Neuroimaging Growth Curves in Children and Young Adults Aged 6–20 Years in the Netherlands: a Population-Based Cohort Substudy of the Generation R Study

Background Current paediatric neuroradiology relies largely on visual inspection and semiquantitative scoring, limiting detection of subtle abnormalities in the developing brain. The present study aimed to develop population-based neuroimaging growth curves for clinical use to assess paediatric neurodevelopmental trajectories and identify deviations from typical brain maturation. Methods The study was conducted within the Generation R Study, a prospective, population-based birth cohort study based in Rotterdam, the Netherlands. A neuroimaging substudy consisted of four neuroimaging waves: 6·1–10·7 years (2010–13), 8·6–12·0 years (2013–15), 12·5–17·1 years (2016–20), and 16·2–20·4 years (2022–23). To assess clinical applicability, percentile curves were calculated for three clinical cohorts. Exclusion criteria included general contraindications for MRI, severe neurological disorders, significant incidental findings, and incomplete or poor-quality scans. Brain volumes were measured using FreeSurfer, whereas microstructural markers, fractional anisotropy, and mean diffusivity, were processed using FMRIB Software Library. Sex-specific neuroimaging growth curves were constructed using quantile regression analyses. Participants were stratified by European and non-European backgrounds. Findings 7316 brain MRI scans from 4551 participants (3587 [49·0%] males, 3729 [51·0%] females) were analysed: 6·1–10·7 years (n=921), 8·6–12·0 years (n=3135), 12·5–17·1 years (n=2118), and 16·2–20·4 years (n=1142). Clinical cohorts included individuals with very preterm birth (VPB; n=32), multiple sclerosis (MS; n=27), and sickle cell disease (SCD; n=18). Total brain volume followed an inverted U-shaped trajectory, peaking at 13–15 years, with subsequent decline attributable mainly to grey matter reduction, while white matter volume increased until late adolescence. Cortical and deep grey matter volumes showed nonlinear developmental patterns, characterised by early increases that reached region-specific peaks. Regional patterns were consistent with these trends, with slightly larger volumes observed in males than in females. Microstructural markers showed increasing fractional anisotropy and decreasing mean diffusivity with age. Clinical cohorts showed notable deviations from normative curves. Interpretation The developed growth curves provide a quantitative reference framework for paediatric brain development and identify subtle but clinically relevant deviations in specific patient groups, and have the potential to enhance diagnostic accuracy, advance understanding of pathophysiological mechanisms, and guide therapeutic strategies. Funding Erasmus Medical Center; Erasmus University Rotterdam; the Netherlands Organization for Health Research and Development (ZonMw); the Netherlands Organization for Scientific Research (NWO); the Ministry of Health, Welfare and Sport; Hartstichting; Intramural Research Program of the National Institute of Mental Health.