503 - Human Milk Insulin is Associated with Markers of Infant Pancreatic Function in the First 5 Months of Life

Poster Number: 503
Publication Number: 503.2115

Background: Human milk (HM) contains elevated concentrations of insulin relevant to maternal circulation. While milk insulin is associated with newborn glycemic outcomes in animal models, the relationship in human infants remains unclear.

Objective: Assess relationship between HM insulin and infant glycemic outcomes.

Design/Methods: Healthy, term, exclusively breastfeeding dyads were recruited into 2 groups: Control = maternal pre-pregnant BMI < 24 kg/m2 and normoglycemia; Test = either maternal pre-pregnant BMI >25 kg/m2 and failing the 28-week oral glucose tolerance test, or pre-pregnant BMI >30 kg/m2. At 2-weeks, mothers collected a full-breast expression after an overnight fast (Control group) or 2-hours after consuming a 50g glucose load (Test group).

At 2-weeks, infants were fed 2.5 oz of mother’s milk and circulating glucose measured at fasting (>2 hours), 30-, and 60-minutes post-feed. All urine produced up to 4 hours post-feed was assayed for urinary C-peptide and creatinine. At 5 months, before solid foods had been introduced, fasting HM was collected. Infants were administered a modified glucose challenge test (1.75 g/kg weight) and circulating glucose and insulin measured at fasting (>2 hours), 30-, and 60-minutes post-dose.

Results: Birth weight, gestational ages and sex were similar between groups.

At 2-weeks, fasting HM insulin was higher in the Test group (11.5± 6.7 vs 7.1 ± 4.6 μU/mL, p=0.03), but infant fasting glucose did not differ between groups. After consuming 2.5 oz of mother’s milk, HM insulin dose was inversely related to a marker of pancreatic insulin production (the change in urinary C-peptide-to-creatinine ratio, p=0.029, R2=0.13, n=38).

At 5-months, fasting HM insulin remained higher in the Test group (16.4± 8.5 vs 7.3 ± 4.3 μU/mL, p< 0.001), and infant fasting glucose was significantly higher in the control group (94.3 ± 8.9 vs 88.7 ± 6.0 mg/dL, p< 0.03).

Prospectively, fasting HM insulin at 2-weeks was inversely associated with infant fasting glucose at 5-months (p=0.077, R2=0.097, n=33). Cross-sectionally, fasting HM insulin at 5-months was inversely associated with infant fasting glucose at 5-months (p=0.087, R2=0.086, n=35). After an oral glucose challenge, infant blood glucose response did not differ between groups, but infants in the Test group exhibited lower insulin concentrations at 60-minutes post glucose dose (4.7 ± 3.5 vs 8.8 ± 5.8, p=0.03).

Conclusion(s): HM insulin may contribute to infant blood sugar regulation, especially in the neonatal period. Together, these results suggest that HM insulin may contribute to infant pancreatic programming and necessitate replication.