Close to 425 million adults may have obstructive sleep apnoea. One well conducted night of recording can be enough to make the diagnosis; the real issue is picking up the snoring, the breathing pauses and the repeated drops in oxygen that go with a higher cardiovascular risk.
If you snore, wake up exhausted or suspect sleep apnoea, Noria Health Hub can arrange a full check-up or a second medical opinion in Brussels, with a report within 24 h, coordinate your care where needed across more than 40 medical pathways in 48 h, then, if your medical condition allows, arrange surgery within 7 days.
One good night can be enough to make the diagnosis
| What was measured | The result |
|---|---|
| 24 studies, 3,250 participants tested more than once | 49% change severity class at least once. 41% see their index vary by more than 10 events per hour. And yet the mean difference between night 1 and night 2 across the whole group is 1.7 events per hour, effectively nil |
| 10,340 adults tested on 3 consecutive nights | 93% of normal tests and 87% of severe forms on the first night are confirmed. But around 20% of mild and 20% of moderate forms are misclassified |
| 67,278 adults, more than 11.6 million nights recorded at home | Diagnostic performance rises from a score of 0.77 with 1 night to 0.94 with 14 nights, then plateaus |
Polysomnography, or a home test in the right patient, can be enough on a single night when it is technically sound and fits the symptoms. The test is mainly repeated when it comes back negative, inconclusive or of insufficient quality while clinical suspicion remains high.
Variability is not explained by the first night alone
The usual explanation is the first-night effect, that night you sleep badly because you are being recorded. The effect is real, it has been measured, and it does not say what people make it say.
A meta-analysis of 53 studies and 1,422 subjects establishes that the first night lengthens sleep onset, reduces total sleep time and increases light sleep. But it finds no significant difference in the apnoea index, nor in deep sleep, nor in periodic limb movements. In other words, the first night degrades sleep architecture, it does not distort the respiratory measurement.
Night-to-night variability is therefore real and physiological. It stems from sleeping position, nasal congestion, alcohol, the sleep stage reached. It will not disappear by repeating the test under better conditions.
Accuracy is not the same from one device to the next
| The type of test | Measured performance |
|---|---|
| Type 3 home polygraphy | Underestimates the respiratory index by 8 events per hour on average against full polysomnography, because it does not measure actual sleep time |
| Simplified 1 to 3 channel sensors | Across 24 studies and 2,068 patients, mean bias ranges from minus 14.8 to plus 10.6 events per hour depending on the device. Specificity falls as low as 25% |
| Peripheral arterial tonometry | Across 13 studies and 1,227 participants, sensitivity of 88% and specificity of 74% at the 15 events per hour threshold. The authors conclude it does not replace polysomnography |
The American Academy of Sleep Medicine guidelines point the same way: a home test that is negative, inconclusive or technically inadequate should be followed by polysomnography; after a negative polysomnography, a further test can be discussed if suspicion of apnoea remains high.
Even experts do not always score events the same way
This is the least known point in the field, and it comes from the largest published assessment programme, over 2,500 scorers and more than 3.2 million scoring decisions.
| What is scored | Agreement between readers |
|---|---|
| Sleep stages, overall | 82.6% |
| Deep sleep | 67.4% |
| Light sleep, stage N1 | 63.0% |
| Respiratory events, overall | 93.9%, but 87.8% of epochs contain no event, which inflates the figure |
| Obstructive apnoea | 77.1% |
| Hypopnoea | 65.4% |
| Central apnoea | 52.4% |
Hypopnoea is precisely the event that tips a patient from one severity class to another. That is where disagreement is greatest.
Part of the variation from one night to the next may therefore come not only from the patient, but also from the way the recording is read. The studies do not always separate these two sources of variability clearly.
Automatic scoring, close agreement, clinical benefit not demonstrated
An automatic scoring system trained on recordings from 15,660 participants across 16 studies reaches accuracy comparable to the best human experts. Independent validation finds an agreement coefficient of 0.79 for the algorithm against 0.78 for trained scorers, a statistical equivalence.
But no randomised trial shows that this improves patient health. All the available literature measures agreement with humans, never a health endpoint. And human scoring, which serves as the reference, agrees 63% of the time on stage N1. No algorithm can be more accurate than the reference it was trained on.
Snoring, oxygen loss and cardiovascular risk
In 2 large American cohorts covering more than 4,500 adults, hypoxic burden, the amount of oxygen loss caused by episodes of obstruction, was one of the factors most strongly associated with cardiovascular events. Another cohort followed for 18 years found a risk of death from any cause around 3 times higher in people with severe apnoea than in those with no sleep-disordered breathing. These data are observational: they show the risk associated with severe apnoea, not proof that a given treatment extends life.
Treatment, benefits demonstrated on symptoms, not on every risk
| What was measured | The result |
|---|---|
| Daytime sleepiness, 47 trials, 7,024 participants | Epworth score reduced by 2.33 points. Solid |
| Quality of life | Modest but significant improvement on mental and physical components |
| Cardiovascular events, 2,717 patients, 3.7 years of follow-up | 17.0% against 15.4%, hazard ratio 1.10. No benefit |
| Meta-analysis of 10 trials, 7,266 patients | No significant association with major cardiovascular events, cardiovascular death or all-cause mortality. And no association with adherence level |
| Road accidents | A 72% reduction reported, but only across 9 before-and-after studies. No randomised trial has found a benefit on crashes |
Questionnaires, useful for sorting patients, not enough to make the diagnosis
Across 47 studies and 26,547 participants, a STOP-BANG score of 3 or more detects 95% of moderate to severe apnoea, but with a specificity of 28%. The Epworth sleepiness scale has a sensitivity of 48%, meaning it misses one patient in two. A negative questionnaire is worth something, a positive one almost nothing.
One point that matters for international patients. STOP-BANG performance falls sharply in East Asia, with an area under the curve of 0.52 against 0.81 to 0.89 in Europe and North America, because it relies partly on body mass index and neck circumference.
What the studies do not allow us to claim
We will not claim that treating sleep apnoea extends life or prevents heart attacks, because randomised trials do not show it. You will not read here that recording several nights improves your health, because the studies show that it reclassifies patients, without having demonstrated that this reclassification produces a better state of health. And we will point out that the most quoted global estimate, 936 million adults affected, is funded by a device manufacturer.
What is demonstrated, and it is enough, is that treatment reduces sleepiness and improves quality of life, and that apnoea never recorded is never treated.
In practice, if this concerns you
You set out your problem, your symptoms or your question. Reports, test results and images already available can be reviewed before you arrive, so that the pathway is prepared. The aim is not to run every possible test, but to select those that can genuinely help confirm a hypothesis, rule out a risk or guide a decision. The findings are then summarised and the next steps are organised. Follow-up is coordinated afterwards.
Need a medical opinion quickly? For a concern that does not call for emergency services, you can ask to speak to a doctor at any time through Noria Health Hub. This on-demand medical consultation is billed separately from the assessment pathway.
Further reading
The companion articles.
Hearing” rel=”noopener” target=”_blank”>https://noriahealth.com/blog/hearing-you-do-not-correct/”>Hearing you do not correct
What” rel=”noopener” target=”_blank”>https://noriahealth.com/blog/what-a-health-assessment-can-actually-find/”>What a health assessment can actually find
Sources
Roeder M et al., Night-to-night variability of respiratory events in obstructive sleep apnoea, systematic review and meta-analysis, Thorax, 2020. https://doi.org/10.1136/thoraxjnl-2020-214544
Punjabi NM et al., Variability and misclassification of sleep apnea severity based on multi-night testing, Chest, 2020. https://doi.org/10.1016/j.chest.2020.01.039
Lechat B et al., Multinight prevalence, variability, and diagnostic misclassification of obstructive sleep apnea, American Journal of Respiratory and Critical Care Medicine, 2022. https://doi.org/10.1164/rccm.202107-1761OC
Ding L et al., A meta-analysis of the first-night effect in healthy individuals for the full age spectrum, Sleep Medicine, 2022. https://doi.org/10.1016/j.sleep.2021.12.007
Abrahamyan L et al., Diagnostic accuracy of level IV portable sleep monitors versus polysomnography, Sleep and Breathing, 2018. https://doi.org/10.1007/s11325-017-1615-1
Yamada T et al., Diagnostic accuracy of home sleep apnea testing using peripheral arterial tonometry, Journal of Sleep Research, 2022. https://doi.org/10.1111/jsr.13682
Kapur VK et al., Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea, Journal of Clinical Sleep Medicine, 2017. https://doi.org/10.5664/jcsm.6506
Rosenberg RS, Van Hout S, The American Academy of Sleep Medicine inter-scorer reliability program, sleep stage scoring, Journal of Clinical Sleep Medicine, 2013. https://doi.org/10.5664/jcsm.2350
Rosenberg RS, Van Hout S, The American Academy of Sleep Medicine inter-scorer reliability program, respiratory events, Journal of Clinical Sleep Medicine, 2014. https://doi.org/10.5664/jcsm.3630
Lee YJ et al., Interrater reliability of sleep stage scoring, a meta-analysis, Journal of Clinical Sleep Medicine, 2022. https://doi.org/10.5664/jcsm.9538
Perslev M et al., U-Sleep, resilient high-frequency sleep staging, npj Digital Medicine, 2021. https://doi.org/10.1038/s41746-021-00440-5
Kevat A et al., Evaluation of automated pediatric sleep stage classification using U-Sleep, Journal of Clinical Sleep Medicine, 2025. https://doi.org/10.5664/jcsm.11362
Feltner C et al., Screening for obstructive sleep apnea in adults, updated evidence report for the US Preventive Services Task Force, JAMA, 2022. https://doi.org/10.1001/jama.2022.18357
McEvoy RD et al., SAVE, CPAP for prevention of cardiovascular events in obstructive sleep apnea, New England Journal of Medicine, 2016. https://doi.org/10.1056/NEJMoa1606599
Yu J et al., Association of positive airway pressure with cardiovascular events and death in adults with sleep apnea, JAMA, 2017. https://doi.org/10.1001/jama.2017.7967
Tregear S et al., Continuous positive airway pressure reduces risk of motor vehicle crash among drivers with obstructive sleep apnea, Sleep, 2010. https://doi.org/10.1093/sleep/33.10.1373
Pivetta B et al., Use and performance of the STOP-Bang questionnaire across geographic regions, JAMA Network Open, 2021. https://doi.org/10.1001/jamanetworkopen.2021.1009
Diagnostic utility of obstructive sleep apnea screening questionnaires, a comprehensive meta-analysis, Sleep and Breathing, 2024. https://doi.org/10.1007/s11325-024-03169-z
Heinzer R et al., Prevalence of sleep-disordered breathing in the general population, the HypnoLaus study, The Lancet Respiratory Medicine, 2015. https://doi.org/10.1016/S2213-2600(15)00043-0
Benjafield AV et al., Estimation of the global prevalence and burden of obstructive sleep apnoea, The Lancet Respiratory Medicine, 2019, study funded by a device manufacturer. https://doi.org/10.1016/S2213-2600(19)30198-5
National Heart, Lung, and Blood Institute (NIH), NIH-funded study explains increased cardiovascular risks in obstructive sleep apnea, 2023. https://www.nhlbi.nih.gov/news/2023/study-explains-increased-cardiovascular-risks-people-obstructive-sleep-apnea
Young T et al., Sleep disordered breathing and mortality: eighteen-year follow-up of the Wisconsin Sleep Cohort, Sleep, 2008. https://pubmed.ncbi.nlm.nih.gov/18714778/