Abstract: Background There's a scare of evidence in association between air pollution exposure and lipid metabolism in Chinese children and adolescents.Objective To explore the association between ambient air pollution exposure and blood lipids metabolism of children and adolescents.DesignCross-sectional study.Methods Students aged 7-18 years were recruited from 4 Shanghai primary, middle and high schools in a cross-sectional study whose fasting blood samples were collected during 2014. Informed consents for participating in the study and blood collection were got from guardians. Serum HDLC, LDLC, TG and TC were examined and dyslipidemia was defined as occurrence of abnormality in any one of them (HDLC≤1.04 mmol·L^-1,LDLC≥3.37 mmol·L^-1,TG≥1.70 mmol·L^-1 or TC≥5.18 mmol·L^-1). With home address provided by the participants and a spatio-temporal model of air pollution, daily average concentrations of PM_2.5, O₃ and NO₂ in the latest 1, 3 and 6 months before blood collection at the participants' home addresses were measured. With linear regression and logistic regression, associations of home exposure to individual air pollutants at different durations with lipids' concentration and risk of dyslipidemia were examined. On the basis of multiple adjustment, pollutants were further mutually adjusted to check independence of association between pollutants exposure and lipids metabolism.Main outcome measures Air pollution exposure of children based on their home location, fasting venous serum lipid profile and dyslipidemia.Results A total of 1,317 students were included in this analysis with 660 girls(50.1%). Mean(SD) concentrations of HDLC, LDLC and TC were (1.52±0.29)mmol·L^-1, (2.35±0.63) mmol·L^-1 and (4.03±0.72) mmol·L^-1, while median (1^st quartile, 3^rd quartile) concentration of TG was 0.59(0.43,0.80) mmol·L^-1, and 142(10.8%) students were identified to have dyslipidemia. Daily concentrations of PM_2.5, O₃ and NO₂ in the past 1, 3 and 6 months varied at ranges of 51.4~61.5, 20.4~22.8 and 11.1~12.8 μg·m^-3. With every IQR increase of PM_2.5 concentration in the latest 3 and 6 months, HDLC decreased by 0.07(95%CI:0.03~0.11)mmol·L^-1 and 0.09(95%CI:0.05~0.14)mmol·L^-1, and OR(95%CI) for risk of dyslipidemia was 1.86(95%CI:1.42~2.44) with each IQR increase of PM_2.5 in the latest 6 months. These associations remained significant after adjusting for exposure of other pollutants, but no independently significant associations of other serum lipids and ambient air pollution exposure were found.Conclusion This study found a significant association of PM_2.5 exposure with adverse lipid profiles in children and adolescents.

Key words: Dyslipidemia, Air pollution, Children and adolescents