In this national database study, we observed a low prevalence of PUD in the adult CF population at 1.19%, as compared to previous studies. To our knowledge, this study represents the first large-sample study to examine the prevalence of PUD in adult PwCF. Historically, data on the prevalence of PUD in people with CF have been limited by small sample sizes, making comparisons to our data challenging. In 1962, an evaluation of 115 patients with cystic fibrosis found a prevalence of 22% for peptic ulcers while an autopsy review in 1975 of 146 patients found a prevalence of 8% for peptic ulcers (16, 17). The PUD rate in our CF population was lower than that of the general population (5–10%), this can be attributable to the “CF paradox”, which remains a subject of controversy in the CF population (1). The prevalence of PUD in PwCF has significantly declined. This decline in CF patients parallels that of the general population and can be attributed to the widespread use of PPI for acid suppression and advancements in H. pylori eradication (18, 19).
Prior studies have shown that Hispanics have a higher prevalence of PUD associated with H. pylori (18.7%) compared to the white population (14%) (20). Despite CF being more prevalent in white populations, there has been an increase in people with CF identifying as Hispanic since 2006, with 9.8% of patients identifying as Hispanic in 2001, which is consistent with our results (8.8%) (21). Multiple disparities affect the Hispanic CF population leading to increased mortality compared to their non-Hispanic counterparts even after adjusting for socioeconomic status and clinical severity (22, 23). Hispanic populations also exhibit decreased pulmonary function, an important predictor of life expectancy in CF (24). McGarry et. al list factors such as lack of representation in clinical trials, implicit bias during care and late diagnosis due to low CFTR mutation detection rates in newborn screening panels as contributors to these disparities (22). The increased rates of PUD in Hispanic CF population could be multifactorial, given the extensive literature on disparities in CF and PUD.
Non-alcoholic steatotohepatitis (NASH, recently renamed metabolic dysfunction-associated steatohepatitis (MASH)) was associated with increased PUD in this population. NASH is a subtype of non-alcoholic fatty liver disease (NAFLD, recently renamed metabolic dysfunction-associated liver disease (MASLD)). Hepatic steatosis has a prevalence of 20–60% and is the most common hepatic manifestation observed in people with CF (25, 26). The mechanism of hepatic steatosis in CF is not well characterized and is appears to be independent of the CFTR mutation (27). A retrospective study of 114 CF patients revealed that 14.9% had steatohepatitis with significant association with overweight status and no association with CF related liver disease (26). PUD is also frequently seen among those with liver cirrhosis; however, there is currently a lack of data on PUD prevalence in MASLD associated hospitalizations. A recent study by Dahiya et al. also observed a rise in inpatient mortality in MASLD hospitalizations with concomitant PUD (13). The noted rise in inpatient mortality in MASLD with PUD, coupled with our findings of increased prevalence of steatohepatitis in PUD, raises concerns about the significance of the relationship between these two diseases. This understanding is crucial for identifying individuals at the highest risk of adverse outcomes and complications, in the prevention of increased morbidity and mortality.
Growing evidence also supports an association between MASLD and H. pylori, one of the most important environmental factors for PUD (28). Possible underlying mechanisms by which H. pylori contributes to MASLD encompass various pathways including insulin resistance (IR), inflammation cytokines and gut dysbiosis (28). First, it is well known that IR is pivotal in MASLD development, and emerging research indicates that H. pylori infection may function as a causative factor for IR through inducing chronic inflammation and activating certain signaling pathway (29). Second, H. pylori infection may trigger chronic low-grade systemic inflammation, leading to elevated levels of inflammatory cytokines such as IL-6 and TNF-α (30). This, in turn, activates the IKK/NF-KB pathway, inducing insulin resistance. Additionally, H. pylori infection may impede the release of leptin from adipose tissue, enhancing the activity of liver stearoyl-CoA desaturase and expediting the production of very-low-density lipoprotein cholesterol (VLDL-C) and fatty deposits in liver tissue (28). Third, the interaction between the stomach and intestines, influenced by H. pylori infection, may contribute to gastrointestinal dysbiosis. Gut dysbiosis has been associated with the development of MASLD (31, 32).
There was a steady increase in the number of admissions associated with CF from 2014 to 2019 (21). According to data from the CF registry (21), the patient count in 2021 stood at 32,100, a notable increase compared to the approximately 15,000 recorded in 1986. This is in line with an earlier investigation conducted by Agrawal et al., examining the NIS database spanning from 2003 to 2013 and revealing a rise in the number of inpatient discharges associated with CF during that period. The majority of these admissions were due to pulmonary examinations (33). These findings are primarily from advancements in therapies like CFTR modulators and improved mucociliary clearance, but also to the adoption of the care center model endorsed by the Cystic Fibrosis Foundation (CFF). These interventions have resulted in fewer pulmonary complications, improved FEV1, and enhanced quality and quantity of life in the CF population. The median predicted survival reached 53.1 in 2021, a significant increase from the mere 16 years observed in 1970 (21). As the CF population ages, managing their gastrointestinal manifestations becomes crucial to continued expansion of life expectancy.
Another pertinent finding is that, concerning payer status, a majority of PwCF were covered by private health insurance compared to other insurance types. This aligns with previous findings from the CFF (21). Medicaid-covered patients might be less inclined to seek care at the onset due to concerns about limited healthcare access (34). Socioeconomic disparities among insured people with CF can impact care, with Medicaid-covered patients having a 3.7-fold higher risk of death compared to those without Medicaid (35). The low number of uninsured patients is promising because CF patients lacking health insurance have been shown to have higher mortality rates.
Our study has several limitations primarily associated with the characteristics of the NIS database and the study’s design. First, our results likely underestimate the true numbers of people with CF hospitalized since the NIS database was designed as an administrative dataset reflecting the coding practices unique to each institution. Therefore, a discharge coded with an alternative diagnosis might not have been reflected in the dataset. The dataset is also devoid of controls for data entry errors. Second, the NIS dataset lacks details about patients, essential to assess for symptoms and signs of PUD, metabolic risk factors that are usually associated with MASLD, and other laboratory and clinical signs of fatty liver disease. As such, this precludes the study of temporal relationship between conditions. Although our study controlled for the use of PPIs, NSAIDS and H2 blockers, one limitation was the inability to identify if the patients diagnosed with PUD were also positive for H. Pylori, another risk factor for peptic ulcer disease. While coinfection is a possibility, current literature suggests that the rates of H. Pylori in cystic fibrosis patients remain the same as the general population making it less likely to be a confounding factor in our study (1).