521 - Electrogastrography Mean Power Ratio as an Objective Measure of Feeding Intolerance in Preterm Infants

Poster Number: 521
Publication Number: 521.1841

Background: Enteral feeding is challenging in preterm infants because of feeding intolerance (FI). Developmental FI (DFI) results from gastrointestinal (GI) immaturity and dysmotility, whereas pathologic FI (PFI) is associated with ileus due to sepsis, necrotizing enterocolitis (NEC), spontaneous intestinal perforation (SIP), and bowel obstruction. Non-specific clinical signs of DFI are often misinterpreted as PFI, leading to feeding cessation or limitation. We must develop specific, objective biomarkers to differentiate DFI and PFI.

Objective: Compare differences in electrogastrography (EGG) measurements of gastric motility between infants with DFI, PFI, and no feeding intolerance (NFI).

Design/Methods: We conducted a longitudinal cohort study of infants < 34 and ≤37 weeks’ gestation who underwent weekly EGGs. We analyzed pre-, during-, and post-feeding data from weeks 1-2. From raw EGG data, we calculated power spectra at frequencies 0.5-9 cycles per minute (cpm). We calculated mean power spectral density (mPSD) to obtain mean power ratios (mPR) between during- and pre-feeding periods (mPRdur/pre) for each group. Ratios ≤1 indicated dysmotility (no response to feeding). Student’s t-test was utilized to compare mPRdur/pre between groups: DFI vs. NFI, PFI vs. NFI, DFI vs. PFI, and (DFI+PFI) vs. NFI.
Infants were classified into 3 groups based on objective criteria. NFI met 2 criteria of sustained feeding tolerance: 1) achieved full feeds (≥120 mL/Kg/day) sustained for ≥5 days without parenteral nutrition and 2) had a growth velocity of ≥10g/Kg/day from the 1st day of full feeds until 34w6d postmenstrual age. DFI met 0-1 criterium WITHOUT GI pathology (ileus due to sepsis, NEC Stage ≥IIa, SIP, or bowel obstruction). PFI developed GI pathology.

Results: 84 infants were included (75 NFI, 5 DFI, 4 PFI); see Table. At 0.5-9 cpm, mPRdur/pre was lower for DFI vs. NFI (0.7 vs. 1.7, p< 0.001) and PFI vs. NFI (1 vs. 1.7, p< 0.001), but not different for DFI vs. PFI (Fig 1a). After combining DFI+PFI into the FI group, mPRdur/pre was lower for FI than NFI (0.85 vs. 1.7, p< 0.001); Fig 1b. NFI had mPRdur/pre >1 consistent with normal motility. At the low frequency band (1-2cpm), PFI, DFI, and FI had mPRdur/pre ≤1 consistent with dysmotility.

Conclusion(s): EGG may help identify infants with FI during the first 2 weeks of life. Strict criteria to classify infants objectively into different cohorts was essential to detect differences between groups. The inability to differentiate between DFI and PFI may be because of sample size. Future studies may need to focus on lower frequency bands 1-2cpm where there is less overlap.