Intimate Partner Homicide of Adolescents

Key Points

Question  What are the characteristics of intimate partner homicides of adolescents?

Findings  In this multistate study of homicides of 2188 individuals aged 11 to 18 years, 150 were perpetrated by an intimate partner. Intimate partner homicide vitims were largely female and killed by a firearm, and homicides often involved broken relationships or jealousy.

Meaning  Intimate partner homicide of adolescents is an important problem that warrants further study and proactive intervention.

Importance  Intimate partner violence during adolescence is widespread, and consequences can be severe. Intimate partner homicide (IPH) is the most extreme form of intimate partner violence, but literature on IPH has almost exclusively focused on adults.

Objectives  To determine the proportion of adolescent homicides that is perpetrated by intimate partners and to describe the victim, perpetrator, and incident characteristics of these IPHs.

Design, Setting, and Participants  Analysis of quantitative and qualitative surveillance data from the National Violent Death Reporting System from 2003 to 2016. Data represent 32 states that contributed to the system for 1 year or longer. There were 8048 homicides of victims aged 11 to 24 years with a known relationship between the victim and perpetrator. For persons aged 11 to 18 years, there were 2188 homicides. Analysis began September 2018.

Main Outcomes and Measures  An incident was identified as an IPH if the relationship between the perpetrator and victim was coded as spouse, ex-spouse, girlfriend or boyfriend, ex-girlfriend or ex-boyfriend, or girlfriend or boyfriend (unspecified current or former). Variables of interest included demographic characteristics (age, sex, race/ethnicity) for the victim and perpetrator, relationship status at time of death, homicide-suicide, homicide method, firearm type, and location of homicide. Contextual categories were created from the qualitative narratives.

Results  Of adolescent homicides, 150 (6.9%) were classified as IPH. A total of 135 victims (90%) were female (mean [SD] age, 16.8 [1.3] years). Overall, 102 perpetrators (77.9%) were 18 years and older (mean [SD] age, 20.6 [5.0] years), and 94 (62.7%) were current intimate partners of the victim. Firearms, specifically handguns, were the most common mechanism of injury. Compared with IPHs of young adults aged 19 to 24 years, perpetrators of adolescent victims were younger and less likely to be a current intimate partner. The most common categories of adolescent IPH homicides were broken/desired relationship or jealousy and an altercation followed by reckless firearm behavior and pregnancy related.

Conclusions and Relevance  Adolescents, particularly girls, in dating relationships may face risk of homicide, especially in circumstances of a breakup or jealousy and when perpetrators have access to firearms. Understanding homicide in early dating relationships can inform prevention and intervention efforts tailored to adolescents.

Intimate partner violence (IPV) is a substantial public health problem with a lifetime economic burden of more than $3.6 trillion for the 43 million adults in the United States affected by IPV.¹ Adolescence is a particularly vulnerable time for experiencing IPV. Among respondents who experienced sexual violence, physical violence, and/or stalking by an intimate partner, the 2011 National Intimate Partner and Sexual Violence Survey estimated that 71.1% of female individuals and 58.2% of male individuals experienced IPV before age 25 years (23.2% of female individuals and 14.1% of male individuals before age 18 years).² Among high schoolers who dated, data from the 2017 national Youth Risk Behavior Survey show that 6.9% experienced sexual violence and 8.0% experienced physical violence by someone they were dating or going out with in the past year.³ According to the National Survey on Teen Relationships and Intimate Violence, more than 60% of adolescents in a current or past-year dating relationship experienced some form of IPV (physical, sexual, and/or psychological abuse).⁴

Intimate partner homicide (IPH) is the most extreme form of IPV. Global estimates show at least 1 in 7 homicides is perpetrated by an intimate partner.⁵ A 2017 study conducted in 18 states across the United States reported that 55% of all homicides of women from 2003 to 2014 were related to IPV.⁶ To our knowledge, prior literature on IPH has almost exclusively focused on adults. While homicide is the third leading cause of death for adolescents,⁷ there is a dearth of research on the scope and circumstances of IPH among adolescents, despite adolescents being at high risk of experiencing many types of IPV.^3,4 A prior study using 3 years of police interviews on all homicides of female individuals aged 11 to 18 years in North Carolina reported that approximately one-third of adolescent femicides were committed by an intimate partner.⁸ Adolescents may be at risk for IPH for a number of reasons. They may have a history of abuse in their families or other family dysfunction, so their parents may be less likely to intervene and they may not have well-developed support systems.^9,10 In addition, adolescence is a time of heightened emotionality and intensity when youth are first entering and exploring romantic relationships.¹¹ Adolescents experience wide emotional swings and their capacity for regulating emotions and using positive relationship skills is still developing, which may impact their behavior in conflicts (eg, when a romantic relationship ends).^12,13 Adolescents experiencing IPV also rarely seek help from adults and instead disclose to their peers, who are often not equipped to give advice or assist in safely extricating their friend from an abusive relationship.¹⁴ Given these unique circumstances, several lead IPV researchers have called for further research to understand the magnitude and risk factors of IPH among younger populations, including adolescents.^10,15

To fill this gap, we sought to understand the proportion of adolescent homicides that is perpetrated by intimate partners and to describe the victim, perpetrator, and incident characteristics of these IPHs. The present study used data from a large national surveillance system that contains quantitative and rich qualitative information, allowing for a mixed-methods analysis that provides more granular details on the circumstances of violent deaths. Understanding homicide in early dating relationships can potentially inform prevention and intervention efforts tailored to adolescents.

It was determined that this study did not need institutional review board approval because all individuals had died and data were deidentified. For these reasons, patient consent was also not obtained.

The National Violent Death Reporting System (NVDRS) is a state-based surveillance system that links data from vital records, coroner/medical examiners, and law enforcement agencies on violent deaths including homicide and suicides.¹⁶ Data for this study came from the NVDRS Restricted Access Database set representing the 32 states that contributed data for 1 or more years from 2003 to 2016 (Alaska, Arizona, Colorado, Connecticut, Georgia, Hawaii, Illinois, Indiana, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Hampshire, New Jersey, New Mexico, New York, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, Utah, Vermont, Virginia, Washington, and Wisconsin). The NVDRS abstractors coded detailed information on demographic characteristics of the victim and perpetrator, incident characteristics, and the circumstances leading up to the violent death. Each incident was also accompanied by 2 qualitative narratives summarizing the findings from the coroner/medical examiner reports and law enforcement records, respectively; these contained information gathered through interviews conducted with the victim’s friends and family, suicide notes, toxicology reports, and other available information.

Incidents were included if (1) victims were aged between 11 and 24 years, (2) the relationship between the victim and perpetrator was known, and (3) the NVDRS manner of death was classified as a homicide (largely based on the International Classification of Diseases codes assigned to the underlying cause of death on the death certificate by the coroner/medical examiner). We accepted the abstractor’s determination about whether cases of death due to possible reckless behavior with a firearm should be classified as homicides (eg, manslaughter) or unintentional firearm fatalities. While the focus of the analysis was on individuals aged 11 to 18 years (referred to as adolescents throughout this article), we included young adults (aged 19 to 24 years) to understand how adolescent IPHs compared with those of young adults. An incident was identified as an IPH if the relationship between the perpetrator and victim was coded as spouse, ex-spouse, girlfriend or boyfriend, ex-girlfriend or ex-boyfriend, or girlfriend or boyfriend (unspecified as to current or former).

Demographic characteristics available for study included sex, race/ethnicity, and age. Perpetrator information is reported for the person identified in NVDRS as the primary suspect in the incident. Age of perpetrator was continuous and also dichotomized as younger than 18 years or 18 years or older. Age difference between the perpetrator and victim was calculated by subtracting the victim’s age from the perpetrator’s age. Relationship status at the time of death was categorized as current, former, or unspecified. An incident was classified as a homicide-suicide if the perpetrator died by suicide within 24 hours of the homicide event, as specified in the NVDRS coding manual.¹⁷ Homicide method was categorized as firearm, sharp or blunt instrument (eg, knives, bats, clubs), hanging/strangulation, or other (eg, hands, poisoning, motor vehicle). If a firearm was used, firearm type was classified as handgun, shotgun, or rifle. Location of the homicide was coded as the victim’s home, other home/apartment, street/sidewalk/alley, motor vehicle/parking lot/garage, or other (eg, natural area, hotel/motel, school). All variables were coded in the NVDRS by the abstractors, and response options were grouped by the authors (with an attempt to have adequate sample size in each category).

Each homicide was assigned to a category indicating the context or motivating factor for the death after a review of all available narrative information from the coroner/medical examiner and law enforcement reports. The categories were initially created by 1 author (A.A.) a priori and generally align with previous research on adolescent homicide and adult IPH.^8,18 Two authors (A.A. and F.P.R.) read all narratives to determine which theme predominated (agreement, 92.7%). We decided to combine 2 of the original categories that were highly related and grouped 1 original category into “other” owing to very few cases. In 11 cases of disagreement, a third author (M.S.V.) read the narratives, and then all 3 authors met and reached a consensus about the most appropriate category. We present example narratives for each context category with slight alteration to protect privacy. In the narrative review, we also created indicator variables for a history of abuse of the victim by the perpetrator, whether substances such as alcohol or drugs were involved in the incident, and whether sexual violence was part of the incident. These factors have been found in prior studies to be commonly involved in IPH.^8,10

The percentage of total homicides perpetrated by an intimate partner was calculated. Descriptive statistics on victim, perpetrator, and incident characteristics were calculated for adolescent and young adult IPHs. Percentages reported do not include missing data; the amount of missing data are noted in the Table. Data were analyzed using Stata 15.1 (StataCorp).

There were 2188 homicides of adolescents aged 11 to 18 years. Of these, 6.9% (n = 150) were classified as IPH. In comparison, there were 5860 homicides of young adults aged 19 to 24 years, and 15.2% (n = 889) were IPHs. The characteristics of all IPHs in our sample and stratified by age group are presented in the Table.

Of 150 adolescent victims, 135 (90.0%) were girls and 63 (42.0%) were white. The mean (SD) age was 16.8 (1.3) years, and the median (interquartile range) age was 17.0 (16.0-18.0) years. Perpetrators of adolescent homicides were 89.9% male (n = 134) and 48.2% black (n = 68) with a mean (SD) age of 20.6 (5.0) years and median (interquartile range) age of 19.0 (18.0-22.0) years. Most perpetrators were 18 years or older (102 [77.9%]). The mean (SD) age difference between the victims and perpetrators was 3.9 (4.7) years; the median (interquartile range) difference was 3.0 (1.0-5.0) years. A total of 94 perpetrators (62.7%) were current intimate partners of the victim, 40 (26.7%) were former intimate partners, and relationship status at the time of death was unspecified for the remaining 10.7% (n = 16). Only 4 victims (2.7%) were noted as being married, in a civil union, or domestic partnership, while the rest were single or never married (data not shown). Firearms were the most common mechanism of injury (90 [61.2%]) followed by sharp or blunt instruments (37 [25.2%]). Of the firearm-related IPHs, 57 (82.6%) were perpetrated with a handgun. In 24 cases (16.0%), the perpetrator also died by suicide, in most instances by firearm. In terms of location, 56 victims (37.8%) were injured at their own residence, and 53 (35.8%) were injured at another home or apartment.

Compared with young adult victims, adolescent victims were more likely to be killed by a partner who was younger than 18 years (29 [22.1%] vs 13 [1.6%]), less likely to be killed by a current intimate partner (94 [62.7%] vs 652 [73.3%]), and less likely to be killed at the victim’s residence (56 [37.8%] vs 490 [57.2%]).

Based on the narratives, we created 4 main categories that covered nearly 70% of the adolescent homicides. We classified another 4.7% as other and 28.7% as unknown based on the narrative review. The 2 most common categories of homicides were broken/desired relationship or jealousy (41 [27.3%]) and altercation (37 [24.7%]). Another 12 homicides (8.0%) were categorized as due to reckless firearm behavior, and 10 homicides (6.7%) were categorized as pregnancy related. We provide definitions and examples of each category below. In addition, any history of abuse was noted in 27 homicides (18.0%), sexual violence specifically was noted as part of the incident in 4 homicides (2.7%) by the victim and/or perpetrator, and substance involvement by the targeted individual and/or perpetrator was noted in 12 homicides (8.0%).

The victim ended a romantic relationship with the perpetrator or would not enter into a relationship, or the homicide was precipitated by the jealousy of the perpetrator. (1) The victim was shot at home by her husband, who then shot himself. The 2 had been having marital difficulties, and the victim was in the process of leaving the relationship. They had been drinking alcohol and smoking marijuana at the time of the incident. (2) The victim was stabbed to death by her boyfriend. She had been having a relationship with another man from work that the perpetrator found out about and then invited her into a motel to talk about her relationship with the other man. He then stabbed her multiple times and fled.

The homicide occurred during the heat of an altercation or argument (nature or subject of argument was not always clear). (1) The victim was shot and killed by her boyfriend after an argument over the perpetrator not wanting the victim to return home as she wished to do. (2) The victim was found inside her home with stab wounds. She was arguing with her boyfriend over a title for a vehicle when he stabbed her. There was a long history of abuse that preceded the homicide.

The homicide was caused by reckless behavior or disregard of firearm safety and lethality (eg, handling a loaded firearm around others). (1) The victim was at home with her boyfriend, who was playing with a shotgun he claimed he believed was not loaded. The gun discharged and shot the victim in the face. (2) The victim was shot by her boyfriend, who claimed the gun was unloaded and was clearing the firearm “military style” when it discharged.

The homicide was precipitated by a pregnancy or suspected pregnancy (eg, perpetrator did not want to have the baby). (1) The victim was pregnant with her boyfriend’s baby. He did not want her to have the child and said he was going to kill the baby. He punched her in the stomach, chest, and head and struck her with a baseball bat. (2) The victim believed she was pregnant, and her boyfriend was upset that if she had the baby, he would be arrested for having sex with a minor. He tried to get her to end the pregnancy but later decided to kill her, stabbing her multiple times and dumping the body.

This study is the largest population-based examination of IPH of adolescents to date and to our knowledge. We found that approximately 7% of adolescent homicides (aged 11 to 18 years) were committed by intimate partners. Female individuals made up most of the IPH victims, in line with prior work that reported that female individuals are at much higher risk for IPH compared with male individuals.¹⁹ In this sample, approximately one-quarter of homicides of female adolescents were perpetrated by intimate partners. We found that the characteristics of adolescent IPHs differ from those of young adults in a few key ways. Adolescents are more likely to be killed by another adolescent and less likely to be killed at home. Interestingly, more than one-third of homicides occurred in a home or apartment that was not the victim’s. While we do not have information about in whose residence the homicide occurred, this may indicate that these adolescents do not cohabitate with the perpetrator but are spending time at the perpetrator’s house where there may be less supervision given most perpetrators of adolescent homicides are older than 18 years.

Firearms were the most common weapon used in adolescent IPHs, which aligns closely with prior literature on adolescent and adult homicide and the greater lethality of firearms.^8,10,20 Laws restricting access to firearms by individuals convicted of misdemeanor domestic violence offenses have been found to reduce IPH.^21,22 However, not all states statutorily define IPV as including nonmarried, noncohabitating intimate partners (commonly known as the boyfriend loophole).²³ Nonmarried, noncohabitating intimate partners are among the most common type of intimate relationships, particularly for adolescents. In this sample, 97% of adolescents were single or never married; even if the perpetrators had been convicted, nearly all of them would still be able to buy or own a gun. Therefore, a broader definition of domestic violence that extends these firearm protections to other partnerships is critical. In addition, reckless behavior with a firearm precipitated 8% of the homicides. Loaded chamber indicators and magazine safeties on firearms may help reduce the risk of fatalities in these types of cases.²⁴

More than one-quarter of homicides in this study were motivated by a broken or desired relationship or jealousy. This is consistent with previous research in adult populations that separation or threats of separation by female intimate partners represents a high-risk period motivated, at least in part, by possessiveness of the male partners.^18,20,25 There needs to be more careful consideration of assessment and intervention for adolescent girls in leaving relationships, perhaps most particularly when their partners have a history of abuse perpetration, jealousy, or possessiveness and access to lethal weapons.²⁰ Individuals as well as families, peers, and other support networks should be aware of the potential risk of violence after separating.²⁶

Although a history of prior IPV was noted in only 18% of these homicides, it is likely this is an underestimate because many adolescents abused by a partner do not disclose the abuse, particularly to adults.^14,27 However, there is a wealth of evidence from large-scale national studies that the prevalence of IPV among adolescents is substantial,^3,4 and there have been a number of interventions for the primary prevention of IPV. Successful interventions have been implemented in school and community settings to prevent IPV among adolescents by addressing topics such as relationship skills, communication strategies, identification of abusive behaviors, safety planning, relevant laws, and the roles of bystanders.^28,29

As a legal intervention, civil protection orders have been shown to be effective in reducing subsequent IPV for women.^30,31 However, adolescents in some states cannot apply for a civil protection order given statutory age or relationship restrictions (ie, those involving marriage, cohabitation, or children in common).^32,33 Additionally, some states require that adolescents obtain parental consent or a parent/guardian file on their behalf.³² Civil protection orders can be an important part of a safety plan overall and preclude legal ownership and possession of a firearm by the abusive partner.³⁴ Adolescents face a number of unique barriers in accessing support and services for IPV³⁵; therefore, all states should strongly consider the adoption of statutes that allow adolescents experiencing IPV to apply for civil protection orders without parental consent.^33,36 Indeed, stronger laws regarding adolescent IPV, specifically civil protection orders, have been found to impact state-level IPV rates.³²

The findings of this study must be considered in light of several limitations. Data for this study come only from 32 states for varying numbers of years from 2003 to 2016, so results may have limited generalizability. The NVDRS is expanding and is now funded to collect data in all 50 states; future research using expanded data will be informative, particularly given the relatively rare occurrence of adolescent IPH. Additional cases may allow for greater insight into differences in male vs female victims, for example. In addition, there is some missing data in the perpetrator variables, and the completeness of the information is limited by the accuracy and depth of detail recorded in the medical examiner and law enforcement reports. There is also no information about family histories and background of these individuals. Given that childhood factors and home environments (eg, parental IPV, childhood abuse) have been identified as risk factors for IPV,³⁷ including this information may shed more light on additional targets for intervention. Nevertheless, these findings may help to increase awareness of adolescent IPH and inform law enforcement and criminal justice efforts to prevent future deaths. Future research examining the impact of various firearm policies and laws around civil protection orders on adolescent IPH would be beneficial.

Using a rich surveillance system of violent deaths, this study identified key characteristics and contextual information that can guide the development of appropriate prevention strategies for adolescent IPH. In particular, adolescent IPH largely affects girls in circumstances of a breakup or jealousy and when perpetrators have access to firearms. Statutes for civil and criminal protection orders, firearm surrender, and other legal mechanisms to protect individuals experiencing IPV should be adapted to apply to younger individuals and to be more inclusive in terms of eligible relationship type. Programs aimed at adolescents involved in IPV should consider the contexts that may put adolescents at increased risk of homicide.

Corresponding Author: Avanti Adhia, ScD, Harborview Injury Prevention and Research Center, 325 Ninth Ave, PO Box 359960, Seattle, WA 98104 (aadhia@uw.edu).

Accepted for Publication: January 1, 2019.

Published Online: April 15, 2019. doi:10.1001/jamapediatrics.2019.0621

Author Contributions: Dr Adhia had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Adhia, Vavilala.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Adhia, Vavilala.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Adhia.

Obtained funding: Rivara.

Administrative, technical, or material support: Vavilala, Rivara.

Supervision: Kernic, Hemenway, Vavilala, Rivara.

Conflict of Interest Disclosures: Dr Adhia reports a grant from Eunice Kennedy Shriver National Institute of Child Health and Human Development during the conduct of the study. Drs Rivara and Vavilala report grants from the National Institutes of Health during the conduct of the study. No other disclosures were reported.

Funding/Support: This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (grant 5T32HD057822-09).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: This research uses data from the National Violent Death Reporting System, a surveillance system designed by the Centers for Disease Control and Prevention’s National Center for Injury Prevention and Control. The findings are based, in part, on the contributions of the funded states and territories that collected violent death data and the contributions of the states’ partners, including personnel from law enforcement, vital records, medical examiners/coroners, and crime laboratories. The analyses, results, and conclusions presented here represent those of the authors and do not necessarily reflect those of the Centers for Disease Control and Prevention.