Internalizing problems of anxiety, depressive symptoms, and withdrawal represent a major obstacle for children’s successful navigation of social and academic demands in their neighborhoods and schools [1
]. Epidemiological studies suggest that low-income, ethnic minority children are at higher risk of elevated levels of internalizing problems than are their more advantaged counterparts [2
Recently, an innovative and rapidly emerging body of evidence across fields of developmental, clinical, and neuroscientific research points to compromises in multiple components of children’s emotional regulation (ER) as key predictors of their risk of internalizing problems [5
]. In the following paper, we draw together methods and findings from these emerging literatures to consider three domains of emotional regulation that reflect an integrated, theoretically-driven definition of ER. Specifically, we draw on recent empirical frameworks suggesting that ER incorporates human attentional orienting to and appraisal of an emotionally evocative stimulus, along with the modulation of accompanying affective response through higher-order executive or cognitive control [8
]. We capitalize on neuroscientific methods usually deployed in laboratory settings to directly assess low-income children’s ER. Drawing from these recent advances in affective neuroscience, we examine whether three candidate neurocognitive mechanisms of emotion regulation (ER)—including: (i) attention bias to threat (ABT); (ii) accuracy of facial emotion appraisal (FEA); and (iii) subjective negative affect (NA)—appear to be disrupted in the context of poverty. We investigate whether they are individually, and additively, significantly predictive of internalizing problems among a clinically at-risk sample of low-income, ethnic minority children, ages 8 to 11 years (N
What prior research serves to inform our choice of these three ER processes? Our focus on children’s ABT is based on the theory that humans are clearly evolutionarily “wired” to orient and attend to stimuli that have an emotional valence, with extensive laboratory studies demonstrating that adults and children tend to orient to threatening stimuli more quickly than to non-threatening stimuli within a given visual array [10
]. To measure those attention processes at neurobiological as well as behavioral levels, neuroscientists have pioneered the use of tasks such as the dot probe paradigm, where individuals are asked to identify the location of a target (an asterisk or “dot”) on a computer screen across a large number of trials. Across those trials, pairs of emotionally neutral versus emotionally negative visual images (such as pictures of a kitchen toaster versus of a ferocious, snarling dog) are first presented for very brief periods of time, allowing investigators to compare participants’ response latencies when the location of the dot is congruent versus incongruent with that of emotionally negative stimuli (see the Measures section below, for greater detail). As reflected through the use of the dot probe paradigm in those studies, emotional information clearly not only captures, but also holds individuals’ attention, making it harder for them to disengage from emotionally valenced stimuli than non-emotional stimuli. Higher ABT using the dot probe paradigm is associated with difficulty modulating anxiety and fear, as illustrated by a growing literature on attention to threat cues among individuals struggling with mood disorders, e.g., high anxiety [12
]. Based on this emerging literature, we expect that children’s attention bias to emotionally negative and threatening cues using the dot probe paradigm is likely predictive of significantly higher risk for internalizing behavioral difficulty [14
]. This process may be heightened for children living in households struggling with chronic poverty; as such, children’s increased risk of exposure to adversity (such as greater levels of community and family violence and instability) may exacerbate (or moderate) the ways that biased attention towards threat may increase their risk for psychopathology [16
As children transition from early through middle childhood, they develop increasing skill in not only orienting to emotional stimuli but also in identifying and labeling others’ emotions [18
]. Individual differences in FEA are predictive of grave social and clinical outcomes, including higher risk of rejection from peers, feelings of failure in social situations, social isolation, and loneliness [6
]. In addition, lower levels of accuracy in identifying emotions by using appropriate emotion labels has been found to be robustly predictive of higher risk of internalizing difficulty among younger children [20
]. We expect to find similar relations between higher accuracy in emotion appraisal and lower levels of parent-reported internalizing problems for our sample of low-income, urban-residing, ethnic minority students in middle childhood. Exposure to chronic poverty-related adversity may exacerbate this process, serving as a statistical moderator of the link between NA and children’s internalizing problems.
A third key component of ER is the extent to which children not only appraise situations as upsetting, distressing or fear-inducing but also appraise their own internal feeling states or mood as more negative [22
]. Children vary greatly in their struggles and successes with felt experiences of NA [23
]. Recent experimental evidence from studies with adults highlights ways in which emotional states can be internally generated through thoughts or internally evoked images, and that those emotional states are accompanied by matching physiological responses (such as changes in eye blink startle) that support the veridicality of subjective report of NA [24
]. Negative mood is a hallmark of clinical diagnoses of anxiety and depression among adults and adolescents; recent analyses of children’s NA in middle childhood suggest that linkages between children’s subjective NA and parent-reported internalizing problem may be more complex, with considerable discrepancies between their reports [25
]. This study provides us with the opportunity to test the predictive role of children’s NA for their internalizing problems, even after controlling for their accuracy in identifying negative emotions with a large community sample of children.
Multiple theories of emotion highlight the coherent neurobiological organization and coordination of these complex processes as fundamental to the framework of “regulation” [26
]. Based on this theoretical framework, children’s behavioral performance on these three components of ER (i.e., ABT, FEA and NA) would be expected to be at least moderately correlated. To date, however, few studies of emotion regulation consider all three of these multiple components, working additively in predicting children’s internalizing problems. Even fewer studies test whether and how these basic models of ER are generalizable to low-income, ethnic minority children at high risk for elevated levels of internalizing problems. This paper attempts to remedy those empirical gaps, carrying out analyses on the ways these ER constructs serve as independent, additive predictors of internalizing problems among the low-income children in our sample in the 5th grade. In addition, we consider whether children’s chronic exposure to poverty-related adversity from early to middle childhood serves to moderate or exacerbate the role of each of those ER processes for their risk of internalizing problems at that 5th grade assessment.
Previous studies have yielded clear evidence that higher exposure to adversities related to living in poverty is associated with alterations in parasympathetic (adrenocortical), emotional, and behavioral response among stressor-exposed children [27
]. Analyses of the National Longitudinal Survey of Youth, for example, demonstrate that children exposed to chronic poverty are at significantly greater risk of internalizing behavioral problems over time [29
]. Although disparate research studies have established links between exposure to chronic poverty, children’s ER skills, and their risk of higher internalizing behavioral difficulties, these studies have not yet been extensively linked or integrated. In the following study, we examine ways that three dimensions of ER and chronic poverty-related adversity may work alone and in combination to predict children’s internalizing problems.
For all sets of analyses, it is important to rule out the possibility of other key hypothesized sources of influence in our models. For example, prior research has also highlighted the role of temperament in contributing to individual differences in children’s ER skills and risk of internalizing behavior problems [30
]. For this reason, we include measurement of symptoms of sadness and anxiety in the preschool period (as indicated by parental reports of internalizing symptoms when children were 2–5 years of age) as a rough proxy for temperamental disposition towards negative reactivity. We also control for a host of additional individual, family, and sociocultural factors that might serve as proxies for unobserved (or “omitted”) variables, including child age, child gender, child‘s racial/ethnic identity as African American versus Latina/o, and parent age at baseline.
In sum, we test a number of hypotheses in the following study. First, we hypothesize that children’s emotional regulatory processes involving attention bias to threat, appraisal of facial emotion, and negative affectivity would be moderately intercorrelated. Second, we hypothesized that individual differences in children’s internalizing problems at that 6-year-follow-up assessment would be significantly predicted by those emotional regulatory processes. Third, we hypothesized that the chronicity of children’s exposure to poverty-related adversity from early to middle childhood would serve as an additional statistically significant predictor (and might also moderate, or exacerbate the role of ER difficulties) in predicting children’s internalizing problems in middle childhood. We sought to test those hypotheses using statistically conservative models that would take into account early childhood environmental and constitutional factors that might serve as potential confounds.
Our results suggest that a small but significant proportion of the children in our CSRP sample were reported by their parents to struggle with persistent feelings of sadness, worry, and withdrawal as they transition through middle childhood. Our inclusion of neurocognitively oriented assessments helped to elucidate that children showed wide variability in their ability to orient and focus their attention to neutral versus threatening stimuli, and in their ability to accurately appraise facial information. In addition, the low-income children in our sample reported generally low levels of state- and trait-level anxiety during direct assessments. Surprisingly, children’s performance on these three candidate ER processes were statistically uncorrelated, suggesting that they operate independently. This finding differs from previous research linking high levels of negative affectivity with more biased attention to threat (particularly for children with lower levels of regulatory skill, or effortful control) [41
]. We look forward to examining whether this set of relations emerges as children transition from middle to high school and into adolescence.
In this study, we tested whether those three neurocognitively-linked processes of emotion regulation worked additively to predict low-income children’s internalizing symptoms. Consonant with prior research, we found that children’s attentional bias away from threat, their lower accuracy in identifying negative facial emotions, and their higher levels of self-reported negative affectivity served as key predictors of their higher behavioral symptoms of sadness, depression, and worry. Results also suggest that children’s chronic exposure to different types of poverty-related adversity over a 6-year period (such as families’ struggles to pay bills and difficulty making ends meet) served as a robust predictor of their risk of internalizing behavior problems in middle childhood. Importantly, poverty-related adversity positively predicted children’s higher levels of internalizing symptoms of sadness, worry, and withdrawal even after statistically taking into account a wide range of potential confounds, including their earlier temperamental negative reactivity in the preschool period. These findings are in keeping with prior research suggesting that insufficient family income (and the multiple stressors that accompany income poverty) take a harsh toll on children’s mental health [42
Our use of multiple methods across multiple reporters to examine low-income children’s emotion regulation is a strength of this study. In addition to its methodological strengths, this study has several limitations. For example, our models of emotion regulation and risk of internalizing problems among low-income children would have benefited from inclusion of more fine-grained clinical diagnostic measures as parents’ reports of children’s internalizing symptoms have been found to underestimate the severity of child psychiatric risk [39
]. In addition, our behavioral data using neuroscientific, lab-developed protocols are likely to be empirically “noisy”, given that multiple sources of measurement error may have been introduced through our effort to collect those data in field-based settings. In our view, these represent a set of important methodological challenges when investigating basic neurobiologically-anchored processes among children facing high risk of exposure to poverty-related adversity. We are confident that these challenges can be overcome in future research and with greater inclusion of low-income children in clinical neuroscientific studies.