[Secular trends of overweight and obesity prevalence between 2007 and 2011 in children and adolescents in Guangzhou].

OBJECTIVE: This study aimed to explore the secular trends of overweight and obesity prevalence between 2007 and 2011 in children and adolescents in Guangzhou. METHODS: The data of physical examination was collected from the routine measurements carried out by the Health Care Facilities of Primary and Secondary schools between 2007 and 2011. Random stratified cluster sampling was conducted, all the students aged 5-18 years old form 19 primary and secondary schools from 4 districts (Tianhe district, Yuexiu district, Baiyun district and Haizhu district) were included in this survey, including 27 944 students in 2007 and 38 284 students in 2011. Body mass index reference norm established by Working Group on Obesity in China (WGOC) and the WHO cut-off criteria were employed for overweight and obesity screening, and the trend was analyzed. RESULTS: Prevalence of obesity in children and adolescents (7-18 years old) significantly increased from 5.96% (1 553/26 055) in 2007 to 6.56% (2 339/35 664) in 2011, and the difference showed statistical significance (χ(2) = 9.195, P < 0.05). Overweight and obesity was more common in boys (overweight: 13.25% (1 766/13 329) in 2007 and 13.87% (2 559/18 451) in 2011; obesity: 7.82% (1 042/13 329) in 2007 and 8.63% (1 592/18 451) in 2011) than in girls (overweight: 7.43% (946/12 726) in 2007 and 8.17% (1 406/17 213) in 2011; obesity: 4.11% (523/12 726) in 2007 and 4.48% (771/17 213) in 2011), and the difference showed statistical significance (overweight:χ(2) = 236.123 in 2007 and χ(2) = 292.892 in 2011; obesity:χ(2) = 158.533 in 2007 and χ(2) = 247.794 in 2011. All P values < 0.05). Further analysis found that significant increases occurred in boys aged 16 and 17 years old and in girls aged 12 years old (boy: 16 years old,χ(2) = 6.820, P < 0.05. 17 years old, χ(2) = 4.893, P < 0.05. girl: 12 years old,χ(2) = 5.921, P < 0.05). RESULTS: of Join-point regression showed that for boys less than 10 years old the prevalence increased with age increasing (in 2007, APC = 3.75; in 2011, APC = 1.76), while over 10 years old the prevalence decreased with age increasing (in 2007, 10-18 years old's APC = -18.58; in 2011, 10-18 years old's APC = -15.95). While for girls the prevalence of obesity increased with age increasing between 7-9 years old (APC = 12.16), decreased with age increasing through 9 to 18 years old (APC = -17.23) in 2007. The prevalence decreased with age increasing for girls in 2011 (APC = -4.66). CONCLUSION: The prevalence of obesity is high and still increasing in children and adolescents in Guangzhou, and it is higher in boys than in girls. It is more likely to become obesity at 10 years for boys, and for girls the prevalence decrease with age increasing.

SP-GNN: Learning structure and position information from graphs.

Graph neural network (GNN) is a powerful model for learning from graph data. However, existing GNNs may have limited expressive power, especially in terms of capturing adequate structural and positional information of input graphs. Structure properties and node position information are unique to graph-structured data, but few GNNs are capable of capturing them. This paper proposes Structure- and Position-aware Graph Neural Networks (SP-GNN), a new class of GNNs offering generic and expressive power of graph data. SP-GNN enhances the expressive power of GNN architectures by incorporating a near-isometric proximity-aware position encoder and a scalable structure encoder. Further, given a GNN learning task, SP-GNN can be used to analyze positional and structural awareness of GNN tasks using the corresponding embeddings computed by the encoders. The awareness scores can guide fusion strategies of the extracted positional and structural information with raw features for better performance of GNNs on downstream tasks. We conduct extensive experiments using SP-GNN on various graph datasets and observe significant improvement in classification over existing GNN models.