Application of the navigation guide systematic review methodology to evaluate prenatal exposure to particulate matter air pollution and infant birth weight.

Low birth weight is an important risk factor for many co-morbidities both in early life as well as in adulthood. Numerous studies report associations between prenatal exposure to particulate matter (PM) air pollution and low birth weight. Previous systematic reviews and meta-analyses report varying effect sizes and significant heterogeneity between studies, but did not systematically evaluate the quality of individual studies or the overall body of evidence. We conducted a new systematic review to determine how prenatal exposure to PM, PM, and coarse PM (PM) by trimester and across pregnancy affects infant birth weight. Using the Navigation Guide methodology, we developed and applied a systematic review protocol [CRD42017058805] that included a comprehensive search of the epidemiological literature, risk of bias (ROB) determination, meta-analysis, and evidence evaluation, all using pre-established criteria. In total, 53 studies met our inclusion criteria, which included evaluation of birth weight as a continuous variable. For PM and PM, we restricted meta-analyses to studies determined overall as "low" or "probably low" ROB; none of the studies evaluating coarse PM were rated as "low" or "probably low" risk of bias, so all studies were used. For PM, we observed that for every 10 µg/m increase in exposure to PM in the 2nd or 3rd trimester, respectively, there was an associated 5.69 g decrease (I: 68%, 95% CI: -10.58, -0.79) or 10.67 g decrease in birth weight (I: 84%, 95% CI: -20.91, -0.43). Over the entire pregnancy, for every 10 µg/m increase in PM exposure, there was an associated 27.55 g decrease in birth weight (I: 94%, 95% CI: -48.45, -6.65). However, the quality of evidence for PM was rated as "low" due to imprecision and/or unexplained heterogeneity among different studies. For PM, we observed that for every 10 µg/m increase in exposure in the 3rd trimester or the entire pregnancy, there was a 6.57 g decrease (I: 0%, 95% CI: -10.66, -2.48) or 8.65 g decrease in birth weight (I: 84%, 95% CI: -16.83, -0.48), respectively. The quality of evidence for PM was rated as "moderate," as heterogeneity was either absent or could be explained. The quality of evidence for coarse PM was rated as very low/low (for risk of bias and imprecision). Overall, while evidence for PM and course PM was inadequate primarily due to heterogeneity and risk of bias, respectively, our results support the existence of an inverse association between prenatal PM exposure and low birth weight.