Socioeconomic Status Inequalities Partially Mediate Racial and Ethnic Differences in Children’s Amygdala Volume

Background: While race/ethnicity and socioeconomic status (SES) impact brain structures such as the amygdala, less is known on whether or not family SES partially explains why amygdala volume is smaller for racial and ethnic minority groups. Purpose: This study tested the mediating effects of family SES on racial and ethnic differences in right and left amygdala volume. Methods: We borrowed the structural Magnetic Resonance Imaging (sMRI) data of the Children Brain Cognitive Development (ABCD) study, a brain imaging investigation of childhood brain development in the US. The total sample was 8977, 9-10-year-old children. The independent variables were race and ethnicity. The primary outcomes were right and left amygdala volume. Age, sex, household size, and marital status were the covariates. Multiple SES indicators such as family income, subjective family SES, parental employment, parental education, and neighborhood income were the mediators. To analyze the data, we used regression models without and with our mediators. Sobel test was used to test if these mediational paths are statistically significant. Results: Black and Latino children had smaller amygdala sizes than non-Latino White children. The effects of race and ethnicity on amygdala volume were partially mediated by SES indicators, suggesting that one of the many reasons Black and Latino children have smaller volumes of right and left amygdala is their lower SES. Conclusions: For American children, lower family and neighborhood SES indicators partially, but not fully, explain smaller amygdala sizes of Black and Latino children compared to nonLatino White children.

Both racial and ethnic minority status and low SES correlate with brain function and structures within (Oshri et al., 2019) and across (Javanbakht et al., 2015) regions. One of the brain regions and structures that correlates with race/ethnicity and low SES is the amygdala (Evans et al., 2016;Javanbakht et al., 2016;Javanbakht et al., 2015). Racial and /ethnicity group membership and low SES are linked to the amygdala (Evans et al., 2016;Javanbakht et al., 2016;Javanbakht et al., 2015). Given the central role of amygdala reactivity in regulating emotions, behaviors, and social relations, there is a particular interest in understanding the additive effects of racial and ethnic group membership and low SES on the amygdala and associated psychopathology and behavioral problems (Dotterer, Hyde, Swartz, Hariri, & Williamson, 2017;Gard, Waller, et al., 2018;Morawetz, Bode, Baudewig, & Heekeren, 2017;Szczepanik et al., 2016;Venta et al., 2018).
From a theoretical perspective, the scarcity hypothesis can explain why low SES and racial and ethnic minority status deteriorate children's brain development. According to this hypothesis, racial minorities' low SES is a proxy of early adversity and a lack of resources that all operate as risk factors for low childhood development. In this view, stress and adversity and low access to resources are the underlying mechanisms that explain the SESbrain development link (Yaple & Yu, 2019 ). Low SES is also a mediator (an underlying mechanism) of racial and ethnic disparities that evolve throughout a child's development (Assari, 2018a;Assari, 2018;Assari, 2018b).
Research has established race/ethnicity and SES differences in the development of various brain regions In a study, individuals with low family income, at age nine, had reduced dorsolateral, ventrolateral, and prefrontal cortex activity. They also showed weaker amygdala activation suppression during effortful emotion regulation at age 24 (P. Kim et al., 2013). Javanbakht et al. (2015) showed that low family SES was associated with a higher amygdala and medial prefrontal cortical response to threatening faces, effects that were independent of adulthood SES. Another study showed a link between family SES during childhood and reduced the amygdala's connectivity with the hippocampus, posterior cingulate, superior frontal cortex, putamen, and lingual gyrus (Barch et al., 2016).
However, there is a need to advance the current literature on the additive effects of race/ethnicity and family SES on the right and left amygdala size. Our study tested the mediating effect of family SES on the association between race/ethnicity and the amygdala volume to extend the existing knowledge on social and economic mechanisms that link race/ethnicity to children's brain structure.
The ABCD is the most extensive brain imaging study of childhood brain development in the US (Alcohol  , 2018). This analysis included 8977 non-twin children between the ages of 9 and 10 who had complete data on all our study variables; however, they could be of any race or ethnicity.

Primary Outcome
The primary outcome was right and left amygdala volume, measured by structural MRI. Amygdala

Independent Variable
Race. Race, a self-identified variable, was coded as White (reference), Black, or other races (Asian, Mixed Race, and any non-European non-African group).
Ethnicity. Parents were asked if they are of Latino ethnic background. This variable was coded as Latino = 1 and non-Latino = 0.

Mediators
Parental Education. Parents reported their years of schooling. This variable was an interval variable ranging from 1 to 21.
Subjective Family SES. Family SES in this study was financial difficulties measured by the following seven items: "In the past 12 months, has there been a time when you and your immediate family experienced any of the following:" 1) "Needed food but could not afford to buy it or could not afford to go out to get it?", 2) "Were without telephone service because you could not afford it?" 3) "Did not pay the full amount of the rent or mortgage because you could not afford it?", 4) "Were evicted from your home for not paying the rent or mortgage?", 5) "Had services turned off by the gas or electric company, or the oil company would not deliver oil because payments were not made?", 6) "Had someone who needed to see a doctor or go to the hospital but did not go because you could not afford it?" and 7) "Had someone who needed a dentist but could not go because you could not afford it?" Responses to each item were either 1 or 0. We calculated a sum score with a potential range between 0 and 7-a higher score

Confounders
Age. Age was a dichotomous variable 9 or 10. Parents reported the age of the children.
Sex. Sex was coded as 1 for males and 0 for females.
Parental Marital Status. Parental marital status was 1 for married and 0 for any other condition (reference).
Parent's Employment. A dichotomous variable recorded the employment status of the parent. The variable was coded as 0 for no employed parent in the household (reference), and 1 for at least one parent employed. This variable was self-reported by the parent who was interviewed.

Data Analysis
We used SPSS 22.00 for data analysis. Frequencies (n and %) and mean [standard deviations (SDs)] were described. To estimate bivariate analyses between the study variables, we used the Pearson correlation in the pooled sample. Then we performed two regression models. Model 1 did not have the mediators.
Model 2 was performed with the mediators-all these models controlled for all confounders. Family SES indicators were the mediators. Both models were performed in the pooled sample. The independent variables were race and ethnicity. Outcomes were right and left amygdala volume. Unstandardized regression coefficient (b), SE, and p-values were reported for each model. A p-value of equal or less than 0.05 was significant. We did not include marital status as an SES indicator, which is an accepted practice (Denney, Rogers, Krueger, & Wadsworth, 2009 (Preacher & Leonardelli, 2001).

Ethical Aspect
Our analysis was exempt from a full review. However, the ABCD study protocol was approved by the

Descriptives
The sample included 8977 9-10 years old children.      Table 3 reports the results of two pooled sample regression models with the right amygdala volume as the outcome. Model 1, which only included race and ethnicity with confounders (age, sex, and marital status),

Unadjusted Bivariate Correlations
showed that race and ethnicity are associated with right amygdala volume. Model 2 showed that family and neighborhood SES indicators partially mediate race and ethnicity's effects on the right amygdala volume.   Neighborhood Income 0.00 -0.02 0.08 -0.17 0.14 -0.20 .842 Outcome: left amygdala volume

Sobel Test of Mediation
Our Sobel tests suggested that SES indicators operate as partial mediator variables, meaning that they significantly carry some of the influences of race and ethnicity (independent variables) on the right and left amygdala volume (dependent variables). As such, some of the effects of race and ethnicity on amygdala volume are indirect and through SES (mediator variable).

Discussion
Two findings were observed. First, race/ethnicity and family SES are associated with the right and left amygdala volume. Second, the smaller amygdala size of Black and Latino children than non-Latino White children is, in part, explains the observed racial and ethnic gaps in amygdala size. However, SES partially, not fully, mediates the racial and ethnic differences in right and left amygdala volume.
Family SES has a profound effect on the brain ( Studies by Javanbakht and others have shown that high family SES shapes the amygdala function in response to negative stimuli. The first study included 52 predominantly White subjects from a prospective longitudinal study and established a link between high family SES (childhood poverty) and lower amygdala response levels to threatening faces. This effect was independent of adulthood income.
They also showed a link between family SES (childhood poverty) and reduced functional connectivity between the left amygdala and the medial prefrontal cortex (Javanbakht et al., 2015). In another study, Javanbakht et al. documented sex differences in childhood family SES effects on amygdala activity.
They found that childhood poverty status may be correlated with amygdala reactivity to fearful faces in females but not males (Javanbakht et al., 2016). In their third study lead by Evans, social adversities showed cumulative (additive) effects on amygdala structure and function (including the amygdala's response to threatening faces) (Evans et al., 2016).
Low SES and poverty are associated with a lower level of connectivity in neural networks involved in emotion regulation, especially for children who receive low levels of supportive parenting (Brody et al., 2019). Childhood poverty is linked to reduced connectivity between the amygdala and putamen, superior frontal cortex, lingual gyrus, and posterior cingulate. Childhood poverty predicts connectivity between the right lingual gyrus and the right amygdala; these brain connectivity indicators explain why early poverty predicts childhood depression (Barch et al., 2016).
Our study findings suggested that some of the racial and ethnic variations on brain structure are due to SES. We found that SES partially explains why racial and ethnic minorities have a smaller amygdala size. Theoretically speaking, one mechanism that can explain the results is everyday exposure of Black and Latino families to discrimination (Assari, Gibbons, & Simons, 2018;Assari, Miller, et al., 2018;Assari, Moghani Lankarani, & Caldwell, 2017).
We, however, did not test discrimination in the data set for a few reasons. In a study, chronic discrimination was associated with altered connections between the amygdala and other brain regions such as the insula, putamen, caudate, medial frontal gyrus, and anterior cingulate (Clark et al., 2018).
It should be emphasized again that similar to SES, we see race and ethnicity as social factors (as proxies of racism, social stratification, exposure to stress, and blocked opportunities). In other terms, race and SES reflect how and individuals and groups are treated by society. We specifically focused on the amygdala, which shapes social relations and emotion processing/ regulation. The amygdala's small size (shrink amygdala) is reported in psychopathologies and chronic exposure to stress (MacMaster et al., 2008).
For our study, we included separate SES indicators rather than a summary score. Future research could create a summary SES score by running principal component analysis. Such an approach could result in a different level of the explanatory power of SES.

Conclusions
In summary, low family and neighborhood SES indicators partially explain why racial and ethnic minority status is correlated with smaller amygdala volume. That means some of the racial and ethnic differences observed in brain structures are shaped by economic disadvantages that racial and ethnic minorities, particularly Black and Latino families, experience.