Discussion:
The relationship of LPR with and its potential role in the pathogenesis of OSAS have not been fully elucidated. The originality of the present study is the realization of both HEMII-pH and PSG in the same time in patients with untreated sleep and LPR findings. We found that patients with LPR and OSAS have different HEMII-pH and clinical patterns compared with LPR individuals without OSAS. Precisely, OSAS patients had a lower number of HREs and less severe otolaryngological symptoms compared with LPR patients. Moreover, there were no significant associations between nighttime HREs and PSG data and between the severity of LPR and OSAS. We only observed associations between ESS, RSS and the number of evening HREs, which supports potential relationship between both conditions without providing pathophysiological explanations. The association between pH-study and PSG findings was investigated in many studies,14-20 which reported controversial results. Some authors observed that patients with severe OSAS had a significant higher number of nocturnal proximal or distal esophageal events compared with patients with mild OSAS,14,16,17 while others did not find any association between reflux and OSAS findings.15,18,19,21-23 The inconsistencies may result from methodological differences across studies.
Some authors used dual- or triple-probe pH monitoring to evaluate reflux events and did not consider nonacid HREs.18,19 Others only included patients with GERD (distal probe findings)14,20,23 or esophagitis.21Among studies reporting gastrointestinal findings, the prevalence of patients with GERD or GERD and LPR varied from 38.9% to 100% of cases.14,15,18,20,21 From an epidemiological standpoint, the use of different inclusion criteria and diagnostic approaches may have a substantial impact on the results of the study.
First, LPR is defined by the occurrence of acid, weakly acid and nonacid HREs.10 Nonacid HREs concern more than 50% of LPR patients and are not detected by pH monitoring without impedance.4 The lack of consideration of nonacid/weakly HREs may lead to a bias selection through the exclusion of patients with nonacid/weakly LPR.
Second, it has been reported that patients with both GERD and LPR have different profile at the HEMII-pH with higher proportion of nocturnal and supine HREs in GERD patients.4 Moreover, patients with GERD may have a significant higher probability to have acid LPR, while patients without GERD had equal proportions of acid, nonacid and weakly acid LPR.4 In that respect, the consideration of GERD as a key inclusion criteria may lead to the introduction of an important bias because many patients with GERD have no LPR or vice versa .1,4 To have a representative sample of LPR population, authors should use HEMII-pH that detects GERD, acid, weakly acid and nonacid HREs.
Third, the severity of LPR was evaluated through different patient-reported outcome questionnaires including GERD-related questionnaires,5,7,15 reflux symptom index,6,14,17,21,22 or LPR-health-related quality of life.22 These patient-reported outcome questionnaires are all validated but do not include similar symptoms and rating approaches, yielding the comparison across studies difficult. According to these reasons, the comparison of our results with the literature remains limited. Only Xiao et al . used HEMII-pH monitoring to investigate the relationship between HREs and PSG findings,15 and they did not find significant objective association, corroborating our observations. Our findings as well as the observations of Xiao et al . reinforce the controverse about the pathophysiological link between LPR and OSAS.
For a long time, it has been proposed that reflux is thought to be induced by decreasing intraesophageal pressure during obstructive sleep apnea, which causes a vacuum-like effect on the gastric contents. This theory was strengthened by the observation that during sleep, there would be a delay in the nocturnal acid clearance, facilitated by the suppression of salivation and esophageal peristalsis.24 However, recent data supported that despite a decrease in esophageal body pressure during obstructive apnea events, compensatory changes in upper esophageal sphincter and gastroesophageal junction pressures prevent reflux.25The lack of significant association between HREs, AHI and arousal findings of the present study may strengthen these findings. According to our observations and those of studies using HEMII-pH or high-resolution pharyngeal manometry,25 we believe that there would exist several pathophysiological mechanisms depending on the lower and upper esophageal sphincter tonicity, the presence of GERD and the type of HREs (liquid versus gaseous). The investigation of the type of HREs makes particularly sense because the supine position would be more favorable for liquid reflux events that are more aggressive for the laryngopharyngeal mucosa because they contain more gastroduodenal enzymes.4 In that way and according to our subgroup analysis, the correlation between reflux and OSAS severity could particularly concern patients with GERD and liquid HREs; the overweight being a favoring factor. Naturally, this hypothesis needs to be confirmed through future studies considering the type (liquid versus gaseous) and the nature of HRE (acid, weakly acid, nonacid) and the time of occurrence of HREs.
In our study, we observed significant association between the number of HREs in the evening, RSS and ESS score. The occurrence of evening HREs and the deposit of gastroduodenal enzymes into the laryngopharyngeal mucosa could be associated with the development of a more significant inflammatory reaction before sleep and related sleep (ESS) disorders. This hypothesis cannot be confirmed with our data but our findings support the need to consider daytime and nighttime periods of occurrence of HREs and not only the nighttime reflux events. Moreover, the measurement of pepsin saliva in the morning could provide additional useful analysis in the understanding of the mechanisms underlying the association between LPR and OSAS.6
With regard to the clinical presentation of patients with OSAS and LPR, our observations supported that OSAS patients may have less severe clinical picture than LPR patients. To the best of our knowledge, only Teklu et al . specifically investigated the symptom profile of LPR patients with or without OSAS, reporting that patients with OSAS had worse symptoms as measured by RSI.7 As above-mentioned, the inclusion of a high number of patients with GERD in the study of Tekly et al . makes the comparison difficult. Another finding highlighted in the present study is the higher proportion of hiatal hernia and potential related esophagitis in LPR group compared with patients with LPR and OSAS. In the LPR general population, esophagitis and hiatal hernia concern 20% to 30% of patients.1 According to our observations, patients with LPR and OSAS may have a lower proportion and the potential mechanisms are still unknown.
In this study, we tried to explore the relationship between the occurrence of arousals/awakenings and HREs. However, we only had 3 patients with nighttime HREs and among them, the sleep events were not associated with HREs. The low number of patients with nighttime HREs is not surprising regarding previous data.4 The lack of association between sleep and reflux events corroborates the findings of a previous study where authors observed that 82% of reflux events occurred during a PSG epoch that was classified as wake.26 Authors also reported that arousals/awakenings preceded almost all reflux events, whereas fewer had an arousal/awakening after the event.26
The primary limitations of our study are the low number of patients and the lack of investigation of laryngopharyngeal signs. In practice, many patients did not accept to benefit from HEMII-pH and PSG in the same time and declined one of two examinations. However, the realization of both examinations in the same period is important to study the relationship between both conditions. To date, only a few studies prospectively investigated the relationship between LPR and OSAS through simultaneous pH-study and PSG.15,16,18,21 This methodological point is particularly important because the reproducibility of the HEMII-pH results from one day to the other is not guaranteed. The number of HREs may depend on the patient diet,4 or, hypothetically, on the patient stress the day of the testing that may be both associated with esophageal sphincter relaxations.1 The primary strengths of the study are the prospective design and the use of HEMII-pH for the diagnosis. The consideration of acid, nonacid and weakly acid HREs led us to include all patients with LPR and not only a subgroup of individuals (acid LPR). Moreover, in this study we deliberately use the full and not the short version of RSS (RSS-12) because the consideration of otolaryngological, digestive and respiratory symptoms makes sense in patients with OSAS that commonly includes respiratory and, in some GERD patients, digestive symptoms.