Air Pollution and Respiratory Permeability in Obstructive Sleep Apnea-A Review

Obstructive sleep apnea (OSA) is a common disorder characterized by recurrent episodes of nocturnal upper airway obstruction during sleep, which can seriously affect sleep quality and cause sleepiness during the daytime. Known risk factors for OSA are numerous, including obesity, age, gender, craniofacial anatomy, smoking, alcohol consumption, and genetic inheritance. Recently, air pollution was linked to an increased risk of OSA severity. Alterations in permeability are considered to be an important factor in the development of OSA; however, the role of air pollution remains unclear. This review article explored the role of air pollution and airway permeability in the pathogenesis of OSA.


INTRODUCTION
Economic burdens related to obstructive sleep apnea (OSA) account for billions of US dollars annually (AlGhanim et al., 2008). Costs associated with OSA mostly arise from OSA diagnoses, treatments, and comorbidities (Kapur, 2010). Estimated costs for undiagnosed OSA range $34 billion to $69 billion annually (Knauert et al., 2015). People with sleep disorders also work less efficiently than those without sleep difficulties (Sherman, 2013). In a population-based study of nearly 3000 subjects over 10 years, men with snoring and excessive daytime sleepiness were more than twice as likely to have occupation-related accidents (95% confidence interval (CI) = 1.3-3.8) (Lindberg et al., 2001). It is also known that the quality of life (QOL) is lower in patients with severe OSA, and QOL substantially improves following clinical therapy such as continuous positive airway pressure (CPAP) (D'Ambrosio et al., 1999;Akashiba et al., 2002;Dutt et al., 2013). Recently, there is emerging evidence of associations of environmental factors such as air pollution on increased risks of OSA; however, the underlying mechanisms of this relationship remain uncertain.
It was estimated that the prevalence of OSA ranged from 9% to 37% in men and from 4% to 50% in women (Franklin and Lindberg, 2015).
OSA is considered a multifactorial disease, including multiple genetic and environmental factors. Obesity is widely known as the most important risk factor for OSA. Older age, male gender and smoking is also strongly associated with a higher risk for OSA. This current review focused on risk factors for OSA and possible mechanisms related to air pollution.

PREVALENCE OF OSA
OSA is a common disorder characterized by recurrent episodes of nocturnal upper airway obstruction associated with repeated cycles of desaturation and re-oxygenation, leading to sleep fragmentation and consequent daytime sleepiness (Senaratna et al., 2017). OSA prevalence in adults more than 18 years of age, which is defined as a score of ≥ 5 on the Apnea Hypopnea Index (AHI), ranged from 9% (Reddy et al., 2009) to 38% (Tufik et al., 2010). Data from the Wisconsin Sleep Cohort showed that almost 25% men and nearly 10% in women of 30-60 years of age had OSA (Young et al., 1993). In a cohort of more than 2000 subjects from Spain, an AHI of ≥ 5 was found in 26% of men and 28% of women (Duran et al., 2001). A population-based study from Korea showed that the prevalence of OSA was more than one-fourth in men and 16% in women (Kim et al., 2004). The estimated prevalence of OSA in Hong Kong was 8.8% for men and 3.7% for women (Ip et al., 2001;Ip et al., 2004), while the prevalence in India was 19.5% (Udwadia et al., 2004). Interestingly, OSA occurs in roughly half of Swedish females aged 20-70 years old (Franklin et al., 2013). When using the American Academy of Sleep Medicine (AASM) 2012 diagnostic criteria, it is postulated that almost 1 billion adults in the world have OSA. Therefore, OSA is an important public health issue worldwide.

RISK FACTORS FOR OSA
Risk factors for OSA are shown in Fig. 1.

Obesity
Obesity is regarded as one of the strongest risk factors for OSA (Franklin et al., 2013). Since the obesity prevalence is growing globally (Punjabi, 2008;Garvey et al., 2015), it is likely that the OSA prevalence is also increasing (Peppard et al., 2013). A prospective cohort involving nearly 700 subjects found that a 10% increase in weight was associated with approximately a 33% increase in the AHI (Peppard et al., 2000). It is believed that the volume of parapharyngeal fatty tissue is related to the severity of OSA (Shelton et al., 1993). However, it is still controversial whether neck or waist circumference is a better estimator of OSA compared to the body-mass index (BMI) alone . A cross-sectional study of 85 male subjects found that the AHI was correlated with abdominal fat (Schafer et al., 2002). Nonetheless, data from the Sleep Heart Health Study showed that in addition to the BMI, neck circumference and waist circumference were also independent risk factors for severe OSA among middle-aged and older adults (Young et al., 2002). In a prospective study of more than 50 patients, Ogretmenoglu et al. (2005) found that the BMI and body fat percentage (determined by a bioelectrical impedance assay) were correlated with the AHI.
Obesity may contribute to OSA through several mechanisms (Punjabi, 2008). Parapharyngeal fat can cause airway narrowing and collapse . Leptin, a hormone produced by adipocytes and which is associated with the body composition, energy homeostasis and feeding behaviors, was significantly correlated with the AHI (Ulukavak Ciftci et al., 2005). The relationship between leptin resistance and obesity was mentioned in a previous study (Carter et al., 2013).
In a study of 18 subjects, patients with OSA had excess parapharyngeal fat deposition compared to non-OSA subjects as measured by the neck circumference (Mortimore et al., 1998). Katz et al. (1990) found that the only contributors to the AHI were the external neck circumference, BMI, and pharyngeal internal circumference. Similarly, a Turkish cross-sectional analysis of 1912 subjects showed that the neck circumference was of greater value than the waist circumference in the association with OSA among men (Onat et al., 2009).

Age
It was shown that an older age is associated with a higher OSA prevalence (Bixler et al., 1998;Gabbay and Lavie, 2012). This can be explained by an increase in parapharyngeal fat and an elongated soft palate (Malhotra et al., 2006). However, the OSA prevalence reaches a plateau after the age of 60 years (Bixler et al., 1998;Young et al., 2002), probably due to a smaller influence of BMI on elderly OSA subjects (Newman et al., 2005). Another possibility is that elderly patients with OSA have a higher mortality than those without OSA.

Gender
It is known that the prevalence of OSA in men is 2-3-fold higher than in women (Punjabi, 2008;Gabbay and Lavie, 2012;Ralls and Grigg-Damberger, 2012). In the Sleep Heart Health Study over a 5-year follow-up, men were more likely to have an increase in the Respiratory Disturbance Index (RDI) than were women (Young et al., 1993). This discrepancy may be explained by a sex difference in reporting symptoms of OSA, since women tend not to report common symptoms (loud snoring, cessation of breathing, and gasping), but instead report symptoms of fatigue and lack of energy (Young et al., 1997). Furthermore, it is more likely for men to be referred to OSA specialists because of the expectation of physicians that OSA predominantly affects men. Sex hormones are also likely to be important in OSA, as the disorder is more common in postmenopausal women (Ralls and Grigg-Damberger, 2012). However, hormone replacement can reduce the severity of OSA in those patients (Shahar et al., 2003;Wesstrom et al., 2005).

Craniofacial Anatomy
It was shown that different craniofacial morphologies are correlated with OSA severity (Fan and Liu, 2010;Vidovic et al., 2013;Fernandez-Salvador et al., 2018;Sutherland et al., 2018). For example, brachycephaly was a strong risk factor for high AHI scores in a white population (Cakirer et al., 2001), while Chinese patients with OSA had retrognathia or micrognathia (Lam et al., 2005). The mechanism for this phenomenon is that several craniofacial morphologies can change the airway shape and increase the likelihood of its collapse during sleep. Furthermore, tonsillar hypertrophy or an enlarged base of the tongue can also result in upper airway narrowing and cause OSA (Schwab, 2003).

Smoking and Alcohol
Current smoking, ex-smoking, and active and passive smoking are all strong factors for snoring and OSA (Khoo et al., 2004;Quan et al., 2014;Varol et al., 2015;Bielicki et al., 2019). Smoking

Genetic Inheritance
Several studies researched the role of genetic factors in the genesis of OSA. Segregation analyses showed that inherited factors can contribute to approximately 35% of the variance in OSA severity. Susceptibility loci for OSA were identified in genome-wide association scans (Patel et al., 2012;Yin et al., 2014;Zhang et al., 2014;Gok et al., 2015;Cade et al., 2016), which showed an association of genetic loci and OSA in African-American and European-American subjects (Patel et al., 2012).

Air Pollution
Air pollution is known to be an important risk factor for cardiopulmonary disease; however, few studies have investigated the effects of air pollution on OSA. In a study of over 3000 subjects from the Sleep Heart Health Study, an increased RDI and decreased sleep efficiency were associated with short-term augmentation in particulate matter (PM) of ≤ 10 µm in aerodynamic diameter (PM10) and temperature in the summer (Zanobetti et al., 2010). Another populationbased study of nearly 2000 adults of 45-84 years of age in six cities of the USA from the Multi-Ethnic Study of Atherosclerosis showed that an increase in annual nitrogen dioxide (NO2) and PM of ≤ 2.5 µm in aerodynamic diameter (PM2.5) was correlated with an increased OR of sleep apnea (Billings et al., 2019). Specifically, each 5-µg m -3 per year increase in PM2.5 led to a 60% higher OR of OSA. Similarly, a large population-based study of more than 4000 individuals in Taiwan also showed an association of 1-year PM2.5 and NO2 exposure with the AHI, which was greater during spring and winter (Shen et al., 2018). In a retrospective study of more than 7000 subjects, the AHI was associated with atmospheric pressure and carbon monoxide (CO) exposure (Cassol et al., 2012). Those authors found that OSA were more prevalent in winter than in other seasons. Interestingly, short-term elevated ozone (O3) and temperature resulted in an increased AHI in a European study (Weinreich et al., 2015). A study of more than 5000 subjects in Taiwan from 2008 to 2015 showed that PM10, O3, sulfur dioxide (SO2), and humidity were only associated with severe OSA in the non-rapid eye movement sleep period (Cheng et al., 2019).

ALTERATIONS IN THE PERMEABILITY OF AIRWAY BY PM
Several studies researched the relationship between a shift in body fluids and OSA in healthy people (Redolfi et al., 2009), and in those with heart failure (Yumino et al., 2010), hypertension (Friedman et al., 2010), and renal failure (Elias et al., 2012). The fact that non-obese patients wearing compression stockings during the day can decrease the AHI suggests a causative role of an overnight fluid shift in OSA (Silva et al., 2017). A study of 36 subjects showed that patients with OSA had more baseline leg fluid than non-OSA patients, which suggests a greater fluid shift in OSA patients (Ding et al., 2014). According to White, during the day, fluid is retained in the legs due to gravity. During night sleep, the fluid is redistributed to the neck also due to gravity, which leads to airway narrowing and consequent OSA (White and Bradley, 2013).
The mechanism for the rostral fluid shift can be explained by Starling's force. Patients with heart failure have high venous pressure, which results in more fluid being retained in the legs during the daytime. Consequently, when a patient lies down, more fluid in the legs will shift into the neck and cause OSA. Similarly, patients with a low protein state, which means low colloid osmotic pressure, will exhibit edema. However, edema also appears in healthy OSA patients, suggesting that OSA can lead to edema (Iftikhar et al., 2008). The mechanism is still unclear and might be explained by an increase in the atrial natriuretic peptide (Kita et al., 1998) and the activation of the renin-angiotensin-aldosterone system.
The association between airway permeability and air pollution was investigated in previous studies. High concentrations of diesel exhaust particles (DEPs) reduce occludin messenger RNA, which has an important role in maintaining the epithelial barrier integrity (Lehmann et al., 2009). PM10 also reduces occludin expression and increases alveolar transepithelial electrical conductance (Caraballo et al., 2011). Cadmium, a chemical element present in air pollutants, was shown to disrupt tight junction integrity (Cao et al., 2015) in vitro via modulating genes associated with tight junctions. PM2.5 also disrupts the nasal epithelial barrier via degradation of tight junction proteins (Zhao et al., 2018). PM2.5 is associated with Clara Cell Protein (CC16), a marker of epithelial barrier permeability (Timonen et al., 2004;Jacquemin et al., 2009) in adolescents (Provost et al., 2014) and elderly men (Madsen et al., 2008). Together, changes in permeability caused by air pollution could be an important pathophysiology of airway disease.

POSSIBLE MECHANISMS OF PM IN OSA
The mechanism of OSA by PM may be explained by oxidative stress and inflammation (Fig. 2). The composition of particles is very important for their toxicity. Thus, transition metals such as iron, copper, nickel, chromium, and vanadium can increase reactive oxygen species (ROS). Furthermore, volatile organic compounds (quinones and polycyclic aromatic hydrocarbons) in PM cause the production of ROS in neutrophils, eosinophils, macrophages, and bronchial epithelial cells. Normally, homeostasis in the body keeps oxidizing substances such as ROS and antioxidant defense mechanisms in a balanced state. However, when there are excess ROS generated during mitochondrial oxidative metabolism, oxidative stress occurs. This phenomenon leads to oxidative lipid degradation, protein oxidation, and consequent inflammation. PM2.5 was reported to increase ROS in human nasal epithelial cells and decrease the viability of those cells (Hong et al., 2016). PM-related oxidative stress and inflammation leads to hyperpermeability in the human upper airway and subsequent airway obstruction, thus causing OSA. Exposure to PM2.5 also led to barrier disruption in the human epithelium by reducing expressions of tight junction proteins and increasing the secretion of inflammatory cytokines, which result in allergic rhinitis and rhinosinusitis (Zhao et al., 2018;Kim et al., 2019;Xian et al., 2020). Furthermore, disruption of tight junctions by DEPs via a ROS-mediated pathway resulted in greater permeability of nasal epithelial cells and seasonal allergic rhinitis as observed in an animal model (Fukuoka et al., 2016). PM2.5 also caused hypersecretion in the rat airway (Harkema et al., 2004). A study in China found a significant association of PM2.5 and PM10 exposure with the prevalence of allergic rhinitis (Teng et al., 2017).

CONCLUSIONS
This review paper demonstrates an association of air pollution and respiratory permeability in OSA. There is now considerable evidence that fluid accumulation in the legs during the daytime and the overnight fluid shift is a risk factor for OSA. Furthermore, airway permeability changes by air pollution play important roles in the pathogenesis of airway diseases. Further work is required to better understand patterns of fluid shifts in OSA and to completely elucidate the mechanisms of air pollution causing OSA.