Abstract
This study was conducted to confirm the negative associations between testosterone and cortisol levels and health and developmental outcomes in very low-birth weight (VLBW) infants after controlling for prenatal glucocorticoid (GC) treatment. Seventy-four VLBW infant-mother pairs were recruited from a neonatal intensive care unit in the Southeastern United States. We divided the pairs into the complete (n = 58) and incomplete (n = 16) GC treatment groups. Data on infants and mothers were obtained at birth, 40 weeks' postmenstrual age, and 3 and 6 months' corrected age. Salivary testosterone and cortisol levels of the pairs were determined at 40 weeks' postmenstrual age using enzyme immunoassay. Log-linear and general linear mixed models showed that gestational age and birth weight were lower when testosterone was 1 pg/mL higher. When cortisol was 1 [mu]g/dL higher, technology dependence at discharge was higher and motor development at 6 months was lower. Mothers with complete GC treatment had greater parity and gravida, more prenatal visits, and more medical complications. The study outcomes supported our hypothesis that steroid hormonal levels are more predictive of infant health and development than GC treatment completeness. Single dose of GC treatment might be just as effective as 2 doses, although further study with more subjects would be needed to confirm. As the associations with steroid hormonal levels lasted longer than the GC treatment associations, we recommend confirming the predictive effects of testosterone and cortisol levels after 6 months.