Risk of Biliary Tract and Pancreatic Cancer Following Obstructive Sleep Apnea Diagnosis: Analysis of a National Health Insurance Database
Article information
Abstract
Background and Objectives
Obstructive sleep apnea (OSA) has been associated with an increased risk of cancer in various organs. OSA is also linked to chronic inflammation in the biliary tract and pancreas, a well-established risk factor for carcinogenesis in these organs. However, its relationship with biliary tract and pancreatic cancers remains unclear and has been rarely investigated. Therefore, we aimed to evaluate whether OSA serves as an independent risk factor for these malignancies by analyzing a nationwide healthcare claims database in South Korea.
Methods
A retrospective cohort study was conducted using the Korean National Health Insurance Service (KNHIS) database. Adults aged ≥20 years who were newly diagnosed with OSA (ICD-10: G47.30) between 2007 and 2014 were identified and propensity score-matched (1:5) with controls based on age, sex, and comorbidities. Individuals with pre-existing cancer diagnoses were excluded. The primary endpoints were the incidence of overall cancer, biliary tract cancer (C23–C24), and pancreatic cancer (C25). Cox proportional hazards regression models were used to calculate hazard ratios (HRs), adjusting for demographic and clinical factors.
Results
A total of 1,191,444 individuals were included, comprising 198,574 patients diagnosed with OSA and 992,870 matched controls. OSA was associated with an increased overall cancer incidence (HR, 1.132; 95% confidence interval [CI], 1.097–1.169); however, it was not significantly associated with pancreatic cancer (HR, 0.941; 95% CI, 0.823–1.072) or biliary tract cancer (HR, 0.931; 95% CI, 0.751–1.142). Subgroup analyses stratified by sex and age revealed no statistically significant associations across these groups.
Conclusion
Our findings do not support OSA as an independent risk factor for biliary tract or pancreatic cancers.
INTRODUCTION
Obstructive sleep apnea (OSA) is among the most common sleep disorders, characterized by recurrent episodes of upper airway collapse during sleep [1]. Airway narrowing and collapse lead to intermittent hypoxia and dysregulation of sympathetic tone, disrupting normal sleep patterns and contributing to various complications [2]. Globally, nearly 1 billion individuals are affected by OSA [3]. In the United States, approximately 25% to 30% of men and 9% to 17% of women are affected, whereas in South Korea, about 27% of men and 16% of women exhibit an apnea-hypopnea index (AHI) exceeding 5 per hour [4,5].
OSA is a major health concern due to its potential to initiate and aggravate numerous chronic comorbidities associated with metabolic syndrome, including hypertension, type 2 diabetes mellitus, stroke, and cardiac arrhythmias [3,6-8]. Furthermore, OSA interacts bidirectionally with inflammatory airway disorders, complicating their clinical management [9,10]. This complexity arises from the intricate interplay between the pathophysiological mechanisms of OSA and the development of these diseases, significantly affecting their progression and therapeutic responses [11].
Beyond these well-established complications, recent studies suggest that OSA may elevate the risk of malignant neoplasms in various organs [11-13]. Analyses of large datasets have demonstrated that the incidence of cancers in the thyroid, breast, lung, colorectal region, central nervous system, and hematologic system is higher in individuals with OSA than in those without the disorder, implying an increased risk of developing various cancers in OSA patients [14-21]. Potential underlying mechanisms include chronic intermittent hypoxia and heightened sympathetic activity, which can alter gene expression related to carcinogenesis and exacerbate inflammation, thereby promoting tumor development and progression [22,23].
Regarding diseases of the biliary tract and pancreas, OSA has been linked to various disorders, such as chronic inflammation in the biliary tract and non-alcoholic fatty pancreatic disease (NAFPD), both of which are established risk factors for cancer in these organs [24,25]. Nevertheless, the association between OSA and the risk of biliary tract and pancreatic cancers remains largely unexplored, with only a few studies specifically investigating this relationship [26,27]. Therefore, this study investigated whether a diagnosis of OSA is an independent risk factor for the development of biliary tract and pancreatic cancers using national healthcare insurance claims data from South Korea.
METHODS
Study design and data source
This retrospective cohort study utilized data from the Korean National Health Insurance Service (KNHIS) database, which contains extensive longitudinal health information including patient demographics, medical diagnoses, prescriptions, procedures, and healthcare utilization. Disease classifications followed the International Classification of Diseases, 10th Revision (ICD-10), as applied within the Korean healthcare system. The study protocol was reviewed and approved by the institutional review board (Investigation No. 2025-02-021), and the requirement for informed consent was waived due to the retrospective, anonymized nature of the dataset.
Study population
The study cohort comprised adults aged 20 years or older who were newly diagnosed with OSA (ICD-10 code G47.30) between January 1, 2007, and December 31, 2014. To ensure comparability, a propensity score-matched control group was selected at a 1:5 ratio based on age, sex, and major baseline comorbidities. Individuals with a prior cancer diagnosis, incomplete follow-up data, or missing clinical variables were excluded. Participants were followed from the index date—defined as the date of OSA diagnosis (or the corresponding entry date for matched controls)—until the earliest occurrence of a cancer diagnosis, death, or the study endpoint (December 31, 2015).
Primary endpoints
The primary outcome of this study was the incidence of overall cancer, as well as biliary tract and pancreatic cancers. Cancer cases were identified in the KNHIS dataset using ICD-10 codes. Any malignant neoplasm (C codes) was considered to define all types of cancer. For biliary tract cancer, gallbladder cancer (C23) was defined as malignancies originating from the gallbladder epithelium, and biliary tract cancer (C24) included intrahepatic bile duct cancer (C24.0) and extrahepatic bile duct cancer (C24.1), encompassing malignancies of both the intra- and extrahepatic biliary system. Pancreatic cancer (C25) included all malignant neoplasms originating from the pancreas. To enhance diagnostic accuracy and minimize misclassification, cases were confirmed by requiring both a corresponding ICD-10 cancer diagnosis and registration in the Korean National Medical Expenses Support Program, a nationwide registry for confirmed cancer cases requiring financial assistance.
Covariates
Demographic and clinical variables included age and sex, which were used as primary matching factors in the propensity score selection, and income level, which was stratified into quartiles based on national health insurance premium brackets. Comorbidities were identified using ICD-10 codes, including hypertension (I10–I15), which encompasses both essential (primary) and secondary hypertensive disorders; diabetes mellitus (E11–E14), covering type 1, type 2, and other specified or unspecified diabetes conditions; dyslipidemia (E78), including disorders of lipoprotein metabolism, hypercholesterolemia, and hyperlipidemia; chronic obstructive pulmonary disease (J41–J44), including chronic bronchitis and emphysema; ischemic heart disease (I20–I25), such as angina pectoris, myocardial infarction, and chronic coronary syndromes; and stroke (I63, I64), which includes ischemic stroke, cerebral infarction, and other cerebrovascular diseases. Standardized mean differences were calculated for all baseline characteristics to assess balance between the OSA and control groups.
Statistical analysis
Cancer incidence hazard ratios (HRs) were calculated as the number of new cancer cases per 1,000 person-years. The association between OSA and cancer risk was assessed using Cox proportional hazards regression models to estimate HRs and 95% confidence intervals (CIs). Additionally, a model incorporating demographic, clinical, and socioeconomic factors as covariates was applied to compute adjusted incidence rates, controlling for potential confounders. Kaplan-Meier survival curves were generated to visualize differences in cumulative cancer incidence, and statistical significance was evaluated using the log-rank test. All statistical analyses were performed using SAS (version 9.4, SAS Institute, Cary, NC, USA) and R (version 3.2.3, The R Foundation for Statistical Computing, Vienna, Austria), with statistical significance defined as a two-tailed p-value<0.05.
A subgroup analysis was also conducted. Patients were stratified by age (<40 years, 40–65 years, and ≥65 years) and sex, as these factors independently influence cancer epidemiology, including incidence, biological characteristics, and clinical outcomes. This categorization aligns with widely used epidemiological classifications, such as those adopted by the Korea Central Cancer Registry in the National Cancer Information Center of South Korea [28]. Sex stratification was performed because of recognized biological differences in cancer susceptibility driven by genetic and hormonal factors, facilitating an evaluation of potential sex-specific variations in the association between OSA and cancer risk.
Ethical declaration
The study protocol was reviewed and approved by the author’s institutional review board (IRB No. 2025-02-021), and the requirement for informed consent was waived due to the retrospective and anonymized nature of the dataset.
RESULTS
A total of 1,191,444 individuals were included in the study, comprising 198,574 patients diagnosed with OSA and 992,870 propensity score-matched controls. The mean follow-up duration was 4.7 years, and there were no significant differences in baseline characteristics, including age and sex distribution, between the groups. The overall cancer incidence rate was higher in the OSA group (5.183 per 1,000 person-years) compared to the control group (4.495 per 1,000 person-years). In the fully adjusted Cox proportional hazards model, which accounted for all demographic and clinical covariates, OSA was associated with an increased risk of overall cancer incidence (HR, 1.132; 95% CI, 1.097–1.169) (Fig. 1A and Table 1).
Cumulative incidence of biliary tract and pancreatic cancer by obstructive sleep apnea (OSA) status. A: Overall cancer incidence in individuals with and without OSA. B: Cumulative incidence of pancreatic cancers, stratified by OSA status. C: Cumulative incidence of biliary tract cancers, stratified by OSA status.
Association between OSA and cancer incidence
For pancreatic cancer, the incidence rate in the OSA group (0.285 per 1,000 person-years) was similar to that in the control group (0.292 per 1,000 person-years). The Kaplan-Meier survival curves showed largely overlapping incidence trajectories between the groups (Fig. 1B). In the fully adjusted model, OSA was not significantly associated with pancreatic cancer incidence (HR, 0.941; 95% CI, 0.823–1.072). Similarly, the incidence rate of biliary tract cancer was 0.113 per 1,000 person-years in the OSA group versus 0.117 per 1,000 person-years in the control group, indicating no significant difference (Fig. 1C). In the fully adjusted model, OSA was not significantly associated with biliary tract cancer incidence (HR, 0.931; 95% CI, 0.751–1.142) (Table 1).
The subgroup analysis, in which subjects were stratified into three age groups (<40 years, 40–64 years, and ≥65 years), did not reveal any significant associations between OSA and cancer incidence in any subgroup (Table 2). For overall cancer, OSA was associated with a higher risk in younger individuals (<40 years, HR, 1.323; 95% CI, 1.200–1.455), a weaker association in middle-aged individuals (40–64 years, HR, 1.137; 95% CI, 1.094–1.181), and no significant association in those aged ≥65 years (HR, 1.040; 95% CI, 0.971–1.114). However, this trend was not observed for pancreatic or biliary tract cancers, with no statistically significant relationships detected across any age group (Table 2).
Subgroup analysis of cancer incidence by age group in OSA patients
When stratified by sex, OSA was associated with a modestly increased overall cancer risk in both males (HR, 1.095; 95% CI, 1.055–1.137) and females (HR, 1.211; 95% CI, 1.142–1.283) (Table 3). However, for pancreatic and biliary tract cancers, no significant differences were observed between males and females, with HRs remaining close to 1, indicating no significant impact of OSA on cancer incidence across sexes.
Subgroup analysis of cancer incidence by sex in OSA patients
DISCUSSION
In this large-scale, nationwide cohort study using the KNHIS database, we investigated the association between OSA and the incidence of biliary tract and pancreatic cancers. Although some previous studies have suggested a potential link between OSA and gastrointestinal malignancies—attributed to its effects on chronic inflammation, oxidative stress, and metabolic dysfunction—other studies have reported no significant association. Our findings do not support OSA as an independent risk factor for biliary tract or pancreatic cancers. Even after adjusting for age, sex, metabolic comorbidities, and other confounders, no significant association between OSA and cancer incidence in these organs was observed. Subgroup analyses stratified by sex and age further confirmed the absence of a statistically significant relationship.
Previous studies investigating the association between OSA and pancreatic cancer have yielded inconsistent results. For instance, Yan et al. [27], using Mendelian randomization, found no evidence of a causal relationship between genetically predicted OSA and pancreatic cancer, suggesting that previously reported associations may be due to shared genetic factors or residual confounding. In contrast, a systematic review and meta-analysis by Wu et al. [29] reported a significant positive association between OSA and cancer risk, including pancreatic malignancy, based primarily on observational cohort studies. In line with Yan et al. [27], our nationwide propensity score-matched retrospective cohort study found no significant association between clinically diagnosed OSA and the subsequent risk of pancreaticobiliary cancer. These discrepancies may be attributable to differences in OSA diagnostic criteria, statistical methods, participant selection, and underlying population characteristics, which could contribute to variations in confounding and study outcomes. Future research involving large-scale populations with standardized OSA diagnostic criteria and diverse genetic backgrounds is warranted to further clarify this relationship.
OSA has also been implicated in various biliary tract and pancreatic diseases, including chronic pancreatitis and NAFPD, where fat accumulation in the pancreas increases the risk of fibrosis and carcinogenesis [24,30]. The underlying mechanisms include OSA-driven hypoxia, metabolic dysfunction, and systemic inflammation, all of which have been implicated in the pathogenesis of biliary tract and pancreatic cancer. Based on these mechanisms, we hypothesized that OSA could elevate the risk of pancreatic and biliary tract cancers [27,30,31]. However, our study did not find a statistically significant association between OSA and pancreatic cancer incidence, even after adjusting for age, sex, and other risk factors. While previous research has reported a more aggressive clinical presentation and poorer prognosis for pancreatic cancers in OSA patients, our findings suggest that OSA does not independently drive carcinogenesis in the pancreas.
One possible explanation for this discrepancy is that OSA may contribute to a worse prognosis in pancreatic cancer rather than to an increased incidence. In this view, OSA-related intermittent hypoxia and inflammation could accelerate tumor growth in individuals with pre-existing pancreatic abnormalities but may not independently cause malignant transformation. This hypothesis aligns with studies suggesting that OSA primarily influences tumor progression rather than incidence [26,32]. Furthermore, although NAFPD is an established risk factor for pancreatic cancer, the lack of a direct association between OSA and pancreatic cancer in our study suggests that other factors, such as genetic predisposition or environmental influences, may modulate carcinogenesis in individuals with OSA. These factors should be elucidated in future studies.
Similar to the findings regarding pancreatic disease, OSA has been associated with various biliary tract disorders, including gallstone disease, cholecystitis, and primary sclerosing cholangitis [27,31]. These conditions are recognized risk factors for cholangiocarcinoma and gallbladder cancer, which led to the hypothesis that OSA-induced chronic inflammation and metabolic dysfunction could contribute to biliary tract carcinogenesis [33,34]. However, our findings did not support a significant association between OSA and biliary tract cancer incidence, even after adjusting for confounding metabolic and inflammatory conditions. One possible explanation is that, although OSA may alter bile composition and promote inflammation in the hepatobiliary system, localized tumor-promoting effects in the biliary tract may require additional carcinogenic stimuli, genetic mutations, or other factors [27,31]. Moreover, gallstone disease and cholecystitis, although common in OSA patients, are not universally linked to cholangiocarcinoma, suggesting that OSA alone may not be sufficient to drive malignant transformation [35,36]. These findings, which are consistent with those of our study, imply that OSA-related metabolic changes may increase the risk of biliary disorders without directly causing oncogenesis in the biliary tract.
Chronic inflammation resulting from OSA can cause persistent local tissue damage, fibrosis, oxidative stress, and dysregulated cellular proliferation, all of which could theoretically increase the risk of carcinogenesis in these organs [37]. Additionally, OSA-induced intermittent hypoxia activates pro-inflammatory pathways, including nuclear factor kappa B (NFκB), tumor necrosis factor-alpha, and interleukin-6, thereby contributing to a pro-tumorigenic environment [11]. Similarly, NAFPD, characterized by fat accumulation in the pancreas, has been linked to chronic inflammation, beta-cell dysfunction, and an increased risk of pancreatic cancer [24]. Given that OSA is associated with lipid dysregulation and ectopic fat deposition in the pancreas, it has been hypothesized that OSA-induced metabolic changes may contribute to pancreatic carcinogenesis [38]. However, the absence of a significant association in our study suggests that while OSA may contribute to systemic metabolic and inflammatory disturbances linked to oncogenesis, it does not act as a direct carcinogen in the pancreas or biliary tract. This finding may be attributed to population-specific factors, inherent differences in genetic susceptibility, or the influence of additional risk modifiers that were not accounted for in our analysis. Further investigation is needed to elucidate the extent to which OSA might function as a co-factor in biliary tract and pancreatic carcinogenesis.
Despite being the first study to examine the relationship between OSA and the incidence of biliary tract and pancreatic cancers in the largest cohort to date—using a nationally representative healthcare registry with propensity score matching and extensive covariate adjustments—several limitations must be acknowledged. First, the retrospective observational design precludes establishing causality, limiting the interpretation of our findings to associations rather than causal relationships. Second, the dataset lacked AHI data, which represents OSA severity, thereby preventing risk stratification based on the degree of intermittent hypoxia and systemic inflammation. Third, OSA diagnoses were identified solely using diagnostic codes from the KNHIS database without confirmation by polysomnography (PSG). Since PSG was not reimbursed in South Korea before 2018, the OSA diagnostic codes from 2007 to 2014 may have relied heavily on physicians’ clinical judgment, potentially introducing selection bias toward patients with more severe symptoms or those financially able to self-fund PSG. Lastly, the lack of histopathological and molecular data on biliary tract and pancreatic cancers precluded analysis of whether OSA preferentially affects specific cancer subtypes, limiting further insights into the potential biological mechanisms linking OSA to carcinogenesis in these organs.
In conclusion, our study does not support OSA as an independent risk factor for biliary tract or pancreatic cancers. Although OSA is associated with chronic inflammation and metabolic dysfunction in the pancreas and biliary tract, its role in pancreaticobiliary oncogenesis remains inconclusive, suggesting that intermittent hypoxia and systemic metabolic alterations may not be sufficient to drive malignant transformation in these organs. Future research should incorporate OSA severity metrics and molecular biomarkers to better determine whether OSA influences carcinogenesis in the biliary tract and pancreas.
Notes
Availability of Data and Material
The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.
Conflicts of Interest
Marn Joon Park and Jae Hoon Cho who are on the editorial board of the Journal of Rhinology were not involved in the editorial evaluation or decision to publish this article. All remaining authors have declared no conflicts of interest.
Author Contributions
Conceptualization: Marn Joon Park, Jae Hoon Cho, Ji Ho Choi. Data curation: all authors. Formal analysis: Marn Joon Park, Seo Jun Kang, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Funding acquisition: Marn Joon Park, Ji Ho Choi. Investigation: Marn Joon Park, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Methodology: Marn Joon Park, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Project administration: Marn Joon Park, Jae Hoon Cho, Ji Ho Choi. Resources: Marn Joon Park, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Software: Marn Joon Park, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Supervision: Marn Joon Park, Jae Hoon Cho, Ji Ho Choi. Validation: Marn Joon Park, Kyung-Do Han, Jae Hoon Cho, Ji Ho Choi. Visualization: Marn Joon Park, Gyu Tae Kim, Seo Jun Kang, Jae Hoon Cho, Ji Ho Choi. Writing—original draft: Marn Joon Park, Ji Ho Choi. Writing—review & editing: Marn Joon Park, Gyu Tae Kim, Seo Jun Kang, Jae Hoon Cho, Ji Ho Choi.
Funding Statement
This work was supported by Inha University Research Grant 2025. This study was supported by the Soonchunhyang University Research Fund.
Acknowledgments
None