A single, common BWAS brain pattern existed across variables that was most reflective of a child's socioeconomics, with SES showing the strongest brain-wide associations strongest in motor and sensory but not cognitive regions.
Key Findings
Results
Socioeconomic status showed the strongest brain-wide associations among 649 variables mapped to children's brains.
649 environmental and behavioral variables were mapped to children's brains and compared with each other and with neurobiological reference patterns.
SES associations were stronger than all other variables tested, including behavioral and environmental measures.
The study used data from children in a large-scale neuroimaging dataset (the ABCD Study).
SES was operationalized using composite measures including income, parental education, and related factors.
Results
SES brain associations were strongest in motor and sensory regions rather than cognitive regions.
The spatial pattern of SES brain associations showed a topography emphasizing motor and sensory cortex.
This pattern was unexpected given that SES is typically linked to cognitive outcomes.
The motor and sensory pattern was shared across many BWAS maps, including IQ.
Cognitive regions did not show the strongest SES-related brain variability.
Results
A single, common BWAS brain pattern existed across variables that was most reflective of a child's socioeconomics.
Comparison of 649 BWAS maps revealed a shared spatial pattern across many different variables.
This common pattern was most strongly correlated with SES measures.
The pattern suggests that many environmental and behavioral variables may share a common neural signature driven by socioeconomic factors.
The common pattern was identified through comparison of resultant BWAS maps with each other and with neurobiological reference patterns.
Results
Adjusting for SES weakened brain-IQ associations and eliminated the BWAS motor and sensory pattern.
When SES was statistically controlled, the strength of brain-IQ associations was reduced.
The motor and sensory spatial pattern in brain-IQ associations was eliminated after SES adjustment.
This suggests the motor and sensory component of brain-IQ associations is mediated by or confounded with SES.
Residual brain-IQ associations after SES adjustment were weaker and spatially distinct from the original pattern.
Results
Brain-with-IQ associations did not generalize when models were trained on higher-SES subsamples.
Predictive models of IQ trained on higher-SES children failed to generalize to lower-SES children.
This generalization failure indicates that brain-IQ associations are SES-dependent rather than universal.
The finding raises concerns about the external validity of BWAS findings derived from non-representative samples.
Lack of generalization suggests brain-behavior associations may be specific to the SES context in which they are identified.
Discussion
Children's brains vary the most with SES, potentially through SES-dependent sleep deprivation and stress.
The authors propose mechanistic pathways including sleep deprivation and stress as potential mediators of SES effects on brain development.
These pathways were identified as consistent with the spatial topography of SES brain associations (motor and sensory regions).
Sleep deprivation and stress are both known to disproportionately affect children from lower-SES backgrounds.
The proposed mechanisms are presented as potential explanations rather than definitively established causal pathways.
Results
The shared BWAS pattern across many variables reflects socioeconomics rather than variable-specific neural signatures.
Many of the 649 variables produced BWAS maps that were spatially similar to each other.
The similarity among maps was most explained by their shared association with SES.
This suggests that previous BWAS findings linking diverse variables to brain structure/function may partially reflect SES confounding.
BWAS maps were compared with neurobiological reference patterns to assess their biological plausibility and specificity.
What This Means
This research suggests that a child's socioeconomic status (SES) — including family income and parental education — is more strongly linked to differences in brain organization than almost any other factor studied. By analyzing 649 different environmental and behavioral variables and their relationships to brain activity patterns in children, the researchers found that most variables produced similar-looking brain maps, and that this common pattern was primarily driven by socioeconomic differences. Importantly, the brain differences associated with SES were largest in regions controlling movement and the senses, not in the regions most associated with thinking and reasoning.
This research suggests that when scientists find that intelligence (IQ) or other traits are associated with brain differences in children, a large part of that association may actually reflect the influence of socioeconomic circumstances. When the researchers statistically accounted for SES, the brain-IQ associations became weaker and lost their characteristic motor-sensory pattern. Furthermore, brain-based predictions of IQ trained on children from higher-SES backgrounds failed to work well for children from lower-SES backgrounds, suggesting that these brain-behavior relationships are not universal but are shaped by a child's socioeconomic environment.
The findings have important implications for how neuroscience research is interpreted and conducted. They suggest that many published findings linking brain differences to cognitive or behavioral outcomes in children may be substantially confounded by socioeconomic factors. The authors propose that SES may affect children's brain development partly through stress and sleep deprivation, which are more common in lower-SES households. This research suggests that future brain imaging studies need to carefully consider socioeconomic context and that broad claims about brain-behavior relationships in children should be interpreted cautiously when SES is not accounted for.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Marek S, Donohue M, Karcher N, Hoyniak C, Chauvin R, Meyer A, et al.. (2026). Patterns of brain-wide associations reflect socioeconomics.. Science (New York, N.Y.). https://doi.org/10.1126/science.aee6213