Serum anion gap (AG) holds potential for clinical application in the prognostic assessment of ill patients. Yet its association with all-cause mortality in critically ill patients with gastrointestinal bleeding (GIB) remains underexplored. This cohort study analyzed data from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database, 2,333 critically ill GIB patients, recording their first-time AG levels. The relationship between AG, ICU mortality, and 90-day mortality in GIB patients was analyzed using the Kaplan-Meier (KM) survival curve, and COX regression model. Restricted cubic splines and subgroup analysis were conducted to enhance the stability and reliability of the findings. A total of 2,333 patients diagnosed with GIB were included in the study. The overall ICU and 90-day mortality in the entire cohort were 9.17% and 27.73%, respectively. KM analysis revealed a significant relationship between high AG and the risk of all-cause mortality (log-rank P < 0.001). After adjusting for all confounders, the association between AG and ICU and 90-day all-cause mortality was significant (HR = 1.07, 95% CI: 1.03–1.12; HR = 1.06, 95% CI: 1.04–1.09, separately). Furthermore, we assessed the potential linear relationship between AG and all-cause mortality (P for non-linearity > 0.05). Subsequent subgroup analysis demonstrated that weight and heart failure could impact the association between the two. AG emerges as a significant biomarker, closely associated with heightened ICU and 90-day mortality in individuals suffering from GIB.
Background Helicobacter pylori (H. pylori) infection is a major pathogen associated with gastric cancer. However, the causal relationship between H. pylori infection and extra-gastric tumors has not been fully elucidated. Methods A two-sample Mendelian randomization (MR) study was conducted using genome-wide association study (GWAS) data on serum H. pylori IgG levels as the exposure variable and GWAS data on 268 benign and malignant tumors as the outcome variable. Single nucleotide polymorphisms (SNPs) associated with the exposure were identified from the OpenGWAS and GWAS Catalog databases, while those linked to the outcome were sourced from the FinnGen database. Independent MR analyses were performed on the discovery and validation cohorts. Sensitivity analyses, including heterogeneity and pleiotropy assessments, were conducted to ensure the robustness of the results. Meta-analysis was used to integrate significant findings from both cohorts. Results Our findings demonstrated negative causal effects of H. pylori infection on benign neoplasms of cranial nerves and extrahepatic bile ducts in both discovery and validation cohorts. Moreover, we found a potentially positive causal link between H. pylori infection and certain malignancies of the hypopharynx, kidney, lip, brain, pancreas, and neuroendocrine system, as well as malignant immunoproliferative diseases and leukemia, although these results were only observed in a single cohort. Sensitivity analyses showed no significant heterogeneity or horizontal pleiotropy in the significant MR directions, supporting the reliability of the results. Meta-analysis further confirmed the protective effect of H. pylori infection on benign cranial nerve and extrahepatic bile duct tumors. Conclusion This study revealed that H. pylori infection is causally associated with extra-gastric tumors, suggesting a potential protective role of H. pylori infection in certain benign tumors.