Introduction

The population pyramid is changing; it is expected that in the future decades, the percentage of older people will be higher and without severe disability [1]. Obstructive sleep apnoea (OSA) is one of the most common respiratory disorders, and it is well known that its prevalence increases with increasing age. Therefore, more than 24% of sleep studies are currently performed in elderly subjects. However, there is little knowledge on the impact of OSA in elderly individuals and the effects of treatment in this age group.

OSA prevalence has been found to be higher in the elderly than in the younger age groups, and existing studies suggest that clinical presentation and consequences could vary with age, being less symptomatic and with lower OSA impact at advanced age. However, there are few data available on this topic, and the results are not always conclusive. Moreover, there are other aspects that should be established in this age group concerning diagnosis and treatment. For example, OSA severity is classified according to the apnoea-hypopnea index (AHI) [2], and there is a debate on whether this parameter could be adequate for elderly individuals and whether the current cut-off values applied to younger patients correctly differentiate between physiological and pathological events in the older group. Moreover, OSA screening questionnaires remain to be validated in elderly patients. Regarding treatment, some studies indicate that it is well tolerated, and that adherence and compliance do not differ significantly from those of younger patients; nevertheless, the results are not always concordant. Moreover, despite the high number of elderly patients with OSA, there are few studies centred on this age group; recommendations have been extrapolated from studies performed in middle-aged adults.

In summary, there is no doubt that OSA is a prevalent disorder in the elderly, but there is scarce scientific evidence. Therefore, this article aimed to analyse the characteristics and impact of OSA in elderly individuals and to review the evidence on the effects of CPAP treatment in this age group.

Changes in Sleep and the Upper Airway with Ageing

Ageing decreases the minor negative pressures needed to produce upper airway collapse, increasing collapsibility [3] and therefore the probability of recurrent apnoeas and hypopneas. The aetiology of obstructive sleep apnoea (OSA) in elderly individuals is frequently multifactorial. On the one hand, there are anatomical factors (fat deposition in the upper airway or loss of teeth), neuromuscular factors (pharyngeal muscle dysfunction) and neuroventilatory factors (changes in sleep architecture) [3]. On the other hand, the aetiology may also be related to comorbidities such as stroke, dementia, oedema in the lower extremities or heart failure.

There are some physiological changes associated with ageing that can facilitate upper airway obstruction, such as an increase in the percentage of superficial sleep and a decrease in deep sleep [4] or an increase in the number of arousals. Some data suggest that an increase in OSA incidence in the elderly could be related to not only the physiological changes associated with ageing but also the increase in central apnoeas associated with greater respiratory instability during sleep. In addition, recent evidence has emerged that dysfunction of the genioglossal nerve and muscle in elderly subjects may have implications for OSA pathogenesis [5]. Moreover, it is important to consider the relevance of polypharmacy in sleep breathing disorders at advanced age and the hormonal changes and obesity (especially around the upper airway) [6] associated with menopause that could also facilitate airway obstruction [7, 8].

However, it is important to note that any of the factors described has been clearly identified as a main cause, and collapse probably depends on the interaction of several factors.

Prevalence of OSA in Older Adults

According to the American Academy of Sleep Medicine (AASM) criteria, in populations over 30 years old, the prevalence rates of mild and moderate-severe OSA are approximately 35.2% and 16.2%, respectively [9]. Higher values have been reported in certain population groups, such as those placed in nursing homes, home dwellings or hospitals [10, 11].

The prevalence of OSA in the elderly population is higher than that in the middle-aged population, and it could be related to the increased airway collapsibility observed in advanced age [12]. It has been described that the number of respiratory events increases with age [13], and OSA syndrome has been found in 18% of men and 7% of women aged 50–70 years. Moreover, a recent study reported that approximately 32% of subjects aged ≥ 60 years have moderate to severe OSA [14].

Male sex, obesity and old age are independent risk factors for OSA [7]. OSA in middle-aged individuals is more prevalent in males, but as the prevalence of OSA in females increases after menopause, the prevalence in both sexes is similar at advanced age.

Clinical Picture of OSA in the Elderly

The clinical presentation of OSA in older adults is different from that in younger adults. Compared to middle-aged adults, snoring and witnessed apnoeas are less frequent because the bed partner can be an older person or may sometimes be absent [3]. Moreover, sleepiness could be perceived as a normal symptom at advanced age, and frequently, there are other causes of somnolence that make it difficult to relate somnolence to OSA syndrome. Thus, in general, elderly individuals are less symptomatic, and anthropometric data have a more limited predictive value at advanced age.

It is important to note that OSA in the geriatric population could be associated with less typical symptoms, such as nocturia, neurocognitive symptoms, functional deterioration with frequent falls [15], sarcopenia [3] or the presence of an excess of cardiovascular events [16, 17].

For more detailed information regarding differences in OSA between age groups, see Table 1.

Table 1 Comparison between OSA characteristics by age group
Full size table

Screening and Diagnosis

The gold standard for diagnosing OSA is overnight polysomnography in a clinical sleep laboratory. Simpler diagnostic tools that can be performed at home, with easier management and lower costs, have been developed and validated [18] in middle-aged individuals. However, their use is more limited in elderly subjects because of the presence of comorbidities, ingestion of psychotropic medication or other sleep disorders such as insomnia or changes in sleep efficiency and architecture [4].

Moreover, screening seniors via questionnaires to detect OSA is problematic because they have not been validated in this age group. Recent studies have evaluated the diagnostic performance of the STOP-BANG and new modifications of this questionnaire called STOP-28 and STOP in older individuals [19]. The authors concluded that they have high sensitivity but low specificity. Moreover, some questionnaires, such as the Epworth Sleepiness Scale (ESS), have not been validated in elderly subjects [20]; therefore, they should be used with caution.

A controversial aspect in the diagnosis, given the high number of respiratory disorders observed in the elderly, is establishing the cut-off point for the apnoea-hypopnea index (AHI) in this age group. This threshold should distinguish between age-related respiratory events (probably physiological) and those that will have harmful consequences for the patient. In this sense, Bliwise et al. [21] suggested that a heuristic model of OSA could exist. On the one hand, there could be a type of OSA that appears at middle age and should be considered pathological. On the other hand, age-dependent OSA is more related to the ageing process, which could be less harmful. However, despite the debate in recent years, no specific diagnostic criteria are available for this population group.

Impact of OSA in the Elderly

Mortality

There are few studies evaluating the mortality of OSA specifically in elderly subjects, and the results are contradictory. Thus, Martinez-Garcia et al. [16] performed a prospective observational study in severe OSA in subjects older than 65 years and observed an increase in cardiovascular mortality. In contrast, Johansson P et al [22] and Mant A et al. [23] did not find an increase in mortality in elderly individuals with OSA.

If we look at studies that do not exclude subjects aged ≥ 65 years, it seems that in elderly subjects, OSA does not have the same impact on mortality as in young subjects. Lavie et al. [24] showed that moderate to severe OSA was associated with an excess mortality rate only in subjects aged < 50 years but not in elderly individuals. Moreover, Punjabi et al. [25] concluded that OSA is associated with mortality, particularly in men aged 40–70 years. In addition, Lavie et al [26] reported lower mortality than expected in the general population of the same age in elderly subjects with moderate OSA. The authors attributed this result to the presence of adaptive mechanisms and described the “hypoxic preconditioning hypothesis”. Nevertheless, it is important to note that differences in mortality should be related to methodological limitations of the studies or related to a survivor bias in patients with OSA and advanced age [27].

Cardiovascular Morbidity

The relationship between OSA and cardiovascular disease (CVD) in middle-aged subjects is well known; however, there are scarce data in older populations. Obstructive sleep apnoea has been associated with endothelial dysfunction [28], ventricular extrasystole [29] and hypertension [30, 31] in some studies, while no association with cardiovascular disease [22, 32] was found by other authors.

The major determinant of cardiovascular morbidity in relation to obstructive sleep apnoea is chronic intermittent hypoxia (CIH), and prior studies have shown that CIH induces changes in the structure of cardiovascular tissues, resulting in cardiovascular diseases [33, 34]. Some studies in animal models have evaluated whether there is a differential effect of OSA on cardiovascular tissues depending on age. A recent study by Castro-Grattoni AL et al. concluded that cardiovascular remodelling induced by severe chronic intermittent hypoxia is affected by age, with more pronounced deleterious cardiovascular effects in younger populations [31].

It is important to note that some of the patients included in studies evaluating OSA, hypertension and cardiovascular diseases are older than 60 years, and the associations observed appear to be greater in younger patients [33, 34]. Thus, the available evidence suggests that cardiovascular consequences could be less common in elderly individuals; nevertheless, evidence is scarce and not conclusive.

Cerebrovascular Morbidity

Clinical studies suggest a bidirectional relationship between cerebrovascular disease and OSA; therefore, suffering a stroke would make the presence of respiratory disorders more frequent during sleep, and having OSA could influence the risk and evolution of cerebrovascular disease [35]. OSA has been associated with a higher prevalence of lacunar or silent infarcts [36] and has been identified in elderly patients as an independent risk factor for incident stroke [37].

Neurocognitive Morbidity

OSA has been associated with negative effects at the cognitive level in experimental and clinical studies. The most frequent cognitive deficits are found in the attention and executive functions domains, while the effects observed in memory have been minor. Domains such as language or global intellectual function are relatively preserved.

The prevalence of OSA in Alzheimer’s patients is more than 45% [38], which is higher than that in populations of the same age with preserved cognitive function. Some authors have reported that mild cognitive impairment and Alzheimer’s disease are associated with breathing disorders [38, 39]. There is a plausible biological link between OSA and cognitive impairment [40,41,42], and it has been described that OSA could provoke disruptions in the amyloid process and the clearance of tau protein. It has been reported that increasing the apnoea-hypopnea index increases the risk of cognitive decline [43], and recent studies have shown a link between neurodegenerative biomarkers (Aβ40, Aβ42, total Aβ and P-tau 181) and OSA [44,45,46].

The results are contradictory regarding the effects of OSA on neurocognitive parameters and disorders. Longitudinal studies have suggested that OSA could be a risk factor for the development of cognitive impairment and dementia [47] due to its onset at an earlier age [48]. However, not all studies have reached such conclusions. Thus, a recent study showed that in patients with mild-moderate Alzheimer’s disease, OSA was not associated with greater cognitive decline after 12 months of follow-up [38].

Cancer

Multiple authors have described a higher incidence of cancer in patients with OSA [49, 50], although others have not found this association. It has been reported that OSA increases malignant tumour properties in breast, kidney and uterine cancers and melanoma [51]; nevertheless, the incidence rates of lung and colorectal cancer attributed to OSA are uncertain [49].

The mechanisms behind an increased risk of cancer and the prevalence of OSA are not completely understood, but intermittent hypoxia and sleep fragmentation have been proposed as key factors [52]. Studies in animal models have shown that intermittent hypoxia and sleep fragmentation can favour tumour growth and invasiveness. Nevertheless, it has been described that the effects of intermittent hypoxia could be modulated by age [53], with lower growth rates at advanced age. Some authors have reported that intermittent hypoxia can alter the immune response in elderly individuals and plays a pivotal role in lung adenocarcinoma progression and aggressiveness [51].

Several observational studies reported an association between OSA and cancer mortality [54], and data from experimental studies showed that there is biological plausibility to explain the link between both disorders and suggest that the effect seems lower at advanced age. However, the studies carried out have important limitations and are not conclusive.

CPAP Treatment in Elderly Patients with OSA

Even though there are a large number of elderly individuals under CPAP treatment, there is little evidence on the effectiveness of this treatment in elderly subjects. Treatment indications are mostly based on the results of studies performed in young adults or in studies including both young and old age groups, with very few studies centred on elderly subjects. Although scientific evidence on the effects of CPAP treatment focused exclusively on elderly individuals is scarce, some observational studies and clinical trials focused on elderly individuals have been published in recent years.

Observational Studies

There are few long-term observational studies performed exclusively in the elderly population, and they are frequently centred on disorders more prevalent at advanced age, such as cognitive disorders or cerebrovascular disease.

Cognitive Disorders

Most of the studies have been performed in subjects with cognitive deterioration or dementia, with results suggesting that CPAP treatment can improve cognitive function and even delay cognitive decline in elderly patients [48, 55, 56].

Studies performed in subjects with preserved cognitive function are much more limited. Aloia MS et al. [57] observed that CPAP treatment produces an improvement in neurocognitive variables after 3 months of CPAP treatment, and Crawford-Achour E. et al. [58] demonstrated that CPAP treatment is associated with the maintenance of memory performance.

Cardiovascular and Cerebrovascular Disease

Martinez-Garcia MA et al. [16] observed a decrease in cardiovascular mortality in patients with good CPAP compliance. Observational studies by López-Padilla et al. [59] (performed in subjects aged ≥80 years) and Ou et al. [60] (including patients with a mean age of 77.8 years) also presented better survival in those with good compliance with CPAP.

Moreover, studies carried out in subjects with stroke reported significant protection against the recurrence of new vascular events [61] with CPAP treatment.

Randomized Controlled Trials

Before 2014, data from only a few trials undertaken exclusively in subjects aged over 60-65 years were available, and those trials were mostly centred on evaluating the effect of CPAP treatment in cerebrovascular diseases.

Hsu CY et al [62] performed a study in 30 patients aged 65–81 years with stroke and severe OSA who were randomized to CPAP or conservative care for 8 weeks. The authors unveiled no effect of CPAP treatment on neurological performance, quality of life or sleepiness. Zhang X et al [61] studied 41 patients aged 60–74 years with moderate-severe OSA and did not find deterioration in analytic variables with respect to healthy controls. Parra O et al [63] conducted a study in 140 patients (aged 64.7 years) with ischaemic stroke and an AHI>20/h who were randomized to receive CPAP or conventional treatment. The authors concluded that nasal CPAP seems to accelerate neurological recovery and delay the appearance of cardiovascular events, although they did not find improvements in patient survival or quality of life.

Since 2014, four randomized clinical trials focusing exclusively on elderly subjects have been published. For more detailed information, see Table 2.

Table 2 Randomized clinical trials consisting exclusively of elderly patients
Full size table

The first published study was the PREDICT trial by McMillan A et al. [64]. The PREDICT was a 12-month, multicentre randomized trial performed in the UK that included 278 patients older than 65 years with newly diagnosed OSA who were randomized to receive CPAP plus supportive care or supportive care alone. The authors concluded that CPAP reduces sleepiness over 12 months compared with supportive care alone. No differences were observed in mood, cognitive function, cardiovascular events or health care costs. It is important to note that subjects included in the PREDICT trial have low adherence, with a mean CPAP compliance lower than 2 h/night.

Dalmases M et al [65] performed a single-centre randomized pilot study in 33 patients aged ≥ 65 years old with severe OSA treated with either conservative care or CPAP for three months. They observed that patients treated with CPAP presented significant improvements in some cognitive domains (episodic and short-term memory, speed of mental processing and mental flexibility). Moreover, neuroimaging revealed an increase in connectivity in the right middle frontal gyrus after 3 months of CPAP treatment and a higher percentage of cortical thinning in the conservative care group. No association was seen between cognition and brain functional connectivity changes within the default mode network. The mean CPAP compliance was 6 h/night.

Martínez-García MA et al. [66] performed an open-label, multicentre clinical trial in a consecutive clinical cohort of 224 elderly patients (aged ≥70 years) with severe OSA who were randomized to receive CPAP or no CPAP for 3 months. The authors concluded that CPAP treatment was effective in improving quality of life (evaluated with the Quebec Sleep Questionnaire), sleep-related symptoms, anxiety, depression and some neurocognitive test scores (digit symbol and Trail making test A). The mean CPAP compliance was 4.9 h/night, and approximately 70% of the patients presented good adherence.

Recently, Ponce S et al. [67] performed an open-label, multicentre study in 145 elderly patients with moderate OSA (IAH>15/h) who were randomized to CPAP or no CPAP for 3 months. The authors concluded that in subjects with moderate OSA, CPAP improves diurnal hypersomnia and some sleep-related symptoms and quality of life domains without changes in neurocognitive tests and blood pressure [67].

Compliance with CPAP

Another important aspect of CPAP in elderly subjects is compliance with the treatment because it seems logical that elderly subjects could present some characteristics that could make CPAP adherence difficult, such as milder symptoms, the presence of cognitive disturbances, living alone, comorbidities or neurological deficits.

It is believed that adherence to CPAP is reduced by age (particularly over 75 years) [68], considering that the first week of adherence and socioeconomic variables are strong predictors of CPAP compliance [69].

Therefore, the indications for CPAP should be individualized in very elderly individuals and, when prescribed, closely monitored to address tolerance and compliance. Identifying predictive factors could be relevant to developing strategies to minimize their negative influence on adherence to CPAP treatment [69].

Current Situation and Future Perspectives

There are many elderly patients with OSA who are under CPAP treatment, and these numbers are expected to increase in future years; nevertheless, our knowledge on the impact of OSA and the benefits of treatment in this age group is very limited. Therefore, the study of OSA in the elderly is a challenge that needs to be addressed in the coming years. It is necessary to perform studies that clarify the impact of OSA in the elderly and the effectiveness of CPAP in this population group to establish specific recommendations for the elderly. Moreover, it could be important to develop new follow-up strategies and educational programs to improve adherence and to facilitate a closer follow-up in this age group, especially in those ages in which lower compliance has been observed. Future studies should also address the validation of specific questionnaires, the use of simplified diagnostic methods for elderly individuals and the evaluation of other therapeutic options for the golden age.