REP 4

Children and Youth Services Review 50 (2015) 53–63

Contents lists available at ScienceDirect

Children and Youth Services Review

journal homepage: www.elsevier.com/locate/childyouth

Why should child welfare pay more attention to emotional maltreatment?☆

Diana English a,⁎, Richard Thompson b, Catherine Roller White a, Dee Wilson a

a Casey Family Programs, Seattle, WA, United States b Richard H. Calica Center for Innovation in Children and Family Services, Juvenile Protective Association, Chicago, IL, United States

☆ The authors would like to thank Casey Family Programs ⁎ Corresponding author at: Casey Family Programs,

Seattle, WA 98121, United States. Tel.: +1 206 282 7300. E-mail address: [email protected] (D. English).

http://dx.doi.org/10.1016/j.childyouth.2015.01.010 0190-7409/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history: Received 15 September 2014 Received in revised form 9 January 2015 Accepted 12 January 2015 Available online 17 January 2015

Keywords: Emotional maltreatment Emotional abuse Emotional neglect Child welfare Psychological maltreatment

A significant body of research indicates that emotional maltreatment (EMT) is harmful to children, resulting in long-term negative impacts on emotional and behavioral development. The child welfare system's focus on physical abuse, physical neglect, and sexual abuse has led a relative lack of attention to EMT. Reported rates of EMT vary widely across states – ranging from 0.2% to 44.9% in a recent national report on child maltreatment – indicating that it is not being measured consistently. This paper uses data collected by the Longitudinal Studies of Child Abuse and Neglect (LONGSCAN) to (1) describe the nature and characteristics of emotional maltreatment experienced by 846 LONGSCAN youth across time, and (2) describe the relation between four subtypes of emo- tional maltreatment (psychological safety and security, acceptance and self-esteem, autonomy, and restriction) and child trauma symptoms and risk behaviors at age 18. Exposure to EMT was related to increased trauma symptoms and risky behaviors. EMT is common, identifiable, harmful, and potentially preventable; and a better understanding of it will help to inform the provision of effective child welfare and mental health services to children and their families. Findings suggest a need for greater understanding of parental behaviors, and the motivations behind them, that result in emotionally harmful outcomes for children, as well as a better under- standing of appropriate interventions for children who experience various types of EMT.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Interest in the topic of emotional maltreatment (EMT) was evident as early as the 1950s, and the past 60 years have been witness to ongoing discussions, refinements and research (Garbarino, 2011). For the past 20 years, significant discussion about definitions of emotional maltreat- ment (variously called psychological abuse, psychological neglect, emo- tional abuse, and emotional neglect) has ensued. While some scholars have noted limited agreement and uncertainty regarding definitions of EMT (Feerick & Snow, 2006, as cited in Trickett, Mennen, Kim, & Sang, 2009), a review of the research literature indicates a growing scholarly consensus and a generally accepted understanding among researchers, mental health practitioners, and child welfare staff and leaders regarding the defining characteristics of EMT (Brassard & Donovan, 2006; Egeland, Sroufe, & Erickson, 1983; Burnett, 1993, as cited in Glaser, 2002; Wolfe & McIsaac, 2011). Related terms include emotional abuse, emotional neglect and psychological maltreatment. As proposed by Glaser (2011), the term maltreatment is preferable as it is inclusive of both acts of omission and commission. The use of emotional vs. psychological is more a matter of

for its support of these analyses. 2001 8th Avenue, Suite 2700,

convention, as it is not clear that distinctions can be made in the two terms; emotional maltreatment is the one most commonly used in an American context (Glaser, 2011).

Examples of EMT definitions include “psychological tactics aimed at undermining emotional security and sense of self that includes guilt induction, and exertion of power through psychologically coercive means” (Bornstein, 2006); a “repeated pattern of behavior that conveys to children that they are worthless, unloved, unwanted, only of value in meeting another's needs, or seriously threatened with physical or psy- chological violence” (Hart, Brassard, & Karlson, 1996), or “excessive and continuing criticism, denigration, terrorizing, repeated blaming insults, or threats” (Brassard & Donovan, 2006). Some argue that the difference between poor parenting and EMT is the degree of chronicity, severity, and potential harm to the child or youth (Wolfe & McIsaac, 2011).

Several models and approaches to classifying EMT exist, with varying levels of research that support using one method or another. In 2006, Brassard and Donovan identified and compared nine different ap- proaches to conceptualizing EMT (Brassard & Donovan, 2006, Fig. 7.1, p. 156). As noted by Brassard and Donovan, the models share many commonalities but have some differences. In 2005, Schneider and col- leagues modified the 27 subtypes of EMT initially developed by Barnett, Manly, and Cicchetti (1993) to create the Longitudinal Studies on Child Abuse and Neglect (LONGSCAN) Modified Maltreatment Classification System (MMCS; English, Bangdiwala, & Runyan, 2005),

54 D. English et al. / Children and Youth Services Review 50 (2015) 53–63

which provides the four categories of EMT utilized in this study: 1) fail- ure to support psychological safety and security; 2) failure to support acceptance and self-esteem; 3) failure to allow age-appropriate auton- omy; and 4) restriction (see the Methods section for a description).

Although a definition of EMT that includes chronicity, severity, and potential harm is generally accepted, some research suggests that even one or two incidents of verbal abuse per year can be harmful to children. In other words, it is not just chronic EMT that is harmful. Donovan and Brassard (2011) argue that the experience of EMT results in variable effects based on differences in intensity and whether the EMT is stable, increasing, or decreasing in frequency across time. Fur- thermore, while definitions differentiate characteristics of emotional maltreatment into emotional abuse (EMT-EA) and emotional neglect (EMT-EN) (Brassard, Hart, & Hardy, 1993; Garbarino, Guttmann, & Seeley, 1986; Iwaniec, 1995; O'Hagen, 1995), research suggests that there are different impacts for hostile, indifferent, and misguided and/or inattentive parenting (Iwaniec, 2006). Until further research is conducted on children's experiences of maltreatment and the related consequences of those experiences, it is difficult to arrive at definitive conclusions about these definitional issues.

1.1. Incidence and prevalence of EMT

Although definitions and classifications vary, findings from a number of research endeavors demonstrate that EMT is common. In 2009, based on a national community sample of 4549 children and youth, Finkelhor, Turner, Ormrod, and Hamby (2009) reported a past-year 6.4% victimiza- tion rate for psychological or emotional abuse, and a lifetime 11.9% victimization rate for psychological or emotional abuse. Among the 14 to 17-year-olds in the sample, 22.6% had experienced psychological or emotional abuse in their lifetime.

The Fourth National Incidence Study (NIS-4) provides information on national rates of EMT. The NIS-4 compared rates of child maltreatment based on the knowledge of “sentinels” of child abuse and neglect cases in 122 counties around the United States. NIS-4 used two standards, a harm standard and a less stringent endangerment standard, to classify cases of child maltreatment identified by sentinels (mostly mandated reporters) and including both cases reported and not reported to Child Protective Services (CPS). NIS-4 found large reductions in physical, sexual, and emotional abuse using both the harm and endangerment standards from 1993 to 2006, but identified a doubling of emotional neglect using the endangerment standard from 1993 to 2006. The in- crease in emotional neglect was much larger for children ages 0–5 than for older children; there was a 259% increase for children ages 0–2 and a 214% increase for children ages 3–5 compared to a 10% increase for 6–8 year olds (Sedlak et al., 2010).

While the increase in emotional neglect as measured by the NIS-4 may reflect increased awareness of EMT, any increase in awareness has not translated into consistent identification and reporting. An exam- ination of states' child welfare system data on the identification and substantiation of EMT reveals widely varying practices in terms of defi- nitions and reports to national databases. The 2012 National Child Abuse and Neglect Data System (NCANDS) report indicates that in 18 states, less than 1% of child victims were substantiated for EMT, and in nine states, 0.5% or less of child victims were substantiated for EMT. However, in six states, more than a quarter of child victims were sub- stantiated for EMT, and in two states, more than 40% of child victims were substantiated for EMT. Such dramatic differences in the percent- ages of child victims substantiated by states' child welfare systems for EMT that exceed, at the extremes, 200 to 1 (44.9% vs. 0.2%) suggest major differences in the way EMT is conceptualized and measured rather than differences in incidence rates per se (U. S. Department of Health et al., 2013). Both NCANDS and NIS studies likely reflect changes in advocates’ and practitioners’ understanding of EMT, not just changes in incidence.

Concerns about the accuracy regarding the incidence of EMT in official records are further supported by research examining Child Protective Services (CPS) records to determine whether EMT was reported but not documented in the official classification of maltreatment. In an examina- tion of CPS records, Trickett et al. (2009) found that nearly 50% of 303 adolescents who had been substantiated for maltreatment experienced EMT, although only 9% were identified as having experienced EMT at the time they were referred to CPS. Furthermore, when asked about their maltreatment experiences, youth reported EMT occurred at a rate more than six times higher than the rate found in official reports for the same youth (Everson et al., 2008).

1.2. Why does the incidence of EMT matter?

A significant body of research indicates that EMT is harmful to children, resulting in negative impacts on both emotional and behavioral development across developmental stages (Claussen & Crittenden, 1991; Hart, Binggeli, & Brassard, 1997; Hart & Brassard, 1987; Miller-Perrin, Perrin, & Kocur, 2009). In their longitudinal study, Erickson and Egeland (1996) found that EMT-EN in the first few years of life predicted later social withdrawal, angry noncompliance, and lower academic achievement. Graham-Berman and Levendovsky (1998) found that exposure to traumatic events such as domestic violence was related to post-traumatic stress symptomatology in children. McGee, Wolfe, and Wilson (1997) found that EMT-EA, such as hostility and rejection, was associated with child behaviors that were negative and hostile. In one longitudinal, multi-informant study of 421 youth from two low income middle schools, verbal aggression (EMT-EA) by parents was associated with depressive symptoms, delinquency, and perceived victimization (Donovan & Brassard, 2011). In another study, Miller-Perrin et al. (2009) found that experiencing EMT-EA as a child predicted higher psychological distress scores in 298 college students and also predicted anxiety, depression, and paranoia. This study also found that psychologi- cal aggression by a parent was the strongest predictor of later psycholog- ical symptoms in young adulthood (Wekerle et al., 2009). Furthermore, EMT can affect physiology, heighten children's stress responses, and neg- atively impact neurodevelopment (Yates, 2007).

In other research, childhood EMT has been linked to poor immune system response in adulthood (Fagundes et al., 2012) and poorer adult health indicators (Rodgers et al., 2004). Furthermore, given that EMT often occurs in conjunction with other types of child maltreatment, EMT may persist even after physical abuse, sexual abuse, or neglect of basic physical needs has stopped. Children may not feel safe in homes where EMT periodically occurs even when they are deemed safe by CPS caseworkers or other professionals, because these professionals are fo- cused on other forms of abuse or neglect. EMT may have long-term toxic effects on the nurturing environment in which other types of child maltreatment occur, and in doing so creates major obstacles to recovery from maltreatment, trauma, and other adversities.

Research has also demonstrated the importance of examining out- comes by subtype of EMT (Taussig & Culhane, 2010). Different types of EMT have been found to be associated with different developmental outcomes. For example, a previous analysis of LONGSCAN data found that among 806 eight-year-olds, subtypes of EMT had distinct effects: psychological threats to safety and security predicted anger/irritability, while threats to a child's sense of self, self-esteem, and autonomy pre- dicted symptoms of post-traumatic stress (Schneider, Ross, Graham, & Zielinski, 2005). In a prospective study of 243 children in out-of-home care, Taussig and Culhane (2010) found that for males, EMT-EA was re- lated to poorer self-esteem, and abandonment was related to higher levels of anxiety and lower levels of life satisfaction after controlling for other types of child maltreatment.

Although EMT is often found in conjunction with other forms of maltreatment, EMT contributes uniquely to negative outcomes (see, for example, Claussen & Crittenden, 1991; Miller-Perrin et al., 2009; Sullivan, Fehon, Andres-Hyman, Lipschitz, & Grilo, 2006; Wright, Crawford, & Del

55D. English et al. / Children and Youth Services Review 50 (2015) 53–63

Castillo, 2009) and that the harm extends into adulthood (Hart et al., 1997).

1.3. What has been done already and what can this paper add to our understanding of EMT?

Numerous articles have been published regarding definitions of emotional maltreatment, and using main effects models to examine the relationship of EMT to outcomes in conjunction with and indepen- dent of other types of maltreatment (Trickett, Kim, & Prindle, 2011). Despite significant progress in understanding EMT, some researchers continue to advocate for increased definitional clarity and useable guidelines for assessment (Brassard & Donovan, 2006; Hart, Brassard, Binggeli, & Davidson, 2002; Smith Slep, Heyman, & Snarra, 2011).

Additionally, researchers have suggested the need for a more nuanced assessment that includes “unpacking” the broad concept of EMT into sub- types (Baker & Festinger, 2011; Glaser, 2002; Schneider et al., 2005; Trickett et al., 2009), especially differences associated with EA and EN (Baker, 2009; Egeland, 2009; Shaffer, Yates, & Egeland, 2009; Wright et al., 2009). In addition, researchers have suggested that additional re- search on age, gender, and ethnicity effects (Baker & Festinger, 2011; Egeland, 1997; Flory et al., 2009; Simmel & Shpiegel, 2013) and develop- mental period or age of onset (Hart et al., 2002; Miller-Perrin et al., 2009; Straus & Field, 2003) is necessary to further understand EMT. Some re- searchers have argued that studies should be conducted to examine other dimensions of maltreatment such as severity and/or chronicity (Pears, Kim, & Fisher, 2008) and whether there are delayed effects across children's developmental stages (Donovan & Brassard, 2011; Herrenkohl, Egolf, & Herrenkohl, 1997). Further, some researchers have suggested a need to include factors from different ecological levels – such as child, family, parent–child interaction, and neighborhood con- text – to understand the context of EMT (Cicchetti & Valentino, 2006).

Finally, Brassard and Hardy (1997) suggested over a decade ago that treatment of EMT is not effective because interventions are not specific to the dynamics of the “subtype” of EMT experienced by a child or youth. Glaser (2002, 2011) suggests that there is a need to determine the “driving” category of EMT within a specific context to determine the most effective intervention, as different subtypes of EMT can require different therapeutic approaches, and the sequencing of interventions may vary.

There are at least nine different models for conceptualizing or grouping specific behaviors into a classification scheme for EMT (See Fig. 7.1, p. 156, Brassard & Donovan, 2006). In their effort to examine commonalities across approaches, Brassard and Donovan (2006) used the American Professional Society on the Abuse of Children (APSAC) categories, including item-level descriptions, to compare models uti- lized in the National Incidence Study (Baily & Baily, 1986; Barnett et al., 1993; Garbarino et al., 1986; Giovannoni & Becerra, 1979; Glaser, 2002; Kairys, Johnson, & Committee on Child Abuse and Neglect, 2002; McGee & Wolfe, 1991; Moran, Bifulco, Ball, Jacobs, & Benaim, 2002; Sedlak & Broadhurst, 1996). An examination of these comparisons reveal that many of the models have common elements; for example, all models include belittling, denigrating or other rejecting behaviors; failing to show affection, caring, or love; or threatening/ perpetrating violence against a child. Other categories in the APSAC model were included in some of the referenced models and not others, but all models include at least one item in each of the six APSAC catego- ries of EMT; that is, spurning, terrorizing, isolating, exploiting/corrupting, denying emotional responsiveness, and mental health/medical/ educational neglect. While a number of different approaches to classify- ing and analyzing EMT are available, for multiple reasons we selected the Maltreatment Classification Scheme initially developed by Barnett et al. (1993) and modified by LONGSCAN. For example, the MCS allowed for the assignment of severity levels for different types of maltreatment, and previous research on the LONGSCAN sample revealed significant

relationships between the subtypes of EMT as outlined in this research and child outcomes in middle childhood.

In summary, a review of research to date suggests that EMT is more common than generally understood. Significant rates of EMT are found in community samples not reported to CPS, as well as for children who are involved in a report to CPS. Further, there has been a significant in- crease in the recognition and/or reporting of EMT by mandated re- porters over the past two decades, which likely reflects an increased understanding of the social, emotional, and behavioral consequences of these behaviors on children's short and long term outcomes. A num- ber of research studies have demonstrated that EMT is harmful to chil- dren, youth, and adults in multiple functional domains, and that additional research is needed on subtypes of EMT and other factors that could improve our understanding of this phenomenon in order to design and conduct effective interventions.

One potential dataset that could inform our understanding of the subtypes of EMT is the LONGSCAN dataset. LONGSCAN is a large, longitu- dinal study tracking maltreated and at risk children from age 4 to age 18, and as noted earlier includes MMCS coding of official reports of EMT. The high-risk nature of the sample allows a great deal of power to identify dif- ferent characteristics and dimensions of EMT, and the longitudinal nature of the sample allows a broader and developmentally informed under- standing of the longer-term implications for child and youth develop- ment. However, it is important to keep in mind that the high-risk nature of the sample may not be representative of the general population of children and families. Other limitations of this approach to understand- ing EMT are discussed near the end of this paper.

EMT is common, identifiable, harmful, and potentially preventable, and a better understanding of it will help inform the provision of effec- tive child welfare and mental health services to children and families. EMT also shapes the nurturing environment in which other types of child maltreatment occur, and affects children's capacity to recover from multiple adversities, including physical abuse, sexual abuse, and neglect of basic needs. Adding to and extending previous research on EMT, this study will:

1. Describe the nature and characteristics of emotional maltreatment experienced by the 846 LONGSCAN children/youth across time, as reported by referents.

2. Describe the relation between subtypes of emotional maltreatment and child trauma symptoms and risk behaviors.

2. Methods

2.1. Procedure

This was a secondary data analysis using data collected by LONGSCAN. LONGSCAN was a consortium of a coordinating center and five regionally- based study sites. LONGSCAN focused on children identified as at risk for maltreatment, or who had been reported to CPS as victims of maltreat- ment. Although the recruitment criteria varied somewhat by site, the sites shared age-specific data collection protocols (for a full description, see Runyan et al., 1998). Target children and their caregivers were recruit- ed into the LONGSCAN study at ages 4 or 6 and were assessed at various follow-up points. The data reported here on maltreatment were collected at ages 4, 6, 8, 12, 14, 16, and 18. The outcome data were collected at age 18.

2.2. Participants

The initial LONGSCAN sample included 1354 children recruited at the age of 4 or 6. The current analysis focused on the portion of the sample that had data available on age 18 outcomes (N = 846, 62.5%). There were no significant demographic or baseline functioning differences be- tween the overall LONGSCAN sample and the sample included in these

Table 1 Participant characteristics, trauma symptoms, and risk behaviors.

Variable M SD % N

Site Eastern 22.9 201 Southeastern 18.4 161 Midwest 16.8 147 Northwest 20.4 179 Southwest 21.6 189 Age 18.51 0.61

Race White 24.5 215 African American 55.8 489 Hispanic 6.6 58 Other 13.1 115

Gender Male 44.6 391 Female 55.4 486

Maltreatment allegations Physical abuse 35.0 387 Sexual abuse 18.7 164 Neglect 58.7 515 Emotional maltreatment 35.6 312

Trauma symptoms (N = 874) Anger/irritability 6.17 6.37 Anxious arousal 3.97 4.65 Depression 4.56 5.47 Intrusive experiences 4.64 5.60 Defensive avoidance 6.26 6.48 Dissociation 4.13 5.07

Risk behaviors (N = 859) Suicidal thoughts 5.1 44 Risky sex 19.0 163 Ever arrested 27.6 237 Cigarette smoking 1.42 2.08 Alcohol 0.80 1.23 Marijuana 0.86 1.65 Other drugs 5.8 50

Note: N = 877. Possible scores on anger/irritability, anxious arousal, depression, intrusive experiences, and defensive avoidance all ranged from 0 to 24, while the range for dissociation was 0 to 27.

56 D. English et al. / Children and Youth Services Review 50 (2015) 53–63

analyses. The demographic description of the analysis sample is presented in Table 1.

2.3. Measures

2.3.1. Control variables Demographic variables were assessed at each interview. Information

on variables that did not vary by time (race/ethnicity, gender, and study site) was gathered at the baseline assessment.

2.3.2. Assessment of emotional maltreatment Official reports to CPS were coded by trained abstractors at each site.

These data were coded using the Maltreatment Classification System (MCS) originally developed by Barnett et al. (1993) and modified by the LONGSCAN investigators (MMCS; English et al., 2005). Each site reviewed CPS administrative records for allegations of child maltreat- ment at least every two years, through the course of the LONGSCAN data collection. Allegations to CPS were coded by trained staff using the MMCS. The measure had acceptable reliability; in addition, the MMCS had expected correlations with child self-reports of maltreat- ment (Everson et al., 2008) and the items assessing EMT correlated in expected ways with child and youth outcomes (Schneider et al., 2005; Trickett et al., 2009). The analyses presented here focused on data re- garding children's experiences before the age 18 assessment.

These coded data included physical abuse, sexual abuse, and neglect, as described below. It also included emotional maltreatment. Twenty-seven classes of coded behavior were included in emotional

maltreatment, including four broad subtypes, described by Schneider et al. (2005): (1) failure to support psychological safety and security; (2) failure to promote acceptance and self-esteem; (3) failure to allow age-appropriate autonomy; and (4) restriction. The specific forms of emotional maltreatment, their frequencies in the analysis sample, and the overall exposures to these forms in the LONGSCAN samples are pre- sented in Table 2, which is discussed further in the results section.

2.3.3. Other forms of maltreatment Information regarding other forms of maltreatment was also

extracted for the time period from birth to the age 18 assessment. The following three types of maltreatment were dichotomized:

1. Physical abuse (inflicted blows or injury to the body); 2. Sexual abuse (sexual exposure, exploitation, molestation, or penetra-

tion); or 3. Neglect (failure to provide for a child's physical well-being, or lack of

adequate supervision).

2.3.4. Outcome measures Trauma Symptom Inventory (TSI). The TSI (Briere, 1995) is a 100-item

test evaluating acute and chronic traumatic symptomatology related to traumatic experiences. The TSI assessment is more broadly defined than symptoms of PTSD per se. The following subscales were used in these analyses:

1. Anger/irritability measures the experience of internalized anger and irritability, and angry thoughts and behavior;

2. Anxious arousal measures the experience of anxiety; 3. Depression measures depressed mood and thoughts; 4. Intrusive experiences measures posttraumatic reactions and symp-

toms such as flashbacks and intrusive thoughts and memories; 5. Defensive avoidance measures conscious attempts to avoid unpleasant

thoughts or memories; 6. Dissociation measures largely unconscious defensive avoidance of

distress.

Possible scores on anger/irritability, anxious arousal, depression, intrusive experiences, and defensive avoidance all ranged from 0 to 24, while the range for dissociation was 0 to 27. Internal consistency alphas for these scales were all very high, ranging from .86 for anxious arousal to .91 for depression.

Risk behaviors: suicidal ideation. Suicidal ideation was assessed using a single item taken from the Youth Risk Behavior Surveillance System (YRBSS), a Centers for Disease Control assessment of national rates of youth risk behavior that is conducted periodically (Brener et al., 2004). This item read, “During the last 12 months, did you ever seriously consider attempting suicide?” In the LONGSCAN studies, this item was administered as part of a larger scale that included a subset of items from the YRBSS (Knight, Smith, Martin, & The LONGSCAN Investigators, 2011).

Risk behaviors: risky sexual behavior. The LONGSCAN administration of risk items from the YRBSS (Brener et al., 2004; Knight et al., 2011) also included several questions about sexual activity. In particular, youth were asked whether they had ever had sexual intercourse. Youth who reported that they had had sexual intercourse were asked in a follow-up question whether they had used condoms during their last sexual intercourse. Youth who reported having had intercourse without a condom were coded as having engaged in risky sex. Youth who reported not having had intercourse or having had intercourse and using condoms were coded as not having engaged in risky sex.

Risk behaviors: arrests. As part of the LONGSCAN assessment of criminal justice involvement (Knight et al., 2011), youth were asked, “How many times in your life have you ever been arrested?” Responses were dichotomized into “never” versus any other number of times.

Table 2 Emotional maltreatment allegations.

Emotional maltreatment subtype and items % (N) of allegations Mean age of exposure (SD) Mean number of allegations per child (SD)

Psychological safety and security 63.6% (479) 6.45 (4.94) 0.54 (1.12) Discipline through intimidation 6.1% (46) 7.71 (5.26) Nonviolent marital conflict 4.2% (32) 9.64 (4.95) Threat of injury 4.0% (30) 10.14 (5.48) Exposure to extreme behaviors 19.4% (146) 5.59 (4.25) Threatens suicide 2.4% (18) 6.54 (6.24) Extreme marital violence 8.2% (62) 6.65 (4.58) Blames for suicide or death of family member 0 – Suicidal attempt 1.2% (9) 3.18 (3.30) Homicidal threat 2.9% (22) 6.50 (4.73) Abandonment 15.1% (114) 5.29 (4.71)

Acceptance and self esteem 25.8% (194) 6.28 (5.11) 0.22 (0.63) Undermines relationships 1.1% (8) 8.97 (4.69) Ridicules child 4.8% (36) 6.88 (5.48) Ignores child 3.6% (27) 2.43 (3.00) Rejection/inattentive to child 3.3% (25) 5.28 (5.44) Blames for marital or family problems 0.3% (2) 0.74 (0.82) Inappropriate expectations 0.5% (4) 6.29 (4.66) Derogatory names 3.3% (25) 8.65 (5.52) Negativity/hostility 8.9% (67) 6.84 (4.58)

Age-appropriate autonomy 7.0% (53) 9.93 (4.70) 0.06 (0.27) Inappropriate responsibility 5.6% (42) 10.46 (4.61) Prohibits age-appropriate socialization 0.5% (4) 10.04 (5.10) Places child in role reversal 0.3% (2) 11.59 (3.66) Thwarts child's development 0.7% (5) 4.68 (2.70)

Restriction 3.6% (27) 6.41 (5.29) 0.04 (0.22) Binding 0.5% (4) 5.04 (6.09) Confinement/isolation 2.0% (15) 5.62 (4.74) Uses restrictive methods for less than 2 h 0.1% (1) 16.18 (−) Extremely restrictive methods for more than 2 h 0.4% (3) 9.20 (4.98) Close confinement 0.5% (4) 6.27 (6.24)

Note: N = 877; N for allegations = 753. Mean number of allegations per child is the mean number of allegations of each subtype of EMT experienced per child.

57D. English et al. / Children and Youth Services Review 50 (2015) 53–63

Risk behaviors: substance use. The LONGSCAN administration of risk items from the YRBSS (Brener et al., 2004; Knight et al., 2011) included several questions about legal and illegal substance use. The following questions about substance use were selected for the current analyses:

1. Cigarette use. “During the past 30 days, on how many days did you smoke cigarettes?” Response options included: 0 = has never smoked cigarettes, 1 = 0 days, 2 = 1 or 2 days, 3 = 3 to 5 days, 4 = 6 to 9 days, 5 = 10 to 19 days, 6 = 20 to 29 days, 7 = All 30 days.

2. Problem drinking. “During the past 30 days, on how many days have you been drunk or very high from drinking alcoholic beverages?” Response options were identical to those listed for cigarette use.

3. Marijuana use. “During the past 30 days, how many times did you use marijuana?” Response options included: 0 = never used marijuana, 1 = 0 times, 2 = 1 or 2 times, 3 = 3 to 9 times, 4 = 10 to 19 times, 5 = 20 to 39 times, 6 = 40 or more times.

4. Other illegal drugs. “In the past year, did you use or try any other illegal drugs or drugs that were not prescribed to you by a doctor?” Responses were dichotomized into “never” versus any other number of times.

2.4. Data analysis

The data analyses summarized in this article were conducted using SPSS, Version 19. The central analyses involved predicting various trauma symptoms and risk behaviors, using a block of control variables (youth gender, race, and exposure to other forms of maltreatment) and the four subtypes of emotional maltreatment (as indicated by the number of allegations for each subtype), all entered simultaneously. This was done using multilevel linear mixed models (Peugh & Enders, 2005), with youth nested in site (site as random effects). This approach accounted for possible shared variance among youth recruited and assessed at the same site.

3. Results

Overall, of the 877 youth in the analysis sample, 36% (n = 312) had an allegation of emotional maltreatment. There was a great deal of overlap between having an allegation of emotional maltreatment and having an allegation of another form of maltreatment (physical abuse, sexual abuse, or neglect). Of the 312 youth with an allegation of emotional maltreatment, 98% (n = 307) had an allegation of another form of maltreatment. Specifically, 70% (n = 217) had an allegation of physical abuse, 34% (n = 107) had an allegation of sexual abuse, and 89% (n = 277) had an allegation of neglect.

3.1. The nature and characteristics of emotional maltreatment

As noted earlier, Table 2 describes proportion of allegations of each type, as well as the frequency and mean age of exposure to the different forms of emotional maltreatment. Of note, roughly two in three allega- tions of emotional maltreatment (64%) were for behaviors that can be classified as “attacks” on children's psychological safety and security (PSS), that is, behaviors that created an environment in which the child feels threatened and unsafe. Within this category, the majority of the psychological safety allegations were related to exposing a child to extreme, unpredictable, and/or inappropriate behaviors (e.g., violence toward other family members other than intimate partner violence), psychotic or paranoid ideation that results in violent outbursts that terrorize the child); this sub-set of PSS accounted for 19% of all EMT allegations. The second largest group of allegations in the PSS category was related to abandonment (15%); the primary caregiver had aban- doned the child for 24 h or longer without any indication of when or whether he/she would return and where they could be located. The third highest group of allegations in the PSS category was for exposure to extreme marital violence (8%).

58 D. English et al. / Children and Youth Services Review 50 (2015) 53–63

The acceptance and self-esteem (ASE) category accounted for roughly one quarter of all allegations of EMT (26%). The most frequent ASE- alleged behaviors were negativity/hostility (9%), ridiculing the child (5%), ignoring the child (4%), and rejecting or being inattentive toward the child (4%). Autonomy and restriction comprised a smaller percent of allegations (11% combined); placing inappropriate responsibility on a child was the largest sub-category of autonomy (6%), and confinement/ isolation was the largest restriction sub-category (2%).

With few exceptions, the mean age for all the reported behaviors was between 5 and 10 years of age, with standard deviations ranging from 3 to almost 8 for some of the low-frequency behaviors. The children in this study reported as emotionally maltreated in the referral narrative ranged in age from toddlers to early- or mid-adolescence. The average number of allegations of each type, per child, is presented in the final column of Table 2. These numbers track closely with the proportion of allegations discussed earlier: the highest frequency was for psychological safety and security, followed by acceptance and self-esteem, with autonomy and restriction both having low average frequencies. These four frequency variables were used as predictors of outcomes.

3.2. Links between emotional maltreatment and outcomes

3.2.1. Trauma symptoms Links between the frequency of EMT subtypes and trauma symptoms

are presented in Tables 3 and 4. In terms of control variables, there were some inconsistent effects for race/ethnicity. There were, however, strong effects for gender, with female youth having more trauma symptoms than male youth for most trauma categories. In addition, sexual abuse uniquely predicted Anxious Arousal, and neglect uniquely predicted Intrusive Experiences, a category that includes post-traumatic stress symptoms.

In terms of types of emotional maltreatment, psychological safety and security uniquely predicted the following trauma symptoms: Anger/ Irritability, Anxious Arousal, and Depression. Acceptance and self-esteem had a significant inverse relationship with Anxious Arousal. There were no significant effects of autonomy or restriction. The absence of findings for these two subtypes of EMT may be due to low base rates for both autonomy and restriction (7.0% and 3.6%, respectively). There were no significant effects of any forms of emotional maltreatment on Intrusive Experiences, Defensive Avoidance, or Dissociation.

Table 3 Models predicting trauma symptoms, part 1.

Fixed effect Anger/irritability Anxious

B SE B B

Intercept 8.48 0.90 9.37⁎⁎⁎ 5.59

Race White −0.42 0.74 −0.56 0.32 African American −1.15 0.71 −1.63 −1.17 Hispanic −1.76 1.04 −1.68 −1.10

Gender Male −0.85 0.44 −1.92 −0.92

Other maltreatment Physical abuse 0.04 0.57 0.07 −0.43 Sexual abuse 1.20 0.63 1.90 1.20 Neglect 0.49 0.51 0.95 0.26

Schneider EMT subtypes Psychological safety 0.52 0.22 2.27⁎ 0.44 Acceptance −0.89 0.42 −2.16⁎ −0.07 Autonomy 1.11 0.84 1.32 0.32 Restriction 0.42 1.02 0.41 0.43

Note: These outcomes are from general linear mixed models, with site entered as a random ef ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

3.2.2. Risk behaviors The links between EMT subtypes and risk behaviors are presented in

Tables 5 and 6. In terms of control variables, there were some protective effects of African American race (on Risky Sex, Cigarette Smoking, and Problem Drinking). There were also some effects of gender: males were significantly less likely to report Risky Sex and more likely to report Arrests, Cigarette Smoking, and Marijuana Use. Neglect was a significant unique predictor of Risky Sex and Problem Drinking.

Psychological safety and security uniquely predicted the following risk behaviors: Suicidal Thoughts, Arrests, Cigarette Smoking and Illegal Drugs. Acceptance and self-esteem predicted Problem Drinking. Autonomy, even with low base rates, predicted Risky Sex. None of the types of emo- tional maltreatment predicted Marijuana Use, and there were no signifi- cant effects of restriction.

4. Discussion

Few studies have utilized longitudinal data to evaluate the long- lasting consequences of EMT (Wright, 2007). This study examined char- acteristics of emotional maltreatment allegations for at-risk and maltreated children across childhood. Based on a typology originally proposed by Schneider et al. (2005), twenty-seven different subtypes of emotional maltreatment were classified into four domains: psycho- logical safety and security, self-acceptance and self-esteem, autonomy, and restriction. In addition to describing the characteristics of EMT in the LONGSCAN sample, this study examined how different types of mal- treatment, including subtypes of EMT, predict both trauma symptoms and risk behaviors.

This study confirms and extends earlier research on the relation be- tween EMT and trauma symptoms in children (Schneider et al., 2005), and it also adds new findings related to subtypes of maltreatment and risk behaviors. The size of the sample, as well as the extent of EMT expe- rienced by study participants, provided an opportunity for a more nuanced examination of demographic differences, as well as the ability to disentangle the general concept of EMT into specific subtypes (Layne et al., 2010). Examining relationships between child maltreat- ment in all forms – including EMT and its effects on trauma symptoms and risk behaviors – is particularly important in a child welfare context. Many of the most troubled youth who engage in risky behaviors also struggle with effects of trauma (Saxe, Ellis, & Kaplow, 2007); treatment of the child trauma may lead to a reduction in risky behavior.

arousal Depression

SE t B SE t

0.68 8.28⁎⁎⁎ 6.42 0.77 8.33⁎⁎⁎

0.53 0.60 0.21 0.63 0.34 0.51 −2.29⁎ −0.60 0.60 −1.01 0.75 −1.48 −0.78 0.88 −0.89

0.32 −2.88⁎⁎ −2.16 0.37 −5.78⁎⁎⁎

0.41 −1.05 −0.19 0.48 −0.40 0.45 2.65⁎⁎ 0.80 0.53 1.50 0.37 0.70 0.69 0.43 1.60

0.16 2.70⁎⁎ 0.54 0.19 2.83⁎⁎

0.30 −0.23 −0.28 0.35 −0.80 0.60 0.54 0.12 0.71 0.18 0.73 0.59 1.23 0.86 1.43

fect. Significant predictors are in boldface.

Table 4 Models predicting trauma symptoms, part 2.

Fixed effect Intrusive experiences Defensive avoidance Dissociation

B SE B B SE t B SE t

Intercept 6.19 0.82 7.58⁎⁎⁎ 7.19 0.95 7.53⁎⁎⁎ 5.44 0.73 7.47⁎⁎⁎

Race White 0.50 0.65 0.76 0.52 0.76 0.69 0.05 0.60 0.08 African American −0.33 0.63 −0.53 0.49 0.73 0.67 −1.10 0.57 −1.95 Hispanic −0.68 0.92 −0.74 −0.19 1.06 −0.18 −1.63 0.82 −1.99⁎

Gender Male −1.02 0.39 −2.62⁎⁎ −1.44 0.45 −3.18⁎⁎ −0.24 0.36 −0.68

Other maltreatment Physical abuse −0.02 0.50 −0.04 −0.08 0.58 −0.13 −0.13 0.45 −0.28 Sexual abuse 1.01 0.55 1.82 0.85 0.64 1.33 0.61 0.50 1.21 Neglect 0.91 0.45 2.01⁎ 0.68 0.53 1.29 0.56 0.41 1.36

Schneider EMT subtypes Psychological safety 0.39 0.20 1.96 0.41 0.23 1.76 0.32 0.18 1.73 Acceptance −0.29 0.37 −0.80 −0.21 0.42 −0.50 −0.11 0.33 −0.34 Autonomy 0.50 0.74 0.68 1.01 0.86 1.17 0.10 0.67 0.14 Restriction 0.93 0.90 1.03 1.97 1.04 1.89 0.73 0.82 0.90

Note: These outcomes are from general linear mixed models, with site entered as a random effect. Significant predictors are in boldface. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

59D. English et al. / Children and Youth Services Review 50 (2015) 53–63

4.1. Demographics, trauma, and risk behaviors

Significant gender effects by subtype of EMT were found, which likely has important implications for both assessment and interventions (Flory et al., 2009). Females in this study experienced more trauma symptoms; males reported themselves as less likely to engage in risky sexual behav- ior than girls, but more likely to engage in risk behaviors that resulted in arrest, and also more likely to use cigarettes and marijuana. These gender differences in outcomes may be related to differences in exposures to different subtypes of EMT and other maltreatment combinations. Gender differences in resulting behaviors may suggest the need for different interventions.

Clearly, EMT can have a profound effect on children's sense of safety. Given that trauma experts almost without exception emphasize that de- veloping a sense of safety is a necessary first step in trauma treatment,

Table 5 Models predicting risk behaviors, part 1.

Fixed effect Suicidal thoughts Risky sex

B SE B B

Intercept 0.07 0.03 2.47⁎ 0.26

Race White 0.04 0.03 1.61 −0.01 African American −0.02 0.02 −0.67 −0.09 Hispanic −0.01 0.04 −0.23 −0.07

Gender Male −0.02 0.02 −1.57 −0.06

Other maltreatment Physical abuse −0.01 0.02 −0.51 −0.02 Sexual abuse 0.03 0.02 1.33 −0.02 Neglect 0.00 0.02 0.23 0.08

Schneider EMT subtypes Psychological safety 0.02 0.01 1.97⁎ 0.01 Acceptance −0.01 0.01 −0.48 −0.02 Autonomy 0.00 0.03 0.07 0.11 Restriction −0.03 0.04 −0.94 0.04

Note: These outcomes are from general linear mixed models, with site entered as a random ef ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

these findings suggest that the psychological safety and security subtype of EMT can undermine the capacity for children's recovery from early trauma and other adversities. EMT appears to have a toxic effect on the nurturing environment in which children live and develop.

4.2. Impact of different domains of emotional maltreatment

This study confirms findings from earlier studies indicating that the majority of EMT allegations received by public child welfare systems (in this case five different CPS systems across the U.S.) are caregiver behav- iors that threaten a child's sense of psychological safety and security (PSS) (Baker & Festinger, 2011; Trickett et al., 2009). While the largest group of allegations clustered into the PSS category, one-third of the allegations were caregiver behaviors that impacted a child's sense of acceptance and self-esteem. This study supports the hypothesis of scholars who

Ever arrested

SE t B SE t

0.05 5.15⁎⁎⁎ 0.33 0.09 3.57⁎⁎

0.05 −0.15 −0.06 0.07 −0.75 0.04 −2.11⁎ −0.07 0.07 −0.99 0.06 −1.08 0.08 0.11 0.44

0.03 2.22⁎ 0.31 0.04 6.89⁎⁎⁎

0.03 −0.68 0.08 0.06 1.39 0.04 −0.58 0.00 0.06 −0.04 0.03 2.56⁎ 0.07 0.05 1.37

0.01 0.39 0.08 0.02 3.29⁎⁎

0.03 −0.61 −0.07 0.04 −1.74 0.05 2.27⁎ 0.12 0.08 1.49 0.06 0.60 −0.05 0.11 −0.43

fect. Significant predictors are in boldface.

Table 6 Models predicting risk behaviors, part 2.

Fixed effect Cigarette smoking Alcohol (drunk) Marijuana use Other illegal drugs

B SE B B SE t B SE t B SE t

Intercept 1.95 0.31 6.25⁎⁎⁎ 0.77 0.16 4.93⁎⁎⁎ 0.51 0.21 2.40⁎ 0.07 0.03 2.33⁎

Race White 0.34 0.23 1.50 0.13 0.14 0.96 0.09 0.19 0.51 0.05 0.03 1.81 African American −0.80 0.22 −3.66⁎⁎⁎ −0.25 0.13 −1.96⁎ −0.05 0.17 −0.30 −0.05 0.02 −1.89 Hispanic −0.46 0.32 −1.42 0.11 0.19 0.58 0.18 0.26 0.70 −0.02 0.04 −0.61

Gender Male 0.52 0.13 3.84⁎⁎⁎ 0.16 0.08 1.91 0.42 0.11 3.74⁎⁎⁎ 0.01 0.02 0.69

Other maltreatment Physical abuse 0.25 0.17 1.46 0.14 0.11 1.35 −0.05 0.14 0.36 0.01 0.02 0.30 Sexual abuse 0.13 0.19 0.69 0.18 0.12 1.56 −0.19 0.16 1.21 0.00 0.02 0.03 Neglect 0.35 0.16 2.20⁎ 0.02 0.09 0.17 0.13 0.13 1.07 0.01 0.02 0.60

Schneider EMT subtypes Psychological safety 0.14 0.07 2.10⁎ −0.04 0.04 −0.88 0.04 0.06 0.80 0.02 0.01 2.55⁎ Acceptance −0.16 0.13 −1.24 0.20 0.08 2.59⁎ 0.00 0.10 0.00 −0.02 0.01 −1.28 Autonomy 0.33 0.24 1.35 −0.07 0.15 −0.50 0.18 0.20 0.92 0.03 0.03 0.92 Restriction −0.38 0.32 −1.18 −0.06 0.20 −0.32 −0.37 0.26 −1.41 −0.05 0.04 −1.38

Note: These outcomes are from general linear mixed models, with site entered as a random effect. Significant predictors are in boldface. ⁎ p b .05.

⁎⁎⁎ p b .001.

60 D. English et al. / Children and Youth Services Review 50 (2015) 53–63

have suggested that different subtypes of EMT can have differential effects (Shaffer et al., 2009). Caregiver behaviors thought to create traumatic stress for their children predict emotional states characterized by anger, anxious arousal, hyper-vigilance, and depression. In addition, children who experienced PSS had more suicidal thoughts and an in- creased risk of acting out in ways that are potentially harmful to them- selves or others. These are significant physiological and emotional consequences that require services to address the child's need for a sense of safety and security.

Potential interventions include: parenting training regarding the harm caused by these specific types of verbal attacks on children/ youth; early screening for possible EMT; services to address caregiver emotion dysregulation associated with aggressive and unpredictable behaviors; and services to address caregiver issues providing their children with the attachment and nurturance needed to meet children's needs for acceptance. However, a recent review of evidence-based parenting programs suggests that available programs do not necessarily or specifically address emotionally abusive or neglectful parenting behaviors (Baker, 2009). Baker's review suggests a need to develop spe- cific components in evidence-based parenting programs that address parenting behaviors that threaten a child's sense of safety and security, a suggestion also made by Glaser (2011) in her discussion of individual- ized treatment approaches based on an understanding of the underlying dynamics of parent behaviors toward the child.

Furthermore, unless a child or youth is acting out (e.g., delinquent or aggressive toward other children and youth), child welfare service interventions are usually focused on the adult caregiver's behavior. If sending parents to parenting classes (which are not necessarily evidenced-based) is the primary intervention, then the current approach in child welfare is not addressing the key dynamic that has resulted in past harm to the child, and without intervention, will likely result in future negative impacts on children's emotional, behavioral, and physical health. In addition, the mean age at which the children in this sample experienced EMT was between 5 and 10, and the emotional harm was still apparent at age 18. These data suggest that EMT results in long- term consequences that should be understood and addressed by service systems when these children are seen (e.g., pediatric clinics, child welfare, juvenile justice, and mental health).

At a minimum, these findings suggest a need for greater under- standing of parental behaviors, and the motivations behind them, that result in emotionally harmful outcomes for children, as well as a better

understanding of appropriate interventions for children who experience various types of emotional maltreatment. These parental behaviors are not simply poor parenting skills of otherwise adequate parents. The acts described in this study were egregious and harmful on multiple levels. National data suggests many children in the community and those served by public child welfare systems experience this type of maltreatment. Whether services for these families are provided in home or the child has been removed and will be reunified, efforts to promote child well- being should include addressing harmful parental behaviors, including creating nurturing parenting behaviors necessary for positive child development.

In addition to interventions to re-direct parenting behaviors, con- sideration should be given to addressing trauma symptoms and risk behaviors to improve child well-being. Even if children are tempo- rarily or permanently removed from their biological families and placed with other families, it is incumbent on public child welfare systems to recognize the impacts of EMT and to provide services to alternative caregivers to assist them in both understanding child/youth behaviors and attitudes, and also employing effective techniques to manage child/youth behaviors associated with these experiences. This approach implies a need for comprehensive as- sessment of EMT, including subtypes related to emotional abuse and emotional neglect (Baker & Festinger, 2011; Shaffer et al., 2009) and for trauma-informed caregiving practices that help children with affect regulation.

To meet this challenge, child welfare systems may be required to create innovative strategies for serving these families and/or sequence services based on the type, severity, and chronicity of EMT following a model such as that proposed by Glaser (Glaser, 2002, 2011). For example, the association of psychological safety and security with both trauma and risk behaviors suggests a potential need for simultaneous approaches that assist children with trauma symptoms to develop trusting relation- ships with caregivers and learn to control emotional states even when reminded of past traumatic events. The need for empathetic and skilled responses to children who feel profoundly unsafe, in part due to EMT, is strongly indicated by this analysis of LONGSCAN data. Recent work by Layne et al. (2010) suggest the need to “unpack” concepts related to predictors, moderators, and mediators of trauma outcomes to further our understanding of the precursors and sequelae of different traumatic experiences. In this study, we found differential prediction of outcomes based on subtypes of EMT experienced by children. Different subtypes

61D. English et al. / Children and Youth Services Review 50 (2015) 53–63

of EMT predicted trauma outcomes and risk behavior outcomes, and sometimes both. Caregiver behaviors that created fearful and distressed emotional states in their children associated with feelings of psycho- logical safety and security predicted clinical level symptoms of anger/ irritability, anxious arousal, and depression, as well as suicidal ideation, arrests, cigarette smoking, and use of marijuana. The developmental impacts of EMT that impact children's sense of psychological safety may be exceptionally severe for children and youth with histories of trauma, including other types of child maltreatment.

EMT is often preceded by or accompanied by other types of maltreat- ment that leave children feeling powerless to protect themselves and un- usually vulnerable to dysregulated states related to deep-seated fears that often appear irrational to observers (Gray, 2007; Saxe et al., 2007). In addition, EMT that undermines a child's or youth's sense of safety is likely to impede recovery from trauma. EMT that attacks children's psychological sense of safety is likely to both exacerbate the effects of traumatic events and reactions and render therapeutic interventions in- effective (Perry & Szalavitz, 2007; Saxe et al., 2007). According to Saxe et al., “the dysregulation of emotional states is a defining feature of child traumatic stress” (2007). Children and youth who experience emo- tional dysregulation are more likely to engage in impulsive and/or risky behaviors and to adopt ways of calming down that increase risk of future harm (e.g., cigarette smoking). A compromised sense of psychological safety may lead to risky behaviors that help children and youth cope with trauma reminders in the short run but endanger health and mental health over the life span.

Although the main type of EMT found in this study is psychological safety and security, a substantial number of children experienced emo- tional maltreatment associated with undermining a child's sense of acceptance and self-esteem. EMT may have a distressing impact on children's internal working models (i.e., beliefs and attitudes about themselves in relation to others). In this study, we also found indepen- dent effects of neglect as a predictor of risky sex, especially for females, and cigarette smoking. Over and above the effect of neglect, acceptance and self-esteem also predicted risk behaviors. Risky sex and problem drinking could co-occur and function as self-soothing behaviors. Attacking a child's sense of acceptance and self-esteem by verbal abuse, withdrawing/ignoring, or becoming emotionally unavailable to a child predicts problem drinking. What is it about a child or youth's low sense of self-worth and feeling unloved and un-lovable that propels him or her toward risky sex and problem drinking? What are the appro- priate interventions to divert these youth from a risk trajectory to a well-being trajectory? At a minimum, these data suggest a need for a comprehensive assessment of the underlying factors associated with frequent engagement in risk behaviors, including EMT, to identify effec- tive interventions and disrupt potential trajectories associated with trauma and utilization of risk behaviors to manage stress and painful memories.

Finally, in addition to re-visiting the content of current parenting programs as suggested by Baker (2009), other possible interventions could address this form of maltreatment. For example, home visitation programs are relationship-based early prevention programs designed for parents and infants with the goal of promoting secure attachments (Egeland, Weinfield, Bosquet, & Cheng, 2000). As a rule, evidenced- based parenting programs promote children's sense of safety, security, and self-worth through positive parenting techniques that increase the frequency of parental praise and other rewards while decreasing use of punishment (Barth & Liggett-Creel, 2014; Timmer et al., 2006).

4.3. Limitations and future directions

This study used a particular approach to classifying EMT and should be interpreted within that context. Other ways to classify EMT have been proposed by other EMT researchers (e.g., using the APSAC classifi- cation scheme), and these alternative approaches are likely to have validity. However, these findings and earlier findings by Schneider

et al. (2005) suggest that the constructs for subtypes used in this study (psychological safety and security, acceptance and self-esteem, autonomy, and restriction) are useful in predicting trauma and risk behavior outcomes. In this research, as in other studies, the majority of EMT allegations were related to what might be called emotional abuse rather than emotional neglect (Shaffer et al., 2009; Trickett et al., 2009). Given that research suggests potentially differential effects related to emotional abuse compared to emotional neglect, additional work should be done on these two different conceptualizations of EMT. Several researchers have suggested that there is co-occurrence and overlap between emotional abuse and emotional neglect that should be considered in future analyses (see, for example, Gardner, 2008). One approach might be to examine neglect allegations – those related to failure to provide and lack of supervision – to determine whether reclassifying some of these behaviors as EMT might improve predictions of trauma and risk behaviors, or other aspects of children social, emotional, behavioral, and physical well-being. In conclusion, this was only one of several potential approaches to classifying EMT, and the LONGSCAN data, now available publicly through the National Data Archive on Child Abuse and Neglect (NDACAN), can be used to explore and compare these different approaches to such classification. Indeed, Trickett et al. (2009), using a subset of the LONGSCAN data, val- idated a very different approach to examining emotional maltreatment.

Additionally, even within the framework used for classifying EMT, further elaborations are possible. The current analyses examined the number of allegations for each subtype across the entire timeframe of the LONGSCAN dataset. Future research could examine age of exposure (or age of onset) to each form of EMT. Further, a trajectory-based anal- ysis of exposure to EMT would yield interesting results. Such endeavors would be complicated by the fact that the bulk of the EMT allegations in this dataset occurred when children were quite young. In other words, the data are somewhat skewed regarding age of exposure. A focus on age of onset would likely make this skew more pronounced.

As noted earlier, very wide variations exist across states in assessing and reporting EMT. Although the current study used trained coders of narrative descriptions of allegations, they depended on reporters noting issues that could be coded as emotional maltreatment, and there may have been state or temporal variations in documentation of allegations. In addition, because this study relied on administrative data, it did not include experiences of children who did not come to the attention of re- porters. Additional work using child and caregiver self-report data would shed additional light on patterns of EMT (Everson et al., 2008).

Finally, these analyses focused on trauma symptoms and risk behav- iors as outcomes. The degree to which EMT, particularly the more neglect- ful behaviors, constitutes trauma is debatable, and it is possible that other outcomes would be more sensitive to the effects of EMT. A host of other outcomes have been linked to EMT, including academic outcomes and self-esteem (Kim & Cicchetti, 2006; Tyler, Johnson, & Brownridge, 2008). Future analyses should explore the impact of EMT on a broader range of youth and child outcomes.

5. Conclusions

Although often overlooked and inconsistently reported, EMT is common, identifiable, harmful, and potentially preventable. A better understanding of it will help inform the provision of effective child welfare and mental health services to children and their families. In the current study, the most common subtype of EMT was violations of psychological safety and security. EMT, especially psychological safety and security, was related to elevations in trauma symptoms and to such risky behaviors as suicidal thoughts, unprotected sex, criminal justice involvement, and substance use. There is a need for greater un- derstanding of parental behaviors, and the motivations behind them, that result in emotionally harmful outcomes for children, as well as a better understanding of appropriate interventions for children who

62 D. English et al. / Children and Youth Services Review 50 (2015) 53–63

experience various types of EMT. EMT is often overlooked, but can have important implications for child and youth outcomes.

References

Baily, T.F., & Baily, W.H. (1986). Operational definitions of child emotional maltreatment. Au- gusta, ME: Maine Department of Human Services.

Baker, A.J.L. (2009). Adult recall of childhood psychological abuse: Definitional strategies and challenges. Children and Youth Services Review, 31(7), 703–714.

Baker, A.J.L., & Festinger, T. (2011). Emotional abuse and emotional neglect subscales of the CTQ: Associations with each other, other measures of psychological maltreatment, and demographic variables. Children and Youth Services Review, 33, 2297–2302.

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 de- velopmental psychology: Child abuse, child development, and social policy (pp. 7–74). Norwood, NJ: Ablex.

Barth, R.P., & Liggett-Creel, K. (2014). Common components of parenting programs for children birth to eight years of age involved with child welfare services. Children and Youth Services Review, 40, 6–23.

Bornstein, M.H. (2006). Parenting science and practice. In K.A. Renninger, I.E. Sigel, W. Damon, & R.M. Lerner (Eds.), Handbook of child psychology, Vol. 4. (pp. 893–949). Hoboken, NJ: Wiley.

Brassard, M.R., & Donovan, K.M. (2006). Defining psychological maltreatment. In M. Feerick, J.F. Knutson, P.K. Trickett, & S. Flanzer (Eds.), Child abuse and neglect. Balti- more, MD: Brookes Publishing.

Brassard, M.R., & Hardy, D.B. (1997). Psychological maltreatment. In M.E. Helfer, R.S. Kempe, & R.D. Krugman (Eds.), Battered child (pp. 392–412) (5th ed.). Chicago: University of Chicago.

Brassard, M.R., Hart, S.N., & Hardy, D. (1993). The psychological maltreatment rating scales. Child Abuse & Neglect, 17, 715–729.

Brener, N.D., Kann, L., Kinchen, S.A., Grunbaum, J.A., Whalen, L., Eaton, D., et al. (2004). Methodology of the Youth Risk Behavior Surveillance System. Morbidity and Mortality Weekly Report, 53(RR12), 1–13.

Briere, J. (Ed.). (1995). Trauma symptom inventory, professional manual. Lutz, FL: Psychological Assessment Resources, Inc.

Burnett, B. (1993). The psychological abuse of latency age children: a survey. Child Abuse & Neglect, 17, 441–454.

Cicchetti, D., & Valentino, K. (2006). An ecological transactional perspective on child mal- treatment: Failure of the average expectable environment and its influence upon child development. (2nd ed.). In D. Cicchetti, & D.J. Cohen (Eds.), Developmental psychology, Vol. 3. (pp. 129–201). New York: Wiley.

Claussen, A.H., & Crittenden, P.M. (1991). Physical and psychological maltreatment: Relations among types of maltreatment. Child Abuse & Neglect, 15(1), 5–18.

Donovan, K.L., & Brassard, M.R. (2011). Trajectories of maternal verbal aggression across the middle school years: Associations with negative view of self and social problems. Child Abuse & Neglect, 35, 814–830.

Egeland, B. (1997). Mediators of the effects of child maltreatment on developmental ad- aptation in adolescence. In D. Cicchetti, & S. Toth (Eds.), The effects of trauma on the developmental process (pp. 403–434). Rochester, NY: University of Rochester Press.

Egeland, B. (2009). Taking stock: Childhood emotional maltreatment and developmental psychopathology. Child Abuse & Neglect, 33, 22–26.

Egeland, B., Sroufe, L., & Erickson, M. (1983). The developmental consequence of different patterns of maltreatment. Child Abuse and Neglect, 7, 459–469.

Egeland, B., Weinfield, N.S., Bosquet, M., & Cheng, V. (2000). Remembering, repeating, and working through: Lessons from attachment-based intervention. In J. Osofsky, & H. Fitzgerald (Eds.), Infant mental health in groups at high risk (pp. 37–89). New York: Wiley.

English, D.J., Bangdiwala, S.I., & Runyan, D.K. (2005). The dimensions of maltreatment: In- troduction. Child Abuse and Neglect, 29, 441–460.

Erickson, M.F., & Egeland, B. (1996). Child neglect. In J. Briere, L. Berliner, J.A. Bulkley, C. Jenny, & T. Reid (Eds.), The APSAC Handbook on child maltreatment (pp. 4–20). Newbury Park, CA: Sage Press.

Everson, M.D., Smith, J.B., Hussey, J.M., English, D., Litrownik, A.J., Dubowitz, H., et al. (2008). Concordance between adolescent reports of childhood abuse and Child Protective Ser- vice determinations in an at-risk sample of young adults. Child Maltreatment, 13(1), 14–26.

Fagundes, C.P., Glaser, R., Johnson, S.L., Andridge, R.R., Yang, E.V., Di Gregorio, M.P., et al. (2012). Basal cell carcinoma: Stressful life events and the tumor environment. Archives of General Psychiatry, 69(6), 618–626.

Feerick, M. M., & Snow, K. L. (2006). Conclusions and future research directions. In M. M. Feerick, J. F. Knutson, P. Trickett, & S. M. Flanzer (Eds.), Child abuse and neglect: Defi- nitions, classifications, and a framework for research (pp. 385–394). Baltimore, MD: Paul H. Brookes Publishing.

Finkelhor, D., Turner, H., Ormrod, R., & Hamby, S.L. (2009). Violence, abuse, and crime exposure in a national sample of children and youth. Pediatrics, 124(5), 1–13.

Flory, J.D., Yehuda, R., Grossman, R., New, A.S., Mitropoulou, V., & Siever, L.J. (2009). Child- hood trauma and basal cortisol in people with personality disorders. Comprehensive Psychiatry, 50, 34–37.

Garbarino, J. (2011). Not all bad treatment is psychological maltreatment. Child Abuse & Neglect, 35(10), 797–801.

Garbarino, J., Guttmann, E., & Seeley, J.W. (1986). The psychologically battered child: Strategies for identification, assessment and intervention. San Francisco, CA: Jossey-Bass.

Gardner, R. (2008). Developing and effective response to neglect and emotional harm to children. London: Center for Research on the Child and Family, School of Social Work

and Psychosocial Sciences, University of East Anglia, National Society for the Preven- tion of Cruelty to Children.

Giovannoni, J.M., & Becerra, R.M. (1979). Defining child abuse. New York, NY: Free Press. Glaser, D. (2002). Emotional abuse and neglect (psychological maltreatment): A conceptual

framework. Child Abuse & Neglect, 26, 697–714. Glaser, D. (2011). How to deal with emotional abuse and neglect—Further development

of a conceptual framework (FRAMEA). Child Abuse & Neglect, 35, 866–875. Graham-Berman, S., & Levendovsky, A. (1998). Traumatic stress symptoms in children of

battered women. Journal of Interpersonal Violence, 13, 111–128. Gray, D.D. (2007). Nurturing adoptions: Creating resilience after neglect and trauma.

Indianapolis, IN: Perspectives Press. Hart, S.N., Binggeli, N.J., & Brassard, M.R. (1997). Evidence for the effects of psychological

maltreatment. Journal of Emotional Abuse, 1(1), 27–58. Hart, S.N., & Brassard, M.R. (1987). A major threat to children's mental health: Psycholog-

ical maltreatment. American Psychologist, 42, 160–165. Hart, S.N., Brassard, M.R., Binggeli, N.J., & Davidson, H.A. (2002). Psychological maltreat-

ment. In J.B. Meyers, L. Berliner, J. Briere, C.T. Hendrix, C. Jenny, & T.A. Reid (Eds.), The APSAC handbook on child maltreatment (pp. 79–104) (2nd ed.). Thousand Oaks, CA: Sage.

Hart, S.N., Brassard, M.R., & Karlson, H.C. (1996). Psychological maltreatment. In J. Briere, L. Berliner, J.A. Bulkley, C. Jenny, & T.A. Reid (Eds.), The APSAC handbook on child maltreatment (pp. 72–89). Thousand Oaks, CA: Sage.

Herrenkohl, R.C., Egolf, B.P., & Herrenkohl, E.C. (1997). Preschool antecedents of adoles- cent assaultive behavior: A longitudinal study. American Journal of Orthopsychiatry, 67, 422–432.

Iwaniec, D. (1995). The emotionally abused and neglected child: Identification, assessment, and intervention. New York: John Wiley & Sons.

Iwaniec, D. (2006). The emotionally abused and neglected child: Identification, assessment, and intervention (2nd ed.). West Sussex, England: John Wiley & Sons Ltd.

Kairys, S.W., Johnson, C.F., & Committee on Child Abuse and Neglect (2002). The psycho- logical maltreatment of children: Technical report. Pediatrics, 109(4), 1–3.

Kim, J., & Cicchetti, D. (2006). Longitudinal trajectories of self-system processes and depres- sive symptoms among maltreated and nonmaltreated children. Child Development, 77, 624–639.

Knight, E.D., Smith, J.B., Martin, L.M., & The LONGSCAN Investigators (2011). Measures for assessment of functioning and outcomes in longitudinal research on child abuse and ne- glect. Young Adulthood, Age 18, Vol. 4, .

Layne, C.M., Baker, A., Isakson, B., Durakovic-Belko, E., Dapo, E., Arlaanagic, B., et al. (2010). Unpacking trauma exposure risk factors and differential pathways of influence: Predicting postwar mental distress in Bosnian adolescents. Child Development, 81(4), 1053–1076.

McGee, R.A., & Wolfe, D.A. (1991). Psychological maltreatment: Toward an operational definition. Development and Psychopathology, 3(1), 3–18.

McGee, R.A., Wolfe, D.A., & Wilson, S.K. (1997). Multiple maltreatment experiences and adolescent behavior problems: Adolescents' perspectives. Development and Psychopathology, 9(1), 131–149.

Miller-Perrin, C.L., Perrin, R.D., & Kocur, J.L. (2009). Parental physical and psychological aggression: Psychological symptoms in young adults. Child Abuse & Neglect, 33, 1–11.

Moran, P.M., Bifulco, A., Ball, C., Jacobs, C., & Benaim, K. (2002). Exploring psychological abuse in childhood: I. Developing a new interview scale. Bulletin of the Menniger Clinic, 66(3), 213–240.

O'Hagen, K. (1995). Emotional and psychological abuse: Problems of definition. Child Abuse & Neglect, 19, 449–461.

Pears, K.C., Kim, H.K., & Fisher, P.A. (2008). Psychosocial and cognitive functioning of chil- dren with specific profiles of maltreatment. Child Abuse & Neglect, 32, 958–971.

Perry, B., & Szalavitz, M. (2007). The boy who was raised as a dog: And other stories from a child psychiatrist's notebook—What traumatized children can teach us about loss, love, and healing. New York: Basic Books.

Peugh, J.L., & Enders, C.K. (2005). Using the SPSS Mixed procedure to fit hierarchical linear and growth trajectory models. Educational and Psychological Measurement, 65, 811–835.

Rodgers, C.S., Lang, A.J., Laffaye, C., Satz, L.E., Dresselhaus, T.R., & Stein, M.B. (2004). The impact of individual forms of childhood maltreatment on health behavior. Child Abuse & Neglect, 28, 575–586.

Runyan, D.K., Curtis, P., Hunter, W.M., Black, M.M., Kotch, J.B., Bangdiwala, S., et al. (1998). LONGSCAN: A consortium for longitudinal studies of maltreatment and the life course of children. Aggression and Violent Behavior, 3(3), 275–285.

Saxe, G., Ellis, B.H., & Kaplow, J.B. (2007). Collaborative treatment of traumatized children and teens: The trauma systems therapy approach. New York: Guildford Press.

Schneider, M.W., Ross, A., Graham, J.C., & Zielinski, A. (2005). Do allegations of emotional maltreatment predict developmental outcomes beyond that of other forms of maltreatment? Child Abuse & Neglect, 29, 513–532.

Sedlak, A.J., & Broadhurst, D.D. (1996). Third National Incidence Study of Child Abuse and Neglect: Final report. Washington, DC: U.S. Department of Health and Human Services.

Sedlak, A.J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., et al. (2010). Fourth National Incidence Study of Child Abuse and Neglect (NIS-4). Report to Congress. Washington, DC: U. S. Department of Health and Human Services, Ad- ministration for Children and Families.

Shaffer, A., Yates, T.M., & Egeland, B.R. (2009). The relation of emotional maltreatment to early adolescent competence: Developmental processes in a prospective study. Child Abuse & Neglect, 33, 36–44.

Simmel, C., & Shpiegel, S. (2013). Describing the context and nature of emotional maltreatment reports in children. Children and Youth Services Review, 35, 626–633.

Smith Slep, A.M., Heyman, R.E., & Snarra, J.D. (2011). Child emotional aggression and abuse: Definitions and prevalence. Child Abuse & Neglect, 35, 783–796.

63D. English et al. / Children and Youth Services Review 50 (2015) 53–63

Straus, M.A., & Field, C.J. (2003). Psychological aggression by American parents: National data on prevalence, chronicity, and severity. Journal of Marriage & Family, 65, 795–808.

Sullivan, T.P., Fehon, D.C., Andres-Hyman, R.C., Lipschitz, D.S., & Grilo, C.M. (2006). Differ- ential relationships of childhood abuse and neglect subtypes to PTSD sympton clus- ters among adolescent inpatients. Journal of Traumatic Stress, 19(2), 229–269.

Taussig, H.N., & Culhane, S.E. (2010). Emotional maltreatment and psychosocial function- ing in preadolescent youth placed in out-of-home care. Journal of Aggression, Maltreatment & Trauma, 19(1), 52–74.

Timmer, S.G., Urquiza, A.J., Herschell, A.D., McGrath, J.M., Zebell, N.M., Porter, A.L., et al. (2006). Parent-Child Interaction Therapy: Application of an empirically supported treatment to maltreated children in foster care. Child Welfare, 85(6), 919–940.

Trickett, P.K., Kim, K., & Prindle, J. (2011). Variations in emotional abuse experiences among multiply maltreated young adolescents and relations with developmental outcomes. Child Abuse & Neglect, 35(10), 876–886.

Trickett, P.K., Mennen, F.E., Kim, K., & Sang, J. (2009). Emotional abuse in a sample of mul- tiply maltreated, urban young adolescents: Issues of definition and identification. Child Abuse & Neglect, 33, 27–35.

Tyler, K.A., Johnson, K.A., & Brownridge, D.A. (2008). A longitudinal study of the effects of child maltreatment on later outcomes among high-risk adolescents. Journal of Youth and Adolescence, 37, 506–521.

U. S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, & Children's Bureau (2013). Child Maltreatment 2012. Washington, D.C.: National Child Abuse and Neglect Data System.

Wekerle, C., Leung, E., Wall, A. -M., MacMillan, H., Boyle, M., Trocme, N., et al. (2009). The contribution of childhood emotional abuse to teen dating violence among child protective services-involved youth. Child Abuse & Neglect, 33(1), 45–58.

Wolfe, D.A., & McIsaac, C. (2011). Distinguishing between poor/dysfunctional parenting and child emotional maltreatment. Child Abuse & Neglect, 35, 802–813.

Wright, M.O. (Ed.). (2007). Childhood emotional abuse: Mediating and moderating processes affecting long-term impact. New York: Routledge.

Wright, M.O., Crawford, E., & Del Castillo, D. (2009). Childhood emotional maltreatment and later psychological distress among college students: The mediating role of maladaptive schemas. Child Abuse & Neglect, 33(1), 59–68.

Yates, T.M. (2007). The developmental consequences of child emotional abuse: A neurodevelopmental perspective. Journal of Emotional Abuse, 7, 9–34.