Temporal graph model predicts teen intelligence from wearable data
Temporal Graph Learning of Wearable Actigraphy and Sleep Traces for Modelling Adolescent Crystallized Intelligence
Artificial Intelligence
Summary
Measuring a teen’s intelligence over time using wearable devices is hard because the data can be inconsistent and complex. The authors created SATURN, a computer model that looks at daily activity and sleep patterns recorded by Fitbits over 21 days, treating each day as a connected point in a graph. This model predicts teens’ verbal and knowledge-based intelligence better than standard methods. Key factors the model found important include light activity, energy use, and how long teens sleep.
What this means in practice
- •For healthcare data analysts: Use wearable activity and sleep data to improve identification of cognitive development trends in adolescents without invasive testing.
- •For public health program teams: Design large-scale adolescent health monitoring programs incorporating temporal graph models to track behavioral factors linked with intelligence more accurately.
Tested on one dataset.
Authors
Md. Tanvir Rahman, Nabil Anan Orka, Asaduzzaman Khan, Mohammad Ali Moni
Abstract
Wearable actigraphy offers a scalable, ecologically valid alternative to episodic clinical assessment. However, predicting continuous adolescent crystallized intelligence ($G_c$) from such traces remains challenging due to irregular device adherence and complex behavioral-environmental interactions. We address this using daily summary data derived from 21-day Fitbit records of 6,091 adolescents in the Adolescent Brain Cognitive Development Study (Release 5.1). We propose SATURN, a Sleep-Activity Temporal Unified Regression Network. It represents participants as 21-node temporal graphs encoding daily behaviors and temporal adjacency. To prevent imputation artifacts, invalid-day edges are dynamically pruned during forward passes. Node embeddings are refined via residual GATv2 layers, aggregated through masked attention pooling, and fused with sociodemographic covariates. Under family-controlled, age-sex-BMI-stratified cross-validation, SATURN achieves $R^2 = 0.2783 \pm 0.0127$, consistently improving upon flattened machine learning (Gradient Boosting, $R^2 = 0.2372$) and sequential deep learning (BiLSTM, $R^2 = 0.2688$) baselines. Explainability analyses identify light activity, metabolic equivalents, and sleep duration as dominant predictors, while Monte Carlo dropout and subgroup analyses confirm equitable performance across sociodemographic strata. Ultimately, SATURN establishes a rigorous computational framework for digital cognitive phenotyping, offering a scalable pathway to complement traditional assessments by highlighting macro-level behavioral anomalies.