From associations and predictions to perturbations and mechanisms: Commentary on “Longer scans boost prediction and cut costs in brain-wide association studies”

Brain-wide association studies (BWAS) are a research tool for discovering reproducible brain–phenotype associations but require massive sample sizes to achieve adequate power. Ooi and colleagues recently demonstrated that longer fMRI scans in fewer participants can outperform short scans in large samples. While acknowledging the practical value of this approach, we highlight three critical considerations: population representativeness, cohort-specific feasibility constraints, and the distinction between predictive optimization and mechanistic understanding. We also outline complementary strategies that can enhance predictive power without solely relying on increased scan duration.