摘要:Background Maternal exposure to wood fuel smoke may lead to impaired fetal growth due to hypoxia and or oxidative stress from smoke constituents such as carbon monoxide and particulate matter. Objectives We studied the risk of low birth weight (LBW) and reduced mean birth weight in relation to reported use of wood for cooking during the prenatal period, compared with natural gas (NG). Methods We studied a historical cohort of women who had a singleton live birth in the years 2000–2002, from a semirural area of Pakistan. Infant’s birth weight was obtained from records, and prenatal records had data for maternal body mass index and parity. Cooking habits, daytime sleep habits, and type of fuel used during the pregnancies in 2000–2002 were ascertained by a survey done in 2004–2005. We performed multiple linear and logistic regression modeling using propensity scores to adjust for confounding variables. Results Unadjusted mean (± SD) birth weight was 2.78 ± 0.45 kg in wood users, and 2.84 ± 0.43 kg ( p < 0.06) in NG users. Infants born to wood users averaged 82 g lighter than infants born to NG users when weight was adjusted for confounders ( p < 0.07). The rate of LBW (< 2,500 g) was 22.7% among wood users compared with 15.0% in NG users ( p < 0.01), for an adjusted relative risk of 1.64 (95% confidence interval, 1.10–2.34). The population attributable risk for LBW explained by wood use was estimated to be 24%. Conclusion Cooking with wood fuel during pregnancy, a potentially modifiable exposure, was associated with LBW and marginally lower mean birth weight compared with using NG.