Childhood Attention Problems Mediate Effects of Child Maltreatment on Decision-Making Performance in Emerging Adulthood

Acknowledgement: Fred A. Rogosch was a senior author of this article but passed away prior to publication. We are deeply grateful to him for his many contributions to this study.
This research was supported by grants received from the National Institute on Drug Abuse (NIDA; R01DA17741), the Spunk Fund, Inc., and the National Institute of Child Health and Human Development (NICHD; P50HD096698).

Throughout the adolescent years and into emerging adulthood, there are rapid changes in biological, social, and psychosocial systems. Priorities shift to identity exploration and pursuing new possibilities related to work, education, and relationships. In emerging adulthood (defined by contemporary society as the developmental period between approximately 18–25 years old succeeding the adolescent years; Arnett, 2000), individuals face developmental tasks (e.g., obtaining employment, pursuing additional schooling, parenthood) in which they engage in more independent decision making as they explore new social and professional roles and experiences. Toward the end of emerging adulthood, roles in one or more domains may become solidified (Tanner & Arnett, 2016). Therefore, insensitive and/or risky decision making during this developmental period can lead to life-altering consequences, including increased odds of contracting a sexually transmitted infection (STI), development of dependence on alcohol or drugs, unwanted pregnancies, loss of employment, incarceration, or even death (Steinberg, 2004). Exploring developmental processes that predispose risky decision making in emerging adulthood is therefore vitally important because choices made during this developmental stage can have profound implications for life trajectories and well-being.

Typical development of decision-making capacities and related executive functions is multiply determined throughout childhood, adolescence, and into adulthood through complex interactions between cognitive and affective development, hormonal changes, and the social context (Romer et al., 2017; Zelazo & Müller, 2010). Sensation-seeking behaviors normatively increase with the onset of puberty. However, reasoning, problem-solving, and impulse-control abilities continue to develop throughout adolescence and into emerging adulthood. (Braams et al., 2015; Giedd, 2004; Steinberg, 2010). Brain development coincides with these behavioral changes. The dorsal lateral prefrontal cortex, a brain region important for controlling impulses and decision making, is one of the last areas of the brain to reach maturity in the early twenties (Giedd, 2004). As such, there is a biological vulnerability that exists in the adolescent and emerging adulthood years. Nevertheless, this span of development also represents a period of growth and improvement in executive functions, such as response inhibition, complex decision making (e.g., using multiple sources of information), and consideration of risk and rewards (Paus, 2005; Steinberg, 2008). The Lifespan Wisdom Model (Romer et al., 2017) proposes that typical levels of adolescent sensation seeking and reward sensitivity increase in a developmentally-adaptive fashion in early adolescence to facilitate exploration. Sensation seeking typically declines thereafter as impulse control develops more fully. Therefore, maladaptive decision making may occur at higher rates for emerging adults when executive functions and inhibitory control lag.

Studies lend support for this model. For example, lower executive function performance relates to higher self-reported risk-taking behaviors for adolescents and emerging adults (Pharo et al., 2011). One study assessed executive functions (i.e., cognitive flexibility and planning) and decision making with performance measures in children and adolescents aged 8–19 years old and similarly found evidence that lower levels of executive functioning abilities were related to riskier decision making (Schiebener et al., 2015). Another study of adolescent males (age 13–17) who were involved with the juvenile justice system found that lower levels of self-reported impulse control were related to risky sexual behavior over time (Knowles et al., 2020). Results indicate that challenges in executive functions increase risky decision making even for individuals with higher propensity for risk-taking behaviors. In contrast, more matured executive functioning may buffer against the continuation of risk behaviors and related outcomes. Although these studies exemplify the link between executive functions and behavioral risk taking, consideration of developmental antecedents and biopsychosocial contributors is vital to understand individual differences in risky decision making.

Substantial variation exists in the stability and safety of the family environment and the ability of a caregiver to provide sufficient support for the development of self-regulatory skills. Sensitive and responsive caregivers provide children with positive formative experiences with the social environment and help children to develop regulation skills needed to competently complete developmental tasks (Groh et al., 2017). In chaotic and/or dangerous environments, parents who interact with their children in frightening, unpredictable, or harmful ways can also be a source of fear, disrupting early regulatory processes (Hesse & Main, 2006).

Childhood maltreatment, defined by behaviors enacted by a caregiver that harm and/or endanger a child, is one kind of severe relational adversity. In 2018, 678,000 children were victims of childhood maltreatment (U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau, 2020). Maltreatment is estimated to occur at a rate of approximately one in 25 children in the United States, with rates of maltreatment 5 times greater for children living in poverty (Sedlak et al., 2010). Maltreating family environments commonly exhibit higher levels of unpredictability, higher rates of parental psychopathology, and higher incidence of intergenerational maltreatment and/or trauma exposure (Rogosch et al., 1995; Stith et al., 2009). The alarming rate of maltreatment in families experiencing financial adversity and other family-level stressors suggests that families involved with Child Protective Services often experience an aggregation of stressors (Drake & Jonson-Reid, 2013). Importantly, decision-making capacities, specifically decisions involving resources or money, are influenced by financial status. Persistent financial challenges associated with poverty (e.g., unexpected costs, insufficient food, lack of job or housing security) affect attentional resources and decision making (Shafir, 2017). The co-occurrence of maltreatment and financial adversity represents an intersection of risks that can leverage changes to the development of decision making.

Advances in operationalization of maltreatment have increased awareness of the importance of considering maltreatment exposures as complex, often overlapping traumatic experiences. Children exposed to maltreatment who are identified by child protective services often experience more than one subtype of maltreatment (sexual abuse, physical abuse, emotional maltreatment, and/or neglect, as described in Barnett et al., 1993) occurring chronically (Rivera et al., 2018; Warmingham et al., 2019). Furthermore, children exposed to multiple subtypes are at greater risk for development of psychopathology (Vachon et al., 2015).

The early caregiving environment is the proximal context for the development of a variety of competencies, including decision making, emotion regulation, and problem solving. The study of children who have been exposed to adverse and/or traumatic early experiences in their caregiving environments provides information about how developmental trajectories may be influenced by lack of sufficiently predictable and supportive caregiving, and by contrast provides information about typical developmental processes (Cicchetti, 1989). Decades of research suggest that chaotic, traumatic, or unpredictable early life environments interact with genetics to shape an individual’s perceptions, learning, and behavior (Del Giudice et al., 2011; McEwen, 2008). For example, Birn and colleagues (2017) found that individuals with greater early life adversity showed lower levels of functional brain activity than low-adversity peers when presented with potential losses and greater activation when faced with actual losses. Patterns of brain activation during a reward-processing task were related to performance on risk-taking tasks and to self-reported risk taking behavior in daily life (Birn et al., 2017). This study supports the hypothesis that there are brain-based differences in processing risk and reward for individuals exposed to early life adversity and underscores the importance of applying a trauma-informed lens to the study of the development of decision making and risk taking.

Maltreatment exposure increases odds of risk behaviors in adolescence and emerging adulthood, including higher rates of risky sexual behaviors (Oshri et al., 2015), more cannabis and alcohol abuse (Shin et al., 2013), more risky gambling (Hodgins et al., 2010), and higher rates of violent crime offending in adulthood (Topitzes et al., 2012). Less research has been conducted investigating decision-making performance (vs self- or parent-report of risk taking) among individuals with a history of maltreatment. One study by Weller and Fisher (2013) specifically investigated decision-making performance using a gambling task and found that maltreated children took greater risks than nonmaltreated peers and made less bet adjustments to match the level of trial risk when compared to their nonmaltreated peers.

Decision-making capacities draw on executive functioning and use of these abilities under different affective states, such as anger, excitement, or sadness (Peters et al., 2006). Maltreatment experiences, which are associated with childhood traumatic stress, affect the development of cortical structures employed in executive functioning, emotion regulation, and impulsivity, all of which are implicated in decision making (Teicher & Samson, 2016). A recent review (Kavanaugh et al., 2017) summarizes findings indicating that individuals with a history of maltreatment displayed neurocognitive weaknesses in a variety of cognitive functions, including cognitive abilities, language, and executive functions. Children exposed to early onset and chronic maltreatment show poorer working memory and inhibitory control abilities during childhood (Cowell et al., 2015) and perform worse on cognitive flexibility tasks and have slower verbal processing speed than their peers in adolescence (Mothes et al., 2015). Maltreatment dimensions, including subtypes, chronicity, and onset, are also related to differential functioning (Kavanaugh et al., 2017). Meta-analytic findings indicate that maltreatment also increases trait impulsivity across the lifespan (Liu, 2019). Together this research suggests that exposure to maltreatment can have a detrimental effect on neurocognitive development, including worse performance on tasks that require skills using complex information to make decisions.

Maltreatment exposure also disrupts emotional and behavioral regulation capacities needed to enact adaptive decision-making abilities (Lavi et al., 2019). Maltreated children often experience greater exposure to negative affective displays in early caregiving environments. Moreover, maltreated children display attentional biases toward affective information, exhibit worse emotion differentiation, and report more negative attributions about the motivations of others (Luke & Banerjee, 2013). These differences in perception may change children’s response to the environment, which, in turn, increases risk for poor peer relations and psychopathology. Additionally, Oshri et al. (2015) found increased avoidant and anxious attachment underlie the effect of maltreatment experiences on increased risk behaviors in college students.

Cognitive and affective processes transact to influence decision making and risk taking. For example, emotion regulation influences the ability to use executive functioning skills from moment to moment in daily life (Peters et al., 2006). There is little evidence that cognitive structures and emotional development should be considered as separately occurring processes in development, particularly because they activate interconnected biological systems (Gross & Jazaieri, 2014; Pechtel & Pizzagalli, 2011). Despite the shared neural processes underlying cognitive and affective regulation, cognitive and affective literature is typically not integrated, making it challenging to understand shared or differential effects on development. Contemporary models of human behavior increasingly integrate cognitive and affective components of the human experience (e.g., Pruessner et al., 2020). Testing coordinated processes that require both cognitive and affective abilities provides a more complete articulation of developmental trajectories. This is particularly true in the study of decision-making capacities, which rely heavily on the employment of cognitive abilities across affective states.

Attention processes are inherently complex and multisystemic. Attention can be conceptualized as the encoding part of perception that initiates the processing of social information and emotion generation and regulation cascades (Crick & Dodge, 1994; Gross, 2015). Shifting and focusing attention are important skills for emotion regulation (Fernandez-Duque et al., 2000; Gross, 2015), with higher negative affect being related to worse attentional control (Derryberry & Rothbart, 1988). Poor attention modulation has been identified as one process by which maltreatment experiences increase risk for emotion regulation difficulties (Shields & Cicchetti, 1998). Maltreatment and difficulty efficiently using executive attention abilities each uniquely predict greater emerging features of borderline personality disorder (BPD), typified by difficulties in emotion regulation and interpersonal relationships (Rogosch & Cicchetti, 2005). Attention is also integral in metacognitive functions, such as planning, inhibitory control, and resource allocation (Fernandez-Duque et al., 2000), areas of functioning negatively affected by maltreatment exposure (Kavanaugh et al., 2017; Liu, 2019).

Neurocognitive difficulties, such as underdeveloped executive functioning, impulsivity, and memory difficulties, are common symptoms of attention deficit hyperactivity disorder (ADHD) in children. Although ADHD is conceptualized as a neurocognitive developmental disorder, it frequently presents with social cognition deficits (Uekermann et al., 2010). Children who have been exposed to relational trauma have increased rates of ADHD, comorbid with mood disorders in childhood. Executive functioning deficits also mediate the effect of maltreatment exposure on social functioning, academics, and behavioral and psychological symptoms (DePrince et al., 2009; Tarren-Sweeney, 2013). It is often challenging to differentiate symptoms of ADHD and childhood responses to traumatic stress due to the significant overlap in symptomatology (Siegfried & Blackshear, 2016). Investigating attention as a mechanism linking childhood maltreatment experiences and decision-making abilities highlights the connection between cognitive and affective developmental processes and emphasizes the importance of trauma-informed treatment approaches for children with complex behavioral challenges.

This investigation has two major aims: (a) to determine if individuals with and without documented histories of maltreatment show different patterns of risk taking in emerging adulthood, and (b) to identify longitudinal associations between childhood maltreatment, attention problems in childhood, and decision-making performance in emerging adulthood in a racially/ethnically diverse sample of individuals recruited as children from families experiencing financial adversity. This study utilizes prospective measurement of maltreatment, ascertained from coded CPS records, and a multimethod approach of assessing attention in childhood. Specifically, a latent construct of attention was modeled with four indicators: two based on behavioral ratings of attention problems, and two based on performance measures that assess executive attention abilities, attention to detail, impulsivity, and memory. This investigation, therefore, provides a multi-method longitudinal approach to study the impact of maltreatment on broad attention processes and subsequent decision making in emerging adulthood. It is well documented that childhood maltreatment co-occurs with poverty (Drake & Jonson-Reid, 2013; Sedlak et al., 2010). Poverty is an important contextual risk factor associated with environmental unpredictability that influences children’s response to potential rewards and losses (Kidd et al., 2013; Sturge-Apple et al., 2017). This study involves a sample of low-income individuals with and without maltreatment exposure to test the effect of maltreatment on decision-making and risk-taking performance beyond the potential confounding effects of socioeconomic disadvantage.

The study of maltreated children’s development across childhood and into emerging adulthood provides an opportunity to understand the organization and adaption, or maladaptation, of multiple developmental systems (Cicchetti & Toth, 1995). Therefore, this study can apply a trauma-informed perspective to identify developmental sequelae resulting from child maltreatment experiences that contribute to variation in decision-making capacities in emerging adulthood. In this study, decision making during emerging adulthood is assessed with a gambling task in order to assess performance, as opposed to self-report of risk taking behaviors, a complementary approach to many studies that have documented higher risk taking behaviors in samples of maltreated individuals (e.g., Oshri et al., 2015).

Participants (n = 379) included emerging adults who participated in a research summer camp as children (Wave 1) and a follow-up study approximately 10 years later (Wave 2). The original study included 659 socioecomically disadvantaged, ethnically diverse maltreated (n = 339) and nonmaltreated children (n = 320) recruited for a summer research camp from 2004–2007. Recruitment of families and children with and without maltreatment was necessary to ascertain groups comparable in size and socioeconomic status. Children in the maltreated group had substantiated investigations of child maltreatment according to Department of Human Services (DHS) Child Protective Services (CPS) records; demographically comparable nonmaltreated children without CPS or preventive records were from families receiving Temporary Assistance to Needy Families. A DHS recruitment liaison contacted a random sample of eligible families from both groups via mail. If families were interested and chose to participate, their contact information was shared with research staff. The demographics of families who declined participation were not disclosed by DHS.

Child participants at Wave 1 were 10–12 years old (Mage = 11.27, SD = .97) and 49.5% were female. The original sample was diverse in regard to race and ethnicity (71.5% African American, 12.0% White, 12.6% Hispanic, 3.9% Other race). Parents provided informed consent for their child’s participation. During the week of summer camp, camp counselors conducted recreational activities with the same groups of 8–10 children (35 hours of direct contact and observation). After providing assent, child participants self-reported on their experiences and behaviors and completed computerized tasks. Camp counselors provided independent ratings of childhood functioning and behavior after the end of the week. For a more detailed description of the summer research camp procedures, see Cicchetti and Manly (1990).

At Wave 2, participants who were children at Wave 1 were recontacted via phone or mail and invited to participate in a voluntary follow-up study. Participants were 18–23 years old at Wave 2. After providing informed consent, Wave 2 participants completed three research visits, which included interviews, computerized tasks, and self-report questionnaires. All procedures for this study were approved by the University of Rochester Research Subjects Review Board (study title: Chronic Stress of Maltreatment: Drug Use Vulnerability; Protocol 46062). A total of 427 participants completed assessments at Wave 2. The present study includes a subset of individuals who completed Wave 2 (n = 379; 51.5% female; Mage = 19.68, SD = 1.12) who completed key measures of interest at Waves 1 and 2. There were 48 individuals who completed some assessments at Wave 2 who were not included in the present study due to missing or incomplete data on performance tasks analyzed herein and/or maltreatment subtype information.

Participants (n = 379) included 199 children who experienced maltreatment during childhood (52.5%) and 180 individuals who did not experience maltreatment in childhood. Subjects were racially/ethnically diverse (77.3% Black/African American, 11.1% White, 7.7% Hispanic, 4.0% Other race). At Wave 1, 86.6% of families were receiving full or partial financial assistance (mean family income: $22,530). The majority (69.4%) of families at Wave 1 were headed by a single caregiver. There were no differences on gender (χ2(1, 659) = .30, p = .58) or maltreatment status (χ2(1, 659) = .41, p = .52) between those included in this study and those who did not complete follow-up measures. Participants included in the present study were similarly racially and ethnically diverse as compared to the original Wave 1 sample.

Wave 1

Maltreatment Classification System

CPS records were coded with the Maltreatment Classification System (MCS; Barnett et al., 1993). The MCS is a comprehensive coding system that reliably quantifies maltreatment subtype, severity, frequency, perpetrator, and developmental timing from written records or by interview. MCS reliable coders scored lifetime CPS records for each child. Based on operational definitions, the MCS identifies four different subtypes of maltreatment: sexual abuse, physical abuse, emotional maltreatment, and neglect, which has four categories: lack of supervision, failure to provide, educational neglect, and moral/legal neglect. A count of subtypes (0–4) was used in analysis, with presence of neglect indicated by presence of any of the categories of neglect described above. Of the 379 children assessed at both waves, 199 children (52.5%) experienced one or more subtypes of maltreatment. Of the children who experienced maltreatment, 43.2% (n = 86) experienced one subtype of maltreatment, 40.2% (n = 80) experienced two subtypes of maltreatment, 14.6% (n = 29) experienced three subtypes of maltreatment, and 2.0% (n = 4) experienced four subtypes.

Consistent with recent literature that documents the common co-occurrence of maltreatment subtypes (Rivera et al., 2018; Vachon et al., 2015; Warmingham et al., 2019), children in this sample experienced complicated patterns of maltreatment subtype. Of the children who experienced one subtype (n = 86), the majority experienced neglect (n = 58), and the rest either experienced emotional maltreatment (n = 14), physical abuse (n = 10), or sexual abuse (n = 4). Children who experience two subtypes (n = 80) most commonly experienced neglect (n = 72), emotional maltreatment (n = 63), followed by physical abuse (n = 22), and sexual abuse (n = 3). For children who experienced three subtypes of maltreatment (n = 29), neglect was the most common experience (n = 28), followed by emotional maltreatment (n = 26), physical abuse (n = 25), and sexual abuse (n = 8). Although individual subtypes have received considerable attention in the literature, subtypes commonly overlap, which creates a problem for parsing individual subtype effects (Jackson et al., 2019; Manly, 2005). Number of subtypes of maltreatment parameterizes a complex and often overlapping set of experiences into a dimensional score that has been related to an array of outcomes (e.g., biological, psychosocial, cognitive) relevant to the current study (Cicchetti & Toth, 2016).

Wechsler Intelligence Scale for Children, 4th edition (WISC-IV; Wechsler, 2003)

Children’s cognitive functioning was assessed using six subtests of the WISC-IV. A Full Scale IQ (FSIQ) was ascertained along with two factors: Verbal Comprehension (subtests: Similarities, Vocabulary, Comprehension) and Perceptual Reasoning (subtests: Block Design, Picture Completion, Matrix Reasoning). In this sample, the mean FSIQ is 87.23, SD = 12.68.

Wave 1 Measures of Child Focus/Attention:

1. Delayed Matching to Sample (DMS). The Cambridge Neuropsychological Testing Automated Battery (CANTAB; Sahakian & Owen, 1992) is a computerized assessment battery used extensively to assess neuropsychological functioning in child and adult populations with and without psychiatric and neurological conditions. The Delayed Matching to Sample (DMS) task was administered to children during the summer camp study at Wave 1. During this task, children were shown a complex visual pattern (the “sample”), then after a brief delay (0, 4, or 12 seconds), children were shown a set of four patterns. One of the patterns was identical to the sample, and the other three were similar but not identical to the sample. The child was instructed to touch the pattern that matched the sample pattern originally displayed. Higher total correct scores (across the 20 trials) is related to better spatial working memory, attentional abilities, and more developed impulse control. The DMS total correct score was then reverse coded so that greater scores indicated worse performance.
2. Attention Network Task (ANT; Rueda et al., 2004). The child version of the ANT is a computer administered performance assessment that measures attention network functioning. In the child version, the stimuli are fish facing one direction or the opposite direction (adult version uses arrows). Either a single fish or a row of five fish appeared on the screen, and children were asked to respond with a left or right mouse click based on the direction that the central fish was facing. The ANT has both congruent (all fish facing the same direction) and incongruent (central fish facing opposite direction) trials. Three independent scores can be calculated based on reaction time across conditions: alerting, orienting, and conflict scores. The conflict score was used in this analysis and was computed as the difference in median reaction time between congruent and incongruent trials. High scores on the conflict dimension indicate more difficulty monitoring and responding to conflictual visual information.
3. Teacher Report Form (TRF)–Attention Problems Subscale (Achenbach, 1991). The Child Behavioral Checklist (CBCL) Teacher Report Form (TRF) is a 113-item reporting scale used to assess behavioral problems in children. Items are scored from 0 = not true, 1 = somewhat or sometimes true or 2= very true or often true. Two camp counselors independently completed this measure for each child after a 35-hour week of direct observation and interaction with children at Wave 1 (average intraclass correlation (ICC): r = .80). The Attention problems subscale was used in the present study. Sample items from the attention problems subscale include: “can’t concentrate,” and “doesn’t carry out tasks.” Higher T-scores on the Attention Problems subscale indicate greater ratings of attention problems relative to other children at the same age.
4. California Child Q-Set (CCQ)–ADHD Sort. The CCQ is a widely used measure to assess dimensions of personality and behavior in children (Block, 1980, 2008). The CCQ consists of 100 items relating to personality and behavior. Two camp counselors completed the CCQ for each child after extensive observation and interaction with children during the week of summer camp (average ICC: r = .85). Raters generated individual profiles of each child by sorting the 100 items into a distribution from 1–9, with higher scores indicating that a given statement is more characteristic of the child being rated. Each child’s item sort values were then correlated with criterion sorts for specific prototypical behavior and personality features. The criterion sort for the attention deficit disorder with hyperactivity (ADHD) sort included high rankings for items such as “is restless and fidgety” and low rankings for items such as “is obedient and compliant.” Higher correlations between a child’s sort and the prototypical ADHD Q-sort indicate greater attention problems.

Wave 2

Cambridge Gambling Task (CGT)

The Cambridge Gambling Task (CGT) was administered as part of the Cambridge Neuropsychological Testing Automated Battery (CANTAB; Cambridge Cognition) at Wave 2. CANTAB is a computerized assessment battery used extensively to assess neuropsychological functioning in adult populations with and without psychiatric and neurological conditions. The CGT is a performance-based task that assesses decision making abilities and risk taking behaviors outside of a learning context. Specifically, the CGT is “hot” executive function task. Unlike some gambling tasks, it can be considered a “decision under risk” paradigm, meaning that the probability of winning varies in a systematic way that is easily discerned by information provided on each trial. Probability of winning one trial is not based on the past trials. Additionally, participants do not benefit from betting on an unlikely outcome. They are able to choose how much they bet on each trial.

During the CGT, participants were asked to bet on the location of a token, which was hidden from view under 10 colored squares (red and blue) presented in a line on the top of the screen. On each trial the colored boxes appeared at the top of the screen in a different ratio of red to blue boxes (see Zois et al., 2014 for a visual depiction). For example, there might be a trial with eight red boxes and two blue boxes at the top of the screen. On the next trial, the box ratio may appear more evenly split, with four red boxes and six blue boxes. The participants selected which color (red or blue) they believed the token is hidden under. Then, the participant selected a wager for each trial. If they selected the correct color, the points they bet on that trial were added to their point total. If they selected the incorrect color, the points they bet were subtracted from their point total. The goal of the task is to accrue as many points as possible by winning bets.

Betting amounts for each trial were calculated by the computer. Bet options included amounts that were 5%, 25%, 50%, 75%, and 95% of the current point total. Bets were automatically shown on the screen, one at a time, either in ascending or descending order for each trial. Participants used the touch screen to select the bet they wish to place. The first and second block of trials completed were ascending trials. On ascending trials, low bets (starting at 5% of total accrued points) appeared on the screen and increased every few seconds. Bet option increases were accompanied by a tone that went up in pitch. Participants selected the bet they wished to place by touching their desired bet when it appeared on screen. If participants failed to select a bet, their bet for that trial was the highest bet available (95% of total accrued points). On descending trials (completed in the third and fourth blocks), the highest bet (95% of total accrued points) was presented first. Bets then decreased every 2 seconds (accompanied by a tone). Participants who did not select a bet automatically wagered the lowest bet (5%) on that trial.The CGT was scored to produce two main variables, Risk Taking and Risk Adjustment, that were of particular interest in this study. Risk taking scores are the average bet (% of total accrued points) placed for trials where participants bet on the majority color (e.g., betting on red when presented with 7 red and 3 blue boxes). Risk taking scores were computed for ascending and descending conditions and for each box ratio (6:4, 7:3, 8:2, 9:1). Higher risk taking scores indicated a greater proportion of points were bet across trials. The risk adjustment score summarizes betting adjustment across box ratio trials and conditions (ascending and descending). Risk adjustment was calculated by summing two times the mean proportion of points risked on 9:1 and 8:2 trials, minus twice the mean proportion of points risked on 7:3 and 6:4 trials, all divided by mean proportion of points bet on all trials. Higher risk adjustment indicates that the participant modified their betting based on the probability of winning or losing, and lower risk adjustment indicates a betting pattern less sensitive to trial risk. Risk adjustment was scored overall and for ascending and descending trials.

Additional scored variables were considered to further investigate performance differences between individuals with and without histories of maltreatment. Quality of Decision Making is the proportion of trials where the participant bet on the more likely color (i.e., the color with a greater number of boxes). Delay Aversion is a score that reflects the difference in betting on ascending and descending trials (Delay Aversion = Risk Taking descending − Risk Taking ascending). Higher delay aversion scores are interpreted as less willingness to wait for low bets on descending trials, when high bets are presented first.

Data Analytic Plan

A set of t-tests and repeated measures ANOVAs were conducted to investigate patterns of betting (measured with CGT Risk Taking) across trial risk (indicated by box ratio) and for ascending and descending conditions on the Cambridge Gambling Task (CGT). T-tests were conducted to investigate maltreatment group differences on CGT quality of decision making, delay aversion, and risk adjustment (ascending and descending). One repeated-measures ANOVA was conducted with trial condition order (ascending/descending) as a within-subjects factor and maltreatment and gender as between-subjects factors. Then, two ANOVAs investigated the effect of box ratio (9:1, 8:2, 7:3, 6:4) for Trials × Maltreatment × Gender (male/female) within ascending and descending trials to investigate patterns of risk taking across trials that differed in probability of winning.

Structural equation modeling was then used to estimate longitudinal pathways between child maltreatment and adjustment of risk taking behavior on the CGT. First, measurement modeling was conducted in Mplus v7.1.4 to establish a latent construct for childhood attention. The four measures of attention (DMS, ANT conflict, TRF attention problems subscale, CCQ ADHD sort) were used as indicators of a latent variable. DMS scores were reversed so that high scores indicated worse performance on the DMS task. Adequate measurement model fit was determined by nonsignificant values of the χ2 statistic, values of the CFI greater than .95, SRMR values less than .08, and RMSEA values smaller than .05 (Hu & Bentler, 1999).

A structural equation model (SEM) was then estimated to test relationships between maltreatment experiences (# subtypes), the latent construct of childhood attention, and performance on the Cambridge Gambling Task, measured by the risk adjustment variable. Childhood attention was predicted by childhood IQ and number of subtypes of maltreatment. Predictors of Wave 2 (emerging adult) risk adjustment included number of maltreatment subtypes, the childhood attention latent construct, and gender. The same indices of fit (χ2, CFI, SRMR, RMSEA) were used to assess global fit in the SEM. Mediation was tested using RMediation (Tofighi & MacKinnon, 2011), which computes asymmetric 95% confidence intervals for the mediated effect. Maximum likelihood robust (MLR) estimation was used in the measurement model and SEM to handle the kurtotic distribution of the TRF attention and ANT conflict measures. Full maximum likelihood estimation (FIML) was used to handle the small amount (< 2%) of missing data on endogenous variables.

There were no differences in Quality of Decision Making between emerging adults who had been maltreated (Mmal = .83, SD = .14) and those who did not experience maltreatment (Mnonmal = .85, SD = .14; Mdiff = −.019, SEdiff = .014; t(377) = −1.35, p =.18), indicating that the groups chose the majority color at similarly high rates. A repeated-measures ANOVA (Bet Order × Gender × Maltreatment) was conducted on the risk taking variable, which is scored as the proportion of points bet on trials when participants chose the more likely outcome. There was a main effect of trial order, indicating that individuals bet significantly more on descending trials, when bets started high and decreased (M = .83, SE = .008) than ascending trials, when bets started low and increased (M = .36, SE = .009; F(1, 375) = 1758.50, p < .001). Males (M = .63, SE = .009) bet significantly more than females (M = .56, SE = .009; F(1, 375) = 28.08, p < .001). Individuals who experienced maltreatment (M = .61, SE = .009) bet more than nonmaltreated peers (M = .58, SE = .009; F(1, 375) = 6.74, p = .010). The only significant interaction was between order and gender (F(1, 375) = 17.31, p < .001). The pattern of means showed that males bet more than females on ascending trials (Mdiff = .12) but this pattern was diminished on descending trials (Mdiff = .022). There was no significant Maltreatment × Order interaction (F(1, 375) = .064, p = .80), Gender × Maltreatment interaction (F(1, 375) = .14, p = .70), or Maltreatment × Gender × Order interaction (F(1, 375) = .002, p = .96). A t-test was conducted on the delay aversion score, and there was no difference between maltreated and nonmaltreated groups (Mdiff = .0092, SEdiff = .023, t(377) = .41, p = .69), indicating that the increase in betting on descending trials, as compared to ascending trials, was similar as across groups.

Additional analyses were conducted to determine if maltreatment history was associated with differential betting patterns (across box ratio) on ascending and descending trials. History of maltreatment was related to worse risk adjustment overall (Mdiff = −.23, SEdiff = .081, t(377) = −2.79, p = .006) and specifically on descending trials (Mdiff = −.16, SEdiff = .070, t(377) = −2.31, p = .022). There were no differences between groups for ascending trial Risk Adjustment (Mdiff = −.19, SEdiff = .14, t(377) = −1.42, p = .16). Repeated measures ANOVAs were conducted to probe differences in risk taking patterns on ascending and descending trials. Box ratio (9:1, 8:2, 7:3, 6:4) was a within-subjects factor; high box ratio (e.g., 9:1) represents a less risky trial than a low box ratio (e.g., 6:4). Between-subject factors included maltreatment status (maltreated/nonmaltreated) and gender (male/female). A Greenhouse-Geisser adjustment was used to correct within-subject tests to account for violation of sphericity (Mauchly’s test ascending: χ2(5) = 129.70, p < .001; descending: χ2(5) = 221.40, p < .001).

On ascending trials, the within subjects effect of box ratio was significant (F(2.39, 893.16) = 147.14, p < .001), indicating participants lowered their bets as the odds of winning became lower. Males bet more than females across ascending trials (F(1, 374) = 34.96, p < .001). Additionally, there was a significant box ratio by gender interaction (F(2.39, 893.16) = 15.94, p < .001), with males adjusting to better odds by betting more on those trials (see Figure 1). There was no main effect of maltreatment status (F(1, 374) = 2.50, p = .11) and no Maltreatment × Box Ratio interaction (F(2.39, 893.16) = .23, p = .83). The interaction of maltreatment and gender (F(1, 374) = .031, p = .86) and the three-way interaction of box ratio, gender, and maltreatment were also not significant (F(2.39, 893.16) = .68, p = .53).

On descending trials, there was also a main within subjects effect of box ratio (F(2.11, 762.69) = 90.03, p < .001), indicating that participants generally decreased bets as the likelihood of winning decreased. Across descending box ratio trials, maltreated individuals bet more than nonmaltreated individuals (F(1, 361) = 5.88, p = .016). The interaction of Box Ratio × Maltreatment was significant (F(2.11,) = 3.22, p = .038) such that maltreated individuals did not adjust their bets (i.e., decrease their bets) as much as nonmaltreated individuals as trials became riskier (see Figure 2). There was no main effect of gender (F(1, 361) = 1.24, p = .27), and the interaction of maltreatment and gender was also not significant (F(1, 361) = .14, p = .71). The interaction between gender and box ratio was not significant (F(2.11, 762.69) = 1.06, p = .35) and the three-way interaction of Box Ratio × Gender × Maltreatment was also not significant (F(2.11, 762.69) = .21, p = .83).

Correlations between variables were investigated prior to conducting the Measurement model and SEM. Worse CGT risk adjustment was significantly correlated with higher scores on indicators of the attention problems latent construct (TRF attention problems, ADHD CCQ, DMS, and ANT conflict; rs = −.12 to −.13, ps < .05). Greater number of maltreatment subtypes was associated with worse risk adjustment (r = −.12, p = .024). Notably, childhood WISC IQ was significantly (ps < .05) correlated with all study variables at the bivariate level. All bivariate correlations between variables included in the SEM are presented in Table 1.

For each indicator of the latent variable, (TRF attention problems, CCQ ADHD, DMS, and ANT), higher scores indicated greater difficulty with facets of attention (See Table 1 for correlations among attention measures). A correlated residual was specified between the TRF and CCQ ADHD indicators to account for the shared method (camp counselor rating) for both the TRF and Q-sort measures. The measurement model was an excellent fit to the data (χ2(1) = 1.64, p = .20, RMSEA = .041, CFI = .99, SRMR = .01). Standardized loadings were all significant (p < .01) and ranged from .35–.42. The correlated residual between the TRF measure of attention problems and the Q-sort ADHD scale was significant (β = .38, SE = .078, p < .001).

The specified model showed good fit to the data (χ2(15) = 35.13, p = .0024, RMSEA = .059, CFI = .92, SRMR = .036). Results indicated that child maltreatment (# subtypes) predicted greater attention problems during childhood (β = .29, SE = .09, p = .001). Lower child IQ simultaneously predicted greater attention problems (β = −.62, SE = .069, p < .001). Greater attention problems in childhood was related to worse risk adjustment on the CGT in emerging adulthood (β = −.35, SE = .073, p < .001). RMediation (Tofighi & MacKinnon, 2011) was used to estimate asymmetric confidence intervals of the indirect effect. The indirect effect of child maltreatment on CGT performance through child attention difficulties was significant (95% CI [−.16, −.014], indicating that child attention problems mediated the relationship between childhood maltreatment experiences and poorer risk adjustment performance in emerging adulthood. Males demonstrated more sensitive risk adjustment than females (β = .11, SE = .05, p = .023; Figure 3).

This longitudinal study investigates antecedents of decision-making performance in emerging adulthood in a sample of individuals with exposure to childhood maltreatment and poverty. Results indicate a mediated pathway from childhood maltreatment experiences to less sensitive risk adjustment during a gambling task in emerging adulthood via attention problems in childhood. The results of this study contribute to an existing literature (e.g., Birn et al., 2017; Oshri et al., 2015; Weller & Fisher, 2013) detailing associations between maltreatment experiences in childhood and subsequent challenges in the development of cognitive and behavioral regulation capacities without using self-report data. This multimethod study extends prior work by using coded child protective record data to characterize child maltreatment objectively and performance tasks and observational ratings to ascertain attention difficulties in childhood in order to establish determinants of risky decision making in emerging adulthood.

The Cambridge Gambling Task (CGT) was used in this study to assess facets of decision making, including risk taking (i.e., bet amounts) and adjustment of risk taking across trials that differed in the probability of winning (risk adjustment). In prior studies the CGT has been used to investigate differences in decision-making performance for individuals with substance abuse and gambling addiction versus healthy controls (Zois et al., 2014) and is considered a “hot” executive function task. Recent functional MRI evidence suggests that participation in this task engages brain regions associated with a range of cognitive abilities, including learning, working memory, and the reward processing network (Yazdi et al., 2019). The CGT is a “decision under risk” paradigm, meaning that the probability of winning varies in a systematic way that can be easily discerned, in this case by attending to the proportion of red to blue boxes on a given trial.

Performance patterns on this task showed that although maltreated and nonmaltreated individuals did not differ in their decision making regarding the box color they bet on, did bet more on the task overall and adjust for odds of winning to a less degree than their nonmaltreated peers. The CGT first presents trials with low bet options that increase (ascending trials) and then descending trials. Clinical and nonclinical samples of individuals both show increased overall bets on descending trials on the CGT (Zois et al., 2014), likely due to a normative aversion to waiting for lower, more calibrated bets. Maltreated individuals did not differ on their change in bet amounts on descending trials (vs ascending trials) but did show less adaptive risk adjustment on descending trials. There are a number of possible explanations and interpretations of these results. Importantly, descending trials were not counterbalanced and were always completed in the second half of the task, which could have increased boredom or frustration during these trials particularly if participants were not successful in winning points. Greater difficulties with behavioral regulation and worse frustration tolerance could contribute diminished attendance to risk on later trials. Although it appears that the difference in performance on descending trials is not solely due to greater impulsivity, it appears that individuals with a history of maltreatment were less attuned to the probability of winning or losing throughout the task, and particularly toward the end of the task.

Throughout adolescence, there is generally a decline in risk taking on “decision under risk” performance tasks (Defoe et al., 2015). Individuals with exposure to maltreatment displayed less mature decision-making capacities in emerging adulthood in situations where greater attention and response modulation is needed to enact measured decisions. This study is consistent with findings from a recent meta-analysis that found that maltreated individuals displayed greater behavioral impulsivity when experiencing strong affect (termed negative urgency), which may contribute to the pattern of behavior exhibited by maltreated individuals on later CGT trials, when boredom or frustration had the potential to be higher (Liu, 2019).

On ascending trials, there was no main or interactive effect of childhood maltreatment. However, there was a significant effect of gender. Males bet more and adjusted their bets based on trial risk more sensitively than females. Males consistently score higher on measures of sensation seeking and exhibit more behavioral risk taking (Cross et al., 2011). The betting pattern seen here is consistent with male heightened sensation seeking and greater risk taking behavior observed in other studies (Cross et al., 2011). Females, on the other hand, bet less overall and adjusted their betting behavior less based on probability of winning or losing. Biological influences as well as socialized gender roles each may contribute and interact to explain the observed behavioral patterns. Current evidence suggests that gender differences in the perception and value of rewards associated with taking risks in different contexts (e.g., social situations, financial decisions) contributes to gender differences in behavior and motivation to engage in risk taking in both daily life and lab paradigms (Figner & Weber, 2011). It may be that males found the CGT more engaging and were more motivated to seek rewards associated with winning trials, while females might have been less motivated by the task and therefore showed more stable and risk-averse betting behavior. More research is needed to examine how gender socialization and motivation could help to explain differences in reward processing, risk taking, and decision making across contexts and socioeconomic status during emerging adulthood.

A main goal of this study was to investigate whether a multimethod assessment of broad executive attentional processes represents a mechanism by which maltreatment negatively impacts decision-making abilities. Results indicated that attentional difficulties, indeed, are one process by which maltreatment experiences negatively affect decision-making performance in emerging adulthood. Attention problems in childhood were measured four ways in this study: two observational measures were completed rating children’s attention deficit symptoms, such as hyperactivity and problems following directions, in a social setting. Two performance tasks were completed in a 1:1 setting: one required memory for details of shapes and sustained focus, and another task that required children to use executive functions to resolve conflict in response to visual stimuli. Confirmation of the latent factor structure provides evidence that behavioral regulation in social settings and focused attention on tasks completed in a controlled environment are influenced by a broader higher-order process, best characterized as executive attention. Consistent with the measurement of this construct, research suggests executive attention has neurobiological underpinnings that influence a broad span of behavior and cognitive abilities, including emotional control, detection and resolution of conflict, and shifting/focusing attention (Fernandez-Duque et al., 2000).

The current findings suggest that maltreatment has a significant and detrimental effect on executive attention in childhood, which, in turn, increases risk for poorer decision-making capacities in emerging adulthood. This finding is consistent with theoretical and empirical literature that suggests that decision making relies on the use of logic and planning as well as emotion regulation and behavioral control (Peters et al., 2006; Yazdi et al., 2019). Findings build upon past work indicating significant relationships between executive functioning weaknesses and riskier decision making (e.g., Romer et al., 2011; Schiebener et al., 2015) by testing longitudinal associations between these constructs from childhood into functioning in emerging adulthood, a developmental period where independent choices become increasingly important to attain goals. The findings contribute to a growing literature documenting developmental risk pathways whereby early maltreatment, a significant interpersonal trauma, disrupts the development of interconnected, neurobiologically rooted executive control processes needed for behavioral, cognitive, and emotional organization in childhood (Cicchetti & Toth, 2016; Del Giudice et al., 2011; Rogosch et al., 1995). Maladaptation in these important executive processes in childhood appears to prolong impulsive risk taking into emerging adulthood. Increased risk taking without adjustment for the realistic likelihood of success has wide-reaching implications for the attainment of developmental competencies and mental and physical health. Less sensitive decision making could predispose individuals to make less measured choices about relationships, jobs, sexual activity, and substance use, and could increase odds of incarceration or poor health outcomes (Mohr-Jensen & Steinhausen, 2016).

This study adds to decades of literature documenting the detrimental effects of maltreatment on multiple domains of functioning throughout childhood and into adulthood (Cicchetti & Toth, 2016). It is essential, therefore, to emphasize the importance of early interventions for families who are involved with or at risk for child abuse and neglect. Furthermore, parents with histories of their maltreatment in childhood may struggle with problems in executive functioning and decision making and may be at heightened risk for continuing the cycle of intergenerational maltreatment and violence (Azar, 1986). Evidence-based relational treatments that provide opportunities for caregivers to acquire positive parenting approaches and build positive relationships with their children can prevent the transmission of violence across generations (Guild et al., 2017).

In childhood, attention problems are common behavioral concerns across school and home contexts. It is important that interventions that aim to help children with problems in attention and executive functions take a contextual view of childhood attention problems that includes the possible strong influence of trauma as well as biological factors influenced by adversity that contribute to challenges with attention and executive functioning (DePrince et al., 2009). Importantly, children who present with attention problems and who have been exposed to maltreatment may have more complex symptom presentations (Tarren-Sweeney, 2013) and engage in risky, violent interpersonal, and criminal behaviors at higher rates later in life (De Sanctis et al., 2012). Assessment of attention and executive functioning symptoms in children should therefore consistently include screening for trauma and/or maltreatment history, and conversely, children treated for trauma exposure should be assessed for executive functioning difficulties so that challenges in these areas can be addressed early in development.

The present findings suggest that executive functioning difficulties that emanate from maltreatment experiences are not isolated to childhood and extend into emerging adulthood. Therefore, childhood interventions that target executive functioning and attention skills, particularly school-based interventions, may provide children with important practice in building executive functioning skills that will continue to benefit them into adulthood. Children who have been exposed to trauma who display behavioral problems may benefit from trauma-informed treatments, such as Trauma-Focused Cognitive Behavioral Therapy (TF-CBT), which is effective in treating symptoms of trauma and improving emotional and behavioral regulation skills (Cohen et al., 2010).

Strengths of the present study include its longitudinal design and use of independent and multi-informant measures of child maltreatment, child attention performance and behavioral ratings, and emerging adulthood decision-making performance in a socioeconomically disadvantaged sample of racially/ethnically diverse individuals. However, limitations should be noted. First and foremost, the present study assesses the contribution of childhood maltreatment on cognitive development but does not take into consideration other likely influences (i.e., effects of parental executive functioning, supportive factors, or other facets of parenting, education, and/or genetic variation) on children’s cognitive development (e.g., Cuevas et al., 2014). Although this sample consists of low-income individuals with diverse race/ethnicities that are largely under-represented in longitudinal developmental studies, the findings herein may be limited in their generalizability and may be more specific to low-income youth living in urban environments. Furthermore, the study design allows for comparisons between maltreated and non-maltreated youth by controlling for socio-economic and demographic variables. However, the role of the broader context (i.e., poverty, school quality, community-based risk and protective factors) on cognitive development is not accounted for in this study. Additionally, laboratory-based risk-taking paradigms, such as the CGT, are thought to be valid assessments of decision making, but do not necessarily include or control for the role of strong affect that often precipitates risky decision making in daily life (Slovic et al., 2005).

Childhood maltreatment represents exposure to parenting dysfunction and interpersonal trauma and affects many facets of development (Cicchetti & Toth, 2016). The present study provides support for one pathway of developmental risk by which childhood maltreatment experiences are associated with greater executive attentional difficulties in childhood, which is predictive of less adaptive decision making in emerging adulthood. Investigating developmental processes in maltreated children and their nonmaltreated peers provides an opportunity for knowledge to be gained about the typical development of cognitive and affective abilities as well as the variation in developmental trajectories that can occur when significant disruption takes place in early caregiving relationships (Cicchetti, 1989). Findings herein contribute to the extant literature documenting the organizational effect that the early caregiving environment has on the development of important regulation capacities. Further research is needed to understand the contribution and interaction of more nuanced contextual factors (e.g., resource availability, protective factors, and socialization) and biological influences on the processing of risk and reward, the development of executive functions and attention, and their relation to decision-making capacities and real-world risk taking behaviors.

Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/4-18, YSR, and TRF profiles. University of Vermont.

Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55(5), 469–480.

Azar, S. T. (1986). A framework for understanding child maltreatment: An integration of cognitive behavioural and developmental perspectives. Canadian Journal of Behavioural Science/Revue Canadienne Des Sciences Du Comportement, 18(4), 340–355. 10.1037/h0079961

Barnett, D., Manly, J. T., & Cicchetti, D. (1993). Defining child maltreatment: The interface between policy and research. In D.Cicchetti & S. L.Toth. (Eds.), Advances in applied developmental psychology: Child abuse, child development and social policy (pp. 7–73). Ablex.

Birn, R. M., Roeber, B. J., Pollak, S. D., & Reyna, V. F. (2017). Early childhood stress exposure, reward pathways, and adult decision making. Proceedings of the National Academy of Sciences of the United States of America, 114(51), 13549–13554. 10.1073/pnas.1708791114

Block, J. (1980). The California Child Q-set. Consulting Psychologists Press.

Block, J. (2008). The California Q-sort procedure described and demonstrated. In J.Block. (Ed.), The Q-Sort in character appraisal: Encoding subjective impressions of persons quantitatively (pp. 17–28). American Psychological Association.

Braams, B. R., van Duijvenvoorde, A. C. K., Peper, J. S., & Crone, E. A. (2015). Longitudinal changes in adolescent risk-taking: A comprehensive study of neural responses to rewards, pubertal development, and risk-taking behavior. The Journal of Neuroscience, 35(18), 7226–7238. 10.1523/JNEUROSCI.4764-14.2015

Cicchetti, D. (1989). How research on child maltreatment has informed the study of child development: Perspectives from developmental psychopathology. In D.Cicchetti & V.Carlson (Eds.), Child maltreatment: Theory and research on the causes and consequences of child abuse and neglect (pp. 377–431). Cambridge University Press.

Cicchetti, D., & Manly, J. T. (1990). A personal perspective on conducting research with maltreating families: Problems and solutions. In G. H.Brody & I.Sigel (Eds.), Methods of family research: Biographies of research projects, Volume 2: Clinical populations (pp. 87–133). Lawrence Erlbaum Associates.

Cicchetti, D., & Toth, S. L. (1995). A developmental psychopathology perspective on child abuse and neglect. Journal of the American Academy of Child & Adolescent Psychiatry, 34(5), 541–565. 10.1097/00004583-199505000-00008

Cicchetti, D., & Toth, S. L. (2016). Child maltreatment and developmental psychopathology: A multilevel perspective. In D.Cicchetti (Ed.), Developmental psychopathology (3rd ed., Vol. 3, pp. 457–512). Wiley.

Cohen, J. A., Berliner, L., & Mannarino, A. (2010). Trauma focused CBT for children with co-occurring trauma and behavior problems. Child Abuse & Neglect: The International Journal, 34(4), 215–224. 10.1016/j.chiabu.2009.12.003

Cowell, R. A., Cicchetti, D., Rogosch, F. A., & Toth, S. L. (2015). Childhood maltreatment and its effect on neurocognitive functioning: Timing and chronicity matter. Development and Psychopathology, 27(2), 521–533. 10.1017/S0954579415000139

Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children’s social adjustment. Psychological Bulletin, 115(1), 74–101. 10.1037/0033-2909.115.1.74

Cross, C. P., Copping, L. T., & Campbell, A. (2011). Sex differences in impulsivity: A meta-analysis. Psychological Bulletin, 137(1), 97–130. 10.1037/a0021591

Cuevas, K., Deater-Deckard, K., Kim-Spoon, J., Watson, A. J., Morasch, K. C., & Bell, M. A. (2014). What’s mom got to do with it? Contributions of maternal executive function and caregiving to the development of executive function across early childhood. Developmental Science, 17(2), 224–238. 10.1111/desc.12073

De Sanctis, V. A., Nomura, Y., Newcorn, J. H., & Halperin, J. M. (2012). Childhood maltreatment and conduct disorder: Independent predictors of criminal outcomes in ADHD youth. Child Abuse & Neglect, 36(11–12), 782–789. 10.1016/j.chiabu.2012.08.003

Defoe, I. N., Dubas, J. S., Figner, B., & van Aken, M. A. G. (2015). A meta-analysis on age differences in risky decision making: Adolescents versus children and adults. Psychological Bulletin, 141(1), 48–84. 10.1037/a0038088

Del Giudice, M., Ellis, B. J., & Shirtcliff, E. A. (2011). The Adaptive Calibration Model of stress responsivity. Neuroscience and Biobehavioral Reviews, 35(7), 1562–1592. 10.1016/j.neubiorev.2010.11.007

DePrince, A. P., Weinzierl, K. M., & Combs, M. D. (2009). Executive function performance and trauma exposure in a community sample of children. Child Abuse & Neglect, 33(6), 353–361. 10.1016/j.chiabu.2008.08.002

Derryberry, D., & Rothbart, M. K. (1988). Arousal, affect, and attention as components of temperament. Journal of Personality and Social Psychology, 55(6), 958–966. 10.1037/0022-3514.55.6.958

Drake, B., & Jonson-Reid, M. (2013). Poverty and child maltreatement. In J. E.Korbin & R. D.Krugman (Eds.), Handbook of child maltreatment (pp. 131–148). Springer. 10.1007/978-94-007-7208-3

Fernandez-Duque, D., Baird, J. A., & Posner, M. I. (2000). Executive attention and metacognitive regulation. Consciousness and Cognition, 9(2), 288–307. 10.1006/ccog.2000.0447

Figner, B., & Weber, E. U. (2011). Who takes risks when and why? Determinants of risk taking. Current Directions in Psychological Science, 20(4), 211–216. 10.1177/0963721411415790

Giedd, J. (2004). Structural magnetic resonance imaging of the adolescent brain. Annals of the New York Academy of Sciences, 1021(1), 77–85. 10.1196/annals.1308.009

Groh, A. M., Fearon, R. M. P., van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., & Roisman, G. I. (2017). Attachment in the early life course: Meta-analytic evidence for its role in socioemotional development. Child Development Perspectives, 11(1), 70–76. 10.1111/cdep.12213

Gross, J. J. (2015). The extended process model of emotion regulation: Elaborations, applications, and future directions. Psychological Inquiry, 26(1), 130–137. 10.1080/1047840X.2015.989751

Gross, J. J., & Jazaieri, H. (2014). Emotion, emotion regulation, and psychopathology. Clinical Psychological Science, 2(4), 387–401. 10.1177/2167702614536164

Guild, D. J., Alto, M. E., & Toth, S. L. (2017). Preventing the intergenerational transmission of child maltreatment through relational interventions. In D.Teti (Ed.), Parenting and family processes in child maltreatment and intervention. Child maltreatment solutions network (pp. 127–137). Springer. 10.1007/978-3-319-40920-7_8

Hesse, E., & Main, M. (2006). Frightened, threatening, and dissociative parental behavior in low-risk samples: Description, discussion, and interpretations. Development and Psychopathology, 18(02), 309–343. 10.1017/S0954579406060172

Hodgins, D. C., Schopflocher, D. P., el-Guebaly, N., Casey, D. M., Smith, G. J., Williams, R. J., & Wood, R. T. (2010). The association between childhood maltreatment and gambling problems in a community sample of adult men and women. Psychology of Addictive Behaviors, 24(3), 548–554. 10.1037/a0019946

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. 10.1080/10705519909540118

Jackson, Y., McGuire, A., Tunno, A. M., & Makanui, P. K. (2019). A reasonably large review of operationalization in child maltreatment research: Assessment approaches and sources of information in youth samples. Child Abuse and Neglect, 87, 5–17. 10.1016/j.chiabu.2018.09.016

Kavanaugh, B. C., Dupont-Frechette, J. A., Jerskey, B. A., & Holler, K. A. (2017). Neurocognitive deficits in children and adolescents following maltreatment: Neurodevelopmental consequences and neuropsychological implications of traumatic stress. Applied Neuropsychology: Child, 6(1), 64–78. 10.1080/21622965.2015.1079712

Kidd, C., Palmeri, H., & Aslin, R. N. (2013). Rational snacking: Young children’s decision-making on the marshmallow task is moderated by beliefs about environmental reliabilityCognition, 126(1), 109–114. 10.1016/j.cognition.2012.08.004

Knowles, A., Rinehart, J. K., Steinberg, L., Frick, P. J., & Cauffman, E. (2020). Risky sexual behavior among arrested adolescent males: The role of future expectations and impulse control. Journal of Research on Adolescence, 30(52), 562–579. 10.1111/jora.12499

Lavi, I., Katz, L. F., Ozer, E. J., & Gross, J. J. (2019). Emotion reactivity and regulation in maltreated children: A meta-analysis. Child Development, 90(5), 1503–1524. 10.1111/cdev.13272

Liu, R. T. (2019). Childhood maltreatment and impulsivity: A meta-analysis and recommendations for future study. Journal of Abnormal Child Psychology, 47(2), 221–243. 10.1007/s10802-018-0445-3

Luke, N., & Banerjee, R. (2013). Differentiated associations between childhood maltreatment experiences and social understanding: A meta-analysis and systematic review. Developmental Review, 33(1), 1–28. 10.1016/j.dr.2012.10.001

Manly, J. T. (2005). Advances in research definitions of child maltreatment. Child Abuse and Neglect, 29(5), 425–439. 10.1016/j.chiabu.2005.04.001

McEwen, B. S. (2008). Central effects of stress hormones in health and disease: Understanding the protective and damaging effects of stress and stress mediators. European Journal of Pharmacology, 583(2–3), 174–185. 10.1016/j.ejphar.2007.11.071

Mohr-Jensen, C., & Steinhausen, H. C. (2016). A meta-analysis and systematic review of the risks associated with childhood attention-deficit hyperactivity disorder on long-term outcome of arrests, convictions, and incarcerations. Clinical Psychology Review, 48, 32–42. 10.1016/j.cpr.2016.05.002

Mothes, L., Kristensen, C. H., Grassi-Oliveira, R., Fonseca, R. P., de Lima Argimon, I. I., & Irigaray, T. Q. (2015). Childhood maltreatment and executive functions in adolescents. Child and Adolescent Mental Health, 20(1), 56–62. 10.1111/camh.12068

Oshri, A., Sutton, T. E., Clay-Warner, J., & Miller, J. D. (2015). Child maltreatment types and risk behaviors: Associations with attachment style and emotion regulation dimensions. Personality and Individual Differences, 73, 127–133. 10.1016/j.paid.2014.09.015

Paus, T. (2005). Mapping brain maturation and cognitive development during adolescence. Trends in Cognitive Sciences, 9(2), 60–68. 10.1016/j.tics.2004.12.008

Pechtel, P., & Pizzagalli, D. A. (2011). Effects of early life stress on cognitive and affective function: An integrated review of human literature. Psychopharmacology, 214(1), 55–70. 10.1007/s00213-010-2009-2

Peters, E., Västfjäll, D., Gärling, T., & Slovic, P. (2006). Affect and decision making: A “hot” topic. Journal of Behavioral Decision Making, 19(2), 79–85. 10.1002/bdm.528

Pharo, H., Sim, C., Graham, M., Gross, J., & Hayne, H. (2011). Risky business: Executive function, personality, and reckless behavior during adolescence and emerging adulthood. Behavioral Neuroscience, 125(6), 970–978. 10.1037/a0025768

Pruessner, L., Barnow, S., Holt, D. V., Joormann, J., & Schulze, K. (2020). A cognitive control framework for understanding emotion regulation flexibility. Emotion, 20(1), 21–29. 10.1037/emo0000658

Rivera, P. M., Fincham, F. D., & Bray, B. C. (2018). Latent classes of maltreatment: A systematic review and critique. Child Maltreatment, 23(1), 3–24. 10.1177/1077559517728125

Rogosch, F. A., & Cicchetti, D. (2005). Child maltreatment, attention networks, and potential precursors to borderline personality disorder. Development and Psychopathology, 17(4), 1071–1089. 10.1017/S0954579405050509

Rogosch, F. A., Cicchetti, D., & Aber, J. L. (1995). The role of child maltreatment in early deviations in cognitive and affective processing abilities and later peer relationship problems. Development and Psychopathology, 7(4), 591–609. 10.1017/S0954579400006738

Rogosch, F. A., Cicchetti, D., Shields, A. M., & Toth, S. L. (1995). Parenting dysfunction in child maltreatment. In M. H.Bornstein (Ed.), Handbook of parenting, Vol.4. Applied and practical parenting (pp. 127–159). Lawrence Erlbaum Associates.

Romer, D., Betancourt, L. M., Brodsky, N. L., Giannetta, J. M., Yang, W., & Hurt, H. (2011). Does adolescent risk taking imply weak executive function? A prospective study of relations between working memory performance, impulsivity, and risk taking in early adolescence. Developmental Science, 14(5), 1119–1133. 10.1111/j.1467-7687.2011.01061.x

Romer, D., Reyna, V. F., & Satterthwaite, T. D. (2017). Beyond stereotypes of adolescent risk taking: Placing the adolescent brain in developmental context. Developmental Cognitive Neuroscience, 27, 19–34. 10.1016/j.dcn.2017.07.007

Rueda, M. R., Fan, J., McCandliss, B. D., Halparin, J. D., Gruber, D. B., Lercari, L. P., & Posner, M. I. (2004). Development of attentional networks in childhood. Neuropsychologia, 42(8), 1029–1040. 10.1016/j.neuropsychologia.2003.12.012

Sahakian, B., & Owen, A. (1992). Computerized assessment in neuropsychiatry using CANTAB: Discussion paper. Journal of the Royal Society of Medicine, 85(7), 399–402.

Schiebener, J., García-Arias, M., García-Villamisar, D., Cabanyes-Truffino, J., & Brand, M. (2015). Developmental changes in decision making under risk: The role of executive functions and reasoning abilities in 8- to 19-year-old decision makers. Child Neuropsychology, 21(6), 759–778. 10.1080/09297049.2014.934216

Sedlak, A. J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., & Li, S. (2010). Fourth national incidence study of child abuse and neglect (NIS-4): Report to Congress. U.S. Department of Health and Human Services, Administration for Children and Families.

Shafir, E. (2017). Decisions in poverty contexts. Current Opinion in Psychology, 18, 131–136. 10.1016/j.copsyc.2017.08.026

Shields, A. M., & Cicchetti, D. (1998). Reactive aggression among maltreated children: The contributions of attention and emotion dysregulation. Journal of Clinical Child Psychology, 27(4), 381–395. 10.1207/s15374424jccp2704_2

Shin, S. H., Miller, D. P., & Teicher, M. H. (2013). Exposure to childhood neglect and physical abuse and developmental trajectories of heavy episodic drinking from early adolescence into young adulthoodDrug and Alcohol Dependence, 127(1–3), 31–38. 10.1016/j.drugalcdep.2012.06.005

Siegfried, C. B., & Blackshear, K. L. (2016). Is it ADHD or child traumatic stress? A guide for clinicians. National Center for Child Traumatic Stress.

Slovic, P., Peters, E., Finucane, M. L., & MacGregor, D. G. (2005). Affect, risk, and decision making. Health Psychology, 24(4, Suppl), S35–S40. 10.1037/0278-6133.24.4.S35

Steinberg, L. (2004). Risk taking in adolescence: What changes, and why?Annals of the New York Academy of Sciences, 1021(1), 51–58. 10.1196/annals.1308.005

Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. Developmental Review, 28(1), 78–106. 10.1016/j.dr.2007.08.002

Steinberg, L. (2010). A dual systems model of adolescent risk-taking. Developmental Psychobiology, 52(3), 216–224. 10.1002/dev.20445

Stith, S. M., Liu, T., Davies, L. C., Boykin, E. L., Alder, M. C., Harris, J. M., Som, A., McPherson, M., & Dees, J. E. M. E. G. (2009). Risk factors in child maltreatment: A meta-analytic review of the literature. Aggression and Violent Behavior, 14(1), 13–29. 10.1016/j.avb.2006.03.006

Sturge-Apple, M. L., Davies, P. T., Cicchetti, D., Hentges, R. F., & Coe, J. L. (2017). Family instability and children’s effortful control in the context of poverty: Sometimes a bird in the hand is worth two in the bush. Development and Psychopathology, 29(3), 685–696. 10.1017/S0954579416000407

Tanner, J. L., & Arnett, J. J. (2016). The emergence of emerging adulthood: The new life stage between adolescence and young adulthood. In A.Furlong (Ed.), Routledge handbook of youth and young adulthood (2nd ed., pp. 34–40). Routledge. 10.4324/9781315753058

Tarren-Sweeney, M. (2013). An investigation of complex attachment- and trauma-related symptomatology among children in foster and kinship care. Child Psychiatry and Human Development, 44(6), 727–741. 10.1007/s10578-013-0366-x

Teicher, M. H., & Samson, J. A. (2016). Annual research review: Enduring neurobiological effects of childhood abuse and neglect. Journal of Child Psychology and Psychiatry, 57(3), 241–266. 10.1111/jcpp.12507

Tofighi, D., & MacKinnon, D. P. (2011). RMediation: An R package for mediation analysis confidence intervals. Behavior Research Methods, 43(3), 692–700. 10.3758/s13428-011-0076-x

Topitzes, J., Mersky, J. P., & Reynolds, A. J. (2012). From child maltreatment to violent offending: An examination of mixed-gender and gender-specific models. Journal of Interpersonal Violence, 27(12), 2322–2347. 10.1177/0886260511433510

U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. (2020). Child maltreatment 2018. https://www.acf.hhs.gov/cb/resource/child-maltreatment-2018

Uekermann, J., Kraemer, M., Abdel-Hamid, M., Schimmelmann, B. G., Hebebrand, J., Daum, I., Wiltfang, J., & Kis, B. (2010). Social cognition in attention-deficit hyperactivity disorder (ADHD). Neuroscience and Biobehavioral Reviews, 34(5), 734–743. 10.1016/j.neubiorev.2009.10.009

Vachon, D. D., Krueger, R. F., Rogosch, F. A., & Cicchetti, D. (2015). Assessment of the harmful psychiatric and behavioral effects of different forms of child maltreatment. JAMA Psychiatry, 72(11), 1135–1142. 10.1001/jamapsychiatry.2015.1792

Warmingham, J. M., Handley, E. D., Rogosch, F. A., Manly, J. T., & Cicchetti, D. (2019). Identifying maltreatment subgroups with patterns of maltreatment subtype and chronicity: A latent class analysis approach. Child Abuse & Neglect, 87, 28–39. 10.1016/j.chiabu.2018.08.013

Wechsler, D. (2003). Wechsler Intelligence Scale for Children–Fourth Edition (WISC–IV). The Psychological Corporation.

Weller, J. A., & Fisher, P. A. (2013). Decision-making deficits among maltreated children. Child Maltreatment, 18(3), 184–194. 10.1177/1077559512467846

Yazdi, K., Rumetshofer, T., Gnauer, M., Csillag, D., Rosenleitner, J., & Kleiser, R. (2019). Neurobiological processes during the Cambridge gambling task. Behavioural Brain Research, 356, 295–304. 10.1016/j.bbr.2018.08.017

Zelazo, P. D., & Müller, U. (2010). Executive function in typical and atypical development In U.Goswami. (Ed.), The Wiley-Blackwell handbook of childhood cognitive development (2nd ed., pp. 574–603). 10.1002/9781444325485.ch22

Zois, E., Kortlang, N., Vollstädt-Klein, S., Lemenager, T., Beutel, M., Mann, K., & Fauth-Bühler, M. (2014). Decision-making deficits in patients diagnosed with disordered gambling using the Cambridge Gambling task: The effects of substance use disorder comorbidity. Brain and Behavior, 4(4), 484–494. 10.1002/brb3.231

Submitted: February 16, 2020 Revised: November 13, 2020 Accepted: December 9, 2020