Predictive associations between brain functional connectivity, motor abilities, and executive function development in early childhood: a longitudinal machine learning study.

Participants were recruited through convenience sampling from kindergartens in a single province of China, limiting generalizability to other geographic, cultural, and socioeconomic contexts. Younger children were scanned during natural sleep while older children were scanned while watching videos, creating a major confound that makes it difficult to separate true brain maturation from state-dependent connectivity differences. The study was observational and predictive, so it cannot establish that brain connectivity causes motor or cognitive development. Prediction accuracies of 76 to 77% leave substantial unexplained variance, meaning environmental and genetic factors not captured in the model contribute meaningfully to development. The 6-minute scan duration and 4-millimeter slice thickness are practical compromises for pediatric imaging that may lower test-retest reliability compared to adult studies.

The text does not report funding sources or author conflicts of interest.

Do not interpret the predictive accuracy as evidence that brain scans can diagnose developmental delays in your clinic. The models achieved 76.8% accuracy in a research setting using specialized MRI protocols and machine learning pipelines that are not available for clinical use. Furthermore, the study compared children scanned in different states, meaning the apparent age-related brain changes may partly reflect sleep versus wakefulness rather than true development.