Data from: Identification of obesity risk factors in 3−12-year-old children and adolescents with prior respiratory tract infections via interpretable machine and deep learning models
Abstract
This dataset contains cross-sectional survey data collected using a stratified sampling method. Data collection was conducted in two rounds between September 2020 and January 2022 and included children and adolescents from Beijing and surrounding areas. The survey included questionnaires on the physical condition and lifestyle habits of children and adolescents, as well as information on their mothers and perinatal history. The study protocol was reviewed and approved by the Ethics Committee of the China-Japan Friendship Hospital and the Ethics Committee of Beijing University of Chinese Medicine, respectively. The parents or guardians of all children signed written informed consent forms. All data involving human subjects in this dataset have undergone rigorous de-identification, retaining no more than three indirect identifiers.
Dataset DOI: 10.5061/dryad.gf1vhhn44
Description of the data and file structure
This dataset contains cross-sectional survey data collected through cluster sampling, covering preschoolers and elementary school students in Beijing, Tangshan, and Hebei. Data collection took place in two rounds, spanning September 2020 to January 2022. The file is named 20_22.csv.
20_22 Data
birthweight: Categorized infant birth weight.
birthheight: Categorized infant birth height.
education_father: Categorical variable indicating the father's educational attainment.
education_mother: Categorical variable indicating the mother's educational attainment.
foodallergy: Binary indicator of food allergy.
drugallergy: Binary indicator of drug allergy.
dentalcaries: Number of dental caries.
gestational_week: Categorized gestational age.
deliverymode: Categorical variable indicating the mode of delivery.
assistedreproduction: Binary indicator of assisted reproductive technology.
pregnancyorder: Categorized pregnancy order.
deliveryorder: Categorized delivery order.
twins: Binary indicator of multiple birth.
infancyfeeding: Categorical variable indicating the infant feeding pattern at 6 months of age.
breastfeeding: Duration of breastfeeding (months).
solidfood: Age at introduction of complementary foods (months).
fastfood: Categorical variable indicating the frequency of fast food and fried food consumption.
sweetfood: Categorical variable indicating the frequency of sweet food consumption.
nightmeal: Categorical variable indicating the frequency of eating before bedtime.
dm_kin: Number of family members with diabetes.
outdoor: Average daily outdoor activity time (hours).
sit: Average daily sedentary time (hours).
screen: Average daily screen time (hours).
sleepduration: Average daily sleep duration, including naps (hours).
eatspeed: Average eating time per meal (minutes).
ethnicity: Categorical variable indicating ethnicity.
income: Categorical variable indicating annual household income.
bmi_father: Categorized paternal body mass index (BMI).
BMI_mother: Categorized maternal body mass index (BMI).
obesity: Binary indicator of childhood obesity.
All missing data in this paper resulted from parents either not providing an answer or being unsure when filling out the questionnaire; these values have been filled with “null” in the dataset.
Data de-identification
To protect participant confidentiality, the publicly available dataset has been de-identified prior to release. Variables with a potential risk of indirect identification, including birth characteristics and parental anthropometric measurements, have been generalized into categorical variables where appropriate. No direct personal identifiers are included in this dataset.
Code/software
Stata 15.0, PyCharm (Community Edition 2024.3.4 x64) embedded in Python (Python Software Foundation) software (Version 3.6.1) under the Windows 10 system, and R coding platform (version 4.3.3)
Human subjects data
This study was reviewed and approved by the Ethics Committee of the China-Japan Friendship Hospital and the Ethics Committee of Beijing University of Chinese Medicine. The parents or legal guardians of all participants signed written informed consent forms. The informed consent process included permission to share de-identified research data in accordance with institutional ethical requirements and relevant regulations. I have ensured that the data has been de-identified and that no personal information about the participants will be disclosed.