• Poster

Familial and genetic susceptibility to sepsis in very low birth weight twins


Boghossian, N. S., & Page, G. P. (2010, May). Familial and genetic susceptibility to sepsis in very low birth weight twins. Presented at Pediatric Academic Societies Annual Meeting, Poster Session: Neonatal Infectious Diseases, Vancouver, British Columbia, May 1-4, .


BACKGROUND: Neonatal sepsis is a major cause of neonatal mortality and morbidity among very low birth weight (VLBW) infants.
Variation in the development of neonatal sepsis and in the response to therapy suggests a major role played by environmental and
genetic factors.
OBJECTIVE: To identify environmental factors associated with late-onset sepsis (LOS) and to use this information to estimate the
heritability in susceptibility to LOS in VLBW twins and higher order multiples.
DESIGN/METHODS: Infants with birth weight 401-1500 g born at or admitted in the first 14 days to a center of the NICHD
Neonatal Research Network during 2002-2008 were studied. Data were obtained on 5,959 offspring of multiple pregnancies and
16,965 singleton pregnancies. Logistic regression was used to identify factors associated with LOS in twin and singleton
pregnancies. Concordance rates in LOS among twins and higher order multiples were assessed both with and without
environmental/covariate effect. The rate of monozygosity and dizygosity was estimated using the Weinberg rule.
RESULTS: A total of 2,417 pairs (twins and pairs from multiples) were examined. The rate of monozygosity was estimated to be
29.2%. The observed rate of LOS in pairs was 23.8%. When LOS occurred in one twin, it was observed in the second twin more
often than expected (45%). When sepsis did not occur in the first twin, it occurred in only 18% of the second twins (crude odds
ratio=3.75, 95% CI 3.04-4.63). The concordance rate for LOS was similar between same-sex pairs (71.5%) and different-sex pairs
(74.0%) (p=0.17). 12.4% of different-sex twin pairs were both affected by LOS, and 11.2% of same-sex twin pairs were both
affected by LOS. Significant factors associated with LOS included center, gender, gestational age, birth weight, bronchopulmonary
dysplasia, antenatal steroids, any surgery, parenteral alimentation, days on the ventilator, and days of supplementary oxygen.
Adjusting for the above covariates increased the concordance among twin pairs but did not cause the concordance rate among
same and different sex twins to be different from one another. The estimated heritability of LOS with and without covariate
adjustment was 0%.
CONCLUSIONS: Numerous factors influenced the development of LOS. The high concordance of LOS among twins suggests that
there may be prenatal factors, as yet unidentified, that influence the development of LOS. Our data however, do not support a
familial component to susceptibility to LOS.