What 6,044 Sleep Studies Reveal About Snoring

By Professor Vik Veer, Consultant ENT Surgeon, Royal National ENT Hospital & Queens Hospital, London - Updated 22.05.2026

I am a Consultant ENT Surgeon at the Royal National ENT Hospital and Queens Hospital in London. A large part of my clinical work involves a procedure called drug-induced sleep endoscopy — DISE — in which I watch patients sleep with a tiny camera through the nose to see what happens to the airway during snoring and sleep apnoea. I have performed thousands of these procedures. The thing that struck me, early on and repeatedly since, was how consistently things improved the moment a patient rolled onto their side.

The tongue, which is a surprisingly heavy structure, stops falling backwards into the throat. The soft palate, which sags down under gravity in the supine position, lifts slightly. The tonsils, which in some patients fall back like boulders across a cave entrance, move out of the way and clear the airway. The change is immediate and, in many patients, dramatic. An airway that had been intermittently collapsing completely opens up.

What frustrated me was that this observation was not translating into clinical practice. Patients were leaving clinic with CPAP machines, or on waiting lists for surgery, when the most clinically significant variable — whether their sleep apnoea was positional in the first place — was barely being discussed. So I went back to the data.

I analysed 6,044 sleep studies from patients who had been referred to our sleep service at the Royal National ENT Hospital and Queens Hospital for snoring and suspected sleep apnoea. What I was looking for was simple: for each patient, how different was their sleep apnoea severity when sleeping on their back compared with sleeping on their side?

The figures were striking. 60.3% of patients with diagnosed sleep apnoea would have had completely normal sleep studies — no snoring, no apnoea events, normal oxygen levels — if they had simply remained on their side throughout the night. These were not patients with trivial sleep apnoea. The group included patients with overall AHI scores of 40, 50, and 60 events per hour when lying on their back, who dropped to normal levels when measured on their side. Patients who would routinely be classified as severe sleep apnoea on the basis of their supine measurements, and would typically be offered CPAP immediately, had no measurable sleep apnoea at all in the lateral position.

A further 85.7% would have been at least 20% better simply by staying on their side. This means that the overwhelming majority had an airway that was, to some degree, position-dependent.

These figures come from a retrospective audit of consecutive NHS referrals to our sleep department. Every patient who presents to us with snoring or suspected sleep apnoea receives a sleep study, regardless of severity or perceived risk. This matters because it means the cohort represents the full clinical spectrum — from patients with mild, occasional snoring to those with severe, symptomatic sleep apnoea — rather than a pre-selected or research-enriched group. The data reflects what happens in real clinical practice, not what happens in a laboratory.

The audit findings prompted a more rigorous analysis. Between July 2017 and December 2024, 1,090 consecutive NHS patients who underwent home sleep testing using the NoxT3 cardiorespiratory monitor at our service formed the basis of a formal cohort study, currently under review at the Thorax Journal. The NoxT3 is a validated Type 3 home sleep-testing device that records nasal airflow, oxygen saturation, respiratory effort, and — critically — body position throughout the night via continuous accelerometry. This allowed us to calculate position-specific AHI values for every patient with sufficient time in both the supine and non-supine positions.

The cohort had a median age of 50 years, was 76.1% male, and had a median overall AHI of 16.1 events per hour — a profile representative of a general clinical sleep population. OSA severity was mild in 46.7%, moderate in 27.0%, and severe in 26.3% of patients.

The headline findings

The results were consistent with the original audit and, in some respects, more striking.

• 86.1% of patients (939 of 1,090) demonstrated at least 10% improvement in AHI when sleeping non-supine compared with supine.

• 83.2% showed at least 20% improvement.

• 66.6% met the Mador criteria for positional sleep apnoea, defined as at least 50% improvement in AHI when off the back.

• 35.9% — more than one in three — achieved complete normalisation: their supine AHI was elevated (above 5 events per hour), but their non-supine AHI was below 5. No measurable sleep apnoea when not on their back. None.

The 35.9% complete normalisation rate deserves particular attention. These are patients who, under current clinical practice, would typically be assessed on the basis of their overall AHI — which reflects a mixture of supine and non-supine sleep — and offered treatment accordingly. Most would be prescribed CPAP. But they have no sleep apnoea that is not caused by lying on their back. For this group, positional therapy is not an adjunct or a partial solution. It is the complete answer.

Not every patient responds to positional change equally. The study identified the factors that predict how much benefit a given patient is likely to see.

BMI: the strongest modifiable predictor

BMI showed a clear and clinically important dose-response relationship with positional benefit. In patients of normal weight (BMI below 25 kg/m²), the median AHI improvement when moving from supine to non-supine sleep was 79.3%. In overweight patients (BMI 25–30 kg/m²), it was 77.0%. Both figures are striking, and both point to a group in whom positional therapy is likely to produce meaningful benefit.

In patients with class I obesity (BMI 30–35 kg/m²), the median improvement fell to 60.6% — still substantial. In patients with class II–III obesity (BMI above 35 kg/m²), the median improvement was 36.1%, and fewer than one in four achieved complete normalisation.

The table below summarises these findings:

Table: Positional AHI improvement by BMI category. Source: Veer et al., under review at Thorax, 2025. n=719 patients with BMI data.

The practical implication is clear: patients with a BMI below 30 kg/m² and mild-to-moderate sleep apnoea are the highest-probability responders to positional therapy. In this group, approximately three in four patients will meet formal positional OSA criteria, and around 38–39% will achieve complete normalisation. For these patients, a conversation about sleep position should precede any discussion of CPAP.

OSA severity: it helps across the spectrum, but less so in severe cases

The degree of positional benefit also varied with overall OSA severity, which is biologically intuitive. Patients with mild OSA (AHI 5–14.9 events per hour) showed a median improvement of 80.0% when sleeping non-supine, and 42.8% achieved complete normalisation. In moderate OSA (AHI 15–29.9), median improvement was 74.6%, with 38.1% achieving normalisation.

In severe OSA (AHI 30 or above), the median improvement fell to 36.6%, and complete normalisation was achieved in 21.3%. This is lower, but it is not negligible. Notably, 36.9% of patients with severe OSA still met Mador positional OSA criteria — meaning their AHI halved when they moved off their back. A 50% reduction in severe sleep apnoea is clinically meaningful, even if it does not normalise the AHI. It may reduce pressure requirements, improve CPAP tolerance, and reduce cardiovascular load.

Age: not a predictor — at all

Perhaps the most clinically important finding for day-to-day practice is what age does not predict. Statistical analysis across age groups from under 30 to over 70 found no significant relationship between age and positional benefit. The Kruskal-Wallis p-value was 0.60, which is about as far from statistical significance as it is possible to be. The proportion of patients meeting positional OSA criteria was consistent across every decade.

This matters because age is often, implicitly or explicitly, used to triage patients away from positional therapy. Older patients are assumed to be less likely to respond, or less likely to be able to change their sleeping habits. Neither assumption is supported by the data. A 70-year-old patient with mild-to-moderate sleep apnoea and a BMI below 30 has, on the evidence, roughly the same probability of responding to positional therapy as a 35-year-old with the same profile.

CPAP remains the gold-standard treatment for sleep apnoea, and I prescribe it regularly. For patients with severe, position-independent sleep apnoea, it is often the right answer. But it is important to be honest about the limitations. Somewhere between 40% and 60% of patients abandon CPAP within the first year — not because they do not understand its importance, but because sleeping with a mask strapped to the face is genuinely difficult for a substantial proportion of people. The mask leaks. The pressure causes aerophagia. The straps create pressure sores. Some people simply cannot tolerate the claustrophobia, night after night.

The patients in this cohort — 35.9% of whom have no non-supine sleep apnoea at all — are being offered a technology that they may struggle to use, when the intervention they actually need is a change in sleeping position. This is not a minor inefficiency. It is a systematic failure to use the most straightforward tool available.

The Cochrane review of positional therapy¹⁵ concluded that there was insufficient evidence to recommend it routinely. But as the authors acknowledged, this reflected a paucity of large, well-designed trials rather than evidence of ineffectiveness. The present data provide the epidemiological case for those trials. We now know the scale of the problem. We know which patients are most likely to respond. The next step is prospective evidence that the improvement observed in sleep studies translates into clinical outcomes over time.

If you snore, or if you have been told you have sleep apnoea, the single most useful question to ask is whether your sleep study includes position-specific data. A modern home sleep test can record your AHI separately for the time you spend on your back and the time you spend on your side. If your doctor has not discussed this with you, it is worth asking.

The reason this matters is practical. If your AHI is substantially higher when supine than when lateral, you have positional sleep apnoea. For many patients, this means that a change in sleeping position is not a lifestyle suggestion — it is the treatment. It carries no adverse effects, no cost, and no technology to tolerate.

The barriers to sustained side sleeping are real and should be taken seriously. Ear discomfort from prolonged pressure on the cartilage, inadequate pillow height for lateral sleeping, shoulder compression, arm numbness, and hip pain are all genuine anatomical problems that cause people to give up and return to their back by morning. The phrase I hear most often in clinic is: "I tried sleeping on my side but it was too uncomfortable." This is not failure. It is unsupported side sleeping failing, which is what unsupported side sleeping does.

There are practical steps that address these problems. Getting the pillow height right — which is considerably higher for side sleeping than for back sleeping — makes a significant difference. A second pillow between the knees reduces hip rotation and lower back discomfort. An awareness that the ear needs to be clear of sustained pressure helps people position themselves more effectively. None of this requires purchasing anything.

If you try sustained side sleeping for two to three weeks and find that you snore less, wake more refreshed, and your partner is no longer being disturbed, that is the most reliable evidence available that positional therapy works for you. At that point, if you are struggling to maintain the position comfortably through the night, it is worth exploring what support mechanisms might help.

The findings from both analyses point to the same conclusion: positional assessment should be routine in sleep medicine, not occasional. Every diagnostic sleep study that records body position — and modern home sleep tests almost all do — should report position-specific AHI as standard. The conversation in clinic should routinely include the question: is this sleep apnoea position-dependent?

For patients with mild-to-moderate sleep apnoea and a BMI below 30 who achieve non-supine normalisation, the evidence now suggests that positional therapy should be the first conversation, not the afterthought. These patients do not have sleep apnoea when they are on their side. The treatment is to be on their side.

For patients with more severe sleep apnoea, or higher BMI, the picture is more nuanced. Positional therapy may reduce the severity of their condition meaningfully, allow CPAP to be used at lower pressures, and improve tolerance. It is not a replacement for other treatment in these patients, but it is a complement that the data suggest most of them would benefit from.

This is not a call to abandon established treatments. CPAP saves lives in patients with severe sleep apnoea, and I would not suggest otherwise. Surgery has a role for the right patients. What the data argue for is a more systematic conversation about position — one that happens before rather than after the prescription pad comes out.

My Side Sleeping Pro is designed to help you sleep on your side. 

1. Benjafield AV, Ayas NT, Eastwood PR, et al. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. Lancet Respir Med. 2019;7(8):687–698.

2. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc. 2008;5(2):173–178.

3. Cartwright RD. Effect of sleep position on sleep apnea severity. Sleep. 1984;7(2):110–114.

4. Mador MJ, Kufel TJ, Magalang UJ, et al. Prevalence of positional sleep apnea in patients undergoing polysomnography. Chest. 2005;128(4):2130–2137.

5. Joosten SA, O'Driscoll DM, Berger PJ, Hamilton GS. Supine position related obstructive sleep apnea in adults: pathogenesis and treatment. Sleep Med Rev. 2014;18(1):7–17.

6. Pevernagie DA, Stanson AW, Sheedy PF, et al. Effects of body position on the upper airway of patients with obstructive sleep apnea. Am J Respir Crit Care Med. 1995;152(1):179–185.

7. Oksenberg A, Silverberg DS, Arons E, Radwan H. Positional vs nonpositional obstructive sleep apnea patients. Chest. 1997;112(3):629–639.

8. Ravesloot MJL, van Maanen JP, Dun L, de Vries N. The undervalued potential of positional therapy in position-dependent snoring and obstructive sleep apnoea. Sleep Breath. 2013;17(1):39–49.

9. Srijithesh PR, Aghoram R, Goel A, Dhanya J. Positional therapy for obstructive sleep apnoea. Cochrane Database Syst Rev. 2019;5:CD010990.

10. National Institute for Health and Care Excellence. Continuous positive airway pressure for the treatment of obstructive sleep apnoea/hypopnoea syndrome. Technology appraisal guidance TA139. London: NICE; 2008.

11. Epstein LJ, Kristo D, Strollo PJ Jr, et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med. 2009;5(3):263–276.

12. Heinzer R, Vat S, Marques-Vidal P, et al. Prevalence of sleep-disordered breathing in the general population: the HypnoLaus study. Lancet Respir Med. 2015;3(4):310–318.

13. De Vries GE, Hoekema A, Doff MH, et al. Usage of positional therapy in adults with obstructive sleep apnea. J Clin Sleep Med. 2015;11(2):131–137.

14. Veer V, et al. Positional improvement in obstructive sleep apnoea is near-universal: a large single-centre real-world cohort study with predictors of response. Thorax (under review).

15. Lan MC, Liu SYC, Lan MY, et al. Lateral pharyngeal wall collapse associates with hypoxemia in obstructive sleep apnea. Laryngoscope. 2015;125(10):2408–2412.

Professor Vik Veer is a Consultant ENT Surgeon at the Royal National ENT Hospital and Queens Hospital, London, and the author of Your Sleep Position eBook (2026). He runs iwantgreatsleep.com.