Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 23.
Published in final edited form as: Dev Psychobiol. 2021 Jan 11;63(5):1589–1596. doi: 10.1002/dev.22096

Childhood Socioeconomic Status is Prospectively Associated with Surface Morphometry in Adulthood

Alexander J Dufford a, Gary W Evans b, Israel Liberzon c, James E Swain d, Pilyoung Kim a
PMCID: PMC13102113  NIHMSID: NIHMS1659442  PMID: 33432574

Abstract

Childhood socioeconomic status (SES) has been associated with brain cortex surface area in children. However, the extent to which childhood SES is prospectively associated with brain morphometry in adulthood is unclear. We tested if childhood SES (income-to-needs ratio averaged across ages 9, 13, and 17) is prospectively associated with cortical surface morphometry in adulthood. Average childhood income-to-needs ratio had a positive, prospective association with cortical thickness in adulthood in the precentral gyrus, postcentral gyrus, and caudal middle frontal gyrus. (p < 0.05, FWE corrected). Childhood income-to-needs ratio also had a positive, prospective association with cortical surface area in adulthood in multiple regions, including the rostral and caudal middle frontal gyri and superior frontal gyrus (p < 0.05, FWE corrected). Concurrent income-to-needs ratio (measured at age 24) was not associated with cortical thickness or surface area in adulthood. The results underscore the importance of addressing poverty in childhood for brain morphological development.

Keywords: surface morphometry, socioeconomic status, childhood, longitudinal

Introduction

Childhood socioeconomic status (SES) has been associated with variations in brain structure in children and adults (Dufford et al., 2020; Hanson, Chandra, Wolfe, & Pollak, 2011; Johnson, Riis, & Noble, 2016; Noble et al., 2015; Staff, Murray, Ahearn, Mustafa, & Whalley, 2011). Studies have primarily focused on the associations between family income and gray matter volume in several brain regions including the hippocampus (Dufford, Bianco, & Kim, 2018; Hanson et al., 2011; Noble et al., 2015), amygdala (Dufford, Bianco, & Kim, 2019; Hanson et al., 2015; Luby et al., 2013), and prefrontal cortex (Hanson et al., 2013; Lawson, Duda, Avants, Wu, & Farah, 2013). Variations in brain structure associated with childhood SES have been further related to individual differences in cognitive, affective, and socioemotional outcomes in adulthood (Dufford et al., 2018; Johnson et al., 2016; Noble et al., 2015; Palacios-Barrios & Hanson, 2018). However, these studies are typically cross-sectional; therefore, with little information on whether childhood SES is prospectively associated with adult brain structure.

Recent studies of brain structure have shifted to focusing on cortical thickness and surface area as they have distinct genetic and developmental contributions (Hanson et al., 2013; Lawson et al., 2013; Noble et al., 2015). Cortical thickness follows a pattern of expansion from infancy to early childhood followed by a linear decline around the age of five (Ducharme et al., 2016; Lyall et al., 2015; Schnack et al., 2015; Wierenga, Langen, Oranje, & Durston, 2014). For surface area, there is an expansion in childhood and early adolescence and decreases in adulthood (Brown & Jernigan, 2012; Wierenga et al., 2014). In a large study of children from 3–20 years old, family income and parental education were positively associated with surface area in the bilateral inferior temporal, insula, right occipital and medial prefrontal cortex (Noble et al., 2015); however, the association between family income and cortical thickness was not as robust. In a sample of children (mean age of 11.4) greater family income was associated with greater cortical thickness in the right anterior cingulate and left superior frontal gyri (Lawson et al., 2013); however, surface area was not tested in this study. Additionally, these studies were cross-sectional and therefore it was not possible to test if childhood SES had a prospective relation with surface morphometry.

Prospective studies are critical to understand the associations between childhood SES and variations in health (both physical and mental) in later life. The association between childhood SES and brain structure in adults (aged 18–25 years) has been examined in just a few cross-sectional studies. In one study, higher childhood socioeconomic status (SES) composite scores (which included the participant’s subjective socioeconomic status rating, income-to-needs ratio, and parental education) were associated with greater hippocampal volume, however this relation was not significant when adulthood SES was incorporated (Yu et al., 2018). In another volumetric study, we found that amygdala volume among young adults mediated links between childhood cumulative risk exposure at ages 9 and 13 and elevated amygdala responses to emotionally neutral facial stimuli (Evans et al., 2016) However, these studies only measured gray matter volume in the hippocampus and amygdala respectfully. In our previous study, we found there was a significant prospective positive association between childhood SES (INR) at age 9, with white matter organization in frontolimbic and association tracts (Dufford et al., 2020) in adulthood. Our study showed that positive trajectories of socioeconomic status during childhood are associated with greater regional white matter organization in adulthood. Building on these studies, it is also critical to examine surface morphometry and utilize a whole-brain analysis to understand how childhood SES may prospectively be associated with brain structure in adults.

In the current study, we measured childhood SES as family income-to-needs ratio at age 9, 13, and 17 years of age. Income-to-needs ratio is calculated by dividing the total family income by the poverty threshold adjusted for the number of individuals living in the household as specified by the United States Census Bureau. In addition to several studies finding associations between family income and surface morphometry (Noble et al., 2015), we focused on family income rather than parental education (another widely used measure of SES) because it more sensitively measures childhood and adulthood SES while parental education tends to be stable across a child’s life (Erola, Jalonen, & Lehti, 2016). Based upon previous studies of the association between SES and brain structure (Johnson et al., 2016; Noble et al., 2015; Piccolo, Merz, He, Sowell, & Noble, 2016), we hypothesized that average childhood income-to-needs ratio, measured at age 9, 13, and 17, would have a prospective and positive association with both cortical thickness and surface area in the prefrontal cortex, specifically the superior frontal gyrus and middle frontal gyrus (Noble et al., 2015). We also examined the association between concurrent income-to-needs ratio, measured at the time of the scan in adulthood (age 24) to better understand developmental timing of risks factors such as childhood poverty. Based upon a previous study examining childhood versus concurrent SES and brain structure (Yu et al., 2018), we hypothesized there would not be an association between concurrent INR and cortical thickness or surface area. To further examine how the variations related to SES are related to behavioral outcomes, guarding for the risk of reverse inference (Ellwood-Lowe, Sacchet, & Gotlib, 2016), we tested whether cortical thickness and surface area in the prefrontal regions associated with average childhood INR (superior frontal gyrus) are negatively associated with a self-report measure of cognitive functioning.

Materials and Methods

Participants.

Fifty-four participants (mean 23.7 years of age, SD = 1.5, 54% male) were recruited from a larger longitudinal study of the relation between income and child development (Evans, 2003). The neuroimaging subsample overlapped 100% with previous studies examining the examining the associations between childhood SES and brain structure/function (Duval et al., 2017; Evans et al., 2016; Javanbakht et al., 2016; Javanbakht et al., 2015; Kim et al., 2013; Kim, Ho, Evans, Liberzon, & Swain, 2015; Liberzon et al., 2015; Sripada, Swain, Evans, Welsh, & Liberzon, 2014). However, the relation between childhood INR and surface morphometry in adulthood has not been examined. Structural images were acquired from fifty-two participants in adulthood; one participant could not tolerate the scanner and one participant had excessive motion during scanning (determined by visual inspection of the structural image). Additionally, adult income-to-needs ratio (at age 24) was not collected for three of the participants. To keep the samples consistent across the analysis (childhood and concurrent), we excluded these participants that did not have a concurrent INR measurement. Due to the missing data the sample size was n = 49 (mean 23.6 years of age, 57.1 % male and 89.8% Caucasian). The neuroimaging subsample for this study closely resembled the larger cohort (mean 23.5 years of age, 51% male, mean adult INR = 2.9). Participants were included in the study if they had no prior or current treatment for psychiatric disorders, neurological conditions, MRI contraindications, and were right-handed.

Procedures.

Data from each age (9, 13, 17) were acquired at home visits with the participants and their mother with a protocol designed to study poverty and socioemotional development (Evans, 2003). The income-to-needs ratio (INR) was averaged across the timepoints (9 ,13, 17) to provide an accurate estimate of INR across childhood in addition to the high degree of multicollinearity making it difficult to make conclusions about one timepoint of INR measurement in childhood versus another (Evans, 2016). The sample was a set of low- and middle-income children recruited in rural area of the northeastern United States. Information concerning the individuals in adulthood is included in Table 1 such as their employment status, student status, and monthly financial support from parents. Data is available upon reasonable request to the authors in compliance with the International Review Board that oversees the data. Analysis code is available upon reasonable request from the corresponding author.

Table 1.

Demographic characteristics for the sample.

N (%) Mean ± SD Range
Average INR (9, 13, 17) 2.1 ± 1.33 0.2 – 5.5
INR (age 9) 1.68 ± 1.05 0.16 – 3.93
INR (age 13) 2.46 ± 1.62 0.22 – 5.97
INR (age 17) 2.49 ± 1.62 0.60 – 1.62
INR (concurrent INR) 3.30 ± 3.43 0.29 – 20.52
Age at scan in years 23.67 ± 1.31 20.00 – 27.00
Sex (female) 21 (42.9)
Race (white/caucasian) 44 (89.8)
Employment (employed) 36 (73.4)
Student (current student) 1 (0.02)
Monthly Financial Support from Parent (yes) 14 (28.5)
Monthly Financial Support from Parent (amount $) 58.8 ± 129.8 0 – 500
Open in a new tab

Income-to-needs Ratio.

INR was calculated by dividing the family’s income as calculated by the US census guidelines by the number of people living in the home. INR was calculated longitudinally at age 9, age 13, and age 17. INR was also collected when the participants completed the neuroimaging scan (referred to as concurrent INR); participants had a mean age of 23.67 ± 1.31 (range = 20–27) for the concurrent INR measurement. Since this measurement of INR was collected closest in time to the neuroimaging scan, it reflects the most accurate measurement of concurrent INR and therefore is used in all the analyses. Childhood INR reflected the child’s family income and family size when they were growing up (used their parent’s income, while the adult INR reflects that individual’s family income and family size when they were scanned). The INR measurement for age 13 was missing for 6 participants while the INR measurement for age 17 was missing data for 5 participants. However, no participants were missing both age 13 and age 17 INR. See Table 1 for the distribution of INR across the timepoints of data collection as well as the distribution of the average childhood INR. An INR of 1 defines the poverty line, therefore 24% of the sample was considered living below the poverty line.

Cognitive Failures Questionnaire.

For the exploratory analysis of childhood SES, adult brain surface morphometry, and behavior, we measured self-reported everyday cognitive functioning using the Cognitive Failures Questionnaire (CFQ) in adulthood (age 24). The CFQ is a self-report measure that assesses the frequency of difficulties in memory, action, perception in everyday life (Wagle et al., 2009). Questions span many topics in everyday cognition such as “Do you find you forget people’s names?”, and “Do you daydream when you ought to be listening to something?”. The CFQ has been widely used in studies as an indicator of cognitive functioning (Wagle, Berrios, & Ho, 1999) and has good psychometric properties (Bridger, Johnsen, & Brasher, 2013).

Structural MRI Data Acquisition and Structural MRI Data Quality Control Procedure.

Structural MRI data was acquired on a 3.0 T Philips scanner (VA Ann Arbor fMRI Center). High-resolution T1-weighted images were acquired with a 3D T1 Turbo field-echo (T1-TFE) pulse sequence (TR = 9.8 ms., TE = 4.60 ms, TI: 3000 ms., FA: 8°, FOV = 256 mm, matrix = 256 × 256, resolution = 1mm, slice thickness = 1 mm3, 180 sagittal slices with whole brain coverage) using an 8-channel SENSE head coil. Each raw T1 weighted image was inspected visually and given a rating of 1–4 based upon the image quality (Blumenthal, Zijdenbos, Molloy, & Giedd, 2002). Based upon the visual inspection, one image was removed to poor quality due to participant motion.

Structural MRI Data Image Processing and Quality Control.

Structural images were processed using Freesurfer’s (Freesurfer version 6.0, https://surfer.nmr.mgh.harvard.edu/fswiki/FreeSurferWiki) ‘recon-all’ pipeline for surface-based morphometry which has been described in detail elsewhere (Fischl & Dale, 2000; Fischl et al., 2002). Preprocessing steps include affine registration to the Talairach atlas, bias correction, skull stripping, and segmentation into gray matter, white matter, and cerebrospinal fluid. Segmentation guides the delineation of gray matter/white matter boundaries and the cortical surface is reconstructed. A surface for each hemisphere undergoes triangular tessellation into approximately 160,000 vertices per hemisphere using the cubic interpolation method and correction for topological defects is performed. Cortical thickness is calculated as the difference in millimeters between spatially equivalent vertices in the pial and gray/white matter surfaces. Surface area is the area of the vertex on the gray matter surface and calculated as the average of the area of the tessellated triangles that are touching that vertex. Freesurfer surface reconstruction was assessed for quality using VisualQC (https://github.com/raamana/visualqc) an assistive tool which provides multiple views of the image being assessed to conduct slice by slice assessment of surface reconstruction quality. VisualQC also uses outlier detection (Pedregosa et al., 2011) implemented via scikit-learn (Pedregosa et al., 2011) to identify potential outliers in the surface reconstruction. For each image, visual quality control of the surface reconstructions was conducted slice-by-slice for the axial, sagittal, and coronal views. Based upon the visual inspection and outlier detection, no image needed manual editing. As a final step, images were smoothed using a full width at half maximum of 3mm. This optimizes FWHM for the threshold-free cluster enhancement algorithm for surface data (Lett et al., 2017).

Vertex-wise Regression Analysis of the Association between Childhood Income-to-needs Ratio and Surface Morphometry in Adulthood.

Vertex-wise regressions were conducted in the tfce_mediation toolbox (Lett et al., 2017). This toolbox uses threshold free cluster enhancement (TFCE) and nonparametric permutation testing to define clusters and correct for multiple comparisons. TFCE avoids arbitrary thresholds and uses spatial neighborhood information in determining clusters. It has been shown to increase sensitivity of analyses compared to more traditional voxel-wise/vertex-wise thresholding (Smith & Nichols, 2007). Regarding false positive rate control, nonparametric permutation testing has been demonstrated to adequately control false positive rates, while parametric cluster-wise correction does not (Greve & Fischl, 2018). Using two vertex-wise regressions, we examined the association between average childhood INR (age 9, age 13, and age 17) and cortical thickness and surface area in adulthood. For each regression, we used the average childhood INR as a predictor of cortical thickness or surface area at each vertex adjusting for participant age at the time of the scan, sex, and race. As an additional analysis, we conducted a multiple regression for both cortical thickness and surface area in which the association between concurrent INR (measured at the time of the scan) and surface morphometry were examined adjusting for participant age at the time of the scan, sex, and race. Regressions are conducted independently for each hemisphere, therefore statistical correction across hemispheres is not necessary. For the exploratory analysis of the associations between cortical thickness/surface area and cognitive functioning (CFQ scores), mean values from the top three largest clusters were extracted and the correlation between them and CFQ were tested. As this analysis was exploratory, we examined the correlations both before and after corrections for multiple testing using the False Discovery Rate implemented in the R package “psych” (Revelle, 2017).

Data Availability.

Due to IRB restrictions the data is only available upon reasonable request from the authors.

Results

Descriptive Statistics.

Average childhood INR (age 9, 13, and 17) was associated with the age at the time of scan (r = −0.40, p < 0.01). Average childhood INR was not significantly different between male and female participants or white/non-white participants (ps > 0.05). Average childhood INR was associated with concurrent INR (measured at the time of the scan) (r = 0.37, p < 0.01), INR at age 9 (r = 0.88, p < 0.001), INR at 13 (r = 0.95, p < 0.001), and INR at 17 (r = 0.91, p < 0.001). For more information about the sample see Table 1. Average childhood INR was significantly associated with CFQ scores (r = −0.23, p < 0.05).

Vertex-wise Regression Analysis of the Association between Childhood Income-to-needs Ratio and Surface Morphometry in Adulthood.

For the multiple regression with average childhood INR as the predictor for cortical thickness in adulthood, there was a significant positive association in the following left hemisphere regions at PFWE < 0.05 (see Figure 1): caudal middle frontal gyrus, precentral gyrus, inferior parietal gyrus, and rostral middle frontal gyrus. Right hemisphere regions for the multiple regression with average childhood INR as the predictor of cortical thickness that reached significance at PFWE < 0.05 included clusters in the rostral middle frontal gyrus, postcentral gyrus, superior frontal gyrus, lingual gyrus, precentral gyrus, and pars triangularis (see Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Cortical thickness results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17), sex, age, race as covariates. Hot regions indicate a significant positive association between average childhood income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Regarding the multiple regression with average childhood INR as the predictor of surface area in adulthood, several cortical regions were significant at PFWE < 0.05 for the left hemisphere including the caudal middle frontal gyrus, superior frontal gyrus, medial orbitofrontal gyrus, lateral orbitofrontal gyrus, paracentral gyrus, frontal pole, and rostral middle frontal gyrus (see Figure 2). As for the multiple regression with average INR as the predictor of surface area in adulthood at PFWE < 0.05 for the right hemisphere, the following regions were significant: the rostral middle frontal gyrus, postcentral gyrus, caudal anterior cingulate, rostral anterior cingulate, postcentral gyrus, superior frontal gyrus, paracentral gyrus, precentral gyrus, and supramarginal gyrus (see Figure 2). These results remained consistent when concurrent INR was included as a covariate (Supplementary Figure 1 and Supplementary Figure 2). Results were also consistent when just INR at age 9 were examined (Supplementary Figure 3 and Supplementary Figure 4). There were no negative associations between average childhood INR and surface area. For the main effect of the concurrent INR, there were no significant associations with cortical thickness or surface area at PFWE < 0.05 for both the positive and negative direction.

Figure 2.

Figure 2.

Open in a new tab

Surface area results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17) as a predictor with sex, age, and race as covariates. Hot regions indicate a significant positive association between income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Exploratory Analysis of Cognitive Functioning.

As an exploratory analysis, we examined the associations between the cortical thickness/surface area mean values extracted from the whole brain analysis of average childhood INR. Using Pearson correlations, we found a significant negative association between left caudal middle frontal gyrus cortical thickness and cognitive functioning (CFQ scores) (r = −0.28, p < 0.05). We also found a significant negative association between the left medial orbitofrontal gyrus surface area and cognitive functioning (r = −0.28, p < 0.05). Associations between the cortical thickness values in the left precentral, left inferior parietal, right rostral middle frontal, right postcentral, right superior frontal gyri, and cognitive functioning were not significant (ps > 0.05). Associations between the surface area values in the left caudal middle frontal, left superior frontal, right rostral middle frontal, right postcentral, right caudal anterior cingulate gyri and cognitive functioning were not significant (ps > 0.05). The associations that were significant did not survive corrections for multiple testing.

Discussion

In this study, we examined whether childhood SES (average of income to needs measured at age 9, 13, and 17) was prospectively associated with surface morphometry in adulthood. The findings have potential implications for how the relation between SES and brain development is understood and highlights the importance of considering childhood SES’s role in processes underlying the developmental of cortical thickness and surface area. The longitudinal data afforded the ability to test associations with surface morphometry for average childhood INR versus concurrent INR. While average childhood INR had association with cortical thickness and surface area, there was no association with concurrent INR.

While longitudinal investigations of the relation between SES and brain development have been discussed as critical for an in-depth understanding of the impacts of SES on brain structure (Brito & Noble, 2014; Noble et al., 2015), most of the studies of this relation to date have been cross-sectional, limiting our understanding of questions related to the role of developmental timing in the development of brain morphology. The longitudinal measures of income-to-needs ratio in the current study enabled us to examine both the prospective and concurrent association between childhood SES and surface morphometry in adults. While concurrent family income has been associated with surface area in children and adolescents (Noble et al., 2015) in some reports, our findings suggest that average childhood INR is associated with both cortical thickness and surface area in early adulthood. We did not find an association between concurrent INR (measured at the time of scan in adulthood) and cortical thickness or surface area. Yu and colleagues (2018) reported similar, cross-sectional results between SES and hippocampal volume in 8–12 years old’s, but no association between adult (18–25 years) SES and concurrent hippocampal volume. Both studies thus offer find converging evidence, one cross-sectional and one longitudinal, that brain structural development may be particularly sensitive to early experiences. As both subcortical and cortical brain structure are undergoing rapid development during middle childhood (Sowell, Trauner, Gamst, & Jernigan, 2002), environmental factors likely have more pronounced associations with brain structure during more dynamic biological system development (Van den Bergh, 2011).

Consistent with previous studies of SES and surface morphometry, the associations found herein are widespread, a number of prefrontal regions, and span regions implicated in several different domains of functioning (Brito & Noble, 2014; Johnson et al., 2016; Noble et al., 2015; Noble, Wolmetz, Ochs, Farah, & McCandliss, 2006; Piccolo et al., 2016). In our cohort, as reported earlier, childhood INR was associated with reduced dorsolateral prefrontal cortex activity in adulthood during an emotion regulation task (Kim et al., 2013). Together this suggests evidence of the relation between childhood INR and both the structure and function of the dorsolateral prefrontal cortex (see Supplementary Figure 3 and Supplementary Figure 4 which show the main effect of childhood SES age 9 and surface morphometry). In addition to prefrontal regions, average childhood INR was associated with surface morphometry in occipital regions involved in visual processes. Recent studies have begun to find associations between childhood SES and regions involved in visual processes (Dufford et al., 2020; Rosen, Sheridan, Sambrook, Meltzoff, & McLaughlin, 2018). Evidence from fMRI, suggests that SES is associated with ventral visual stream activity (Rosen et al., 2018) which may have downstream impacts on visual attention and executive functioning. Future studies will be needed to examine how the structural findings in the occipital lobe are related to behaviors such as visual attention and executive functioning. While the direction of the association between SES and surface morphometry is similar to previous studies (Noble et al., 2015; Piccolo et al., 2016), the particular neuronal processes underlying these associations are yet unknown. This makes it difficult to ascertain whether the associations between childhood SES and surface morphometry is ‘maladaptive’. Future studies, perhaps using animal models, will be needed to examine how deprivation may be associated with the cellular processes underlying changes in thickness or surface area. Additionally, we conducted an exploratory analysis to examine if the cortical thickness/surface area associated with average childhood INR was correlated with cognitive functioning in adulthood. However, the negative associations between morphometry and cognitive functioning did not survive corrections for multiple comparisons. Given our sample size, the data set may be underpowered to adequately evaluate these post-hoc tests and future studies should focus on testing these associations.

The findings of the current study are not without limitations and the following should be considered. First, childhood SES was first measured at age 9, therefore, we did not have measurements of infancy or early childhood which could be prospectively associated with adult surface morphometry. Future longitudinal studies will be needed to examine these associations in infancy and early childhood as well as potential include measures of the prenatal environment. As discussed, due to the high collinearity of income measurements across the study, it is difficult to examine questions related to the timing of childhood SES. Second, INR measured at the age of 24 is difficult to measure as discussed in previous studies (Chan et al., 2018) as this age range is mixture of working adults and students. However, this limitation is less applicable for the current sample as the majority were working at the time of the measurement. Lastly, the sample for this study was Caucasian and relatively healthy and did not have a current or historical diagnosis of a psychiatric disorder. Individuals for this study were specifically chosen to have experienced low SES in childhood but have a minimal amount of psychopathology in adulthood. Due to this, other studies of this sample have not found robust associations with symptoms of psychopathology. These sample characteristics limit the generalizability of the results and require replication in more diverse samples particularly more ethnically diverse individuals and those experiencing varying degrees of psychopathology.

Conclusions

We found a positive and prospective relation between average childhood SES and adult surface morphometry (both cortical thickness and surface area) primarily in the caudal and rostral middle frontal gyrus and superior frontal gyrus. The study did not find evidence of an association between concurrent SES in adulthood with surface morphometry in adulthood but did find associations with average childhood SES. These findings establish long-lasting associations between average childhood SES and brain structural morphometry in adulthood and highlight the importance of addressing poverty in childhood.

Supplementary Material

Fig S4

Supplementary Figure 4. Surface area results from the vertex-wise regression for childhood income-to-needs ratio (age 9) as a predictor with sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Fig S3

Supplementary Figure 3. Cortical thickness results from the vertex-wise regression for childhood income-to-needs ratio (age 9), sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between average childhood income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Fig S2

Supplementary Figure 2. Surface area results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17) as a predictor with sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Fig S1

Supplementary Figure 1. Cortical thickness results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17), sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between average childhood income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

Acknowledgements

We thank Erika Blackburn, Sarah Garfinkel, S. Shaun Ho, and Robert Varney for assistance with data collection. The current study was supported by National Institutes of Health Grant RC2MD004767 (GWE, IL, JES), R01HD090068 (PK), R01DA074336 (JES), T32MH18268 (AJD), the William T. Grant Foundation, the John D. and Catherine T. MacArthur Foundation Network on Socioeconomic Status and Health, and the Robert Wood Johnson Foundation.

Footnotes

Declarations of Interest

The authors have no conflicts of interest to disclosed.

References

  1. Blumenthal JD, Zijdenbos A, Molloy E, & Giedd JN (2002). Motion artifact in magnetic resonance imaging: implications for automated analysis. Neuroimage, 16(1), 89–92. [DOI] [PubMed] [Google Scholar]
  2. Bridger RS, Johnsen SÅK, & Brasher K (2013). Psychometric properties of the cognitive failures questionnaire. Ergonomics, 56(10), 1515–1524. [DOI] [PubMed] [Google Scholar]
  3. Brito NH, & Noble KG (2014). Socioeconomic status and structural brain development. Frontiers in neuroscience, 8, 276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brown TT, & Jernigan TL (2012). Brain development during the preschool years. Neuropsychology review, 22(4), 313–333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chan MY, Na J, Agres PF, Savalia NK, Park DC, & Wig GS (2018). Socioeconomic status moderates age-related differences in the brain’s functional network organization and anatomy across the adult lifespan. Proceedings of the National Academy of Sciences, 115(22), E5144–E5153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ducharme S, Albaugh MD, Nguyen T-V, Hudziak JJ, Mateos-Pérez J, Labbe A, . . . Group, B. D. C. (2016). Trajectories of cortical thickness maturation in normal brain development—The importance of quality control procedures. Neuroimage, 125, 267–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dufford AJ, Bianco H, & Kim P (2018). Socioeconomic disadvantage, brain morphometry, and attentional bias to threat in middle childhood. Cognitive, Affective, & Behavioral Neuroscience, 1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dufford AJ, Bianco H, & Kim P (2019). Socioeconomic disadvantage, brain morphometry, and attentional bias to threat in middle childhood. Cognitive, Affective, & Behavioral Neuroscience, 19(2), 309–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dufford AJ, Evans GW, Dmitrieva J, Swain JE, Liberzon I, & Kim P (2020). Prospective associations, longitudinal patterns of childhood socioeconomic status, and white matter organization in adulthood. Human Brain Mapping. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Duval ER, Garfinkel SN, Swain JE, Evans GW, Blackburn EK, Angstadt M, . . . Liberzon I (2017). Childhood poverty is associated with altered hippocampal function and visuospatial memory in adulthood. Developmental cognitive neuroscience, 23, 39–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Erola J, Jalonen S, & Lehti H (2016). Parental education, class and income over early life course and children’s achievement. Research in Social Stratification and Mobility, 44, 33–43. [Google Scholar]
  12. Evans GW (2003). A multimethodological analysis of cumulative risk and allostatic load among rural children. Developmental psychology, 39(5), 924. [DOI] [PubMed] [Google Scholar]
  13. Evans GW (2016). Childhood poverty and adult psychological well-being. Proceedings of the National Academy of Sciences, 113(52), 14949–14952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Evans GW, Swain JE, King AP, Wang X, Javanbakht A, Ho SS, . . . Liberzon I (2016). Childhood cumulative risk exposure and adult amygdala volume and function. Journal of Neuroscience Research, 94(6), 535–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fischl B, & Dale AM (2000). Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proceedings of the National Academy of Sciences, 97(20), 11050–11055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, . . . Klaveness S (2002). Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron, 33(3), 341–355. [DOI] [PubMed] [Google Scholar]
  17. Greve DN, & Fischl B (2018). False positive rates in surface-based anatomical analysis. Neuroimage, 171, 6–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hanson JL, Chandra A, Wolfe BL, & Pollak SD (2011). Association between income and the hippocampus. PloS one, 6(5), e18712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hanson JL, Hair N, Shen DG, Shi F, Gilmore JH, Wolfe BL, & Pollak SD (2013). Family poverty affects the rate of human infant brain growth. PloS one, 8(12), e80954. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hanson JL, Nacewicz BM, Sutterer MJ, Cayo AA, Schaefer SM, Rudolph KD, . . . Davidson RJ (2015). Behavioral problems after early life stress: contributions of the hippocampus and amygdala. Biological Psychiatry, 77(4), 314–323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Javanbakht A, Kim P, Swain JE, Evans GW, Phan KL, & Liberzon I (2016). Sex-specific effects of childhood poverty on neurocircuitry of processing of emotional cues: a neuroimaging study. Behavioral Sciences, 6(4), 28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Javanbakht A, King AP, Evans GW, Swain JE, Angstadt M, Phan KL, & Liberzon I (2015). Childhood poverty predicts adult amygdala and frontal activity and connectivity in response to emotional faces. Frontiers in behavioral neuroscience, 9, 154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Johnson SB, Riis JL, & Noble KG (2016). State of the art review: poverty and the developing brain. Pediatrics, 137(4), e20153075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kim Evans, G. W., Angstadt M, Ho SS, Sripada CS, Swain JE, . . . Phan KL (2013). Effects of childhood poverty and chronic stress on emotion regulatory brain function in adulthood. Proceedings of the National Academy of Sciences, 110(46), 18442–18447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kim Ho, S. S., Evans GW, Liberzon I, & Swain JE (2015). Childhood social inequalities influences neural processes in young adult caregiving. Developmental Psychobiology, 57(8), 948–960. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lawson GM, Duda JT, Avants BB, Wu J, & Farah MJ (2013). Associations between children’s socioeconomic status and prefrontal cortical thickness. Developmental Science, 16(5), 641–652. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lett TA, Waller L, Tost H, Veer IM, Nazeri A, Erk S, . . . Vollstädt‐Klein S (2017). Cortical surface‐based threshold‐free cluster enhancement and cortexwise mediation. Human Brain Mapping, 38(6), 2795–2807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Liberzon I, Ma ST, Okada G, Shaun Ho S, Swain JE, & Evans GW (2015). Childhood poverty and recruitment of adult emotion regulatory neurocircuitry. Social Cognitive and Affective Neuroscience, 10(11), 1596–1606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Luby Belden, Botteron K, Marrus N, Harms MP, Babb C, . . . Barch D(2013). The effects of poverty on childhood brain development: the mediating effect of caregiving and stressful life events. JAMA pediatrics, 167(12), 1135–1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lyall AE, Shi F, Geng X, Woolson S, Li G, Wang L, . . . Gilmore JH (2015). Dynamic development of regional cortical thickness and surface area in early childhood. Cerebral Cortex, 25(8), 2204–2212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Noble KG, Houston SM, Brito NH, Bartsch H, Kan E, Kuperman JM, . . . Libiger O (2015). Family income, parental education and brain structure in children and adolescents. Nature neuroscience, 18(5), 773–778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Noble KG, Wolmetz ME, Ochs LG, Farah MJ, & McCandliss BD (2006). Brain–behavior relationships in reading acquisition are modulated by socioeconomic factors. Developmental Science, 9(6), 642–654. [DOI] [PubMed] [Google Scholar]
  33. Palacios-Barrios EE, & Hanson JL (2018). Poverty and self-regulation: Connecting psychosocial processes, neurobiology, and the risk for psychopathology. Comprehensive Psychiatry. [DOI] [PubMed] [Google Scholar]
  34. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, . . . Dubourg V (2011). Scikit-learn: Machine learning in Python. Journal of machine learning research, 12(Oct), 2825–2830. [Google Scholar]
  35. Piccolo LR, Merz EC, He X, Sowell ER, & Noble KG (2016). Age-related differences in cortical thickness vary by socioeconomic status. PLoS One, 11(9), e0162511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Revelle WR (2017). psych: Procedures for personality and psychological research. [Google Scholar]
  37. Rosen ML, Sheridan MA, Sambrook KA, Meltzoff AN, & McLaughlin KA (2018). Socioeconomic disparities in academic achievement: A multi-modal investigation of neural mechanisms in children and adolescents. NeuroImage, 173, 298–310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schnack HG, Van Haren NE, Brouwer RM, Evans A, Durston S, Boomsma DI, . . . Hulshoff Pol HE (2015). Changes in thickness and surface area of the human cortex and their relationship with intelligence. Cerebral Cortex, 25(6), 1608–1617. [DOI] [PubMed] [Google Scholar]
  39. Smith S, & Nichols T (2007). Threshold-free cluster-enhancement addressing the problem of threshold dependence in cluster inference. Paper presented at the 13th Annual Meeting of the OHBM, Chicago, Illinois. Neuroimage. [DOI] [PubMed] [Google Scholar]
  40. Sowell ER, Trauner DA, Gamst A, & Jernigan TL (2002). Development of cortical and subcortical brain structures in childhood and adolescence: a structural MRI study. Developmental Medicine & Child Neurology, 44(1), 4–16. [DOI] [PubMed] [Google Scholar]
  41. Sripada RK, Swain JE, Evans GW, Welsh RC, & Liberzon I (2014). Childhood poverty and stress reactivity are associated with aberrant functional connectivity in default mode network. Neuropsychopharmacology, 39(9), 2244–2251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Staff R, Murray A, Ahearn T, Mustafa N, & Whalley L (2011). Socioeconomic deprivation and brain size: A study using the Aberdeen 1932 birth cohort. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, 7(4), S425–S426. [Google Scholar]
  43. Van den Bergh BR (2011). Developmental programming of early brain and behaviour development and mental health: a conceptual framework. Developmental Medicine & Child Neurology, 53, 19–23. [DOI] [PubMed] [Google Scholar]
  44. Wagle A, Berrios G, & Ho L (1999). The cognitive failures questionnaire in psychiatry. Comprehensive Psychiatry, 40(6), 478–484. [DOI] [PubMed] [Google Scholar]
  45. Wierenga LM, Langen M, Oranje B, & Durston S (2014). Unique developmental trajectories of cortical thickness and surface area. Neuroimage, 87, 120–126. [DOI] [PubMed] [Google Scholar]
  46. Yu Q, Daugherty AM, Anderson DM, Nishimura M, Brush D, Hardwick A, . . . Ofen N (2018). Socioeconomic status and hippocampal volume in children and young adults. Developmental science, 21(3), e12561. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig S4

Supplementary Figure 4. Surface area results from the vertex-wise regression for childhood income-to-needs ratio (age 9) as a predictor with sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

NIHMS1659442-supplement-Fig_S4.pdf (5.2MB, pdf)
Fig S3

Supplementary Figure 3. Cortical thickness results from the vertex-wise regression for childhood income-to-needs ratio (age 9), sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between average childhood income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

NIHMS1659442-supplement-Fig_S3.pdf (4.7MB, pdf)
Fig S2

Supplementary Figure 2. Surface area results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17) as a predictor with sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

NIHMS1659442-supplement-Fig_S2.pdf (2.6MB, pdf)
Fig S1

Supplementary Figure 1. Cortical thickness results from the vertex-wise regression for average childhood income-to-needs ratio (age 9, 13, and 17), sex, age, race, and adult income-to-needs ratio as covariates. Hot regions indicate a significant positive association between average childhood income-to-needs ratio and cortical thickness at p < 0.05, family-wise error corrected, with hotter regions indicating greater statistical significance.

NIHMS1659442-supplement-Fig_S1.pdf (2.5MB, pdf)

Data Availability Statement

Due to IRB restrictions the data is only available upon reasonable request from the authors.

RESOURCES