Introduction

Adolescent-to-parent violence is reported to occur in between 5 and 22% of the population (Cornell and Gelles 1982; Holt 2015; Peek et al. 1985; Routt and Anderson 2011); however, it is likely an underreported problem (Holt and Retford 2012). Although the full extent of this form of violence is unknown, it is clear that youth who exhibit violence toward caregivers represent an important population in need of effective prevention and intervention services. These violent behaviors confer risk for deleterious developmental, psychological, academic, and legal consequences for the adolescents (Kennedy et al. 2010). A collective, collaborative, and interdisciplinary effort by clinical scientists, mental health practitioners, and the legal community is greatly needed. Development of empirically-informed practices for screening, prevention, and intervention will maximize the likelihood that at-risk youth will be connected with evidenced-based services to positively alter negative trajectories and reduce risk of legal system involvement.

Two published review articles have evaluated the state of the literature concerning adolescent-to-parent violence (Hong et al. 2012; Kennair and Mellor 2007). In the first article, the authors summarized major findings, identified empirical gaps in the literature and concluded that, despite growing interest in the topic, conceptual models were underdeveloped, empirical findings regarding reliable risk factors were largely equivocal, and evidence for effective interventions were profoundly lacking. The authors called for movement toward greater public awareness of adolescent-to-parent violence as a social problem and rigorous research investigating efficacious treatment approaches (Kennair and Mellor 2007).

The more recent review article synthesized existing evidence on sociodemographic risk factors of adolescent-to-parent violence, and situated findings within an ecological framework to promote a more comprehensive understanding of contributing factors (Hong et al. 2012). The authors examined existing evidence regarding risk factors across levels of influence (i.e., micro-, meso-, exo-, macro-, and chrono-systems) and organized data indicating risk of adolescent abuse toward parents as a function of sociodemographic (i.e., age, gender, SES, ethnicity), and social-ecological (i.e., familial and peer relationships, adverse experiences, media and cultural influences) factors. The study found that, although some consistency exists regarding increased risk for adolescent-to-parent violence as a function of age, gender, and ethnicity, the phenomenon is more comprehensively understood by considering the interactive and complex set of risk and protective factors occurring at multiple levels of the youth’s environment (Hong et al. 2012). The authors posited that Bronfenbrenner’s ecological model is most conducive to understanding such complexities and urge the research community to look beyond the sociodemographic and micro-system levels of influence. Further, practice implications are detailed and specific suggestions are offered with regard to targeting multiple levels of influence in assessment and intervention procedures.

The current article offers an updated review and synthesis of existing literature on adolescent-to-parent violence. We review the status of theoretical frameworks that guide research and practice, discuss the impact of variability in the research methodology on the current state of knowledge, integrate new empirical data on risk and protective factors, and review developments in assessment procedures and intervention approaches. The two existing review articles discussed previously contributed to progress in this field by providing a survey of existing research and highlighting gaps in knowledge that can be addressed by improving our conceptual models and using theory to guide analysis of data and influence intervention development. We build on these reviews by providing an overview of the collective evidence inclusive of new findings and detailing limitations in existing data. We also extend efforts made by these authors in urging the research community to organize efforts toward a more comprehensive and integrated literature well equipped to inform policy and practice. Specifically, we advocate for a new research agenda rooted in collaborative, multidisciplinary efforts toward encouraging theory development and establishing best practices in assessment and intervention for families at risk for adolescent-to-parent violence.

This article presents a narrative analysis of existing data on adolescent-to-parent violence. At the present time, there does not exist an adequate number of studies to conduct a meta-analysis. Instead, we reviewed all empirical and scholarly work identified by database searches of Google Scholar and PsycInfo using the following key words in all combinations: parent, adolescent, violence, aggression, youth, teen, caregiver, and abuse. We subsequently collected and reviewed related articles found in the reference lists of originally identified articles. Throughout this manuscript, the terms adolescent-to-parent violence, adolescent aggression, and adolescent violence are used interchangeably to mirror the language used in the studies discussed.

Underlying Theoretical Frameworks

Several theoretical frameworks are offered to understand the underlying causal and maintenance factors associated with adolescent-to-parent violence. The majority of these frameworks have focused specifically on how parent and child factors interact and predict later violence. According to the contextual social-cognitive model, adolescents learn to rely on aggression as a general strategy when interacting with others through distorted cognitive processes, impaired problem-solving, and negative perception of social cues that develop in the context of suboptimal parenting (Lochman and Wells 2002; Lochman et al. 2000). Social learning theory (Bandura 1978) posits that youth aggression results from observing and then imitating aggressive and violent behavior in others (Brezina 1999). Patterson’s reciprocal-coercion theory (as cited in Brezina 1999), an extension of social learning theory, specifies that youth aggression is learned and reinforced through coercive and harsh parenting practices (Snyder and Patterson 1995); a pattern of repeatedly reinforced and reciprocal interactions that are characterized by coercion is created between parent and child.

Alternative frameworks have focused on the broader environmental influences on the child’s development, and continued use, of aggressive behaviors. The subculture of violence theory (Wolfgang and Ferracutti 1967, 2002) suggests that aggressive youth are influenced by a culturally-bound belief system that considers aggression an appropriate and adaptive response to perceived threat, whereas proponents of stress theory (Kratcoski 1985; Strasburg 1978) posit that aggression is a reserve strategy utilized only under circumstances of intolerable stress. The most integrative of these theories is ecological systems theory (Bronfenbrenner 1979). Proponents of this theory assert that human development is driven by a complex set of interactions that occurs between individuals as they change and adapt to their direct and indirect environments. Thus, adolescent-to-parent violence is the result of multiple, interactive risk (e.g., family violence, abuse, hostile parenting, negative peers, violence in media, gender-based socialization for violence) and protective factors (e.g., positive parenting, prosocial peers, positive role models) (Hong et al. 2012).

Scholars have long struggled to situate adolescent-to-parent violence within the larger family violence literature as particular aspects mirror existing conceptualizations of adult intimate partner violence, whereas other facets are distinctively based on developmental and systemic factors unique to adolescence. Existing arguments fall along a continuum of adolescent and parent dynamics that places normative conflict and defiance, rooted in developmental tasks of individuation and autonomy, at one extreme, and adolescent’s attempts to coerce, control, and intimidate parents at the other extreme (Coogan 2001; Cottrell 2001; Holt 2013; Paterson et al. 2002; Tew and Nixon 2010).

The “family conflict” conceptualization is broad and inclusive of any abusive behavior that occurs in response to disagreement among family members, “regardless of intent or impact” (Holt 2015, p. 2; Howarth and Feder 2013). This viewpoint centralizes systemic issues as the driving factor of violence in the family and, thus, considers the function of adolescent-to-parent violence as responsive to strain, stress, conflict, or distorted hierarchy in the family (Brezina 1999; Harbin and Madden 1979; Kratcoski 1985). Arguments supporting this perspective primarily focus on nationally representative, community-based, parent and child self-report survey data and draw on findings of increased rates of verbal versus physical aggressive acts and similar instigation rates across sex (Agnew and Hugeley 1989; Browne and Hamilton 1998; Ibabe and Bentler 2016; McCloskey and Lichter 2003; Nock and Kazdin 2002; Pagani et al. 2004, 2009; Straus et al. 1980). Proponents of this view caution against conceptualizations derived by extrapolating from existing theories of family violence and argue that the unique set of power dynamics calls for explanatory theories specifically geared toward the developmental circumstances of adolescence and parent–child dynamics (Paterson et al. 2002; Peek et al. 1985; Routt and Anderson 2011). This conceptualization identifies families as the target for intervention and aims to explore problematic interactional patterns, promote respectful relationships, and improve conflict resolution skills (Paterson et al. 2002; Routt and Anderson 2011).

Conversely, others define adolescent-to-parent violence more narrowly, as “a pattern of behavior that uses verbal, financial, physical or emotional means to practice power and exert control over a parent” (Holt 2013, p. 1). This notion rests on the parallel themes evident in adolescent-to-parent violence and intimate partner violence, including sex-based rates of perpetration (i.e., males are more often perpetrators) and use of similar victimization tactics such as threats, manipulation, and intimidation. Further, there are analogous behaviors seen in victim responses, including isolation and withdrawal, and equivalent impacts that include both immediate and long-term harm to victims and other family members (Cottrell and Monk 2004; Holt 2015). This perspective draws heavily from criminal justice data and qualitative studies that indicate lower overall prevalence rates and suggest that males are more likely to instigate adolescent-to-parent violence (Condry and Miles 2014; Holt 2015; Kethineni 2004; Routt and Anderson 2011; Walsh and Kreinert 2009). Proponents of this view urge researchers to use this framework to form research questions that will better inform how sex-based power dynamics maintain violence in families. However, they do not advocate for treating youth as criminals, recognizing that developmentally appropriate interventions are likely to be the most productive (Holt 2015).

These ostensibly competing ways of conceptualizing adolescent-to-parent violence are best understood as alternative viewpoints that fall along a continuum, and contribute to a comprehensive understanding of the complex phenomenon. Both carry unique advantages and drawbacks (Holt 2015). There is overwhelming evidence that adolescent-to-parent violence manifests differently across families and subcultures and that different kinds of empirical data offer unique and useful perspectives on the problem. Accordingly, each viewpoint is accurately explanatory in some instances. Any given case may be the result of coping deficits in the face of overwhelming stress, a chaotic family system, unbalanced family power dynamics, individual psychopathology, or any combination of these elements. Given the current state of the literature, the most optimal framework is one that resists attempts to pigeonhole adolescent-to-parent violence into existing paradigms and instead recognizes important areas of overlap with other forms of family violence, while distinguishing it as its own phenomenon with unique and variable manifestations.

There are many existing theories that are relevant and potentially informative, but continued theory development is essential to formulate specific, testable, theory-driven hypotheses that will guide future research toward a better understanding of factors integral to the development, assessment, and treatment of adolescent-to-parent violence. Theory development and advancement in empirical knowledge are best established through an iterative process of specifying and testing theoretically-derived hypotheses, and applying those findings to refine the overarching theoretical framework. No one theory comprehensively explains development, maintenance, or severity of aggressive and violent behaviors in adolescence. Instead, the literature has identified multiple factors that interact to differentially predict adolescent-to-parent violence trajectories. The challenge for clinical scientists is to develop strong theories to guide research and to integrate the available information to improve assessment instruments and interventions that simultaneously maximize the identification of at-risk youth while considering the great variability in needs and trajectories.

Definitions and Prevalence Rates

Currently, there is great variability in the way adolescent-to-parent violence is defined, measured, and investigated across disciplines, their theoretical scholars, and empirical scientists. As this body of literature moves forward, it will be important to establish consensus regarding a universal way to define and measure adolescent-to-parent violence as current variation undermines the ability of scholars and scientists to effectively organize collective knowledge and build on existing evidence. For example, over the past 30 years, several terms have been used in the literature (see Table 1 for examples). Some terms are broad in scope (e.g., “child-initiated family violence”), whereas others focus on either the age of the youth (e.g., adolescent-initiated parent abuse, child-to-parent violence) or the sex of the victim (e.g., child-to-mother or child-to-father). Some terms describe the behavior (e.g., “violence”) and others use labels that connote a consistent, maladaptive classification such as “abuse” and “syndrome”.

Table 1 Descriptors in selected published articles
Full size table

In addition to the convolutions surrounding definition, contemporary prevalence rates of adolescent-to-parent violence are complex, varying as a function of data collection methods and variables of interest. Current data collection methods include community-based victim and perpetrator self-report surveys, law enforcement and service-user records, and qualitative data from focus groups, structured interviews, and case studies (Routt and Anderson 2011). Although multiple methods and informants build a comprehensive and rich understanding of adolescent-to-parent violence, and contribute to our understanding of how prevalent these behaviors are in youth and families, variation in sampling and reporting procedures heavily impact the interpretability, comparability, and usefulness of currently available prevalence rates.

As mentioned previously, there is a large spread in reported prevalence rates (5–22%), making it difficult to pinpoint the scope of adolescent-to-parent violence (Cornell and Gelles 1982; Holt 2015; Peek et al. 1985; Routt and Anderson 2011). Studies conducted with specialized samples yield rates at the higher end of the range. Examples of these samples include: clinic-referred youth (12%; Nock and Kazdin 2002), youth from homes with domestic violence (12.6%; McCloskey and Lichter 2003), justice center referred youth (19%; Kratcoski 1985), undergraduate college students (14.5%; Browne and Hamilton 1998), and families recruited for a study on family conflict (22%; Margolin and Baucom 2014). Prevalence rates also depend on the period of time of reported behaviors (9.4% in the last 3 years vs. 5% in the last 1 year; Agnew and Huguley 1989), the reporter (7–8% youth report vs. 11–12% parent-report; Calvete et al. 2015a), type of aggression (7–12% physical vs. 91–99% psychological; Calvete et al. 2015a; 22% physical, 59% property damage, and 75% verbal; Margolin and Baucom 2014), and severity (2–5% severe vs. 5–13% overall; Browne and Hamilton 1998; Cornell and Gelles 1982; Ibabe and Bentler 2016).

To further complicate the overall picture, studies vary with regard to measurement of adolescent-to-parent violence. They range from one-item interview questions (e.g., “Have you ever hit your parent?”; Paulson et al. 1990; Peek et al. 1985), to multi-item questionnaires (e.g., 20-item Likert scaled survey assessing the various ways that children and their parents resolve conflict; Browne and Hamilton 1998). Behaviors included under an umbrella term of adolescent-to-parent violence also vary, as some researchers include only physically abusive acts (McCloskey and Lichter 2003; Nock and Kazdin 2002), others consider both physical and psychological aggression (Calvete et al. 2015b), and still others include physical violence, verbal aggression, and property damage (Margolin and Baucom 2014).

Statistics drawn from national databases highlight the magnitude of adolescent-to-parent violence through total reported incidents. Examples include statistics such as: 18,000 children under the age of 21 reported to law enforcement for adolescent-to-parent violence across 22 states during 1 year; 108,000 acts of adolescent-to-parent violence over a 10-year period (Walsh and Krienert 2009); and 56,000 instances of assault or intimidation against a parent or step-parent across 34 states during 1 year (Puzzanchera et al. 2015). Some researchers analyze police data from specific jurisdictions and report findings such as: 9% of all domestic violence offenders are juvenile and 51% victimized at least one parent (Synder and McCurley 2008); 63% of juvenile domestic violence charges were against a parent (Routt and Anderson 2011); and 5% of all domestic violence cases with juveniles listed as the primary aggressor were against a parent or caregiver.

Moreover, studies indicate variable and nuanced comparison rates of adolescent-to-parent violence in certain subgroups, particularly regarding sex differences. For example, physical aggression is reported to be much less common than verbal aggression overall (12 vs. 60%, respectively; Pagani et al. 2004, 2009). However, one study found that youth-report data reveal no sex differences in aggression type, yet parent-report data indicate that boys are more physically aggressive (15–18 vs. 7%; Calvete et al. 2015b). Similarly, an analysis of a clinical sample found no sex differences in therapist-reported physical aggression when considering the entire sample (11.4% boys, 14.6% girls, non-significant), but there were more boys than girls in the group of physically aggressive youth (Nock and Kazdin 2002). Generally speaking, survey-based studies tend to find equivalent rates of perpetration across sexes (Agnew and Hugeley, 1989; Browne and Hamilton 1998; Ibabe and Bentler 2016; McCloskey and Lichter 2003; Nock and Kazdin 2002; Straus et al. 1980), while criminal justice data show higher rates of perpetration among adolescent boys (63–83%, Condry and Miles 2014; Pagani et al. 2004; Routt and Anderson 2011; Walsh and Krienert 2009). Thus, the reporter (i.e., adolescent versus parent) and the data source (i.e., survey versus court system) may be important considerations when interpreting sex differences.

Current data regarding sex of the parent-victim are also complex. Many sources indicate disproportional victimization of mothers, with ratio estimates as high as 8:2 mother versus father victims and percentages as high as 72% mother versus father victims (Cottrell and Monk 2004; Holt 2015; Evans and Warren-Sohlberg 1988; Ibabe and Jaureguizar 2010; Routt and Anderson 2011). However, proportions of mother versus father victimization vary greatly across samples (47–82%; Evans and Warren-Sohlberg 1988; McCloskey and Lichter 2003; Nock and Kazdin 2002; Paulson et al. 1990; Walsh and Krienert 2009). Many studies report differential patterns in reports of severe adolescent-to-parent violence (1.7% toward mothers vs. 2.8% toward fathers; Browne and Hamilton 1998) while other studies report nuances within subgroups. For example, some studies find higher rates of father victimization among high school boys (2–6% toward mothers, 5–8% toward fathers; Peek et al. 1985) whereas other studies find that girls almost exclusively aggress toward mothers (Evans and Warren-Sohlberg 1988). Researchers suggest that these discrepancies may reflect differences in certain subgroups (mothers are more likely to be victims, except in samples comprised of older sons; Ibabe and Jaureguizar 2010; Pagani et al. 2004) or in overall reporting tendencies (mothers are more likely than fathers to make police reports; Condry and Miles 2014).

Finally, comparison rates also differ as a function of offense type and outcome. An analysis of the National Incident-Based Reporting System found sex comparison rate variation across aggravated assault (31% female), simple assault (38% female), intimidation (27% female), but no difference with respect to rates of injury, use of weapons, or influence of drugs and alcohol (Walsh and Krienert 2009). Estimates from samples with unique family demographics may not be representative of the larger population. For example, in a study drawing participants from one county, family composition was related to adolescent-to-parent violence; however, divorced families made up almost half of the sample and 43% of juvenile offenders lived in mother-headed households (Routt and Anderson 2011).

There is no perfect method for gaining a comprehensive view of adolescent-to-parent violence prevalence. Answers to different questions require data sources analyzed in multiple ways. However, the existing limitations of prevalence data will be minimized by establishing clear and consistent definitions, and validated and reliable scales that assess the frequency, severity, and context of a wide range of behaviors.

Risk Factors and Predictors

Many researchers have begun to investigate how adolescent-to-parent violence manifests in particular families by collecting data on adolescent characteristics, family dynamics, behavior patterns, and distal and proximal factors that provoke and maintain problematic behaviors. Regardless of the definition used, researchers and practitioners interested in implementing interventions to reduce adolescent-to-parent violence focus on discerning the factors that predict the development of such behaviors. Risk factors that predict maladjustment in adolescence and ongoing difficulties into adulthood present multiple points of intervention; however, certain factors differentially inform screening, prevention, intervention, and maintenance efforts (see Fig. 1).

Fig. 1
figure 1

Conceptual model of how various types of risk factors may inform prevention and intervention services

Full size image

Etiological and demographic factors place certain individuals at higher risk for negative outcomes relative to the general population. Thus, screening and early prevention efforts are likely most useful for families who have been identified as potentially at-risk from broad estimates of population prevalence. Contextual factors, which tend to be proximal and specific to family dynamics and individual characteristics, identify those who may benefit from early intervention to prevent persistence or exacerbation of initial problems. Finally, factors found to maintain aggressive or violent behaviors within the family system can become the targets of interventions to prevent further occurrence of the problematic behaviors. The following sections review the current state of the science regarding these various risk factors or predictors of maladjustment.

Etiological Predictors

A multitude of studies document a positive and strong relationship between childhood victimization or witnessing family domestic violence and adolescent aggression toward self and others (Brezina 1999; Browne and Hamilton 1998; Calvete et al. 2014; Duke et al. 2010; Ibabe et al. 2013; Maxwell and Maxwell 2003; Pagani et al. 2009; Ulman and Straus 2003). Scholars have proposed multiple pathways to explain this relationship. Modeling occurs when adolescents observe their parents’ behaviors and learn that aggression is an appropriate means of conflict resolution and/or an effective strategy to control another’s behavior. Adolescents can also become angry at the perpetrator (i.e., retribution-based violence) or try to protect the abused parent (i.e., protection-based violence) (Brezina 1999; Browne and Hamilton 1998; Cottrell and Monk 2004; Wells 1987). Documentation of high rates of co-occurring child abuse and intimate partner violence supports a mechanistic pathway through social learning, or modeling (Ibabe et al. 2013), but theories of retribution- or protection-based adolescent-to-parent violence have been more difficult to evaluate due to a dearth of longitudinal data. In one study using a national sample of high school males, researchers found an overall correlation between adolescent and parent aggression with diverging time course patterns (Brezina 1999). Findings supported their hypothesis that adolescent-to-parent violence may function as a deterrent of parent-to-youth aggression, as parent aggression at time 1 predicted more youth aggression at time 2, but youth aggression at time 1 predicted less parent aggression at time 2 (Brezina 1999).

Although a wealth of evidence suggests that bi-directionality in offspring and parent violence is common, reliance on cross-sectional and retrospective data makes it difficult to identify exactly how these behaviors develop and influence one another over time. Recent longitudinal reports begin to shed light on the development of adolescent-to-parent violence in the context of child abuse. One study explored relations between parent and adolescent aggressive behaviors in a nonclinical and non-court-referred sample by gathering mother, father, and adolescent reports of various behaviors over 8 years (i.e., at ages 10, 11, 12, 15, and 18) (Margolin and Baucom 2014). Measures of parental aggression included both historical parent-to-child and parent-to-parent aggression reported retrospectively for the previous 12 months at the first three waves of the study and concurrent parent-to-child and parent-to-parent aggression reported during the last two waves of the study. Results indicated a strong relationship between parental aggression, both historical and current, and adolescent-to-parent violence. Adolescents who reported low levels of exposure to historical parental aggression (aggregated across types) were much less likely to exhibit all types of adolescent-to-parent violence (3.2% physical, 29.0% property damage, 38.7% verbal) relative to those who reported medium or high total levels of exposure (22% physical, 74.2% property damage, 77.4% verbal) (Margolin and Baucom 2014). When the types of historical parental aggression (i.e., mother-to-child, father-to-child, mother-to-father, and father-to-mother) were assessed separately, mother-to-adolescent aggression predicted adolescent-to-parent physical aggression and property damage and father-to-mother aggression predicted adolescent-to-parent verbal aggression (Margolin and Baucom 2014). These effects held even after controlling for concurrent parental aggression. This finding refutes the hypothesis that adolescent-to-parent violence may be merely a reaction to recent or current conflict and demonstrates temporal precedence of historical parental aggression as a predictor of adolescent-to-parent violence, beyond what would be predicted by the presence of reciprocal aggression between parent and adolescent.

Moreover, the study also revealed that adolescent characteristics, including emotion dysregulation (e.g., “goes to pieces under stress”) and positive attitudes about aggression toward parents (e.g., “it is okay for a child to hit a parent if the parent hits the child first”), did not predict child-to-parent aggression (physical, verbal or property damage) at any time point. Interestingly, adolescents who were exposed to historical parent aggression were more likely to endorse positive attitudes about violence toward parents, but not to actually display these behaviors at higher rates (Margolin and Baucom 2014). Notably, parent-to-adolescent aggression was measured in this study using only two items (i.e., “slapped on the arm or leg” and “shook”), which significantly limits our understanding of the breadth of parents’ behaviors in childhood that might influence later aggression in adolescence (Margolin and Baucom 2014).

Demographic Predictors

As discussed previously, the existing evidence regarding demographic factors that reliably predict adolescent-to-parent violence is severely limited. This problem is due to wide variation in descriptors used for tracking and compiling data and variation in sampling methods, as well as the inherently complex interactions that exist among all factors that influence adolescent problem behaviors.

For estimating prevalence, discrepancies regarding sex, ethnicity, age, and diagnosis-related predictors illustrate the complexities of using demographic factors to reliably predict risk of adolescent-to-parent violence. For example, as discussed above, some studies have documented similar rates of adolescent-to-parent violence perpetration across sexes, whereas others have reported differential rates conditional on type of aggression (i.e., verbal or physical, more or less severe, involving injury or not), targets of aggression (i.e., caregiver-specific or others more broadly), and presence of internalizing and externalizing problems (i.e., oppositional behaviors, substance use, delinquency, low self-esteem, low empathy) (Ibabe and Jaureguizar 2010; Ibabe et al. 2013; Kennedy et al. 2010; Nock and Kazdin 2002; Walsh and Krienert 2009). Across studies, findings concerning differential prevalence rates of adolescent-to-parent violence related to race, ethnicity, and SES, are mixed and severely limited by sampling variation and disparity in data collection methods (Kennedy et al. 2010).

The effect of age on prevalence of and risk for persistence of adolescent-to-parent violence is similarly complex. It is well-documented that aggressive behaviors (e.g., hit, kick, bite) during early childhood are universal, and that typical development includes an increase in self-regulation concomitant with reduction in displays of aggression (Alink et al. 2006; Keenan et al. 1998; Tremblay 2000). When childhood aggression persists beyond the toddlerhood and preschool years, it can lead to disruptions in parent–child relationships, social development, and educational achievement, as well as legal consequences (Keenan and Wakschlag 2000; Tremblay et al. 2004). Thus, childhood aggression is best understood as a risk factor for later maladjustment as it is correlated with many adverse outcomes in adolescence, including conduct problems, substance use, poor educational attainment, and delinquency (Hinshaw et al. 1993; Nagin and Tremblay 1999; Olweus 1979). Exactly how powerful these risk factors are in predicting persistent and pervasive aggression in adolescence is not unequivocal. Researchers have identified that certain early indicators of problematic behavior, such as high oppositional behaviors and hyperactivity, increase the likelihood that these behaviors will become chronic issues and there is some evidence that aggression is fairly stable over time (Nagin and Tremblay 2001; Olweus 1979; Ulman and Straus 2003). However, factors associated with desistance, such as the severity of aggressive acts, should not be overlooked and are, in fact, critically important to informing intervention services (Loeber and Stouthamer-Loeber 1998).

Another risk factor of interest is how medical and psychological conditions impact risk for adolescent-to-parent violence. There is some preliminary evidence that medical and physical conditions in infancy (e.g., prenatal exposure to toxins, low resting heart rate) may impact future risk of exhibiting violent behaviors (Ortiz and Raine 2004; Williams and Ross 2007). Similarly, psychological characteristics commonly originating in childhood (e.g., oppositional behaviors, attention difficulties, impulsivity, peer rejection, antisocial or low prosocial behaviors) have been associated with future development of aggression (Lahey et al. 1999; Moffitt 1993; Nagin and Tremblay 1999; Valois et al. 2002). Initial attempts to disentangle these often co-occurring behaviors have found that academic motivation and engagement and general aggressiveness reliably differentiate children who engage in adolescent-to-parent violence from peers with other types of externalizing problems (Cottrell and Monk 2004; Nock and Kazdin 2002; Paulson et al. 1990; Pagani et al. 2004). One study documented that most youth involved in adolescent-to-parent violence-related police calls had prior involvement with the juvenile justice or mental health system (Evans and Warren-Sohlberg 1988). Overall, evidence suggests that youth who commit violence against their parents are more likely to have co-occurring medical and mental health problems, including histories of psychiatric medication use, hospitalization, and suicide attempts (Kennedy et al. 2010; Evans and Warren-Sohlberg 1988).

Contextual Predictors

Empirical evidence regarding the context in which adolescent-to-parent violence occurs has identified several common themes that more proximally predict the circumstances under which adolescents are more likely to be aggressive toward parents/caregivers. Themes discussed in the literature include parenting style, parent–child relationship, family structure, community influences, school, and peers.

Parenting Style

The relative emphasis that parents place on control, punishment, and responsiveness in regard to childrearing strategies is reliably detectable and predictive of various important child outcomes (Baumrind 1991). A disciplinary strategy labeled as “power-assertive,” characterized by corporal and psychological aggression, in combination with high levels of supervision and punishment was positively related to adolescent-to-parent violence (Ibabe and Bentler 2016). Further analysis of the same dataset found “power-assertive” parenting strategies to mediate the association between family relationships and adolescent behavior such that families with less positive family relationships, who also employed power-assertive parenting strategies, had higher levels of adolescent-to-parent violence (Ibabe and Bentler 2016). Different forms of problematic parenting styles, including overly permissive and overly controlling practices, are linked to adolescent-to-parent violence (Calvete et al. 2014; Paulson et al. 1990; Peek et al. 1985). Qualitative research suggests that the relation between parenting style and adolescent violence may also be partially explained by a mismatch between parenting strategies and developmental stage of the adolescent, or discordant parenting practices between co-parents (Cottrell and Monk 2004).

Parent–Child Relationships

Several studies document an inverse relationship between attachment or emotional engagement with parents and occurrence of adolescent-to-parent violence (Agnew and Huguley 1989; Calvete et al. 2014; Paulson et al. 1990; Peek et al. 1985). Positive family relationships characterized by warmth and positive communication were associated with less violent and more prosocial behaviors in Spanish adolescents (Ibabe and Bentler 2016). Further, results suggested a bidirectional relation between family relationships and adolescent violence. The cross-sectional study design and sole reliance on adolescent reports precludes firm conclusions regarding the development of adolescent violence over time, but the authors suggest that positive family relationships may lead to fewer incidences of adolescent violence, and adolescent violence may lead to deteriorating family relationships (Ibabe and Bentler 2016). Other researchers also document that juvenile offenders who commit adolescent-to-parent violence, relative to those who do not, more often report relational problems with parents and other family members (Kennedy et al. 2010).

Family Structure

Empirical findings regarding associations between family structure, socioeconomic status (SES), and risk for adolescent-to-parent violence are complex (Charles 1986; Peek et al. 1985). Juvenile offenders with charges of violence toward parents were more likely to come from a single-parent household than their counterparts who did not aggress against their parents. Further, single-parent households were overrepresented in police cases of adolescent-to-parent violence (Evans and Warren-Sohlberg 1988; Ibabe and Jaureguizar 2010; Kennedy et al. 2010). Interestingly, this pattern did not hold for a juvenile justice sample in Spain (Contreras and Cano 2014). Most cases of adolescent-to-parent violence existed in two-parent households. However, a higher proportion of adolescents from single-parent homes fell in the group who were violent toward their parents, relative to juvenile offenders who had other types of offenses (Contreras and Cano 2014).

The quality of the parent–child relationship, parenting practices, and individual parental factors such as criminality, psychopathology, and discipline techniques, may be particularly influential and preclude finding simple relations among these factors. For example, it may be that single-parent families confer higher risk, but only in the context of specific parenting practices (e.g., less supervision and monitoring) or limited financial resources. Adolescents who come from teenage single-parent families, compared to other types of single-parent families, are more likely to carry additional risk factors, such as low SES, low birth weight infants, and lower levels of education (Corcoran 1998).

Community

Exposure to poverty and disorganized communities characterized by low neighborhood attachment increases risk for aggression in adolescents (Valois et al. 2002). Neighborhood poverty has been linked to many externalizing problems, including violence, aggression, and delinquency, but data are mixed as to the underlying mechanisms of this relationship. Existing evidence suggests that the effect of poverty on externalizing problems in youth may be influenced by factors such as parents’ subjective perception of the neighborhood, gender, and peer influences (Ewart and Suchday 2002; Murray et al. 2011). For example, living in neighborhoods with high drug use, weapon use, and community or gang violence imparts risks for adolescents to engage in these same behaviors (Valois et al. 2002). Although low SES is a risk factor in adolescent aggressive behaviors, the influence of SES on adolescent-to-parent violence specifically is not yet well characterized (Kennair and Mellor 2007). In a study that conducted parent interviews, poverty emerged as a contributing factor to adolescent-to-parent violence: a lack of financial resources was perceived by parents to fuel anger and resentment in their children, feelings which were then expressed in the form of parent-directed aggression (Cottrell and Monk 2004). In the same study, service providers echoed resentment-based, parent-directed anger but also indicated that overwhelming poverty-related stressors (e.g., childcare, social support, limited finances) often undermine a family’s capacity to cope effectively with conflict (Cottrell and Monk 2004).

School

Children spend a majority of time in school; therefore, factors related to experiences at school highly influence child development. Low academic achievement is related to poor self-concept, increased rates of drop-out, negative attitudes about school, and low commitment to school (Finn and Rock 1997). Suspensions and expulsions increase risk for violence and aggression, as this form of punishment makes situations more difficult for students in several ways, including struggles to keep up with the school curriculum, and more time spent without adult supervision (Valois et al. 2002). For some youth, these experiences lead to chronic truancy and dropping out of school.

Peers

Developing adolescents are highly influenced by peers, and violent behaviors are no exception. Adolescents who have a strong allegiance to a delinquent peer group are at the highest risk (Elliott and Menard 1996). Gang membership is especially linked to participating in aggressive acts; exposure to gang activity in early adolescence is particularly risky. In a sample of juvenile offenders, adolescent-to-parent violence was strongly associated with gang membership, and associating with gang members and peers who owned guns (Kennedy et al. 2010). Researchers who have attempted to elucidate peer influence with regard to aggressive behaviors posit that this relationship may be driven by positive attitudes concerning the acceptability and utility of using violence in relationships (Valois et al. 2002). Some research suggests that adolescents behave aggressively toward their parents due to displaced anger about victimization by peers (Agnew and Huguley 1989; Cottrell and Monk 2004). Other literature suggests that peers who perpetrate abusive behaviors serve as role models for when and how to use aggression for personal gain (Cottrell and Monk 2004). Lastly, participation in deviant activities with peers (e.g., substance use) may increase conflict in the home if parents try to establish stricter limits. In turn, more frequent conflict between adolescents and parents, in the absence of effective conflict resolution skills, may increase the likelihood that the adolescent exhibits aggressive behaviors (Cottrell and Monk 2004).

Maintenance Factors

A clear understanding of the factors that maintain adolescent aggression toward parents is likely to advance the effectiveness of intervention and relapse prevention efforts. Treatments can then be developed and tested by targeting these maintenance factors. Once established, this knowledge will help service providers to know the targets of treatment most likely to produce clinically meaningful change in these families. Much of our current knowledge of these factors is preliminary and based largely on qualitative research that has elicited views from parents of youth who have been violent in the home and from professionals who work with these families.

Substance Use

The influence of substance use on adolescent-to-parent violence is well-documented, yet the directionality of the relation is not fully understood. Recent longitudinal data, however, indicate a temporal relationship where substance use predates and predicts adolescent-to-parent violence, although this pattern was only observed in boys (Calvete et al. 2015b). Conversely, adolescent violence toward parents did not prospectively predict substance use over an 18-month period (Calvete et al. 2015a). Evidence suggests that substance use may represent a common topic of disagreement between parents and youth and such conflict escalates into physical violence in the absence of alternative and effective conflict resolution strategies (Cottrell and Monk 2004; Evans and Warren-Sohlberg 1988). One study found that substance use in 15- and 16-year olds over the previous 6 months increased the risk of verbal aggression by over 50% and doubled the risk of physical aggression toward fathers (Pagani et al. 2009). In this same sample, substance use also predicted mother-directed adolescent violence, and an interesting interaction emerged. Specifically, adolescent substance use increased the risk of verbal aggression in youth whereas parent substance use predicted youth-initiated physical aggression (Pagani et al. 2004).

Inadequate Support Systems

A perceived inability to access helpful and responsive community supports is a recurrent theme across studies with parents who experience violence by their adolescents. One qualitative study in the UK found that the majority of adoptive parents reported feeling particularly unsupported by community service providers: instead of support and assistance, parents felt confronted with judgment and blame (Selwyn and Meakings 2015). In cases of re-involvement with child protection services, parents often reported feeling excluded from the decision-making process and undermined by professionals who reinforced the unbalanced power dynamics between parent and child by defending the behavior of, or “siding with,” the child (Selwyn and Meakings 2015). Parents in another study provided several examples of variable responses from law enforcement and justice system personnel wherein such professionals were either unresponsive or very slow to respond, leaving the parents without interim solutions (Cottrell and Monk 2004). Further, many parents reported a lack of consistent guidance, such that the police officers who responded to their crisis calls offered very different solutions than did child protection workers who followed up on the cases (Cottrell and Monk 2004; Tew and Nixon 2010).

Stigma

Qualitative research with biological and foster parents demonstrates that parents often fear being stigmatized when disclosing family secrets (Cottrell and Monk 2004; Selwyn and Meakings 2015; Tew and Nixon 2010). In a study of adoptive families, adolescent violence toward parents was the leading cause for adoption disruptions or failed adoptions. Adoptive parents described feeling ashamed, hopeless, and inadequate in parenting adolescents who were aggressive toward them (Selwyn and Meakings 2015). Similarly, researchers in Canada found that biological parents commonly blame themselves for the violence and, thus, minimize or deny its existence. They may also avoid letting others know of these behaviors due to fear of retaliation from their children, or the wish to circumvent the potential legal consequences for the teen, including ongoing involvement with the justice or child welfare systems (Cottrell and Monk 2004).

Power Imbalances

Dysfunctional family power dynamics are a defining feature in some families with adolescent-to-parent violence. In these families, youth use “verbal, financial, emotional or physical means to exert power and control over a parent” (Holt 2013, p. 1). Holt et al. proposed that adolescent-to-parent violence is a distinct form of domestic violence that shares the same sex-based nature with higher rates of boys committing violence toward mother victims than toward father victims. Specifically, the presence of power and control dynamics combined with the use of multiple forms of violence is analogous to adult fathers’ violence toward mothers (Holt 2013). Some researchers argue that unhealthy power dynamics are more common in single-mother headed households, particularly between mothers and teenage sons (Holt 2013; Tew and Nixon 2010). This view is supported by data from police records, in which rates of violence are highest among adolescent boys toward their mothers (Condry and Miles 2014).

Conflict Resolution Skills

What separates normative parent-adolescent interactions from those that include violence may be a difference in the ability to effectively manage typical, developmentally appropriate disagreements. To date, there are no published studies in which conflict resolution skills have been measured. Nevertheless, existing findings do suggest that the precursors to adolescent-to-parent violence include disagreements over adolescent freedom, rules, friends, or other common conflict areas (Calvete et al. 2015a; Evans and Warren-Sohlberg 1988). For example, a systematic review of police records revealed that the most common precipitating factors to disagreements between youth and caregivers were related to chores, transportation, sibling conflict, threats to run away, allowances, and substance use (Evans and Warren-Sohlberg 1988). In a sample of clinic-referred youth, violence toward a parent was found to be more common in youth who exhibited lower frustration tolerance and less capacity to adapt to stressors (Nock and Kazdin 2002). Less is known about these same qualities in parents; however, it is important that these skills are assessed for all members of families as they are likely to influence one another. Although disagreements between parents and youth about these common topics appear to be ubiquitous to parenting adolescents, a better understanding of the role of conflict resolution skills, or lack thereof, may be useful in guiding effective targets for treatment.

Risk factors associated with adolescent-to-parent violence are wide-ranging and differentially informative depending on how well such information can direct services for particular families. A great deal of attention has been paid to demographic risk factors, such as age, sex, and mental health status, but results are mixed and conclusions are nuanced and complex. Etiological risk factors, such as exposure to childhood maltreatment, are an important aspect of social history that may inform treatment for a particular adolescent, but are not useful in prospectively predicting which adolescents will engage in violence against a parent, as many adolescents who are maltreated as children do not become violent. There exists a dearth of research investigating the contextual and maintenance factors underlying adolescent-to-parent violence, and these are the components most likely to inform useful assessment and effective treatment services. A better understanding of these dynamics will advance our ability to identify which families, under which circumstances, are most likely to be affected by adolescent-to-parent violence, as well as isolate potent treatment targets for families who consequently enter the juvenile justice or mental health system.

Assessment and Intervention Practices

The recognition of adolescent-to-parent violence as a pressing social problem has sparked interest and action by mental health and legal professionals to address the needs of families. As a result, a proliferation of new screening tools, assessment procedures, mental health interventions, and specialized responses from the legal system has emerged over the past two decades. The following section reviews the development of specific assessment procedures and intervention approaches, discusses the current state of empirical support for existing programs, and calls for efforts toward successful implementation of evidence-based practices in existing community systems.

Screening and Assessment

A move toward building effective interventions for families with adolescent-to-parent violence rests on practitioners’ ability to accurately assess these behaviors in families. Many scholars have insisted that adolescent-to-parent violence is a distinct form of family violence (Wilson 1996) and patterns of interaction for all subsystems should be explored, including parental, sibling, and parent–child interactions (Wells 1987; Wilson 1996). The structure and functioning of the family system should be considered in the context of individual risk factors (e.g., criminal/delinquency behaviors, peer group involvement, mental health issues, and substance use), as well as the developmental stage of the family and individual needs of all family members (Wilson 1996). While many clinicians and scholars advocate for screening of family violence to be standard practice, some have suggested that these behaviors only be explored in the presence of certain warning signs, such as a family that expresses emotions solely through anger (Wilson 1996).

In the broader research literature, many instruments have emerged to measure the frequency, severity, and general proclivity toward violence in adolescents, but few address adolescent-to-parent violence specifically (Cornell and Gelles 1982; Tarter et al. 2002; Wilson 1996). One group of researchers adapted existing behavior-based measures to inquire about family violence between all possible dyads (i.e., parent–parent, child-to-parent, parent-to-child, sibling–sibling) within the family system, by making wording modifications to the same set of items (Ulman and Straus 2003). More recently, new questionnaires have been developed in which community professionals can quickly assess the incidence, experiences, and triggers of adolescent-to-parent violence (Edenborough et al. 2011).

The National Center for Juvenile Justice recognizes the need to reform the way adolescent-to-parent violence is assessed in the juvenile court system and has called for important advances toward developing optimal methods of responding to families (Siegel and Halemba 2015). One county developed an internal screening tool to delineate offenders “types” and improve system responses in adolescent domestic battery cases (Siegel and Halemba 2015). This effort inspired the National Youth Screen and Assessment Project, a multi-site research initiative to empirically validate these typologies (Siegel and Halemba 2015). The result was the Adolescent Domestic Battery Typologies Tool (ADBTT), which is now available as a framework to help inform court decision-making for youth who enter the juvenile justice system on charges of adolescent-to-parent violence (Nussbaum et al. 2015).

The principles that underlie the ADBTT include placing adolescent-to-parent violence in a developmental context that recognizes the core differences between adolescent- and adult-perpetuated family violence and the heterogeneity in background issues, mental health status, and treatment needs (Nussbaum et al. 2015). The ADBTT offers professionals a valid and reliable tool for understanding the unique set of characteristics for individual youth. Dispositional recommendations can then be tailored to reduce risk and achieve better outcomes (Nussbaum et al. 2015). Research findings from the ADBTT project support a framework that encompasses four distinct types of youth who exhibit violence toward parents/caregivers: (1) “defensive” youth who direct violence toward a parent in response to victimization; (2) “isolated incident” youth who utilize violence as conflict resolution strategies in the face of atypical stressful circumstances; (3) “family chaos” youth who exhibit violence in the context of inconsistent parental authority; and (4) “escalating” youth whose pattern of violence has effectively shifted familial power dynamics to position the youth being in control over the parent/caregiver (Nussbaum et al. 2015).

The ADBTT represents an ongoing, collective effort toward systematic methods of identifying, classifying, and coordinating legal responses to youth who are arrested for adolescent-to-parent violence across juvenile justice settings. Empirical evidence supports the predictive validity of the typology in identifying who is at the greatest risk for re-offending. Comparative studies find evidence that the ADBTT improves general risk assessment measures for this population, as it specifies risks of persistent violence toward a parent or caregiver (Nussbaum et al. 2015).

Response and Intervention

Despite recognition of adolescent-to-parent violence as a critical problem facing families, mental health professions, the court system, and the community, professionals have long struggled with determining best practices for these cases. Although empirically-supported interventions exist to treat frequently co-occurring behaviors, such as general aggression, delinquency, and substance use (Bartle-Haring et al. 2015; Kolko et al. 2009; Larson and Lochman 2002; Timmons-Mitchell et al. 2006; Weinblatt and Omer 2008), specific interventions for youth who aggress against a parent/caregiver are scarce and many lack solid empirical support. The following discussion details existing interventions for families with adolescent-to-parent violence by law enforcement, the juvenile justice system, and mental health professionals.

Law Enforcement

Police responses to adolescent-to-parent violence are inconsistent across jurisdictions and challenges to streamlining responses are rooted in complex issues. Parents are reluctant to request law enforcement assistance and are often ambivalent about the best course of action for their families. Parents report that they find it difficult to make a call to police about their child and thus adolescent-to-parent violence is likely to be an underreported problem (Holt and Retford 2012). Also, parents do not consistently seek the same response. For example, some parents may want the child removed (Holt 2011), whereas others want the police to simply talk to their child (Holt and Retford 2012). Many parents report feeling that police minimize the situation, and thus leave them with a sense of hopelessness about the possibility of lasting change (Cottrell and Monk 2004; Miles and Condry 2016).

Police also struggle with how to most effectively respond to a report of adolescent-to-parent violence. Studies show that police lack guidance on how to proceed during these calls, and thus rely heavily on individual discretion (Holt and Retford 2012; Miles and Condry 2016). Even when police take an action such as making an arrest, many parents end up dropping the charges or recanting their statements to prevent further legal consequences for their child (Miles and Condry 2016). Overall, there are limited policy and practice guidelines for officers responding to calls of adolescent-to-parent violence (Cottrell and Monk 2004; Holt 2011; Holt and Retford 2012; Miles and Condry 2016).

Juvenile Justice

Similar to the police, other community or governmental organizations lack guidance on how to respond to adolescent-to-parent violence (Holt 2011; Holt and Retford 2012). This may result in inconsistent responses between agencies (Cottrell and Monk 2004; Holt 2011) and by different representatives within agencies. However, some jurisdictions have attempted to transform how adolescent-to-parent violence is handled in the legal system and have demonstrated positive outcomes (Siegel and Halemba 2015). Paragraphs in this section provide examples of diversion programs implemented to reduce recidivism and divert youth from detention.

Two counties in California led the movement toward juvenile justice system responses to youth with charges of violence toward a parent (Uekert et al. 2006). Specialized courts were created to offer assessment, specialized legal professionals, intensive supervision, dedicated dockets, victim services, and offender programs to families with adolescent-to-parent violence. A multi-site program evaluation project found that the youth in the two counties that used specialized court programs were less likely to be re-arrested, relative to a county without such a program, but only for first time offenders (Uekert et al. 2006). Recidivism rates for those with extensive arrest histories were not different in counties with and without specialized programming (Uekert et al. 2006). Investigators also found that that youth who did not violate probation, were engaged in electronic monitoring, and were younger upon program entry were more likely to successfully complete the diversion program (Uekert et al. 2006). Outcome information with regard to treatment moderators is imperative to advancing our understanding about how best to use specialized court programming. Preliminary data suggests that the benefits of these programs differ as a function of many factors, such as age and arrest history. This type of outcome information can be very useful in determining tailored responses by the juvenile justice system.

The Domestic Violence Alternative Center (DVAC) in Pima County, Arizona was created to provide an alternative to juvenile detention. Instead of immediately incarcerating an offending youth, DVAC provides a network of available services to eligible youth and families, particularly mental health services. This program has resulted in increased police, family, and community referrals to DVAC and decreased the number of juveniles physically brought into the detention facility (Siegel and Halemba 2015). A similar program, the Florida Domestic Violence Respite Care Services, provides all youth charged with domestic violence temporary placement in respite care or a shelter, rather than being detained or returning home, in order to create individualized behavioral treatment goals and to ensure safety. Family reunification is typically built into each youth’s treatment plan; however, if the youth is unable to make significant gains and cannot safely return to their families, they may be eligible to receive other services (unspecified) prior to reunification (Siegel and Halemba 2015). In Illinois, the Dupage County Department of Probation and Court Services allows for creating individualized diversion programs. If a youth is placed on probation, participation in a treatment program could be required. If a youth fails to participate in or complete the diversion program, detention becomes an option (Siegel and Halemba 2015). Although it appears that these programs may be successful in diverting youth from juvenile detention, it is not clear how they influence long-term outcomes and rates of re-entry or prolonged involvement with the juvenile justice systems. It is imperative that the implementation of such programs coincides with data collection to understand the impact of contextual factors, such as the environment to which a youth will return, as well as the effectiveness of referral processes and follow-up services.

Mental Health

Recent years have seen an influx in specialized interventions for adolescents and parents impacted by adolescent-to-parent violence. Although some interventions appear to be promising, there is a dearth of empirical research demonstrating the effectiveness of these interventions.

The “Step-Up” program was developed in King County, Washington, after adolescent-to-parent violence was noted as a problem by the King County Department of Judicial Administration (Routt and Anderson 2011). The curriculum utilizes a cognitive-behavioral and skills-based approach to reduce violence and conflict in the family and teaches participants how to better respond to adolescent aggression (“Step-Up Program” 2015). The King County Juvenile Court and the Dupage County Department of Probation and Court Services have begun to make referrals to this program as a diversion from detention (Siegel and Halemba 2015). Step-Up was evaluated by comparing outcomes between participants and matched controls who did not participate in the program (Organizational Research Services, 2005). The evaluation demonstrated promising results at 12-month follow-up, including a lower percentage of felony referrals (9.9% in Step-Up group versus 20.2% in comparison group) and lower average number of adolescent-to-parent violence referrals (0.06 in Step-Up group versus 0.28 in comparison group). The teens in the Step-Up group self-reported decreases in aggression and violence (“Step-Up Program” 2015). While these data provide important preliminary evidence of positive effects of Step-Up, it is important to note that there were important differences between completers and non-completers in the Step-Up group and the comparison group was heterogeneous with regard to participation in community services that may be similar in unknown ways to Step-Up. Randomized trials and long-term follow-up data on behavioral outcomes and recidivism rates are needed to draw more firm conclusions about program effects.

In Australia, a parent group intervention was developed to help mothers who have violent adolescents (Paterson et al. 2002). This parent group is both educational and therapeutic as it addresses topics such as safety, child development, anger, violence, and conflict resolution. Self-reports and qualitative data demonstrated a decrease in the number of physical and verbal violent events in the home. Further, mothers self-reported less anxiety, somatic symptoms, and fatigue. However, mothers still reported significant levels of depression even with completion of the group intervention (Paterson et al. 2002).

The Youth Offender Diversion Alternative (YODA) uses a solution-focused treatment approach that aims to reduce violent behaviors in youth who have been charged by the juvenile justice system with assault against a family member (Bolton et al. 2015). YODA, developed in a collaborative partnership among university researchers, is a voluntary diversion option which treats youth between the ages of 17–25, using a three-phase diversion model. The three phases include an initial assessment of self-sufficiency needs, followed by individual therapy and case management aimed at improving self-sufficiency behaviors and skills, and concluded with solution-focused family therapy sessions. The justice system partnered with providers to increase compliance and motivation by offering youth an opportunity to have assault charges dismissed after successful program completion and the charges expunged after a period of 1 year without additional incidents reported (Bolton et al. 2015). Preliminary program evaluation data suggest that YODA participants showed positive short-term outcomes on measures assessing mental health symptoms, anger intensity and regulation, and impulsive, violent behaviors (Jordan et al. 2013). Currently available data on YODA is solely descriptive, although investigators are preparing to launch adequately-powered and well-controlled comparison studies (Whitehill 2012).

The Family Violence Intervention Program (FVIP) is housed in the juvenile justice system to provide youth arrested due to violence toward a parent/caregiver with counseling, case management, and conflict management skill-building (Nowakowski and Mattern 2014). Youth enrolled in FVIP are obligated to comply with program requirements such as improved school attendance, no additional violence-based arrests, no substance use, random drug testing, and family therapy. Family members are also required to agree to specified involvement in the program (Nowakowski and Mattern 2014). A follow-up study found that youth with more extensive juvenile justice records and poorer school attendance were more likely to prematurely terminate from FVIP (Nowakowski and Mattern 2014).

Our own research group is developing a novel group intervention targeting adolescent-to-parent violence and its behavioral and mental health correlates. Guided by a new approach to behavior change (Acceptance and Commitment Therapy or ACT; Hayes et al. 1999), we developed an intervention that targets the factors shown to reduce family violence and associated mental health problems. The new intervention, which comprises eight, 3-hour sessions, has demonstrated effectiveness with adults (e.g., Zarling et al. 2015). Feasibility studies have been completed with 20 adolescent participants. Results indicated that facilitators can effectively understand and teach ACT techniques and concepts to adolescent group participants. Moreover, participants were able to understand key ACT concepts and skills, and many participants identified ways in which these concepts and skills might be applicable in their daily lives.

We are currently conducting a series of pilot studies in which we collect questionnaire data pre- and post-treatment with intervention participants and a matched, non-treatment control group. Specifically, we are collecting mental health symptom data along with data regarding ACT concepts and skills. Finally, we will analyze recidivism data for all pilot study participants 3 months post-treatment. The long-term goal is to establish this new intervention as the first empirically supported intervention for violence and comorbid mental health issues among adolescent boys and then disseminate it statewide to boys who are part of the juvenile court system.

Summary and Future Directions

An important body of evidence on adolescent-to-parent violence has accumulated over the last several decades. In this article, we reviewed several theories discussed in the existing literature and highlighted that few studies have tested theoretically-derived hypotheses. Adolescent-to-parent violence does not fit with the prevailing models of family violence and, thus, the lack of strong theoretical models impedes professionals’ ability to adequately address the problem (Selwyn and Meakings 2015).

A considerable number of studies have documented factors that predict adolescent-to-parent violence. However, we found that much of the existing literature relies on large-scale, aggregate statistics to extricate reliable sociodemographic risk factors, and a few qualitative studies attempt to uncover the nuanced defining features and individual differences in presentations of adolescent-to-parent violence. Inconsistencies in research methodologies, definitions, and measurement tools have impeded progress toward effectively identifying and connecting at-risk and in-need families with evidence-based treatments. Moreover, much of this work has been and continues to be conducted in academic silos of criminal justice, family violence, child psychology, and juvenile justice. Without integration of findings across disciplines, the existing research remains underutilized.

Over the past several years, preliminary steps have been taken to improve assessment and intervention procedures. The Adolescent Domestic Battery Typologies Tool (Nussbaum et al. 2015) is a leading example of a cross-discipline and collaborative effort toward developing a comprehensive and practical framework for addressing adolescent-to-parent violence that will be informative and useful across involved disciplines. In addition, new intervention programs such as “Step Up” in King, County, Washington and our novel, ACT-based program in Tucson, Arizona are being developed and empirically tested in community-researcher partnerships. The emergence of these treatment programs and partnerships to test them over the last several years demonstrates that this need is increasing across the law enforcement, juvenile justice, and mental health systems.

In addition, important advancements have been made in the form of court programs with specialized units to divert youth from the juvenile justice system. These programs connect youth with community services and skills-based, targeted treatment programs designed to encourage alternative behaviors in youth and effective responses from parents. While preliminary results appear promising, systematic and controlled investigations of these newly-implemented court and mental health programs are yet to be reported in the literature.

The empirical findings, related discussion, and conclusions reached in this article must be considered in the context of study limitations. This study is a narrative analysis based on a review of literature available in well-known academic electronic databases. The authors made considerable effort to gather all relevant articles, but it is possible that some were not found or inadvertently overlooked. The current state of the literature precludes conducting a meta-analysis at this time; however, in the future, when more empirically based studies exist, such a methodology would provide considerable benefits, including a more rigorous synthesis of existing data and the ability to draw firmer conclusions regarding collective evidence.

Adolescent-to-parent violence is a multi-faceted and systemic family problem and is likely caused by a complex set of interrelated factors. A future research agenda will include the development of flexible and comprehensive theoretical models that elucidate overall risk patterns and account for individual differences. With these theoretical models, research scientists can specify testable hypotheses and test them with adequately powered longitudinal study designs. Clinical scientists will then be well-equipped to explore dynamic and static risk factors, developmental trajectories, and key components that maintain and interrupt violent behavior patterns in families wherein youth aggress against a parent/caregiver.

With this increased knowledge, interventions can be developed and modified to target the dynamic and static risk factors and developmental trajectories identified. Researchers can also uncover which key factors are positively related to treatment engagement and completion, as subgroups of adolescent-to-parent violence offenders have been found to prematurely terminate from existing treatment programs (Nowakowski and Mattern 2014; Uekert et al. 2006). Also needed is the development of well-designed protocols to organize community responses.

The usefulness of this new data will be maximized if multi-method and multi-informant longitudinal research protocols clarify both ideological and nomothetic patterns within a collaborative interdisciplinary framework. In particular, efforts should be directed toward establishing valid, reliable, and useful assessment measures that can inform community agency responses and determine the best route for targeted, evidence-based treatment services for individual families.

Conclusions

Adolescent-to-parent violence is an emerging area of study with important cross-discipline implications for mental health professionals, as well as the juvenile justice and child welfare systems. The high prevalence rates indicate that many families face youth violence and aggression in their homes, and consequently become involved with community agencies and systems, including law enforcement, mental health, and juvenile justice. A comprehensive understanding of the key risk factors contributing to the development and maintenance of violent and aggressive behaviors toward parents/caregivers, as well as protective factors that reduce the risk of development of such violent behaviors, is a crucial contribution to the study of adolescence as these violent behaviors significantly impede healthy adolescent development across social, interpersonal, mental health, and academic domains of functioning.

We know a fair amount about the phenomenon of adolescent-to-parent violence; however, future progress depends upon a collective, cross-discipline, systematic shift in focus. This will include the development of theories that explain the nuances of these behaviors for families, in conjunction with tests of specific hypotheses based on these theories. While theories are developing and being tested, adolescents are acting violently toward their parents and caregivers; therefore, society cannot wait for the results of these theory-driven studies to act. We must continue to develop system-wide responses that flexibly apply the best existing evidence-based assessment and treatment approaches to reduce the detrimental impact that adolescent-to-parent violence has on entire families and revise these responses as data are published from theory-based research.