Abstract 1474 Poster Session III, Monday, 5/3 (poster 52)

Hyaline membrane disease (HMD) is the most common form of lung disease among premature newborn infants. Although HMD is thought to occur because of immaturity of the pulmonary surfactant system, variations in the incidence of HMD are only partially accounted for by gestational age. Studies among siblings and unselected twin pairs have suggested clustering of HMD within families, but these studies have been rather small-scale due in part to the small fraction of informative sibships. We adopted an alternate strategy and studied pairs of premature twins (birth weights ≤ 1500 g) admitted during a 15-year period to the Vanderbilt NICU (n=166 pairs). The rate of HMD was 61% compared with a gestational age and gender standardized rate of 41% in the corresponding singleton population (n=1881). Among the twin pairs, 49% had 2 affected, 25% had 1, and 26% had 0; the (unadjusted) odds ratio measuring the within-pair association for HMD was 7.9. This measure is likely influenced by both environmental and genetic factors. One factor is twin birth order (in the absence of c-section delivery): second born twins had a higher rate of HMD (61/83, 73%) than firstborns (42/87, 48%). To account for such factors, the paired HMD outcomes were modeled using bivariate logistic regression with covariates including relevant factors in the singleton population plus twin-related factors: gestational age, gender, race, antenatal steroids, inborn/outborn, c-section, birth order in the absence of c-section, and 1-minute Apgar score. With zygosity information unavailable, separate adjusted odds ratios for association were estimated among pairs of like- (23.2) and unlike-sex (3.2) twins (ratio of like/unlike sex = 7.2; 90% CI: 0.9, 53.4). These findings support a positive association for HMD within low birth weight twin pairs. The elevated association observed among like-sex twins is consistent with genetic factors influencing HMD, though alternate explanations such as the existence of unmeasured factors related to placentation are not ruled out. Since the normal production of pulmonary surfactant depends on many different phospholipid and protein metabolic pathways, it is possible that mutations in the genes controlling these pathways alter susceptibility to HMD. This study provides further justification to pursue allele sharing and other linkage studies of these and other genes to identify genes that predispose to HMD.