Introduction
Chronic cough is defined as a cough lasting more than eight weeks. Worldwide, the major causes of chronic cough include upper airway cough syndrome (UACS), asthma, nonasthmatic eosinophilic bronchitis (NAEB), and gastroesophageal reflux disease (GERD), or a combination of these disorders (1). In contrast, cough-variant asthma (CVA), sinobronchial syndrome (SBS), atopic cough and GERD are the major causes of chronic cough in Japan (2).
Recent studies have suggested an association between chronic cough and obstructive sleep apnea (OSA). Although different diagnostic criteria and measurement methods have been used across studies, OSA is common in the general population, with prevalence estimates of 9% in women and 24% in men among U.S. adults aged 30-60 years (3) and 32.7% among Japanese adults aged 30-69 years (4). Several studies conducted outside Japan have reported a high prevalence of OSA among patients with chronic cough (5, 6).
One study revealed that 44% of patients with chronic cough had OSA, and 93% of those treated with CPAP improved their symptoms (5). Another study also demonstrated that the prevalence of chronic cough in the OSA group was significantly greater than that in the non-OSA group (39.4% vs. 12.5%, p=0.005) (6).
In that study, the OSA group had a higher body mass index (BMI, kg/m²) (28.9±4.1 vs. 24.9±4.3) than the non-OSA group did, and the prevalence of GERD was also significantly greater in the OSA group, while the prevalence of UACS and asthma was similar between the two groups. Because obesity is an important risk factor shared by both OSA and GERD, these findings suggest that obesity-related factors may contribute to the association between OSA and chronic cough.
However, because obesity-independent mechanisms of OSA are common in Japan, analyses from this population may provide unique insights into the association between chronic cough and OSA. In this study, we investigated the prevalence and clinical characteristics of suspected OSA in patients with chronic cough using a portable sleep apnea monitoring system.
Patients and Methods
Study design and patients. This was a multicenter, prospective observational study. Outpatients with chronic cough were recruited at three institutions in Japan between March 2023 and February 2025.
The eligibility criteria were as follows: patients aged 20 years or older with a cough lasting more than eight weeks and those who provided written informed consent. The exclusion criteria were as follows: those with abnormalities on chest radiography or computed tomography, those with dementia or psychiatric conditions requiring treatment, or those already diagnosed with OSA. Clinical data, including age, sex, body mass index (BMI), duration of cough, allergy history, medication history, smoking history, causes of chronic cough, Epworth Sleepiness Scale (ESS), visual analog scale (VAS) scores for cough severity, and the Japanese version of the Leicester Cough Questionnaire (J-LCQ), were obtained from medical records and questionnaires.
The LCQ is a 19-item questionnaire comprising three health domains –physical, psychological and social (7)- and the J-LCQ has demonstrated good validity in Japanese patients (8).
Outcome. The primary outcome was the prevalence of suspected OSA, which was assessed using the respiratory event index (REI) in Japanese patients with chronic cough. Exploratory analyses included comparisons of REI across causes of chronic cough and the associations between REI and age, BMI, VAS scores for cough severity, and J-LCQ scores.
Sleep study. Participants were subjected to overnight home sleep apnea testing (HSAT) with a portable sleep apnea monitoring system (SAS-2200; Nihon Kohden, Tokyo, Japan; distributed by Teijin Pharma, Tokyo, Japan) to assess sleep-disordered breathing. The evaluation focused on parameters such as nasal airflow and oxygen saturation (SpO₂).
The REI was defined as the total number of apneas and hypopneas per hour of monitoring time. In accordance with the American Academy of Sleep Medicine (AASM) criteria (9), apnea and hypopnea events were defined as a ≥10-second reduction in airflow amplitude, with hypopnea additionally requiring an associated ≥3% decrease in SpO₂. OSA was suspected when the REI was ≥5 events/h.
We also assessed the 3% oxygen desaturation index (3% ODI) and minimum SpO2. To examine whether patient characteristics and cough-related symptoms differed according to OSA severity, we divided the patients into two groups on the basis of an REI of 15 events/h, which is the threshold separating mild from moderate-to-severe OSA.
Statistical analysis. All statistical analyses were performed using EZR software (10), version 1.70 (Saitama Medical Center, Jichi Medical University, Saitama, Japan). Data are expressed as medians and interquartile ranges (IQRs).
The Mann-Whitney U test was used for continuous variables. Categorical variables were compared by Fisher’s exact tests. Spearman’s rank correlation coefficient was used to assess the associations between the REI and clinical variables, including age, BMI, VAS scores for cough severity, and J-LCQ scores. p-Values <0.05 were considered to indicate statistical significance.
Ethics approval and consent to participate. Written informed consent was obtained from all participants. This study was approved by the Kobe University Ethics Committee (reference number: B220175) and by the other participating institutions.
This study was conducted in accordance with the Declaration of Helsinki.
Results
Patient characteristics and assessment of cough-related symptoms. In total, 23 outpatients with chronic cough were enrolled in the study across three institutions. As summarized in Table I, 11 (47.8%) patients were men, and 5 (21.7%) were smokers.
The median age was 56.0 years (IQR=43.5-69.0), and the median BMI was 25.2 kg/m2 (IQR=22.3-28.4). The major causes of chronic cough included asthma/CVA, GERD, and UACS, followed by SBS. The median duration of cough was 11.2 months (IQR=5.2-29.5).
The median VAS score for cough severity was 50.0 mm (IQR=33.0-73.5), and the median J-LCQ total score was 12.4 out of 21 (IQR=8.3-15.6), indicating reduced cough-related quality of life; the normal range of the J-LCQ score was 16-21. The median ESS score was 7.0 (IQR=4.0-9.0), and most patients were within the normal range (≤10 points).
Sleep apnea monitoring parameters. The median REI was 15.6 events/h (range=2.3-103.0; IQR=11.0-26.1). An REI ≥5 events/h, which is indicative of OSA, was observed in 20 of 23 patients (87.0%).
An REI ≥30 events/h was present in four patients (Figure 1). The median 3% ODI was 18.2 events/h, and the median minimum SpO2 was 72.0% (Table II). A 3% ODI ≥5 events/h was observed in 22 of 23 patients (95.7%).
When patients were grouped according to the cause of chronic cough, the median REIs for asthma/CVA, GERD, and UACS were 15.1, 18.7, and 16.5 events/h, respectively. The median 3% ODIs for asthma/CVA, GERD, and UACS were 20.7, 21.8, and 23.3 events/h, respectively.
Patient characteristics and cough-related symptoms according to OSA severity. When the data were stratified by REI <15 events/h and REI ≥15 events/h, patients in the REI ≥15 events/h group were significantly older (p=0.04), whereas BMI did not significantly differ between the groups (p=0.32) (Table III). The duration of cough, VAS score, and J-LCQ score also did not significantly differ between the two groups.
Spearman’s rank correlation coefficient analysis revealed a significant association between age and REI (ρ=0.58, p=0.003;), whereas J-LCQ scores were not significantly associated (ρ=0.17, p=0.43). BMI and VAS score for cough severity were also not significantly correlated with the REI (BMI: ρ=0.26, p=0.22; VAS: ρ=0.04, p=0.85).
Discussion
In our study, an REI ≥5 events/h was observed in 20 of 23 patients with chronic cough (87.0%), and a 3% ODI ≥5 events/h was observed in 22 of 23 patients with chronic cough (95.7%). Previous reports have shown that approximately 40% of men and 20% of women in the general Japanese population have a 3% ODI ≥5 events/h (11, 12). In comparison, the proportion of patients with a 3% ODI ≥5 events/h in our cohort was markedly greater, suggesting a potentially high prevalence of OSA.
Notably, the median ESS score was 7.0 (IQR=4.0-9.0), which was within the normal range, despite the high prevalence of elevated REI. These findings indicate that OSA may be highly prevalent even in patients without daytime sleepiness, supporting the need to consider OSA screening in the evaluation of chronic cough.
When REI across different causes of chronic cough was compared, it tended to be high not only in patients with GERD, but also in those with asthma/CVA and UACS. Although the small sample size limited the statistical analysis, our findings suggest that OSA may occur in patients with chronic cough regardless of the underlying cause. The median BMI of the patients in our study was 25.2 kg/m² (IQR=22.3-28.4), which was lower than the mean BMI of 35 ± 7 kg/m² reported for the OSA group with chronic cough in a previous study (5).
Nevertheless, the high prevalence of OSA observed in our study suggests that factors other than obesity may contribute to the development of OSA in patients with chronic cough. Furthermore, BMI did not significantly differ between the groups with an REI <15 events/h and an REI ≥15 events/h (p=0.32), indicating that obesity alone does not account for the elevated REI.
The median age of our cohort was 56.0 years. When stratified by OSA severity, the REI <15 events/h group had a median age of 45.0 years, whereas the REI ≥15 events/h group had a median age of 66.0 years. In comparison, a previous study reported similar mean ages among patients with chronic cough in the non-OSA and OSA groups (48.3 and 52.2 years, respectively) (6), indicating that our cohort was relatively older.
Therefore, the relatively older age of our cohort may have contributed to the high prevalence of suspected OSA. Patients in the REI ≥15 events/h group were significantly older (p=0.04), and age was positively correlated with the REI (ρ=0.58). These findings are consistent with those of previous epidemiological studies reporting that OSA severity tends to increase with advancing age (13).
This association may be explained by age-related changes that increase upper-airway collapsibility, including loss of teeth, reduced pharyngeal muscle responsiveness, unstable respiratory control, and a lower arousal threshold (14). These findings suggest that, even among patients with chronic cough, advanced age may be an important factor contributing to a high REI.
Although the prevalence of obesity is lower in Japan than in Western countries, several studies have shown that Japanese individuals exhibit a relatively high prevalence of OSA. These observations may be partially explained by the craniofacial features of Japanese patients, such as a smaller, posteriorly positioned mandible, an inferiorly positioned hyoid bone, and upper-airway narrowing (15, 16). A non-obese clinical phenotype of OSA has also been identified in Japanese populations, suggesting that obesity-independent mechanisms are common (17).
The mechanisms linking chronic cough and OSA remain unclear, but both direct and indirect pathways have been proposed. The direct pathway involves upper-airway mechanical stress and inflammation. Repetitive airway closure and reopening not only lead to local mechanical stress but also result in inflammation of the mucosa (18).
An increase in the levels of proinflammatory cytokines, including IL-6 and IL-8, in pharyngeal lavages was reported in patients with OSA compared with those in healthy controls (19). OSA has been shown to be associated with upper-airway inflammation, which may contribute to postnasal drip and result in UACS-related chronic cough (5). In addition to this direct mechanism, several indirect pathways have been proposed.
In the GERD-related pathway, recurrent obstructive events generate marked negative intrathoracic pressure, promoting gastroesophageal reflux. Sleep fragmentation and intermittent hypoxia may impair lower esophageal sphincter function through autonomic dysregulation (20). In the asthma-related pathway, obstructive events can increase airway resistance and trigger vagally mediated bronchoconstriction, directly worsening lower-airway narrowing.
Intermittent hypoxia and oxidative stress may promote airway inflammation, contributing to noneosinophilic asthma phenotypes (21). A recent systematic review and meta-analysis demonstrated that patients with asthma have a significantly increased risk of OSA (22). On the basis of our findings, the high prevalence of OSA in patients with chronic cough may not be explained simply by obesity.
Taken together, these direct and indirect pathways suggest that OSA may contribute to chronic cough not only through obesity-related mechanisms but also through the induction or exacerbation of UACS, GERD, or asthma/CVA, as well as through direct effects on the airway.
Study limitations. First, this prospective observational study included only 23 patients from three institutions, resulting in substantial statistical imprecision and susceptibility to selection bias, and the study was not blinded. We did not include an age-, sex-, and BMI-matched control group without chronic cough; therefore, we cannot determine whether suspected OSA is more prevalent in patients with chronic cough than in comparable individuals without chronic cough.
The small sample also precluded reliable adjustment for age, sex, BMI, and other OSA risk factors, and therefore does not establish obesity-independent mechanisms. In addition, cough etiologies were overlapping and some subgroups were very small, limiting interpretation of cause-specific REI values.
Second, OSA was assessed by home respiratory polygraphy without electroencephalography rather than full polysomnography; consequently, REI was based on monitoring rather than confirmed sleep time and OSA severity may have been misclassified. Cough severity and J-LCQ scores may also have been influenced by treatment received before enrollment and by variation in the timing of assessment. Finally, patients were not followed after OSA treatment, so the study cannot determine whether treatment of OSA improves chronic cough or establish the direction or causality of the association.
The present findings should therefore be considered preliminary and hypothesis-generating.
Conclusion
In this study, the REI was high among patients with chronic cough, suggesting a high prevalence of OSA that may be related to factors other than obesity. Because OSA may contribute to chronic cough, routine screening for OSA should be considered in this population, even in the absence of daytime sleepiness.
Conflicts of Interest
All Authors declare that they have no conflicts of interest in relation to this study. The SAS-2200 devices were provided in-kind by Teijin Pharma. Teijin Pharma had no role in the study design, analysis, interpretation of data, or manuscript preparation.
Authors’ Contributions
NI drafted the manuscript. NI and MT performed the statistical analyses. NK provided advice on the study design and implementation.
TaN supervised the overall study. KO and TeN collected the data. All Authors contributed to the conception or design of the study and to interpretation of the data, and all Authors reviewed and approved the final manuscript.
Acknowledgements
The Authors gratefully acknowledge the cooperation and support of the staff at the following institutions: Division of Respiratory Medicine, Kobe University Hospital; Division of Respiratory Medicine, Akashi Medical Center; and Division of Respiratory Medicine, Kakogawa Central City Hospital. We sincerely appreciate the permission to use the J-LCQ granted by the original author, Dr. Surinder Birring, and by the translators, Dr.
Akio Niimi and Dr. Haruhiko Ogawa. The Authors also thank Teijin Pharma Limited for providing the SAS-2200 devices used in this study.
Funding
This study received no funding support.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.