Chapter 102 Snoring

Abstract

Snoring is common, occurring in about 40% of the adult population. Snoring occurs when normal physiologic changes in pharyngeal mechanics during sleep become superimposed on underlying pharyngeal abnormalities present during wakefulness. Although snoring is the most common reason for referral to sleep laboratories, its main clinical significance is that it is a marker of obstructive sleep apnea syndrome and upper airways resistance syndrome. To date, there is no convincing evidence of a causal relationship between nonapneic snoring and daytime sleepiness or cardiovascular disease. The best way to treat nonapneic snoring is to modify the risk factors, including obesity and alcohol ingestion; body position training can be beneficial in selected cases. If snoring persists, oral appliances and continuous positive airway pressure are the noninvasive treatments of choice. Nasal lubricants, decongestants, and nasal dilators do not have proven efficacy. Surgical approaches may be warranted in patients whose snoring is problematic and persistent despite noninvasive treatments, but there is a need for better-quality studies using objective measurement of snoring as an endpoint.

Historical Perspective

Before the 1970s, snoring was thought to be simply an acoustic nuisance. However, with the recognition of obstructive sleep apnea syndrome, snoring began to assume a new importance as the cardinal symptom of this disorder. Investigators began to understand snoring and obstructive sleep apnea as a continuum of sleep-disordered breathing ¹ and question whether this acoustic nuisance in fact had serious health consequences. As polysomnography became more commonplace, objective evaluation of snoring evolved from a research oddity to an accessible, real-world tool. The 1990s led to the recognition that some nonapneic snorers have frequent respiratory arousals and sleep fragmentation, and the term upper airways resistance syndrome (UARS) was coined to describe these patients (see Chapter 105).

In conjunction with increasing scientific evidence regarding the health significance of snoring and public awareness surrounding the importance of sleep disorders, a panoply of treatment options for snoring has become available to patients. Today, snoring remains the most common reason for referral to most sleep laboratories.

Definitions

The International Classification of Sleep Disorders defines snoring as a “respiratory sound generated in the upper airway during sleep that typically occurs during inspiration but may also occur in expiration.”² The manual also lists a number of alternate names for nonapneic snoring, including “primary snoring,” “simple snoring,” “benign snoring,” “habitual snoring,” and “heavy snoring.”

Pathophysiology and Measurement of Snoring

Snoring is a sound produced by the vibrating structures of the upper airway. It is not caused by a single, localized abnormality within the airway. Endoscopic and imaging observations of the upper airway in snorers, during natural or induced sleep, demonstrate that any membranous part of the upper airway that lacks cartilaginous support can vibrate.³ This includes the soft palate, uvula, faucial pillars, pharyngeal walls, and the rest of the upper airway, almost to the level of the vocal cords. This diffuse involvement of the upper airway has made successful treatment of snoring more difficult, as many treatments address only a particular airway segment.

Several theoretical models of the human upper airway have been developed (see Chapter 101).⁴^–⁶ These models include the effects of airway wall compliance, gas density, and airway dimensions on the pressure–flow relationships in the upper airway. They predict that depending on the mechanical properties of the pharynx and the density of gas flowing through it, the walls of the upper airway can be stable (normal breathing), can vibrate or flutter (snoring), or can narrow or collapse completely (obstructive hypopnea or apnea, respectively). Some predictions have been verified experimentally. For example, alcohol ingestion promotes snoring by reducing pharyngeal dilator muscle tone and increasing airway wall compliance, in effect making the pharynx more floppy. On the other hand, breathing a lower-density, helium-containing gas mixture can improve snoring by promoting laminar over turbulent airflow, favorably altering the pressure–flow relationship in the pharynx and thereby reducing vibratory sounds.

Spectral analysis of the sound power spectrum can be used to determine the frequency components of snoring sounds. The power spectrum describes how much power is contributed by each constituent frequency component. Because snoring may be generated at several sites along the airway, and sometimes at multiple sites simultaneously, the power spectrum of snoring is wide, encompassing a range of frequencies up to 10,000 Hz. The spectral characteristics of snoring sounds are influenced by the oral or nasal breathing route, sleep stage, whether the sleep is natural or induced, posture, body weight, neck circumference, airway wall mass, and elasticity. The extent to which each factor is responsible for the power spectrum of snoring sound is not known.

Although sound recordings have become a routine component of nocturnal polysomnography, there is no standardized technique for recording these signals and reporting snoring. There is no consensus regarding the choice of instrumentation, where the instrument should be placed, the method of analysis of sound recordings, or the biologic calibration (i.e., validation that the measured sound is what a listener would identify as snoring). Apneas and hypopneas have standardized, objective definitions, and the distribution of such events across stages of sleep and body positions is well delineated in polysomnogram reports; in contrast, snoring is often only commented upon in general terms, rather than measured quantitatively.

Methods of measuring snoring vary,⁷^,⁸ but they generally include one or several of the following parameters: maximum sound intensity, percentage of sleep time spent above a threshold sound level, and number of sound intensity spikes above threshold level. However, these measurements are very specific to an individual laboratory’s configuration and are not standardized. In many laboratories, “measurement” of snoring simply consists of a sleep technologist’s subjective comment. It is therefore not surprising that many patients, particularly those referred only because of their snoring, leave the physician’s office rather disappointed: when they ask about their snoring, all they are told is that they do not have sleep apnea.

Until a uniform method for measuring, analyzing, and reporting snoring is developed, a combination of objective sound measurements and subjective assessment with questionnaires, preferably with input from the bed partner, should be used to characterize snoring. In our laboratory, we currently record snoring using a microphone suspended approximately 40 cm above the head. It is connected to a sound meter whose output is analyzed and is a part of the scored polysomnogram report. We report the maximum, mean, and minimum sound intensity, its distribution as a percentage of sleep position, and the histogram of sound intensity duration with respect to sleep time; the ambient noise in our laboratory is 40 dB, and we report percent of total sleep time spent in 5-dB bins starting at 41 dB, with the last bin being 76 to 80 dB.

Epidemiology of Snoring

Snoring is common, but estimates of its prevalence are imprecise and vary widely among studies. The prevalence of snoring ranges from 5% in men and 2% in women ⁹ to 86% in men and 57% in women.¹⁰ Although ethnic differences in snoring prevalence undoubtedly exist, most of the variability is very likely a reflection of differences in data-collecting methodology. Snorers rely on a bed partner or family member to describe their snoring. Such reports are inherently subjective and inconsistent. Furthermore, some studies do not even employ bed-partner questionnaires, using only the subject’s impression of his or her own snoring to determine the prevalence of the condition. Night-to-night variability further compounds the difficulty in interpreting reports of snoring. Not surprisingly, therefore, there is commonly a discrepancy between patient reports of snoring and objective snoring recordings. In one study, 15% of self-confessed snorers did not snore at all when studied in the sleep laboratory. On the other hand, 15% of those who reported never snoring actually snored for at least 50% of the night.

The male predominance observed in all prevalence studies of snoring remains unexplained. Possible contributing factors include differences in the perception and reporting of snoring between male and female bed partners,¹¹ hormonal factors influencing upper airway tone, and differences in pharyngeal anatomy and function. The apparent reduction in snoring with age, observed in some studies, may be related to the loss of a bed partner, altered perception of snoring, survival bias, or other methodologic problems.¹² A few special conditions appear to be particularly associated with snoring, such as pregnancy,¹³^,¹⁴ presence of nasal polyps,¹⁵ or nasal obstruction from other causes.¹⁶

It is clinically observed that snorers very often have a history of at least one other affected family member. Hence, genetic factors (see Chapter 103) can influence the prevalence of snoring, perhaps through obesity or through inherited oropharyngeal anatomy.

Health Effects of Nonapneic Snoring

The most commonly cited adverse health consequences associated with snoring are excessive daytime sleepiness and cardiovascular disease. However, the majority of studies evaluating the health effects of snoring fail to exclude obstructive sleep apnea and UARS with polysomnography. As a result, it is difficult to ascertain how much of the health risk is attributable to the snoring alone.

Snoring and Daytime Sleepiness

Some of the evidence for a relationship between snoring and daytime sleepiness comes from the Sleep Heart Health Study, a cross-sectional study of 5777 community-dwelling adults. Participants underwent full polysomnography at home. The authors showed a correlation between daytime sleepiness, as measured by the Epworth Sleepiness Scale (ESS), and self-reported snoring; the correlation was independent of the Respiratory Disturbance Index (RDI). The ESS increased progressively (reflecting increasing degrees of perceived sleep propensity during daily activities) with increasing snoring severity across the study population, from 6.4 in nonsnorers to 9.3 in the heaviest snorers.¹⁷ A more recent study also described a correlation between objectively quantified snoring and ESS.¹⁸ However, the overall correlation was quite weak, as snoring intensity and AHI together only explained 15% of the variance in ESS. Furthermore, the correlation was only present in patients with an AHI of more than 15 events/hour. This suggests that snoring is not a major cause of sleepiness in patients with mild OSA or nonapneic snoring.

Some patients classified in the past as sleepy, nonapneic snorers would probably now be reclassified as having UARS based on the presence of arousals related to respiratory effort (Videos 102-1 to 102-4). Others might now be reclassified as having mild sleep apnea, because of improvements in our ability to detect hypopneas. Clinicians should examine the polysomnogram data carefully for indicative findings. In the sleepy, truly nonapneic snorer without evidence of UARS, we would still hesitate to attribute the somnolence to the snoring alone and would recommend further workup for other causes of sleepiness.

Snoring and Cardiovascular Disease

Many authors have postulated that there is a relationship between snoring and hypertension. In 1975, Lugaresi and colleagues demonstrated that episodes of snoring were associated with acute, simultaneous increases in blood pressure.¹⁹ Many epidemiologic, questionnaire-based studies since then have demonstrated an association between snoring and hypertension. However, without available polysomnographic data it is unclear whether the hypertension is associated with nonapneic snoring, obstructive sleep apnea, or both. Several attempts have been made to clarify this issue. In support of the association between snoring and hypertension is the study of Bixler and colleagues,²⁰ who performed polysomnography in 1741 community dwellers and reported that when compared to nonsnorers, nonapneic snorers had an odds ratio of 1.6 (confidence interval [CI], 1.09 to 2.20) for having hypertension (defined either as taking antihypertensive medication or having elevated blood pressure measured at the time of polysomnography). On the other hand, Stradling and colleagues examined a cohort of snorers from general practice who were unlikely to have significant sleep apnea because they had fewer than two episodes of 4% drops in oxygen saturation per hour of sleep; they found no relationship between snoring and hypertension.²¹ From the available evidence, it would appear that the association between nonapneic snoring and hypertension is, if anything, modest.

Similarly, although epidemiologic studies have demonstrated an association between snoring and cardiovascular disease and stroke, none have convincingly demonstrated that snorers are at risk in the absence of apnea. Marin and colleagues compared long-term cardiovascular outcomes in nonapneic snorers (apnea–hypopnea index [AHI] less than 5 events per hour of sleep), subjects with obstructive sleep apnea, and population-based healthy controls matched for age and body mass index.²² They found that among the nonapneic snorers, there was no increased risk for cardiovascular death and no increase in the rates of nonfatal cardiovascular events. On the other hand, there is intriguing new evidence that snoring may be associated with carotid artery atherosclerosis. In a study of 110 subjects, mainly community volunteers who had nonapneic snoring or mild nonhypoxic obstructive sleep apnea (median AHI 11 events per hour; median 3% oxygen desaturation index 1.5 events per hour, 0% of total sleep time spent with oxygen saturation lower than 91%), the prevalence of carotid atherosclerosis on Doppler ultrasound increased significantly with snoring severity, independent of the AHI.²³ This relationship was not seen in the femoral artery, and the authors postulate that transmitted vibrations to the carotid artery might in fact lead to endothelial damage and atherogenesis.

Clinical Features and Diagnostic Assessment

Clinical History

The history is best obtained in the presence of the bed partner. A complete general medical and sleep history should be taken, but several key components to the snoring history require specific attention.

How bad is the snoring? The clinician should determine the loudness and frequency of the snoring, and whether it is positional. The time course of the snoring should be delineated. What was the age of onset of the snoring, and can this be correlated with the development of risk factors, such as weight gain? Is the snoring worsening or has it only become an issue since sleeping with a new bed partner? The history should also explore the effects of the snoring on the bed partner’s sleep, particularly if this is the reason for seeking evaluation. The degree of bed partner distress often affects the patient’s motivation to initiate and comply with therapy.

What risk factors for snoring are present? Weight gain should be identified. Alcohol consumption should be quantified, including the usual time of ingestion. Similarly, sedative medication use should be assessed. The clinician should determine if there are symptoms of nasal congestion and discharge or if there is a history of nasal trauma or surgery. A history of recurrent tonsillitis can alert the clinician to hypertrophy of the tonsils. A variety of other medical conditions can also cause oropharyngeal crowding including hypothyroidism, acromegaly, and congenital conditions such as trisomy 21, achondroplasia, and fusion of cervical vertebrae (Klippel-Feil syndrome).

Does the patient in fact have obstructive sleep apnea? Much of the clinical assessment of the snorer is directed toward determining the probability of obstructive sleep apnea or UARS. This is discussed in detail in Chapter 105. The clinician should enquire about nighttime symptoms of obstructive sleep apnea such as frequent awakenings, restless sleep, witnessed pauses in breathing, choking and gasping spells, diaphoresis, and nocturia. Daytime symptoms must also be assessed. Patients with obstructive sleep apnea can experience excessive daytime sleepiness, neurocognitive symptoms such as poor concentration and memory, and mood symptoms such as irritability and depression. Daytime sleepiness can be quantified using a standardized tool such as the Epworth Sleepiness Scale (see Chapter 143). Finally, the history should document medical conditions that may be associated with obstructive sleep apnea, such as hypertension and cardiovascular disease.

Physical Examination

The snoring history defines the problem and helps the clinician estimate the risk of obstructive sleep apnea; physical examination helps to further refine such estimations and identify potential sites for intervention.

The general physical examination should include measurement of blood pressure. Obesity may be assessed with the body mass index (BMI); regional obesity more specific to sleep-disordered breathing is documented with the neck circumference.

The nose should be inspected for evidence of rhinitis and other causes of obstruction, such as nasal polyps or a deviated nasal septum. The dimensions of the oropharynx should be noted. Is it small and crowded, with a low-lying soft palate? Is there retrognathia or micrognathia? Is there obstruction from hypertrophy of the tonsils? Is the tongue large? Some physicians report the Mallampati score, a numerical value based on the tongue size and position. The uvula may be inflamed and bulky in snorers, sometimes hardly lifting off the base of the tongue during phonation.

Although the ability to predict a patient’s response to treatment (particularly surgery or an oral appliance) based on physical examination of the oropharynx is an attractive one, at present there is no evidence that the presence or absence of specific features improves the success rate of treatment.

Clinical and Laboratory Investigations

Upper Airway Assessment

The upper airway can be assessed during wakefulness using various imaging techniques such as x-ray cephalometry, computed tomography, or magnetic resonance imaging. The results, however, are neither predictive of snoring nor helpful in the selection of surgical candidates, so these techniques are not indicated for the routine assessment of snorers. Only patients who might undergo maxillofacial surgery (a rare occurrence for treatment of nonapneic snoring) require radiographic imaging of the airway.

The anatomic structures of the upper airway can be visualized using fiberoptic nasopharyngoscopy, which is usually carried out by an otolaryngologist. Airway pressure measurements, simulated snoring, and Müller’s maneuver may be performed during endoscopic examination, but there is actually little evidence that such maneuvers can accurately determine the site of obstruction or predict surgical success. In fact, even when nasopharyngoscopy is done during sleep, the findings are a poor predictor of success in palatal surgery for nonapneic snoring.²⁴

Polysomnography

Figure 102-1 demonstrates some polysomnographic findings of nonapneic snoring. Snoring was measured in this instance by a snoring microphone, positioned above the patient’s head. Note the snoring in the bottom channel, in the absence of a reduction in flow, oxygen desaturation, or arousal.