Baseline characteristics were tabulated to describe general characteristics of the mothers and infants, as well as for smoking habits and other environmental factors such as home fuel resources usage, garbage burning, and household insecticides usage. Continuous variables were expressed as mean and standard deviation or median and interquartile range if distributions were skewed. Categorical variables were expressed as number of subject and its percentage.
Socio-economic status (SES) i.e. household income and level of education, maternal age, parity [6 (link), 18 (link), 19 (link)], BMI increment during pregnancy (Δ BMI) [19 (link)], mother’s working status, mother’s active and passive smoking during pregnancy [8 (link), 19 (link)], and pesticides exposure in pregnancy were a priori considered as possible confounders. All of the possible confounders were treated as categorical, excluding mother’s age and Δ BMI. Level of education was categorized as elementary, high school, and under-post graduate, while family income was categorized as below the minimum monthly wedges per capita in Jakarta (< 290 USD) or above or equal to the minimum monthly wedges per capita in Jakarta (≥ 290 USD). Maternal co-morbidity/gestational complication [20 (link)] and infant gender were considered as possible effect modifiers [18 (link), 19 (link)].
Multiple linear regression models adjusted for gestational age and subsequently all potential covariates were specified to evaluate the association between outdoor concentration of PM2.5, soot, NOx, and NO2 with birth weight and birth length. Logistic regression was used to explore the association between outdoor air pollutants concentration with low birth weight, expressed as OR with 95% confidence interval. We performed sensitivity analysis to test the robustness of our results when restricting the cohort to women with infant born at gestational age ≥ 37 weeks. We additionally calculated the difference between PM2.5 and soot to further differentiate effects of PM2.5 and soot. This difference was highly correlated (R = 0.99) with PM2.5. We also isolated the effect of soot from PM2.5 by generating new variable as the residual of soot [21 (link)] and further, we tested the association of this new variable (the residual of soot) with birth anthropometry. All effect estimates were expressed for an interquartile increase of each air pollutant i.e. 7.14 μg/m3 for PM2.5, 0.75 × 10− 5 per m for soot, 4.68 μg/m3 for NOx, and 3.74 μg/m3 for NO2. Statistical significance was assumed if 95% confidence intervals did not include the estimation of null values, corresponding to two-sided p values < 0.05. Statistical analyses were conducted using IBM SPSS version 24 for Mac, while the LUR models were developed using R 4.1.2.
Socio-economic status (SES) i.e. household income and level of education, maternal age, parity [6 (link), 18 (link), 19 (link)], BMI increment during pregnancy (Δ BMI) [19 (link)], mother’s working status, mother’s active and passive smoking during pregnancy [8 (link), 19 (link)], and pesticides exposure in pregnancy were a priori considered as possible confounders. All of the possible confounders were treated as categorical, excluding mother’s age and Δ BMI. Level of education was categorized as elementary, high school, and under-post graduate, while family income was categorized as below the minimum monthly wedges per capita in Jakarta (< 290 USD) or above or equal to the minimum monthly wedges per capita in Jakarta (≥ 290 USD). Maternal co-morbidity/gestational complication [20 (link)] and infant gender were considered as possible effect modifiers [18 (link), 19 (link)].
Multiple linear regression models adjusted for gestational age and subsequently all potential covariates were specified to evaluate the association between outdoor concentration of PM2.5, soot, NOx, and NO2 with birth weight and birth length. Logistic regression was used to explore the association between outdoor air pollutants concentration with low birth weight, expressed as OR with 95% confidence interval. We performed sensitivity analysis to test the robustness of our results when restricting the cohort to women with infant born at gestational age ≥ 37 weeks. We additionally calculated the difference between PM2.5 and soot to further differentiate effects of PM2.5 and soot. This difference was highly correlated (R = 0.99) with PM2.5. We also isolated the effect of soot from PM2.5 by generating new variable as the residual of soot [21 (link)] and further, we tested the association of this new variable (the residual of soot) with birth anthropometry. All effect estimates were expressed for an interquartile increase of each air pollutant i.e. 7.14 μg/m3 for PM2.5, 0.75 × 10− 5 per m for soot, 4.68 μg/m3 for NOx, and 3.74 μg/m3 for NO2. Statistical significance was assumed if 95% confidence intervals did not include the estimation of null values, corresponding to two-sided p values < 0.05. Statistical analyses were conducted using IBM SPSS version 24 for Mac, while the LUR models were developed using R 4.1.2.
Free full text: Click here
