The Many Faces of Obstructive Sleep Apnea

Abstract

Obstructive sleep apnea is a common disorder with many different comorbidities. Patients can present with these comorbidities even when sleep apnea has not been diagnosed due to the low diagnostic rates for obstructive sleep apnea worldwide. This presents a concern for clinicians as unless sleep apnea is treated along with a presenting comorbidity, the patient will not have an optimal health outcome. This chapter addresses several of the most common and worrisome comorbidities of sleep apnea and discusses the relationships and pathophysiology of comorbidities including insomnia, treatment resistant hypertension, restless legs syndrome, depression, gastroesophageal reflux, asthma, cognitive disorder, REM sleep behavioral disorder and paroxysmal atrial fibrillation. Clinical studies documenting the relationships of each of these comorbidities to obstructive sleep apnea are presented with putative pathophysiologic discussion of how obstructive sleep apnea aggravates or leads to the development of each of these illnesses.

1. Introduction

There are many comorbidities associated with obstructive sleep apnea, several of which will be discussed in this chapter. The problem for clinicians providing care to patients is that patients may present with the comorbidity rather than the sleep related breathing disorder. Indeed, underdiagnosis of sleep related breathing disorders has been shown in epidemiological studies [1, 2]. There are thus many patients who present for medical care of a comorbidity rather than the causative or aggravating sleep related breathing disorder.

The reason for the low diagnosis and treatment rates for sleep related breathing disorders (SRBD) is speculative as most of the literature deals with the problems of continuous positive airway pressure (CPAP) management rather than resistance to presentation for evaluation and treatment for the symptoms of SRBD [3, 4, 5]. Anecdotal hints are provided by primary care physician comments and patient encounters in the author’s 35 years of evaluating and managing patients with SRBD [6]. After a lecture to primary care physicians one of the physicians raised patient concerns about cost, especially if the evaluation did not show a SRBD. This cost issue discouraged him from referring patients who did not have a high probability of SRBD. One of our fellows, whose family immigrated from Vietnam described a cultural concept that viewed snoring positively as a sign of deep restorative sleep, not a phenomenon requiring medical evaluation. The author has seen a lack of knowledge of the risks associated with untreated SRBD among patients presenting to the sleep clinic. In addition, anxiety about the ability to tolerate a CPAP mask secondary to claustrophobia [7] or cosmetic concerns, i.e., head gear producing indentations on their cheeks or forehead or loss of romantic attractiveness in bed. Some non-compliant patients in my clinic report that they feel that they sleep well, and it is only the concern of their family and friends that brought them to the sleep center. The author had an enlightening experience in London. I and my wife were having dinner at a pub by Victoria Station in 2018. The pub was filled with Americans, and we were sitting with another couple from Minnesota. I had mentioned that I was a sleep medicine physician and I’m not sure if I was overheard. Across the room a Caucasian gentleman stood up to leave with his group. He appeared to be in his mid-sixties with an estimated body mass index (BMI) of 33. He complained in a loud voice that he had been referred by his cardiologist for a sleep study due to a heart rhythm problem (I assumed paroxysmal atrial fibrillation from his description). He stated “I sleep fine. I don’t need a sleep study. Sleep studies are just another way for doctors to make money.” All these examples are anecdotal. Improving technology may lower the cost of diagnostic evaluations and hopefully one of the barriers to treatment of patients.

There are several comorbidities of SRBD. This chapter will review those comorbidities that are common or concerning in the author’s clinical practice. These consist of insomnia [8, 9], hypertension [10, 11], depression [12, 13], restless legs syndrome [14, 15], REM sleep behavioral disorder [16, 17], nocturnal asthma [18, 19], cognitive disorders [20, 21], gastroesophageal reflux disease [22, 23], and finally, paroxysmal atrial fibrillation [24, 25]. Additional comorbidities occur, such as nocturia, sleepwalking, diabetes, metabolic syndrome, and congestive heart failure, but are not discussed in this chapter to allow a more thorough discussion of the listed comorbidities. This chapter discusses the relationship in some detail between each listed comorbidity and SRBD and the need to consider SRBD in various diagnostic differentials.

2. Insomnia

The relationship with chronic insomnia and sleep related breathing disorder or obstructive sleep apnea is the most complex of the list. Insomnia is defined in the International Classification of Sleep Disorders, Third Edition [26] as requiring three components, including persistent sleep difficulty, adequate sleep opportunity, and associated daytime dysfunction. The recognized daytime symptoms of fatigue, decreased mood or irritability, general malaise and cognitive impairment overlap with the daytime symptoms of SRBD [27]. Impaired social or job/school performance, heightened risk of automobile and work accidents and cardiovascular disorders also overlap [8]. Chronic insomnia disorder of adults is present in an estimated 10% of the population but is more common in women, those with secondary insomnia from medical, psychiatric, and substance abuse disorders and in lower income individuals. Cho et al. [28] performed a clinical research study in South Korea utilizing two university sleep centers and enrolling 476 patients with polysomnogram documented obstructive sleep apnea (OSA). The investigators used Korean versions of a health survey and the Insomnia Severity Index (ISI-K) [29, 30], the Pittsburgh Sleep Quality Index (PSQI-K) [31], the Epworth Sleepiness Scale (ESS-K) [32], and the Beck Depression Inventory (BDI-K) [33]. The investigators used the ISI-K to separate the subjects with OSA plus insomnia (OSA + I) and OSA without insomnia (OSA-I) based on an ISI-K score of less than 15 for the OSA-I group. Of the 476 patients 29.2% or 139 were in the OSA + I group. Whereas females accounted for 24.2% of the total number of subjects, females accounted for 35.3% of the OSA + I group. There was also a significant difference in the BDI-K scale indicating an increase in depressive mood in the OSA + I group. The OSA + I group also showed a statistically significant increase in heart disease. While both OSA and chronic insomnia have documented cardiovascular risk, this study supported prior concerns that the combined risk is greater. This Korean study pointed to three conclusions. Women are more likely to have combined OSA and insomnia and individuals with OSA plus insomnia are more likely to have depression and heart disease.

The causation of OSA plus insomnia is speculative. These two sleep disorders are very common, however, the prevalence of insomnia found in OSA is higher than the prevalence of insomnia in the general population [28]. Several authors [8, 34, 35, 36] have reviewed the relationship of co-morbid insomnia and sleep apnea. Over the past 5 years the abbreviation OSA + I has evolved to the eponym ‘COMISA.’ One clear fact is that these patients are more difficult to manage [37, 38], requiring treatment of both sleep disorders in order to reach an optimal clinical improvement. This suggests that COMISA is not a complication of OSA alone, though the sleep fragmentation and activation of the sympathetic nervous system and hypothalamic-pituitary-adrenal axis appear to be aggravating factors. Chronic sleep deprivation of insomnia may play an additional role, compromising upper airway dilator muscle tone [39, 40] thus worsening OSA.

The treatment of COMISA is more complex than management of OSA without insomnia. COMISA patients have a greater prevalence of psychiatric disorders, including anxiety and claustrophobia that reduce CPAP adherence. Several authors [9, 38, 41] recommend treating insomnia first or concomitantly with cognitive behavioral therapy for insomnia (CBT-I) and/or non-benzodiazepine sedative hypnotics such as zolpidem and eszopiclone. CPAP produces improvement in sleep maintenance insomnia, but not sleep onset insomnia at two-year follow-up [42].

3. Hypertension

Obstructive sleep apnea is the leading contributor to treatment resistant hypertension [43]. The exact prevalence of resistant hypertension is estimated at 20–30% [44]. The strongest risk factors are older age and obesity. Logan et al. [45] performed a study with 41 participants (24 men and 17 women) who completed an overnight polysomnographic study. Drug-resistant hypertension was defined as a clinic blood pressure greater or equal to 140/90 mmHg while on a sensible combination of three or more antihypertensive drugs in maximally recommended doses. A long list of exclusionary criteria included a prior diagnosis of obstructive sleep apnea, ‘white coat’ hypertension, a history of poor compliance with drug treatment, use of substances that raise blood pressure or interfere with antihypertensive agents, renal insufficiency, excessive alcohol use, liver enzyme levels greater than twice the upper limit of normal, anatomic abnormalities of the upper airway, pregnancy, and significant aortic or mitral valve disease. The participants had a mean age of 57.2 ± 1.6 years, were 85% white, and had a mean body mass index 34.0 ± 0.9 kg/m². They were taking an average of 3.6 ± 0.1 different antihypertensives. The mean office blood pressure was a systolic of 168 ± 4.4 over a diastolic of 94.0 ± 2.3 mm Hg. The overall prevalence of OSA in this resistant hypertension group was 83%. There was a gender difference with OSA being seen in 96% of the male participants compared to 65% of the females. This gender difference was also reflected in the severity of OSA. The average apnea hypopnea index was 32.2 ± 4.5 events per hour in the men (severe OSA is greater than 30 events per hour). This compared to 14.0 ± 3.1 events per hour in the women (mild OSA is 5–15 events per hour).

The pathogenesis whereby OSA is causative of resistant hypertension has been addressed by several authors [10, 11, 46, 47, 48, 49]. The etiology appears to be multifactorial. Clearly, there is persistent sympathetic activation in both experimental animals [50] and humans [51, 52] from recurrent episodes of hypoxia. Activation of the renin-angiotensin-aldosterone system occurs through mechanisms that are not completely understood [53, 54]. In addition, oxidative stress has been shown to impair endothelial-dependent vasodilation [55].

The causal relationship of OSA to treatment resistant hypertension is one of the clearest relationships in this chapter.

4. Restless legs syndrome

Sleep apnea is often comorbid with restless legs syndrome (RLS) and the associated polysomnographic finding of periodic limb movements of sleep [14]. There is also an overlap with COMISA discussed previously. Pistoris et al. [14] investigated 202 patients retrospectively who had been studied in the Sleep Disorder Center in Regensburg, Germany from 2015 to 2016. Inclusion criteria were mild symptomatic obstructive sleep apnea (OSA) with an apnea hypopnea index (AHI) greater than five events per hour but less than 15 events per hour or moderate to severe OSA (AHI of greater than 15 events per hour) with or without symptoms. Patients with RLS were diagnosed clinically in a face-to-face interview and examination conducted by a sleep specialist. Criteria from the International Classification of Sleep Disorders [56] were used including (1) urge to move the legs, usually accompanied by discomfort in the lower limbs, (2) occurrence or worsening of symptoms in situations of rest or inactivity, (3) partial or total relief of symptoms by movement, and (4) symptoms that are worse or solely occur in the evening. Patients were excluded from the retrospective study if they had severe psychiatric disorders including psychosis, cognitive disorder, alcohol abuse, Parkinson’s disease, of difficulties with the German language. Patients with a prior diagnosis of RLS or who were treated with CPAP were also excluded. Fourteen patients were excluded for the above reasons plus another 20 were excluded due to inadequate information on retrospective review. The patients’ mean age was 55 ± 11 years. Women were 35% of the total number of patients. The mean body mass index (BMI) was 31 ± 6 kg/m². Of the 202 patients in the retrospective review 42 patients (21%) had comorbid RLS. The percentage of 21% is higher than the 5–10% prevalence of RLS in North American and European populations [56]. Of these 42 patients 25 or 60% were women, significantly higher than the percentage of women (29%) in the entire group of 202 patients.

The etiology of an increase in the prevalence of RLS in OSA is speculative and not without controversy. Lakshminarayanan et al. [57] showed prevalence of RLS in 60 sequential OSA patients with an AHI of greater than 10 events per hour of only 8.3%, within the expected range for the general population. Gothi et al. [58] reported findings like Pistoris et al. [14] described above. Of Gothi’s 249 OSA patients 61 patients had comorbid RLS for an increased prevalence of 28.5%.

The pathophysiology of RLS is a decrease of dopamine receptor density [59]. Some patients report a family history and younger age of onset. Familial genetic linkage studies [60] have revealed some sequence variants, but how these genes affect the pathogenesis of RLS is unclear. In addition, RLS can be secondary to iron deficiency, renal failure, pregnancy, peripheral neuropathy, chronic myelopathy and a variety of medications including tricyclic antidepressants, selective serotonin re-uptake inhibitors, lithium, antipsychotic drugs and dopamine receptor antagonists [15].

Though there are rationales for the pathophysiology of RLS in some of the afore mentioned, the rationale for an increased prevalence of RLS in OSA is unclear. The overlap with chronic insomnia suggests a common pathophysiology in the relationship with OSA. In the author’s personal experience, and as described by Gothi et al. [58], the combination of RLS and OSA (as with insomnia and OSA) makes these patients more complex to manage. This is especially true as two options for the management of treatment resistant RLS, benzodiazepines and opiates, are respiratory suppressants. These drugs suppress arousals, thus prolonging apneas and deepening oxygen desaturations in patients not using CPAP every night [61]. Rodrigues et al. [61] reported that 17 patients with RLS + OSA improved both their OSA and RLS symptoms with CPAP. Myc et al. [62] reported similar improvement in a single case report. The long-term results of consistent CPAP use in RLS + OSA patients need further research.

5. Depression

Several investigators [13, 63, 64] have surveyed newly diagnosed patients with OSA for depressive and anxiety symptoms. Velescu et al. [63] utilizing the Patient Health Questionnaire-9 Depression Scale found a prevalence of depressive symptoms in 48.48% of 99 consecutive new patients. At 1 year follow-up of CPAP treatment 92.9% of their patients had experienced significant improvement. Akberzie et al. [64] examined 45 patients’ records who presented for polysomnography over a five-month period and were diagnosed with obstructive sleep apnea (AHI greater than 5). The patients had completed a Hospital Anxiety and Depression Scale. There was a female predominance with 29 of the 45 patients being female. Of the 45 patients 29 were positive for depression (64.4%). Shoib et al. [13] performed the mini international neuropsychiatric interview plus scale and Hamilton Depression Rating Scale (HAM-D) on 182 patients undergoing polysomnography over a 2 year period. Patients were excluded if they were on nocturnal oxygen supplementation, CPAP or mandibular advancement devices, had upper airway surgery or had unstable cardiopulmonary, neurological, or psychiatric disease. Of the 182 patients, 47 had depression. Of the depressed patients 44 or 93.6% had either mild (3 or 6.8%), moderate (18 or 40.9%), or severe (23 or 52.3%) OSA. The HAM-D scale was significantly greater (p = 0.0001) in the depressive patients than the non-depressive patients.

Patients with OSA have a higher prevalence of depression [12, 65]. These depressive symptoms may gradually improve with CPAP treatment [66, 67]. Zheng et al. [68] showed a statistically significant (p = 0.031) reduced odds of depression care in the Sleep Apnea Cardiovascular Endpoints (SAVE) trial of 2410 patients with moderate to severe OSA and established cardiovascular disease in patients randomly allocated to CPAP plus usual care versus usual care alone and followed for 3–7 years.

The pathophysiology of OSA and depression’s association is speculative. Direct pathophysiology through aggravation of each other’s symptom complexes, as well as, indirect pathophysiology through molecular processes (hormonal and inflammatory) have been postulated.

6. Gastroesophageal reflux

Nocturnal gastroesophageal reflux (nGER) is common in patients with obstructive sleep apnea (OSA) [23, 69, 70, 71]. Green et al. [22] studied 331 patients diagnosed with OSA over a seven-year period from 1993 to 2000 who graded their nGER frequency from 1 (never) to 5 (always). All their patients were prescribed CPAP. Prior to treatment nGER was present in 204 patients (62%). Follow-up was obtained from 181 patients. Of these 181 (91%) were still using CPAP and 16 (9%) were not using CPAP. The compliant group’s frequency score fell from a mean of 3.38 to 1.75 (p < 0.001). Patients not using CPAP had no significant change, 3.56–3.44. (p = 0.55). Interestingly, there was a correlation between CPAP pressure and improvement in the frequency score (r = 0.70; p < 0.001).

Ing et al. [69] studied 63 patients with OSA (AHI > 15) and 41 patients without OSA (AHI < 5) with esophageal pH monitoring simultaneous with polysomnography. Patients with OSA had more gastroesophageal reflux events than patients without OSA (115 vs. 23; p < 0.001). In addition, OSA patients spent a higher percentage of their recording time at a pH of less than 4.0 (21.4 vs. 3.7%; p < 0.001).

The pathophysiology of the relationship of nGER to OSA is suggested by the effect of CPAP pressure. Intrathoracic negative pressure, as compared to abdominal pressure can show marked negativity with attempts to breath against a closed upper airway. CPAP treatment not only removes upper airway obstruction, thus reducing negative intrathoracic pressure, but also increases intrathoracic pressure with the use of higher CPAP pressures. This has a direct effect as esophageal pressure exceeds gastric pressure.

7. Asthma

A bidirectional link between bronchial asthma and obstructive sleep apnea (OSA) has been reported since the initial single case report by Hudgel and Shucard in 1979 [72].

Teodorescu et al. [73] recruited 472 subjects out of the allergy and pulmonary subspecialty clinics of the University of Wisconsin-Madison with a diagnosis of asthma who did not have co-morbid lung disease or were not under treatment for OSA. The subjects completed the Asthma Control Questionnaire (ACQ) [74] and the Sleep Apnea scale of the Sleep Disorders Questionnaire (SA-SDQ) [75]. Abnormal scores were ≥ 1.5 for the ACQ and ≥ 36 (men) or ≥ 32 for women for the SA-SDQ. High SA-SDQ scores were associated with 3.60 times higher odds of having a high ACQ score. With adjustments for obesity, race, nasal polyps, GERD, and psychopathology the odds of not-well-controlled asthma were still 2.87 times higher in the high OSA risk group.

Teodorescu et al. [76] performed a prospective study with 547 participants recruited from a random sample of Wisconsin state employees who were enrolled in the Wisconsin Sleep Cohort Study [2]. The presence or absence of OSA was assessed by polysomnography performed on enrollment and every 4 years thereafter. None of the 547 participants had OSA on their first polysomnogram. The presence of asthma in 81 participants was assessed by a questionnaire. At their first 4-year interval 22 (27%) participants with asthma had developed OSA while 16% of the asthma-free participants had developed OSA. This was statistically significant (p = 0.02).

The pathophysiology that links asthma and OSA are hypothetical. Dixit [77] postulated pathways that included increased parasympathetic tone, hypoxia-related reflex bronchospasm, altered nocturnal neurohormonal secretion, increased inflammatory mediators, gastroesophageal reflux, and obesity. Also included were adverse actions of inhaled corticosteroids producing increased upper airway adiposity or localized steroid myopathy of dilator muscles. The pathophysiology of this overlap continues to be discussed [19, 78].

8. Cognitive disorder

Several authors have reviewed the association of obstructive sleep apnea (OSA) with cognitive disorders [79, 80, 81]. The risk of OSA for the future development of Alzheimer’s disease remains prominently discussed [79]. The question of whether treatment of OSA with CPAP will improve cognitive status and improve risk of further deterioration is only beginning to be answered.

Dalmases et al. [82] performed a randomized, evaluator-blinded, parallel-group, single center study of 31 patients (69.7% male) who were at least 65 years old. The two arms of the study looked at newly diagnosed OSA patients (apnea hypopnea index mean average of 55.49 ± 17.63 events per hour). Patients were excluded if they had a Mini Mental Status Examination of less than 24, respiratory failure, neurologic or psychiatric disorders, chronic heart failure, unstable illnesses, other sleep disorders, contraindications for MRI, and inability to respond to questionnaires. Out of the 51 patients assessed, 18 patients were excluded (3 for declining to participate). The 33 eligible patients were divided into two groups. One group was managed with sleep hygiene and dietary counseling and the second group was placed on continuous positive airway pressure (CPAP) in addition to sleep hygiene and dietary counseling. All patients received neuropsychological examination and T1 high resolution MRI and 5-minute resting state functional MRI testing at baseline and at 3 months. One patient in each group was lost to follow-up and not included in the final assessment. While there were no significant differences between the two groups at baseline, the CPAP treatment group showed improvement in short-term memory (p = 0.032) and executive functioning (p = 0.014). The CPAP treatment group showed improvement in the speed of mental processing (p = 0.007) and mental flexibility (p = 0.008). Functional MRI revealed a significant increase in the intensity of connectivity (p = 0.012) between the posterior cingulate cortex and precuneus, the parahippocampal gyrus, and the middle and medial frontal gyrus (p = 0.013). There were no significant changes in the Epworth Sleepiness Scale (ESS) between groups.

Werli et al. [83] looked specifically at the effect of residual excessive daytime sleepiness (ESS > 10) in patients (ages from 30 to 65 years old) with moderate to severe OSA (AHI > 20) all of whom were being treated with CPAP. The comparison groups were 15 patients with excessive daytime sleepiness (EDS) versus a control group of 15 patients without excessive daytime sleepiness. There was no significant difference in the baseline AHI and ESS, however the multiple sleep latency test did reveal the average sleep latency in the control group to be 9.4 ± 2.7 minutes versus the average sleep latency in the EDS group of 2.9 ± 2.34. There was no significant difference in CPAP adherence to therapy. The EDS group performed poorly in executive functions, including such functions as handling of information, inadequate planning, judgment, and decision making, plus inflexibility, impulsivity, and difficulty maintaining motivation. The control group did not show evidence of executive function deficits. The prognostic implications of excessive daytime sleepiness in OSA continue to be an area of study and discussion.

There are several rationales for the finding of cognitive deficits in OSA [84, 85, 86]. Edwards et al. [84] showed that higher night-time cortisol levels were associated with greater cognitive impairment in patients with moderate to severe OSA (mean AHI 30.3 ± 21.7 events per hour). Ciccone and Mehra [85] presented an unpublished study showing the number of minutes during the sleep period with low oxygen levels was significantly associated with lower scores on the Montreal Cognitive Assessment (MOCA) test. There is a growing literature on cognition and OSA [20, 21]. OSA must be considered as an aggravating factor in patients presenting with a mild cognitive disorder.

9. REM sleep behavioral disorder

Rapid eye movement (REM) sleep behavioral disorder is a sleep parasomnia associated with dream enactment. It has been associated with neurodegeneration of long tracts originating in pontomedullary regions that send excitatory signals to the glycinergic neurons of the spinal ventral horns to hyperpolarize spinal motor neurons producing paralysis due to loss of muscle tone [16]. Idiopathic REM sleep behavioral disorder (RBD) is caused by a group of neurodegenerative disorders called the α-synucleinopathies, including Parkinson’s disease, multiple system atrophy, and dementia with Lewy bodies [87].

Secondary disorders mimicking idiopathic RBD include dream enactment behaviors caused by certain drugs, such as selective serotonin reuptake inhibitors and brainstem injuries, tumors, vascular lesions, or inflammation [16]. Iranzo and Santamaria [88] described 16 patients who were identified with dream enactment behaviors in addition to snoring and excessive daytime sleepiness. Polysomnograms did not show REM sleep without atonia but did show severe obstructive sleep apnea (OSA) with mean apnea hypopnea index (AHI) of 67.5 ± 18.7 events per hour. The abnormal sleep behaviors occurred only during apnea-induced arousals. The authors felt that the polysomnographic finding of REM sleep without atonia in the neurodegenerative disorders allowed dream enactment of neurodegenerative disorders to be distinguished fromdream behaviors of severe OSA. Dream enactment associated with OSA gained the name “pseudo-RBD.” Gabryelska et al. [17] mailed a questionnaire to 120 patients (85% male) with diagnosed RBD. One hundred and seven (89.2%) of the patients had a diagnosis of OSA with an AHI greater than 5. Of the 72 patients who responded to the questionnaire, 27 were using CPAP. Of the 27, 45.8% reported CPAP therapy improved their dream enactment.

Pseudo-RBD can present as RBD, however, the treatment of pseudo-RBD with clonazepam or another benzodiazepine can be contraindicated due to suppression of arousals to breathe, especially in the absence of CPAP.

10. Paroxysmal atrial fibrillation

Paroxysmal atrial fibrillation has been recognized as having a strong association with obstructive sleep apnea (OSA) [24, 25, 89, 90]. Atrial fibrillation (chronic and paroxysmal) is the most common cardiac arrhythmia. It can result in embolism to the brain and body. The Sleep Heart Health Study compared 228 patients with sleep-disordered breathing (SDB) (mean AHI 44.7 ± 13.1; 50.88% male) with 338 patients without SDB (mean AHI 2.7 ± 1.4; 47.00% male). This comparison showed a higher incidence of atrial fibrillation in patients with OSA than in the general population [91]. If adjustments were made for age, body mass index (BMI), hypertension, and congestive heart failure, the odds ratio for atrial fibrillation in the OSA group was 4.02.

Anter et al. [24] studied 86 patients with paroxysmal atrial fibrillation including 43 patients with moderate to severe OSA (AHI ≥ 15) and 43 patients without OSA (AHI < 5). The two groups underwent detailed electrophysiologic mapping and ablation protocol which included pulmonary vein (PV) isolation plus ablation of extra-PV triggers. An atrial fibrillation trigger site was defined as a site that produces an atrial premature depolarization (APD) triggering episodes of atrial fibrillation (AF) lasting ≥30 seconds. At baseline the PV was the most frequent trigger site in both groups and PV isolation was achieved in all patients. On repeating the protocol to identify additional triggers 18 of 43 patients with OSA had residual extra-PV triggers compared to 5 of 43 patients without OSA. The association of AF and OSA was felt to be secondary to electric and structural bi-atrial remodeling predominantly in the anterior septum.

The pathophysiology of atrial fibrillation is becoming clearer from animal studies. The heart has dual autonomic nervous system innervations. The sympathetic nervous system causes the heart to speed up and vagal nerve parasympathetic nervous system causes the heart to slow down among other effects. Occlusion of the upper airway leads to exaggerated negative intrathoracic pressure swings, intermittent hypoxia, atrial stretching, and cortical arousals with proarrhythmic cycles of parasympathetic and sympathetic nervous activities.

Figure 1 is a routine electrocardiogram (ECG) from a patient in our laboratory. The ECG is normal without any evidence of an abnormality that would predispose to a cardiac arrhythmia. Figure 2 is the initial recording from the diagnostic polysomnogram. Figure 3 shows the appearance of prolonged apneas, however, there is continuation of normal sinus rhythm. Figure 4 shows the appearance of two atrial premature contractions occurring quickly during a prolonged apnea with initiation of atrial flutter. Figure 5 shows atrial flutter deteriorating to atrial fibrillation. Figure 6 is post initiation of CPAP and atrial fibrillation spontaneously converts to normal sinus rhythm.

Figure 1.

This is the baseline electrocardiogram showing normal sinus rhythm with no atrial abnormality.

Figure 2.

The polysomnogram begins and the patient falls asleep with early signs of respiratory arousals.

Figure 3.

Prolonged apneas develop with occasional premature beats.

Figure 4.

After two premature beats occurring consecutively, the rhythm changes to atrial flutter.

Figure 5.

CPAP is initiated. Atrial flutter with a rapid ventricular response changes to atrial fibrillation.

Figure 6.

Spontaneous conversion back to normal sinus rhythm on CPAP at `5 cm of water. The patient would need to be treated with a final CPAP pressure of 10 cm of water.

11. Conclusion

Obstructive sleep apnea is a common illness that often flies under the radar of clinicians, but the comorbidities discussed above may be common in clinical practice. The danger for clinicians is that patients may not mention their sleep problems in the outpatient evaluation and management unless specifically asked. Patients may also be concerned about the cost of testing and treatment and resistant to referral. Treatment of the comorbidity alone may fail to resolve the co-morbidity in the absence of treating sleep related breathing, underlying or aggravating the presenting complaint.

The increasing use and sophistication of inexpensive personal monitors and home testing devices may be very helpful for the cost concern, however, public education about the risks of sleep related breathing disorders and their comorbidities will require continuing efforts.

Conflict of interest

The author declares no conflict of interest.