Associations between Socioeconomic Status, Cognition, and Brain Structure: Evaluating Potential Causal Pathways Through Mechanistic Models of Development

Abstract

Differences in socioeconomic status (SES) correlate both with differences in cognitive development and in brain structure. Associations between SES and brain measures such as cortical surface area and cortical thickness mediate differences in cognitive skills such as executive function and language. However, causal accounts that link SES, brain, and behavior are challenging because SES is a multidimensional construct: correlated environmental factors, such as family income and parental education, are only distal markers for proximal causal pathways. Moreover, the causal accounts themselves must span multiple levels of description, employ a developmental perspective, and integrate genetic effects on individual differences. Nevertheless, causal accounts have the potential to inform policy and guide interventions to reduce gaps in developmental outcomes. In this article, we review the range of empirical data to be integrated in causal accounts of developmental effects on the brain and cognition associated with variation in SES. We take the specific example of language development and evaluate the potential of a multiscale computational model of development, based on an artificial neural network, to support the construction of causal accounts. We show how, with bridging assumptions that link properties of network structure to magnetic resonance imaging (MRI) measures of brain structure, different sets of empirical data on SES effects can be connected. We use the model to contrast two possible causal pathways for environmental influences that are associated with SES: differences in prenatal brain development and differences in postnatal cognitive stimulation. We then use the model to explore the implications of each pathway for the potential to intervene to reduce gaps in developmental outcomes. The model points to the cumulative effects of social disadvantage on multiple pathways as the

Parental education and income are linked to offspring cortical brain structure and psychopathology at 9–11 years

“…First, regarding our SES estimations, we did not include subjective or cultural aspects of the construct (W. M. Liu et al., 2004 ; Yosso, 2005 ) nor did we investigate the unique contributions of our SES metrics. Also, income and education are distal markers for proximal causal pathways including stress, cognitive stimulation, obstetric complications, prenatal care, toxins and nutrition (Evans & Kim, 2013 ; Farah, 2017 ; Thomas & Coecke, 2023 ). Second, our study design can only capture correlational relationships and cannot infer causation.…”

“…The socioeconomic environment of a child can shape many central aspects of their life, including life expectancy, present and prospective cognitive abilities, school performance, and susceptibility for mental health struggles (Thomas & Coecke, 2023 ). Childhood is also marked by extensive brain development.…”

“…We then assessed the linear relation between parental education and parental income and each LICA component, and whether components could moderate the negative relationship between parental SES and general psychopathology. We hypothesized that parental education and income would show positive associations with global and frontal surface area (Noble et al., 2015 ; Rakesh & Whittle, 2021 ; Thomas & Coecke, 2023 ), and with widespread GWC for parental education (Norbom et al., 2022 ). We expected no parental SES ‐ cortical thickness relations (Norbom et al., 2022 ; Rakesh & Whittle, 2021 ), and had no specific hypotheses pertaining to curvature or multimodal coupling.…”

Parental education and income are linked to offspring cortical brain structure and psychopathology at 9-11 years

“…First, regarding our SES estimations, we did not include subjective or cultural aspects of the construct (W. M. Liu et al, 2004; Yosso, 2005) nor did we investigate the unique contributions of our SES metrics. Also, income and education are distal markers for proximal causal pathways including stress, cognitive stimulation, obstetric complications, prenatal care, toxins and nutrition (Evans & Kim, 2013; Farah, 2017; Thomas & Coecke, 2023). Second, our study design can only capture correlational relationships and cannot infer causation.…”

“…We then assessed the linear relation between parental education and parental income and each LICA component, and whether components could moderate the negative relationship between parental SES and general psychopathology. We hypothesized that parental education and income would show positive associations with global and frontal surface area (Noble et al, 2015; Rakesh & Whittle, 2021; Thomas & Coecke, 2023), and with widespread GWC for parental education (Norbom et al, 2022). We expected no parental SES - cortical thickness relations (Norbom et al, 2022; Rakesh & Whittle, 2021), and had no specific hypotheses pertaining to curvature or multimodal coupling.…”