Short-term effects of wildfire-specific fine particulate matter and its carbonaceous components on perinatal outcomes: A multicentre cohort study in New South Wales, Australia

Background: Epidemiological evidence on the association between wildfire-specific fine particulate matter (PM2.5) and its carbonaceous components with perinatal outcomes is limited. We aimed to examine the short-term effects of wildfire-specific PM2.5 and its carbonaceous components on perinatal outcomes. Methods: A multicentre cohort of 9743 singleton births during the wildfire seasons from 1 September 2009 to 31 December 2015 across six cities in New South Wales, Australia were linked with daily wildfire-specific PM2.5 and carbonaceous components (organic carbon and black carbon). Adjusted distributed lag Cox regression models with spatial clustering were performed to estimate daily and cumulative adjusted hazard ratios (aHRs) during the last four gestational weeks for preterm birth, stillbirth, nonvertex presentation, low 5-min Apgar score, special care nursery/neonatal intensive care unit (SCN/NICU) admission, and caesarean section. Results: Daily aHRs per 10 mu g/m(3) PM2.5 showed nearly inverted 'U'-shaped positive associations and daily cumulative aHRs that increased with increasing duration of the exposures. The aHRs for lag 0-6 days were 1.17 (95 % CI: 1.04, 1.32) for preterm birth, 1.40 (95 % CI: 1.11, 1.78) for stillbirth, 1.20 (95 % CI: 1.08, 1.33) for nonvertex presentation, 1.12 (95 % CI: 0.93, 1.35) for low 5-min Apgar score, 0.99 (95 % CI: 0.83, 1.19) for SNC/NICU admission, and 1.01 (95 % CI: 0.94, 1.08) for caesarean section. Organic carbon and black carbon components for lag 0-6 days showed positive associations. The highest component-specific aHRs were 1.09 (95 % CI: 1.03, 1.15) and 4.57 (95 % CI: 1.96, 10.68) for stillbirth per 1 mu g/m(3) organic carbon and black carbon, respectively. The subgroups identified as most vulnerable were female births, births to mothers with low socioeconomic status, and births to mothers with high biothermal exposure. Conclusions: Positive associations of short-term wildfire-specific PM2.5 exposure and its carbonaceous components with adverse perinatal outcomes suggest that policies to reduce exposure would benefit public health.