795 - Umbilical cord inflammation and birth size in the population-based Projahnmo pregnancy cohort in rural Bangladesh

Publication Number: 795.6723

Sara Cherkerzian, Harvard Medical School, Boston, MA, United States; Ingrid Olson, Brigham and Women's Hospital, Boston, MA, United States; Salahuddin Ahmed, Projahnmo Research Foundation, Dhaka, Dhaka, Bangladesh; Rasheda Khanam, Johns Hopkins University, Baltimore, MD, United States; Anne CC. Lee, The Warren Alpert Medical School of Brown University, Wayland, MA, United States

Background: Maternal malnutrition and infection have been linked to maternal-fetal inflammation and may contribute to high rates of fetal growth restriction observed in low- and middle-income countries.

Objective: To examine the association between cord blood inflammation and birth size in a population-based, pregnancy cohort from rural Bangladesh.

Design/Methods: . In Sylhet district, Bangladesh, we consecutively enrolled 299 pregnancies and collected prenatal data and umbilical cord blood spots at birth (07/2016-03/2017). Cord blood was analyzed for interleukin (IL)-1α, IL-1β, IL-6, IL-8, and C-reactive protein. Z scores for weight, length, weight-to-length ratio, and head circumference at birth were calculated using the Intergrowth-21 standard. Small for gestational age (SGA) was defined as < 10% birth size for gestational age(GA)/sex. The association between elevated biomarkers (>67% vs <=67%) and birth size was assessed using logistic regression adjusted for potential confounding (add specific covariables). Analyses were performed in the total cohort as well as stratified by GA categories, early term/preterm ( < 39 weeks) versus full/late term (>39 weeks) GA.

Results: The average maternal age was 23.8 (+/- 4.6) years, and there was a high prevalence of maternal malnutrition (35% with mid-upper arm circumference < 22 cm, 19% with height < 145 cm). The majority (n=247, 89%) of newborns were born full term (>=37 weeks), and 41.3% were SGA. Among infants < 39 weeks GA, elevation of IL-1β and IL-8 were associated with greater odds of SGA (IL-1β. OR:2.55, 95%CI:0.99,6.59; IL-8. OR:4.47, 95%CI:1.67,11.95) and weight-to-length ratio (IL-8 only. OR:3.71, 95%CI: 1.36,10.16). However, among infants >= 39 weeks gestation, elevated IL-1β was associated with lower odds of SGA (OR:0.47, 95%CI:0.23,0.96) and weight-to-length ratio (OR:0.44, 95%CI:0.21,0.95).

Conclusion(s): Among infants in rural Bangladesh, elevation in pro-inflammatory cord biomarkers IL-1β and IL-8 was associated with SGA and weight-to-length ratio but the direction of the effect was modified effect by GA.

Table 1. Demographic and clinical characteristics among a cohort of n = 278 mother-infant dyads in rural Bangladesh
Table 1_v2.pdf1. Of the 299 pregnancies in the cohort, information on both birth size (n = 278) and inflammation biomarkers (n = 298) was available for n = 278 (93.0%)
2. Z scores calculated using Intergrowth-21 standard


Figure 1. Relationship between cord blood IL-1β concentration (ln transformed) and birth size ((A)weight and (B)weight-to-length ratio z scores). Spearman correlations (i) and box and whiskers plots by gestational age and SGA and weight-to-length ratio status (ii)
Figure 1.pdf1. Preterm (GA <= 37 wks GA), Early term (37<= wks GA < 39), Full term (39<= wks GA < 41), Late-/post-term (wks GA >= 41)

Table 2. High levels of cord inflammtory markers ( > 67th%ile v.<= 67th%ile) and birth size < 10th percentile among infants born ≥ 39 and < 39 weeks gestation. Logistic regression
Table 2.pdf1. Model 2 adjusted for maternal upper arm circumference < 22 cm, parity, educational attainment, depression score (PHQ), infant sex, and gestational age of birth