Chapter 60 Obstructive Sleep Apnea
Epidemiology, Risk Factors, and Pathophysiology
Obstructive sleep apnea (OSA) is a disorder characterized by repetitive collapse of the upper airway during sleep, resulting in changes in ventilation and intermittent hypoxemia and arousals, which may result in diurnal sleepiness and may lead to cognitive impairment and cardiovascular morbidity. The OSA syndrome is defined on the basis of recognition of symptoms (especially daytime sleepiness) and the objective measurement of disordered breathing during sleep.
Although obstructive sleep apnea clearly is a common disorder within the general population, its incidence is hard to establish, because methodologic differences among the various epidemiologic studies have made comparisons difficult. First, different tests have been used to diagnose OSA. Overnight polysomnography (PSG) is considered the “gold standard” diagnostic modality, but assessments have been made using other tests instead, such as unattended in-home PSG or respiratory polygraphy, pulse oximetry, and even clinical questionnaires (see Chapter 61). Second, variability in the definitions of different respiratory events (especially hypopnea) and the apnea-hypopnea index (AHI) cutoff value that defines OSA, or clinically significant OSA, is well recognized. In this sense, the chosen oxyhemoglobin desaturation threshold, typically 3% or 4%, used to define hypopnea can lead to different AHI scores; accordingly, estimates of disease severity will vary. Third, differences in sampling of populations (for example, the percentages of elderly and female subjects included) have been noted. Fourth, disparities in signal processing and a lack of standardization in the quantification of airflow (including thermistor, inductance plethysmography, and nasal cannula–pressure transduction) are common. Finally, the quality of validation of equipment and the conclusions of some studies have been questioned because of methodologic limitations such as small sample sizes or inadequate controls for potential confounding variables.
Apart from these differences, epidemiologic studies have focused on two levels of abnormal sleep quantification: OSA, when defined physiologically as increased obstructive breathing events (apneas or hypopneas) during sleep, usually with an AHI of 5 or more events/hour; and the clinical syndrome (the combination of an AHI of 5 or more events/hour and significant self-reported symptoms, especially daytime sleepiness). The real prevalence of OSA probably is underestimated, because a majority of studies used subjective sleepiness as the sole self-reported symptom for definition, although OSA is well known to be associated with other negative outcomes, such as sleep disruption, deterioration of quality of life, cognitive impairment, or cardiovascular disease, which also should be included in the definition of the syndrome.
It is estimated that 24% of men and 9% of women have OSA, as defined by an AHI of 5 or greater, and that 15% of men and 5% of women between 30 and 60 years have an AHI of 10 or higher. With use of daytime sleepiness as a clinical syndrome, the prevalence ranges between 3% and 7% in men and 2% and 5% in women in the general population (Box 60-1). Until recently, population-based epidemiologic studies of OSA were available only for North America, Europe, and Australia. However, more recent studies undertaken in other countries, including China, India, and Korea, report similar prevalence rates. The overall incidence of moderate to severe OSA (defined by an AHI of 15 or higher) occurring over a 5-year period is 11% and 5% in men and in women, respectively, which persists even after adjustment for confounding variables. This means that, even in the absence of any weight change, approximately 20% of men and 10% of women will develop moderate to severe OSA in that period of time.
Epidemiologic Features of Obstructive Sleep Apnea (OSA)
• 24% of men and 9% of women have OSA.
• 3% to 7% men and 2% to 5% women have OSA syndrome.
• 65% of older men and 56% of older women have OSA.
• 20% of older men and 15% of older women have OSA syndrome.
• In postmenopausal status, women reach similar incidence rates of OSA than men.
• The overall 5-year incidence of moderate to severe OSA ranges between 5% and 11%.
• Referral to sleep centers for further investigation is four- to eight times more frequent in men than in women.
Most population-based studies have found a two- to three-fold higher prevalence of OSA in males than in females. The ratio is even higher for men treated in sleep centers, with reported ratios between 4 : 1 and 8 : 1 or higher. This higher ratio may be the result of multiple factors: Women do not show the “classic” OSA symptomatology—they typically have more comorbid illnesses, use more psychoactive drugs in the absence of a correct diagnosis, and often present with vague, nonspecific symptoms, which widens the differential diagnosis and leads to a higher level of underdiagnosis or misdiagnosis of OSA. An important finding from epidemiologic studies is that gender disparities in prevalence seem to decrease with age, and when women reach postmenopausal status (and are not receiving hormonal replacement treatment), incidence rates for men and women become similar. Table 60-1 summarizes the most important sleep apnea prevalence studies in middle-aged populations.
The prevalence of OSA in adults increases with age as a result of greater collapsibility of the upper airway and probably reaches a plateau after the age of 65. It is estimated that 65% of older men and 56% of older women between 65 and 95 have OSA as defined by AHI of 10 or greater, and 26% of men and 21% of women between 71 and 100 years have an AHI of 30 or greater. Finally, 20% of older men and 15% of older women have the OSA syndrome (AHI of 10 or higher plus daytime hypersomnia). With age comes an increase in the frequency of both obstructive and central respiratory events. The main problem is to identify the AHI cutoff point that marks the limit of a physiologic and abnormal increase, in order to determine the real prevalence of clinically relevant OSA in the elderly population. In this sense, some investigators state that OSA represents different and distinct clinical entities in middle-aged adults and in older adults, based on morbidity and mortality data, although this position is controversial. Perhaps there is a more complex model of OSA that varies with the patient’s age.
In any case, the two proposed types of OSA consist of (1) a pathologic form of OSA that appears in middle age in those patients who usually are diagnosed in sleep laboratories and (2) OSA that appears after the age of 60 years, with some overlap between the two, mainly caused by physiologic changes (increase in pharyngeal collapsibility) associated with aging, and of less clinical importance. Data also suggest that the interaction between body weight and OSA in elderly persons may be different from that in younger adults. Because of the population-wide increase in longevity, the proportion of elderly persons being treated at sleep units also is increasing; currently one in four sleep studies are performed in patients older than 65 years of age. This scenario will present a scientific challenge in the future, in view of the lack of scientific evidence available on OSA in the elderly. Table 60-2 summarizes the most important epidemiologic studies describing the prevalence of OSA in older people.
Risk Factors for Obstructive Sleep Apnea
Obesity, aging, and male gender are the main risk factors for the development of OSA. It has been estimated that approximately 30% to 40% of AHI variance can be explained by genetic factors. Other risk factors have been proposed, related to increased anatomic or physiologic upper airway collapsibility.
According to recent estimates, 60% of adults in industrialized nations are overweight (body mass index [BMI] of 25 kg/m2 or higher) and at least 15% are obese (BMI of 30 kg/m2 or higher). Obesity is a common clinical finding and is present in more than 60% of patients referred for diagnostic sleep evaluations. Twin studies have shown that up to 70% of the variance in obesity within a population may be attributable to genetic factors. Although candidate genes for obesity are numerous, only a few single-gene mutations causally related to obesity have been convincingly detected, including the leptin receptor gene, the leptin gene, the pro-opiomelanocortin gene, the prohormone convertase 1 gene, and the melanocortin MC4 receptor gene, but very few studies on candidate genes associated with weight loss or weight gain have been performed, and no work in this area has been carried out specifically in the OSA population.
The prevalence of OSA in obese subjects is as high as 45%. Obesity explains 30% to 50% of the variance in AHI and is the only variable that can be modified. Several researchers have speculated that obesity and OSA may share a common genotype, and linkage analysis identified candidate regions at least on chromosomal arms 2p and 19p, but further studies are needed to confirm these results. Some major epidemiologic studies from around the world have consistently identified body weight as the strongest risk factor for OSA and have demonstrated a positive correlation between changes in OSA incidence and changes in weight over time. The Wisconsin Sleep Cohort showed that an increase of one standard deviation in BMI was associated with a four-fold increase in the prevalence of OSA. The longitudinal analysis component of the same study demonstrated that OSA severity changed approximately 3% for every 1% change in weight over a 4-year period. The Sleep Heart Health Study found that the change in AHI with weight changes over a 5-year period was more pronounced in men, and that the change was greater when associated with a weight increase than with a weight decrease. Corroborating this strong association, some studies have shown that dietary or surgical weight loss leads to reduced OSA severity in many patients and that OSA can even be completely cured in some cases. In other words, patients with mild OSA who gain 10% of their baseline weight are at a six-fold increased risk for progression of OSA, and an equivalent weight loss can result in a more than 20% improvement with respect to OSA severity.
Whether OSA predisposes affected persons to the preferential accumulation of visceral fat remains to be determined. More evidence is necessary to determine which specific measure of peripheral or central fat distribution is less compromising for OSA. Some investigators have shown that neck circumference, waist circumference, and BMI are independently associated with OSA severity at all ages, although it seems that neck size is the strongest predictor of sleep-disordered breathing, indicating that upper body obesity (fat deposition around the upper airway or fat deposited in the parapharyngeal fat pads), rather than a more generalized distribution of body fat, is important for the development of OSA. In any case, it is advisable to obtain information from all three measures in clinical practice.
The fundamental mechanisms by which obesity leads to an increase in the number of respiratory events during sleep, and weight loss decreases them, are unknown. It has been postulated that fat deposits surrounding the airway play a role in increasing the critical closing pressure. Other potential contributing factors include the reduction of functional residual capacity, ventilatory control system instability, alterations in neural compensatory mechanisms that maintain airway patency, and functional impairment of the upper airway muscles. Obesity also reduces chest wall compliance and increases whole-body oxygen demand, again predisposing affected persons to development of OSA. The degree to which common conditions associated with obesity, such as diabetes, may cause vascular or neuropathic damage to the dilator pharyngeal muscles and reduced upper airway sensation remains to be fully elucidated.
Aging in itself is associated with numerous physiologic changes, one of them being an increase in upper airway collapsibility, leading to a higher prevalence of OSA. The critical closing pressure of the upper airway in older people is −8.3 ± 2.3 cm H2O, whereas in middle-aged people it is −16 ± 6.9 cm H2O, independent of BMI. This difference may be due to various factors: an age-related decrease in the genioglossus response, an increase in airflow resistance, a decrease in upper airway dilation reflex mechanisms, reduced response to the stimulus of hypoxia, changes in the bony structure surrounding the pharynx, descent of the hyoid bone, increase in soft palate length, or an increase in pharyngeal fat pads. All of these changes lead to a generalized age-related decrease in the size of the upper airway lumen, specifically in men, and a consequent increase in airflow resistance. Age also is associated with an increased incidence of comorbid conditions, postmenopausal hormone status in women, dental disorders, decreased quality of sleep, and intake of psychoactive drugs. All of these factors may increase upper airway collapsibility. Finally, an additional factor that may predispose older persons to development of OSA is the aging-related increase in arousal frequency. Arousals from sleep lead to hyperventilation and relative hypocapnia, which in turn can promote respiratory instability and periodic breathing during subsequent periods of sleep onset.
It is not known which mechanisms explain the finding that OSA risk is twice as high in men as in women. Some studies have implicated several factors. First, men have increased fat deposition around the upper airways walls. Although data suggest that, in percentage terms, more women (33.4%) than men (27.5%) have a BMI ≥30 kg/m2, magnetic resonance imaging (MRI) studies have shown a decreased proportion of pharyngeal fat and soft tissue volume in the neck of obese women in comparison with obese men. Women in general have lower Mallampati scores, suggesting that the fat does not play as large a role in the female tongue as it does in the male tongue. Second, the upper airway in men is more prone to collapse as a result of its greater overall length with a longer vulnerable segment. Third, PSG characteristics of sleep and breathing patterns differ between women and men. Women tend to have a lower AHI in non–rapid eye movement (REM) sleep but have a similar AHI in REM sleep. Moreover, disordered-breathing events in women are of shorter duration and are associated with less oxyhemoglobin desaturation than in men. Finally, several mechanisms have been proposed to explain how male- or female-specific hormones would affect the propensity to develop OSA. One hypothesis is that the different hormones affect the distribution of body fat. Android body fat distribution (in the upper body and trunk) increases with both age and years after onset of menopause, which is a risk factor for the development of OSA. Hormone levels also have been hypothesized to affect central and neural respiratory control mechanisms. In this sense, progesterone has been shown to be a respiratory stimulant, which might protect premenopausal women from OSA; moreover, combined estrogen-progesterone treatment leads to a decrease in the number of apneic and hypopneic episodes during sleep. On the other hand, lower levels of testosterone may be protective against the development of OSA in women. Some researchers have shown that exogenous androgen therapy in men and women can aggravate OSA severity. It is possible that differing levels of hormones, starting from puberty and further modified by later maturational changes, can affect the development of OSA.