Dynamic prediction model to identify young children at high risk of future overweight: Development and internal validation in a cohort study

Predicting the onset of overweight in Chinese high school students: a machine-learning approach in a one-year prospective cohort study

OBJECTIVE

This study aimed to develop and evaluate machine-learning models for predicting the onset of overweight in adolescents aged 14‒17, utilizing easily collectible personal information.

METHODS

This study was a one-year prospective cohort study. Baseline data were collected through anthropometric measurements and questionnaires, and the incidence of overweight was calculated one year later via anthropometric measurements. Predictive factors were selected through univariate analysis. Six machine-learning models were developed for predicting the onset of overweight. The SHapley Additive exPlanations (SHAP) was used for global and local interpretation of the models.

RESULTS

Out of 1,241 adolescents, 204 (16.4%) were identified as overweight after one year. Nineteen features were associated with the overweight incidence in univariable analysis. Participants were randomly divided into a training group and a testing group in a 7:3 ratio. The Light Gradient Boosting Machine (LGBM) algorithm achieved outperformed other models, achieving the following metrics: Accuracy (0.956), Recall (0.812), Specificity (0.983), F1-score (0.855), AUC (0.961). Importance ranking revealed that the top 11 minimal feature set can maintain the stability of model performance.

CONCLUSIONS

The onset of overweight in adolescents was accurately predicted using easily collectible personal information. The LGBM-based model exhibited superior performance. Oversampling technique notably improved model performance. The model interpretation technique provided innovative strategies for managing adolescent overweight/obesity.

Prediction models for children/adolescents with obesity/overweight: A systematic review and meta-analysis

The incidence of obesity and overweight in children and adolescents is increasing worldwide and becomes a global health concern. This study aims to evaluate the accuracy of available prediction models in early identification of obesity and overweight in general children or adolescents and identify predictive factors for the models, thus provide a reference for subsequent development of risk prediction tools for obesity and overweight in children or adolescents. Related publications were obtained from several databases such as PubMed, Embase, Cochrane Library, and Web of Science from their inception to September 18th, 2022. The novel Prediction Model Risk of Bias Assessment Tool (PROBAST) was employed to assess the bias risk of the included studies. R4.2.0 and Stata15.1 softwares were used to conduct meta-analysis. This study involved 45 cross-sectional and/or prospective studies with 126 models. Meta-analyses showed that the overall pooled index of concordance (c-index) of prediction models for children/adolescents with obesity and overweight in the training set was 0.769 (95% CI 0.754-0.785) and 0.835(95% CI 0.792-0.879), respectively. Additionally, a large number of predictors were found to be related to children's lifestyles, such as sleep duration, sleep quality, and eating speed. In conclusions, prediction models can be employed to predict obesity/overweight in children and adolescents. Most predictors are controllable factors and are associated with lifestyle. Therefore, the prediction model serves as an excellent tool to formulate effective strategies for combating obesity/overweight in pediatric patients.

Prediction of childhood overweight and obesity at age 10-11: findings from the Studying Lifecourse Obesity PrEdictors and the Born in Bradford cohorts

BACKGROUND

In England, 41% of children aged 10-11 years live with overweight or obesity. Identifying children at risk of developing overweight or obesity may help target early prevention interventions. We aimed to develop and externally validate prediction models of childhood overweight and obesity at age 10-11 years using routinely collected weight and height measurements at age 4-5 years and maternal and early-life health data.

METHODS

We used an anonymised linked cohort of maternal pregnancy and birth health records in Hampshire, UK between 2003 and 2008 and child health records. Childhood body mass index (BMI), adjusted for age and sex, at 10-11 years was used to define the outcome of overweight and obesity (BMI ≥ 91st centile) in the models. Logistic regression models and multivariable fractional polynomials were used to select model predictors and to identify transformations of continuous predictors that best predict the outcome. Models were externally validated using data from the Born in Bradford birth cohort. Model performance was assessed using discrimination and calibration.

RESULTS

Childhood BMI was available for 6566 children at 4-5 (14.6% overweight) and 10-11 years (26.1% overweight) with 10.8% overweight at both timepoints. The area under the curve (AUC) was 0.82 at development and 0.83 on external validation for the model only incorporating two predictors: BMI at 4-5 years and child sex. AUC increased to 0.84 on development and 0.85 on external validation on additionally incorporating maternal predictors in early pregnancy (BMI, smoking, age, educational attainment, ethnicity, parity, employment status). Models were well calibrated.

CONCLUSIONS

This prediction modelling can be applied at 4-5 years to identify the risk for childhood overweight at 10-11 years, with slightly improved prediction with the inclusion of maternal data. These prediction models demonstrate that routinely collected data can be used to target early preventive interventions to reduce the prevalence of childhood obesity.

Trans-ethnic polygenic risk scores for body mass index: An international hundred K+ cohorts consortium study

BACKGROUND

While polygenic risk scores hold significant promise in estimating an individual's risk of developing a complex trait such as obesity, their application in the clinic has, to date, been limited by a lack of data from non-European populations. As a collaboration model of the International Hundred K+ Cohorts Consortium (IHCC), we endeavored to develop a globally applicable trans-ethnic PRS for body mass index (BMI) through this relatively new international effort.

METHODS

The polygenic risk score (PRS) model was developed, trained and tested at the Center for Applied Genomics (CAG) of The Children's Hospital of Philadelphia (CHOP) based on a BMI meta-analysis from the GIANT consortium. The validated PRS models were subsequently disseminated to the participating sites. Scores were generated by each site locally on their cohorts and summary statistics returned to CAG for final analysis.

RESULTS

We show that in the absence of a well powered trans-ethnic GWAS from which to derive marker SNPs and effect estimates for PRS, trans-ethnic scores can be generated from European ancestry GWAS using Bayesian approaches such as LDpred, by adjusting the summary statistics using trans-ethnic linkage disequilibrium reference panels. The ported trans-ethnic scores outperform population specific-PRS across all non-European ancestry populations investigated including East Asians and three-way admixed Brazilian cohort.

CONCLUSIONS

Here we show that for a truly polygenic trait such as BMI adjusting the summary statistics of a well powered European ancestry study using trans-ethnic LD reference results in a score that is predictive across a range of ancestries including East Asians and three-way admixed Brazilians.

Validity of Scottish predictors of child obesity (age 12) for risk screening in mid-childhood: a secondary analysis of prospective cohort study data-with sensitivity analyses for settings without various routinely collected predictor variables

OBJECTIVE

To analyse the Growing Up in Scotland cohort for predictors of obesity at age 12, present at school entry (age 5-6).

METHODS

The initial model included literature-based risk factors likely to be routinely collected in high-income countries (HICs), as well as "Adverse/Protective Childhood Experiences (ACEs/PCEs)". Missing data were handled by Multiple Chained Equations. Variable-reduction was performed using multivariable logistic regression with backwards and forwards stepwise elimination, followed by internal validation by bootstrapping. Optimal sensitivity/specificity cut-offs for the most parsimonious and accurate models in two situations (optimum available data, and routinely available data in Scotland) were examined for their referral burden, and Positive and Negative Predictive Values.

RESULTS

Data for 2787 children with full outcome data (obesity prevalence 18.3% at age 12) were used to develop the models. The final "Optimum Data" model included six predictors of obesity: maternal body mass index, indoor smoking, equivalized income quintile, child's sex, child's BMI at age 5-6, and ACEs. After internal validation, the area under the receiver operating characteristic curve was 0.855 (95% CI 0.852-0.859). A cut-off based on Youden's J statistic for the Optimum Data model yielded a specificity of 77.6% and sensitivity of 76.3%. 37.0% of screened children were "Total Screen Positives" (and thus would constitute the "referral burden".) A "Scottish Data" model, without equivalized income quintile and ACEs as a predictor, and instead using Scottish Index of Multiple Deprivation quintile and "age at introduction of solid foods," was slightly less sensitive (76.2%) but slightly more specific (79.2%), leading to a smaller referral burden (30.8%).

CONCLUSION

Universally collected, machine readable and linkable data at age 5-6 predict reasonably well children who will be obese by age 12. However, the Scottish treatment system is unable to cope with the resultant referral burden and other criteria for screening would have to be met.

Predicting the earliest deviation in weight gain in the course towards manifest overweight in offspring exposed to obesity in pregnancy: a longitudinal cohort study

BACKGROUND

Obesity in pregnancy and related early-life factors place the offspring at the highest risk of being overweight. Despite convincing evidence on these associations, there is an unmet public health need to identify "high-risk" offspring by predicting very early deviations in weight gain patterns as a subclinical stage towards overweight. However, data and methods for individual risk prediction are lacking. We aimed to identify those infants exposed to obesity in pregnancy at ages 3 months, 1 year, and 2 years who likely will follow a higher-than-normal body mass index (BMI) growth trajectory towards manifest overweight by developing an early-risk quantification system.

METHODS

This study uses data from the prospective mother-child cohort study Programming of Enhanced Adiposity Risk in CHildhood-Early Screening (PEACHES) comprising 1671 mothers with pre-conception obesity and without (controls) and their offspring. Exposures were pre- and postnatal risks documented in patient-held maternal and child health records. The main outcome was a "higher-than-normal BMI growth pattern" preceding overweight, defined as BMI z-score >1 SD (i.e., World Health Organization [WHO] cut-off "at risk of overweight") at least twice during consecutive offspring growth periods between age 6 months and 5 years. The independent cohort PErinatal Prevention of Obesity (PEPO) comprising 11,730 mother-child pairs recruited close to school entry (around age 6 years) was available for data validation. Cluster analysis and sequential prediction modelling were performed.

RESULTS

Data of 1557 PEACHES mother-child pairs and the validation cohort were analyzed comprising more than 50,000 offspring BMI measurements. More than 1-in-5 offspring exposed to obesity in pregnancy belonged to an upper BMI z-score cluster as a distinct pattern of BMI development (above the cut-off of 1 SD) from the first months of life onwards resulting in preschool overweight/obesity (age 5 years: odds ratio [OR] 16.13; 95% confidence interval [CI] 9.98-26.05). Contributing early-life factors including excessive weight gain (OR 2.08; 95% CI 1.25-3.45) and smoking (OR 1.94; 95% CI 1.27-2.95) in pregnancy were instrumental in predicting a "higher-than-normal BMI growth pattern" at age 3 months and re-evaluating the risk at ages 1 year and 2 years (area under the receiver operating characteristic [AUROC] 0.69-0.79, sensitivity 70.7-76.0%, specificity 64.7-78.1%). External validation of prediction models demonstrated adequate predictive performances.

CONCLUSIONS

We devised a novel sequential strategy of individual prediction and re-evaluation of a higher-than-normal weight gain in "high-risk" infants well before developing overweight to guide decision-making. The strategy holds promise to elaborate interventions in an early preventive manner for integration in systems of well-child care.

Development of a Prediction Model to Identify Children at Risk of Future Developmental Delay at Age 4 in a Population-Based Setting

Our aim was to develop a prediction model for infants from the general population, with easily obtainable predictors, that accurately predicts risk of future developmental delay at age 4 and then assess its performance. Longitudinal cohort data were used (N = 1983), including full-term and preterm children. Development at age 4 was assessed using the Ages and Stages Questionnaire. Candidate predictors included perinatal and parental factors as well as growth and developmental milestones during the first two years. We applied multiple logistic regression with backwards selection and internal validation, and we assessed calibration and discriminative performance (i.e., area under the curve (AUC)). The model was evaluated in terms of sensitivity and specificity at several cut-off values. The final model included sex, maternal educational level, pre-existing maternal obesity, several milestones (smiling, speaking 2–3 word sentences, standing) and weight for height z score at age 1. The fit was good, and the discriminative performance was high (AUC: 0.837). Sensitivity and specificity were 73% and 80% at a cut-off probability of 10%. Our model is promising for use as a prediction tool in community-based settings. It could aid to identify infants in early life (age 2) with increased risk of future developmental problems at age 4 that may benefit from early interventions.

Dynamic prediction model to identify young children at high risk of future overweight: Development and internal validation in a cohort study

BACKGROUND

Primary prevention of overweight is to be preferred above secondary prevention, which has shown moderate effectiveness.

OBJECTIVE

To develop and internally validate a dynamic prediction model to identify young children in the general population, applicable at every age between birth and age 6, at high risk of future overweight (age 8).

METHODS

Data were used from the Prevention and Incidence of Asthma and Mite Allergy birth cohort, born in 1996 to 1997, in the Netherlands. Participants for whom data on the outcome overweight at age 8 and at least three body mass index SD scores (BMI SDS) at the age of ≥3 months and ≤6 years were available, were included (N = 2265). The outcome of the prediction model is overweight (yes/no) at age 8 (range 7.4-10.5 years), defined according to the sex- and age-specific BMI cut-offs of the International Obesity Task Force.

RESULTS

After backward selection in a Generalized Estimating Equations analysis, the prediction model included the baseline predictors maternal BMI, paternal BMI, paternal education, birthweight, sex, ethnicity and indoor smoke exposure; and the longitudinal predictors BMI SDS, and the linear and quadratic terms of the growth curve describing a child's BMI SDS development over time, as well as the longitudinal predictors' interactions with age. The area under the curve of the model after internal validation was 0.845 and Nagelkerke R² was 0.351.

CONCLUSIONS

A dynamic prediction model for overweight was developed with a good predictive ability using easily obtainable predictor information. External validation is needed to confirm that the model has potential for use in practice.