RESUMO
BACKGROUND: Longitudinally conserved microbe-microbe interactions may provide insights to understand the complex dynamic system of early-life gut microbiota among preterm infants. OBJECTIVES: We aimed to profile the covarying network of gut microbiota among preterm infants and investigate its potential influence on host growth (2-5 y). METHODS: We collected time-series stool samples (n = 717 from children and n = 116 from mothers) among 51 preterm and 51 full-term infants from birth up to 5 y of age and among 53 mothers. The included infants underwent time-series measurements of early-life gut microbiota (0-5 y) and growth (2-5 y) from June 2014 to April 2017. The covarying taxa that exhibited consistent covariation from day 1 to year 5 were defined as conserved features in the development of gut microbiota. Childrens' height-for-age z score (HAZ) and weight-for-age z score were calculated according to World Health Organization Child Growth Standards. RESULTS: We observed distinct dynamic patterns of both microbial alpha and beta diversity comparing preterm infants with full-term controls during the very early stage (<3 mo). Moreover, we identified a covarying network containing 10 taxa as a conserved gut microbial feature of these preterm infants from birth to 5 y old. This covarying network was distinctive between preterm and full-term infants before 3 mo of age (P < 0.001) and tended to be similar as the infants grew up. Several covarying taxa of the network during early life (<3 mo) were associated with childhood growth (2-5 y) (eg, Clostridium_sensu_stricto_1 with HAZ, ß = -0.32, q < 0.01), and the human milk feeding duration was a main modulating factor. CONCLUSIONS: Preterm born children possess conserved and distinct covarying microbiota during very early life, which may have a profound influence on their growth later in life. This trial was registered at clinicaltrials.gov as NCT03373721.