Opportunities and challenges of using big data for global health

Clinical data mining: challenges, opportunities, and recommendations for translational applications

Clinical data mining of predictive models offers significant advantages for re-evaluating and leveraging large amounts of complex clinical real-world data and experimental comparison data for tasks such as risk stratification, diagnosis, classification, and survival prediction. However, its translational application is still limited. One challenge is that the proposed clinical requirements and data mining are not synchronized. Additionally, the exotic predictions of data mining are difficult to apply directly in local medical institutions. Hence, it is necessary to incisively review the translational application of clinical data mining, providing an analytical workflow for developing and validating prediction models to ensure the scientific validity of analytic workflows in response to clinical questions. This review systematically revisits the purpose, process, and principles of clinical data mining and discusses the key causes contributing to the detachment from practice and the misuse of model verification in developing predictive models for research. Based on this, we propose a niche-targeting framework of four principles: Clinical Contextual, Subgroup-Oriented, Confounder- and False Positive-Controlled (CSCF), to provide guidance for clinical data mining prior to the model's development in clinical settings. Eventually, it is hoped that this review can help guide future research and develop personalized predictive models to achieve the goal of discovering subgroups with varied remedial benefits or risks and ensuring that precision medicine can deliver its full potential.

Associations between long-term exposure to PM2.5 chemical constituents and allergic diseases: evidence from a large cohort study in China

BACKGROUND

Exposure to air pollutants may cause immune responses and further allergic diseases, but existing studies have mostly, if not all, focused on effects of short-term exposure to PM2.5 on allergic diseases.

OBJECTIVES

We estimated associations of long-term exposure to PM2.5 chemical constituents with allergic disease risks and effect modification.

METHODS

We used the baseline of a newly established, provincially representative cohort of 51,480 participants in southwest China. The presence of allergic rhinitis, allergic asthma, urticaria, and allergic conjunctivitis was self-reported by following a formed questionnaire in face-to-face interviews. The average concentrations of PM2.5 chemical constituents (NO3-, SO42-, NH4+, organic matter [OM], and black carbon [BC]) over participants' residence were estimated using machine learning models. Logistic regression with double robust estimator and weighted quantile sum regression were used to estimate the effects of PM2.5 chemical constituents on allergic disease risks, as well as relative importance of each PM2.5 chemical constituent.

RESULTS

Per interquartile range increase in the concentration of all PM2.5 chemical constituents was associated with the elevated risks for allergic asthma (OR = 1.79 [1.41-2.26]), allergic conjunctivitis (1.54 [1.19-2.00]), urticaria (1.36 [1.25-1.48]), and allergic rhinitis (1.18 [1.11-1.26]). NO3- contributed more to risks for allergic asthma (weight = 46.05 %), urticaria (72.29 %), and allergic conjunctivitis (47.65 %), while NH4+ contributed more to allergic rhinitis (78.07 %). OM contributed most to the risks for allergic asthma (30.81 %) and allergic conjunctivitis (31.40 %). BC was also associated with allergic rhinitis, urticaria, and allergic conjunctivitis, only with a considerable weight for urticaria (24.59 %). Joint effects of PM2.5 chemical constituents on risks for allergic rhinitis and urticaria were stronger in minorities and farmers than their counterparts.

CONCLUSION

Long-term exposure to PM2.5 chemical constituents was associated with the increased allergic disease risks, with NO3- and NH4+ accounting for the largest variance of the associations. Our findings would serve as scientific evidence for developing more explicit strategies of air pollution control.

Associations of screen use with physical activity and social capital amid the COVID-19 pandemic: A network analysis of youths in China

Inconsistent correlations of screen use with physical activity (PA) and social capital (SC) in youths have been observed in existing cross-sectional studies. This study aimed to elucidate associations among variables in screen use, PA, and SC domains during COVID-19, to improve the prediction and prevention of suboptimal health status in youths. An online survey based on the nationwide COVID-19 Impact on Lifestyle Change Survey (COINLICS) was conducted in China, and 10,082 youths reported their screen use, PA, and SC in the months immediately before, during, and after the COVID-19 lockdown. Cross-sectional and longitudinal network models were used to identify associations of variables in domains of screen use with PA and SC. Effect modifications of bridges and predictors in the associations were examined. The network models suggested that individual SC was a bridge that strongly connected other types of SC, and domains of screen use and PA before lockdown, while phone use became such a bridge during and after lockdown. More PC/TV use before lockdown predicted less household-related PA during lockdown (β = -0.142); more phone use during lockdown was a predictor for higher levels of household-related PA (β = 0.106), active transport (β = 0.096), and individual SC (β = 0.072) after lockdown. Phone use was negatively associated with PA through PC/TV use in the more phone use subgroup. Relationships among screen use, PA, and SC dynamically changed during COVID-19, and phone use that was identified as a bridge and a predictor may be the potential action point for health intervention in youths during lockdown.

Green space access in the neighbourhood and childhood obesity

Access to green space may influence individual physical activity (PA) and subsequently weight status, as increased exposure to green space could improve health by increasing opportunities and the actual levels of PA. However, whether such associations hold empirically remains inconclusive. This study reviewed articles that analysed the association between access to green space and weight-related behaviours/outcomes among children, published before 1 January 2019. The sample sizes ranged from 108 to 44 278. Four cohorts and 17 cross-sectional studies conducted in nine countries were identified. Overall, evidence showed a positive association between access to green space and PA and a negative association between access to green space and television-watching time, body mass index (BMI) and weight status among children. Distance to the nearest green space, measured by geographic information system (GIS) in 10 studies, was often used to represent access to the nearest green space. It still remains difficult to draw a clear conclusion on the association between access to green space and BMI. Longitudinal studies can directly estimate the strength of the association between exposure and disease, which is needed to determine the causal association between access to green space and weight status.

Natural environment and childhood obesity: A systematic review

The associations between built and food environments and childhood obesity have been studied extensively. However, the association between the natural environment and childhood obesity has received too little scholarly attention. This study reviewed the literature published before 1 January 2019, which described associations between a full range of natural environmental factors (e.g., rainfall, temperature, sunlight, natural disasters, flood and drought) and weight-related behaviours and childhood obesity. Five cross-sectional studies and one longitudinal study were identified. Measures of natural environmental factors varied across six included studies, falling into five broad categories: weather conditions, altitude, natural disaster risk, air quality and day length. It was found that temperature was a significant weather indicator in most included studies and was associated with a reduction of daily physical activity. Children living in high-altitude areas were more likely to be shorter and heavier than their counterparts in low-altitude areas. Findings of this study will contribute to helping multiple stakeholders, including policy makers and urban planners, and formulate health policies and interventions to mitigate the detrimental impact of the natural environment on childhood obesity. More longitudinal studies should be designed to confirm these effects and explore the potential health effects of more natural environmental factors.

Mapping Fine-Scale Urban Spatial Population Distribution Based on High-Resolution Stereo Pair Images, Points of Interest, and Land Cover Data

Fine-scale population distribution is increasingly becoming a research hotspot owing to its high demand in many applied fields. It is of great significance in urban emergency response, disaster assessment, resource allocation, urban planning, market research, and transportation route design. This study employed land cover, building address, and housing price data, and high-resolution stereo pair remote sensing images to simulate fine-scale urban population distribution. We firstly extracted the residential zones on the basis of land cover and Google Earth data, combined them with building information including address and price. Then, we employed the stereo pair analysis method to obtain the building height on the basis of ZY3-02 high-resolution satellite data and transform the building height into building floors. After that, we built a sophisticated, high spatial resolution model of population density. Finally, we evaluated the accuracy of the model using the survey data from 12 communities in the study area. Results demonstrated that the proposed model for spatial fine-scale urban population products yielded more accurate small-area population estimation relative to high-resolution gridded population surface (HGPS). The approach proposed in this study holds potential to improve the precision and automation of high-resolution population estimation.