A high arousal threshold is associated with nocturnal gastroesophageal reflux in obstructive sleep apnea

Abstract

Objective

Nocturnal gastroesophageal reflux (nGER) was reported to be associated with obstructive sleep apnea (OSA). However, pathophysiological traits in OSA patients with nGER were poorly understood, and arousal threshold (ArTH) might play an important role in this relationship. This study aimed to identify the clinical characteristics of patients with OSA comorbid with nGER, and investigate the association between ArTH and nGER in patients with OSA.

Methods

482 adult patients with OSA diagnosed by polysomnography (PSG) at Peking University People's Hospital were recruited in this cross-sectional study. nGER presence was defined as typical nGER symptoms (heartburn or acid regurgitation occurring at night which affects sleep, with or without medication treatment) at least once a week in the past four weeks. A low ArTH was predicted based on the following PSG features: a total score of 2 or more, with one point assigned for each of the following criteria: apnea-hypopnea index (AHI) < 30/h, nadir oxygen saturation > 82.5 %, the proportion of hypopneas > 58.3 %, otherwise the ArTH was classified as high. Demographic and anthropometric characteristics, PSG parameters, and questionnaires related to clinical symptoms and comorbidities were obtained. The associations between nGER and clinical features were examined by multivariate logistic regressions.

Results

The prevalence of typical nGER symptoms in patients with OSA was 17.8 %. High ArTH (P = 0.006), BMI ≥28 kg/m² (P = 0.023), the Epworth Sleepiness Scale (ESS) Scores >16 (P = 0.006), and age over 60 years old (P = 0.002) were associated with a markedly increased risk of nGER in patients with OSA. Notably, in non-obese patients with OSA, a high ArTH not only significantly added to the presence of nGER (OR = 2.905, 95 % CI: 1.352–6.241, P < 0.01) after controlling for age, sex, ESS scores, comorbidities (including hypertension, diabetes mellitus, hyperlipidemia, coronary artery disease, and stroke) and total arousal index, but also associated with nGER symptom frequencies (OR = 2.856, 95 % CI = 1.118–7.294, P = 0.028) after adjusting for age, ESS scores, and medication use.

Conclusions

A high ArTH is independently associated with nGER in patients with OSA. For non-obese patients with OSA, a high ArTH is not only associated with the presence of nGER but also correlated with nGER symptom frequencies.