*2.4. Statistical Analysis*

We analyzed data with the SPSS software (SPSS Inc., Chicago, IL, USA). We examined the distributions of all continuous variables (age, body weight, height, body mass index [BMI]) for skewness. Data of the variables showing rightward skewing were presented as geometric mean ± standard deviation (SD) values. We used age-adjusted logistic regression analysis to derive the prevalence odds ratio (POR) for high blood Pb levels (≥ 5 μg/dL) and for abnormal growth. We used the generalized linear model (GLM) analysis to derive the age-adjusted mean blood Pb and age-adjusted BMI. We also used GLM to evaluate an effect of parental occupation and the child's gender on blood Pb levels and BMI. *p* values ≤ 0.05 for two-tailed tests were assumed to identify statistical significance.
