Characterizing the Relationship Between Cortical Gradients and Cognitive Traits in Children
Poster No:
1707
Submission Type:
Abstract Submission
Authors:
Mia Zwally1, Dustin Moraczewski1, Ka Chun Lam2, Adam Thomas1
Institutions:
1National Institute of Mental Health, Data Science and Sharing Team, Bethesda, MD, 2National Institute of Mental Health, Machine Learning Team, Bethesda, MD
First Author:
Co-Author(s):
Introduction:
Cortical gradients, or axes of variance in cortical structure, provide a novel perspective on understanding brain functional connectivity, and while previous work has reached a consensus that the primary gradient in adults displays a unimodal to transmodal axis (Margulies et al., 2016), comparatively less work has investigated the development of functional gradients and their associations with behavioral and cognitive measures. In a developmental sample, Dong et al. found that the primary and secondary gradients of children before the age of 12 were flipped compared to their adult counterparts, with the unimodal to transmodal gradient being the secondary contributor in children while ranking as primary in adults (Dong et al., 2021).
This study aims to replicate the findings of the flipped primary-secondary gradient within the larger (11,000 subjects) and more heterogeneous Adolescent Brain Cognitive Development dataset (ABCD) which was collected at 21 different sites across the United States. We also initiate the investigation of the complex, multidimensional relationship between functional gradient profiles and behavioral and cognitive traits in youth. Prior to beginning the analysis, we filed a pre-registration on OSF: https://doi.org/10.17605/OSF.IO/T9DHK
This study aims to replicate the findings of the flipped primary-secondary gradient within the larger (11,000 subjects) and more heterogeneous Adolescent Brain Cognitive Development dataset (ABCD) which was collected at 21 different sites across the United States. We also initiate the investigation of the complex, multidimensional relationship between functional gradient profiles and behavioral and cognitive traits in youth. Prior to beginning the analysis, we filed a pre-registration on OSF: https://doi.org/10.17605/OSF.IO/T9DHK
Methods:
All code used in the creation of this abstract is publicly available in this repository: https://github.com/MIZwally/gradients-and-behavior
This study uses the baseline resting state fMRI data from the ABCD dataset preprocessed with fMRIPrep 20.2.0 (Esteban et al., 2018) and 25 measures of behavior and cognition distributed in ABCD Annual Release 5. For comparisons with adult gradients, we used the Human Connectome Project (HCP) connectivity matrices distributed in the BrainSpace toolbox (Vos de Wael et al., 2020). After quality control, our ABCD sample consisted of 7,179 children. Using the 400-region, 17-network Schaefer atlas (Schaefer et al., 2018) we constructed a pairwise connectivity matrix for every participant, on which we calculated the top 10 gradients using the BrainSpace toolbox. Group gradients for both datasets were created from a connectivity matrix resulting from the average of all individual matrices.
To investigate the relationship between gradient profiles and cognitive traits, we first calculated individual differences from the group via the Spearman's correlation of each individual's gradient to the group gradient (Mckeown et al., 2020). We then conducted a sparse PCA on the behavior and cognition measures, retaining the top 8 components that account for 95% of the variance. The individual to group correlations and the PCA components were entered into a canonical correlation analysis (CCA). Confidence intervals for all tests were created using 1000 bootstrap samples.
This study uses the baseline resting state fMRI data from the ABCD dataset preprocessed with fMRIPrep 20.2.0 (Esteban et al., 2018) and 25 measures of behavior and cognition distributed in ABCD Annual Release 5. For comparisons with adult gradients, we used the Human Connectome Project (HCP) connectivity matrices distributed in the BrainSpace toolbox (Vos de Wael et al., 2020). After quality control, our ABCD sample consisted of 7,179 children. Using the 400-region, 17-network Schaefer atlas (Schaefer et al., 2018) we constructed a pairwise connectivity matrix for every participant, on which we calculated the top 10 gradients using the BrainSpace toolbox. Group gradients for both datasets were created from a connectivity matrix resulting from the average of all individual matrices.
To investigate the relationship between gradient profiles and cognitive traits, we first calculated individual differences from the group via the Spearman's correlation of each individual's gradient to the group gradient (Mckeown et al., 2020). We then conducted a sparse PCA on the behavior and cognition measures, retaining the top 8 components that account for 95% of the variance. The individual to group correlations and the PCA components were entered into a canonical correlation analysis (CCA). Confidence intervals for all tests were created using 1000 bootstrap samples.
Results:
Confirming previous findings, the correlation comparing the ABCD child primary and the HCP adult secondary was 0.418, 99% CI [0.416, 0.421], and the child secondary to adult primary 0.724, 99% CI [0.719, 0.728]. These primary-secondary correlations were greater than the primary-primary and secondary-secondary correlations of -0.011, 99% CI [-0.024, -0.001] and 0.168, 99% CI [0.162, 0.177], respectively. See Figures 1 and 2 for hemisphere visualizations of primary and secondary gradients from both ABCD and HCP samples.
In addition, we found that the primary mode from the CCA had a correlation value of 0.220, 95% CI [0.206 0.240], suggesting that there is indeed a complex, multidimensional relationship between gradient profiles and measures of behavior and cognition.
In addition, we found that the primary mode from the CCA had a correlation value of 0.220, 95% CI [0.206 0.240], suggesting that there is indeed a complex, multidimensional relationship between gradient profiles and measures of behavior and cognition.
Conclusions:
Our results show strong support for the flipping of the primary and secondary gradients in children through the replication of Dong et al. (2021) in a larger, more heterogeneous sample. In addition, we show that there is a complex, multidimensional relationship between functional gradient profiles and measures of behavior and cognition. Future work should focus on examining the nature of this relationship in greater detail.
Lifespan Development:
Normal Brain Development: Fetus to Adolescence 2
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 1
Task-Independent and Resting-State Analysis
Keywords:
Cognition
Development
FUNCTIONAL MRI
Open Data
Open-Source Code
Other - Gradients
1|2Indicates the priority used for review
Provide references using author date format
Esteban, O., Markiewicz, C. J., Blair, R. W., Moodie, C. A., Isik, A. I., Erramuzpe, A., Kent, J. D., Goncalves, M., DuPre, E., Snyder, M., Oya, H., Ghosh, S. S., Wright, J., Durnez, J., Poldrack, R. A., & Gorgolewski, K. J. (2018). fMRIPrep: A robust preprocessing pipeline for functional MRI. Nature Methods, 16(1), 111–116. https://doi.org/10.1038/s41592-018-0235-4
Goyal, N., Moraczewski, D., Bandettini, P. A., Finn, E. S., & Thomas, A. G. (2022). The positive–negative mode link between brain connectivity, demographics and behaviour: A pre-registered replication of Smith et al. (2015). Royal Society Open Science, 9(2). https://doi.org/10.1098/rsos.201090
Margulies, D. S., Ghosh, S. S., Goulas, A., Falkiewicz, M., Huntenburg, J. M., Langs, G., Bezgin, G., Eickhoff, S. B., Castellanos, F. X., Petrides, M., Jefferies, E., & Smallwood, J. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. Proceedings of the National Academy of Sciences of the United States of America, 113(44), 12574–12579. https://doi.org/10.1073/pnas.1608282113
Mckeown, B., Strawson, W. H., Wang, H.-T., Karapanagiotidis, T., Vos de Wael, R., Benkarim, O., Turnbull, A., Margulies, D., Jefferies, E., McCall, C., Bernhardt, B., & Smallwood, J. (2020). The relationship between individual variation in macroscale functional gradients and distinct aspects of ongoing thought. NeuroImage, 220, 117072. https://doi.org/10.1016/j.neuroimage.2020.117072
Schaefer, A., Kong, R., Gordon, E. M., Laumann, T. O., Zuo, X.-N., Holmes, A. J., Eickhoff, S. B., & Yeo, B. T. (2017). Local-global parcellation of the human cerebral cortex from intrinsic functional connectivity MRI. Cerebral Cortex, 28(9), 3095–3114. https://doi.org/10.1093/cercor/bhx179
Vos de Wael, R., Benkarim, O., Paquola, C., Lariviere, S., Royer, J., Tavakol, S., Xu, T., Hong, S.-J., Langs, G., Valk, S., Misic, B., Milham, M., Margulies, D., Smallwood, J., & Bernhardt, B. C. (2020). BrainSpace: A toolbox for the analysis of macroscale gradients in neuroimaging and Connectomics datasets. Communications Biology, 3(1). https://doi.org/10.1038/s42003-020-0794-7