Search

Abstract Background Adolescent weight problems have become a growing public health concern, making early prediction of non-normal weight status crucial for effective prevention. However, few temporal prediction tools for adolescent four weight status have been developed. This study aimed to predict the short- and long-term weight status of Hong Kong adolescents and assess the importance of predictors. Methods A population-based retrospective cohort study of adolescents was conducted using data from a territory-wide voluntary annual health assessment service provided by the Department of Health in Hong Kong. Using diet habits, physical activity, psychological well-being, and demographics, we generated six prediction models for successive weight status (normal, overweight, obese and underweight) using multiclass Decision Tree, Random Forest, k-Nearest Neighbor, eXtreme gradient boosting, support vector machine, logistic regression. Model performance was evaluated by multiple standard classifier metrics and the overall accuracy. Predictors’ importance was assessed using Shapley values. Results 442,898 Primary 4 (P4, Grade 4 in the US) and 344,186 in Primary 6 (P6, Grade 6 in the US) students, with followed up until their Secondary 6 (Grade 12 in the US) during the academic years 1995/96 to 2014/15 were included. The XG Boosts model consistently outperformed all other model in predicting the long-term weight status at S6 from P4 or P6. It achieved an overall accuracy of 0.72 or 0.74, a micro-averaging AUC of 0.92 or 0.93, and a macro-averaging AUC of 0.83 or 0.86, respectively. XG Boost also demonstrated accurate predictions for each predicted weight status, surpassing the AUC values obtained by other models. Weight, height, sex, age, frequency and hours of aerobic exercise were consistently the most important predictors for both cohorts. Conclusions The machine learning approaches accurately predict adolescent weight status in both short- and long-term. The developed multiclass model that utilizing easy-assessed variables enables accurate long-term prediction on weight status, which can be used by adolescents and parents for self-prediction when applied in health care system. The interpretable models may help to provide the early and individualized interventions suggestions for adolescents with weight problems particularly.

BackgroundEarly identification of high-risk individuals for weight problems in children and adolescents is crucial for implementing timely preventive measures. While machine learning (ML) techniques have shown promise in addressing this complex challenge with high-dimensional data, feature selection is vital for identifying the key predictors that can facilitate effective and targeted interventions. This study aims to utilize feature selection process to identify a robust and minimal set of predictors that can aid in the early prediction of short- and long-term weight problems in children and adolescents.MethodsWe utilized demographic, physical, and psychological wellbeing predictors to model weight status (normal, underweight, overweight, and obese) for 1-, 3-, and 5-year periods. To select the most influential features, we employed four feature selection methods: (1) Chi-Square test; (2) Information Gain; (3) Random Forest; (4) eXtreme Gradient Boosting (XGBoost) with six ML approaches. The stability of the feature selection methods was assessed by Jaccard's index, Spearman's rank correlation and Pearson's correlation. Model evaluation was performed by various accuracy metrics.ResultsWith 3,862,820 million student-visits were included in this population-based study, the mean age of 11.6 (SD = 3.64) for the training set and 10.8 years (SD = 3.50) for the temporal test set. From the initial set of 38 predictors, we identified 6, 9, and 13 features for 1-, 3-, and 5-year predictions, respectively, by the best performed feature selection method of Chi-Square test in XGBoost models. These feature sets demonstrated excellent stability and achieved prediction accuracies of 0.82, 0.73, and 0.70; macro-AUCs of 0.94, 0.86, and 0.83; micro-AUCs of 0.96, 0.93, and 0.92 for different prediction windows, respectively. Weight, height, sex, total score of self-esteem, and age were consistently the most influential predictors across all prediction windows. Additionally, several psychological and social wellbeing predictors showed relatively high importance in long-term weight status prediction.ConclusionsWe demonstrate the potential of ML in identifying key predictors of weight status in children and adolescents. While traditional anthropometric measures remain important, psychological and social wellbeing factors also emerge as crucial predictors, potentially informing targeted interventions to address childhood and adolescence weight problems.

Abstract Background How weight status changes with time may affect self-esteem was seldom studied. We identified the distinct trajectories of overweight/obesity over age and assessed their associations with different domains of self-esteem in Hong Kong Chinese children. Methods Territory-wide longitudinal data of 48,558 children (girls: 50.0%; 6.3 ± 0.51 years) studying Primary 1 in the academic cohorts of 1995/96 and 1996/97 followed till Primary 6 were obtained from the Student Health Service of Hong Kong. Weight was annually measured and categorized as underweight/normal and overweight/obese and self-esteem was measured in Primary 6. Distinct trajectories of weight status were first identified using growth mixture modeling and their associations with low self-esteem were assessed by logistic regression. Results Four distinct overweight/obesity trajectories were identified: never (76.8%), late-onset (8.1%), early-onset (4.2%) and chronic (10.9%) overweight/obesity. Compared with children who were never overweight/obese, more of those in the late-onset or chronic overweight/obesity group showed low self-esteem and specific domains including general, social and academic/school-related (adjusted odds ratios: 1.20 - 1.43, all P < 0.001) except parent/home-related self-esteem (P = 0.36), whereas children being in the early-onset overweight/obesity group showed no significant difference (P ≥ 0.53) except a lower risk of low social self-esteem (adjusted odds ratio = 0.82, P = 0.03). Conclusions Late-onset or chronic overweight/obesity predicted low general, social and academic/school-related self-esteem. Children who successfully reduced weight may have equal levels of self-esteem or even better social self-esteem than those being always underweight/normal weight. Overweight/obese children had a vulnerability to self-esteem in non-domestic environments.

Background: Breakfast is deemed the most important meal of the day. We examined the prospective associations of breakfast habits with emotional/behavioral problems in adolescents and potential effect modification. Methods: 115,217 Primary 6 students (United States Grade 6; mean age, 11.9; standard deviation [SD], 0.59 years) who attended the Student Health Service of Department of Health in Hong Kong in 2004/05, 2006/07, 2008/09 were followed till Secondary 6 (United States Grade 12). Emotional/behavioral problems were biennially examined using Youth Self-Report since Secondary 2 (United States Grade 8). Lifestyles were biennially examined using standardized questionnaires since Primary 6. Prospective associations of breakfast habit with emotional/behavioral problems and potential effect modification were examined using generalized estimating equations. Results: Compared with eating breakfast at home, eating breakfast away from home was significantly associated with total emotional/behavioral problems and seven syndromes, including withdrawal, somatic complaints, anxiety/depression, thought problems, attention problems, delinquent behaviors, and aggressive behaviors (adjusted odds ratios [AORs] 1.22–2.04), while skipping breakfast showed stronger associations with the above problems and social problems (AORs 1.34–2.29). Stronger associations were observed in younger students for total and attention problems (P < 0.03) and in those with lower weight status for delinquent behaviors (P = 0.005). Conclusion: Eating breakfast away from home and especially skipping breakfast were prospectively associated with adolescent emotional/behavioral problems. The associations weakened with increasing age for total emotional/behavioral and attention problems, and weakened with higher weight status for delinquent behaviors, highlighting the vulnerability of younger and underweight children. If the associations are causal, increasing home breakfast may reduce adolescent emotional/behavioral problems and benefit psychosocial health.

Abstract Background Sedentary behaviors are prevalent in Chinese children, however, the studies on their trends and socioeconomic disparities are scarce. We examined the time trends of daily television (TV) viewing and video game playing and the associated socioeconomic factors in Chinese children in Hong Kong, the most developed and westernized city in China. Methods In a panel data study involving 538,300 primary four and 510,294 primary six students from 1999/2000 to 2008/09, data on socioeconomic status, sedentary behaviors (TV viewing and video game playing) and other lifestyle habits were collected using a self-administered questionnaire. Trends in sedentary behaviors over time were assessed. Their socioeconomic disparities were examined by interactions in generalized estimating equations with the adjustment for weight status and extracurricular physical activities. Results The age and sex-standardized prevalence of ≥2 h daily TV viewing decreased from 51.4% (95% confidence interval [CI] 51.1–51.8%) in 1999/2000 to 43.8% (95% CI 43.4–44.2%) in 2008/09 (P for trend