Background: Air pollution exposure during pregnancy has been associated with adverse birth outcomes. Uncertainties remain about the effect at very low exposure levels. The aim of this study was to explore the association of maternal exposure to air pollutants during pregnancy at very low exposure levels with birth weight and estimate the health impact.

Methods: The MATEX birth cohort (226,551 singleton births in 2012–2016) was linked with eight modelled air pollutants (PM2.5, PM10, PMcoarse, NO2, NOx, CO, SO2, O3) at home address during pregnancy. Multiple regression was used to estimate the change in birth weight (in g) associated with individual-level mean exposure during pregnancy. We tested different adjustment models and conducted sensitivity analyses. We also estimated the potential number of low birth weight cases attributable to PM2.5 to quantify the public health issues at the prevailing low exposure levels.

Results: PM2.5 was associated with the largest reduction of birth weight (-6.5 g per 1 µg/m3) followed by PMcrs (-4.9 g) and PM10 (-3.0 g). Among the gaseous pollutants the strongest reduction in birth weight was observed for NO2 (-0.8 g), followed by CO (-0.5 g), NOx (-0.4 g) and SO2 (-0.2 g). On the contrary, O3 was associated with a modest increase in birth weight (+ 0.9 g). Effects on births weight were observed also below WHO guideline values. When accounting for the prevailing exposure levels in Finland, CO was associated with the biggest reduction in birth weight. The effect of PM2.5 exposure on birthweight corresponds to a loss of 30 g at mean exposure. Assuming a causal relationship, about 700 cases of low birth weight could be attributable to PM2.5 in Finland during the study period.

Conclusions: No clear evidence on safe exposure level was found in this study. All pollutants were associated with reduced birthweight except ozone. Causality and confounding due to correlations warrant specific attention.