The Impact of Household Dysfunction on Dating Violence Perpetration Among Adolescents in the United States: A Scoping Review

Introduction

Adolescent dating violence (DV) is a serious public health concern in the United States (CDC, 2022; Peskin et al., 2019; Reidy et al., 2015; Vagi et al., 2013). Adolescent DV is a type of intimate partner violence (IPV) that includes (a) physical DV, (b) sexual DV, (c) psychological/emotional DV (i.e., threatening, verbal, relational, and digital/cyber), and (d) stalking (CDC, 2022; Shorey et al., 2019). The prevalence of DV victimization is likely underestimated among adolescents in the United States due to underlying factors such as social stigma, fear, and measurement-related limitations regarding how DV questions are assessed across national surveys (Basile et al., 2006; Peskin et al., 2019; Rothman & Xuan, 2014). However, the 2021 CDC Youth Risk Behavior Surveillance System survey reported that among high school adolescents who were in a relationship in the past year, nearly 9.7% reported sexual DV and 8.5% reported physical DV (CDC, 2023; Clayton, 2023). Further, DV perpetration rates are scant compared to victimization rates, and existing literature tends to concentrate on victimization rather than comprehensively exploring risk factors for perpetration, including adverse childhood experiences (ACEs) among adolescents (Boyce et al., 2023).

ACEs, such as childhood maltreatment, household dysfunction (HD), and community violence, have been associated with DV victimization and perpetration both in adolescence and adulthood; however, there is limited empirical research examining the relation between the HD domain and DV perpetration, especially among adolescents (Afifi et al., 2020; Duke et al., 2010; Haynie et al., 2013; Jouriles et al., 2012; Malhi et al., 2020; Navarro et al., 2022; Wolfe et al., 2005). HD is one of the domains of ACE that has been described as children growing up in an environment that threatens a child’s stability and welfare due to witnessing parental IPV (Afifi et al., 2020). Due to HD’s multifaceted and subjective nature, there is no set definition of the HD construct. However, according to the original ACE study, HD encompasses witnessing parental IPV, parental divorce/separation, living with household member(s) with substance use and mental health disorders, and family history of incarceration (Felitti et al., 2019). In addition, recent studies have expanded upon the definition and categories of ACE. They identified additional forms of HD, including family chaos, parental harsh discipline, family violence, death in the family, and living with household member(s) with disabilities (Afifi et al., 2020; Bussemakers et al., 2022; Davis et al., 2019; Felitti et al., 2019; Gottfredson et al., 2022; Morris et al., 2015). It is important to note that most of the existing ACE measures only include some of these additional forms of HD and usually assess them as individual constructs rather than assessing HD’s overall impact. Since romantic or dating relationships usually begin during middle school, which is a critical period of time when adolescents develop their cognitive abilities, interpersonal relationships, independence, social connectedness, and communication skills, adverse experiences such as HD can further hinder adolescents from initiating and maintaining their healthy dating relationships (Collins, 2003; Malhi et al., 2020; Simon et al., 2008).

Further research is warranted regarding the association between experiencing HD, specifically, and increased risk for DV perpetration among adolescents. For example, a study that explored the association between various forms of ACEs and adolescent DV perpetration indicated that approximately 28.9% of adolescents reported at least one ACE, primarily substance use disorder by a household member, which was significantly correlated with adolescents’ DV perpetration, bullying, physical fights, and other maladaptive behaviors (Duke et al., 2010). In addition, research has shown that adolescents living in a dysfunctional environment or suffering from any negative experiences are at heightened risk for extreme emotional distress resulting in relationship problems, breakups, a decline in cognitive skills, challenges with self-esteem and autonomy, and other psychological problems (Malhi et al., 2020; Navarro et al., 2022; Peskin et al., 2019; Wincentak et al., 2017).

Previous scoping reviews have supported the link between ACEs (including HD) and DV. For example, one scoping review that investigated DV perpetration by male adolescents and young adults aged 10 to 24 years within an international context found that individuals who experienced ACEs were more likely to perpetrate violence in relationships (Malhi et al., 2020). However, that scoping review was limited because it did not assess the effect of HD specifically on both male and female DV perpetration, particularly among adolescents in the United States (Malhi et al., 2020). Another scoping review examined the association between ACEs within the family context (i.e., witnessing IPV, parental neglect, etc.) and physical and online DV among adults. Yet, this review reported direct associations only between particular types of ACEs and DV victimization and perpetration (Navarro et al., 2022). Navarro et al. (2022) also found that emotional, sexual, and physical abuse, witnessing IPV, parental mental illness, HD, and poor relationships with parents were strongly associated with DV perpetration among adults. However, this review did not address the gap in evaluating the relation between HD and any type of DV perpetration (i.e., physical, sexual, psychological, online, etc.) simultaneously among adolescents in the United States.

Existing studies that identified ACEs as a predictor of DV perpetration have predominantly examined ACEs’ impact on IPV in adulthood rather than assessing the concurrent impact of HD on adolescents’ dating perpetration behaviors (Duke et al., 2010; Fonseka et al., 2015). Moreover, no comprehensive reviews have been undertaken to investigate the impact of HD on various types of DV perpetration, especially among adolescents in the United States. Thus, the primary goal of this scoping review is to summarize results from existing research studies on how HD impacts DV perpetration behaviors among adolescents ages 10 to 17 years in the United States, address research gaps and limitations, and discuss the next steps for future research.

Methods

This scoping review followed the five-step methodological framework by Arksey and O’Malley (2005). The five steps of the framework include (a) identification of the research question; (b) identifying relevant studies; (c) selection of studies; (d) charting the data; and (e) collating, summarizing, and reporting the results (Arksey & O’Malley, 2005). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist was also used as a reference for study sections to follow the guidelines for conducting and publishing scoping reviews.

Eligibility Criteria

We used the PECO (Population, Exposure, Comparator, and Outcome) framework to develop the study eligibility criteria. The eligibility criteria for this scoping review are as follows: (a) adolescents aged between 10 and 17 years of age; (b) quantitative observational studies (i.e., cross-sectional, cohort, case-control, longitudinal, and trials (which includes baseline data)); (c) studies investigating at least one type of HD as an exposure (i.e., witnessing IPV/domestic violence/household violence, family violence, family chaos, family conflict, harsh parenting, parental conflict, parental aggression, parental divorce, separation, absence, parental death/sibling death/household member death, household member substance use disorder, household member mental health illness, household member incarceration, jail, criminal history, and household member with physical illness or disability); (d) at least one type of adolescent DV perpetration has to be included as one of the outcomes in the studies (i.e., physical DV, sexual DV, psychological or emotional DV, threatening DV, digital or cyber DV, and stalking); (e) published in English in a peer-reviewed journal; (f) published between 2013 and August 2023; and (g) conducted in the United States. The detailed study eligibility criteria are presented in Table 1.

OPEN IN VIEWER

Table 1.

Scoping Review Study Eligibility Criteria.

PECO Framework	Inclusion	Exclusion
Population	Adolescents aged between 10 and 17 years.	The mean or median age of children/adolescents is not within the inclusion age range.
Exposure (IV)	Any form of household dysfunction (HD): • Witnessing intimate partner violence/domestic violence/household violence. • Family violence, family conflict, harsh parenting, parental conflict, parental aggression. • Parental divorce, separation, absence. • Parental death/sibling death/household member death. • Household member substance use disorder. • Household member mental health illness. • Household member incarceration, jail, criminal history. • Household member with physical illness or disability.	Only includes all other Adverse childhood experiences (ACEs): • Sexual or physical abuse/neglect. • Neighborhood or community violence. • Peer victimization. • Bullying • Natural disasters. • War-related trauma. • ACEs are reported as an overall composite score where HD is included but not directly measured in association with the outcome. • The HD category is measured as a moderator or a mediator and not measured as a single exposure or an independent variable regarding dating violence (DV) perpetration (outcome).
Comparator	N/A	N/A
Outcome (DV)	Any form of DV perpetration or aggression. • Physical DV • Sexual DV • Psychological or Emotional DV • Threatening DV • Digital or Cyber DV • Stalking	Only includes: • DV victimization • DV perpetration and victimization are combined into a composite score. • Bullying • Violence that is not related to dating or romantic relationships.
Study design	Quantitative studies • Cross-sectional • Cohort • Case-control • Longitudinal, and • Trials (which includes baseline data)	• Reviews • Qualitative studies • Research letters • Editorials • Commentaries • Reports • Dissertations, Conference • Proceedings/abstracts. • Trials that do not include findings from baseline data.
Limits	• Location: United States. • Published: 2013–2023 (August 16)—when the final search was conducted. • Language: English	• Other countries besides the United States • Published before 2013 • Languages other than English

Steps 4 and 5: Charting the Data and Collating, Summarizing, and Reporting the Results

The Garrard Matrix method was used to extract, chart, and synthesize the findings from the included studies (Garrard, 2020). The matrix consists of study number, study author(s)/publication year, study aim(s), study design, sample size, type(s) of HD exposure measured in the studies (i.e., witnessing domestic violence, divorce, household member substance use disorder, etc.) along with the instrument(s) used, type(s) of DV perpetration measured as an outcome(s) in the studies (i.e., physical, sexual, psychological, stalking, threatening, verbal/emotional, relational, and digital) and the instrument(s) used, and the key findings of the studies. This information was charted in an Excel spreadsheet SC developed and reviewed by all authors.

The lead author, SC, evaluated the rigor and quality of the study using the Critical Appraisal Skills Programme (CASP) checklist (Center for Evidence-Based Management, 2014). This checklist has been utilized by previous scoping reviews (Navarro et al., 2022; M. E. Reyes et al., 2021). The CASP checklist implemented in this scoping review encompasses the following criteria: (a) clarity of stated study aims and objectives; (b) appropriateness of study methods; (c) adequate description of the methodology; (d) was there any bias in sample selection; (e) representativeness of the sample for generalizability of study results; (f) utilization of statistical power analysis for sample size calculation; (g) response rate; (h) utilization of reliable and valid measures; (i) examination for statistical significance; and (j) inclusion of confidence intervals (CI) in study findings (Center for Evidence-Based Management, 2014). The response to each criterion is classified as “yes,” “no,” or “unsure,” and a quality score for each study was derived by summing the number of “yes” responses. However, for item 4, the reverse score “no” was counted within the study rigor indices that contributed to the overall quality score of each included study (Center for Evidence-Based Management, 2014; M. E. Reyes et al., 2021). The CASP checklist findings are presented in Table 2.

OPEN IN VIEWER

Table 2.

Critical Appraisal Skills Programme (CASP) Checklist for Assessing Study Rigor and Quality.

Study #	Study Aim(s)	Study Design	Selection of Subjects	Selection Bias a	Sample Generalizability	Stat Power	Response Rate	Valid Measures	Stat Sig	CI	Quality Score
1	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6
2	Yes	Yes	Yes	Unsure	Unsure	No	No	Yes	Yes	Yes	6
3	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6
4	Yes	Yes	Yes	Unsure	Unsure	No	Yes	Yes	Yes	Yes	7
5	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	No	5
6	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6
7	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6
8	Yes	Yes	Yes	Yes	No	No	No	Yes	Yes	No	5
9	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	No	5
10	Yes	Yes	Yes	Yes	No	No	No	Yes	Yes	Yes	6
11	Yes	Yes	Yes	Unsure	Yes	No	No	Yes	Yes	Yes	7
12	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6
13	Yes	Yes	Yes	Unsure	Unsure	No	No	Yes	Yes	No	5
14	Yes	Yes	Yes	Unsure	No	No	No	Yes	Yes	Yes	6

Note. CI = confidence intervals; quality score = total number of indicators reported by each study, which is marked as yes; Sig = significance; Stat = statistical.

a

Indicator scored in a reversed manner.

Results

The literature search yielded a total of 5,038 studies (Figure 1 PRISMA diagram). Consequently, 1,720 studies were excluded due to duplication, and 3,318 studies were extracted for title and abstract screening. A total of 3,258 studies were excluded due to not meeting this review’s eligibility criteria, and 60 studies were retained for full-text screening. During the full-text review phase, 46 studies were excluded due to ineligibility, and 14 eligible studies were retained for analysis for this scoping review.

OPEN IN VIEWER

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart.

Discussion

To our knowledge, this is the first scoping review to describe existing literature regarding the impact of HD on DV perpetration among adolescents aged 10 to 17 years in the United States. We reviewed 14 studies and found that most (n = 11) of the included studies that were conducted from 2013 to August 2023 found a significant association between HD and DV perpetration among adolescents in the United States (A. Mennicke et al., 2021; Foshee et al., 2015, 2016; Gottfredson et al., 2022; Reyes et al., 2015; Latzman et al., 2015; Moretti et al., 2014; Morris et al., 2015; Ontiveros et al., 2020; Temple, Shorey, Fite, et al., 2013; Temple, Shorey, Tortolero, et al., 2013). Findings from this review would help inform primary and secondary DV perpetration prevention interventions for adolescents and youth.

In this scoping review, the three most common types of HD that were assessed as an independent predictor of adolescent DV perpetration were family conflict, witnessing parental IPV or interparental violence, and harsh discipline. This exploration of the HD domain across these studies indicated that HD can be measured as an independent construct rather than only a sub-construct of ACEs. Often, HD is embedded under the broader paradigm of ACEs and is measured compositely, with individuals receiving an overall score based on their ACE exposures rather than domain-specific scores (McLennan et al., 2020; Mersky et al., 2017). However, it should be noted that these studies did not specifically define the items as a part of the HD domain but measured them as individual constructs using various instrument(s) that have several similarities with the HD domain that are usually assessed under ACEs. It is evident that HD is an integral component of ACEs; however, it can be challenging to define the HD construct due to the lack of a clear definition in the existing literature and the nuances of how youth internalize or externalize their traumatic experiences. Furthermore, as HD is a latent construct, it is challenging to encapsulate every aspect within a single definition. Thus, our HD definition is grounded in existing literature evidence. It should also be noted that although the studies only examined family conflict, witnessing IPV, and harsh discipline as individual constructs, most of these predictors used similarly phrased items as the corresponding HD items included across many ACEs instrument(s) (BRFSS, 2021; NSCH, 2021; PEARLS, 2019). Furthermore, one of the included studies examined overall ACEs and exposure to interparental violence as separate predictors of DV perpetration, which shows the complexity of these constructs and how more psychometric research is needed to set clear definitions and sub-categories of these latent constructs (Ontiveros et al., 2020).

The included studies did not explore other aspects of the HD domain (i.e., death in the family, household member(s) with mental and/or substance use disorder, divorce/separation, family criminal history, and household member(s) living with a disability). This may be because there is more precedence regarding the association between adolescents witnessing IPV or family conflict in their household and adapting those learned violent behaviors and applying them in their romantic relationships (A. Mennicke et al., 2021; Foshee et al., 2015, 2016; Gottfredson et al., 2022; Reyes et al., 2015; Latzman et al., 2015; Moretti et al., 2014; Morris et al., 2015; Ontiveros et al., 2020; Temple, Shorey, Fite, et al., 2013; Temple, Shorey, Tortolero, et al., 2013). Thus, it is imperative for future research to explore other exposures within the HD domain and examine their impact on DV and other health-related outcomes, particularly among adolescents, to have a broader understanding of the overall impact of different types of HD, as the literature remains scant.

Regarding DV perpetration, most of the included studies examined physical and psychological DV perpetration among adolescents (Foshee et al., 2015, 2016; Gottfredson et al., 2022; A. Mennicke et al., 2021; Moretti et al., 2014; Morris et al., 2015; Ontiveros et al., 2020; Reyes & Foshee, 2013; Reyes et al., 2015; Temple, Shorey, Fite, et al., 2013; Temple, Shorey, Tortolero, et al., 2013). This may be because adolescents tend to witness these types of violent behaviors in their households more than witnessing sexual DV, stalking, or digital DV behaviors among their household members. Furthermore, compared to the various instruments that the studies used to measure HD, the included studies primarily employed the Safe Dates Abuse measures and the CADRI to examine several forms of DV perpetration as the outcome(s). This demonstrates that there are already existing reliable and valid instruments to measure DV behaviors, especially for adolescents. However, one of the included studies discussed the significance of how DV is measured, whether as a continuous variable with a sum score or as a categorical variable (Ontiveros et al., 2020). This is of interest because Ontiveros et al. (2020) found that when DV perpetration was scored in a summative manner, female adolescents were found to perpetrate DV at a higher level to their partners compared to males; yet, when DV was measured categorically, the amplification of these scores reduced drastically, and females were no longer with the higher DV perpetration scores. Thus, it is important to consider a pragmatic method and justifications from the literature to select an appropriate measurement approach to minimize the chances of reporting misleading results.

There are other methodological concerns regarding some of the included studies, which limited the generalizability of their findings to all adolescents in the United States. For example, although most of the studies included a moderate to large sample size, their findings had limited generalizability to the particular regions (i.e., North Carolina, Kentucky, Rio Grande Valley, Houston, youth residing in rural counties, or high-risk metropolitan areas) and were not representative of all adolescents (Davis et al., 2019; Foshee et al., 2015; Gottfredson et al., 2022; Latzman et al., 2015; A. Mennicke et al., 2021; Ontiveros et al., 2020; Reyes et al., 2015; Temple, Shorey, Fite, et al., 2013; Temple, Shorey, Tortolero, et al., 2013). Another study utilized a convenience sampling approach to recruit adolescents exposed to parental IPV from low socioeconomic status (SES), which is not representative of adolescents who belong to higher SES also exposed to parental IPV (Foshee et al., 2016). These findings indicate that while the studies examined the impact of HD in DV perpetration in high-risk populations across various regions with low SES backgrounds, the generalizability of the findings is limited to adolescents with these contextual factors and may not extend to all adolescents in the United States.

The relationship between HD and DV perpetration is multifaceted, as many factors have been shown to contribute to this association among adolescents (Davis et al., 2019; A. Mennicke et al., 2021; A. M. Mennicke et al., 2022; Moretti et al., 2014; Temple, Shorey, Tortolero, et al., 2013). Around 29% (4/14) of the included studies assessed several moderators and mediators regarding the relationship between HD and DV perpetration, including social support, parental monitoring, school belonging, DV acceptance, rape myths, sensitivity to rejection, and acceptability of female and male violence; 80% of these studies reported significant associations between HD and DV perpetration by analyzing the aforementioned variables as moderators or mediators (Davis et al., 2019; A. Mennicke et al., 2021; A. M. Mennicke et al., 2022; Moretti et al., 2014; Temple, Shorey, Tortolero, et al., 2013) (Supplemental Material Appendix B: Garrard Matrix of Included Articles (N = 14)). Previous research has also indicated family/household chaos to be a significant contributor to risky behaviors among adolescents (Chatterjee et al., 2015; Delker et al., 2020). Therefore, future research should explore these protective and risk factors among this vulnerable population to inform early intervention efforts and trauma-informed care to prevent and reduce the burden of HD-related outcomes such as DV perpetration.

Conclusion

This scoping review accumulated and summarized the current literature evidence regarding the impact of HD on DV perpetration among adolescents in the United States. Most of the included studies found a significant association between HD and DV perpetration among adolescents; however, the review also identified gaps particularly related to measurement issues regarding HD and the need for additional investigations regarding the HD construct to understand its overall impact on DV and other health-related outcomes. Future research addressing these issues may help establish valid and reliable psychometric measures, promote early interventions, and develop preventative policies to help foster resilience among this vulnerable population locally and globally. The critical findings and the implications for practice, policy, and research are presented in Tables 3 and 4.

OPEN IN VIEWER

Table 3.

Critical Findings.

Critical Findings

• Seventy-nine percent of the included studies found a significant association between HD and various types of DV perpetration (i.e., physical, psychological, sexual, relational, verbal, and threatening). • The three most common types of HD that were assessed as an independent predictor of DV perpetration were family conflict, witnessing parental IPV, and harsh discipline. • Other aspects of HD (i.e., death in the family, household member(s) with mental and/or substance use disorder, divorce/separation, family criminal history, and household member(s) living with a disability) were not examined. • The included studies did not use an HD measure; instead, they utilized various instruments (with similarly phrased items) specific to the individual constructs of HD they assessed. • The majority of the studies used the CADRI or Safe Dates Abuse measure to assess DV perpetration. • The relationship between HD and DV perpetration is multifaceted, as there are underlying factors that may contribute to this association. For example, 29% of the studies evaluated the role of several moderators and mediators (i.e., protective factors, rape myths, DV acceptance, and more); 80% of those studies reported these factors to have a significant role regarding the relation between HD and DV perpetration among adolescents.

Note. CADRI = Conflict in Adolescent Dating Relationships Inventory; DV = dating violence; HD = household dysfunction; IPV = intimate partner violence.

OPEN IN VIEWER

Table 4.

Implications for Practice, Policy, and Research.

Implications for Practice, Policy, and Research

• There is limited empirical evidence regarding the impact of HD on DV perpetration behaviors among adolescents. Thus, future research is warranted regarding examining additional HD-related elements to better understand its cumulative effect on these behaviors. • There is a need for improved domain-specific age-appropriate HD measures for adolescents. Future research is needed to establish valid and reliable psychometric HD measures to understand its overall impact on DV behaviors and other health-related outcomes. • HD- and DV-related early interventions and the development of preventative policies should be undertaken to mitigate immediate and long-term consequences related to these issues and foster resilience among adolescents.

Note. DV = dating violence; HD = household dysfunction.

Supplemental Material