[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Download PDF
Figure 1.
Conceptual Model Testing Multiple Mediators of the Hypothesized Association Between Income-to-Needs Ratio and Variation in Brain Volume
Conceptual Model Testing Multiple Mediators of the Hypothesized Association Between Income-to-Needs Ratio and Variation in Brain Volume

aMeasured at baseline.
bMeasured after baseline but before scan.
cBetween baseline and time of scan.
dTime of scan.

Figure 2.
Caregivers’ Education, Supportive/Hostile Parenting, and Children’s Experiences of Stressful Life Events as Mediators of the Relation Between Income-to-Needs Ratio and Hippocampus Volumes
Caregivers’ Education, Supportive/Hostile Parenting, and Children’s Experiences of Stressful Life Events as Mediators of the Relation Between Income-to-Needs Ratio and Hippocampus Volumes

Values shown are standardized regression coefficients. The top model is for the left (L) hippocampus volume, while the model at the bottom represents the right (R) hippocampus volume. Both models include whole-brain volume and sex as covariates.
aMeasured at baseline.
bP < .001.
cMeasured after baseline but before scan.
dP < .01.
eP < .05.
fAfter adding parents’ education, supportive/hostile parenting, and children’s stressful life events to the model.
gTime of scan.

Table 1.  
Demographics for Current Sample
Demographics for Current Sample
Table 2.  
Income-to-Needs Ratio Predicting Total White Matter and Cortical Gray Matter Volumes
Income-to-Needs Ratio Predicting Total White Matter and Cortical Gray Matter Volumes
Table 3.  
Hierarchical Regression: Income-to-Needs Ratio Variable Predicting Hippocampus and Amygdala Volumes
Hierarchical Regression: Income-to-Needs Ratio Variable Predicting Hippocampus and Amygdala Volumes
1.
Carneiro  PM, Heckman  JJ. Human capital policy. In: Heckman  JJ, Krueger  AB, Friedman  BM, eds. Inequality in America: What Role for Human Capital Policies? Cambridge, MA: MIT Press; 2003.
2.
Brooks-Gunn  J, Duncan  GJ.  The effects of poverty on children. Future Child. 1997;7(2):55-71.
PubMedArticle
3.
Yoshikawa  H, Aber  JL, Beardslee  WR.  The effects of poverty on the mental, emotional, and behavioral health of children and youth: implications for prevention. Am Psychol. 2012;67(4):272-284.
PubMedArticle
4.
Lipina  SJ, Martelli  MI, Vuelta  B, Colombo  JA.  Performance on the A-not-B task of Argentinean infants from unsatisfied and satisfied basic needs homes. Rev Interam Psicol. 2005;39:49-60.
5.
Hart  B, Risley  TR. Meaningful Differences in the Everyday Experience of Young American Children. Baltimore, MD: Paul H Brookes Publishing; 1995.
6.
US Census Bureau. Child poverty in the United States 2009 and 2010: selected race groups and Hispanic origin. 2011. http://www.census.gov/prod/2011pubs/acsbr10-05.pdf. Accessed February 21, 2013.
7.
van Praag  H, Kempermann  G, Gage  FH.  Neural consequences of environmental enrichment. Nat Rev Neurosci. 2000;1(3):191-198.
PubMedArticle
8.
Noble  KG, Houston  SM, Kan  E, Sowell  ER.  Neural correlates of socioeconomic status in the developing human brain. Dev Sci. 2012;15(4):516-527.
PubMedArticle
9.
Hanson  JL, Chandra  A, Wolfe  BL, Pollak  SD.  Association between income and the hippocampus. PLoS One. 2011;6(5):e18712.
PubMedArticle
10.
Jednoróg  K, Altarelli  I, Monzalvo  K,  et al.  The influence of socioeconomic status on children’s brain structure. PLoS One. 2012;7(8):e42486.
PubMedArticle
11.
National Research Council and Institute of Medicine. Family resources. In: Shonkoff  JP, Phillips  D, eds. From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington, DC: National Academy Press; 2000.
12.
Vyas  A, Mitra  R, Shankaranarayana Rao  BS, Chattarji  S.  Chronic stress induces contrasting patterns of dendritic remodeling in hippocampal and amygdaloid neurons. J Neurosci. 2002;22(15):6810-6818.
PubMed
13.
Eiland  L, Ramroop  J, Hill  MN, Manley  J, McEwen  BS.  Chronic juvenile stress produces corticolimbic dendritic architectural remodeling and modulates emotional behavior in male and female rats. Psychoneuroendocrinology. 2012;37(1):39-47.
PubMedArticle
14.
Sánchez  MM, Ladd  CO, Plotsky  PM.  Early adverse experience as a developmental risk factor for later psychopathology: evidence from rodent and primate models. Dev Psychopathol. 2001;13(3):419-449.
PubMedArticle
15.
Sapolsky  RM.  Mothering style and methylation. Nat Neurosci. 2004;7(8):791-792.
PubMedArticle
16.
Nelson  CA  III, Zeanah  CH, Fox  NA, Marshall  PJ, Smyke  AT, Guthrie  D.  Cognitive recovery in socially deprived young children: the Bucharest Early Intervention Project. Science. 2007;318(5858):1937-1940.
PubMedArticle
17.
Tottenham  N, Hare  TA, Quinn  BT,  et al.  Prolonged institutional rearing is associated with atypically large amygdala volume and difficulties in emotion regulation. Dev Sci. 2010;13(1):46-61.
PubMedArticle
18.
Lupien  SJ, Fiocco  A, Wan  N,  et al.  Stress hormones and human memory function across the lifespan. Psychoneuroendocrinology. 2005;30(3):225-242.
PubMedArticle
19.
Bremner  JD, Randall  P, Scott  TM,  et al.  MRI-based measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. Am J Psychiatry. 1995;152(7):973-981.
PubMed
20.
Bremner  JD, Randall  P, Vermetten  E,  et al.  Magnetic resonance imaging–based measurement of hippocampal volume in posttraumatic stress disorder related to childhood physical and sexual abuse: a preliminary report. Biol Psychiatry. 1997;41(1):23-32.
PubMedArticle
21.
Luby  JL, Barch  DM, Belden  A,  et al.  Maternal support in early childhood predicts larger hippocampal volumes at school age. Proc Natl Acad Sci U S A. 2012;109(8):2854-2859.
PubMedArticle
22.
Luby  JL, Belden  AC, Pautsch  J, Si  X, Spitznagel  E.  The clinical significance of preschool depression: impairment in functioning and clinical markers of the disorder. J Affect Disord. 2009;112(1-3):111-119.
PubMedArticle
23.
McLoyd  VC.  Socioeconomic disadvantage and child development. Am Psychol. 1998;53(2):185-204.
PubMedArticle
24.
Egger  H, Ascher  B, Angold  A. Preschool Age Psychiatric Assessment (PAPA): Version 1.1. Durham, NC: Center for Developmental Epidemiology, Department of Psychiatry and Behavioral Sciences, Duke University Medical Center; 1999.
25.
Angold  A, Costello  EJ.  A test-retest reliability study of child-reported psychiatric symptoms and diagnoses using the Child and Adolescent Psychiatric Assessment (CAPA-C). Psychol Med. 1995;25(4):755-762.
PubMedArticle
26.
Angold  A, Costello  EJ.  The Child and Adolescent Psychiatric Assessment (CAPA). J Am Acad Child Adolesc Psychiatry. 2000;39(1):39-48.
PubMedArticle
27.
Carskadon  MA, Acebo  C.  A self-administered rating scale for pubertal development. J Adolesc Health. 1993;14(3):190-195.
PubMedArticle
28.
Petersen  AC, Crockett  L, Richards  M, Boxer  A.  A self-report measure of pubertal status: reliability, validity and initial norms. J Youth Adolesc. 1988;17(2):117-133. doi:10.1007/BF01537962.Article
29.
Carmichael-Olson  H, Greenberg  M, Slough  N. Manual for the Waiting Task. Seattle, WA: University of Washington; 1985.
30.
Cole  PM, Teti  LO, Zahn-Waxler  C.  Mutual emotion regulation and the stability of conduct problems between preschool and early school age. Dev Psychopathol. 2003;15(1):1-18.
PubMedArticle
31.
Cole  PM, Dennis  TA, Smith-Simon  KE, Cohen  LH.  Preschoolers' emotion regulation strategy understanding: relations with emotion socialization and child self-regulation. Soc Dev. 2009;18(2):324-352. doi:10.1111/j.1467-9507.2008.00503.x.Article
32.
Mirabile  SP, Scaramella  LV, Sohr-Preston  SL, Robison  SD.  Mothers’ socialization of emotion regulation: the moderating role of children’s negative emotional reactivity. Child Youth Care Forum. 2009;38(1):19-37.
PubMedArticle
33.
Csernansky  JG, Wang  L, Jones  D,  et al.  Hippocampal deformities in schizophrenia characterized by high dimensional brain mapping. Am J Psychiatry. 2002;159(12):2000-2006.
PubMedArticle
34.
Haller  JW, Banerjee  A, Christensen  GE,  et al.  Three-dimensional hippocampal MR morphometry with high-dimensional transformation of a neuroanatomic atlas. Radiology. 1997;202(2):504-510.
PubMed
35.
Beg  MF, Miller  MI, Trouvé  A, Younes  L.  Computing large deformation metric mappings via geodesic flows of diffeomorphisms. Int J Comput Vis. 2005;61(2):139-157. doi:10.1023/B:VISI.0000043755.93987.aa.Article
36.
Hayes  AF. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York, NY: Guilford Press; 2013.
37.
Preacher  KJ, Hayes  AF.  Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879-891.
PubMedArticle
38.
MacKinnon  DP, Lockwood  CM, Hoffman  JM, West  SG, Sheets  V.  A comparison of methods to test mediation and other intervening variable effects. Psychol Methods. 2002;7(1):83-104.
PubMedArticle
39.
Biglan  A, Flay  BR, Embry  DD, Sandler  IN.  The critical role of nurturing environments for promoting human well-being. Am Psychol. 2012;67(4):257-271.
PubMedArticle
40.
Seckl  JR, Meaney  MJ.  Glucocorticoid programming. Ann N Y Acad Sci. 2004;1032:63-84.
PubMedArticle
41.
Heckman  JJ.  Skill formation and the economics of investing in disadvantaged children. Science. 2006;312(5782):1900-1902.
PubMedArticle
42.
Ge  X, Lorenz  FO, Conger  RD, Elder  GH, Simons  RL.  Trajectories of stressful life events and depressive symptoms during adolescence. Dev Psychol. 1994;30(4):467-483. doi:10.1037//0012-1649.30.4.467.Article
Original Investigation
December 2013

The Effects of Poverty on Childhood Brain DevelopmentThe Mediating Effect of Caregiving and Stressful Life Events

Author Affiliations
  • 1Department of Psychiatry, Washington University School of Medicine in St Louis, Missouri
  • 2Department of Radiology, Mallinckrodt Institute of Radiology, Washington University School of Medicine in St Louis, Missouri
  • 3Department of Psychology, Washington University in St Louis, Missouri
  • 4Program in Neuroscience, Washington University in St Louis, Missouri
JAMA Pediatr. 2013;167(12):1135-1142. doi:10.1001/jamapediatrics.2013.3139
Abstract

Importance  The study provides novel data to inform the mechanisms by which poverty negatively impacts childhood brain development.

Objective  To investigate whether the income-to-needs ratio experienced in early childhood impacts brain development at school age and to explore the mediators of this effect.

Design, Setting, and Participants  This study was conducted at an academic research unit at the Washington University School of Medicine in St Louis. Data from a prospective longitudinal study of emotion development in preschool children who participated in neuroimaging at school age were used to investigate the effects of poverty on brain development. Children were assessed annually for 3 to 6 years prior to the time of a magnetic resonance imaging scan, during which they were evaluated on psychosocial, behavioral, and other developmental dimensions. Preschoolers included in the study were 3 to 6 years of age and were recruited from primary care and day care sites in the St Louis metropolitan area; they were annually assessed behaviorally for 5 to 10 years. Healthy preschoolers and those with clinical symptoms of depression participated in neuroimaging at school age/early adolescence.

Exposure  Household poverty as measured by the income-to-needs ratio.

Main Outcomes and Measures  Brain volumes of children’s white matter and cortical gray matter, as well as hippocampus and amygdala volumes, obtained using magnetic resonance imaging. Mediators of interest were caregiver support/hostility measured observationally during the preschool period and stressful life events measured prospectively.

Results  Poverty was associated with smaller white and cortical gray matter and hippocampal and amygdala volumes. The effects of poverty on hippocampal volume were mediated by caregiving support/hostility on the left and right, as well as stressful life events on the left.

Conclusions and Relevance  The finding that exposure to poverty in early childhood materially impacts brain development at school age further underscores the importance of attention to the well-established deleterious effects of poverty on child development. Findings that these effects on the hippocampus are mediated by caregiving and stressful life events suggest that attempts to enhance early caregiving should be a focused public health target for prevention and early intervention. Findings substantiate the behavioral literature on the negative effects of poverty on child development and provide new data confirming that effects extend to brain development. Mechanisms for these effects on the hippocampus are suggested to inform intervention.

The deleterious effects of poverty on child development have been well established in psychosocial research, with poverty identified as being among the most powerful risk factors for poor developmental outcomes.1,2 Children exposed to poverty have poorer cognitive outcomes and school performance, and they are at higher risk for antisocial behaviors and mental disorders.3 Notably, developmental deficits associated with poverty have been detected as early as infancy.4,5 Despite these established and alarming poor developmental outcomes, to date, there have been little neurobiological data in humans to inform the mechanism(s) of these relationships. This represents a critical gap in the literature and an urgent national and global public health problem based on statistics that more than 1 in 5 children are now living below the poverty line in the United States alone.6

The tangible effect of early environmental exposures on brain development has been well established in laboratory animals. Animals exposed to enriched environments high in stimulation have been shown to display increased hippocampal cell proliferation and neurogenesis compared with those reared in relative deprivation.7 Poverty represents a form of human deprivation that may parallel this animal model, raising the question of whether low levels of stimulation and relative psychosocial neglect associated with poverty have a similar negative effect on human brain development. A few studies have directly investigated the relationship between poverty and childhood brain development. Consistent with animal data, Noble and colleagues8 detected a smaller hippocampus and amygdala in 5- to 17-year-old children living in poverty. In a large community sample, Hanson et al9 reported smaller hippocampal gray matter volumes among children from lower-income backgrounds. Lower socioeconomic status was associated with smaller hippocampal gray matter volumes bilaterally in a small sample of healthy 10-year-old children.10

These findings suggest that exposure to poverty has deleterious effects on human amygdala and hippocampal development. These brain regions, involved in stress regulation and emotion processing, are known to be sensitive to environmental stimuli. However, what remains unclear, and critical to addressing this public health problem, are the specific factors that mediate this association in humans. Poverty is strongly associated with a number of risk factors implicated in poor developmental outcomes in behavioral studies, such as unsupportive parenting, poor nutrition and education, lack of caregiver education, and high levels of traumatic and stressful life events, making the income-to-needs ratio a good proxy for cumulative developmental stress.11 These and other associated factors could serve as mechanisms mediating the negative impact of poverty on brain development. It is unclear whether such mediators of risk are also operative at the neurobiological level in humans.

Experimental studies of the neurobiological impact of poverty cannot be conducted in humans for obvious ethical reasons. However, the negative effect of early unsupportive parenting in the form of maternal deprivation and stress on hippocampal and amygdala development has been well established in rodents. Stress paradigms in rodent models have been associated with elevated anxiety and contrasting alterations in neuronal morphology in the hippocampus and amygdala, with dendritic atrophy observed in the hippocampus and increased dendritic arborization in the amygdala.12,13 Developing rodents deprived of maternal nurturance show decreased hippocampal volume and altered stress reactivity.14 An epigenetic mechanism for this effect has been elaborated.15 Importantly, controlled trials that have randomized institutionalized toddlers to early therapeutic foster care vs institutionalization have documented the deleterious effects of early relative deprivation on cognitive outcomes.16

A few studies have investigated the effects of early caregiving on amygdala and hippocampal volumes in children. Consistent with animal data, Tottenham et al17 showed an association between early institutional rearing and larger amygdala volumes. While animal data would suggest that institutional rearing would lead to reduced hippocampal volume, some investigators have suggested that such effects may not become evident in humans until later in life.18 Consistent with this, decreased hippocampal volumes have been found in numerous studies of adults who experienced high levels of childhood stress/trauma.19,20 In spite of this hypothesized delayed hippocampal effect, a positive impact of early supportive parenting on hippocampal development has been detected as early as school age.21

To investigate the effects of poverty on childhood brain development and to begin to inform the mediating mechanisms of these negative effects, we investigated associations between poverty and total white and total cortical gray matter volume, as well as hippocampus and amygdala volumes, in a sample of children ages 6 to 12 years followed up longitudinally since the preschool period. Based on the behavioral data in humans and the neurobiological data in animals, we hypothesized that an effect of poverty on these brain volume outcomes would be found. We also hypothesized that key variables associated with poverty and known to negatively impact child development outcomes, including caregiving support, caregiver education, and stressful life events, would mediate the association between poverty and brain volumes.

Methods
Participants

A total of 145 right-handed children were recruited from a larger sample enrolled in the 10-year longitudinal Preschool Depression Study (N = 305 at baseline). The larger sample was recruited from metropolitan St Louis day cares and preschools using a screening checklist to include healthy children and to oversample preschoolers with depressive symptoms. Subjects and their caregivers participated in 3 to 6 comprehensive annual diagnostic and developmental assessments prior to the first neuroimaging session (see article by Luby et al22 for full description). Subjects were screened for standard imaging contraindications. There were no significant differences on demographic variables between the imaging subsample and the original sample. Table 1 shows the characteristics of the study sample. All study procedures were reviewed and approved by the institutional review board at the Washington University School of Medicine in St Louis. Written informed consent was obtained from parents, and assent was obtained from children.

Measures

The income-to-needs ratio was operationalized as the total family income divided by the federal poverty level based on family size in the year most proximal to data collection.23 The value was calculated through baseline Preschool Depression Study data of caregiver-reported total family income and total number of people living in the household.

Psychiatric Diagnostic Status, Stressful Life Events, and Caregivers’ Education

Subjects were assessed annually using the Preschool Age Psychiatric Assessment (parent interview, age 3-8 years) and Child and Adolescent Psychiatric Assessment (parent/child interview, age ≥9 years).24 Both measures also reliably capture experiences of stressful and traumatic life events.25,26 Life events between baseline and time of scan were used for the current analysis.

Tanner Staging Questionnaire

The Tanner staging questionnaire was used to measure children’s pubertal status at the time of the scan.27,28

Parental Supportive/Hostile Caregiving

At the second assessment wave (ages 4-7 years), parent-child dyads were observed interacting during the waiting task, a structured task designed to elicit mild dyadic stress.29 This laboratory task requires the child to wait for 8 minutes before opening a brightly wrapped gift within arm’s reach. Children are told that they can open the gift once their caregiver completes questionnaires. Blind raters, trained to reliability, coded the interaction for caregivers’ use of both supportive (eg, praising the child for waiting) and hostile (eg, threats about negative consequences) strategies. This task has acceptable psychometric properties and is a well-validated and widely used parenting measure.2932 Hostility scores were subtracted from support scores to provide a difference score.

Magnetic Resonance Image Acquisition

Two 3-dimensional T1-weighted magnetization-prepared rapid gradient echo scans were acquired on a Siemens 3.0-T Tim Trio scanner without sedation (sagittal acquisition; repetition time = 2300 milliseconds; echo time = 3.16 milliseconds; inversion time = 1200 milliseconds; flip angle = 8°; 160 slices; 256 × 256 matrix; field of view = 256 mm; 1.0-mm3 voxels; total time = 12:36 min).

Image Analyses
Whole Brain

Total gray and white matter volumes were obtained using FreeSurfer version 5.1. The white and pial FreeSurfer surfaces were visually inspected and were regenerated with manual intervention when necessary. Cortical gray matter volume was defined as the volume between the pial and white matter surfaces. White matter volume was calculated by subtracting the subcortical and ventricular volumes from the volume bounded by the white matter surface.

Amygdala and Hippocampus

The hippocampus was segmented by an automated high-dimensional template-based transformation. The manual template, delineated on 1 subject with typical anatomy, was reviewed by neuroanatomical gold standard experts following boundary definitions.33,34 The gold template surface, generated from the manual template, included gray and white matter. Subject images, landmarked by an experienced rater blind to subject characteristics, were aligned to the template through an affine transformation followed by a nonlinear large deformation transformation to increase alignment precision. After matching subject-template voxel intensities, a high-dimensional subject-template transformation was generated through large deformation diffeomorphic metric mapping.35 Results were blindly reviewed (C.B.) for surface quality. The reliability of this process is well established.34 The amygdala segmentation paralleled the methodology of the hippocampus.

Statistical Analyses
Potential Covariates

Pearson correlations and t tests were conducted to explore variation in brain volumes related to children’s sex, age, pubertal status, history of psychiatric disorders (yes/no), and children’s history of psychotropic medication use (yes/no). Covariates were included in the final analyses if significant for that particular region.

Associations Between Income-to-Needs Ratio and Brain Volume

Hierarchical multiple linear regression analyses were conducted to test whether the income-to-needs ratio predicted brain volumes. For all models, covariates were entered at step 1 and the income-to-needs ratio was entered at step 2.

Mediators of the Hypothesized Associations Between Income-to-Needs Ratios and Brain Volumes

Three variables were tested as possible mediators of the relations between baseline income-to-needs ratios and children’s brain volumes (Figure 1). Mediators were tested by calculating bias-corrected 95% CIs using bootstrapping with 10 000 resamples via the Process procedure for SPSS.36,37 Given that our data could not establish temporal precedence between caregivers’ income-to-needs ratio and highest level of education, we chose to use baseline data for both variables.

Results

eTable 1 in Supplement shows the results of analyses testing potential covariates. Based on these results, sex was included as a covariate in all analyses except those examining right hippocampal volume. For analyses of white matter volume, children’s age and pubertal status were also included as covariates. None of the brain volumes differed significantly in relation to children’s history of DSM-IV Axis I disorder or psychotropic medication exposure. For all analyses examining hippocampus or amygdala volumes, children’s total cortical brain volume (total white + total cortical gray) was included as a covariate to assess specificity.

Income-to-Needs Ratio Predicting Total White and Cortical Gray Matter Volumes
White Matter Volume

Children’s age, sex, and pubertal status were entered at step 1. The income-to-needs ratio was entered at step 2 and was a positive predictor of white matter volume, accounting for a significant increase in variance (change F1,137 = 8.12, P = .005). The R2adjusted for each step of the model, as well as the unstandardized regression coefficients (B), standard error (SE), and standardized regression coefficients (β), are reported in Table 2.

Cortical Gray Matter

Sex was included at step 1 of the model. The income-to-needs ratio was entered at step 2 and was a positive predictor of gray matter volume, accounting for a significant increase in variance (change F1,142 = 21.79, P < .001) (Table 2).

Income-to-Needs Ratio Predicting Left and Right Hippocampus and Amygdala Volumes

Covariates, including whole-brain volume, were entered in step 1. As seen in Table 3, for children’s left hippocampus volume, including the income-to-needs ratio at step 2 resulted in a significant increase in the amount of variance accounted for (change F1,115 = 5.76, P = .02). The income-to-needs ratio was a positive predictor of children’s left hippocampus volumes. For the right hippocampus, the increase in variance accounted for after including the income-to-needs ratio at step 2 only approached significance (change F1,119 = 2.94, P = .09). For children’s left amygdala volume, including the income-to-needs ratio at step 2 resulted in a significant increase in the amount of variance accounted for (change F1,120 = 6.28, P = .01). The income-to-needs ratio was a positive predictor of children’s left amygdala volumes. For right amygdala volumes, the increase in variance accounted for after including the income-to-needs ratio at step 2 only approached significance (change F1,127 = 2.79, P = .09).

Caregivers’ Education, Parenting, and Stressful Life Events as Mediators of the Associations Between Income-to-Needs Ratio and Brain Volumes

The analyses just described established a relationship between the income-to-needs ratio and later brain volumes. We hypothesized that there would also be indirect (ie, mediated) effects through caregivers’ education, observed use of supportive/hostile parenting, and children’s experience of stressful life events. Figure 1 provides a conceptual diagram of the meditational analyses conducted. MacKinnon and colleagues38 suggested that mediation analyses be conducted when there is a relation between a predictor and mediator (paths A1, A2, and A3 in Figure 1), as well as a relation between a mediator and outcome (paths B1, B2, and B3 in Figure 1). To be considered a mediator, the strength of the direct relation between predictor and outcome (path C in Figure 1) will be diminished when the mediator is entered into the analysis (path C’ in Figure 1). Covariates included in the meditational analyses were parallel with prior analyses and were only applied to outcome variables. Here we first established the relationship between the predictor (income-to-needs ratio) and the potential mediators (caregiver education, parenting, and life events), and then examined the relationships of the mediators to the outcome (brain volume) and, when significant, whether they reduced the direct effect of income-to-needs ratio on brain volumes.

Income-to-Needs Ratio Predicting Potential Mediators

Regression analyses confirmed that the income-to-needs ratio was significantly associated with caregivers’ education (path A1; ranges across all regions: P < .001 in all models), predicted caregiving support/hostility assessed 1 year after baseline controlling for caregivers’ education (path A2, P < .001), and predicted children’s experience of stressful life events between baseline and time of scan when covarying for caregivers’ education and supportive/hostile parenting (path A3, P < .001 in all models).

Mediators of Total White Matter and Cortical Gray Matter Volumes

Paths B1, B2, and B3 from the mediators to white matter and cortical gray matter volume were all nonsignificant (all P > .05). Thus, neither caregiving behaviors, education, nor life stress mediated the relationship between the income-to-needs ratio and cortical gray or white matter volume.

Mediators of Hippocampal Volumes

Figure 2 illustrates that 2 of the mediating variables, stressful life events (path B1) and caregiving behaviors (path B3), positively predicted children’s left hippocampus volumes. For right hippocampus volume, caregiving behavior (path B3) was the only significant mediator. When mediators were included in the model, the direct paths (ie, path C’) from the income-to-needs ratio to the left hippocampus (P > .51) and right hippocampus (P > .55) volumes were no longer significant, indicating full mediation (Figure 2). In Supplement, eTable 2 shows the mediated effects of the income-to-needs ratio on left and right hippocampus volumes.

Mediators of Amygdala Volumes

Paths B1, B2, and B3 from the mediators to left and right amygdala volumes were all nonsignificant (P > .14).

Discussion

These study findings demonstrated that exposure to poverty during early childhood is associated with smaller white matter, cortical gray matter, and hippocampal and amygdala volumes measured at school age/early adolescence. These findings extend the substantial body of behavioral data demonstrating the deleterious effects of poverty on child developmental outcomes into the neurodevelopmental domain and are consistent with prior results.8,9 Furthermore, these study findings extend the available structural neuroimaging data in children exposed to poverty by informing the mechanism of the effects of poverty on hippocampal volumes. Findings indicated that the effects of poverty on hippocampal volumes were mediated by caregiving support/hostility on both the left and right hippocampus. On the left, stressful life events also emerged as significant mediators. Caregiver education was not a significant mediator. As exposure to poverty is well known to be strongly associated with a variety of negative life experiences, the role that these risk factors appeared to play in the relationship between poverty and alterations in brain development elucidates more specific targets for prevention.

Notably, alterations in brain volume associated with poverty were detected more globally in cortical gray and white matter volume, although mediation in these regions was not identified. The finding that mediation associated with parenting and life stress was selective to the hippocampus suggests regional specificity to these mechanistic relationships. The key role of caregiver nurturance in hippocampal development and its relationship to adaptive stress responses has been well established in animal studies. Consistent findings have been provided from an earlier subgroup of this study sample suggesting that supportive parenting also plays a key role in child hippocampal development independent of income.21 Thus, the current findings add to and extend the literature underscoring the critical role of nurturance for childhood well-being.39 The finding that experiences of stressful life events also mediated the relationship between poverty and left hippocampal volume is consistent with the extensive body of animal data that have elucidated the negative effects of early stress on hypothalamic-pituitary-adrenal function and hippocampal volume.40 Understanding these mechanisms is key to the design of more targeted interventions, providing a feasible alternative to changing psychosocial status itself, a much more challenging goal that vulnerable rapidly developing young children do not have time to await.

Limitations of the current data were that the original study sample was oversampled for preschoolers with symptoms of depression, limiting generalizability. Furthermore, the relationships in the mediation model may be bidirectional. A sample with multiple waves of imaging data starting earlier in development would be necessary to adequately test directionality. Future studies with such designs and more detailed assessments of the correlates of poverty, such as nutrition, parental psychopathology, and genetic factors, are needed to further elucidate the mechanisms of risk.

We believe these findings may be useful to inform preventive interventions for this high-risk population facing a multitude of psychosocial stressors and suggest that caregiving should be a specific target. The importance of early interventions that target caregiving is underscored by studies demonstrating high cost-effectiveness through greatly enhanced long-term outcomes.41 Furthermore, children who receive more nurturing caregiving may also be protected from exposure to stressful life events, suggesting this central target may have positive ramifications on brain development.42 Considering these issues, study findings are relevant to the public policy debate on the importance of early preschool programs for young children living in poverty. The finding that the effects of poverty on hippocampal development are mediated through caregiving and stressful life events further underscores the importance of high-quality early childhood caregiving, a task that can be achieved through parenting education and support, as well as through preschool programs that provide high-quality supplementary caregiving and safe haven to vulnerable young children.

Back to top
Article Information

Corresponding Author: Joan Luby, MD, Early Emotional Development Program, Department of Psychiatry, Washington University School of Medicine in St Louis, Washington University Medical Center, Campus Box 8134, 660 S Euclid Ave, St Louis, MO 63110-1093 (lubyj@wustl.edu).

Accepted for Publication: June 9, 2013.

Published Online: October 28, 2013. doi:10.1001/jamapediatrics.2013.3139.

Author Contributions: Dr Luby had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Luby, Botteron, Barch.

Acquisition of data: Luby, Botteron, Babb, Nishino, Barch.

Analysis and interpretation of data: Luby, Belden, Botteron, Marrus, Harms, Barch.

Drafting of the manuscript: Luby, Belden, Barch.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Belden, Harms, Barch.

Obtained funding: Luby, Botteron, Barch.

Administrative, technical, or material support: Luby, Botteron, Marrus, Harms, Babb, Nishino.

Study supervision: Luby, Botteron, Barch.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by grants 2R01 MH064769-06A1 (Dr Luby) and PA-07-070 NIMH R01 (Drs Luby, Barch, and Botteron) from the National Institutes of Health. Dr Belden’s work on this manuscript was supported by grant 5K01MH090515-04 from the National Institutes of Health.

Role of the Sponsor: The funding source had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: We wish to acknowledge our child participants and their parents whose participation and cooperation made this research possible. We thank Michael I. Miller, PhD, and Tilak Ratnanather, PhD, of the Center for Imaging Science, Johns Hopkins University, for their technical assistance and insights with the implementation of the large deformation diffeomorphic metric mapping to quantify the hippocampus and amygdala volumes.

References
1.
Carneiro  PM, Heckman  JJ. Human capital policy. In: Heckman  JJ, Krueger  AB, Friedman  BM, eds. Inequality in America: What Role for Human Capital Policies? Cambridge, MA: MIT Press; 2003.
2.
Brooks-Gunn  J, Duncan  GJ.  The effects of poverty on children. Future Child. 1997;7(2):55-71.
PubMedArticle
3.
Yoshikawa  H, Aber  JL, Beardslee  WR.  The effects of poverty on the mental, emotional, and behavioral health of children and youth: implications for prevention. Am Psychol. 2012;67(4):272-284.
PubMedArticle
4.
Lipina  SJ, Martelli  MI, Vuelta  B, Colombo  JA.  Performance on the A-not-B task of Argentinean infants from unsatisfied and satisfied basic needs homes. Rev Interam Psicol. 2005;39:49-60.
5.
Hart  B, Risley  TR. Meaningful Differences in the Everyday Experience of Young American Children. Baltimore, MD: Paul H Brookes Publishing; 1995.
6.
US Census Bureau. Child poverty in the United States 2009 and 2010: selected race groups and Hispanic origin. 2011. http://www.census.gov/prod/2011pubs/acsbr10-05.pdf. Accessed February 21, 2013.
7.
van Praag  H, Kempermann  G, Gage  FH.  Neural consequences of environmental enrichment. Nat Rev Neurosci. 2000;1(3):191-198.
PubMedArticle
8.
Noble  KG, Houston  SM, Kan  E, Sowell  ER.  Neural correlates of socioeconomic status in the developing human brain. Dev Sci. 2012;15(4):516-527.
PubMedArticle
9.
Hanson  JL, Chandra  A, Wolfe  BL, Pollak  SD.  Association between income and the hippocampus. PLoS One. 2011;6(5):e18712.
PubMedArticle
10.
Jednoróg  K, Altarelli  I, Monzalvo  K,  et al.  The influence of socioeconomic status on children’s brain structure. PLoS One. 2012;7(8):e42486.
PubMedArticle
11.
National Research Council and Institute of Medicine. Family resources. In: Shonkoff  JP, Phillips  D, eds. From Neurons to Neighborhoods: The Science of Early Childhood Development. Washington, DC: National Academy Press; 2000.
12.
Vyas  A, Mitra  R, Shankaranarayana Rao  BS, Chattarji  S.  Chronic stress induces contrasting patterns of dendritic remodeling in hippocampal and amygdaloid neurons. J Neurosci. 2002;22(15):6810-6818.
PubMed
13.
Eiland  L, Ramroop  J, Hill  MN, Manley  J, McEwen  BS.  Chronic juvenile stress produces corticolimbic dendritic architectural remodeling and modulates emotional behavior in male and female rats. Psychoneuroendocrinology. 2012;37(1):39-47.
PubMedArticle
14.
Sánchez  MM, Ladd  CO, Plotsky  PM.  Early adverse experience as a developmental risk factor for later psychopathology: evidence from rodent and primate models. Dev Psychopathol. 2001;13(3):419-449.
PubMedArticle
15.
Sapolsky  RM.  Mothering style and methylation. Nat Neurosci. 2004;7(8):791-792.
PubMedArticle
16.
Nelson  CA  III, Zeanah  CH, Fox  NA, Marshall  PJ, Smyke  AT, Guthrie  D.  Cognitive recovery in socially deprived young children: the Bucharest Early Intervention Project. Science. 2007;318(5858):1937-1940.
PubMedArticle
17.
Tottenham  N, Hare  TA, Quinn  BT,  et al.  Prolonged institutional rearing is associated with atypically large amygdala volume and difficulties in emotion regulation. Dev Sci. 2010;13(1):46-61.
PubMedArticle
18.
Lupien  SJ, Fiocco  A, Wan  N,  et al.  Stress hormones and human memory function across the lifespan. Psychoneuroendocrinology. 2005;30(3):225-242.
PubMedArticle
19.
Bremner  JD, Randall  P, Scott  TM,  et al.  MRI-based measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. Am J Psychiatry. 1995;152(7):973-981.
PubMed
20.
Bremner  JD, Randall  P, Vermetten  E,  et al.  Magnetic resonance imaging–based measurement of hippocampal volume in posttraumatic stress disorder related to childhood physical and sexual abuse: a preliminary report. Biol Psychiatry. 1997;41(1):23-32.
PubMedArticle
21.
Luby  JL, Barch  DM, Belden  A,  et al.  Maternal support in early childhood predicts larger hippocampal volumes at school age. Proc Natl Acad Sci U S A. 2012;109(8):2854-2859.
PubMedArticle
22.
Luby  JL, Belden  AC, Pautsch  J, Si  X, Spitznagel  E.  The clinical significance of preschool depression: impairment in functioning and clinical markers of the disorder. J Affect Disord. 2009;112(1-3):111-119.
PubMedArticle
23.
McLoyd  VC.  Socioeconomic disadvantage and child development. Am Psychol. 1998;53(2):185-204.
PubMedArticle
24.
Egger  H, Ascher  B, Angold  A. Preschool Age Psychiatric Assessment (PAPA): Version 1.1. Durham, NC: Center for Developmental Epidemiology, Department of Psychiatry and Behavioral Sciences, Duke University Medical Center; 1999.
25.
Angold  A, Costello  EJ.  A test-retest reliability study of child-reported psychiatric symptoms and diagnoses using the Child and Adolescent Psychiatric Assessment (CAPA-C). Psychol Med. 1995;25(4):755-762.
PubMedArticle
26.
Angold  A, Costello  EJ.  The Child and Adolescent Psychiatric Assessment (CAPA). J Am Acad Child Adolesc Psychiatry. 2000;39(1):39-48.
PubMedArticle
27.
Carskadon  MA, Acebo  C.  A self-administered rating scale for pubertal development. J Adolesc Health. 1993;14(3):190-195.
PubMedArticle
28.
Petersen  AC, Crockett  L, Richards  M, Boxer  A.  A self-report measure of pubertal status: reliability, validity and initial norms. J Youth Adolesc. 1988;17(2):117-133. doi:10.1007/BF01537962.Article
29.
Carmichael-Olson  H, Greenberg  M, Slough  N. Manual for the Waiting Task. Seattle, WA: University of Washington; 1985.
30.
Cole  PM, Teti  LO, Zahn-Waxler  C.  Mutual emotion regulation and the stability of conduct problems between preschool and early school age. Dev Psychopathol. 2003;15(1):1-18.
PubMedArticle
31.
Cole  PM, Dennis  TA, Smith-Simon  KE, Cohen  LH.  Preschoolers' emotion regulation strategy understanding: relations with emotion socialization and child self-regulation. Soc Dev. 2009;18(2):324-352. doi:10.1111/j.1467-9507.2008.00503.x.Article
32.
Mirabile  SP, Scaramella  LV, Sohr-Preston  SL, Robison  SD.  Mothers’ socialization of emotion regulation: the moderating role of children’s negative emotional reactivity. Child Youth Care Forum. 2009;38(1):19-37.
PubMedArticle
33.
Csernansky  JG, Wang  L, Jones  D,  et al.  Hippocampal deformities in schizophrenia characterized by high dimensional brain mapping. Am J Psychiatry. 2002;159(12):2000-2006.
PubMedArticle
34.
Haller  JW, Banerjee  A, Christensen  GE,  et al.  Three-dimensional hippocampal MR morphometry with high-dimensional transformation of a neuroanatomic atlas. Radiology. 1997;202(2):504-510.
PubMed
35.
Beg  MF, Miller  MI, Trouvé  A, Younes  L.  Computing large deformation metric mappings via geodesic flows of diffeomorphisms. Int J Comput Vis. 2005;61(2):139-157. doi:10.1023/B:VISI.0000043755.93987.aa.Article
36.
Hayes  AF. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York, NY: Guilford Press; 2013.
37.
Preacher  KJ, Hayes  AF.  Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879-891.
PubMedArticle
38.
MacKinnon  DP, Lockwood  CM, Hoffman  JM, West  SG, Sheets  V.  A comparison of methods to test mediation and other intervening variable effects. Psychol Methods. 2002;7(1):83-104.
PubMedArticle
39.
Biglan  A, Flay  BR, Embry  DD, Sandler  IN.  The critical role of nurturing environments for promoting human well-being. Am Psychol. 2012;67(4):257-271.
PubMedArticle
40.
Seckl  JR, Meaney  MJ.  Glucocorticoid programming. Ann N Y Acad Sci. 2004;1032:63-84.
PubMedArticle
41.
Heckman  JJ.  Skill formation and the economics of investing in disadvantaged children. Science. 2006;312(5782):1900-1902.
PubMedArticle
42.
Ge  X, Lorenz  FO, Conger  RD, Elder  GH, Simons  RL.  Trajectories of stressful life events and depressive symptoms during adolescence. Dev Psychol. 1994;30(4):467-483. doi:10.1037//0012-1649.30.4.467.Article
×