Neuroimaging was completed as part of the Philadelphia Neurodevelopmental Cohort (46 (link)). All participants, or their parent or guardian, provided informed consent, and minors provided assent; study procedures were approved by the institutional review boards of both the University of Pennsylvania and the Children’s Hospital of Philadelphia. All participants included in this study were medically healthy, were not taking psychotropic medication at the time of study, and passed strict quality-assurance procedures for four imaging modalities including T1-weighted structural images, diffusion-weighted imaging, resting-state functional MRI (fMRI), and n-back fMRI. The final sample included 727 youths ages 8 to 23 y (420 females; mean = 15.9 y, SD = 3.2 y). From the original study sample, 147 typically developing youths returned for longitudinal neuroimaging assessments ∼1.7 y after baseline (83 females; 294 total scans). For further details regarding image preprocessing and brain network construction see SI Appendix, SI Methods .
To evaluate the relationship between structure–function coupling and previously characterized cortical hierarchies, evolutionary cortical areal expansion (3 (link)) and the principal gradient of intrinsic functional connectivity (2 (link)) were extracted from publicly available atlases. The significance of the spatial correspondence between brain maps was estimated using a conservative spatial permutation test, which generates a null distribution of randomly rotated brain maps that preserve spatial covariance structure of the original data (23 (link)).
We used penalized splines within a GAM to estimate linear and nonlinear age-related changes in structure–function coupling for each brain region. Importantly, the GAM estimates nonlinearities using restricted maximum likelihood, penalizing nonlinearity in order to avoid overfitting the data (47 ). To evaluate regional associations between structure–function coupling and executive function, executive performance was measured as a factor score summarizing accuracy across mental flexibility, attention, working memory, verbal reasoning, and spatial ability tasks administered as part of the Penn Computerized Neurocognitive Battery (SI Appendix, SI Methods ).
Longitudinal developmental change in structure–function coupling was evaluated with two approaches. First, we estimated longitudinal age effects on coupling within a linear mixed effects model, including a random subject intercept in addition to other covariates. Second, we used linear regression models with longitudinal change scores. Longitudinal intraindividual change in coupling (ΔCoupling) and the participation coefficient (ΔPC) were calculated as the difference in regional brain measures between time points. Baseline age, sex, mean relative framewise displacement, and the number of years between time points were included as additional covariates in linear regression models.
The data reported in this paper have been deposited in the database of Genotypes and Phenotypes under accession number dbGaP: phs000607.v2.p2 (https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000607.v2.p2 ).
To evaluate the relationship between structure–function coupling and previously characterized cortical hierarchies, evolutionary cortical areal expansion (3 (link)) and the principal gradient of intrinsic functional connectivity (2 (link)) were extracted from publicly available atlases. The significance of the spatial correspondence between brain maps was estimated using a conservative spatial permutation test, which generates a null distribution of randomly rotated brain maps that preserve spatial covariance structure of the original data (23 (link)).
We used penalized splines within a GAM to estimate linear and nonlinear age-related changes in structure–function coupling for each brain region. Importantly, the GAM estimates nonlinearities using restricted maximum likelihood, penalizing nonlinearity in order to avoid overfitting the data (47 ). To evaluate regional associations between structure–function coupling and executive function, executive performance was measured as a factor score summarizing accuracy across mental flexibility, attention, working memory, verbal reasoning, and spatial ability tasks administered as part of the Penn Computerized Neurocognitive Battery (
Longitudinal developmental change in structure–function coupling was evaluated with two approaches. First, we estimated longitudinal age effects on coupling within a linear mixed effects model, including a random subject intercept in addition to other covariates. Second, we used linear regression models with longitudinal change scores. Longitudinal intraindividual change in coupling (ΔCoupling) and the participation coefficient (ΔPC) were calculated as the difference in regional brain measures between time points. Baseline age, sex, mean relative framewise displacement, and the number of years between time points were included as additional covariates in linear regression models.
The data reported in this paper have been deposited in the database of Genotypes and Phenotypes under accession number dbGaP: phs000607.v2.p2 (
