Measures of neonatal glycemia from blood glucose concentrations and neurodevelopmental outcomes at 2 years

Neonatal hypoglycemia is common and associated with adverse neurodevelopmental outcomes, but it is unclear which glycemic measures best explain this association.

Blood glucose concentrations (BGC) from three prospective cohorts of late preterm and term babies born at risk of hypoglycemia, all screened and treated to maintain BGC ≥ 2.6 mmol/L, were used to derive both continuous (maximum, minimum, range, mean, median, standard deviation and variability metrics) and dichotomous glycemic measures (any, severe, recurrent and late hypoglycemic episodes). Generalized linear models adjusted for socioeconomic deprivation, site, primary risk of hypoglycemia and multiple birth related glycemic measures to outcomes at 2 years.

We analyzed 18,947 BGCs from 2,894 babies. Higher gradient variability was associated with poorer neurodevelopment, but dichotomous indicators were more strongly predictive; e.g. 1 SD increase in gradient variability was associated with -0.9-point adjusted mean Bayley motor score (95% confidence interval -1.4, -0.3), compared to -3.1 (-4.8, -1.5) for severe and -2.1 (-3.6, -0.6) for recurrent hypoglycemia. Associations were similar for Bayley cognitive and language and Brief P global executive composite scores.

Continuous glycemic measures derived from intermittent capillary sampling are not stronger predictors of neurodevelopment at 2 years than dichotomous measures, although glycemic variability warrants further investigation.

Beyond simple dichotomous classifications (e.g., hypoglycemic or not), little is known about how blood glucose profiles derived from intermittent heel-prick measurements relate to later neurodevelopmental outcomes in infants at risk of neonatal hypoglycemia. Prespecified glycemic measures were related to neurodevelopmental outcomes at 2 years using data from three longitudinal studies of children born at risk of neonatal hypoglycemia. While the dichotomous indicators of severe (<2.0 mmol/L) and recurrent (≥3 episodes) hypoglycemia are most strongly related to neurodevelopmental outcome, measures of glycemic variability also relate to neurodevelopmental risk in infants at risk of neonatal hypoglycemia.