619 - Associations between neonatal glycemia and neurodevelopment in neonates of mothers with and without dysglycemia in pregnancy

Publication Number: 619.3866

Carmen Monthe-Dreze, Brigham and Women's Hospital, Boston, MA, United States; Sara Cherkerzian, Harvard Medical School, Boston, MA, United States; Ayanna Coburn-Sanderson, Brigham and Women's, Somerville, MA, United States; Isabella Lawandy, Women & Infants Hospital of Rhode Island, Somerville, MA, United States; Laurie Foster, Brigham & Women's Hospital, Boston, MA, United States; Rana Abdel-Rahman, CDC, Atlanta, GA, United States; Camille E. Powe, Massachusetts General Hospital, Boston, MA, United States; Elizabeth Rosenfeld, Children's Hospital of Philadelphia, Philadelphia, PA, United States; Michael SD. Agus, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States; Sarah Lassey, Brigham and Women's Hospital, Boston, MA, United States; Mohamed El-Dib, Harvard Medical School - Brigham and Women's Hospital, Boston, MA, United States; Alfonso Galderisi, Yale School of Medicine, New Haven, CT, United States; Sarbattama Sen, Women & Infants Hospital of Rhode Island, providence, RI, United States

Background: Neonatal hypoglycemia (NH), affecting 50% of newborns with risk factors, is associated with adverse neurodevelopmental outcomes (NDO), even when treated to maintain euglycemic targets. Higher cumulative glucose exposure, glucose variability and maternal metabolism have been hypothesized as contributory. Yet, existing studies are limited by intermittent neonatal glucose monitoring and do not consider the role of the prenatal milieu.

Objective: Determine associations of infant interstitial glucose (IG) patterns with NDO and examine the moderating role of maternal body mass index (BMI) and glycemia in pregnancy.

Design/Methods: Infants (n=34) were enrolled in the ongoing observational prospective longitudinal LAMMBS study in Boston, MA. Blinded continuous IG monitoring was performed from delivery to discharge, and the following exposure variables were derived for the time periods between 4-24 hours and 24-48 hours of life (HOL): (1) mean IG, (2) IG coefficient of variation (CV), and Poincaré analysis measures of (3) short-term (SD1) and (4) long-term (SD2) glucose variability (GV). NDO measures included (1) NeoNatal Neurobehavioral Scale (NNNS) scores (primary outcome) assessed prior to nursery discharge, and Ages and Stages Questionnaire (ASQ) scores at (2) 6-months, and (3) 12-months. Maternal BMI, diabetes status, and glucose loading test (GLT) data were obtained from medical records. We fit confounder-adjusted linear regression models to examine associations of exposures with NDO and included interaction terms to examine modification by maternal BMI, GLT, and diabetes category.

Results: Cohort characteristics are presented in Table 1. Fifty percent of infants had mothers with overweight/obesity (OWOB) or diabetes (IDM) and 25% had mothers with GLT ≥140 mg/dl. Higher mean IG at 4-24 HOL was associated with lower NNNS quality of movement (ẞ -0.03, 95% CI:-0.05, -0.004). Higher mean IG and greater measures of glucose variability (CV, SD1, and SD2) at 24-48 HOL were associated with abnormal neurobehavior and lower developmental scores (Table 2). The associations between mean IG and 6-month ASQ motor scores were only observed among IDM (P-int=0.08) and infants of mothers with GLT ≥140 mg/dl (P-int =0.04, Fig). Maternal OWOB status did not modify these associations.

Conclusion(s): Higher mean glucose and greater glucose variability during the first 48 HOL were associated with adverse NDO, with associations observed only among infants with more severe maternal dysglycemia. Larger cohort studies should examine implications for long-term NDO and the potential impact of optimizing metabolic health in pregnancy.

Table 1: Participant Characteristics
Data presented as mean ± SD for normally distributed variables, median (IQR) for non-normally
distributed variables or n (%) for categorical variables. Sample size defined as infants with non-missing
IG data who had any NDO data available. Abbreviations: SD, standard deviation; IQR, interquartile range;
GLT, 50-gram glucose loading test; SGA, small-for-gestational age; BMI, body mass index; LGA,
large-for-gestational age; NICU, neonatal intensive care unit; IG, neonatal interstitial glucose; HOL,
hour of life; ASQ, Ages and Stages Questionnaire.
a Suspected developmental delay in at least 1 domain based on two cutoffs ( <1SD or <2SD)


Table 2: Associations between neonatal interstitial glucose and infant neurodevelopment
Data are reported as beta estimates from linear regression models adjusted for maternal BMI; gestational age; (+infant sex for NNNS and 6-month ASQ models).
Abbreviations: IG, interstitial glucose; CV, coefficient of variation; HOL, hour of life; NNNS, NeoNatal Neurobehavioral Scale; ASQ, Ages and Stages questionnaire.
a SD2: The standard deviation along the major axis of the ellipse in Poincaré analysis refers to long-term variability, or fluctuations occurring over a longer period of time
b SD1: The standard deviation along the minor axis of the ellipse in Poincaré analysis refers to short-term variability, or fluctuations occurring within a short period of time
c Higher scores = less favorable NNNS measures of handling, arousal/reactivity and tone; Lower scores = less favorable for attention, self-regulation and quality of movement.
*P < 0.05; **P < 0.01


Figure: Moderating role of maternal BMI and glycemia on associations between neonatal interstitial glucose and neurodevelopment.
Results are displayed as estimated means with 95%CIs from linear regression analyses adjusted for gestational age and sex, including a maternal metabolic status*IG interaction term. Abbreviations: BMI, body mass index; IDM, infants of mothers with diabetes; GLT, glucose loading test. Figures A, C, E: Effect modification by maternal BMI, IDM and GLT category on 6-month total ASQ scores; Figures B, D, F: Effect modification by maternal BMI, IDM and GLT category on 6-month ASQ motor scores. P-value for interaction term provided for each model. P-value for interaction was considered significant at P < 0.2.