Related Experiment Video
Updated: Oct 28, 2026

A Neonatal BALB/c Mouse Model of Necrotizing Enterocolitis
Published on: November 30, 2021
Association Between Methemoglobin Levels and Necrotizing Enterocolitis in Preterm Infants: A Retrospective Cohort
Muhammet Asena1, Duygu Tuncel2, Leyla Şero2
1Department of Pediatrics, Istanbul Medipol University, Medipol Mega University Hospital, Istanbul 34214, Turkey.
Abstract:
Background: Necrotizing enterocolitis (NEC) is a leading cause of morbidity and mortality in preterm infants, and there remains a need for simple, easily measured parameters reflecting its prevalence and severity. This study tested the hypothesis that methemoglobin (MetHb), a molecule implicated in nitric oxide metabolism, is associated with the presence and severity of NEC, as well as with its complications. Methods: We performed a single-center, retrospective cohort study on 291 premature infants (<32 weeks' gestational age). A total of 217 had no NEC and 74 had Bell stages II-III NEC. MetHb expressed as a percent of the total hemoglobin (% MetHb) was studied for its relationship with NEC and outcomes. Logistic regression and receiver operating characteristic analysis were performed. Results: % MetHb was higher among neonates who developed NEC compared to controls [2.55% (1.70-4.95) vs. 2.20% (1.70-2.80); p = 0.040]. For every twofold increase in % MetHb, there was an association with the development of NEC after adjustment for sex and gestational age (aOR: 1.74; 95% CI: 1.24-2.45; p = 0.001). Discrimination of NEC was poor, with an area under the curve of 0.580 (95% CI: 0.493-0.667); at a cutoff of ≥4.6% (exploration purposes), the test had a sensitivity of 28.4% and a specificity of 94.5%. The test failed to differentiate stage II from stage III NEC, and its association with all-cause death and death with an NEC association, after adjustment, failed to reach significance. Conclusions: Elevated MetHb levels were independently associated with NEC, although they lacked adequate standalone predictive and discriminatory ability; these data do not support current utilization for prediction or diagnosis, and further prospective study with standardized sampling, external validation, and a multimarker approach is needed.

