# A novel corrected anion gap links metabolic acid attack to peptic ulcer disease

## TL;DR

A new corrected anion gap is linked to peptic ulcer disease, suggesting metabolic acid load may contribute to its development.

## Contribution

The study introduces a novel corrected anion gap as a potential biomarker for peptic ulcer disease risk.

## Key findings

- AGcorr was significantly associated with increased PUD risk in adjusted models.
- HCO₃− showed a negative correlation with PUD but with lower statistical power.
- AGcorr outperformed other acid-base metrics in predicting PUD risk.

## Abstract

Physiological evidence indicates that systemic acid-base balance may be associated with gastric acid secretion, which is a key factor in peptic ulcers disease (PUD). However, the epidemiological relationship of this association with PUD risk remains unclear.

This study investigates a novel corrected anion gap (AGcorr, reflecting metabolic acid load) and bicarbonate (HCO₃−, reflecting alkaline reserve) in relation to PUD.

This retrospective case-control study analyzed endoscopy-confirmed 64 PUD patients and 541 non-PUD controls with normal hepatic and renal function from a Fujian tertiary hospital (Jan 2023–Jun 2025). The anion gap (AG) was corrected to fully-corrected AGcorr and albumin-corrected AG (ACAG). Four models (A: AGcorr; B: HCO₃−; C: AG; D: ACAG) were adjusted for sex, age, and Na to study the relationships between AGcorr, ACAG, AG, HCO3, and PUD. Post-hoc power and sensitivity analysis were conducted.

Model A showed that AGcorr was significantly positively associated with PUD risk (OR = 1.167, 95% CI: 1.050–1.298, P = 0.045), with a high statistical power of 89%. Sensitivity analysis with an expanded control group (n = 1915) confirmed a significant AGcorr-PUD association (OR = 1.138, 95% CI: 1.029–1.259, P = 0.012), with post hoc power of 0.81. The sensitivity analysis with case-control matching for sex and age showed that AGcorr remained significantly associated with PUD (P = 0.020). In contrast, Model C and Model D were both significantly correlated (P < 0.05) but showed poor fit (H-L P < 0.05). However, their statistical power differed, with values of only 0.64 and 0.84, respectively. Model B indicated a negative correlation between PUD and HCO₃− (OR = 0.894, 95% CI: 0.802–0.996, P = 0.043), but with low statistical power (65%) and instability in sensitivity analysis.

After adjusting for sex, age, and Na, AGcorr remained an independent risk factor for PUD in individuals with normal liver and kidney function. The fully-corrected AG outperformed both ACAG and uncorrected AG. Metabolic non-volatile acid attack may be an important pathogenic mechanism in PUD. HCO₃− may have a potential protective effect against PUD. The findings contribute to improving PUD care but require further validation in diverse groups.