Effect of ambient temperature and air pollutants on the risk of preterm birth, Rome 2001-2010

Environ Int. 2013 Nov;61:77-87. doi: 10.1016/j.envint.2013.09.005. Epub 2013 Oct 5.


Introduction: Although the prevalence of preterm births ranges from 5 to 13% and represents the leading cause of perinatal mortality and morbidity in developed countries, the etiology of preterm birth remains uncertain. We aimed to evaluate the effect of short-term exposure to high and low temperatures and air pollution on preterm delivery and to identify socio-demographic and clinical maternal risk factors enhancing individual susceptibility.

Methods: We analyzed all singleton live births by natural delivery that occurred in Rome in 2001-2010. A time-series approach was used to estimate the effect of exposure to minimum temperature, maximum apparent temperature, heat waves, particulate matter with an aerodynamic diameter of 10μm or less (PM10), ozone, and nitrogen dioxide in the month preceding delivery; the analysis was conducted separately for cold and warm seasons. Socio-demographic and clinical risk factors were included as interaction terms.

Results: Preterm births comprised nearly 6% of our cohort. An increase of 1.9% (95% confidence interval (CI) 0.86-2.87) in daily preterm births per 1°C increase in maximum apparent temperature in the 2days preceding delivery was estimated for the warm season. Older women, women with higher education levels, and women with obstetric or chronic pathologies reported during delivery had a lower effect of temperature on the risk of preterm birth, while women with a chronic disease in the two years before delivery and mothers<20years showed a higher effect. A +19% (95% CI 7.91-31.69) increase in preterm births was observed during heat waves. Temperature had no effect during the cold season. We detected a significant effect of PM10 on preterm-birth risk at a lag period of 12-22days during the warm season (+0.69%; 95% CI 0.23-1.15, for 1μg/m(3) increase of pollutant); women with obstetric pathologies and with a higher education level showed a higher risk.

Conclusions: Our results suggest a possible short-term effect of heat and a more delayed and prolonged effect of PM10 exposure on preterm-birth risk, as well as the existence of more susceptible subgroups of women. Our observations support the few reported investigations, and may help to increase awareness among public-health stakeholders and clinicians regarding the role of these environmental exposures as risk factors for premature birth and health consequences for children later in life.

Keywords: Air pollutants; CI; DLM; Gestational age; HW; IQR; MAT; Maternal exposure; NO(2); Newborn; O(3); PM10; Premature birth; TMIN; Temperature; confidence interval; distributed lag model; heat wave; interquartile range; maximum apparent temperature; minimum temperature; nitrogen dioxide; ozone; particulate matter with an aerodynamic diameter of 10μm or less.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Air Pollutants / adverse effects*
  • Air Pollutants / analysis
  • Environmental Exposure / adverse effects
  • Female
  • Humans
  • Infant, Newborn
  • Italy / epidemiology
  • Middle Aged
  • Nitrogen Dioxide / analysis
  • Ozone / analysis
  • Particulate Matter / adverse effects*
  • Particulate Matter / analysis
  • Pregnancy
  • Premature Birth / epidemiology*
  • Risk Factors
  • Seasons
  • Temperature*
  • Young Adult


  • Air Pollutants
  • Particulate Matter
  • Ozone
  • Nitrogen Dioxide