Clinical and laboratory correlates of low fecal elastase-1 levels: A retrospective analysis
Neziha Hacihasanoglu Cakmak1, Gozde Ulfer2, Pakize Yigit3, Turkan Yigitbasi4, Gulbanu Canbaloglu Erkan51Medical Laboratory Techniques, Istanbul Medipol University Vocational School of Health Services, Istanbul, Türkiye2Department of Medical Biochemistry, Istanbul Medipol University Faculty of Medicine, Istanbul, Türkiye
3Department of Management Information Systems, Istanbul Medipol University Faculty of Business and Management Sciences, Istanbul, Türkiye
4Genetic Diseases Evaluation Center, Istanbul Medipol University Faculty of Medicine, Istanbul, Türkiye
5Department of Gastroenterology, Istanbul Medipol University Faculty of Medicine, Istanbul, Türkiye
INTRODUCTION: Fecal elastase-1 (FE-1) is widely used as a noninvasive, stool-based test to evaluate pancreatic exocrine function. This study compared radiological and laboratory findings between patients with low and normal FE-1 levels and investigated variables independently associated with low FE-1 levels.
METHODS: The records of 1,133 patients who underwent FE-1 testing were retrospectively reviewed. Patients were clas-sified into normal (≥200 µg/g, n=961) and low (<200 µg/g, n=172) FE-1 groups. Group comparisons and multivariable logistic regression analyses were performed. To assess the incremental discriminative contribution of pancreatic am-ylase while avoiding confounding due to differences in sample size, four additional models were evaluated using the same complete-case subset (n=396). Areas under the receiver operating characteristic curves (AUCs) were compared using paired DeLong tests in jamovi version 2.6.2.0.
RESULTS: Low FE-1 levels were more frequent among patients with pathological radiological findings than among those with normal findings (31.7% vs. 12.1%; χ²=45.2, p<0.001; OR=3.38, 95% CI: 2.33–4.89). Among the same 396 complete cases, Model A (age, sex, and radiology) yielded an AUC of 0.694 (95% CI: 0.619–0.769). Adding pancreatic amylase (Model B) increased the AUC to 0.727 (95% CI: 0.657–0.798), although this increase was not statistically significant (ΔAUC=0.0333; DeLong p=0.100). Model C (age, sex, radiology, ALT, glucose, BUN, IgA, and lipase) yielded an AUC of 0.740 (95% CI: 0.672–0.807). Adding pancreatic amylase (Model D) increased the AUC to 0.766 (95% CI: 0.705–0.827), but the increase was not statistically significant (ΔAUC=0.0264; DeLong p=0.107).
DISCUSSION AND CONCLUSION: Low FE-1 levels were associated with radiological, metabolic, nutritional, and laboratory abnormalities. Pancreatic amylase was independently associated with low FE-1 levels in the expanded regression model; however, its incremental contribution to AUC was not statistically significant when the models were compared using data from the same patients. These findings reflect factors associated with low FE-1 levels and should not be interpreted as establish-ing a diagnostic model for exocrine pancreatic insufficiency (EPI).
Keywords: Biomarkers, exocrine pancreatic insufficiency, feces, logistic models, pancreatic elastase
Manuscript Language: English