Previous studies have shown that active smoking during pregnancy can reduce the level of neonatal cord blood leptin, and thereby affect birth weight. However, few studies have studied the association of passive smoking during pregnancy with leptin in neonatal cord blood and birth weight.

To explore the effects of passive smoking in varied pregnancy stages and entire pregnancy on neonatal cord blood leptin level and birth weight in a certain rural area of Yunnan, and potential mediating role of cord blood leptin.

Based on a prospective prenatal cohort study conducted in Xuanwei County, Yunnan Province, a total of 545 mother-infant pairs were included in this study from early pregnancy enrollment to delivery. The demographic information and reproductive history of the subjects were collected by questionnaire. The urine samples of pregnant women in the first, second, and third trimesters of pregnancy were collected during regular prenatal examinations. Umbilical cord blood samples were collected from newborns at birth. The concentration of urine cotinine (UC) was measured by ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS). According to the results of UC level during pregnancy, the study subjects were divided into three groups: negative group (<LOD), low exposure group (LOD-M), and high exposure group(>M). The level of leptin in cord blood was detected by ELISA. Multiple linear regression was used to analyze the effect of passive smoking on umbilical cord blood leptin in newborns during pregnancy. Path analysis was used to explore the relationship among passive smoking during pregnancy, neonatal cord blood leptin, and birth weight.

The average exposure rate of passive smoking during pregnancy was 87.28%, and the exposure rate for entire pregnancy was 76.88%. The median concentration of leptin in neonatal cord blood was 4.17 μg·L⁻¹. After adjusting for maternal age, ethnicity, educational level, pre-pregnancy body mass index (BMI), gestational weight gain, parity, annual household income, infant sex, and birth weight, we found that low level (b=−3.388, P=0.001) and high level (b=−2.738, P=0.006) of passive smoking in the first trimester of pregnancy had negative associations with leptin concentration of cord blood by multiple linear model. The path analysis results showed that passive smoking in the first trimester and pre-pregnancy BMI directly affected leptin levels, and the sizes of direct effects were −0.073 and −0.087 (both P<0.05) respectively. Passive smoking in late pregnancy, gestational weight gain, premature, newborn girls, parity, and pre-pregnancy BMI directly affected birth weight, and the sizes of direct effects were −0.063, 0.191, −0.301, −0.128, −0.121, and 0.167 (all P<0.05), respectively. No mediating role of leptin was found in the effect of passive smoking on neonatal birth weight.

Passive smoking exposure during pregnancy is common among rural women in Yunnan Province. Passive smoking in the first trimester may be key in decreasing the leptin level of neonatal cord blood. Passive smoking in third trimester may lead to a decrease in birth weight. No evidence shows that leptin mediates the relationship between passive smoking and birth weight.