Table 2.
Exposure, outcome, and statistical information of the 15 included studies on the association between air pollution and childhood obesity.
| Study ID | Author (Year) | Exposure | Duration | Exposure Assessment | Outcome Definition | Statistical Model | Adjusted Covariates |
|---|---|---|---|---|---|---|---|
| 1 | Zheng et al. (2021) | PM10, PM2.5, O3, NO2 | Long-term | Monitoring stations | Age-and-sex specific BMI cut-offs (Chinese national standard) | Multivariate regression model | Sex, age, paternal, sugar-sweetened beverage consumption, sweetened food consumption, frequency of having breakfast, fried food consumption, physical activity duration |
| 2 | Zhang et al. (2021a) | PM10, PM2.5, PM1, NO2 | Long-term | Satellite-based spatial-temporal model | Age-and-sex specific BMI cut-offs (Chinese national standard) | Mixed-effects linear and logistic regression models | Age, physical activity, fruit & vegetable intake, parental smoking, parental education, north or south, urban residency, regional GDP per capita |
| 3 | Zhang et al. (2021b) | PM10, PM2.5,PM1, NO2 | Long-term | Satellite-based spatial-temporal model | Waist circumference (Chinese national standard) | Generalized linear mixed-effects models | Age, sex, weight status, temperature, relative humidity, parental education level achieved, parental smoking status, parental alcohol consumption, family history of type 2 diabetes, hypertension, obesity, or cerebrovascular disease, outdoor physical activity time, diet of high fat, SSBs intake. |
| 4 | Tamayo et al. (2021) | PM2.5 | Long-term | Hybrid spatio-temporal model | Age-specific BMI (WHO standard) | Logistic regression models | Age, sex, SES, and smoking status |
| 5 | Bont et al. (2021) | PM10, PM2.5, NO2 | Long-term | Land use regression model | Age-and-sex specific BMI (WHO standard) | Cox proportional hazards models | Sex, deprivation index, nationality, deprivation index, and had age (1-year categories) in the strata statement. |
| 6 | Vrijheid et al. (2020) | NO2 | Long-term | Land use regression model | Age-and-sex specific BMI (WHO standard) | Linear regression models, and logistic regression models | Sex, maternal BMI, maternal education, maternal age at conception, parity, parental country of origin, breastfeeding, and birth weight |
| 7 | Guo et al. (2020) | PM2.5 | Long-term | Machine-learning model | Age-and-sex specific BMI cut-offs (Chinese national standard) | Logistic regression models | Sex, age, urbanity, boarding school or not, economic level, maternal occupation, maternal education, vegetable intake, fruit intake, beverages intake, activity times, ventilation, cooking fuel type, household heating fuel type, school heating fuel type, and secondhand smoke duration |
| 8 | Bont et al. (2020) | PM10, PM2.5, NO2 | Long-term | Land use regression model | BMI z-scores (WHO standard) | Linear spline multilevel model | Sex, age, deprivation index, nationality |
| 9 | Chen et al. (2020) | NO2 | Long-term | Land use regression model | Age- and sex-specific z scores for BMI (WHO standard) | Generalized estimating equation models, Distributed lag nonlinear models | Maternal age, maternal education, annual household income and residence area |
| 10 | Bont et al. (2019) | PM10, PM2.5, NO2 | Long-term | Land use regression model | Age- and sex-specific z scores for BMI (WHO standard) | Multilevel mixed linear and ordered logistic models | Maternal and paternal education, maternal and paternal country of birth, paternal employment status, number of siblings, household status and maternal smoking during pregnancy |
| 11 | Bloemsma et al. (2019) | PM10, PM2.5, NO2 | Long-term | Land use regression model | Age-and-sex specific BMI (International Obesity Task Force cut-offs) | Generalized linear mixed models | Age, sex maternal level of education, paternal level of education, maternal smoking during pregnancy, parental smoking in child’s home and neighborhood socioeconomic status and region |
| 12 | Kim et al. (2018) | NOx | Long-term | California line-source dispersion model | BMI (US CDC criteria) | Linear mixed effects models | Age, sex, race/ethnicity, parental education, and Spanish baseline questionnaire |
| 13 | Fioravanti et al. (2018) | PM10, PM2.5, NO2 | Long-term | Land use regression model | Age- and sex-specific z scores for BMI (WHO standard) | Logistic regression models, Generalized Estimating Equation models and linear regression models | Maternal and paternal education, maternal pre-pregnancy BMI, maternal smoking during pregnancy, gestational diabetes, maternal age at delivery, gestational age, childbirth weight, breastfeeding duration, age at weaning and inversely weighted for the probability of participation at baseline and at the two follow-ups, respectively |
| 14 | McConnell et al. (2015) | NOx | Long-term | California line-source dispersion model | Age-and-sex specific BMI (US CDC criteria) | Multilevel linear model | Sex, ethnicity, community, year of enrollment, and age |
| 15 | Dong et al. (2014) | PM10, NO2, SO2, O3 | Long-term | Monitoring stations | Age-and-sex specific BMI standards (Chinese CDC criteria) | Logistic regression | Age, gender, parental education, breastfeeding, low birth weight, area of residence per person, house decorations, home coal use, ventilation device in kitchen, air exchange in winter, passive smoking exposure, and districts |
Abbreviations: PM10, particulate matter with the diameter ≤ 10 mm; PM2.5, particulate matter with diameter ≤ 2.5 mm; PM1, particulate matter with the diameter ≤ 1 mm; NO2, nitrogen dioxide; NOx, nitrogen oxides; SO2, sulfur dioxide; O3, ozone; BMI, body mass index; WHO, World Health Organization; US, The United States; CDC, Center for Disease Control and Prevention.