The Proportion of Preterm Birth Attributable to Modifiable Risk Factors: A Retrospective Cohort Study in Nova Scotia, 2005-2019

Abstract

Preterm birth (PTB), occurring at <37 weeks’ gestation, is a leading cause of child mortality. The objective was to estimate the population attributable risk (PAR%) and population impact fraction (PIF%) of modifiable risk factors for PTB in Nova Scotia. A population-based retrospective cohort of women and singleton infants delivered from 2005 to 2019 was conducted. Eleven theoretically modifiable risk factors were studied. Poisson regressions were used with 10 imputed datasets to estimate the probability of PTB (p0) in the population under scenarios where risk factors were removed (PAR%) or reduced (PIF%); calculated as 100*(pobs-p0)/pobs. A total of 123607 infants were included. The largest PAR% was for smoking during pregnancy, 5.8% of PTB. Results varied for risk factors across time. A small proportion of PTB was estimated to be attributable to the risk factors studied. These findings can inform which risk factors may be targeted to reduce PTB in Nova Scotia.