Obstructive sleep apnea (OSA) is the most prevalent form of sleep-disordered breathing with a global prevalence rate of 9–38% (1, 2). OSA can lead to intermittent hypoxemia, sleep fragmentation, sympathetic hyperactivity, variations in intrathoracic pressure, and disturbances in physiological balance (3). These symptoms greatly decrease one’s quality of life and increase morbidity of other diseases (4, 5).

The quality of sleep and sleep-related breathing disorders constitute significant factors contributing to the development of Gastrointestinal diseases (GDs) (6). Observational cohort studies have consistently indicated a higher prevalence of GDs among individuals with OSA compared to those without, encompassing conditions such as gallstones (7), non-alcoholic fatty liver disease (NAFLD) (8), inflammatory bowel disease (IBD) (9) and gastroesophageal reflux disease (GERD) (10–12). A retrospective study involving 586,377 adults in the USA revealed an independent association between OSA and GERD, accounting for multiple potential confounding variables (13). Continuous positive airway pressure (CPAP) therapy has demonstrated efficacy in alleviating gastroesophageal reflux symptoms and stabilizing NAFLD progression in patients with OSA (14, 15). However, these findings have encountered scrutiny in recent research (16–18). At present, there is limited evidence of associations between OSA and other common gastrointestinal disorders. In summary, the causal link between OSA and GDs remains contentious, with debates centering on potential confounders and reverse causation biases. A comprehensive evaluation of this causal relationship is crucial to unraveling the mechanisms underlying disease co-occurrence and identifying therapeutic strategies that benefit both conditions.

Mendelian randomization (MR) provides a more precise evaluation of causality for exposure–outcome associations by leveraging genetic variations as instrumental variables (IVs) (19–21). By generating genotypes through randomly assigning parental alleles during conception, MR minimizes the possibility of potential residual confounders and reverse causalities that could be influenced by environmental and self-imposed factors (22). MR has been widely used to predict causal pathways in studies examining the effects of OSA on atrial fibrillation (23), interleukin levels (24), and neurodegenerative disease (25). However, the effects of OSA on a broad range of gastrointestinal outcomes have not been investigated.

We performed a two-sample MR analysis and multivariate Mendelian randomization (MVMR) to evaluate whether OSA and common GDs are causally related. In addition, we used the same MR analysis methods in the reverse direction to determine whether GDs increased susceptibility to OSA.

For this investigation, the globally significant threshold for IVs screening was set at p<1×10^-5, r^{2 =} 0.001, and a clumping window >10,000 kb. Based on these criteria, 747 SNPs were selected as IVs to evaluate causal relationships linking OSA to all 19 GDs, using OSA as the exposure factor. Unfortunately, when GDs were utilized as exposure factors during reverse MR analysis, the IVs that could be extracted from the genome-wide association study (GWAS) summary data on calculus of the bile duct and acute pancreatitis were insufficient. Consequently, no causal association between these two conditions and OSA was identified. The included IVs are detailed in Supplementary Table S1 . Supplementary Table S2 presents the IVs featured in the MR analysis that identified a potential causal link between OSA and GDs.

This MR analysis is the first to systematically evaluate the cause-and-effect relationship between OSA and digestive disorders. Two-sample MR analysis found a strong correlation between OSA and the incidence of gastroduodenal ulcer and GERD. A nominal causal association was found between OSA and increased risk of calculus of bile duct, calculus of gallbladder, chronic gastritis, and ulcerative colitis. The MVMR analysis found a reciprocal causal link between OSA and GERD. After adjusting for four risk factors (obesity, T2D, smoking, and alcohol consumption) in MVMR analysis, OSA increases the risk of gastroduodenal ulcer and calculus of the bile duct independently.

Numerous epidemiological studies have provided compelling evidence linking several gastrointestinal disorders to sleep disorders such as OSA (6, 10, 44). OSA can induce intermittent hypoxemia, sleep fragmentation, oxidative stress, systemic inflammatory response, and disrupt intestinal flora, potentially contributing to the onset of gastrointestinal diseases. Conversely, nocturnal awakenings, sleep deprivation and sleep fragmentation due to gastrointestinal diseases may reciprocally influence the occurrence of OSA. However, the causal relationship remains unclear due to limitations in temporal sequencing in existing observational studies. To address this gap, we conducted bidirectional MR and MVMR analyses to robustly evaluate the causal connections between OSA and common GDs, providing robust evidence of their causal associations. This findings underscores the importance of early screening for these disorders in OSA-diagnosed patients.

Global population–based studies observed an 8–33% incidence rate of GERD (11). However, individuals with OSA have much greater rates, ranging from 40% to 60% (45, 46). A retrospective study of 586,377 adults in the USA discovered an independent link between OSA and GERD while considering multiple potential confounders (13). However, another multicenter, cross-sectional, observational study conducted in Turkey (n=1104) found that the severity of OSA did not independently predict the occurrence of GERD (17). Recently, an MR study discovered an implied bidirectional relationship between OSA and GERD (47); however, it was unclear whether the results were affected by shared risk factors, particularly obesity. Our findings, which were based on MR techniques that accounted for obesity, smoking, alcohol use, and T2D, indicated a bidirectional, independent causal link between GERD and OSA. This finding can benefit future research on the comorbidity of the two diseases and the optimization of a joint treatment plan.

The supine position itself increases the occurrence of nocturnal GER (48). GER events can be caused by declining intrathoracic negative pressure and elevating abdominal pressure during apneic episodes in OSA patients. This pressure fluctuation also relaxes and opens the lower esophageal sphincter, worsening GERD symptoms (10, 49). Common symptoms of GERD include heartburn or regurgitation, non-cardiac chest pain, chronic cough, and sore throat (50). Nocturnal uncomfortable symptoms can lead to heightened sleep arousal and diminished sleep duration by influencing the interaction of various neural pathways (51). The gastric content of reflux can irritate edema and cause spasms in the upper airway and larynx, thereby contributing to the collapse of the upper airway (10). Additionally, reflux can stimulate the vagus nerve, cause bronchial constriction (52), and affect normal systolic function of the respiratory muscles (53), potentially leading to sleep apnea. The complex pathophysiology underlying their causal relationship warrants further investigation.

The present investigation found previously unrecognized causal links between a genetic predisposition to OSA and a greater incidence of gastroduodenal ulcer and chronic gastritis. An earlier study conducted by Wu et al. found that OSA could affect duodenal morphology via increased oxidative stress and activation of transcription factors, which can lead to the disruption of intestinal tight junctions, culminating in intestinal damage (54). This disruption may lead to the development of gastroduodenal ulcer. Furthermore, multiple cross-sectional studies have shown that shorter sleep durations and poor sleep quality increase the likelihood of gastroduodenal ulcer recurrence (55, 56). The MVMR analysis showed that OSA’s effect on chronic gastritis was diminished after accounting for obesity, possibly indicating that OSA and obesity synergistically contribute to chronic gastritis.

Inflammatory bowel diseases (IBD), including Crohn’s disease and ulcerative colitis, represent chronic inflammatory condition of the gastrointestinal tract. The pathogenesis of intestinal lesions in IBD remains elusive. The bidirectional interactions between sleep disorders and IBD have recently received considerable attention (57, 58). OSA causes intermittent hypoxia, which results in periodic reductions in the oxygen partial pressure gradient within the lumen, promoting intestinal dysbiosis and inflammation (59). Disruption of the circadian rhythm can impair normal intestinal immune system activity, leading to an increased release of inflammatory factors and exacerbating intestinal inflammation (60–62). A retrospective study using US-wide diagnostic coding data independently linked IBD to a higher prevalence of OSA (63), consistent with the findings of this MR analysis. The potential causal relationship between genetically predicted OSA and ulcerative colitis identified in this study may provide new perspectives to explore the pathogenesis and treatment of IBD.

Moreover, we observed a positive link between OSA and gallstones, including calculi in the gallbladder and bile duct. This outcome is consistent with the finding from a cross-sectional study conducted by Chen et al. on 3,827 patients (7). OSA may cause gallstones by impeding gallbladder contraction through stimulating sympathetic nerves and enhancing bile supersaturation through inflammation associated with chronic intermittent hypoxia (64, 65). A previous Mendelian analysis showed that 1−standard deviation increase in the body mass index resulted 1.631 times increased odds of developing gallstones (66). However, the current MVMR analysis found that the causal relationship between OSA and gallbladder calculi was no longer significant after considering obesity; the causal relationship with bile duct calculi remained. This result suggests that OSA may affect gallbladder calculi, but not an independent risk factor.

Recent research has revealed a connection between OSA and the initiation and progression of NAFLD (67), indicating that patients with OSA face poor prognoses and are susceptible to other liver conditions (68, 69). However, MR results did not reveal a significant causal association between genetic predisposition to OSA and the development of NAFLD or cirrhosis. In addition, separate epidemiological studies by Nieto et al. (70) and Marshall et al. (71) reported significantly higher cancer mortality rates among patients with sleep apnea. However, Kendzerska et al. (72) concluded that OSA severity does not increase cancer incidence. Our study found no correlations between genetically predicted OSA and gastrointestinal cancers, including esophageal, gastric, or colorectal cancers. These findings suggest that the observed epidemiological relationships may stem from shared genetic components or unmeasured confounding factors.

This study had several advantages. First, this study is the first to comprehensively assess the cause-and-effect relationship underlying OSA with major GDs. We utilized two-sample MR based on SNPs from large sample populations (33,423 cases and 307,648 controls for OSA; an average of 9,530 cases and 299,113 controls for each GD); the sample sizes for exposure and outcomes did not overlap. Second, multiple MR statistical methods and multivariate MR analyses were applied to minimize confounding bias and guarantee the reliability of the findings, including adjustments for obesity, T2D, smoking, and alcohol consumption.

This study had limitations. First, the summary-level data from the GWASs included trans-ancestry populations among the source populations, as shown in Table 1 . This inclusion could bias results because of the possibility of population mixture confounding. Second, MVMR analysis was performed for common risk factors for OSA and GDs, but other potential confounders may have a potential impact on the independence of causal association. Third, completely excluding the possibility that the SNPs linked to GERD affect OSA via additional causative pathways (i.e., horizontal pleiotropy) was challenging.

This MR analysis provides compelling evidence of an independent causal relationship between genetically predicted OSA and an elevated risk of multiple inflammation-related GDs. In summary, this study provides strong evidence that OSA and GERD are independent risk factors for each other. In addition, a novel independent causal association was found between genetically predicted OSA and gastroduodenal ulcer. The increased risk of chronic gastritis, ulcerative colitis, and gallstones (including calculus in the gallbladder and bile duct) may also be correlated directly with OSA. Understanding the biological pathways and mechanisms underlying these correlations will aid physicians and researchers in developing preventive and therapeutic strategies for patients with OSA presenting with gastrointestinal symptoms.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

WY: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing – original draft. JZ: Data curation, Investigation, Validation, Writing – original draft. MJ: Conceptualization, Supervision, Validation, Writing – review & editing. YK: Data curation, Formal analysis, Writing – original draft. HQ: Data curation, Visualization, Writing – original draft. YQ: Data curation, Formal analysis, Writing – original draft. SW: Conceptualization, Investigation, Supervision, Validation, Writing – review & editing. JT: Conceptualization, Funding acquisition, Writing – review & editing.