Obstructive Sleep Apnea and Reflux Disease: Bedfellows at Best FREE TO VIEW

Gastroesophageal reflux disease (GERD) is very common. The more loosely defined, the more common it is. In a telephone survey accessing >20,000 randomly selected US residents, 22% report having heartburn at least once a month.¹ Even using a more restrictive definition, GERD is common. A population-based endoscopy study of 1,000 Scandinavians found a 15.5% prevalence of esophagitis.² Herein lies the problem in examining the relationship between GERD and obstructive sleep apnea (OSA). Are we simply looking at two common conditions that may well coexist in an individual? Do they commonly coexist because of shared risk factors? Or does a causal relationship exist between them?

OSA occurs in about 4% of adult men and 2% of adult women.³ One recent analysis of the risk factors for OSA identified male sex (odds ratio 3.7, CI 1.63-8.19), age >65 years (odds ratio 2.4, CI 0.65-8.54), and BMI (odds ratio 1.5, CI 0.68-3.16) as being the most significant.⁴ Among these, the most obvious shared risk with GERD is obesity.⁵ Also supporting an association between GERD and OSA is a 24-h esophageal pH monitoring study of 16 patients with OSA, which demonstrated 80% of them had abnormally high esophageal acid exposure times.⁶ Furthermore, a study of 16 patients with both OSA and GERD found that improving their breathing mechanics with continuous positive airway pressure (CPAP) normalized the esophageal acid exposure in 81% of them and reduced the mean acid exposure time of the group from 12.4% before CPAP to 6.8% with CPAP.⁷ Thus, it seems reasonable to conclude that GERD is more common in patients with OSA than in the general adult population and that improving the ventilation of patients with OSA with CPAP not only improves their breathing but also reduces the occurrence of gastroesophageal reflux during sleep. Or does it?

It is precisely this question that Kuribayashi et al⁸ seek to shed light on in this issue of CHEST (see page 769). They performed a detailed physiologic investigation using concurrent high-resolution manometry spanning from the pharynx to the abdomen, pH-impedance reflux monitoring, and polysomnography in patients with and without all possible combinations of OSA and GERD to evaluate unique physiologic mechanisms for reflux that were operational among the patients with OSA. Instead, they found that physiologic compensations prevented reflux during apneic events even though the magnitude of the abdominal-to-chest pressure differential increased with labored breathing, as had been previously reported.⁹ Specifically, crural diaphragm contractions became increasingly vigorous, augmenting the antireflux barrier and preventing reflux during apneic events. This is, in fact, what one would predict from prior physiologic studies of the crural diaphragm. The crural diaphragm is a very responsive component of the esophagogastric junction, protecting against gastroesophageal reflux in instances of abdominal straining, coughing, and augmented respiratory effort.¹⁰

How then to explain the negative findings of Kuribayashi et al? One need only examine Table 1 of their treatise to resolve this seeming contradiction. The median number of reflux events during sleep among the 20 patients with GERD (with or without OSA) studied (one event) was no different than the number of reflux events observed in the 15 control subjects (two events) or in the subjects with OSA and without GERD (one event). Evidently, for whatever reason, these patients had minimal reflux during the night of the study. Further suggesting that they had mild disease, many of them were defined as patients with GERD on the basis of exhibiting symptomatic response to therapy rather than by pH monitoring, none of them had a fixed hiatus hernia, and as a group, their mean BMI was no different than for the control population. Contrast this to the patients studied by Tawk et al,⁷ who were awakening from sleep with heartburn at least once per week, had a mean BMI of 35.1, and had a baseline esophageal acid exposure time of 12.4%, making it a virtual certainty that most if not all of them had a hiatal hernia. Quite simply, the two studies were looking at dissimilar GERD populations.

So, does a causal relationship between OSA and GERD exist? Probably not. Rather, I would concur with the conclusions of Morse et al⁴ that GERD and OSA are common entities that share similar risk factors without being causally linked. Both populations tend to be overweight or obese. Obesity causes increased intraabdominal pressure, an increased pressure differential between the abdomen and chest during inspiration, and increased anatomic disruption of the esophagogastric junction (hiatus hernia).¹¹ Hiatus hernia predisposes to reflux by the strain mechanism (exemplified by labored breathing), with the magnitude of the risk being proportional to the size of the hernia.¹² Consequently, it is probably very unusual that reflux causes OSA or even apneic events (apart from the occasional episode of overt aspiration). The argument is much more compelling that people with OSA are more likely to be obese and more likely to have hiatus hernia. Obese people with hiatus hernia are more prone to nocturnal reflux with prolonged esophageal acid exposure and more prone to strain-induced reflux events. Obese people with hiatus hernia and OSA more frequently challenge their esophagogastric junction with straining on account of their frequent episodes of labored breathing during apnea. Basically, it is the hernia, and although the patients studied by Kuribayashi et al may have met the definition of reflux disease,¹³ none of them had significant hernias.