BACKGROUND: Studies suggest that exposure to ambient air pollution during pregnancy may be associated with increased risks of birth defects (BDs), but conclusions have been inconsistent. This study describes the ethnic distribution of major BDs and examines the relationship between air pollution and BDs among different ethnic groups in Liuzhou city, China. METHODS: Surveillance data of infants born in 114 registered hospitals in Liuzhou in 2019 were analyzed to determine the epidemiology of BDs across five major ethnic groups. Concentrations of six air pollutants (PM2.5, PM10, SO2, CO, NO2, O3) were obtained from the Liuzhou Environmental Protection Bureau. Logistic regression was used to examine the associations between ambient air pollution exposure and risk of BDs. RESULTS: Among 32,549 infants, 635 infants had BDs, yielding a prevalence of 19.5 per 1000 perinatal infants. Dong ethnic group had the highest prevalence of BDs (2.59%), followed by Yao (2.57%), Miao (2.35%), Zhuang (2.07%), and Han (1.75%). Relative to the Han ethnic group, infants from Zhuang, Miao, Yao and Dong groups had lower risks of congenital heart disease, polydactyly, and hypospadias. The Zhuang ethnic group had higher risks of severe thalassemia, cleft lip and/or palate, and syndactyls. Overall BDs were positively correlated with air pollutants PM10 (aOR =1.14, 95% CI:1.12 ~ 2.43; aOR =1.51, 95% CI:1.13 ~ 2.03 for per 10mug/mg3 increment) and CO (aOR =1.36, 95% CI:1.14 ~ 2.48; aOR =1.75, 95% CI:1.02 ~ 3.61 for every 1 mg /m3 increment) in second and third month of pregnancy. SO2 was also significantly associated with BDs in the second month before the pregnancy (aOR = 1.31; 95% CI: 1.20 ~ 3.22) and third month of pregnancy (aOR =1.75; 95% CI:1.02 ~ 3.61). Congenital heart disease, polydactyl, cleft lip and/or palate were also significantly associated with PM10, SO2 and CO exposures. However, no significant association was found between birth defects and O3, PM2.5 and NO2 exposures (P > 0.05). CONCLUSION: This study provides a comprehensive description of ethnic differences in BDs in Southwest China and broadens the evidence of the association between air pollution exposure during gestation and BDs.

BACKGROUND: Precision prevention is increasingly important in HIV prevention research to move beyond universal interventions to those tailored for high-risk individuals. The current study was designed to develop machine learning algorithms for predicting adolescent HIV risk behaviours. METHODS: Comprehensive longitudinal data on adolescent risk behaviours, perceptions, peer and family influence, and neighbourhood risk factors were collected from 2564 grade-10 students at baseline followed for 24 months over 2008-2012. Machine learning techniques [support vector machine (SVM) and random forests] were applied to innovatively leverage longitudinal data for robust HIV risk behaviour prediction. In this study, we focused on two adolescent risk behaviours: had ever had sex and had multiple sex partners. Twenty percent of the data were withheld for model testing. RESULTS: The SVM model with cost-sensitive learning achieved the highest sensitivity, at 79.1%, specificity of 75.4% with AUC of 0.86 in predicting multiple sex partners on the training data (10-fold cross-validation), and sensitivity of 79.7%, specificity of 76.5% with AUC of 0.86 on the testing data. The random forest model obtained the best performance in predicting had ever had sex, yielding the sensitivity of 78.5%, specificity of 73.1% with AUC of 0.84 on the training data and sensitivity of 82.7%, specificity of 75.3% with AUC of 0.87 on the testing data. CONCLUSION: Machine learning methods can be used to build effective prediction model(s) to identify adolescents who are likely to engage in HIV risk behaviours. This study builds a foundation for targeted intervention strategies and informs precision prevention efforts in school-setting.