Sleep apnea test: home study vs lab, and what the AHI number means

Short answer

A sleep apnea test is either a home sleep apnea test you wear in your own bed or an in-lab polysomnography study. Both count breathing events per hour of recording to produce your AHI. In adults, under 5 is normal, 5 to 14.9 is mild, 15 to 29.9 is moderate, and 30 or more is severe (1, 2).

If you have spent years being told you are just tired, you are not imagining it. Sleep apnea is one of the most under-recognized conditions in American medicine: a Frost & Sullivan analysis commissioned by the American Academy of Sleep Medicine (AASM) estimated that about 23.5 million US adults with obstructive sleep apnea — roughly 80% of everyone who has it — have never been diagnosed (3). Being exhausted at 3 p.m. is not laziness and not a character flaw, and the only way to settle the question is a test your doctor orders. This page explains the tests, the numbers they produce, and what a normal result does and does not rule out. Your doctor decides which test fits you.

Note: this article explains how sleep apnea is diagnosed and is not medical advice or a treatment guide; a clinician orders and interprets the test. Call 911 for chest pain or pressure, trouble breathing at rest, blue or gray lips or face, fainting, a seizure, a very irregular or racing heartbeat, or sudden weakness, numbness, slurred speech or a drooping face. Get urgent medical care the same day for falling asleep while driving or at work, sleepiness that has suddenly got much worse, morning headaches with confusion, or swelling in both legs with breathlessness when you lie flat (4, 5).

Sleep apnea test at a glance

Two tests, one number, and a rule about what happens when the number comes back low.

The single most useful line in the AASM guideline is the one people most often miss: a low number on a home test does not close the case. We come back to that below.

Home sleep apnea test or in-lab study: which one will you get?

It depends on how "uncomplicated" you are. The AASM's clinical practice guideline, published in the Journal of Clinical Sleep Medicine in 2017 and still the current US guideline while a focused update is in development, makes a STRONG recommendation that either polysomnography or a technically adequate home sleep apnea test be used "in uncomplicated adult patients presenting with signs and symptoms that indicate an increased risk of moderate to severe OSA" (1, 10). The VA/DoD clinical practice guideline for chronic insomnia disorder and obstructive sleep apnea, version 3.0, published in January 2025, says the same thing in plainer words: diagnose with polysomnography or home testing (Strong for), and home testing is a reasonable alternative in appropriate patients (Weak for) (11).

What counts as "increased risk of moderate to severe OSA"

The guideline is specific. It means excessive daytime sleepiness plus at least two of these three: habitual loud snoring; witnessed apnea, gasping or choking at night; or diagnosed high blood pressure (1). If that describes you and you have no complicating conditions, a home test is a legitimate first step. If it does not describe you but your clinician still suspects apnea, the guideline points to the lab.

For the symptom side of this — the signs that are obvious and the ones that are not, including why apnea is so often missed in women — see our guide to the signs of sleep apnea and obstructive sleep apnea symptoms.

When the guideline says you need the lab, not the kit

Recommendation 4 of the AASM guideline is STRONG: use polysomnography rather than a home test in patients with any of these (1):

• Significant cardiorespiratory disease — for example moderate to severe lung disease, or heart failure.

• Potential respiratory muscle weakness from a neuromuscular condition.

• Awake hypoventilation or suspected sleep-related hypoventilation, including obesity hypoventilation.

• Chronic opioid use.

• History of stroke.

• Severe insomnia.

Commercial payer policy follows the same logic and spells it out further. Aetna's clinical policy bulletin on obstructive sleep apnea in adults, last reviewed 6 October 2026, lists attended in-lab polysomnography as medically necessary for people with moderate to severe lung disease on nocturnal oxygen, neuromuscular disease, stroke with residual respiratory effects, epilepsy, heart failure of NYHA class III or IV or an ejection fraction below 45%, pulmonary hypertension, chronic opioid use, or a BMI above 45 — and also for periodic limb movement disorder, atypical parasomnias, severe insomnia, narcolepsy, central or complex sleep apnea, a negative or technically inadequate home test, a low pretest probability of apnea, or an inability to handle the home equipment safely (12).

The reason is mechanical, not bureaucratic. A home kit measures airflow, effort and oxygen. It cannot see a brain-wave arousal, cannot tell apnea from a leg-movement disorder, and cannot distinguish obstructive events (the airway collapses while you keep trying to breathe) from central events (the brain briefly stops sending the signal to breathe), which matter in heart failure, stroke and opioid use (4, 1).

What "technically adequate" means, and why it matters to you

A home test only counts if the recording holds up. The AASM defines a technically adequate home study as one that captures at least 4 hours of usable oximetry and airflow data, recorded across your habitual sleep period, on a device with the right sensors: nasal pressure plus chest and abdominal effort belts plus oximetry, or peripheral arterial tone with oximetry and actigraphy (1). Raw data must be reviewed by a qualified clinician, not just auto-scored — the guideline says the test is to be administered by an accredited sleep center under the supervision of a board-certified or board-eligible sleep medicine physician (1). If a cannula fell out at 2 a.m. and the device logged three hours, that is not a negative test. It is no test.

What happens in the lab

You arrive in the evening and a technologist applies the sensors, which takes 45 to 60 minutes. MedlinePlus lists what a full study records: breathing rate, blood oxygen, heart rate, brain waves, eye movements and leg movements (6). Practical preparation from the same source: no alcohol or caffeine in the afternoon and evening beforehand, no daytime nap, no lotion, hair gel, makeup, nail polish or artificial nails, and ask your clinician whether to take your usual sleep medication (6).

Some labs run a split-night study: if enough apnea shows up in the first part of the night, the technologist wakes you and fits a CPAP mask for the second part, so diagnosis and pressure setting happen in one visit. The AASM suggests this where clinically appropriate (a WEAK recommendation), with the condition that moderate to severe OSA is seen during at least 2 hours of recording and at least 3 hours remain for the titration (1). Treatment itself — CPAP and everything else — is a separate subject, covered in sleep apnea treatment options.

What does the AHI number actually mean?

AHI is the apnea-hypopnea index: the average number of breathing events per hour of sleep. An apnea is airflow essentially stopping for at least 10 seconds. A hypopnea is a partial collapse — airflow dropping substantially for at least 10 seconds — that either drops your blood oxygen or jolts your brain into a brief arousal (13, 14). Add the apneas and hypopneas, divide by hours of sleep, and that is your AHI.

What is a normal AHI in adults

AASM's own quality measure for severity assessment at initial diagnosis, Quality ID #277 version 9.0, dated December 2024, states the thresholds plainly (2):

Medicare's national coverage determination 240.4 turns those bands into a payment rule: CPAP is covered when AHI or RDI is 15 or more, or 5 to 14 with documented excessive daytime sleepiness, impaired cognition, mood disorder or insomnia, or documented hypertension, ischemic heart disease or a history of stroke (13). This is why the exact number on your report matters so much, and why two reports that sound similar can lead to two different conversations.

One number that is not on the list: AHI does not measure how tired you feel. Plenty of people with an AHI of 40 insist they sleep fine, and plenty of people with an AHI of 8 are wrecked. Severity on paper and severity in your life are two different axes.

AHI, RDI, REI and ODI: four numbers people mix up

Your report may carry several indices. They are not interchangeable.

RERAs are the quiet ones: your airway narrows, you work harder to breathe, and your brain surfaces briefly without the oxygen dropping enough to count as a hypopnea. They fragment sleep without showing up in the AHI. That is why the AASM says RDI matters, and recommends treating symptomatic patients when polysomnography shows an AHI of 5 or more or an RDI of 5 or more (14).

The 3% versus 4% rule that can change your diagnosis

Here is a detail almost nobody is told. The AASM scoring manual contains two ways to count a hypopnea. The recommended rule counts a drop in airflow accompanied by either a brain-wave arousal or a 3% oxygen desaturation. The acceptable rule counts only drops with a 4% desaturation and does not allow arousal-based scoring at all, and the AASM says it should be used only when payer policy demands it (14).

The same night, scored both ways, can produce two different AHIs and two different diagnoses. The AASM position statement, published in JCSM in 2018, says so directly: having two criteria "leads to differences among lab reports, patient treatment, and payer policies," and events that cause an arousal without a 4% desaturation can still cause daytime sleepiness, fatigue, insomnia and cognitive symptoms (14). The AASM's advice is that if a lab must use the 4% rule, it should also score the recommended hypopneas and report them. If your report shows an AHI just under the threshold that would unlock treatment, this is a fair and specific question to ask: which hypopnea rule was used, and what is my AHI under the other one?

Why can a home sleep apnea test come back normal when you still have apnea?

Because a home test is built to confirm apnea, not to rule it out — and the maths of the test are quietly stacked towards a lower number. Four separate things push a home result down.

The denominator problem

A home kit usually does not know when you fell asleep. It divides your events by the whole recording time, including the hour you spent awake reading and the stretch you lay staring at the ceiling. The AASM states this plainly: because sleep-staging channels are not typically monitored, the home test "uses the recording time rather than sleep time to define the denominator of the respiratory event index" (1). A bigger denominator means a smaller index for the same number of events.

How much smaller? A 2023 study in Sleep Advances took 1,561 polysomnography recordings from clinics in Australia and Israel and calculated both numbers from the same nights. Median REI was 8.4 against an AHI of 19.4 in one dataset and 12.1 against 17.1 in the other. Overall severity classification matched the AHI only 42.1% and 72.8% of the time. Most striking: 18.6% of patients who had moderate OSA by AHI were classified as having no OSA at all by REI, and 14.2% of those with severe OSA by AHI were graded mild (15). That is not a rounding error. That is a person going home with a clean bill of health and a collapsing airway.

Your apnea is not the same every night

Even a perfect test only sees one night. A 2020 systematic review and meta-analysis in Thorax pooled 24 studies and 3,250 participants who had more than one diagnostic sleep study. On average, 41% of participants had their event count shift by more than 10 per hour from one night to the next, and 49% changed severity category (16). A 2026 prospective study in npj Digital Medicine followed 92 people with suspected apnea for three months with a nightly in-home sensor alongside a single polysomnography night, and found that the people whose results disagreed between the two methods were precisely those with high night-to-night variability (17).

Sleeping position and REM sleep are a large part of this. Lie on your back and events multiply; spend a long stretch in REM and they multiply again. A night in an unfamiliar bed with wires on your face is not a typical night either, in either direction.

Sensors come off, and arousal-only events are invisible

The AASM lists the mechanical reasons a home study runs low: sensors dislodging, poor signal quality, and the inability of conventional home sensors to detect hypopneas that end in an arousal rather than a desaturation. "All these factors can result in the underestimation of the 'true' AHI, and may result in the need for repeated studies" (1). The AASM's scoring position statement adds the obvious corollary: because most home devices cannot record arousals, a patient at risk whose home test comes back negative should have a polysomnography (14).

The rule that protects you, if someone applies it

Recommendation 3 of the AASM guideline is STRONG and unambiguous: if a single home sleep apnea test is negative, inconclusive or technically inadequate, polysomnography should be performed (1). The guideline explicitly advises against simply repeating the home test, on the grounds that a second one is likely to be negative, inconclusive or technically inadequate again, and that people drop out of the diagnostic process along the way. A repeat is allowed only when the clinician expects it to succeed and the patient prefers it (1).

The VA/DoD guideline reaches the same destination by a slightly different route. Its Recommendation 4 (Strong for) says a non-diagnostic home test calls for further testing — either in-lab polysomnography or another home study — and it treats an event index under 5 on a home test as not ruling apnea out. It also flags the grey zone of 5 to 15 events per hour as a judgement call that should weigh symptoms, occupation and other conditions, and names 15 per hour as the cutoff it considers definitive on a home test (11).

Put together: a home test that finds apnea is trustworthy. A home test that does not find apnea, in someone with real symptoms, is an unfinished sentence.

What STOP-BANG and Epworth actually tell you before the test

They tell your clinician how worried to be. They do not tell you whether you have sleep apnea. The first recommendation in the AASM guideline is a STRONG one, and it is a warning: clinical tools, questionnaires and prediction algorithms "should not be used to diagnose OSA in adults, in the absence of polysomnography or home sleep apnea testing" (1). The guideline reviewed the Berlin Questionnaire, the Epworth Sleepiness Scale, STOP and STOP-BANG and found their specificity low and their false negatives too many to rule anything out.

STOP-BANG: eight yes-or-no questions

STOP-BANG is an acronym. One point for each yes (18):

• S — Snoring — Do you snore loudly, loud enough to be heard through a closed door?

• T — Tired — Do you often feel tired, fatigued or sleepy during the day?

• O — Observed — Has anyone observed you stop breathing, choke or gasp during sleep?

• P — Pressure — Do you have, or are you treated for, high blood pressure?

• B — BMI — Is your BMI over 35?

• A — Age — Are you over 50?

• N — Neck — Is your neck circumference over 16 inches (women) or 17 inches (men)?

• G — Gender — Are you male?

A 2021 systematic review and meta-analysis in JAMA Network Open pooled 47 studies and 26,547 sleep-clinic patients. A STOP-BANG score of 3 or more had sensitivity above 90% and was good at excluding moderate-to-severe and severe apnea, with negative predictive values of 77% and 91% respectively; diagnostic accuracy for moderate-to-severe apnea was above 0.80 in every region studied except East Asia, where it fell to 0.52 (18). Note what that means: a high score raises suspicion, and a low score lowers it, but neither is a result. The VA/DoD guideline picked the shorter four-item STOP for routine screening because it performs similarly and is quicker to administer, and reported pooled sensitivities of 88% for STOP-BANG and 87% for STOP against 54% for the Epworth scale (11).

The Epworth Sleepiness Scale measures sleepiness, not apnea

The Epworth scale, introduced by Murray Johns in Sleep in 1991, asks how likely you are to doze off in eight everyday situations — reading, watching TV, sitting in a meeting, stopped in traffic — scored 0 to 3 each, for a total out of 24 (19). It is widely used, and in the VA/DoD pooled analysis it caught only about half the people who had apnea (11). Plenty of people with severe apnea score low on it, particularly women, whose symptoms more often present as fatigue and insomnia than as frank sleepiness. A low Epworth score is not a reason to skip the test if everything else points at apnea.

Insurers do use these numbers, though. Aetna's policy, for example, lists an Epworth score above 10 among the qualifying symptoms that let someone with an AHI in the 5 to 14 range get CPAP covered (12). So filling the questionnaires in honestly, rather than minimising, is worth doing.

How much does a sleep study cost, and what does insurance cover?

The home test is the cheap one, by a wide margin, and almost everyone with symptoms is covered for something. Medicare's own 2024 claims data, published in the CMS Medicare Physician & Other Practitioners — by Geography and Service dataset, gives the national averages for what clinicians billed and what Medicare allowed (9):

• In-lab diagnostic sleep study (95810) — Average submitted charge: $1,631; Average Medicare allowed: $432

• In-lab study with CPAP titration or split-night (95811) — Average submitted charge: $1,716; Average Medicare allowed: $446

• Home sleep apnea test, airflow and effort (95806) — Average submitted charge: $387; Average Medicare allowed: $64

• Home sleep apnea test with sleep time, e.g. PAT devices (95800) — Average submitted charge: $469; Average Medicare allowed: $125

Two caveats on those figures. They are national averages for the office or independent-facility setting, and they cover what the physician or testing facility bills under Part B — a hospital outpatient sleep lab bills a separate facility fee on top, which is why a hospital-based study can land far higher on a commercial plan. And what you actually pay depends on your plan, not on these numbers: Medicare.gov says Part B covers Type I, II, III and IV sleep tests when you have clinical signs and symptoms of sleep apnea and your doctor orders the test, that you pay the Part B deductible and then 20% of the Medicare-approved amount, and that Type I tests are covered only in a sleep lab facility (20).

On commercial insurance the practical pattern is this: a home test usually sails through, an in-lab study often needs prior authorization and a documented reason. The reasons that work are the ones in the guideline and in the payer's own list — heart or lung disease, neuromuscular disease, stroke, opioid use, severe insomnia, suspected central apnea, a BMI above 45, or a previous home test that was negative or technically inadequate (12). If your clinician writes one of those down, the lab study is much easier to get approved. Aetna also states that repeat home testing on multiple consecutive nights has no proven value, which is a useful thing to know if you are offered a second kit instead of a lab night (12).

Three practical questions to ask before you book: is this billed as a hospital outpatient service or an independent facility; will the interpretation be billed separately; and does my plan require a home test first. Direct-to-consumer home kits sold without a clinician's order sit outside all of this — they are usually not covered, and the AASM's position is that testing should be administered by an accredited sleep center under a sleep physician's supervision (1).

What about the sleep apnea alert on a smartwatch?

Treat it as a nudge, not a number. Several smartwatches now carry FDA-authorized features that flag possible signs of moderate-to-severe sleep apnea over a month of nights. The AASM, reporting the first such authorization in February 2024, notes the manufacturer's own framing: the feature "is not intended to replace diagnosis and treatment by a clinician," is not meant to help clinicians diagnose sleep disorders, and is not for people already diagnosed with apnea (21).

The AASM's broader position statement on consumer sleep technology, published in JCSM in 2018, is blunter still: these devices "cannot be utilized for the diagnosis and/or treatment of sleep disorders at this time," though they "may be utilized to enhance the patient-clinician interaction" when used inside a proper clinical evaluation (22). That is the honest ceiling. A watch can tell you something is worth checking. It cannot tell you your AHI, and a quiet month on a watch is not a negative test. For what trackers can and cannot see at night, see the best sleep trackers and sleep stages explained.

Sleep apnea test for children: why the home kit usually is not it

Children get the lab. The AASM published a dedicated position paper on this in JCSM in October 2017, and the position statement is one sentence: "Use of a home sleep apnea test is not recommended for the diagnosis of obstructive sleep apnea in children" (7). The American Academy of Pediatrics clinical practice guideline on childhood obstructive sleep apnea, published in Pediatrics in 2012, recommends that all children and adolescents be screened for snoring, and that polysomnography be performed in those with snoring plus symptoms or signs of apnea; if polysomnography is not available, it allows alternative diagnostic tests or referral to a specialist for a more extensive evaluation (8).

The numbers are different too, which matters if you are reading a child's report with adult thresholds in your head. In children, AASM criteria set the diagnostic threshold at an obstructive AHI of 1 or more events per hour — a fifth of the adult cutoff — and severity grading commonly runs: under 5 mild, 5 to 9.9 moderate, 10 or more severe (23). A child with an AHI of 6 is not "almost normal." A pediatric study also measures carbon dioxide to catch obstructive hypoventilation, which home kits do not do (23).

The AASM has a task force working on an update to the pediatric home-testing position paper, so this may change (10). Until it does, a home kit on a child is off-label by guideline, and a negative one tells you very little. Your child's pediatrician or a pediatric sleep specialist decides.

Living with suspected sleep apnea: the questions people ask next

My home test said AHI 4.2. Is that normal?

By the adult cutoff, yes — under 5 is normal (2). But remember what that 4.2 is: a respiratory event index over your whole recording time, from a device that cannot see arousal-only events, on one night. The AASM's rule for a negative home test in someone with symptoms is polysomnography, not reassurance (1). If you still wake unrefreshed, ask what the next step is rather than accepting the number as the end of it.

I am thin and I do not snore. Can I still have it?

Yes. BMI over 35 and loud snoring are two of eight STOP-BANG points, not prerequisites (18). Jaw and airway shape, nasal obstruction, enlarged tonsils and menopause all raise risk independently of weight, and Aetna's own policy treats a low pretest probability — normal BMI, no snoring, small neck — as a reason to go to the lab rather than the home kit, precisely because a home test performs poorly in that group (12).

Do I need to sleep badly on test night for it to count?

No, but you do need to sleep. A home test needs at least four hours of usable airflow and oximetry data over your habitual sleep period (1). Keep your normal bedtime, skip the alcohol — it relaxes the airway and can inflate the result — skip the afternoon caffeine and the daytime nap, and do not take a sleeping pill unless your clinician told you to (6). If you sleep only on your side out of habit, mention it: position changes the count a lot.

How long until I get results?

Days to a few weeks. The raw recording has to be downloaded, scored and interpreted by a sleep physician before anyone can tell you a number (1). Ask at the time of the test who will call you, in what timeframe, and whether you will get the full report rather than a one-line summary. Ask for the report in writing — the AHI, the RDI if there is one, the ODI, the time spent supine, and which hypopnea rule was used.

Why do I keep waking up with a headache?

Morning headache is on the standard list of apnea symptoms, but it is far from specific — dehydration, teeth grinding, medication overuse, caffeine withdrawal, blood pressure and alcohol all produce it. We go through the possibilities in why you wake up with a headache. If it comes with loud snoring, witnessed pauses or daytime sleepiness, say so when you ask about testing.

My test was positive. What changes, and what happens to my body?

That is the next page, not this one. Treatment options and how to tell whether they are working are in sleep apnea treatment options, and what people actually notice in the first weeks and months is in body changes after CPAP.

Which doctor do I even go to?

Start with your primary care clinician, who can order a home test directly in many systems. If the picture is complicated, or the home test comes back negative and you still have symptoms, you want a sleep medicine physician. The AASM runs a public directory of accredited sleep centers by state (24); accreditation means the center is led by a sleep physician and meets AASM standards.

How to bring this up with your doctor

Go in with specifics, not adjectives. "I'm tired" starts one conversation; "I fall asleep at red lights, my partner says I stop breathing, and my resting heart rate at night has been running 8 beats higher than it did in July" starts another.

Bring, if you have it: how long this has been going on; what a bed partner has witnessed (pauses, gasping, choking, snoring loud enough to hear through a door); your morning symptoms (headache, dry mouth, waking unrefreshed); whether you doze off while driving, in meetings or watching TV; your blood pressure readings; your neck circumference; your medications, especially opioids, sedatives or muscle relaxants; and any heart, lung, neuromuscular or stroke history, because those decide which test you need (1, 12).

What to say. "I think I should be tested for sleep apnea. I have daytime sleepiness plus [loud snoring / witnessed pauses / high blood pressure]. Can we do a home sleep apnea test, or do my other conditions mean I need an in-lab study?"

What to ask for. A referral for a home sleep apnea test or polysomnography; the full report, not a summary; and the specific numbers — AHI, RDI, REI, ODI, time supine, and which hypopnea scoring rule the lab used (14).

If the home test comes back negative and you still feel terrible, this is the sentence to use: "The AASM guideline says that if a single home sleep apnea test is negative, inconclusive or technically inadequate, polysomnography should be performed. Can we arrange that?" (1) If the answer is "let's repeat the home test," you can ask why, since the guideline advises against it in most cases.

If you are brushed off, ask for the reasoning to be written in your notes — "declined sleep study despite daytime sleepiness, witnessed apneas and hypertension" — and ask for a referral to a sleep medicine physician. A second opinion from an accredited sleep center is a reasonable request (24).

Red flags, same as above, same size type. Call 911 for chest pain or pressure, trouble breathing at rest, blue or gray lips or face, fainting, a seizure, a very irregular or racing heartbeat, or sudden weakness, numbness, slurred speech or a drooping face. Seek same-day care for falling asleep while driving or at work, sleepiness that suddenly got much worse, morning headaches with confusion, or swelling in both legs with breathlessness lying flat (4).

How Welltory helps — and what it cannot do

The limits first, because on this topic they matter more than anything else. Welltory does not detect sleep apnea and is not a substitute for a diagnostic test. Welltory is a general wellness product, not a medical device. It does not diagnose, detect, predict, monitor, prevent, treat or mitigate any condition. It does not measure your breathing or your blood oxygen, it does not count respiratory events, and it cannot produce an AHI, an REI or anything resembling one. No consumer tracker can: the AASM's position is that consumer sleep technology "cannot be utilized for the diagnosis and/or treatment of sleep disorders" (22), and even the FDA-authorized breathing-disturbance notifications now built into some watches are explicitly not intended to replace diagnosis by a clinician (21). A sleep apnea diagnosis comes from a test your doctor orders. Full stop. Welltory's help center also notes that measurements may be less accurate if you are under 18 or pregnant, because those bodies work differently (25).

What it can do is give you a record instead of an impression. The hardest part of asking for a sleep study is that the evidence lives in your memory, and your memory of how tired you were in August is bad. Welltory records the same things every day — resting heart rate, heart rate variability, stress minutes, Battery and sleep — so three months later you have a line, not a feeling.

You note what happened; the app keeps the rest. Tap the plus at the top of the feed to add a tag: "snored, partner woke me," "woke with headache," "fell asleep in the 2 p.m. call," "slept on my back." When the Today screen marks a stress or rest stretch it asks "What happened?" — pick a suggested tag, type or talk, and correct the tags before saving.

Sleep analysis shows how your heart rate moved across the night (detailed sleep analysis needs an Apple Watch; Battery and stress minutes need iOS with an Apple Watch or Oura). That is a heart-rate curve, not a breathing curve. It cannot show an apnea. What it can show is whether your nights look different from your own nights two months ago.

Over time, My Patterns looks for what your tagged days have in common (iOS with an Apple Watch or Oura; patterns start appearing at around seven tagged events in a month). Any pattern it finds is an association to discuss, never a cause and never a diagnosis.

A record you can hand over at the appointment. The Journal (Premium) keeps your measurements, tags and notes, and from the web app you can export a CSV (Dashboard → choose a chart → Export). Walking in with "I snore, I wake up wrecked, and here is what my nighttime heart rate has done for two months" is a reason for a clinician to order a test. It is not a replacement for one, and nothing in the app should delay seeing someone.

The same red flags apply here, in the same type size: chest pain, trouble breathing at rest, blue or gray lips, fainting, a seizure, a racing irregular heartbeat, or sudden weakness, numbness, slurred speech or a drooping face means 911 — whatever your data says. Falling asleep at the wheel needs care today.

How we made it

Built from the primary guidance and the peer-reviewed literature, not from secondary summaries: the AASM clinical practice guideline for diagnostic testing for adult OSA (Kapur et al., JCSM 2017, still current while a focused update is in development) for who gets which test, the four comorbidity exclusions and the rule on a negative home test; the VA/DoD clinical practice guideline v3.0 (January 2025) for the current US federal position; AASM Quality ID #277 v9.0 (December 2024) for the severity bands; the AASM position statement on arousal-based scoring (JCSM 2018) for the 3% versus 4% hypopnea problem; Pitkänen et al. (Sleep Advances 2023) and Roeder et al. (Thorax 2020) for how much a single home night can understate things; Pivetta et al. (JAMA Network Open 2021) for STOP-BANG; the AASM pediatric position paper (JCSM 2017) and the AAP guideline (Pediatrics 2012) for children; CMS's NCD 240.4, Medicare.gov and the CMS 2024 claims dataset for coverage and cost; and Aetna's clinical policy bulletin (last reviewed 6 October 2026) as a worked example of commercial payer criteria. Every URL was checked. Where a detail was not clearly in a primary source, we left it out.

This article contains no Welltory user data and no chart built on our own cohort, and that is deliberate: Welltory does not measure breathing or blood oxygen, so there is nothing in our data that corresponds to a scored respiratory event or could sit next to an AHI. The description of the app covers only what it measures.

Made with AI tools, then edited and fact-checked by the Welltory team. See our Editorial & AI policy.

Jane Smorodnikova, co-founder of Welltory, built the methodology behind how we read physiological data.

Written by Tatsiana Yashyna.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data