Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: Neuroimage. 2020 Jan 11;209:116536. doi: 10.1016/j.neuroimage.2020.116536

Neighborhood poverty predicts altered neural and behavioral response inhibition

Rachel C Tomlinson [a], S Alexandra Burt [b], Rebecca Waller [a],[c],[d], John Jonides [a], Alison L Miller [e], Ashley N Gearhardt [a], Scott J Peltier [a],[f],[g], Kelly L Klump [b], Julie C Lumeng [h], Luke W Hyde [a]
PMCID: PMC7065021  NIHMSID: NIHMS1559422  PMID: 31935521

Abstract

Socioeconomic disadvantage during childhood is associated with a myriad of negative adult outcomes. One mechanism through which disadvantage undermines positive outcomes may be by disrupting the development of self-control. The goal of the present study was to examine pathways from three key indicators of socioeconomic disadvantage – low family income, low maternal education, and neighborhood poverty – to neural and behavioral measures of response inhibition. We utilized data from a representative cohort of 215 twins (ages 7–18 years, 70% male) oversampled for exposure to disadvantage, who participated in the Michigan Twins Neurogenetics Study (MTwiNS), a study within the Michigan State University Twin Registry (MSUTR). Our child-friendly Go/No-Go task activated the bilateral inferior frontal gyrus (IFG), and activation during this task predicted behavioral inhibition performance, extending prior work on adults to youth. Critically, we also found that neighborhood poverty, assessed via geocoding, but not family income or maternal education, was associated with IFG activation, a finding that we replicated in an independent sample of disadvantaged youth. Further, we found that neighborhood poverty predicted response inhibition performance via its effect on IFG activation. These results provide the first mechanistic evidence that disadvantaged contexts may undermine self-control via their effect on the brain. The broader neighborhood, beyond familial contexts, may be critically important for this association, suggesting that contexts beyond the home have profound effects on the developing brain and behaviors critical for future health, wealth, and wellbeing.

1. Introduction

Socioeconomic disadvantage during childhood predicts a myriad of negative outcomes, including lower earnings, poorer mental and physical health, and higher rates of criminal behavior (Cohen et al., 2008; Duncan, Ziol-Guest, & Kalil, 2010). One way disadvantage may exert these effects is by undermining the development of self-control, or the broad capacity to regulate behavior, thoughts, and emotions (Baumeister, Heatherton, & Tice, 1994). Consistent with this notion, disadvantage predicts poorer self-control during childhood (Hackman, Gallop, Evans, & Farah, 2015; Last, Lawson, Breiner, Steinberg, & Farah, 2018; Lengua et al., 2015) and, in turn, lower levels of self-control during childhood and adolescence predict poor short and long-term outcomes (Mischel & Ayduk, 2004; Moffitt et al., 2011).

Self-control is a broad construct comprised of a number of related functions, each with different underlying neural architecture (Mischel et al., 2011; Nee & Jonides, 2008). One such function is response inhibition, a construct that indexes the ability to suppress an unwanted or inappropriate response (Miyake et al., 2000; Munakata et al., 2011). Response inhibition (e.g., performance on a Go/No-Go task) has been linked to adult health and well-being (Mahmood et al., 2013; Young et al., 2009), is correlated with socioeconomic disadvantage (Noble, McCandliss, & Farah, 2007), shows stable individual differences from childhood to adulthood (Casey et al., 2011), and can be measured in non-human animals, providing clear neural regions of interest (Bari & Robbins, 2013). In humans, the inferior frontal gyrus (IFG) is critical for response inhibition (Chambers et al., 2007): damage to the IFG has been linked to impaired response inhibition (Aron, Fletcher, Bullmore, Sahakian, & Robbins, 2003; Swick, Ashley, & Turken, 2008), and the IFG is consistently activated by response inhibition tasks in adults (Aron & Poldrack, 2006).

Response inhibition demonstrates a protracted developmental trajectory, with adolescence representing a critical juncture for the development of its underlying neural architecture (Blakemore & Choudhury, 2006; Davidson, Amso, Anderson, & Diamond, 2006). Consequently, inhibition-related brain regions (e.g., the IFG) may be particularly susceptible to environmental influence during adolescence. Indeed, environmental factors, including socioeconomic disadvantage, appear to contribute significantly to variance in response inhibition performance in youth (Baumeister et al., 1994; Friedman et al., 2016). Thus, the current study aims to identify the neural mechanisms underlying this pathway by testing the hypothesis that socioeconomic disadvantage impacts response inhibition performance via disrupting the functioning of the IFG during adolescence (Falk et al., 2013).

Importantly, socioeconomic disadvantage can take many forms, and the few existing studies linking disadvantage to brain structure and function have focused on family income and maternal education (Hackman et al., 2015; Lengua et al., 2015), each of which could differentially influence neurobehavioral development and imply different mechanisms (Duncan & Magnuson, 2012). Additionally, youth in families with fewer resources often live in more impoverished neighborhoods; a key third unmeasured variable. Neighborhood contexts confer additional risk beyond low family income because they increase exposure to additional adverse experiences (e.g., neighborhood danger, under-resourced schools, toxicant exposure (Evans, 2004; Leventhal & Brooks-Gunn, 2000)).

Thus, the current study examined pathways from three different sources of socioeconomic disadvantage – low family income, low maternal education, and neighborhood poverty – to neural and behavioral measures of response inhibition in a relatively large sample of youth (age 7–18 years; N=215). This sample was recruited via birth records to be representative of southeast Michigan with an oversampling for exposure to neighborhood impoverishment, an important innovation over most existing studies that are based on smaller, convenience samples with few families facing significant disadvantage and without a clear population to generalize to (Falk et al., 2013). Using this unique sample, we measured inhibition-related activation during a child-friendly Go/No-Go task and examined whether socioeconomic disadvantage predicted neural reactivity and behavioral performance on this task. We leveraged an independent sample of youth (age 9–11 years; N=29) residing, on average, in even more impoverished neighborhoods, to examine whether findings replicated in a relevant population. Finally, we tested a mechanistic pathway through which disadvantage may become biologically embedded to undermine the development of self-control, by testing whether socioeconomic disadvantage predicted behavioral response inhibition via inhibition-related brain activation.

2. Materials and Methods

2.1. Participants

The primary sample included 215 youth from 140 twin pairs participating in the Michigan Twins Neurogenetics Study (MTwiNS), a new and on-going study within the broader Michigan State University Twin Registry (MSUTR; (Burt & Klump, 2013). Youth in MTwiNS ranged in age from 7 to 18 years (mean age = 13.6 years, SD 2.5 years, 70% male) and were living in south-central Michigan. These youth were recruited from the Twin Study of Behavioral and Emotional Development – Child (TBED-C), which identified twins for two cohorts via birth records (a strong sampling frame). The first cohort was sampled to represent families with twins living within 120 miles of Michigan State University, an area that includes Detroit, Flint, Lansing and other urban areas, as well as substantial swaths of rural Michigan. The second cohort was recruited from the same area, but only included families living in neighborhoods with over 10.5% of families living below the poverty line (the median for the state of Michigan at the time; e.g., Burt, Klump, Gorman-Smith, & Neiderhiser, 2016), thus representing families living in neighborhoods with above average poverty. For the present study, 140 twin pairs were re-recruited from these two samples at age 7 to 18 years with an additional oversample for boys in the neuroimaging sample, resulting in a sample that represents families living in south-central Michigan with substantial oversampling for families living in impoverished neighborhoods and families with boy twin-pairs. This sampling approach yields advantages over many previous studies of socioeconomic context and the brain that were based on small, affluent convenience samples by providing a sample from a clear sampling frame (birth records) for generalizability, but also one that has substantial numbers of families living in disadvantaged context. These families have in the past been the least likely to recruited to participate in fMRI research, but have the experiences most important to the research question at hand. In the current MTwiNS sample, 68% of twin families lived in neighborhoods with >10.5% of families living below the poverty line (mean percentage of families below the poverty line in the neighborhood was 21% and ranged as high as 69%). Participants included in the present analyses met basic fMRI eligibility criteria, such as the absence of metal in their body and willingness to participate in the scanning session (Table 1).

Table 1:

Summary of data included in MTwiNS analyses

Number Lost Participants With Usable Data
Original Sample 280 (140 sets)
Sample with Go/No-Go Behavioral Data 226
 • Declined scan 14
 • Ended scan before completing Go/No-Go task 11
 • Too large for scanner 2
 • Structural abnormality found during high-resolution scan 1
 • Ineligible for study (e.g. braces) 25
 • Experimenter error 1
Total Lost 54
Sample with Usable Go/No-Go Behavioral Data 215
 • Go Accuracy < 90% 11
Total Lost 11
Sample with Go/No-Go Functional Imaging Data 212
 • Ended scan early 1
 • MRI technical error 2
Total Lost 3
Sample with Usable Functional Imaging Data 185 (from 113 sets)
 • No usable structural scan 2
 • Frontal lobe coverage <85% 16
 • Failed visual inspection (prefrontal artifacts) 9
 • >10% motion outliers identified using ART 0
Total Lost 27
Open in a new tab

Based on the sampling frame and oversample for families in impoverished neighborhoods, twins from under-represented minority backgrounds were somewhat over-represented compared to the local area. Parents reported the race of the twins included in this report as follows: 76% White/Caucasian, 11% Black/African American, 3% Biracial, 1% Native American, 4% Hispanic, and 5% Other. This distribution of twin race contains somewhat more minority participants than the average for the State of Michigan (e.g., 76% identified as white, versus 80% in Michigan; https://www.census.gov/quickfacts/mi). This sample also contains a substantial number of families with low-income and lower maternal education than average. Maternal education ranged from 8 years (no high school degree) to 18 years (graduate degree), with a mean of 15 years. Median reported family annual income for this sample was $60,000 to $69,999 and ranged from less than $4,999 to greater than $90,000. 12% of MTwiNS families reported an annual income below the 2017 federal poverty line of $24,600 per year and 59% reported annual income below the living wage for a family of 4 in Michigan (http://livingwage.mit.edu/states/26). Parents provided informed consent and youth provided assent in compliance with the policies of the Institutional Review Board of the University of Michigan.

To assess the reproducibility and generalizability of our findings we replicated our results in an independent sample of 29 children age 9–11. These children, part of the ABC Brains study, were originally recruited into the Appetite, Behavior, and Cortisol (ABC) Preschool cohort at ages 3–4 years from 2009–2011 from Head Start, a federally-funded preschool program, with a goal of obtaining a large portion of families living in poverty (see Lumeng et al., 2014 for more details about the sample). Exclusion criteria for the original ABC Brains sample were the following: parent with ≥ 4-year college degree; parent or child not English-speaking; child in foster care, with food allergies, significant medical problems or perinatal complications; gestational age < 35 weeks; or use of medication known or hypothesized to affect cortisol. The 29 children included in the present analyses reflect the ABC Brains participants who had usable MRI data out of an original MRI sample of 56. Sample breakdown is presented in Table S1. Consistent with the recruitment goals, the sample was highly impoverished, even more so than the MTwiNS sample, with 55% of the sample reporting family income below the federal poverty line ($25,100 for a family of 4 in 2018). Median reported family annual income for this sample was $20,000 and ranged from $0 to above $75,000. Parents reported the race/ethnicity of the children included as follows: 45% White/Caucasian, 17% Black/African American, 17% Hispanic, and 21% Other.

2.2. Procedure

In the primary sample, children and their primary caregivers took part in a day-long visit to the University of Michigan which included a one-hour fMRI scan for each youth. Youth were scanned using blood-oxygen-level-dependent (BOLD) fMRI while completing several tasks, including the Go/No-Go task described below. Additionally, primary caregivers completed a demographic interview with an examiner. Most primary caregivers were mothers (96%). All procedures (including fMRI acquisition, task, fMRI analysis and preprocessing) were identical for the replication sample.

2.3. Measures

2.3.1. Socioeconomic Disadvantage.

Neighborhood poverty was defined using geocoding of family addresses to assess the proportion of neighborhood residents living below the poverty line between 2011 and 2015 in each family’s census tract, (www.census.gov). Family income was defined via primary caregiver reported monthly household gross income, including outside additional sources such as government assistance or child support. Maternal education was defined via the primary caregiver’s highest completed level of education. Though most primary caregivers were mothers, a small percentage (4%) were fathers. In this case, the father’s highest completed level of education was used.

2.3.2. Go/No-Go Task.

The task of interest for the current study was a child-friendly Go/No-Go task adapted from Casey et al. (Casey et al., 1997), in which neural reactivity during response inhibition is elicited via a “whack-a-mole” game (stimuli courtesy of Sarah Getz and the Sackler Institute for Developmental Psychobiology; task downloaded from http://fablab.yale.edu/page/assays-tools). In the present task, participants were instructed to press a button as quickly as possible in response to one stimulus (“go”, a mole) and avoid responding to a less frequent non-target (“No-Go”, a vegetable). The target stimuli (moles) were modified with various “disguises” to make the task more interesting and difficult given the relatively slow speed of stimuli due to MRI scanning requirements. The task consisted of four runs, each with approximately 55 trials, for a total of 255 trials of which 55 were No-Go (21.6% No-Go). Each No-Go trial was preceded by 1–5 go trials. Each trial lasted 2300ms, including a maximum of 1800ms stimulus presentation, 400ms feedback, and 100 – 1000ms fixation to account for individual differences in reaction time (Figure 1). Participants practiced the task in an MRI simulator before the MRI scan.

Figure 1: A child-friendly Go/No-Go task reliably activated inhibitory control neural architecture and activity predicted behavioral response inhibition.

Figure 1:

Open in a new tab

(a) Child-friendly Go/No-Go task. Participants were instructed to “whack” the moles by pressing a button, and to refrain from “whacking” the eggplants. Feedback was provided. Moles and eggplants had disguises in some images to increase the difficulty. The task included 255 trials, 55 No-Go, for a length of about10 minutes. (b) Brain regions exhibiting inhibition-related activation during No-Go > Go trials (N=185). Coordinates provided using the Montreal Neurological Institute (MNI) online atlas. Results shown from a one-sample t-test for all regions active during No-Go>Go trials in SPM12. False positive rate is controlled across the whole brain using 3dClustSim for cluster-level correction (punc<0.001, alpha<0.05). (c) Right inferior frontal gyrus, anterior cingulate cortex, and thalamus activation predict better performance during response inhibition in a large sample of youth (N=185). These areas were identified via an SPM12 regression which regressed efficiency (Go/No-Go accuracy / RT) on brain activation. False positive rate is controlled across the whole brain using 3dClustSim for cluster-level correction (punc<0.001, alpha<0.05). (d) Right inferior frontal gyrus activation correlates with Go/No-Go Efficiency (r=0.19, N=185). Right IFG main effects activation was extracted from an anatomical mask and correlated with efficiency (accuracy / RT). Outliers were determined to have a Cook’s d<0.5 and were retained for analyses.

For each participant, a behavioral inhibitory efficiency score based on both accuracy and reaction time was calculated as a measure of behavioral Go/No-Go performance. Accuracy was calculated as the percent of “No-Go” trials for which a participant correctly avoided responding. Reaction time was calculated as the average reaction time for “go” trials, excluding failed “go” trials (no response). Inhibitory efficiency (henceforth called “response inhibition” for clarity) was calculated by dividing accuracy by reaction time. This measure accounts for the potential of two individuals to obtain the same accuracy score while one individual trades reaction time for accuracy. Efficiency scores have been used in similar prior studies (Hirose et al., 2012) and provide greater variability in performance for the current task given its slow, fMRI-friendly structure (i.e., accuracy is relatively high with less variability). Participants with a hit rate below 90% on “go” trials were not considered to be consistently participating in the task and were therefore dropped from all analyses (n=11; see Table 1). Of these 11 participants, 9 had usable fMRI data. Including these dropped participants in SPM models produced effectively identical results to those reported in this paper.

2.4. fMRI Image Acquisition and Processing

Functional imaging data were acquired using a GE Discovery MR750 3T scanner located at the University of Michigan Functional MRI Laboratory. Data were acquired using an 8-channel GE head coil. One run of 274 volumes was collected for each participant. BOLD functional images were acquired using a gradient-echo reverse spiral sequence (repetition time = 2000ms, echo time = 30ms, flip angle = 90°, FOV = 22cm). Images included 43 interleaved oblique slices of 3mm thickness with 3.44×3.44mm2 in-plane resolution. High-resolution T1-weighted images (156 slices, slice thickness = 1mm, in-plane resolution of 1×1mm2) were also collected for use during normalization.

Functional data were preprocessed and analyzed using Statistical Parametric Mapping software version 12 (SPM12; Wellcome Department of Imaging Neuroscience, London, England). To allow for stabilization of the MR signal, we did not collect the first four volumes of each run. Raw k-space data were de-spiked before reconstruction to image space. Slice timing correction was performed using the 23rd slide as the reference slice. Functional data were then spatially realigned to the 10th slice of the volume. These spatially realigned data were coregistered to a high-resolution T1-weighted image, and segmented and spatially normalized into standard stereotactic space (MNI template). Finally, functional data were smoothed using a 6mm Gaussian kernel. After preprocessing, the Artifact detection Tools toolbox (ART; https://www.nitrc.org/projects/artifact_detect/) was used to detect translation or rotational motion outlier volumes that remained after earlier QA (>2mm movement or 3.5 rotation) and to create regressors accounting for the possible effects of these volumes. In addition, preprocessed images were visually inspected for major artifacts, and frontal lobe coverage was checked using the WFU PickAtlas “frontal lobe” structural mask (Maldjian, Laurienti, Kraft, & Burdette, 2003). A participant’s fMRI data were considered unusable if they contained obvious prefrontal artifacts or had less than 85% coverage of the frontal mask (Table 1). Families recruited into this study with 0, versus 1, versus 2 twins with usable fMRI data did not differ on twin age, twin race, or maternal education. However, families with two twins included in the fMRI analyses did report higher family incomes than families with one or fewer twins included (p= 0.025) and thus this variable was included in all models. Twins included in the fMRI analyses did not differ from the full sample in gender distribution. Note that these considerations were only important for analyses run in SPM12 in which missing data cannot be accommodated. In indirect effects models in MPlus v. 7.2, all participating youth were included in analyses using Maximum Likelihood estimation which provides efficient and unbiased estimates, even in the face of missing data.

2.5. Experimental Design and Statistical Analysis

2.5.1. Functional Data Analysis.

Functional data were modeled using the general linear model in SPM12. Three conditions were modeled: correct No-Go trials, in which a participant correctly withheld a response to a No-Go stimulus; incorrect No-Go trials, in which a participant incorrectly responded to a No-Go stimulus; and Go trials, in which a participant saw a Go stimulus. Incorrect Go trials were not modeled due to the expected high hit rates for Go trials (median 100%). For each participant, the main contrast of interest was all No-Go > Go.

2.5.2. Aim 1: Neural correlates of response inhibition.

To identify neural regions showing inhibition-related activation during the Go/No-Go task, consistent with past work in adults (Criaud & Boulinguez, 2013), individual contrast images for No-Go > Go trial were used in second-level random effects models using a one-sample t-test. Additionally, to identify neural regions responding to the task and predicting behavioral performance, we ran a whole-brain regression model with Go/No-Go performance (efficiency) as a continuous regressor. Across all analyses, we used the monte-carlo simulation program 3dClustSim to achieve p<.05 corrected for multiple comparisons across the whole brain (voxel level p<0.001). We replicated the main effects of inhibition-related IFG activation in the replication sample by examining group-level activation for the No-Go > Go contrast within an anatomically-defined IFG ROI from WFU PickAtlas (Maldjian et al., 2003). Additionally, we created a region-of-interest mask in the primary sample of the voxels in the IFG that were activated by the task and correlated with efficiency scores to create a mask to extract activation in the independent replication sample (i.e., to confirm if the same voxels in the replication sample also correlated with task performance). We then examined the correlation between inhibition-related activation in those voxels and task performance within the replication sample using the statistical package R v.3.3.2 (2016).

2.5.3. Aim 2: The relationship between socioeconomic disadvantage and neural activation during response inhibition.

To determine the unique relationship between each socioeconomic context variable (family income, maternal education, neighborhood poverty) and brain activation, we ran regression models within SPM12 for each variable predicting activation across the whole brain, while controlling for other socioeconomic context variables. We replicated these findings in the replication sample by examining the correlation of these three socioeconomic context variables with extracted activation from the IFG (as described above).

2.5.4. Single-twin subsample sensitivity analyses

To address the issue of nesting of twins within families, and therefore neighborhoods, we re-ran the socioeconomic context regression model to check for similar patterns of results in a single-twin subsample, which included no related twins. This subsample was made up of all twins that did not have a co-twin scanned and then a randomly selected twin from all complete pairs, including a total of 113 twins.

2.5.5. Aim 3: Indirect effects of socioeconomic disadvantage on response inhibition via IFG reactivity.

We extracted activation from the main effects of No-Go>Go (i.e., first eigenvariate) from the bilateral anatomical IFG mask for each participant using SPM12 for use in structural equation models (SEM) in MPlus v.8.1 (Muthén & Muthén, 1998–2011). These values were extracted from the main effect of task, not from regressions including socioeconomic disadvantage, nor performance variables, and thus not susceptible to bias via double-correlation. We tested the indirect pathway SEM (see Figure 2c) on the full MTwiNS sample (N=215), using Maximum Likelihood estimation with robust standard errors (MLR) and the CLUSTER command to account for nesting within families. Our conservative model controlled for all other SES-related variables, as well as age.

Figure 2: Neighborhood poverty is associated with behavioral response inhibition via decreased inhibitory control-related inferior frontal gyrus (IFG) activity.

Figure 2:

Open in a new tab

(a) Neighborhood poverty predicts lower activation in bilateral IFG during response inhibition (N=176).Brain regions in which lower inhibition-related activation during No-Go > Go correlates with greater neighborhood poverty (calculated as the percent of families in the neighborhood living below the poverty line according to 2011–2015 census data) from a regression in SPM12 that regressed neighborhood poverty onto brain activity (No-Go>Go) while controlling for family income and maternal education. The analysis used 3DClustSim to correct for multiple comparisons (p<0.05) across the entire brain. (b) Correlation between bilateral IFG activation and neighborhood poverty (N=176). Bilateral IFG activation extracted from an anatomical IFG region (as defined by the WFU Pickatlas; Maldjian et al., 2003) correlates with neighborhood poverty (r=−0.19). Outliers were determined to have Cook’s d <0.5 and were therefore retained for analyses. (c) Indirect effects of socioeconomic context on response inhibition via brain activation (N=215). We examined associations from socioeconomic context variables to response inhibition via inhibition-related bilateral IFG activation in 215 twins aged 7–18 years old using MLR estimation in Mplus version 8.1. IFG activation was extracted from an anatomical mask retrieved from WFU PickAtlas (Maldjian et al., 2003). All socioeconomic context variables were allowed to co-vary within the model. We utilized the MPlus CLUSTER command to account for nesting of twins within families. Model fit was excellent (χ2 = 3.338, p=0.34; RMSEA=0.023; CFI=0.996; TLI=0.988; R2 response inhibition=0.307). There was an indirect pathway from neighborhood poverty to lower response inhibition via decreased inhibition-related IFG activation (αβ=−0.034, SE=0.017, β=−2.063, p=0.039; bootstrapped CI not available when using CLUSTER command)

2.6. Data sharing statement

Full anonymized functional imaging and behavioral data for the entire MTwiNS study will be made available to the research community via NIMHdb. Relevant scripts will be made available to readers upon request to the corresponding author (lukehyde@umich.edu).

3. Results

3.1. Zero-order associations

Descriptive statistics and bivariate correlations between socioeconomic context variables and response inhibition for MTwiNS are presented in Table 2, and for the replication sample in Table S2. As expected, contextual variables were all related to each other, with moderate to modest positive correlations among maternal income, maternal education, and neighborhood poverty (r’s=0.09–0.34). In addition, as expected, response inhibition performance was strongly correlated with age (r=0.53, p<0.001) and modestly correlated with neighborhood poverty (r=−0.15, p=0.026), family income (r = 0.15, p = 0.028), and maternal education (r = 0.13, p = 0.054).

Table 2:

Descriptive statistics and bivariate correlations for study variables in the MTwiNS sample.

N M (SD) Range Response Inhibition Family Income Maternal Education Neighborhood Poverty Age IFG
Response Inhibition 215 1.69 (0.32) 0.83 – 2.8 1
Family Income 210 4724 (2190) 0 – 7500 0.15* 1
Maternal Education 215 14.75 (2.29) 7 – 18 0.13 0.24* 1
Neighborhood Poverty 215 0.21 (0.17) 0 – 0.69 −0.15* −0.34* −0.09 1
Age 215 13.63 (2.51) 7 – 18 0.53** 0.04 0.09 −0.13 1
IFG Activation (No-Go>Go) 185 1.11 (1.54) −7.9 – 9.4 0.19* −0.05 0.08 −0.19* 0.06 1
Open in a new tab

N = 215.

**

p<0.01,

*

p<0.05.

Response inhibition refers to performance on the Go/No-Go task, determined as an efficiency score (No-Go accuracy / Go reaction time). Family income is parent report of monthly gross household income, including sources such as government assistance or child support. Maternal education is parental report of years of education completed. Neighborhood poverty is census tract data indicating the percent of families with children under 18 within the census tract who were living below the poverty line from 2011 – 2015. IFG activation was extracted from a bilateral PickAtlas anatomical IFG mask.

3.2. What are the neural correlates of response inhibition in youth?

In the main sample of 215 twins (age 7–18), the child-friendly Go/No-Go task yielded reliable inferior frontal gyrus (IFG) and anterior cingulate cortex (ACC) activation (for No-Go>Go trials; Figure 1b, Table 3). We replicated this task eliciting IFG activation in the replication sample of 29 children age 9–11. Extending results from adults to youth (Hirose et al., 2012), we found that individual differences in neural activation were related to response inhibition performance in both the right IFG (t=5.25, k=249) and in the ACC (t=4.99; k=577; 3dClustSim punc<0.01, alpha<0.05; Figure 1c) such that greater activation in these areas related to better task performance. Importantly, we found that activation in the same voxels of the IFG activated by the task in the main sample of twins was also correlated with task performance in the replication sample (r=0.38, Table S2).

Table 3:

Go/No-Go task main effects (No-Go > Go trials)

Name Peak (x, y, z) T k (# voxels)
Left Inferior Frontal Gyrus −34, 18, −14 12.54 2815
Right Inferior Frontal Gyrus 42, 18, −2 12.11 4878
Cingulate 4, 28, 26 11.84 22744
Right Superior Temporal Gyrus 56, −44, 18 9.86 4111
Left Precentral Gyrus −42, −2, 44 5.82 503
Left Superior Parietal Lobule −32, −50, 44 5.53 233
Open in a new tab

Brain regions showing greater activation for No-Go trials compared to Go trials in a sample of 185 twins. False positive rate is controlled across the whole brain using 3dClustSim for cluster-level correction (punc < 0.001, alpha < 0.05, k > 209).

3.3. Which aspects of socioeconomic disadvantage predict neural activation during response inhibition?

Low family income and neighborhood poverty were each related to poorer behavioral response inhibition performance in zero-order correlations. However, only neighborhood poverty was related to inhibition-related brain activation both in zero-order models and multivariate models that examined the unique contribution of each aspect of disadvantage. Family income and maternal education were not related to differences in inhibition-related activation in zero-order whole brain analyses (i.e., even when we did not control for the other disadvantage variables). Moreover, higher levels of neighborhood poverty were not only associated with reduced inhibition-related bilateral IFG activation, but also reduced activation in the bilateral medial temporal gyrus, the bilateral caudate, and the ventral striatum (for No-Go>Go trials; Figure 2; Tables 2 and 4).

Table 4:

Brain regions showing reduced activation to No-Go > Go with increasing neighborhood poverty

Name Peak (x, y, z) T k (# voxels)
Left Inferior Frontal Gyrus
Left Ventral Striatum		 −32, 16, −20 5.38 959
Right Medial Temporal Gyrus 56 −46 −8 4.83 1089
Left Medial Temporal Gyrus −46 −44 −8 4.77 620
Right Ventral Striatum 10 6 0 4.59 379
Right Inferior Frontal Gyrus −42, −2, 44 4.46 392
Open in a new tab

Brain regions showing reduced activation to No-Go > Go with increasing neighborhood poverty in a sample of 176 twins. False positive rate is controlled across the whole brain using 3dClustSim for cluster-level correction (punc < 0.001, alpha < 0.05, k > 220; Clusters labeled via peak activation and SPM12 Neuromorphometrics).

3.3.1. Confirming effects in a single-twin subsample

To confirm that these results were not affected by the nesting of twins, we re-examined a single-twin random subsample and found that the relationship of interest between IFG activation and neighborhood poverty persisted. Effect sizes were similar and estimates reported remained statistically significant (results available upon request).

3.3.2. Replicating results in an independent sample (ABC Brains)

Further, we replicated these results by finding that activation in the same voxels in the IFG identified in the main twin sample was correlated with neighborhood poverty in the independent replication sample (r=−0.39, Table S2).

3.4. Does socioeconomic disadvantage predict response inhibition via IFG activation?

We examined whether our three measures of socioeconomic disadvantage were related to response inhibition via IFG activation during inhibition. In line with the results reported in the main text, using bilateral IFG activation extracted from an anatomic ROI during No-Go > Go trials in MPlus structural equation models, there was a pathway from neighborhood poverty to IFG activation (B=−0.23, SE=0.07, β=−3.07, p<0.01) and IFG activation to lower response inhibition (B=0.15, SE=0.07, β =2.21, p=0.027) such that those living in more impoverished neighborhood showed lower IFG activation during response inhibition and in turn showed worse inhibition performance. Moreover, there was an indirect pathway from neighborhood poverty to lower response inhibition via decreased inhibition-related IFG activation, supporting the hypothesis that neighborhood disadvantage is linked to poorer inhibitory control via neural activity (B=−0.034, SE=0.017, β=−2.063, p=0.039; Figure 2c). There were no indirect effects from maternal education or lower family income to response inhibition via IFG activation, though these predictors were retained in the overall path model to perform a more conservative test of the indirect effects. Note that in indirect effects models, we examined data extracted from a conservative bilateral IFG mask, as we had no hemispheric hypotheses and using a bilateral mask allowed us to decrease model complexity and decrease multiple comparisons by avoiding two correlated neural variables in the same model.

3.5. Sex-specific effects post-hoc analyses

Given that our sample was 70% male due to oversampling of boys in the early pilot phase of the study, we examined whether controlling for gender would affect results. Thus, in sensitivity analyses, we re-ran the SEM model described in 3.4, including gender as a covariate (i.e., it was regressed on all other variables). We found that the indirect effect from neighborhood to inhibition via brain activation persisted in this model (β =−2.027, p=0.043). Indirect effects for the other socioeconomic disadvantage variables did not emerge, nor did the overall pattern of findings change. We also calculated separate zero-order correlations as discussed in section 3.1 for male and female participants (Table S3). The primary correlations of interest for this paper, the correlations between neighborhood and IFG and between IFG and inhibition, do not significantly differ between boys and girls (z=0.48 and 0.62, p=0.63 and 0.54, respectively). However, the correlation between family income and inhibition did differ between boys and girls (r=0.22 for boys and r=−0.08 for girls, z=2.01, p=0.04, Table S3). Otherwise, no two correlations were significantly different between these groups.

4. Discussion

Within a large, well-sampled cohort of youth from neighborhoods with a wide range of poverty levels (including enrichment for high poverty neighborhoods), we found that living in a more impoverished neighborhood was associated with lower levels of bilateral inhibition-related IFG activation, even when controlling for family income and maternal education. We replicated these effects in an independent sample of youth living in even more disadvantaged contexts. Moreover, we found that neighborhood poverty predicted decreased response inhibition via decreased bilateral IFG activation, suggesting that one mechanism through which disadvantage may undermine self-control is via decreased activation in the IFG. Finally, we extended brain-behavior results in adults to youth by showing that performance on a child-friendly Go/No-Go task correlates with activation within the neural architecture of inhibitory control (i.e., the IFC and cingulate cortex). Overall, these results are striking given the relatively large, representative sample with substantial numbers of families living in impoverished neighborhoods, the use of extensive controls for confounds, geocoded measures of neighborhood poverty, and a replication sample. Our findings shed light on the specific neural mechanisms through which disadvantaged contexts undermine the development of self-control. The findings also highlight the critical role disadvantage in the broader neighborhood, beyond the familial context, may play on poor future outcomes.

We examined the neural correlates of response inhibition in a large well-sampled cohort of youth enriched for neighborhood poverty, a type of sample rarely studied using fMRI (Falk et al., 2013). We found evidence that the neural architecture of response inhibition is supported by the IFG and cingulate cortex among other areas, in line with previous results in adults (see Swick, Ashley, & Turken, 2011 for a meta-analysis) and children (Bunge, Dudukovic, Thomason, Vaidya, & Gabrieli, 2002; Casey et al., 1997). We also found that activity in these areas is related to behavioral performance, such that children who performed more efficiently on the task had stronger inhibition-related activation. This novel finding extends previous evidence of performance-related activation in the IFG in adults (Hirose et al., 2012) and supports the behavioral relevance of this neural activation.

Strikingly, and in contrast to previous work that has focused on family income and maternal education as important for the neural mechanisms underlying self-control (Hackman et al., 2015; Lengua et al., 2015), we found clear evidence for a link between neighborhood poverty and inhibition-related brain activation, over and above any family-level effects. Moreover, though we found associations between family income and behavioral response inhibition performance, neither family income, nor maternal education predicts neural activity in zero-order or multivariate models. These results, consistent across two samples, demonstrate that higher levels of neighborhood poverty predict worse self-control via disruption of inhibition-related IFG activation. These results are significant for two reasons: First, they provide evidence for the biological embedding of adverse experiences in childhood and a mechanism through which early adversity may undermine the development of self-control. Second, this is the first study to date to investigate the relationship between neighborhood poverty and inhibition-related brain activation. Our finding that neighborhood poverty predicts brain activation over and above family income or maternal education indicates that studies of socioeconomic context and the neurobiology of self-control may be missing a key, unmeasured variable: the neighborhood. Indeed, the current findings are consistent with recent work linking neighborhood disadvantage to emotion-related neural reactivity over and above family income and education (Gard et al., 2017), as well as a growing body of evidence that links neighborhood disadvantage to health outcomes (Minh, Muhajarine, Janus, Brownell, & Guhn, 2017). Taken together, evidence is mounting that the broader neighborhood context is critical to neurobehavioral development, above and beyond the effects of related familial contexts. These findings have major policy implications as they suggest that broader, non-familial contexts undermine brain and behavioral development above and beyond more proximal factors. It is clear that where children live can have lasting impacts on their chances in life and thus should be a target for preventative interventions aimed at ameliorating the effects of disadvantage.

Key for policy makers is that our geocoded, census-derived measure of neighborhood poverty may tap a variety of experiences and contexts, as an individual’s lived experience in impoverished neighborhoods is different from their more advantaged peers’ in countless ways. To identify the specific ‘active ingredients’ of neighborhood poverty that are mechanistically driving this association, future work should investigate mediators of the association between neighborhood disadvantage and neurobiological development, such as school quality, exposure to violence, or exposure to neurotoxicants, all of which cluster in neighborhoods with high levels of poverty. For example, recent work has indicated that better school climate is associated with increased global cortical thickness and better executive functioning (Piccolo, Merz, Noble, Pediatric Imaging, & Genetics, 2018). Though many preventative interventions for high risk youth have targeted family processes (e.g. Dishion et al., 2008), the current findings underscore the need to target more distal neighborhood processes.

That we did not find significant associations between family income and education and IFG activation was surprising. Future, larger studies with additional power may uncover significant associations. Future studies are also needed to disentangle the genetic and environmental mechanisms underlying this apparent relationship between socioeconomic disadvantage and poorer response inhibition. Though we show evidence that the impact of neighborhood poverty on response inhibition performance is not inflated by including twin pairs in the sample, additional work can be done to address potential gene-environment correlation and interplay implicated in the complex relationship between socioeconomic disadvantage and self-control (with our current sample size and twin nesting within neighborhood, we were unable to leverage the twin design in a meaningful way). In particular, it will be important to examine potential confounding variables like IQ that may explain the associations documented here. Moreover, a major limitation to this current work is that the analyses were cross-sectional. As MTwiNS is an ongoing longitudinal study, we look forward to examining these associations longitudinally in the future. Additionally, due to sampling methods implemented early in the study, our sample had more boys than girls (70% boys) and therefore results may not generalize as well to girls. In a post-hoc analysis we also found some limited evidence for gender differences in the effects of socioeconomic disadvantage on behavioral response inhibition that warrant future investigation. Importantly, our primary findings did replicate in an independent sample which was more gender balanced (51% boys), arguing for the generalizability of these findings. As some work has suggested that neighborhood effects of poverty may be sex-specific during adolescence (McBride Murry, Berkel, Gaylord-Harden, Copeland-Linder, & Nation, 2011), future work should consider conceptually replicating these findings in other large datasets, including publically available ones like ABCD (Hagler et al., 2019).

In sum, we provide important evidence that neighborhood disadvantage impacts both neural and behavioral measures of response inhibition. These findings provide a model by which a seemingly distal experience, the poverty level of a family’s neighbors, may affect the developing brain and subsequent performance on a critical indicator of self-control. Such a model serves to underscore the impact of adversity on children’s neurocognitive abilities, moving us closer to understanding the pathway through which socioeconomic context could undermine lifelong outcomes. Moreover, the extent to which findings were specific to neighborhood poverty highlight that where children live can have a profound impact on their biology, behavior, and chances in life. Ultimately, these results highlight the neural embedding of early adversity and provide evidence for a mechanistic neural pathway through which neighborhood disadvantage may undermine children’s future health, wealth, and wellbeing.

Supplementary Material

1

Acknowledgements

Research reported in this publication related to MTwiNS was supported by the National Institute of Mental Health of the National Institutes of Health (NIMH) and the Office of the Director National Institute of Health (OD), under Award Number UG3MH114249 to SAB and LWH. The ABC Brains study was supported primarily by an MCubed grant from the University of Michigan to JCL, LWH, ANG, and AM as well as by R21DK090718 to AM and JCL, RC1DK086376 to JCL, R01HD061356 to JCL, R01DK098983 to AM and JCL, and OBSSR/NICHD UH2HD087979 to AM. The content is solely the responsibility of the authors and does not necessarily represent the views of the National Institutes of Health. Additional MTwiNS funding was provided by the Avielle Foundation via The Conway Family Award for Excellence in Neuroscience (to LWH and SAB), a NARSAD young Investigator Grant from the Brain and Behavior Foundation (to LWH), and institutional funding provided by the University of Michigan (to LWH). This manuscript was supported by the National Science Foundation Graduate Research Fellowship under Grant No. DGE 1256260 to RCT. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Neuroimaging took place at the Functional MRI Laboratory of the University of Michigan which is supported by NIH Grant 1S10OD012240-01A1 (PI Noll).

We would like to thank the staff of the MTwiNS and ABC Brains studies for their hard work, and we thank the families who participated in MTwiNS and ABC Brains for sharing their lives with us.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interests Statement: The authors declare no competing financial interests.

5. References

  1. Aron AR, Fletcher PC, Bullmore ET, Sahakian BJ, & Robbins TW (2003). Stop-signal inhibition disrupted by damage to right inferior frontal gyrus in humans. Nat Neurosci, 6(2), 115–116. doi: 10.1038/nn1003 [DOI] [PubMed] [Google Scholar]
  2. Aron AR, & Poldrack RA (2006). Cortical and subcortical contributions to Stop signal response inhibition: role of the subthalamic nucleus. J Neurosci, 26(9), 2424–2433. doi: 10.1523/JNEUROSCI.4682-05.2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bari A, & Robbins TW (2013). Inhibition and impulsivity: Behavioral and neural basis of response control. Progress in Neurobiology, 108, 44–79. doi: 10.1016/j.pneurobio.2013.06.005 [DOI] [PubMed] [Google Scholar]
  4. Baumeister RF, Heatherton TF, & Tice DM (1994). Losing control: How and why people fail at self-regulation. San Diego, CA, US: Academic Press. [Google Scholar]
  5. Blakemore S-J, & Choudhury S (2006). Development of the adolescent brain: implications for executive function and social cognition. Journal of Child Psychology and Psychiatry, 47(3–4), 296–312. doi: 10.1111/j.1469-7610.2006.01611.x [DOI] [PubMed] [Google Scholar]
  6. Bunge SA, Dudukovic NM, Thomason ME, Vaidya CJ, & Gabrieli JDE (2002). Immature frontal lobe contributions to cognitive control in children: Evidence from fMRI. Neuron, 33(2), 301–311. doi: 10.1016/s0896-6273(01)00583-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burt SA, & Klump KL (2013). The Michigan State University Twin Registry (MSUTR): An update. Twin Research and Human Genetics, 16, 344–350. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Burt SA, Klump KL, Gorman-Smith D, & Neiderhiser JM (2016). Neighborhood disadvantage alters the origins of children’s non-aggressive conduct problems. Clinical Psychological Science. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Casey BJ, Somerville LH, Gotlib IH, Ayduk O, Franklin NT, Askren MK, … Shoda Y (2011). Behavioral and neural correlates of delay of gratification 40 years later. Proc Natl Acad Sci U S A, 108(36), 14998–15003. doi: 10.1073/pnas.1108561108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Casey BJ, Trainor RJ, Orendi JL, Schubert AB, Nystrom LE, Giedd JN, … Rapoport JL (1997). A developmental functional MRI study of prefrontal activation during performance of a Go-No-Go task. Journal of Cognitive Neuroscience, 9(6), 835–847. doi: 10.1162/jocn.1997.9.6.835 [DOI] [PubMed] [Google Scholar]
  11. Chambers CD, Bellgrove MA, Gould IC, English T, Garavan H, McNaught E, … Mattingley JB (2007). Dissociable mechanisms of cognitive control in prefrontal and premotor cortex. J Neurophysiol, 98(6), 3638–3647. doi: 10.1152/jn.00685.2007 [DOI] [PubMed] [Google Scholar]
  12. Cohen P, Chen H, Gordon K, Johnson J, Brook J, & Kasen S (2008). Socioeconomic background and the developmental course of schizotypal and borderline personality disorder symptoms. Dev Psychopathol, 20(2), 633–650. doi: 10.1017/S095457940800031X [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Criaud M, & Boulinguez P (2013). Have we been asking the right questions when assessing response inhibition in go/no-go tasks with fMRI? A meta-analysis and critical review. Neurosci Biobehav Rev, 37(1), 11–23. doi: 10.1016/j.neubiorev.2012.11.003 [DOI] [PubMed] [Google Scholar]
  14. Davidson MC, Amso D, Anderson LC, & Diamond A (2006). Development of cognitive control and executive functions from 4 to 13 years: Evidence from manipulations of memory, inhibition, and task switching. Neuropsychologia, 44(11), 2037–2078. doi: 10.1016/j.neuropsychologia.2006.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dishion TJ, Shaw D, Connell A, Gardner F, Weaver C, & Wilson M (2008). The family check-up with high-risk indigent families: preventing problem behavior by increasing parents’ positive behavior support in early childhood. Child Dev, 79(5), 1395–1414. doi: 10.1111/j.1467-8624.2008.01195.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Duncan GJ, & Magnuson K (2012). Socioeconomic status and cognitive functioning: moving from correlation to causation. Wiley Interdisciplinary Reviews: Cognitive Science, 3(3), 377–386. doi: 10.1002/wcs.1176 [DOI] [PubMed] [Google Scholar]
  17. Duncan GJ, Ziol-Guest KM, & Kalil A (2010). Early-childhood poverty and adult attainment, behavior, and health. Child Dev, 81(1), 306–325. doi: 10.1111/j.1467-8624.2009.01396.x [DOI] [PubMed] [Google Scholar]
  18. Evans GW (2004). The environment of childhood poverty. Am Psychol, 59(2), 77–92. doi: 10.1037/0003-066X.59.2.77 [DOI] [PubMed] [Google Scholar]
  19. Falk EB, Hyde LW, Mitchell C, Faul J, Gonzalez R, Heitzeg MM, … Schulenberg J (2013). What is a representative brain? Neuroscience meets population science. Proc Natl Acad Sci U S A, 110(44), 17615–17622. doi: 10.1073/pnas.1310134110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Friedman NP, Miyake A, Altamirano LJ, Corley RP, Young SE, Rhea SA, & Hewitt JK (2016). Stability and change in executive function abilities from late adolescence to early adulthood: A longitudinal twin study. Dev Psychol, 52(2), 326–340. doi: 10.1037/dev0000075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gard AM, Waller R, Shaw DS, Forbes EE, Hariri AR, & Hyde LW (2017). The long reach of early adversity: Parenting, stress, and neural pathways to antisocial behavior in adulthood. Biol Psychiatry Cogn Neurosci Neuroimaging, 2(7), 582–590. doi: 10.1016/j.bpsc.2017.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hackman DA, Gallop R, Evans GW, & Farah MJ (2015). Socioeconomic status and executive function: developmental trajectories and mediation. Dev Sci, 18(5), 686–702. doi: 10.1111/desc.12246 [DOI] [PubMed] [Google Scholar]
  23. Hagler DJ Jr., Hatton S, Cornejo MD, Makowski C, Fair DA, Dick AS, … Dale AM (2019). Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. Neuroimage, 116091. doi: 10.1016/j.neuroimage.2019.116091 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hirose S, Chikazoe J, Watanabe T, Jimura K, Kunimatsu A, Abe O, … Konishi S (2012). Efficiency of go/no-go task performance implemented in the left hemisphere. J Neurosci, 32(26), 9059–9065. doi: 10.1523/JNEUROSCI.0540-12.2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Last BS, Lawson GM, Breiner K, Steinberg L, & Farah MJ (2018). Childhood socioeconomic status and executive function in childhood and beyond. Plos One, 13(8), e0202964. doi: 10.1371/journal.pone.0202964 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lengua LJ, Moran L, Zalewski M, Ruberry E, Kiff C, & Thompson S (2015). Relations of growth in effortful control to family income, cumulative risk, and adjustment in preschool-age children. J Abnorm Child Psychol, 43(4), 705–720. doi: 10.1007/s10802-014-9941-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Leventhal T, & Brooks-Gunn J (2000). The neighborhoods they live in: The effects of neighborhood residence on child and adolescent outcomes. Psychological Bulletin, 126(2), 309–337. doi: 10.1037/0033-2909.126.2.309 [DOI] [PubMed] [Google Scholar]
  28. Lumeng JC, Miller A, Peterson KE, Kaciroti N, Sturza J, Rosenblum K, & Vazquez DM (2014). Diurnal cortisol pattern, eating behaviors and overweight in low-income preschool-aged children. Appetite, 73, 65–72. doi: 10.1016/j.appet.2013.10.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mahmood OM, Goldenberg D, Thayer R, Migliorini R, Simmons AN, & Tapert SF (2013). Adolescents’ fMRI activation to a response inhibition task predicts future substance use. Addict Behav, 38(1), 1435–1441. doi: 10.1016/j.addbeh.2012.07.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Maldjian JA, Laurienti PJ, Kraft RA, & Burdette JH (2003). An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage, 19(3), 1233–1239. [DOI] [PubMed] [Google Scholar]
  31. McBride Murry V, Berkel C, Gaylord-Harden NK, Copeland-Linder N, & Nation M (2011). Neighborhood Poverty and Adolescent Development. Journal of Research on Adolescence, 21(1), 114–128. doi: 10.1111/j.1532-7795.2010.00718.x [DOI] [Google Scholar]
  32. Minh A, Muhajarine N, Janus M, Brownell M, & Guhn M (2017). A review of neighborhood effects and early child development: How, where, and for whom, do neighborhoods matter? Health Place, 46, 155–174. doi: 10.1016/j.healthplace.2017.04.012 [DOI] [PubMed] [Google Scholar]
  33. Mischel W, & Ayduk O (2004). Willpower in a cognitive-affective processing system: The dynamics of delay of gratification In Baumeister RF & Vohs KD (Eds.), Handbook of self-regulation: Research, theory, and applications. (pp. 99–129). New York, NY, US: Guilford Press. [Google Scholar]
  34. Mischel W, Ayduk O, Berman MG, Casey BJ, Gotlib IH, Jonides J, … Shoda Y (2011). ‘Willpower’ over the life span: decomposing self-regulation. Soc Cogn Affect Neurosci, 6(2), 252–256. doi: 10.1093/scan/nsq081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, & Wager TD (2000). The unity and diversity of executive functions and their contributions to complex “Frontal Lobe” tasks: a latent variable analysis. Cogn Psychol, 41(1), 49–100. doi: 10.1006/cogp.1999.0734 [DOI] [PubMed] [Google Scholar]
  36. Moffitt TE, Arseneault L, Belsky D, Dickson N, Hancox RJ, Harrington H, … Caspi A (2011). A gradient of childhood self-control predicts health, wealth, and public safety. Proc Natl Acad Sci U S A, 108(7), 2693–2698. doi: 10.1073/pnas.1010076108 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Munakata Y, Herd SA, Chatham CH, Depue BE, Banich MT, & O’Reilly RC (2011). A unified framework for inhibitory control. Trends Cogn Sci, 15(10), 453–459. doi: 10.1016/j.tics.2011.07.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Muthén LK, & Muthén BO (1998–2011). Mplus User’s Guide (Sixth Edition ed.). Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  39. Nee DE, & Jonides J (2008). Dissociable interference-control processes in perception and memory. Psychol Sci, 19(5), 490–500. doi: 10.1111/j.1467-9280.2008.02114.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Noble KG, McCandliss BD, & Farah MJ (2007). Socioeconomic gradients predict individual differences in neurocognitive abilities. Developmental Science, 10(4), 464–480. doi: 10.1111/j.1467-7687.2007.00600.x [DOI] [PubMed] [Google Scholar]
  41. Piccolo LR, Merz EC, Noble KG, Pediatric Imaging N, & Genetics S (2018). School climate is associated with cortical thickness and executive function in children and adolescents. Dev Sci, e12719. doi: 10.1111/desc.12719 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Swick D, Ashley V, & Turken AU (2008). Left inferior frontal gyrus is critical for response inhibition. BMC Neurosci, 9, 102. doi: 10.1186/1471-2202-9-102 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Swick D, Ashley V, & Turken U (2011). Are the neural correlates of stopping and not going identical? Quantitative meta-analysis of two response inhibition tasks. Neuroimage, 56(3), 1655–1665. doi: 10.1016/j.neuroimage.2011.02.070 [DOI] [PubMed] [Google Scholar]
  44. Team, R. C. (2016). R: A language and environment for statistical computing. Vienna, Austria: Retrieved from https://www.r-project.org/ [Google Scholar]
  45. Young SE, Friedman NP, Miyake A, Willcutt EG, Corley RP, Haberstick BC, & Hewitt JK (2009). Behavioral disinhibition: liability for externalizing spectrum disorders and its genetic and environmental relation to response inhibition across adolescence. J Abnorm Psychol, 118(1), 117–130. doi: 10.1037/a0014657 [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

1
NIHMS1559422-supplement-1.docx (43.7KB, docx)

RESOURCES