About 80 % of OSAS patients are obese and obesity is an established risk factor for OSAS. A very tight relationship has been observed between body weight change and AHI: a 10 % weight gain has been shown to predict an approximate 32 % increase in the AHI, a 10 % weight loss predicts a 26 % decrease in the AHI, and a 10 % weight gain predicts a sixfold increase in the odds of developing moderate to severe OSAS [34]. Obesity causes upper airway narrowing as a result of excess fat in the (peri- and para)pharyngeal tissues [35]. Despite the strong relationship with obesity, it is important to remember that not all subjects who are obese or have a large neck circumference suffer from relevant sleep apnea and that one-third of OSAS patients are not obese [36, 37].

Several papers have shown a higher prevalence of OSAS in the elderly. In the Sleep Heart Health Study, it was shown that 25 % of males and 11 % of females in the age group 40–98 years had an AHI of higher than 15 events per hour [38]. However, daytime symptoms may be less common with advancing age [39]. The influence of male gender and BMI on OSAS tends to wane with age, and the overall prevalence of symptomatic OSAS seems to stabilise after age 65 years, while the age distribution of OSAS seems to increase regardless of age [9]. On the other hand, the age distribution of patients first diagnosed with OSAS generally peaks at the age of 50. Anyway, the high prevalence of sleep apnea in the elderly has led to a debate regarding its causes and consequences in older people [40, 41]. It could be suggested that older OSAS patients may be habituated to the added disease-related sleep disruption of OSAS and, therefore, do not suffer symptoms of daytime sleepiness in the same way as younger patients.

Epidemiological studies have reported that OSAS is much more prevalent in males. Possible explanations include the effects of hormonal influences affecting upper airway musculature and its ability to collapse, differences in body fat distribution and sex differences in pharyngeal structure and function [36].

A referral bias and gender differences in clinical presentation may have resulted in more males than females being diagnosed [42]. Males with OSAS are more likely to have symptoms of loud snoring, witnessed apneas or sleepiness.

Symptoms expressed by female OSAS patients are less typical and encompass insomnia, fatigue, morning headache and depression, while male bed partners are less likely to report snoring and apneas to their family physician as compared to their female counterpart. Therefore, female OSAS patients are less likely to be diagnosed and treated as compared to their male counterpart. Currently, there is increasing recognition that the disease is also prevalent in females, particularly after the menopause, and that the clinical manifestations may differ from those in males [43, 44].

The role of smoking as an established risk factor for OSAS remains controversial. Wetter et al. found a dose-response relationship between smoking and AHI, while smokers in the Sleep Heart Health Study displayed less sleep apnea than nonsmokers [45, 46]. These discrepancies remain unexplained, and there is a need to unravel the link between tobacco consumption and OSAS. Solid epidemiological studies related to chronic alcohol intake and sleep apnea are missing, due to a lack of reliable instruments for estimating alcohol use. Some population-based cross-sectional studies have reported a significant association between chronic alcohol intake and OSAS, whereas other cross-sectional or longitudinal studies did not [30, 47].

Although data are scarce, African-Americans and Asians appear to be at higher risk of developing OSAS than Caucasians [48]. This finding could at least partly be explained by differences in craniofacial structure [49].

Obstructive sleep apnea is mainly characterised clinically by loud snoring. It is often more cumbersome for the bed partner than for the patient himself and is often existing for a long time. Four or five loud snores followed by a silence (apnea) and another series of loud snores is a very suggestive description of a subject with obstructive apnea. The resumption of ventilation can be associated with loud stridorous breathing (“gasping”). Typically, these symptoms are more prominent in the supine position or after alcohol consumption. Occasionally, however, snoring may not be so obvious even in the presence of severe sleep apnea [52]. There is also a growing body of evidence that snoring might cause daytime sleepiness in the absence of OSAS [1]. This might be explained by the upper airway resistance syndrome that is characterised by episodes of increased respiratory effort followed by arousals and daytime sleepiness or by upper airway inflammation due to snoring-induced vibrations within the pharynx [53, 54].

Pathological daytime sleepiness is the second key symptom in the diagnosis of sleep apnea [51, 55, 56]. It is caused by loss of deep sleep, which presents 20 % of total sleep time in young subjects. The obstruction of the upper airways leads to the activation of the central nervous system, the so-called arousal, with sleep fragmentation as a consequence. A more fragmented sleep will result in a more sleepy patient during daytime. In general, patient and environment will not take much attention towards this daytime sleepiness, as far as it does not lead to repetitive car accidents or occupational accidents. Extreme sleepiness is characterised by falling asleep inadvertently during motor activity (talking, eating). Unequivocal sleepiness means falling asleep at rest, or while driving. The extension of normal diurnal sleepiness is considered as mild sleepiness. Tests have been developed to measure sleepiness more objectively [57]. The multiple sleep latency test (MSLT), maintenance of wakefulness test (MWT) and vigilance tests are used for this purpose. The use of the self-administered questionnaire, the Epworth Sleepiness Scale (ESS), which was validated with MSLT, enhances the accessibility of the quantification of EDS. There is however some controversy, since the ESS does not correlate well with the MSLT and other vigilance tests. Also the AHI only weakly correlates with quantified measures of sleepiness, which indicates that some individuals cope better with sleep fragmentation than others, or could be related to brain susceptibility and subjective perception of the consequences of hypoxemia [58]. Using a cut-off value of 18 events per hour including all apneas, hypopneas and flow limitation events, a sensitivity of 71 % and a specificity of 60 % for identifying subjects with excessive daytime sleepiness were obtained. When flow limitations are not taken into account, the sensitivity/specificity is even far worse [59]. Hence, for clinical purposes, it must be clear that the respiratory disturbance index is a more reliable parameter, but even then, its correlation with daytime symptoms also remains suboptimal.

In the Sleep Heart Health Study, also a weak correlation between the AHI and sleepiness was reported: the ESS only rose from 7.2 to 9.3 when the AHI changed from less than 5 to more than 30 [6]. It is however important to decide whether CPAP therapy should be instituted to treat daytime sleepiness. In case of sleep apnea, MSLT and vigilance testing are most often restricted to patients with persistent hypersomnolence despite adequate continuous positive airway pressure (CPAP) therapy or surgical or oral device therapy. On the other hand, routine use of these tests could be recommended for medicolegal reasons [60]. While tasks arising at regular intervals can even be performed at decreased levels of vigilance, fulfilling the criteria of light sleep, this is not the case for unexpected tasks and events. Therefore, up to 9 % of all traffic accidents are ascribed to sleepiness. In those car accidents in which sleepiness was the obvious cause, the rate of deaths was three times as high as in other accidents. Falling asleep while driving seems to be the cause of 3 % of the accidents causing material damage, 20 % of the accidents causing injuries and 50 % of the accidents causing death [61]. Not rarely, a recent accident can give occasion to consult a physician. A polysomnographic study revealed that socially disturbing loud snoring is associated with OSAS (AHI > 10) in 20 % of the patients referred to exclude sleep apnea. The combination of loud snoring with excessive daytime sleepiness reveals OSAS in 35 % of the referred patients [62]. The difficulty particularly in moderately severe OSAS is to identify EDS resulting from causes other than sleep apnea. Depression and mood disturbance are among the most important confounding factors [63–65]. Obesity alone may also interfere through adipokines and chemokines being activated even in the absence of OSAS. In OSAS, stress activation involving both the HPA axis and the sympathetic system may also play some role.

Abnormal motor activity, problems with maintaining sleep and awakening too early in the morning, nightmares, nocturnal dyspnea, nocturnal suffocation and nycturia are often reported. OSAS patients often sleep restless and turn and toss. Apneas are often associated with movements, which can sometimes be limited to mild movements, but abrupt more pronounced arm and leg movements with involuntary kicking and beating of the bed partner can occur [66]. Due to this nocturnal motor activity, OSAS patients often suffer from severe nocturnal transpiration. Nevertheless, the patient is often convinced that he/she is sleeping well at night and is unconscious of the breathing disturbances.

Detailed history taking of patient and bed partner can unravel the association between main symptoms and the sleep apnea syndrome: matinal headache (due to nocturnal CO₂ retention), being not refreshed in the morning, behaviour changes, decreased intellectual performance, depression, anxiety, automatic behaviour, social problems, marital problems, impotence (men), decreased libido, unexplained muscle discomfort and—last but not least—decreased quality of life. The absence of symptoms does not exclude OSAS. On the other hand, it is remarkable that, despite these typical symptoms, the interval between the first symptoms and the final diagnosis can often take some years.

It should be kept in mind that most studies have been performed in middle-aged subjects (range 50–60 years), clearly overweight male subjects with moderate to severe OSAS. Hence, the symptoms and neurobehavioural as well as cardiovascular and metabolic sequelae in OSAS mainly apply to this cohort. Nevertheless, OSAS can also have a large impact on daytime sleepiness and quality of life in the elderly (>70 years old). Surprisingly, history and daytime symptoms can be less specific in this age category [67]. A gender bias also takes place in the OSAS population, as discussed earlier.

OSAS is a serious health hazard being recognised as an independent risk factor for arterial hypertension, stroke, cardiac arrhythmias and coronary artery disease [7, 68, 69].

Observational cohort studies indicate that untreated patients with OSAS have an increased risk of fatal and nonfatal cardiovascular events, an increased risk of sudden cardiac death during the sleeping hours and a higher risk of stroke or death from any cause [7, 68]. According to He et al., the probability of cumulative 8-year survival was 0.96 for patients with an apnea index <20 and 0.63 for those with an apnea index >20. Difference in mortality related to apnea index was particularly true in the patients less than 50 years of age, in whom mortality from other causes is uncommon [118]. In the study of Marin et al., multivariate analysis, adjusted for potential confounders, showed that untreated severe OSAS significantly increased the risk of fatal (OR 2.87 [95 % CI 1.17–7.51]) cardiovascular events compared with healthy participants [7]. Treatment with tracheostomy or CPAP attenuated this risk [7, 118].

Since the group treated with CPAP received more intensive follow-up during the first year after diagnosis (two additional visits), outcome could be improved in this group independently of the CPAP treatment. Moreover, these results were only applicable to men. In patients with coronary artery disease but also in stroke, the occurrence of OSAS was a significant predictor of (early) death [119, 120]. These studies have the inherent limitation of lacking a randomised controlled design, which clearly limits the evidence level. In one prospective study, it was found that the apnea index was a predictor of excess mortality in the fourth and fifth decade, but not in the elderly [121]. In some population-based cohorts, a decrease in survival was reported with increasing OSAS severity, with an OR of 3.0 (95 % CI 1.4–6.3) (Wisconsin) and 1.46 (1.14–1.86) (Sleep Heart Health Study) in subjects with an AHI ≥ 30 compared with those with an AHI < 5 [122, 123]. However, after stratification by age and gender in the Sleep Heart Health Study, the OR remained only significant in males aged <70 years. Lavie et al. proposed a survival advantage in moderate OSAS, suggesting, as a potential mechanism, that chronic intermittent hypoxia during sleep may activate adaptive pathways in the elderly [124]. For example, older subjects have a reduced acute cardiovascular response to arousal from sleep, compared with younger people [125]. Hence, the poorer cardiovascular reactivity of older adults may, paradoxically, reduce the impact of arousals from sleep and protect against cardiovascular morbidity and mortality. However, it has to be reminded that the cardiovascular consequences of sleep apnea in older people may also be influenced by survival bias, as middle-aged, hypertensive OSAS patients may not survive into old age. Discrepancies among studies could potentially be explained by the heterogeneity of the patients included in the elderly populations.

A number of OSAS metabolic consequences have been identified. Some studies found increased insulin resistance and impaired glucose tolerance in OSAS patients, independent of body weight [126–128], and a worsening of insulin resistance with increasing AHI [129]. However, other investigations failed to demonstrate an independent effect of AHI owing to the major impact of obesity [130]. In cohort of the general population, both the Wisconsin Cohort Study and the Sleep Heart Health Study have identified OSAS as an independent risk factor for insulin resistance, after adjustment for potential confounding variables, such as age, sex and BMI [131, 132]. However, subjects with an AHI ≥ 15 did not differ significantly from those with an AHI < 5 when it came to the risk of developing diabetes over a 4-year period (OR 1.62 [95 % CI 0.7–3.6]), after adjustment for age, sex and BMI [132]. Intermittent hypoxia and sleep fragmentation are thought to play a key role in the development of metabolic disturbances through the activation of the sympathetic nervous system and pro-inflammatory pathways. It was shown that even lower-grade desaturations are strongly linked with glycaemic status abnormalities [133]. Furthermore, the increase in risk for developing diabetes overtime is increased in OSAS independent of the decrease in arterial oxygen saturation [134]. Sleep fragmentation due to cortical and automatic arousals accounts for alterations in sympathetic/parasympathetic activity, and the simultaneous presence of metabolic syndrome and OSAS further increases sympathetic activity and worsens glycaemic control, even after adjustment for body mass index [135]. Alteration in glucose metabolism and sympathovagal balance has been observed previously in normal subjects following two nights of experimental sleep fragmentation [136].

It comes as no surprise that OSAS with the consecutive sleep disturbance has detrimental effects on cerebral function. Memory, attention and learning ability have been reported to be abnormal in some OSAS patients [137–140]. Both “lower-level” processes (arousal and alertness) and “higher-level” cognitive processes (e.g. executive attention) have been disturbed, including the ability to inhibit inappropriate behaviours and thoughts, regulate attention and plan and organise for the future [50, 141, 142]. Specific cognitive impairments (like thinking, perception, memory, communication or the ability to learn new information) are present in 76 % of OSAS patients [143]. Studies using functional MRI or PET indicate sleep loss as the primary cause of neurocognitive deficits, mainly a basal slowing in information processing, more so than hypoxemia [142]. Whether there are not only functional but also anatomical sequels of OSAS is more difficult to evaluate, since the profound effects of decreased alertness on higher cognitive functioning mimic cerebral damage due to hypoxia. OSAS can promote axonal dysfunction or loss, as well as myelin metabolism impairment in the frontal periventricular white matter, which causes cognitive executive dysfunction [144]. A therapeutic challenge with CPAP can learn that not all cognitive function alterations reverse, and they may represent neuronal damage. Moreover, when cognitive function is preserved in OSAS patients, functional brain imaging has revealed that brain activation is increased, compared with the activation that occurs in healthy controls performing the same task. The association between preserved cognitive function and greater activation in OSAS patients suggests that increased cerebral recruitment (overrecruitment) is required to maintain cognitive performance [145, 146]. Currently, it is also well established that cognitive and attentional deficits may occur in OSAS in the absence of perceived subjective EDS [147]. It questions the sensitivity of the tools used for evaluating EDS. Moreover, attentional deficits may occur without objective EDS [147]. Reaction time and sustained and divided attention tasks may be altered in the absence of sleepiness. It should further be studied whether these attentional deficits may impair driving ability and other social and professional abilities. If confirmed, it raises the question of who should be treated, since the absence of perceived sleepiness would then not be required for treating OSAS.

Obesity is linked with OSAS. According to various data, the simultaneous occurrence of obesity and OSAS is approximately 35 %, and their mutual relationship is reciprocal. Deposition of fat even in the neck area, especially in men with central type of obesity, contributes to upper airway obstruction and to easier development of apneas during sleep.

Controversy remains as to whether specific anthropometric indices of body habitus, such as neck or waist circumference, are better predictors of OSAS as compared with BMI alone [36, 149].

However, despite the strong relationship with obesity, it is important to remember that not all subjects who are obese or have a large neck circumference suffer from sleep apnea and that some one-third of OSAS patients are not obese [36].

Several large population-based cross-sectional studies have reported an independent link between arterial hypertension and OSAS, when controlling for multiple potential confounding variables. OSAS and systemic hypertension commonly coexist: OSAS is present in at least 30 % of hypertensive patients, while about half of OSAS patients suffer from systemic hypertension [88]. In patients with refractory hypertension, up to 85 % has OSAS. Currently, OSAS is also considered a risk factor in the hypertension management guidelines [150, 161].

Studies have suggested an increased prevalence of OSAS as well as CSA in patients with CHF, even in patients with asymptomatic left ventricular dysfunction, between 20 and 50 % [152]. A failing heart with reduced left ventricular ejection fraction, diastolic dysfunction and increased filling pressures is more vulnerable to stressors such as increased blood pressure (afterload) or sympathetic activation as compared with a healthy heart. It explains why OSAS is particularly detrimental in patients with established heart failure. These data imply that strategies to recognise sleep apnea are highly warranted in CHF patients. Predictive factors for the presence of CSA were male gender, the presence of atrial fibrillation, daytime hypocapnia or age >60 years. The predictive factors for OSAS include an increased BMI (BMI > 35 kg/m²) in men and age >60 in women [162]. Surprisingly, the majority of patients with CHF and comorbid OSAS do not complain of excessive daytime sleepiness, possibly owing to chronically elevated sympathetic activity [163].

In a study comparing patients with atrial fibrillation (AF) (n = 151) to patients with no history of AF in a general cardiology practice (n = 312), Gami et al. demonstrated that the proportion of patients with OSAS was significantly higher in the patients with AF than in the patients without AF (49 % vs. 32 %, p = 0.0004) [154]. The adjusted OR for AF was 2.19 in OSAS subjects (p = 0.0006). In a multivariate analysis, BMI, neck circumference, hypertension, diabetes mellitus and AF remained significantly associated with OSAS, and the OR was largest for AF. A higher recurrence of atrial fibrillation after cardioversion has also been reported in those with untreated OSAS compared to those without OSAS [164]. Therefore, patients referred for the evaluation of significant tachyarrhythmia or bradyarrhythmia should be questioned about symptoms of sleep apnea.

There is a high prevalence of OSAS among patients with angiographically proven coronary artery disease and an increased incidence of coronary artery disease in patients free of coronary symptoms at the time of OSAS diagnosis [155]. Data suggest that patients with coronary artery disease should be particularly questioned for symptoms and signs of sleep apnea. If there is even the slightest suspicion of sleep apnea, the patients should undergo a polysomnography [162].

OSAS is common in many endocrine and metabolic conditions, like insulin resistance, type 2 diabetes and the metabolic syndrome [51, 156, 165]. Given the high prevalence of OSAS among these patient categories, specialists treating metabolic disorders and abnormalities should consider investigation for OSAS in their patients.