Background: The incidence of both the survival of premature infants and type 1 diabetes (T1D) is rising worldwide. The pathophysiology of T1D depends on the differentiation of cellular and humoral immune systems recognition of the self, which occurs during the neonatal period and is influenced by environmental exposures. Aim: To assess the association between nutritional, antibiotic and parenteral exposures during the neonatal period and the development of pediatric T1D. Methods: a multicenter, paired case-control study. Preterm subjects who were born between 1990-2013, and developed T1D before the age of 18 years (T1D group), were paired with subjects who didn't develop TID (Control group) at a ratio of 1:3 by: gender, gestational age (GA), month of birth and birth medical center . This preterm population is a model of a neonatal population with strictly meticulous follow up data available in all medical centers in Israel. Data retrieved from charts included maternal history, delivery method, ethnic origin, weight for GA, NICU length of hospitalization, medications, parenteral fluid types, feeding modes and their timing. Univariate and multivariate analysis via Generalized Estimating Equations (GEE) using a binary logistic regression model were performed to evaluate the association between T1D and the assessed exposures. Results: T1D group included 52 subjects, 26 males, median GA 35 (26-36), mean BW 2307.6±534.5. Control group included 133 subjects, 69 males. median GA 35 (27-36), mean BW 2094.5±484.1. In univariate analysis, the odds ratio (OR) for T1D increased with each extra 100 grams in BW (OR 1.13, 95%CI 1.057-1.206, p < 0.001), later day of parenteral glucose initiation (OR 1.61, 95%CI 1.05-2.48, p 0.03), and parenteral cephalosporines initiated beyond the first week of life (OR 3.25, 95%CI 1.504-7.04, p 0.003). Treatment with narrow spectrum beta-lactam antibiotics during the first week of life was associated with lower risk for T1D development (OR 0.23, 95%CI 0.12-0.47, p < 0.001). Multivariate analysis revealed significant association between both treatment with cephalosporins beyond the first week of life and later day of parenteral glucose initiation and the development of pediatric T1D (OR 6.49, 95%CI 1.54-27.44, p 0.011 and OR 1.61, 95%CI 1.05-2.48, p 0.002, respectively). No association with feeding modality (oral/nasogastric), timing and type of feeding (breast or bottle) was found. Conclusions: This is the first report indicating a significant association between parenteral exposure and timing of variable antibiotic treatments and glucose solution administration during the neonatal period with development of T1D. This report indicates the need for a larger study to conclude clinical implications of management and exposure to these substances during the neonatal period and its relevance to the neonatal microbiome.
Read full abstract