Obstructive sleep apnoea (OSA) is common and is underdiagnosed in the community. Epidemiological studies have estimated that 4% of men and 2% of women over 50 years have symptomatic OSA [1]. Benjafield et al. estimated that the global prevalence of OSA was almost 1 billion people with 936 million adults aged 30–69 years having mild to moderate OSA and 450 million adults aged 30–69 years having moderate to severe OSA [2]. OSA is a major chronic health burden as it causes nocturnal hypoxia (decreased oxyhaemoglobin saturation (SpO2)), impacting cognition and mood, and is associated with cardiovascular diseases. OSA may be associated with other diseases, especially chronic ageing conditions found in similar populations, yet little research has been done to look at associations in populations with specific ageing conditions such as age-related macular degeneration (AMD).

Overnight pulse oximetry (OPO) has been used as a screening and diagnostic tool to detect changes in blood oxygenation, both in primary care and hospital settings. It provides portable, high-quality information about a patient’s blood oxygen level [3]. OPO is used as a diagnostic and screening tool for specific respiratory diseases such as chronic obstructive pulmonary disease (COPD) [4], interstitial lung disease [5], and neuromuscular diseases [6]. More recently, pulse oximetry has been used for monitoring blood oxygen levels in COVID patients as part of remote management [7]. Given the ease of use of OPO and widespread applicability in screening, diagnosis, and management of sleep disordered breathing (SDB), and the ability to self-administer in the home, it could be used to screen for OSA when determining associations between nocturnal hypoxia with other diseases. However, it is important to understand the variability in measurements obtained from night to night in an older, asymptomatic, cohort, to be able to determine the most robust screening protocol.

Many simple, validated screening questionnaires such as the STOP-BANG Questionnaire (SBQ) and the Epworth Sleepiness Scale (ESS) are used as a screening tool in primary care and sleep medicine to identify patients at high-risk for OSA [8, 9]. El-Sayed et al. compared four commonly used sleep questionnaires (ESS, SBQ, Berlin, and STOP) and determined that SBQ, Berlin, and STOP questionnaires had the highest sensitivity to predict OSA (97.55%, 95.07%, and 91.67%, respectively), whilst the ESS had the highest specificity to predict OSA (75%) but the lowest sensitivity (72.55%) [9]. In another study comparing commonly used questionnaires in an at-risk population, the SBQ appeared superior to the Berlin and OSA50, but all tests were noted to have significant limitations [10]. More objective screening tools are needed to screen for OSA and monitor nocturnal blood oxygenation levels.

Polysomnography, a laboratory-based, single-night intensive test, is considered the ‘Gold Standard’ diagnostic test for OSA [11]; however, it is expensive, and there are long waiting times for an assessment. Hence, assessments with a home-based OPO is an alternative, objective method that has been used where there is a high suspicion of OSA [11,12,13,14,15]. Vázquez et al. evaluated the diagnostic performance of OPO with polysomnography in an adult symptomatic population and found that the polysomnography derived apnoea hypopnoea index (AHI) and the OPO derived respiratory disturbance index (RDI) were highly correlated (R = 0.97) [11]. Another study by Jonas et al. comparing OPO with polysomnography in the paediatric population, with SDB found satisfactory diagnostic performance of OPO in the detection of moderate to severe OSA and a high positive predictive value [15]. The strong correlations of OPO with polysomnography data, along with OPO being widely available suggests that it might be a robust, objective, simple technique to measure overnight blood oxygenation, as a screening tool for OSA in a general population. However, there is a need to understand the inter-night variability of OPO within normal populations to inform the most appropriate protocols for screening.

The inter-night variability of self-administered, OPO parameters has been investigated in systemic conditions such as COPD [16] and chronic heart failure [17], but not in the normal older, adult population. Studies have investigated inter-night variability in OPO parameters both in adults and children with prior diagnosis of OSA, although findings vary across the different OPO parameters [18,19,20]. Hoppenbrouwer et al. reported minimal inter-night variability of OPO parameters including ODI in children with OSA [20]. However, Stöberl et al. reported high inter-night variability of the ODI in adult OSA patients who were withdrawn from continuous positive airway pressure therapy [18]. To our knowledge, the inter-night variability of OPO parameters in an older adult population, without any respiratory disease nor specific symptoms of a respiratory disease or SDB, is currently unknown.

Greater knowledge of the variability around nightly measurement of an individual’s oxyhaemoglobin level and nocturnal hypoxia status is required if it is to be used when determining association with other chronic disease in the elderly. Furthermore, although both sleep questionnaires and OPO are used for screening of OSA, a correlation between OPO parameters and sleep questionnaires in this asymptomatic population has not been established. We had an interest in using OPO and sleep questionnaires to investigate if there was an association between nocturnal hypoxia and AMD. We were therefore interested in assessing the inter-night variability of the self-administered, home OPO in an older population to determine if single recordings of OPO were suitable to use in assessing nocturnal hypoxia or whether multiple recordings were likely to provide more robust results.

The primary aim of this study was to determine the inter-night variability in OPO parameters in an older, asymptomatic, adult population. A secondary aim was to examine the correlations between the oximetry parameters and OSA risk, as determined from common sleep questionnaires, in this asymptomatic older population.