Colin Rohrer
A patient presents for a routine primary care appointment. The nurse will go over the usual checklist, including temperature, blood pressure, pulse, weight, and sometimes pulse oximetry. You probably won’t be able to sleep. If so, it will be a secondary explanation, and if the patient says he or she is “not feeling very well,” this is often followed by a brief look of sympathy and the familiar advice to relax a little before going to bed.
In other words, here’s what sleep looks like in the most common diagnostic interactions in American medicine. Don’t worry. In other countries, not much is better. Other essentials are marked with numbers and sleep is marked with chatter. To call this a minor gap is to miss the point.
According to the evidence
Sleep is one of the strongest behavioral and physiological predictors of chronic disease, cognitive decline, mental health outcomes, and burnout.
A recent Stanford University study found that just one night of sleep data (from a hospital sleep laboratory) processed into a baseline model called SleepFM could flag increased risk with high accuracy across 130 disease categories. The list of outcomes is not trivial and includes all-cause mortality, dementia, myocardial infarction, and heart failure.
A 2025 umbrella review that integrated 29 systematic reviews found bidirectional physiologically mediated associations between sleep and depression and anxiety, as well as a long catalog of cardiometabolic diseases.
And researchers at Washington State University have published the longest, most objective description of sleep in chronic insomnia to date. Eight weeks of continuous at-home measurements point to something clinicians have struggled to capture for years. In other words, day-night fluctuations in sleep efficiency, sleep latency, and intermittent awakenings are at the core of this disorder. Sleep diaries and overnight laboratory studies continue to miss such patterns.
The clinical basis for measuring sleep is established, but what remains unclear is whether medicine intends to act as if it believes its own evidence.
Take a look at your current settings. Obstructive sleep apnea affects approximately 960 million people worldwide, and 80% of moderate to severe cases remain undiagnosed. Chronic insomnia affects more than 800 million people worldwide. Both disorders have costly and common downstream consequences, such as cardiovascular disease, depression, automobile crashes, workplace injuries, dementia, and more. Both are treatable. However, routine primary care typically does not screen for either.
The American College of Physicians has recommended cognitive behavioral therapy for insomnia as first-line treatment since 2016. Despite this, most people with chronic insomnia do not receive CBT-I at all. This is partly because it is not confirmed in the first place. Clinicians cannot treat what is not revealed, and they often do not even ask the questions that will surface.
The void filled by consumer technology
Working professionals, parents of young children, menopausal women, older adults, teenagers, etc., I talk to just about everyone and they all agree that they fall asleep quickly. People know it’s important. They read about it, monitor it with a watch, and take it to the doctor. And when the clinical system has nowhere to put those concerns, they increasingly look elsewhere.
In over 16 years of working in sleep science and health technology, the biggest change I have observed is the change in what patients do when they run out of medication.
Consumer technology has moved into the space left open by health care. People measure their sleep—sometimes they sleep well and sometimes they sleep poorly—through wearables, phone apps, and bedside devices. Apple, Google and the broader consumer market have helped make sleep a “countable” feeling worth paying attention to. That’s real progress.
However, a problem arises in the next step. When a patient’s clock reports a steady decline in deep sleep over a six-month period, there is usually no clinical pathway to that signal. Most primary care practices are not designed to receive them. Doctors often have little training in interpreting them. Insurance companies rarely pay for the time and work required to conduct an investigation.
Data is available, but the machines that turn it into insights and management are missing.
So patients end up making their own interpretations, usually with mixed results, and often surrounded by wellness content that ranges from thoughtful to careless. The gap is not a consumer problem. This is for medical use.
What medicine should do
This example is more practical than you might think. Medicine doesn’t need to swallow the entire world of consumer wearables for it to take sleep seriously. You need to do four specific things:
First, incorporate validated sleep measures into routine primary care along with other essentials. At population scale, tools already exist for standardized wearable data aggregation benchmarked against smartphone-based measurements, clinical-grade bedside sensors, and polysomnography. Science is not a bottleneck. Includes reimbursement, workflow, and training.
Second, we continue to screen for the three most common and most underdiagnosed sleep disorders, obstructive sleep apnea, chronic insomnia, and restless legs syndrome, especially in high prevalence groups. Primary care is an obvious home for this, but so is obesity medicine, cardiology, endocrinology, mental health and women’s health. None of these areas perform reliably today.
Third, build functioning referral and treatment pathways. If sleep measurements indicate a clinical problem, the patient must have a place to go. This means higher doses of sleep medication, wider access to CBT-I, and closer collaboration between sleep specialists and the rest of the care team. Currently, this route often passes through too few sleep laboratories and even fewer sleep physicians or behavioral sleep clinicians, leaving patients waiting or not receiving care. Capacity needs to be increased.
Fourth, treat sleep data already collected from people as legitimate input. Tens of millions of Americans track their sleep every night. Data quality varies and interpretation is often uncertain. But when you add validated measurements and clinical context, the signal becomes much clearer. This is made easier with high-quality harmonization tools. When a patient walks into an appointment with months of data collected, the system is doing something it hasn’t officially asked anyone to do. Medicine must take this seriously.
Science has advanced enough. What remains is the operational task of sorting the wheat from the chaff, creating workflows, defending coverage, training clinicians, expanding capacity, and treating sleep with the same gravity we’ve given to other vital signs for a century.
Sleep is also a clear entry point into bigger questions. How should medicine harness the power of persistent physiological signals in everyday practice? Sleep is becoming easier to measure, it feels so important, it feels personal, and it’s the biggest gap between what we know and what we actually do.
If health care systems cannot figure out how to measure and respond to sleep that is universal to patients, intuitive, and supported by incontrovertible evidence, the larger prospects for preventive medicine based on physiological data appear shaky. We’re debating whether sleep is important. We’re already proving that technology can measure it. The remaining questions are both simpler and more difficult. Is medicine willing to treat sleep like a vital sign?
This one has ‘slept’ long enough.
Colin Lawlor is the founder and CEO. sleep.aiHe has spent more than a decade developing proven sleep measurement and intelligence technologies.
