Intrinsic brain mapping of cognitive abilities: A multiple-dataset study on intelligence and its components

• Published in: NeuroImage (Orlando, Fla.) Vol. 309; p. 121094
• Main Authors: Di Plinio, Simone, Perrucci, Mauro Gianni, Ferrara, Grazia, Sergi, Maria Rita, Tommasi, Marco, Martino, Mariavittoria, Saggino, Aristide, Ebisch, Sjoerd JH
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.04.2025; Elsevier Limited; Elsevier
• Subjects: Adult; Brain - diagnostic imaging; Brain - physiology; Brain architecture; Brain mapping; Brain Mapping - methods; Cognition & reasoning; Cognition - physiology; Cognitive ability; Connectome - methods; Datasets; Efficiency; Executive Function - physiology; Factor analysis; Female; Fluid reasoning; Functional connectivity; Functional magnetic resonance imaging; Humans; Hypotheses; Intelligence; Intelligence - physiology; Magnetic Resonance Imaging - methods; Male; Memory; Mental task performance; Middle Aged; Modularity; Nerve Net - diagnostic imaging; Nerve Net - physiology; Neural coding; Neurocognition; Neurosciences; Quantitative psychology; Radiology/Diagnostic Imaging; Sensorimotor system; Standard scores; System theory; Wechsler Scales; Young Adult; Functional connectivity; Functional magnetic resonance imaging; Intelligence; Fluid reasoning; Neurocognition
• ISSN: 1053-8119; 1095-9572; 1095-9572
• DOI: 10.1016/j.neuroimage.2025.121094

•Distributed efficiency as well as selective integration impact cognitive abilities.•Sensorimotor efficiency underpins fluid intelligence across cognitive batteries.•Efficient visual processing supports faster cognition and memory.•Multilevel brain modularity encodes general to specific cognitive abilities.•DMN and limbic integration weakens cognitive skills across cognitive batteries. This study investigates how functional brain network features contribute to general intelligence and its cognitive components by analyzing three independent cohorts of healthy participants. Cognitive scores were derived from 1) the Wechsler Adult Intelligence Scale (WAIS-IV), 2) the Raven Standard Progressive Matrices (RPM), and 3) the NIH and Penn cognitive batteries from the Human Connectome Project. Factor analysis on the NIH and Penn cognitive batteries yielded latent variables that closely resembled the content of the WAIS-IV indices and RPM. We employed graph theory and a multi-resolution network analysis by varying the modularity parameter (γ) to investigate hierarchical brain-behavior relationships across different scales of brain organization. Brain-behavior associations were quantified using multi-level robust regression analyses to accommodate variability and confounds at the subject-level, node-level, and resolution-level. Our findings reveal consistent brain-behavior relationships across the datasets. Nodal efficiency in fronto-parietal sensorimotor regions consistently played a pivotal role in fluid reasoning, whereas efficiency in visual networks was linked to executive functions and memory. A broad, low-resolution 'task-positive' network emerged as predictive of full-scale IQ scores, indicating a hierarchical brain-behavior coding. Conversely, increased cross-network connections involving default mode and subcortical-limbic networks were associated with reductions in both general and specific cognitive performance. These outcomes highlight the relevance of network efficiency and integration, as well as of the hierarchical organization in supporting specific aspects of intelligence, while recognizing the inherent complexity of these relationships. Our multi-resolution network approach offers new insights into the interplay between multilayer network properties and the structure of cognitive abilities, advancing the understanding of the neural substrates of the intelligence construct.