OBJECTIVE: To examine the distribution and time trends of admission temperature in very preterm infants, effects of temperature-preserving interventions and associations between hypothermia and outcomes. METHODS: Population-based cohort study of infants <
32 gestational weeks without lethal malformations admitted to a neonatal intensive care unit within 2 hours of birth between 1 January 2012 and 31 December 2021. Hypothermia was defined as mild (36.0°C-36.4°C), moderate (34.5°C-35.9°C) or severe (<
34.5°C). Associations between admission temperature and outcomes were investigated using multivariate logistic regression, adjusting for sex, weight Z-score, gestational age, antenatal steroids, surfactant and hospital. RESULTS: In total, 2457/4879 (50.4%) had normal admission temperatures between 36.5°C and 37.5℃, 1314 (26.9%) had mild hypothermia and 886 (18.2%) had moderate to severe hypothermia, with a significant interhospital variance of 7%-40%. Moderate to severe hypothermia decreased from 24% in 2012 to 16% in 2021 (p CONCLUSIONS: Although the incidence of hypothermia decreased during the study period, it remains a concern due to significant associations with adverse outcomes. Areas for further improvement include addressing interhospital variance and enhancing temperature monitoring in the delivery unit.