three US adults sleeps less than seven hours a night, and most of the conditions behind those nights go undiagnosed. Your patients already know they sleep badly. They bought the ring, the mattress cover, and the magnesium. Some of them tape their mouths shut. What they cannot do is tell you whether any of it works, or whether the real problem is a disease their wearable was never built to catch.
A readiness score tells a patient they slept poorly. It does not tell them why. Circadian misalignment, an undiagnosed apnea, a late workout, a 10mg melatonin gummy that fragmented the back half of the night. Each one produces a low score that looks identical on the app. Reading them apart is clinical work. And clinical work is exactly what the consumer sleep market skips.
Why a better sleep score is not the goal
Sleep is now a consumer category. The sleep technology market reached about $25 billion in 2024, and wearables make up roughly three-quarters of it. Rings, watches, and smart mattresses all promise a better night. Almost none of them are built to find what's actually wrong. A low score is simply another symptom. Until you know what is driving it, optimizing the score is guesswork.
The CDC found that 12.9% of adults use a sleep aid most nights, split across prescriptions, over-the-counter supplements, and cannabis. Meanwhile around 30 million US adults have obstructive sleep apnea, and roughly 80% remain undiagnosed. A 2024 estimate puts the prevalence higher still, at 83.7 million US adults, about a third of the adult population. These people do not need a better sleep score. They need a diagnosis. The optimization economy is not built to give them a diagnosis, but you are.
The clinical risk hiding in the readiness score
Let's start with what a low readiness score is actually telling you. It could be circadian drift, an undiagnosed apnea, or something the patient tried last week that made things worse. One trend that has been coming up in patient conversations lately is mouth taping. Mouth taping went viral a few years ago as a fix for snoring and mouth breathing. A 2025 systematic review looked at 10 studies covering 213 patients and found mouth taping had minimal benefits and a real risk of asphyxiation, concentrated in exactly the people most likely to try it: those with undiagnosed obstructive sleep apnea. Taping shut the mouth of a patient who already struggles to move air at night is not a biohack. Before anyone tapes anything, rule out apnea.
Supplements deserve the same scrutiny. Melatonin is the default sleep aid, and most of what your patients buy is not what the label claims. A JAMA analysis found 22 of 25 melatonin gummies were inaccurately labeled, ranging from 74% to 347% of the stated dose, with one product containing no melatonin at all. Pediatric poison control calls for melatonin rose 530% between 2012 and 2021. Magnesium has a cleaner profile and modest evidence. A 2024 trial of magnesium bisglycinate found improved sleep quality at 28 days, and a 2021 meta analysis found patients fell asleep about 17 minutes faster, though on low quality evidence.
A patient can chase a better number with tape, supplements, and a cooler mattress for a year while moderate apnea goes untreated. Rule out apnea first, then optimize.
Reading sleep across your whole panel
The newest clinical AI does not just display a sleep score. It reads the streams underneath it and tells you which patients need a closer look. Three streams matter:
The circadian stream. Duration is the wrong target. A 2024 analysis of more than 60,000 UK Biobank participants found that sleep regularity predicted all-cause mortality more strongly than sleep duration, with the most regular sleepers showing 20 to 48% lower mortality risk. Regularity is also easier to change than total hours. Wearables already capture the raw material: bedtime, wake time, sleep midpoint, and overnight temperature. Upload a patient's wearable export to a clinical AI tool and it surfaces patterns a single score hides. One common example: a patient whose sleep midpoint swings two hours across the week and whose body never settles into a rhythm.
The apnea stream. This is where wearables earned real clinical weight. In February 2024 the FDA cleared Samsung's Galaxy Watch sleep apnea feature, the first of its kind, for moderate to severe OSA. That September it cleared Apple's sleep apnea notification, which catches roughly two-thirds of cases. These are screening signals that give you a direction. The play is to treat a positive notification as a prompt. Pull the patient in, order a validated home sleep test, and stop the optimization spiral before it buries a treatable disease. AI lets you watch the whole panel for these flags instead of waiting for one patient to mention theirs.
The response stream. Once a patient starts an intervention, the data tells you whether it works. Did the magnesium move sleep onset? Did dropping the late workout lift deep sleep? Did the melatonin dose actually fragment the back half of the night? Feed the before and after into a clinical AI tool and the answer stops being a patient's vague recap and becomes a trend you can read.
Where this is heading is bigger than any single score. Stanford researchers just built SleepFM, a foundation model trained on 585,000 hours of sleep data from 65,000 patients. SleepFM predicted more than 100 future conditions from a single night, including Parkinson's, dementia, and heart disease, several with a concordance index above 0.8. Be honest about the gap. It runs on full polysomnography, not an Oura export, and it is research grade today. It signals where sleep data is going, not a tool in your stack this week. The SleepFM team is already working to fold in wearable data. Plan for the panel you will read soon.
Before you trust the output
AI can invent a trend that is not there. A wearable can flag an apnea a home test rules out. Treat every output as a draft you verify.
Before you upload any patient data, confirm three things: the vendor has a signed BAA, you know their data retention policy, and the tool runs on a compliant deployment.
What to try this week
Pick one play and run it with five patients.
Build a sleep intake that reads the wearable. Ask five patients to export two weeks of ring or watch data and a short symptom log. Run it through a clinical AI tool or a custom GPT that flags three things: circadian drift (a swinging sleep midpoint), an apnea signal (a notification or heavy overnight desaturation), and intervention response (whether their supplement or tape changed anything). You walk into the visit already knowing which of the three is the real problem.
Screen before you optimize. For every patient asking about mouth tape, melatonin, or a fancier mattress, run a 60-second apnea risk check first. STOP-Bang answers plus their wearable's apnea history, summarized by an AI tool. The patient who screens high gets a sleep test.
The patient wearing a $300 ring who still sleeps badly is not unreachable. A short message offering to actually interpret their data is the easiest sleep visit you will book this month.
Be a modern clinician with the help of Ultralight, the AI-native EHR built specifically for functional, integrative, and longevity medicine.
In the news
UpDoc Becomes First FDA-Cleared Clinical AI Platform. UpDoc announced FDA clearance this week for what it calls the first agentic clinical AI platform cleared for real-time patient care delivery and care coordination. The clearance sets a regulatory precedent for AI that actively coordinates care, and signals the FDA is ready to move on this category. Watch this space.
The FDA just blinked on wearable blood pressure. In July 2025, the FDA warned Whoop for releasing a Blood Pressure Insights feature without clearance, calling it a medical device. Whoop pushed back, calling it a wellness feature. Last week the FDA dropped enforcement after Whoop made minor tweaks. The line between wellness tool and medical device just got blurrier, and your patients are on the other side of it asking what their numbers mean.
Researchers just found a gene that may explain why some families stay healthier for decades longer. A study of long-lived families presented at the European Society of Human Genetics identified a rare mutation in the CGAS gene that appears to temper chronic inflammation without compromising the immune system. Middle-aged adults from these families developed cardiometabolic diseases an average of 13 years later than their peers. The finding is early and needs replication, but it points to inflammation control as one of the most promising levers in longevity medicine.
ARPA-H just funded a program to build closed-loop sleep technology for the home. The REST program aims to develop in-home systems that measure sleep physiology with clinical fidelity and adapt interventions in real time across the night. The target: lift insomnia treatment response from roughly 50 percent today to at least 90 percent. Sleep is being reclassified from a wellness metric to a disease target.
Upcoming Conferences & Events
Oct 8–10, A4M Women's Health Summit · San Antonio, TX · The best clinical education on hormone, metabolic, and midlife women's health you will see this year. The room to be in if you are growing the perimenopause and menopause side of your practice.
Oct 21–24, NAMS Annual Meeting · San Diego, CA · The single most practice-changing meeting of the year for midlife women's health. Your protocols will look different after this one.
Nov 5–8, Eudēmonia Summit · West Palm Beach, FL · One of the most talked-about longevity gatherings in the U.S. Experientials, hands-on demos, and the best place to try the emerging frameworks your patients will ask you about next year. Ovation and Ultralight team will be there!
Nov 5-7 — Private Physicians Alliance Annual Meeting · St. Petersburg, FL The gathering for independent, cash-pay, and concierge physicians navigating practice independence. Practical and peer-driven. Ultralight will be there!
Nov 8-11 — American College of Lifestyle Medicine Conference · Orlando, FL Lifestyle medicine's main annual event — evidence-based approaches to behavior change, chronic disease, and healthspan. Growing overlap with the longevity medicine community.
Dec 11–13, A4M Longevity Fest · Las Vegas, NV · The biggest longevity event in the U.S. The room spans clinicians, industry, founders, and the people building next year's platforms, and the connections from this one tend to compound through the rest of your year. Ultralight will be there!
Know of an event we should add? Reply and tell us.
Until next week
Pick one patient this week. Pull their wearable export, screen for apnea, and find the one thing worth acting on. Reply and tell us what you found, or what got in the way. Browse past issues in the archive if you missed one.
Until next week, keep building the practice you imagined when you started. We are building it with you.
— Sunita and Dr. G