What Position are Humans Designed to Sleep in? The Science & Evolution of Sleeping Positions

By Vik Veer, Consultant ENT Surgeon, Royal National ENT Hospital, London

Updated 21.5.2026

I have spent years looking at human airways during sleep, performing Drug-Induced Sleep Endoscopy procedures in which I anaesthetise a patient, observe exactly what their throat does during simulated sleep, and then watch what happens when I turn them from their back to their side. The transformation is, to put it mildly, striking. In one position, the throat collapses. In the other, it opens. The physics of this is not subtle. And yet the broader question of what position humans are actually designed to sleep in has received surprisingly little systematic attention.

This article sets out to answer the question - What Position are Humans Designed to Sleep in? I have drawn on evidence from evolutionary biology, anthropology, sleep science, and clinical medicine. The conclusion, supported by a substantial body of peer-reviewed research, is that back sleeping is almost certainly a product of the modern flat mattress rather than anything inherent to human biology. We are, in all probability, designed to sleep predominantly on our sides.

What Our Evolutionary History Tells Us

The Evidence From Our Closest Relatives

The most direct window into ancestral human sleep behaviour comes not from any study of people, but from the study of great apes. Chimpanzees, bonobos, gorillas, and orangutans are our closest evolutionary relatives, sharing between 96 and 99 per cent of our DNA. All of them, without exception, build sleeping nests every single night. This behaviour has never been observed in any other non-human primate species. A comprehensive review of 46 studies spanning all great ape species and subspecies, published in the American Journal of Physical Anthropology by Fruth and colleagues in 2018, confirmed that nest-building is a universal behaviour across the entire great ape family.¹ These platforms are not flat. They are constructed from branches, leaves, and vegetation in ways that create natural contours: a slight depression for the body, raised edges for support, and structural variation that accommodates the shape of the sleeping animal.

Key fact: Every great ape species on Earth builds a structured sleeping nest every night. None sleeps on a flat surface.

The critical question is what posture our relatives adopt once settled on these platforms. The only published study to directly measure this used infrared video recording to observe six captive chimpanzees over two nights, accumulating 95 hours of observational data. The findings were clear: adult chimpanzees preferred the lateral posture above all others.² They changed position approximately five times per hour, which closely mirrors human behaviour (adults typically shift position between eleven and forty-five times per night), and the duration of a single held posture ranged from just 0.1 minutes to 60.8 minutes. Crucially, there was no coaxing, no mattress designed for back sleeping, and no sleep advice. Given the choice, a chimpanzee lies on its side.

A comparison of orangutan and baboon sleep published by Samson and colleagues in 2015 adds an important dimension to this picture.³ Over 128 nights of orangutan observation and 45 nights of baboon observation using infrared videography at Indianapolis Zoo, the researchers found that orangutans, who construct sleeping platforms, slept in what the authors described as "insouciant, relaxed positions" on their constructed materials. Baboons, which lack access to any augmented sleeping surface, were compelled to sleep in guarded, upright positions balanced on their ischial callosities, unable to fully recline. The implication is significant: the ability to sleep in a relaxed lateral posture is directly enabled by having an appropriate surface. Remove the surface, and the animal cannot adopt the position. Provide it, and the lateral posture is the natural default.

The Transition to Ground Sleeping and What It Changed

Our own lineage diverged from that of our great ape cousins and, over time, made the transition from sleeping in trees to sleeping on the ground. Coolidge and colleagues proposed in 2006 that this transition, which likely became complete with the emergence of Homo erectus approximately 1.9 million years ago, represented one of the most significant cognitive leaps in human evolution.⁴ Freed from the need to grip a branch throughout the night, early terrestrial humans could enter deeper stages of REM sleep, which in turn enhanced memory consolidation, procedural learning, and creative thought. The tree-to-ground transition was not merely a change of venue. It was a biological upgrade.

But even after Homo erectus began sleeping on the ground, the surface being slept on was nothing resembling a modern flat mattress. Evidence from chimpanzee behaviour is instructive here. Research by Koops and colleagues in 2012, studying unhabituated chimpanzees in Guinea's Nimba Mountains, found that ground-nesting was widespread and habitual in two separate communities, with males constructing elaborate nocturnal nests directly on the ground.⁵ These were structured sleeping platforms at ground level, not simply lying flat on bare earth. The implication for understanding human sleep is that even as hominins transitioned to terrestrial sleeping, they almost certainly continued to construct or select sleeping surfaces with natural contours. The completely flat, uniform sleeping surface is a product of the last century or two, not of the preceding two million years.

Hunter-Gatherers: The Modern Window Into Ancestral Sleep

Living hunter-gatherer populations offer perhaps the closest available analogue to how humans slept for the overwhelming majority of their evolutionary history. In 2015, Yetish and colleagues published a landmark study in Current Biology, examining sleep patterns in three pre-industrial societies: the Hadza of Tanzania, the San of Namibia and Botswana, and the Tsimane of Bolivia.⁶ The study, which accumulated 1,165 days of actigraphy data across all three populations, found that these groups slept for between 5.7 and 7.1 hours per night on average, which is broadly similar to industrialised populations despite having no electricity, no screens, and no alarm clocks. All three groups showed strikingly similar sleep organisation, suggesting this represents something close to the core human sleep pattern. Critically, all three populations sleep on ground-level surfaces: animal skins, woven mats, or bare earth, with none of the engineered comfort of a modern mattress.

Key fact: Three geographically separate hunter-gatherer populations all show similar sleep patterns, sleeping on natural ground-level surfaces with none of the cushioning of a modern mattress.

A detailed study of Hadza sleep by Samson and colleagues in 2017, which followed 33 adults over 393 nights using actigraphy, found a mean sleep duration of 6.25 hours with a sleep efficiency of 68.9 per cent.⁷ The Hadza sleep in small grass huts with minimal bedding, on surfaces that retain the natural variation of the ground beneath them. There are slight depressions and elevations, areas of softness and firmness, all of which naturally accommodate the body's shape during lateral sleeping in a way that a completely flat surface cannot.

No published study has directly measured sleep position in hunter-gatherer populations, and intellectual honesty requires acknowledging this gap. It is not possible to state with certainty that hunter-gatherers sleep predominantly on their sides. What can be stated with confidence is that their sleeping surfaces make sustained back sleeping inherently uncomfortable in a way that a modern flat mattress does not, for reasons that will become clear in the following section.

The Hard Ground Problem: Why Back Sleeping Is a Modern Invention

This brings us to what I consider the most compelling and most overlooked argument in the entire debate about sleep position. When a human body lies on its back on a hard or firm surface, the weight of the body is distributed across a remarkably small contact area: the back of the skull (occiput), the shoulder blades (scapulae), the lower back (sacrum and coccyx), the calves, and the heels. In 1980, Garfin and colleagues published a detailed pressure mapping study of the human body across multiple sleeping surfaces.⁸ Using 65 pressure transducers distributed across a body-shaped sheet, they found that on hard and moderately soft surfaces, the body in the supine position was essentially supported at only these five bony points. The skin pressures measured over these bony prominences significantly exceeded the critical threshold for ischaemia of 30 mmHg, meaning they would restrict blood flow and cause pain.

Key fact: On a firm surface, the supine sleeping body concentrates pressure on five bony points — the back of the skull, shoulder blades, sacrum, calves, and heels — all at pressures that exceed the threshold for tissue ischaemia.

A 2025 clinical study by Borges and colleagues, measuring interface pressures on bony prominences across 448 assessments in each position, confirmed that pressure distribution varies substantially with position and surface type.⁹ Research by Mohamadi and colleagues, published in Ergonomics in 2025, measured body pressure distribution across supine, prone, and lateral positions on surfaces including a basic sleeping mat, and found that surface type dramatically altered the pressure experienced at bony landmarks.¹⁰ On minimal sleeping surfaces, the sustained supine position is simply not tolerable for long periods. The body signals, through discomfort and pain, that it needs to move.

This is the essential point. Before the invention of thick foam and sprung mattresses, sleeping on your back through the night was physiologically uncomfortable in a way that sleeping on your side was not. The lateral position distributes weight across a far larger surface area: the entire length of the outer thigh and hip, the lateral aspect of the shoulder, the knee, and the side of the lower leg. The contact points are broad and well-muscled rather than narrow and bony. On hard ground, lateral sleeping is simply more sustainable.

The modern flat mattress, by redistributing pressure so effectively that the supine position becomes comfortable, has in effect removed the body's own protective feedback signal. It has made a position tolerable that the body would otherwise reject. In doing so, it may have inadvertently encouraged a sleeping position that conflicts with several important biological systems.

When I see patients sleeping rough in the winter months, or observe people sleeping on hard surfaces in airports or on floors, there is a consistent pattern: virtually none adopt the supine position for sustained periods. Without the cushioning of a modern mattress, the body returns to what appears to be its default. It curls to its side, knees drawn up, weight distributed across the outer hip and shoulder. This is not a scientific study. But it is a consistent observation, and it points in the same direction as all the evidence described above.

The Biology: What Happens to Your Body When You Lie on Your Back

Evolutionary and anthropological arguments, whilst compelling, are ultimately circumstantial. The biological evidence, however, is anything but. There are now multiple well-evidenced physiological mechanisms that make lateral sleeping superior to supine in a range of circumstances.

Your Airway: The Gravity Problem

I spend a substantial portion of my clinical practice examining what happens to the human upper airway during sleep. The relevant anatomy is straightforward. The tongue, soft palate, and surrounding soft tissues of the throat are mobile structures that respond to gravity. In the supine position, gravity acts directly on these structures, pulling them posteriorly towards the back of the throat. The result is narrowing and, in susceptible individuals, complete collapse of the airway.

Key fact: In the supine position, 88.2% of obstructive sleep apnoea patients show complete collapse at the soft palate, and 53.2% at the tongue base. Turning to the side resolves tongue base obstruction in 94.9% of cases.

Research by Landry and colleagues, published in Sleep Medicine Reviews in 2023, established that supine-related obstructive sleep apnoea (OSA) is the most common clinical phenotype of the condition, defined by respiratory events occurring at twice the frequency in the supine position compared to non-supine positions.¹¹ A study by Park and colleagues in 2024, examining 500 OSA patients using polysomnography, found that 63.4 per cent had positional OSA, with an average apnoea-hypopnoea index (AHI) of 23.3 events per hour in positional patients compared to 43.9 events per hour in non-positional patients.¹² In a large prospective study of 689 OSA patients by Cheng and colleagues, published in the European Respiratory Journal in 2024, 75.8 per cent were identified as having supine-predominant OSA.¹³

The physical reality of what gravity does to the throat is most vividly demonstrated through Drug-Induced Sleep Endoscopy. In research by Mladoňová and colleagues published in 2025, examining 186 patients with a median AHI of 19.3 events per hour, the supine position produced complete obstruction at the soft palate in 88.2 per cent of cases, at the tongue base in 53.2 per cent, at the oropharynx in 33.3 per cent, and at the epiglottis in 15.6 per cent.¹⁴ When these same patients were placed in the lateral position, tongue base obstruction resolved completely in 94 of 99 cases: a resolution rate of 94.9 per cent, which was statistically significant at p less than 0.001. The tongue, when not subject to direct gravitational pull towards the spine, simply does not collapse.

Joosten and colleagues quantified the physiological changes that accompany lateral positioning in a study of 20 severe OSA patients published in Sleep in 2015.¹⁵ Lateral positioning significantly improved passive airway collapsibility, with pharyngeal critical closing pressure (Pcrit) moving from 2.02 cmH₂O in the supine position to minus 1.92 cmH₂O in the lateral position, a change that is of considerable clinical significance. Functional residual lung capacity also improved in the lateral position.

Crucially, this is not simply an issue for people with diagnosed sleep apnoea. Research by Rayward and colleagues, published in the Journal of Sleep Research in 2024, examined 41 mature-aged adults without sleep-disordered breathing using two nights of home polysomnography on a standardised firm mattress.¹⁶ In the supine position compared to the lateral position, these healthy participants showed a 379 per cent increase in respiratory arousals, a 108 per cent increase in the overall arousal index, and a 107 per cent increase in the wake index. All differences were statistically significant at p less than 0.001. The researchers also found a very strong correlation between lateral preference on the first night and lateral preference on the second (β = 0.95, p less than 0.001), confirming that individual position preference is consistent and that those who are most disturbed supinely compensate by spending more time lateral.

Key fact: Even in adults without sleep apnoea, the supine position produces 379% more respiratory arousals than the lateral position.

Your Brain: The Glymphatic System

Of all the biological findings linking sleep position to health, none is more striking than the discovery of the glymphatic system and its relationship to body posture. The glymphatic system is a network of channels surrounding the brain's blood vessels through which cerebrospinal fluid flows during sleep, clearing metabolic waste products including amyloid-beta, the protein whose accumulation in the brain is one of the hallmarks of Alzheimer's disease.

In 2015, Lee and colleagues published research in the Journal of Neuroscience examining how body posture affects the efficiency of this waste clearance system.¹⁷ Using dynamic-contrast-enhanced MRI and kinetic modelling to measure CSF-ISF exchange rates in anaesthetised rodents in the supine, prone, and lateral positions, they found that glymphatic transport was most efficient in the lateral position. In the prone position, tracer was retained and clearance was slower. The lateral position outperformed both alternatives. The study has accumulated 330 citations in the decade since its publication, making it one of the most-cited papers in sleep science. The authors proposed directly that "the most popular sleep posture (lateral) has evolved to optimise waste removal during sleep."

This finding has been replicated and extended by subsequent research into the glymphatic system's role in neurodegenerative disease. A 2020 review by Reddy and van der Werf in Brain Sciences, drawing on the accumulated evidence, identified sleep position as one of the key lifestyle factors modulating glymphatic clearance, alongside exercise, alcohol consumption, and dietary patterns.^17a The lateral sleep position was explicitly identified as beneficial. The implication, which should not be overstated given that the primary data comes from animal models, is that sleeping on your side may, over a lifetime, reduce the accumulation of the proteins associated with neurodegenerative disease.

Key fact: Brain waste clearance — including amyloid-beta linked to Alzheimer's disease — is most efficient in the lateral sleeping position, according to the most-cited study in sleep science on this topic.

Your Digestive System: Acid Reflux

Gastro-oesophageal reflux disease (GORD) affects a substantial proportion of the population and is significantly worsened during sleep. The relationship between sleep position and nocturnal acid reflux was examined in a study by Schuitenmaker and colleagues published in the American Journal of Gastroenterology in 2021, which followed 57 patients with concurrent pH-impedance monitoring and sleep position tracking throughout the night.¹⁸

The results were precise and clinically meaningful. In the left lateral decubitus position, the median oesophageal acid exposure time was 0.0 per cent, with an interquartile range of 0.0 to 3.0 per cent. In the right lateral position, this rose to 1.2 per cent (p = 0.022 compared to left lateral). In the supine position, acid exposure time was 0.6 per cent (p = 0.022 compared to left lateral). Oesophageal acid clearance time showed an even clearer gradient: left lateral 35 seconds, supine 76 seconds (p = 0.030), right lateral 90 seconds (p = 0.002). The mechanism is anatomical: in the left lateral position, the gastro-oesophageal junction sits above the level of gastric fluid, reducing reflux. In the right lateral position, fluid pools directly at the junction.

Key fact: Sleeping on the left side reduces oesophageal acid exposure time to near zero and clears acid in 35 seconds. Sleeping on the right side allows acid to pool at the gullet junction and takes 90 seconds to clear it.

For anyone who takes antacid medication, sleeps propped up on pillows, or wakes with a burning sensation in the chest or throat, this finding alone is worth knowing about.

Your Heart: The Heart Failure Picture

The relationship between sleep position and heart health is both well-evidenced and nuanced. Szollosi and colleagues published a pivotal study in Sleep in 2006, analysing polysomnographic data from 20 consecutive heart failure patients with Cheyne-Stokes respiration, a form of central sleep apnoea characterised by cyclical breathing cessation that occurs in approximately 40 per cent of heart failure patients with reduced ejection fraction.¹⁹ With a total apnoea-hypopnoea index of 26.4 events per hour, these patients were significantly symptomatic.

The lateral position reduced the AHI in every single sleep stage. In Stage 2 sleep, events fell from 43.3 to 14.4 per hour (p less than 0.001). In REM sleep, they fell from 38.0 to 11.0 per hour (p less than 0.001). Oxygen desaturation associated with apnoeas improved from 4.7 per cent supine to 3.0 per cent lateral (p less than 0.001). The benefit is thought to arise from improved pulmonary oxygen stores rather than any direct effect on the upper airway, since Cheyne-Stokes respiration is a central rather than obstructive phenomenon.

A subsequent study by Joho and colleagues, examining 71 heart failure patients, found that a single night of positional therapy in patients with positional Cheyne-Stokes respiration reduced the supine sleep time from 57 per cent to 13 per cent and the AHI from 23 to 13 events per hour (p less than 0.05), with a trend towards reduced BNP (a marker of cardiac stress) from 323 to 243 pg/ml.^19a

There is, however, an important nuance. Research by Leung and colleagues, published in the Journal of the American College of Cardiology in 2003 and examining 75 heart failure patients alongside 75 controls, found that heart failure patients spontaneously and consistently avoided the left lateral decubitus position during sleep.²⁰ No such pattern was seen in healthy controls. Patients with larger left ventricular end-diastolic diameter, higher pulmonary capillary wedge pressure, and lower cardiac output showed the greatest avoidance of the left side. The proposed mechanism is that in a significantly enlarged heart, the apical heartbeat becomes palpable and uncomfortable when lying on the left side — a phenomenon known as trepopnea. The right lateral position is therefore generally better tolerated in heart failure, whilst still being substantially superior to the supine position for breathing.

Key fact: Heart failure patients with Cheyne-Stokes respiration show a reduction in breathing events from 43.3 to 14.4 per hour simply by turning from their back to their side.

Pregnancy: A Risk That Is Entirely Preventable

The relationship between maternal sleep position and late stillbirth is one of the most important findings in obstetric research of the past two decades, and it remains less widely known than it deserves to be.

An individual participant data meta-analysis by Cronin and colleagues, published in EClinicalMedicine in 2019, pooled data from five international case-control studies encompassing 851 stillbirths at 28 weeks of gestation or more and 2,257 controls.²¹ The analysis found that supine going-to-sleep position was associated with an adjusted odds ratio of 2.63 (95% CI 1.72 to 4.04, p less than 0.0001) for late stillbirth compared with falling asleep on the left side. Falling asleep on the right side was equally safe as the left (adjusted odds ratio 1.04, 95% CI 0.83 to 1.31). The population attributable risk calculation suggested that 5.8 per cent of late stillbirths could be prevented if every pregnant woman over 28 weeks of gestation settled to sleep on her side rather than her back.

Earlier case-control research by McCowan and colleagues in New Zealand, published in PLoS ONE in 2017 and involving 164 stillbirth cases and 569 controls across seven health regions, found a supine going-to-sleep position was associated with an adjusted odds ratio of 3.67 for late stillbirth overall, rising to 10.26 at term.^21a A UK multicentre study by Heazell and colleagues, examining 291 stillbirth cases and 733 controls across 41 maternity units, found an adjusted odds ratio of 2.31 for supine going-to-sleep compared to the left side.^21b

The mechanism was clarified by Couper and colleagues in an MRI-based study published in the Journal of Physiology in 2020, examining 20 healthy pregnancies at 34 to 38 weeks of gestation in both supine and left lateral positions.²² The supine position caused a 23.7 per cent reduction in total internal iliac arterial blood flow to the uterus (p less than 0.0001) and a 6.2 per cent reduction in oxygen movement across the placenta (p = 0.038). A separate study by Humphries and colleagues found a 16.4 per cent reduction in cardiac output in the supine position compared to the left lateral, accompanied by an 85.3 per cent reduction in inferior vena cava blood flow at its origin and a 220 per cent compensatory increase in azygos vein blood flow, confirming significant circulatory compromise.^22a

Key fact: Sleeping on the back in late pregnancy reduces blood flow to the uterus by 23.7% and oxygen transfer across the placenta by 6.2%, compared to sleeping on the left side.

The message is not that the right side is dangerous — it is not, based on current evidence. The message is that for pregnant women beyond 28 weeks of gestation, falling asleep on either side rather than on the back is associated with a meaningful reduction in the risk of late stillbirth.

Obesity and the Respiratory Burden of Lying Flat

Obesity and its most severe respiratory consequence, obesity hypoventilation syndrome (OHS), deserve separate consideration. OHS is defined by the combination of obesity (BMI at or above 30), daytime hypercapnia (elevated carbon dioxide levels), and sleep-disordered breathing without another identifiable cause. A comprehensive review by Masa and colleagues in the European Respiratory Review in 2019 established that OHS has an estimated prevalence of approximately 0.4 per cent of the adult population, with 90 per cent of those affected also having obstructive sleep apnoea.²³

The supine position is particularly damaging in obesity because the mechanical load of abdominal fat acts directly on the diaphragm and chest wall when the body is horizontal and facing upwards. In a study of nine severely obese patients (mean BMI 49.7) by Cheng and colleagues in 2023, dynamic compliance of the respiratory system in the supine position fell to 66 per cent of the upright baseline in those without OHS, and remained severely impaired in all positions in the two patients with confirmed OHS.^23a The intuitive reality is straightforward: lying on your back with a substantial abdominal load directly compresses the respiratory machinery. Turning to the side removes this gravitational burden.

Clinically, supine positioning in severely obese patients is now used as a diagnostic screen for sleep hypoventilation: measuring blood gases in the supine awake state can predict whether a patient will hypoventilate during sleep, serving as a simple and practical marker of risk.

What Happens After a Stroke

For the sake of completeness, it is worth noting the particularly troubling situation of stroke patients, for whom positional therapy in the acute phase is both most beneficial and most difficult to implement. A study by Brown and colleagues, published in Stroke in 2008 and examining 30 acute ischaemic stroke inpatients, found that the median percentage of total sleep time spent in the supine position was 100 per cent, and that 63 per cent of patients spent no time whatsoever in any non-supine position.²⁴ Of these patients, 73 per cent had obstructive sleep apnoea with an AHI of 5 or above. The more severe the stroke, the more time patients spent supine, and the worse the breathing. The paralysis and weakness caused by stroke forces supine sleep at precisely the moment when OSA, which is known to worsen neurological recovery, is most prevalent. Research by Camilo and colleagues confirmed that 66.7 per cent of 66 first-stroke patients spent their entire first night supine, with OSA present in 78.8 per cent.^24a

 

A Position-by-Position Summary of the Evidence

The table below consolidates the research discussed above. It is deliberately simple and should not be used as a clinical guide for any individual medical condition, but it provides a useful overview of the balance of evidence.

Condition or concern	Supine (back)	Left lateral	Right lateral	Recovery position	Prone (front)
Snoring and OSA	Worst	Good	Good	Best	Moderate
General airway and breathing	379% more arousals	Best	Good	Best	Moderate
Brain waste clearance (glymphatic)	Moderate	Best	Best	Good	Worst
Acid reflux	Moderate	Best (0% acid exposure)	Worst (90s clearance)	Good	Neutral
Heart failure and Cheyne-Stokes	Worst	Poorly tolerated in enlarged hearts	Best tolerated	Good	Limited data
Pregnancy (third trimester)	2.6x stillbirth risk	Best	Equally safe as left	Good	Avoid
Obesity and OHS	Worst	Good	Good	Good	Limited data
Low back pain	Best (lowest disc pressure)	Needs good support	Needs good support	Moderate	Worst
Shoulder pain	Best	Avoid affected side	Avoid affected side	Less loading than pure lateral	Moderate
Stroke (acute phase)	Worst	Good	Good	Good	Specialist guidance only
Raised intracranial pressure	Flat supine is worst	Moderate	Moderate	Head-elevated best	Complex

 

A Closer Look at Each Position

Back Sleeping (Supine)

Back sleeping is the position most commonly recommended by practitioners concerned with spinal alignment, and there is some justification for this in the context of low back pain. A systematic review by Saini and colleagues published in Musculoskeletal Care in 2025, examining six studies on sleep posture and low back pain, found that the supine position supports spinal alignment and was associated with lower LBP prevalence compared to prone sleeping.²⁵ Biomechanical measurements from intervertebral disc pressure studies, summarised in a review by Sewell and colleagues, suggest that the lumbar disc at L4/L5 sustains approximately 0.1 MPa of pressure in the supine position, compared to 0.12 MPa laterally and 0.3 MPa sitting. These are differences of some significance in a patient with significant disc pathology.

The problems with supine sleeping, however, are substantial and well-evidenced. It is the worst position for the airway, producing substantially more respiratory arousals even in healthy non-apnoeic individuals. It is unfavourable for glymphatic brain waste clearance. It significantly worsens reflux compared to the left lateral position. In pregnancy beyond 28 weeks, it is associated with reduced uteroplacental blood flow and an elevated risk of late stillbirth. In obesity, it mechanically impairs respiratory function. Back sleeping is the position most of us think we should be sleeping in, and it is the position that the biology least supports for the majority of people.

Key fact: 70.9% of men and 72.6% of women report sleeping mainly on their sides, according to a telephone survey of 812 people, despite most sleep advice pointing towards back sleeping.

Left Lateral

The left lateral position is, for most people and most conditions, the most beneficial sleeping position. It is optimal for acid reflux, protective in pregnancy, effective for OSA and snoring, and associated with the most efficient glymphatic brain clearance. Its principal disadvantages are a tendency to worsen trepopnea in heart failure patients with significantly enlarged hearts, and the potential for shoulder compression on the left side in those with existing left-shoulder pathology.

Right Lateral

The right lateral position is broadly comparable to the left for OSA and airway protection. It is equally safe in pregnancy. It is the preferred lateral position for heart failure patients because of the discomfort that can occur on the left with an enlarged heart. Its significant disadvantage is acid reflux: it is the worst sleeping position for those with GORD, producing acid clearance times two and a half times longer than the left lateral position.

The Recovery Position

The recovery position, used in first aid to protect the airway of an unconscious person, deserves recognition as something rather more than an emergency measure. It is, in effect, the optimal implementation of lateral sleeping: the body is in a semi-prone lateral position with the top knee drawn forward to prevent rolling supine, the lower arm extended, and the head in a slightly extended position that maintains an open airway. The neck is in gentle extension rather than flexion, which matters considerably for the airway.

In my book The Pillow Book, I argue that the first aid community rediscovered the recovery position through necessity rather than invention. When faced with the need to protect an airway without any mechanical devices, experienced practitioners converged on this position because it is what human anatomy actually requires. The evidence from Drug-Induced Sleep Endoscopy, showing near-complete resolution of tongue base obstruction in the lateral position, makes clear why it works. The recovery position is, essentially, lateral sleeping done properly.

The main practical challenge with the recovery position as a sleeping posture is comfort. The arm cannot remain fully extended throughout the night without going numb, and the exact positioning of the head requires support that a standard pillow typically fails to provide. These are engineering problems rather than biological ones, and they are solvable.

Prone (Face Down)

Prone sleeping has the fewest advantages of any position for general use. The systematic review by Saini and colleagues identified prone sleeping as associated with the highest risk of low back pain due to lumbar strain.²⁵ A study by Ylinen and colleagues, examining 375 consecutive patients with chronic low back pain, found that 42 per cent specifically avoided the prone position because of pain, making it the most commonly avoided position among this group.^25a The Lee et al. glymphatic research identified the prone position as producing the slowest brain waste clearance of all three positions studied.¹⁷ Prone sleeping decreases with age in the population at large, suggesting that the body naturally moves away from it over time.

There are niche contexts where prone positioning has documented benefits: in acute respiratory distress syndrome (ARDS), prone positioning significantly improves oxygenation in mechanically ventilated patients. Prone sleeping in premature infants in neonatal intensive care is associated with improved sleep architecture in that specific context. Neither of these represents a recommendation for healthy adults sleeping at home.

The Shoulder Problem: An Important Caveat

Side sleeping is not without its complications, and the most significant is shoulder pain. Research by Richards and colleagues, published in Arthroscopy, Sports Medicine, and Rehabilitation in 2024 and examining 58 patients with confirmed partial or full-thickness rotator cuff tears of insidious onset, found that 52 of 58 patients (89.7 per cent) were side sleepers, compared to just one back sleeper and four stomach sleepers, with a chi-squared test significant at p less than 0.0001.²⁶ A separate study by Çabuk and colleagues, comparing 87 patients with rotator cuff pathology against 93 healthy controls, found that 83.9 per cent of rotator cuff patients preferred the lateral decubitus position compared to 61.3 per cent of controls, a statistically significant difference at p = 0.003.^26a

These are associative findings, not causal ones, and it is worth noting that the causality may run in both directions: people who already have shoulder pathology may specifically choose to sleep on the affected side because of pain referral patterns, and people who have shoulder pain may be more likely to recall their sleeping position accurately. Nevertheless, the hypothesis that sustained lateral pressure on the shoulder increases subacromial pressure and may impair microvascular circulation in the rotator cuff is biologically plausible.

The key practical point is this: the problem is not lateral sleeping per se, but lateral sleeping without adequate shoulder support. Pure 90-degree lateral sleep places the full body weight through the dependent shoulder. The recovery position, with its semi-prone lean and forward knee placement, reduces this direct loading and allows the shoulder blade to sit more freely. A pillow designed to accommodate the shoulder rather than compress it further changes the biomechanics substantially. Anyone with existing shoulder pathology should sleep on the unaffected side rather than the affected one, with a pillow of appropriate height to maintain neck neutrality.

What Most People Actually Do When Left to Their Own Devices

It is worth stepping back from all of this to note what the observational data actually shows when researchers simply watch what people do during sleep. A five-year retrospective study by Yu and colleagues, examining 309 adults without sleep-disordered breathing who underwent full polysomnography with position monitoring, found that the most prevalent sleeping position was lateral at 59 per cent of total sleep time, followed by supine at 37 per cent and prone at 5 per cent.²⁷ A telephone survey of 812 people by Oksenberg and colleagues found that 72.6 per cent of women and 70.9 per cent of men reported sleeping mainly on their sides.²⁸

Key fact: In an objective sleep laboratory study of 309 adults, participants spent 59% of the night in the lateral position, 37% supine, and just 5% prone — without any instruction or advice about position.

The body, when not instructed otherwise, overwhelmingly chooses the lateral position. This is consistent across different sleep stages: Yu and colleagues found no significant difference in position between REM and non-REM sleep, meaning this is not a conscious preference but a deep physiological one.

The Oksenberg paper also noted that with advancing age, prone sleeping decreases and side sleeping increases. This is broadly consistent with what we know about the body's increasing need for airway protection as muscle tone diminishes with age, and about the increasing prevalence of reflux, cardiac, and respiratory conditions that are improved by lateral positioning.

What to Look for in a Pillow

A pillow is not a trivial purchase for a side or recovery position sleeper, and it is worth being precise about what the biology actually requires.

For Side Sleeping and the Recovery Position

The single most important measurement is height. A side-sleeping pillow must fill the space between the head and the mattress, which for most adults is between ten and fifteen centimetres depending on shoulder width, mattress firmness, and individual anatomy. A pillow that is too low causes the head to drop and the neck to flex laterally, loading the cervical joints asymmetrically. A pillow that is too high forces the neck into the opposite lateral bend. Neither is appropriate for sustained sleep.

Beyond height, the pillow must accommodate the shoulder. Standard pillows, being flat rectangles, do not account for the fact that the shoulder exists as a three-dimensional structure that needs somewhere to rest that is not compressed further between the head and the mattress. A pillow with a shoulder channel or recess, or one specifically designed to sit in the space between the head and the top of the shoulder rather than on top of the shoulder, addresses this problem.

Neck extension rather than neck flexion is required for airway protection, which means the pillow should support a slightly extended head position rather than one where the chin is drawn towards the chest. A pillow that is too soft, or that sags significantly under the weight of the head, will produce the latter.

Ear comfort over a full night is also a practical consideration that is regularly overlooked. Sustained pressure on the ear cartilage from a firm pillow surface produces real discomfort and is one of the most common reasons patients in my clinic have abandoned side sleeping after an initial attempt. Memory foam with adequate ear-area compliance, or a cut-out design, addresses this.

For those using CPAP for sleep apnoea, the mask creates an additional challenge in the lateral position, as standard pillows do not accommodate the device adequately and patients frequently dislodge the mask during the night, returning to the supine position as a result. A pillow designed with a CPAP mask recess in mind is more likely to maintain lateral positioning through the night.

Body stability matters too. Without something to prevent backward rolling from the lateral position into the supine during sleep, many people find themselves supine by the time they wake regardless of how they settled. The mechanism used in the recovery position to prevent this is the bent top knee. A pillow or wedge that provides gentle posterior support to the body can serve the same purpose.

For Back Sleeping (Where It Is Genuinely Appropriate)

If you have been assessed by a sleep specialist, confirmed not to have obstructive sleep apnoea, do not have significant acid reflux, and are not in the third trimester of pregnancy, there are circumstances where back sleeping is entirely reasonable. The most common is significant lumbar disc pathology, where the neutral spinal alignment achievable supine on a firm mattress is genuinely preferable.

For back sleeping, the mattress matters more than for any other position. A medium-firm pocket spring mattress provides the underlying structural support that the lumbar spine requires, preventing the sag that produces loss of the natural lumbar lordosis. Adding a memory foam topper of four to six centimetres thickness on top of the spring mattress gives the pressure redistribution across those five bony points that the body requires without sacrificing the structural integrity. This is a well-validated combination: a 2020 study by Lee and colleagues in the International Journal of Industrial Ergonomics, examining 40 healthy males on different mattress topper combinations, found that a soft 30mm topper produced significantly lower body pressure, better spinal alignment in the lower thoracic and lumbar regions, and lower paravertebral muscle activity than a hard topper.^28a

The pillow height for back sleeping should be substantially lower than for side sleeping, approximately seven to ten centimetres rather than twelve to fifteen, since the shoulder is no longer interposed between head and mattress. A pillow that is too thick for back sleeping will push the head forward into cervical flexion, which closes the throat and worsens snoring even in individuals who do not have OSA.

Anyone who snores, even occasionally and even mildly, should read this as a caution. Snoring is the sound of a narrowed and partially obstructed airway. In the supine position, it is substantially worse than in the lateral position for the biomechanical reasons described throughout this article. If you snore on your back and not on your side, the advice is clear: sleep on your side.

Conclusion: The Case for Rethinking What Is Natural

The cumulative evidence reviewed in this article points towards a consistent conclusion. For most of human evolutionary history, sleeping on a completely flat, uniform surface was not possible. The sleeping surfaces available to our ancestors, from the structured nests of great apes to the ground-level mats and natural contours used by hunter-gatherer populations today, encouraged lateral sleeping by virtue of their physical properties. The body's own pressure-point anatomy provides a further argument: sustained back sleeping on a hard surface is genuinely uncomfortable in ways that sustained lateral sleeping is not, because the contact points in the supine position are narrower, bonier, and closer to critical blood pressure thresholds.

The modern padded mattress has, quite inadvertently, made a physiologically suboptimal position comfortable enough to sustain through the night. The biological evidence is clear about what this has cost: worse airway function, less efficient brain waste clearance, more significant reflux, greater risk in pregnancy, and substantially compromised breathing in the setting of obesity or heart failure.

Key fact: For most of human evolutionary history, the flat sleeping surface that makes sustained back sleeping comfortable simply did not exist. Our biology evolved for a different surface entirely.

This does not mean that back sleeping is inherently pathological or that anyone who sleeps on their back is doing something wrong. It does mean that the assumption that back sleeping is the natural human default has been tested against the evidence and found wanting. The great apes lie on their sides. The hunter-gatherers sleep on surfaces that accommodate the lateral position. The body's pressure-point anatomy argues for the lateral position on any firm surface. The airway, the brain, the oesophagus, the pregnant uterus, and the overloaded diaphragm all function better in the lateral position.

The recovery position has been used in first aid training for generations precisely because it is the position that best protects the airway of a person who cannot protect it themselves. We teach it to schoolchildren as an emergency measure. It has not occurred to most of us to consider whether it might also be the best position in which to sleep.

The answer, as the evidence in this article suggests, is that it probably is.

 

My Side Sleeping Pro support system has been designed to help you sleep on your side at night.

References

1. Fruth B et al. Sleep and nesting behavior in primates: A review. American Journal of Physical Anthropology. 2018. 69 citations. https://consensus.app/papers/details/bc8ae10b14135d4093b4e2eb551adf3f/
2. Zamma K et al. Sleeping postures of captive chimpanzees. International Primatological Society Supplement. 2010. https://consensus.app/papers/details/58f50bc010d9585396df7b0d90e1937b/
3. Samson D et al. Orangutans (Pongo spp.) have deeper, more efficient sleep than baboons (Papio papio) in captivity. American Journal of Physical Anthropology. 2015. 29 citations. https://consensus.app/papers/details/72145757ea185988b5bd905e2fe61d9f/
4. Coolidge FL, Wynn T. The effects of the tree-to-ground sleep transition in the evolution of cognition in early Homo. Before Farming. 2006. 31 citations. https://consensus.app/papers/details/2385766cbb675856b7e4dc229d12bcc8/
5. Koops K et al. Terrestrial nest-building by wild chimpanzees (Pan troglodytes): implications for the tree-to-ground sleep transition in early hominins. American Journal of Physical Anthropology. 2012. 51 citations. https://consensus.app/papers/details/4d4836a62f785de4b134c27e5a28d864/
6. Yetish G et al. Natural sleep and its seasonal variations in three pre-industrial societies. Current Biology. 2015. 310 citations. https://consensus.app/papers/details/4630e72f6e2d587982a55f154e181599/
7. Samson D et al. Hadza sleep biology: Evidence for flexible sleep-wake patterns in hunter-gatherers. American Journal of Physical Anthropology. 2017. 76 citations. https://consensus.app/papers/details/2806fb43c81c5535bf50073174978cd8/
8. Garfin SR et al. Surface pressure distribution of the human body in the recumbent position. Archives of Physical Medicine and Rehabilitation. 1980. 20 citations. https://consensus.app/papers/details/ccf8eca0310452babdacf5c752ab727c/
9. Borges E et al. Pressure on bony prominences in the dorsal, lateral, and ventral decubitus: a clinical trial. Revista da Escola de Enfermagem da USP. 2025. https://consensus.app/papers/details/af6a84cfdd1156148c88ea6846bcfdd1/
10. Mohamadi P et al. Sex-specific variations in body pressure distribution across different sleep surfaces and positions. Ergonomics. 2025. https://consensus.app/papers/details/88176a17f0bb56059a076193377c5b77/
11. Landry SA et al. A review of supine position related obstructive sleep apnea: Classification, epidemiology, pathogenesis and treatment. Sleep Medicine Reviews. 2023. 28 citations. https://consensus.app/papers/details/583c11361b1c58c8a80fa1cec671b535/
12. Park J et al. Polysomnographic Phenotype of Positional Obstructive Sleep Apnea. Journal of Rhinology. 2024. https://consensus.app/papers/details/c91591d2db035335838e108f0b5e32b1/
13. Cheng WJ et al. Endotypic traits of supine position and supine-predominant obstructive sleep apnoea in Asian patients. European Respiratory Journal. 2024. 6 citations. https://consensus.app/papers/details/a992103b079452ef869ee6e30f31c213/
14. Mladoňová M et al. The role of positional changes in optimizing OSA treatment: evidence from DISE. European Archives of Oto-Rhino-Laryngology. 2025. https://consensus.app/papers/details/2979cd2cb5b9510497f28bf04dfc5de2/
15. Joosten SA et al. The effect of body position on physiological factors that contribute to obstructive sleep apnea. Sleep. 2015. 93 citations. https://consensus.app/papers/details/f9cf1b521796549f845944252c24a38f/
16. Rayward L et al. Sleep disruption and sleep position: Increased wake frequency in supine predicts lateral position preference. Journal of Sleep Research. 2024. https://consensus.app/papers/details/72919c6cbef1571fad906ab46b045c07/
17. Lee H et al. The Effect of Body Posture on Brain Glymphatic Transport. Journal of Neuroscience. 2015. 330 citations. https://consensus.app/papers/details/75e15dd11b535511b043669c89812cb3/

17a. Reddy OC, van der Werf YD. The Sleeping Brain: Harnessing the Power of the Glymphatic System through Lifestyle Choices. Brain Sciences. 2020. 188 citations. https://consensus.app/papers/details/75df74c8d3465aa19aec6eff40cdbb8b/

18. Schuitenmaker JM et al. Associations Between Sleep Position and Nocturnal Gastroesophageal Reflux. American Journal of Gastroenterology. 2021. 31 citations. https://consensus.app/papers/details/b119bfe5573f5a2b94bbdf86eaca42f7/
19. Szollosi I et al. Lateral sleeping position reduces severity of central sleep apnea/Cheyne-Stokes respiration. Sleep. 2006. 132 citations. https://consensus.app/papers/details/484f337539d358f5ad32afe10e227a53/

19a. Joho S et al. Impact of sleeping position on central sleep apnea/Cheyne-Stokes respiration in patients with heart failure. Sleep Medicine. 2010. 52 citations. https://consensus.app/papers/details/3b668940adfc56c9b9e7a81cae30b8f7/

20. Leung RST et al. Avoidance of the left lateral decubitus position during sleep in patients with heart failure: relationship to cardiac size and function. Journal of the American College of Cardiology. 2003. 58 citations. https://consensus.app/papers/details/3334d4d6f47f5ed8b72563d14fff8ca4/
21. Cronin RS et al. An Individual Participant Data Meta-analysis of Maternal Going-to-Sleep Position, Interactions with Fetal Vulnerability, and the Risk of Late Stillbirth. EClinicalMedicine. 2019. 60 citations. https://consensus.app/papers/details/2aa868e3972e5fcc80c234ee3cb8b04c/

21a. McCowan LME et al. Going to sleep in the supine position is a modifiable risk factor for late pregnancy stillbirth. PLoS ONE. 2017. 87 citations. https://consensus.app/papers/details/6af0cd525915578ea413247aa60d1340/

21b. Heazell AEP et al. Association between maternal sleep practices and late stillbirth. BJOG. 2017. 83 citations. https://consensus.app/papers/details/1df6b498f0c15dc49df2a9d9c483355f/

22. Couper S et al. The effects of maternal position, in late gestation pregnancy, on placental blood flow and oxygenation: an MRI study. Journal of Physiology. 2020. 29 citations. https://consensus.app/papers/details/239150ed7cb65adea74427c22050ef71/

22a. Humphries A et al. The effect of supine positioning on maternal hemodynamics during late pregnancy. Journal of Maternal-Fetal and Neonatal Medicine. 2018. 101 citations. https://consensus.app/papers/details/ee65c35e86c35345b6775baf2a26458e/

23. Masa JF et al. Obesity hypoventilation syndrome. European Respiratory Review. 2019. 178 citations. https://consensus.app/papers/details/c51d52edf42f52b0b9c33a4bd5febdca/

23a. Cheng M et al. Positional Changes in Respiratory Mechanics in Patients with Severe Obesity. Clinical Respiratory Physiology, Exercise and Functional Imaging. 2023. https://consensus.app/papers/details/6ae7afeb70b05ed88e7dbf81efdfb263/

24. Brown DL et al. High Prevalence of Supine Sleep in Ischemic Stroke Patients. Stroke. 2008. 45 citations. https://consensus.app/papers/details/84f48fbe768f5493ab4e482021d8d0dd/

24a. Camilo MR et al. Supine sleep and positional sleep apnea after acute ischemic stroke and intracerebral hemorrhage. Clinics. 2012. 18 citations. https://consensus.app/papers/details/616a9fe596485049a230daa6e22475ce/

25. Saini Y et al. Relationship Between Sleep Posture and Low Back Pain: A Systematic Review. Musculoskeletal Care. 2025. https://consensus.app/papers/details/1ce306e51c5f5edbb7d6a874977aa5af/

25a. Ylinen J et al. Preferences and Avoidance of Sleeping Positions Among Patients With Chronic Low Back Pain: A Cross-Sectional Study. Cureus. 2024. https://consensus.app/papers/details/3a72a2b5ca2858479f64a9bc2caab2f1/

26. Richards DP et al. Rotator Cuff Tears Are Related to the Side Sleeping Position. Arthroscopy, Sports Medicine, and Rehabilitation. 2024. https://consensus.app/papers/details/8da2a2b9f62c51588fe2a4bf36a71c83/

26a. Çabuk H et al. Effects of Sleeping Positions on the Rotator Cuff Pathology. Medicine Science: International Medical Journal. 2015. https://consensus.app/papers/details/37272f0bb35d5325a6d71b0e6a497b50/

27. Yu C et al. In what position do people sleep in the sleep laboratory? ERS Congress Proceedings. 2022. https://consensus.app/papers/details/abe430b6499f5c0c874ec5ae1faef5bb/
28. Oksenberg A, Silverberg DS. Breathe well, sleep well: avoid the supine and adopt the lateral posture. Sleep Health. 2016. https://consensus.app/papers/details/3b1c17ea2cc0594396ea8c5de1357f5e/

28a. Lee YC et al. Better combination of thickness and hardness of mattress topper for supine sleeping posture: A physiological measurements evaluation. International Journal of Industrial Ergonomics. 2020. 17 citations. https://consensus.app/papers/details/28308974aaf254ecb2d707f328f2ddae/

Vik Veer is a Consultant ENT Surgeon at the Royal National ENT Hospital and Queens Hospital, London. He is the author of The Pillow Book (2026) and the founder of iwantgreatsleep.com. His clinical and research interests focus on sleep-disordered breathing, positional therapy, and the relationship between sleep posture and upper airway function.