Obstructive sleep apnea (OSA) is a sleep-disordered breathing (SDB) characterized by repetitive episodes of total (apnea) or partial (hypopnea) obstructions of the upper airway during sleep. These alterations in normal ventilation cause intermittent hypoxia, changes in intrathoracic pressure and sleep fragmentation, which contribute to the development of significant impairments on health [1]. It is estimated that OSA is present in nearly one billion people worldwide, with a prevalence of moderate-to-severe OSA between 3–9% in females and 10–17% in males [2, 3, 4].

Overnight, in laboratory polysomnography (PSG) is the gold standard test for the diagnosis of OSA [5]. However, PSG is a complex test that is not always available, meaning alternative diagnostic tests may be performed [5]. The American Academy of Sleep Medicine (AASM) guidelines recommend home sleep apnea tests (HSAT) for the diagnosis of adult patients with a high probability of moderate-to-severe OSA [6] and without serious comorbidities. HSAT usually measures airflow, respiratory effort, oxygen saturation (SpO${}_2$) and heart rate. Since it has limitations because it does not include electroencephalogram, electrooculogram or electromyogram, a negative result should be considered for further investigation with PSG [1].

OSA is measured through the apnea-hypopnea index (AHI), defined as the total number of respiratory events recorded per hour of sleep and is obtained from sleep studies. The International Consensus Document considers OSA when: (1) the presence of an AHI $\ge$15/h, and is predominantly obstructive; (2) the presence of an AHI $\ge$5/h accompanied by one or more of the following factors: excessive sleepiness during the day, restless sleep, excessive fatigue and/or impaired sleep-related quality of life, not justifiable by other causes [7]. OSA classification is divided as: no OSA when AHI is less than 5 events per hour, mild OSA when AHI is between 5 and 15 events per hour, moderate OSA if AHI is between 15 and 30 events per hour, and severe OSA when AHI is greater than 30 events per hour.

The most effective treatment is continuous positive airway pressure (CPAP), which mitigates AHI and hypoxia. However, other treatments are available when CPAP is not recommended or ineffective. Among alternatives to CPAP, body weight loss should be recommended for all overweight or obese patients with OSA [8]. In a subgroup of patients, oral appliance devices and CPAP are similarly effective in terms of their impact on symptoms and quality of life [9, 10], being higher adherence to these devices. Finally, surgical and medical treatments (positional therapy, anti-inflammatory medication and hypoglossal nerve stimulation) could be considered, although they are less frequent [11].

Therefore, diagnosis and severity of OSA and treatment management are mostly determined by their AHI. However, this sole parameter does not reflect the complex heterogeneity of the disease, which is a multifactorial illness that includes different pathogenic mechanisms, clinical phenotypes [12, 13] and cardiovascular (CV) consequences (Fig. 1). An example could be that individuals with similar AHI values have different patterns of hypoxia, the main deleterious factor involved in the principal consequences of OSA [14], and distinct clinical daytime repercussions [15]. In this sense, there has been an effort to move beyond AHI for OSA clinical management. First, cluster analyses have identified groups of OSA patients with similar characteristics including symptoms, sleep disturbances and comorbidities, which may be helpful to establish prognostic factors and personalized therapeutic strategies in OSA management [12, 13, 16]. Another important approach in OSA management has been the identification of biomarkers for patients who may be at high CV risk.

Fig. 1.

Comparison between current management and proposed management for obstructive sleep apnea based on the heterogeneity of the disease. Abbreviations: OSA, obstructive sleep apnea; AHI, apnea-hypopnea index; SpO${}_2$, oxygen saturation; ODI, oxygen desaturation index; T90, % of time with oxygen saturation below 90%.

Observational studies indicate that OSA is associated with adverse morbidities, being the most scientific evidence on CV disease, and that alleviation of obstructive events with CPAP may improve such CV outcomes [17]. In this CV sphere, OSA has been predominantly associated with high blood pressure (HBP) and CV mortality, but also with other CV complications such as heart failure, coronary artery disease and stroke [18]. OSA is highly prevalent in HBP patients, between 30 to 50%, a prevalence which is increased in resistant HBP (more than 85%), and OSA is the most frequent factor of secondary HBP [19, 20]. The increase in blood pressure (BP) levels and the worsening of its control, has been suggested that is slightly reverted with treatment with CPAP [21]. Besides, although moderate-to-severe OSA has been associated with increased CV mortality [18, 22], randomized clinical trials (RCT) have failed to prove the reduction of such adverse CV events with treatment with CPAP [23, 24, 25, 26] in populations with CV disease concomitant with OSA. One of the possibilities is that this lack of effect would be related to the parameter chosen (AHI) in these RCTs for OSA characterization.

Hence, different studies have questioned the utility of AHI in OSA, highlighting that it is not the best parameter for its management, including CV disease prediction. Since there is a need to search for new parameters that better reflect this heterogeneity and predict which patients would benefit most from treatment, novel biomarkers for OSA severity have been proposed. In this review, we will focus on hypoxic burden (HB), described as a measure of OSA that captures all dimensions of OSA-related desaturations, including frequency, duration and depth of the respiratory events. HB was first defined by Azarbarzin et al. [27] in 2019 as the total area under the respiratory event-related desaturation curve. HB is obtained by adding all individual desaturation areas and dividing it by the total sleep time, being the units of HB (%min)/h [27].

To assess whether HB has been used to evaluate OSA, we conducted a systematic review to find publications about the heterogeneity of OSA measured by HB and if it is associated with future disease.

This systematic review was conducted using PubMed and Web of Science, and based on Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [28]. The main focus was to identify scientific knowledge of HB as a novel measurement for OSA severity, and its clinical implication as a predictor of increased CV risk.

The terms “sleep apnea” and “hypoxic burden” were used to look for all the publications from the date of inception to August 15, 2023. Inclusion criteria were: articles in English published in peer-reviewed journals. Exclusion criteria included: (1) not available publications; (2) duplicated articles; (3) letters, editorials, and congress communications; and (4) articles not including information about HB as a specific biomarker of OSA. Two independent reviewers screened the titles and abstracts of all candidates according to eligibility criteria. Data were extracted independently and included the type of study and main CV outcomes related to HB. The flow diagram is illustrated in Fig. 2.

Fig. 2.

Schematic flow diagram for the selection of reports. Abbreviations: HB, hypoxic burden; OSA, obstructive sleep apnea.

There is enough evidence to affirm that AHI by itself is outdated to characterize OSA, focusing on the remodeling of OSA management in developing novel markers that consider symptoms and comorbid illnesses. Currently, classifying OSA severity based on the number of observed events oversimplifies the pathophysiology of the disease. Different variables such as HB collect more information about the respiratory event.

The strengths of HB comprise, firstly, that this parameter describes the frequency, duration and depth of the desaturation associated with the obstructive episodes, which presents a higher relationship with associated consequences and morbidities than classical metrics.

Secondly, as intermittent hypoxia is considered the principal damaging factor for developing CV disease, HB is an innovative measurement specific to OSA and has recently been used to characterize it. Until now, conventional measures including ODI and T90 are better related to CV outcomes than AHI [55], but they share similar limitations. On the one hand, ODI only measures frequency but not the duration or depth of the events, and there is not a consensus on the percentage drop to be considered, being 3% or 4% depending on the study criteria. T90 measures persistent hypoxemia but not intermittent hypoxia that is not specific to OSA and that may be correlated with chronic airway diseases. Finally, similar to AHI, both metrics do not take into account the basal saturation of the patients. Hence, the HB could provide additional clinical information rather than classic values. OSA categorization in severity groups employing HB could differentiate patients with high HB and low HB, at the same time that this population is grouped according to their associated CV risk. More RCTs are needed to establish the cut-off point of OSA patients with high HB, although the first reports have shown a limit of 60 %min/h [32].

Thirdly, in the same way as the classical methods, HB can also be obtained from simplified methods, since it is calculated from the data acquired from the flow and saturation signals. Moreover, it can be calculated from automatic analyses of the sleep studies, facilitating the work of clinicians and accelerating the diagnosis. A recent publication [41] has assessed the relationship between manual-scored HB and automatically-scored HB, showing a high correlation between them and with adverse CV outcomes in a similar way as reported. Thus, it could be used to contribute to reducing the underdiagnosis observed in this disease.

Fourthly, it has been proposed that HB could identify the patients that would benefit from treatment and represent an important target for improving CV risk. Despite the beneficial effects of CPAP in CV risk reduction (improvements in endothelial dysfunction, blood pressure and first signs of atherosclerosis) in OSA patients, there is controversy about the role of CPAP in the prevention of adverse CV events. In the three principal RCTs [23, 24, 26] evaluating CPAP in CV outcomes, there were no significant associations between CPAP and adverse CV events or death. A possible explanation could be the selection of the patients included in such studies, as they had previous CV illnesses, suggesting a possible decreased beneficial effect in secondary CV risk prevention. The heterogeneous pathophysiology of OSA stands out that a personalized approach accounting for specific endotypes of patients with OSA is necessary and worthwhile. If we properly determine the different OSA phenotypes that exist, using HB among other variables, we could be able to identify patients who could benefit from CPAP as secondary prevention. Thus, more studies are needed to evaluate the possible HB utilization for detecting those specific profiles of patients. In this line, a recent study from Pinilla L. et al. 2023 [56] reported that in a population with acute coronary syndrome and OSA, the group with higher HB had a significant reduction in the incidence of CV events when they were treated with CPAP. These results demonstrate the possibility of implementing HB as a measure of OSA severity, differentiating patients in which CPAP could have a long-term protective effect on CV prognosis.

Lastly, the different responses to treatment and different prognoses could be evaluated through HB. It has been reviewed that different drugs and other non-pharmacological treatments alternative to CPAP lower HB but not AHI in all cases. A possible explanation could be that they decrease the duration and depth of the desaturation but not the number of respiratory events. Consequently, HB would help in the selection of the treatment in terms of reducing the risk of future disease.

However, the limitations of HB need to be commented on. The main disadvantage is the inability to differentiate long and more superficial desaturations from short but deep ones. For example, an HB of 60 (%min)/h could correspond to 10 min of 6% desaturation or 20 min of 3% desaturation per hour of sleep. Besides, it cannot detect the arousals associated with the respiratory event, which may underestimate the diagnosis. Another weakness is the incapacity to distinguish apneas from hypopneas. Finally, there is no consensus on how to use it in practice or to calculate it to ensure that everyone obtains the same HB. Nevertheless, these limitations could be complemented with other parameters, symptoms and other illnesses, creating a score that includes various aspects of the disease.

In summary, HB has more advantages compared to current management, which justifies studying its feasibility to be implemented in clinical practice.

HB has been used in numerous studies to assess OSA, emphasizing its correlation with the CV domain. It appears to surpass traditional measurements in terms of diagnosis based on comorbidities and making clinical therapeutic decisions for OSA patients. Integrating this measure into sleep units could enhance OSA management and contribute to a shift towards more personalized medicine in clinical practice. Furthermore, this approach not only addresses the immediate concerns associated with OSA but also recognizes and manages potential CV implications.

OSA, obstructive sleep apnea; SDB, sleep-disordered breathing; PSG, polysomnography; AASM, American Academy of Sleep Medicine; HSAT, home sleep apnea test; SpO${}_2$, oxygen saturation; AHI, apnea-hypopnea index; CPAP, continuous positive airway pressure; CV, cardiovascular; HBP, high blood pressure; BP, blood pressure; RCT, randomized clinical trials; HB, hypoxic burden; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analysis; CKD, chronic kidney disease; ODI, oxygen desaturation index; T90, time with oxygen saturation below 90%; DBP, diastolic blood pressure; SBP, systolic blood pressure; ICU, intensive care unit; MESA, Multi-Ethnic Study of Atherosclerosis.

Conception and design: OM, CC, ES, SR; data adquisition and systematic review: OM, CC, ES, MC, LS, SL, PR, IC, MS, OM; manuscript draft: CC, ES, OM; manuscript review: SR, MC, LS, SL, PR, IC, MS; final approval of the version to be published: CC, ES, SR, MC, LS, SL, PR, IC, MS, OM. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research was funded by the Instituto de Salud Carlos III (ISCIII: PI18/00565 and PI22/01653) and co-funded by the European Regional Development Fund (ERDF) “A way to make Europe”, the Spanish Respiratory Society (Sociedad Española de Neumología y Cirugía Torácica- SEPAR) (535-2018 and 1073-2020) and Sociedad Madrileña de Neumología y Cirugía Torácica (NEUMOMADRID).

The authors declare no conflict of interest.