Associations of Early Socio-familial Stress with Maladaptive and Adaptive Functioning in Middle Childhood: Roles of Executive Control and Foundational Cognitive Abilities

Abstract

This study examined whether cognitive processes in preschool, conceptualized as a unitary construct of executive control (EC) as well as foundational cognitive abilities (FCA), predict both maladaptive and adaptive functioning in middle childhood and mediate associations between early childhood socio-familial stress and those functional outcomes. Performance-based, multi-dimensional, and age-appropriate measures of EC and FCA were collected in a laboratory setting from 313 preschool-age children at age 5, along with questionnaire data from children and their parents on three dimensions of early socio-familial stress and parent smoking. Parent, teacher, and child self-report data on 285 of these children were obtained when they were in Grade 3 or 4. Middle childhood data were used to create indices of maladaptive and adaptive functioning. A bi-factor structural equation modeling analysis captured distinct dimensions of preschool EC and FCA and was used to test the hypothesized pathways. EC had a statistically significant negative association with later maladaptive functioning. FCA, but not EC, served as a mediator in links between each type of family stressor and both maladaptive and adaptive functioning in middle childhood. Results suggest that EC may play a role in predicting maladaptation, whereas early childhood FCA may operate as an intervening variable in pathways from early family stressors to subsequent maladaptation as well as adaptation. Findings point to the need to address FCA by reducing early family stressors. Early interventions that enhance cognitive abilities may help reduce maladaptive and promote adaptive functioning later in childhood, thereby potentially preventing, in turn, later behavioral problems.

Children’s exposure in their first few years of life to socio-familial stress – such as poverty, poor parenting, and toxic home environments – can have both immediate and enduring associations with adverse developmental outcomes (e.g., Harden & Whittaker, 2011; Landry, Smith, Swank, Assel, & Vellet, 2001; Leventhal & Brooks-Gunn, 2000). As highlighted by a developmental psychopathology perspective (Cicchetti, 1989), which seeks to understand the multiple and multi-level (neurobiological, individual, and contextual) pathways involved in both atypical and typical development, such exposure can initiate a cascade that disrupts cognitive development and leads to maladaptive behavioral, social, and emotional functioning later in childhood (Masten & Cicchetti, 2010) that, in turn, is often a precursor to health-compromising behaviors, such as substance use and violence, in adolescence (Hawkins, Catalano, & Miller, 1992; Stoddard et al., 2013). Children’s exposure to parent substance use has also been implicated in the development of adolescent health-compromising behaviors, including substance use (Hawkins et al., 1992). From this perspective, the notion of a developmental cascade suggests that children exposed to early socio-familial stress can experience further vulnerability due to compromised development of adap tive functioning, such as social competence and academic skill acquisition (Masten et al., 2005), thereby leaving children vulnerable to negative outcomes.

To decrease children’s vulnerability, it is imperative to establish nurturing environments early in development that reduce toxic conditions and teach pro-sociality to promote well-being (Biglan, Flay, Embry, & Sanders, 2012). Although a robust literature identifies malleable risk and protective factors that predict behavioral health problems, a need remains to identify early predictors of both maladaptation and adaptation later in childhood, and the intervening processes linking socio-familial stress with subsequent functioning. This is especially true given that prior prevention research has been limited by relying primarily on broad survey-based measures of psychosocial constructs to establish basic predictive relationships. More rigorously identifying nuanced predictive pathways can inform the development of optimally-timed, preventive interventions designed to reduce risk, enhance protection, and redirect pathways away from problem behaviors and toward healthy functioning (Masten & Cicchetti, 2010).

One set of candidate predictors of later functioning that can be measured with performance-based tasks in a controlled laboratory setting is childhood cognitive processes, including executive control (EC) and foundational cognitive abilities (FCA). EC (which is also often termed executive function;¹ see Diamond, 2013) is a set of “top-down” mental abilities for intentionally directing thoughts and behaviors. EC is typically conceptualized as a multifaceted construct consisting of the related abilities of working memory, inhibitory control, and flexible shifting (Garon, Bryson, & Smith, 2008). FCA, on the other hand, refers to a more general set of abilities that is drawn upon in completing a wide range of cognitive tasks (including, but not limited to, executive tasks), and includes visual/spatial perception, sensory processing, verbal abilities, and concept formation. The distinction between EC and FCA is, therefore, hierarchical in nature. Across development, recruitment of foundational cognitive processes becomes more automatic, which lays the foundation for the more deliberate enactment of higher-order EC abilities in situations that demand goal-directed, flexible responses (Espy, 2016).

A focus on EC and its associations with maladaptive and adaptive functioning may be particularly informative for prevention science, as EC has been shown to be modifiable (Diamond, 2013; Hillman et al., 2014), whereas FCA, once established, generally shows greater stability across development (Deary, Whalley, Lemmon, Crawford, & Starr, 2000; Gow et al., 2011). However, measuring EC in isolation, which is common in the literature, risks conflating the role of specific executive abilities with that of more foundational skills that are also drawn upon in typical EC tasks. For example, non-executive features, such as verbal instructions or stimuli color and shape, may interfere with executive task demands depending on the proficiency of children’s foundational cognitive processes. To address this issue of task impurity, and to better distinguish EC and FCA, Espy (2016) implemented a bi-factor modeling approach in which latent factors represented orthogonal general (FCA) and specific (EC) constructs.

Research indicates that the preschool years represent a critical period in which the development of EC accelerates and becomes organized as a unitary factor (Espy, 2016; Fuhs & Day, 2011; Weibe et al., 2011). Of course, early EC formulation occurs within ecological context, and research shows that childhood socio-familial stress can compromise the development of EC (Nelson et al., 2015). Numerous studies have found negative associations between socio-familial stress and children’s EC (for a review, see Clark, James, & Espy, 2016). However, highlighting the importance of considering FCA, our prior work using bi-factor modeling with a sample that shares about 2/3 of the participants in the sample used in the current study (Clark et al., 2016; Espy, 2016; Nelson et al., 2018a) found that the associations between dimensions of socio-familial stress and EC were small and not statistically significant, whereas the associations between socio-familial stress and FCA were negative and significant. The current study goes beyond earlier project studies by considering EC and FCA as pathways linking early risk to later child outcomes reflecting both maladaptive and adaptive functioning.

In addition to studies linking early socio-familial stress with childhood EC, there is a rich tradition of research examining the potential consequences of either childhood EC abilities for positive development (e.g., social competence; Jacobson, Williford, & Pianta, 2011) or EC deficits for maladaptive development (e.g., academic difficulties; Blair & Razza, 2007). Nelson et al. (2019) provide a review of relevant literature, finding evidence for associations between deficits in EC (both globally and specific aspects of EC) and health-compromising behaviors (Espy, Sheffield, Wiebe, Clark, & Moehr, 2011; Han et al., 2016; Kertz, Belden, Tillman, & Luby, 2016). Much of this literature has been limited by cross-sectional or short-term longitudinal designs and a narrow focus on specific aspects of EC (e.g., inhibitory control) studied in isolation. A smaller literature has leveraged longer-term longitudinal samples and comprehensive measurement of EC using multifaceted, performance-based task batteries (e.g., Rohlf, Holl, Kirsch, Krahe, & Elsner, 2018; Sulik et al., 2015). Studies further employing a bi-factor measurement model to distinguish the unique contributions of EC and FCA to later child outcomes are especially rare. As part of our ongoing longitudinal study, Nelson et al. (2018b) found that poor EC in preschool, represented in a bi-factor model with FCA, predicted both greater depression and anxiety symptoms in late elementary school. However, no prior research, including our own, has examined early EC and FCA as joint predictors of comprehensively measured maladaptive functioning and, simultaneously, adaptive functioning later in childhood.

The current study examines unitary EC and FCA in preschool, represented in a bi-factor model with performance-based measures assessed in a laboratory setting, as predictors of composite maladaptive functioning and adaptive functioning indices in late elementary school, and as potential intervening variables linking early socio-familial stress with those indices. Analyses controlled for child sex and parent smoking. We expected that higher levels of EC and FCA would be associated with higher levels of adaptive functioning and lower levels of maladaptive functioning. Consistent with the theoretical framework articulated by Clark et al. (2016), only FCA might operate as an intervening variable because socio-familial stressors are expected to have broad effects on the development of general cognitive abilities rather than more specific effects on executive systems. Still, findings regarding EC may be particularly useful for prevention practice, as EC may be relatively more modifiable than FCA. Identifying child cognitive processes as predictors of later functioning, and as potential mediators in pathways leading from socio-familial stress and exposure to parent smoking, would suggest that preventive interventions to address these processes could help reduce risk and enhance protection in children prior to adolescence to reduce the likelihood of health-compromising behaviors.

Method

Participants

We used data on 313 children (48.9% female) who are part of an ongoing study of the development and consequences of executive control (Espy, 2016). Families with preschool age children were recruited into the project between 2005 and 2012. Recruitment took place in a small Midwestern city. Because performance on EC and FCA measurement tasks can be affected by language ability, bilingual children, for whom English might not be the primary language spoken, and children diagnosed with speech or language delays at or subsequent to enrollment were excluded. Also, children with a diagnosed developmental or behavioral disorder were excluded at enrollment, as were families who were planning to move from the area. Children who were diagnosed with a behavioral or emotional disorder after enrollment were not excluded.

The sample for the current study consists of children who completed an age 5 years and 3 months laboratory visit during the preschool phase of the project and who had a series of follow-up assessments. The sample is 63.6% European American, 13.4% Hispanic, 3.8% African American, 0.3% Asian, and 18.8% multiracial, which was similar to the population of the area in which the project took place. Because we oversampled for families with sociodemographic risk, a majority (53.7%) received public medical assistance and a substantial portion of the households (37.1%) were headed by one parent. The median household income was $42,000 per year, and 39.8% of mothers had a college or post-graduate degree. More information regarding the preschool age period of the study can be found in James, Choi, Wiebe, and Espy (2016).

Procedures

Measures of socio-familial stress and parent smoking were gathered at enrollment from in-home observation data collection and from a background survey completed during the first laboratory visit. The current sample was enrolled in four cohorts (ages 3 years, 3 years 9 months, 4 years 6 months, and 5 years 3 months) and assessed every nine months until the age 5 years and 3 months laboratory session. At the end of the preschool phase (age 5 years 3 months for 244, but age 6 years for 69 in this sample), children completed standardized tests to assess FCA. Children then were followed annually from grades one through four, including laboratory visits and surveys of teachers. We used outcome data from the last assessment, which was Grade 4 for most children in this sample but Grade 3 for 23 children. Out of the 313 children who participated at 5 years 3 months, 285 (91%) had a child and parent survey and 275 (88%) had a teacher survey at Grade 3 or Grade 4. Those missing a Grade 3 or 4 assessment were primarily from families who had moved out of the area in which participants were recruited into the project. Those missing at the Grade 3 or 4 follow-up time point did not significantly differ from the 313 preschool phase families with respect to sex, parent smoking, or measures of socio-familial risk (see below) assessed at enrollment. All adult participants gave informed consent, and child participants gave assent during the elementary school phase of the study. The Institutional Review Board of the University of Nebraska-Lincoln approved all procedures.

Measures

Preschool executive control (EC) and foundational cognitive abilities (FCA).

EC was measured with nine tasks, administered during a laboratory session, that assess working memory, inhibitory control, and flexible shifting (Espy, 2016; James et al., 2016). See Table 1 for descriptions of these tasks. Indicators of the preschool FCA factor were measured at the end of the preschool period (age 5 years 3 months or 6 years) with the Verbal Comprehension, Concept Formation, and Visual Matching subtests from the Woodcock-Johnson-III (WJ-III) Brief Intellectual Assessment (Woodcock et al., 2001).

Table 1.

Tasks Used to Measure Executive Control at Age 5 Years and 3 Months

Name	Description	Citation
Working Memory
Delayed Alternation	Children must find a small reward hidden beneath cups. After correctly locating the reward, the location alternates to the opposite cup. After a 10-second delay, youth have to locate the reward again, requiring them to recall the previous location.	Espy et al. (1999); Goldman et al. (1971)
Nine Boxes	Children search for small figures hidden under nine boxes with lids of varying shapes and colors. They can only open one box per trial, and boxes are scrambled between trials, requiring them to remember previously searched boxes in order to locate the figurines in the fewest possible trials.	Adapted from Diamond et al. (1997)
Nebraska Barnyard	Nine animal pictures are displayed on a touch-screen computer, which produce a corresponding animal sound when buttons are pressed. After training, the animal pictures are removed, and children must press the colored boxes corresponding to the animals from memory.	Adapted from Hughes et al. (1998)
Inhibitory Control
Big-Little Stroop	Children are presented with drawings of everyday objects with smaller objects embedded inside them; the smaller objects either match or conflict with the identity of the larger objects. Youth must the name the smaller object, requiring them to suppress the name of the larger object during conflicting trials.	Adapted from Kochanska et al. (2000)
Go/No-Go	Children are instructed to “catch” fish on a computer screen with an animated net by pressing a button. During the no-go trials, a shark appears instead of a fish, and youth must not press the button or the net will break.	Adapted from Simpson & Riggs (2006)
Shape School (Inhibit Task)	Children are shown cartoon stimuli with a happy or a sad face. They must name aloud the color of the cartoon if the face is happy, but have to inhibit their response by remaining silent if the cartoon face is sad.	Espy (1997); Espy et. al. (2006)
Modified Snack Delay	Children are presented with a candy reward. They must refrain from eating the reward until the examiner rings a bell.	Adapted from Kochanska et al. (1996) and Korkman et al. (1998)
Flexible Shifting
Shape School (Switching Task)	Children are shown cartoon stimuli whose bodies came in two colors (red or blue) and two shapes (squares or circles). In the switching condition, children must press a button matching the cartoons’ color or shape, depending on whether the cartoon is wearing a hat.	Espy (1997); Espy et al. (2006)
Trails (Switching Condition)	Children are shown a sheet of paper with pictures of dogs and bones of varying sizes. They must alternate stamping the dogs and then the matching-sized bones, in size order from smallest to biggest.	Espy & Cwik (2004)

Preschool socio-familial stress.

Prior research from this ongoing project (Clark et al., 2016) provides the basis for our measures of socio-familial stress in the current study. A background interview with the parent included questions on income, household size, parent education, and family routines. The Life Stressors and Social Resources Inventory (LISRES; Moos & Moos, 1994) included 200 items and generated scales including financial stressors and resources, home and neighborhood stressors, family and extended family stressors and resources, and friend and social stressors and resources. The Satisfaction With Parenting Scale (SWPS) from the Inventory of Parent Experiences (Ragozin, Basham, Crnic, Greenberg, & Robinson, 1982) is a 17-item checklist that assessed caregivers’ subjective levels of parental stress and satisfaction with their parenting role, as well as their available levels of social and professional support. The Early Childhood HOME Observation for Measurement of the Environment (EC-HOME; Caldwell & Bradley, 1984) provided a direct observational and interview-based assessment of the child’s home environment conducted during the home visit at study entry. The EC-HOME scales assessed the availability of learning materials, such as books and toys; language stimulation, such as exposure to the alphabet; the physical and aesthetic quality of the residence and neighborhood setting; the level of maternal responsiveness to children’s questions and interests; the academic stimulation provided to the child, including direct teaching of numbers, shapes, and patterns; appropriate modeling of behaviors such as manners; the variety of exposure to new places and activities; and, finally, the acceptance that the parent shows for the child, a scale that focuses particularly on the use of harsh discipline strategies.

Clark et al. (2016) conducted an exploratory factor analysis using these measures and identified three latent factors: distal financial stress, proximal household stress, and parent social stress. Distal financial stress included: income to needs ratio, maternal education, household crowding, LISRES financial resources, LISRES financial stressors, and LISRES home and neighborhood stressors. Proximal household stress included: EC-HOME learning materials, EC-HOME language stimulation, EC-HOME academic stimulation, EC-HOME variety, EC-HOME responsivity, EC-HOME modeling, and parent report of frequency of reading to the child. Parent social stress included: SWPS satisfaction with parenting, LISRES family stressors, LISRES family resources, LISRES friend stressors, and LISRES friend resources. Scales were coded so that higher scores represented higher stress. Due to the smaller sample size and increased model complexity in the current study, we obtained factor scores in a confirmatory factor analysis from the model identified in Clark et al. (2016) to represent the socio-familial factors.

Parent substance use.

We used a measure of parent smoking based on items in the background interview that asked whether the child’s mother and father were current smokers at enrollment. If either parent smoked, this measure was coded as 1; if neither smoked, the measure was coded as 0. Nearly half (42.8%) of the children had a parent who smoked at study entry.

Maladaptive functioning and adaptive functioning in middle childhood.

Composite indices of maladaptive functioning and adaptive functioning utilized parent, child, and teacher report data collected when children were in Grade 3 or Grade 4. Both indices consisted of 5 components each (Maladaptive Functioning: Externalizing Problems, Hyperactivity, Internalizing Problems, Aggression, and Peer Victimization; Adaptive Functioning: Prosocial Behavior, Recipient of Prosocial Behavior, Social Competency, Effortful Control, and Academic Skill), and most of these components involved combinations of scale scores. See Table 1S in online supplement for information on the specific measures within the 10 components, including psychometric data. When multiple scale scores were combined within a given component, they were first z-scored and then averaged. Employing the approach used by Stoddard et al. (2013), the continuous measure for each component was then coded 0, 1, or 2 based on cut points set at 1 standard deviation unit below and above the mean. These ordinal measures were then added together to generate maladaptation scores and, separately, adaptation scores.

Child sex.

Child sex, coded male=1 and female=0, was included as a covariate.

Analysis

Models were estimated with Mplus version 8 (Muthén & Muthén, 1998–2018) via maximum likelihood robust (MLR) estimation, which uses cases with partially missing data and cases with complete data under the assumption that data are missing at random after accounting for other model variables (Graham, 2009); therefore, the analysis sample was 313. A confirmatory factor analysis assessed the overall associations among all variables. A structural equation model then assessed unique paths. This model included paths from child sex, parent smoking, and each type of family stressor to EC, FCA, and maladaptation and adaptation, and from EC and FCA to maladaptation and adaptation. Mediation was based on joint statistical significance tests (Leth-Steensen & Gallitto, 2016), that is, whether both links in the specified mediational pathways were statistically significant. The residual error terms for the two Shape School indicators (i.e., the Inhibit and Switching conditions) were correlated because of shared stimuli and response format. Model fit indices CFI and TLI around ≥ .95 and RMSEA around ≤ .06 represent good model fit (Hu & Bentler, 1999).

Results

The confirmatory factor analysis had good fit (χ²(df) = 166.67 (114), p=.001, RMSEA = .038; CFI = .96). Overall associations, shown in Table 2, were in the expected direction. Higher scores on all three types of socio-familial stress, parent smoking, and male sex correlated with higher maladaptive functioning scores and lower adaptive functioning scores in middle childhood. Higher EC and FCA were associated with less maladaptation and more adaptation in middle childhood, although the association between EC and adaptation was not significant. All types of socio-familial stress and parent smoking were significantly associated with lower levels of FCA, but only parent smoking was significantly associated with EC. Correlations between the different types of socio-familial stress and parent smoking ranged from .26 to .79. Note that, following the specifications of the bi-factor model (Espy, 2016), EC and FCA were not allowed to correlate in order to create orthogonal general (FCA) and specific (EC) latent constructs.

Table 2.

Associations Among Variables Based on the Confirmatory Factor Analysis Model

	Male	Parent smoked	Distal financial stress	Proximal household stress	Parent social stress	EC	FCA	Mal- adapt
Parent smoked	.045	---
Distal financial stress	−.014	.583*	---
Proximal household stress	.074	.528*	.792*	---
Parent social stress	−.022	.260*	.431*	.529*	---
EC	−.179	−.168*	.006	.037	−.010	---
FCA	−.064	−.357*	−.530*	−.552*	−.163*	---	---
Maladaptive functioning	.227*	.308*	.268*	.323*	.255*	−.258*	−.296*	---
Adaptive functioning	−.284*	−.224*	−.332*	−.386*	−.295*	.155	.360*	−.594*

^*p<.05.

EC=executive control, FCA=foundational cognitive abilities.

The structural model also fit the data well (χ²(df) = 175.22 (118), p<.001, RMSEA = .039; CFI = .94). Standardized coefficients are shown in Figure 1. The factor loadings on the EC and FCA latent factors using the specifications of a bi-factor model are similar to those reported in earlier studies that used some of the same data (Clark et al., 2016; Espy, 2016). The fairly high loadings of some EC tasks on the FCA latent factor reflect that performance on these tasks is strongly related to cognitive processing abilities, which is consistent with theory that EC tasks also draw on more general cognitive abilities. Factors of socio-familial stress more strongly predicted FCA than EC. Sex and parent smoking significantly predicted EC, with being male and having a parent who smoked uniquely associated with lower EC. All three types of socio-familial stress uniquely predicted FCA, with a positive parent social stress association and negative financial stress and household stress associations. Both EC and FCA, as well as parent social stress and sex, uniquely predicted maladaptive functioning in middle childhood. FCA, parent social stress, and sex uniquely predicted adaptive functioning in middle childhood. Joint statistical significant criteria indicated mediation for each of the three types of socio-familial stress on both maladaptation and adaptation in middle childhood through FCA, but not EC. There also was evidence of mediation of sex and parent smoking on maladaptation through EC. Variables in the model explained an estimated 9% and 35% of the variance in EC and FCA, respectively, as well as an estimated 24% and 28% of the variance in maladaptive functioning and adaptive functioning, respectively.

Figure 1.

Standardized estimates for path model linking socio-familial stress in early childhood, executive control and foundational cognitive abilities at age 5 years 3 months, and maladaptive and adaptive functioning in middle childhood. Bold typeface and * indicate p<.05. 9B = Nine Boxes, DA= Delayed Alteration, NB= Nebraska Barnyard, BL = Big-Little Stroop, GNG=Go/No-Go, SSI=Shape School (Inhibition), mSD=Modified Snack Delay, SSS= Shape School (Switching), TRB=Trails (Switching), VC= Verbal Comprehension, CF=Concept Formation, VM=Visual Matching.

Discussion

Guided by developmental psychopathology (Cicchetti, 1989), we examined childhood cognitive processes, including executive control (EC) and foundational cognitive abilities (FCA), during the preschool years as predictors of both maladaptation and adaptation in middle childhood, and also tested EC and FCA as intervening variables linking early childhood socio-familial stress with subsequent functioning to capture a developmental cascade (Cicchetti & Masten, 2010). Using a bi-factor structural equation model, this study is an advancement over those that have addressed the role of early EC or FCA alone without considering their joint influences (e.g., Sulik et al., 2015), have included each construct at the structural rather than measurement level (e.g., Rohlf et al., 2018), or have relied on questionnaire measures rather than more objective performance-based assessments (e.g., Ghassabian et al., 2014).

Confirmatory factor analysis showed that each of the three measures of socio-familial stress (distal financial stress, proximal household stress, and parent social stress) had a statistically significant negative association with preschool FCA. The overall associations between FCA and both distal financial and proximal household stress were substantial (rs > .5), while the associations between EC and all three dimensions of socio-familial stress were close to zero. Interestingly, in the structural model, the path for parent social stress became positive and statistically significant, which may be a statistical artifact (e.g., suppression) due to the large number of parameter estimates. Still, in general, findings suggest that socio-familial stress may compromise the development of FCA early in development, perhaps by diminishing the types of enriching social interactions and household environments that promote optimal cognitive development for young children. The measures of socio-familial stress were not associated with EC. This finding contrasts with some prior research establishing such associations (e.g., Nelson et al., 2015), but is consistent with Clark et al. (2016) who showed that family socioeconomic status predicts FCA in early childhood but not EC within a bi-factor modeling approach. Associations between family sociodemographic factors and EC that have been reported in the literature may be spurious due to the overlap of EC with FCA that has not been accounted for in prior studies. Compared to FCA, EC abilities are just beginning to emerge in preschool and continue to develop and differentiate in subsequent years; therefore, possible predictive effects of early family stressors may emerge later in development and relate to specific EC aspects (e.g., inhibitory control). Further longitudinal research is needed to test this possibility.

EC in preschool had a statistically significant negative association with maladaptive functioning in middle childhood, but its unique association with adaptive functioning, though positive, was small and not statistically significant. This is consistent with a growing literature showing that EC deficits account for variation in health and behavioral problems (Nelson et al., 2019). Some research has shown links between EC and positive outcomes (e.g., social competence; Jacobson et al., 2011); however, this study advances the literature by considering EC while accounting for FCA, and examining longer-term pathways leading to both maladaptive and adaptive functioning. Interestingly, we found that FCA had both a statistically significant negative association with maladaptation and a statistically significant positive association with adaptation, and served as an intervening variable linking socio-familial stress with the outcomes. Consistent with the notion of experiential canalization (Blair & Raver, 2012), which describes processes by which biological and environmental influences interact to selectively shape development and optimize functioning, this suggests that early family stressors can disrupt the development and expression of foundational cognitive processes that children need to promote competence and reduce risk for early emerging difficulties, independent of EC.

Only parent social stress had persistent, adjusted long-term associations with maladaptive functioning (β = .16) and adaptive functioning (β = −.19). This measure captured early childhood parenting and family relationship stressors, well documented predictors of child developmental outcomes (Morris et al., 2017). It is likely that additional intervening variables, such as poor academic performance and peer relationship difficulties, would further explain relationships between early socio-familial stress and later functional outcomes. Finally, parent smoking had a significant negative association with EC but not with FCA, whereas male sex had expected negative associations with EC and adaptation and a positive association with maladaptation.

There are some noteworthy study limitations. The sample was drawn from a small city in one region of the United States and was predominantly White; results might not generalize to diverse children from other regions of the country. The rigorous bi-factor modeling strategy provided a stringent test of the independent associations of EC and FCA, but afforded little opportunity to include additional intervening variables in the analysis due to considerations related to the large number of parameter estimates and relatively small sample size. Data collection for this study is ongoing, and future analyses will extend into adolescence with direct measures of health-compromising behaviors, such as substance use and delinquency.

Notwithstanding these limitations, findings from the current study have implications for prevention practice. System changes that reduce family stressors, such as poverty and poor parenting, early in development are needed to promote positive, nurturing environments for children (Biglan et al., 2012). Such changes may affect the development of children’s cognitive abilities, which, in turn, may stabilize and have lasting influences on their health and well-being. Accounting for FCA, we found that preschool EC was a predictor of middle childhood maladaptive functioning; therefore, intervening to enhance EC abilities during the preschool years could help reduce the types of risks that portend the emergence of costly and debilitating problems in adolescence. Efforts to facilitate the development of FCA, which early environmental factors still have the potential to influence (Sternberg, 2012), might further promote the development of adaptive functioning in middle childhood. Together, these findings suggest the value of decreasing family stressors by both directly increasing family resources (e.g., more generous income and in-kind transfer programs) and incorporating curricula that promote cognitive development, broadly speaking, into existing school- and family-based preventive interventions to facilitate children’s health and well-being.