Abstract
Background
This study evaluates associations of commonly co-occurring childhood adversities with physical violence in dating relationships to identify potential strategies for refining and targeting dating violence prevention programmes.
Methods
Data on 5130 adult respondents to a nationally representative survey with at least one dating relationship before the age of 21 years were analysed. Logistic regression models assessed associations between 12 childhood adversities and physical dating violence (PDV).
Results
Adjusting for the number of co-occurring adversities, 10 of the 12 childhood adversities were significantly associated with PDV perpetration or victimisation (OR 1.5–2.8). The population attributable risk proportion of PDV due to all 12 childhood adversities was 53.4%. Childhood adversities with the highest attributable risk proportions were sexual abuse (13.8%), interparental violence (11.6%) and parent mental illness (10.7%). Multivariate prediction equations ranked respondents by their childhood adversity risk profiles; 46.4% of PDV cases occurred in the top two risk deciles.
Conclusions
Assessment of a broad range of childhood exposures to familial adversities may help to identify adolescents at particularly high risk of PDV and to guide prevention efforts.
Physical violence in dating relationships (physical dating violence; PDV) is common among adolescents and young adults in the USA.1–3 Studies reporting associations of PDV with witnessing interparental violence in childhood support prevention efforts targeted to individuals with this particular adverse childhood experience.4–8 Similar findings have been reported in studies of violence in adult intimate relationships, supporting theories of PDV as learnt behaviour.9–18 However, the specificity of the effect of interparental violence on PDV has been called into question by studies examining a broader range of frequently co-occurring childhood adversities.19,20 Several studies have found associations between PDV and maltreatment (including childhood physical and sexual abuse and neglect).4,5,7,8,21 One study of adults found that the number of types of violent experiences in childhood, regardless of the particular types, was associated with relationship violence among adults.22 In another study, which included the broadest assessment of childhood adversities, the association between interparental violence and subsequent relationship violence among young adults was not significant in a multivariate model with statistical adjustment for the effects of co-occurring childhood adversities.23 Taken together, these findings suggest that the increased PDV risk attributed to experiences of interparental violence may be due to a broader constellation of childhood adversities, some of which may be potential targets for refining PDV prevention programmes that focus on universal education. For example, a classroom-based PDV prevention programme has been found to have some effect in reducing boys’ perpetration of PDV, with no significant effect on girls’ PDV perpetration.24 Identifying subgroups at increased risk of PDV may be useful for developing targeted interventions to supplement such universal prevention efforts.
The goal of this study is to examine the joint predictive effects on PDV of a broad range of childhood adversities, including interparental violence, in order to examine their implications for PDV prevention. The current study includes a larger, more diverse set of childhood adversities than previous studies and examines both the distinct association of each individual childhood adversities with PDV and modification of these associations when multiple childhood adversities co-occur. The contributions of each independent childhood adversity and all childhood adversities taken together to PDV prevalence are then estimated, taking into account each childhood adversity’s unique association with PDV and its tendency to co-occur with other childhood adversities in the US population.
METHODS
Sample
Data come from the National Comorbidity Survey Replication (NCS-R), a survey of the prevalence and correlates of mental disorders in a nationally representative sample of the English-speaking adult (18+ years) non-institutionalised civilian population.25–27 Computer-assisted face-to-face interviews were conducted by trained non-clinician interviewers. Field procedures, described in detail elsewhere,25,27 were approved by the Human Subjects Committees of both Harvard Medical School and the University of Michigan. Data and documentation are available at http://www.icpsr.umich.edu/CPES.
All 9282 respondents were administered a part I core diagnostic interview, for a response rate of 71%, and a subsample of 5692 part I respondents were administered a part II interview that assessed correlates and disorders of secondary focus. The sample was weighted to adjust for the differential within-household probability of selection and undersampling of hard-to-reach cases, as well as demographic and geographical distributions.27,28 The analyses presented here were carried out in the subset of part II respondents who were 21 years of age or older at the time of interview and reported having had at least one dating relationship before the age of 21 years (n=5130).
Measures
Physical violence in dating relationships
Dating relationships were defined as romantic relationships involving at least one date, with or without sexual activity. Respondents with at least one dating relationship before the age of 21 years were asked whether they ever were a victim or perpetrator of moderate (‘pushed, grabbed or shoved, threw something, slapped or hit’) or severe (‘kicked, bit, or hit with a fist, beat up, choked, burned or scalded, or threatened with a knife or gun’) physical violence in any of these relationships.29 Methodological studies suggest that this instrument is limited in its ability to distinguish between victims and perpetrators of violence, as some respondents report actions taken in self-defence as perpetration of violence while other respondents underreport perpetration due to social undesirability.7,30 To address this limitation, we conducted parallel analyses of three outcomes: victimisation, perpetration, and either victimisation or perpetration.
Childhood adversities
Twelve childhood adversities were assessed in the NCS-R: (1) parental death; (2) parental divorce; (3) other long-term parental separation (eg, adopted after age 2 years, foster care, juvenile detention, lived with relatives for 6 months or more); (4) parent mental illness (major depression, generalised anxiety disorder, or panic disorder); (5) parental substance use disorder; (6) parental criminality; (7) interparental violence; (8) serious physical illness in childhood; (9) physical abuse; (10) sexual abuse; (11) neglect; and (12) family economic adversity. Parent mental illnesses and substance use disorders were assessed using the family history research diagnostic criteria, for either male or female adult caregiver.31–33 Interparental violence and physical abuse were assessed as moderate physical violence between parents or adult caregivers or towards the respondent by a parent or adult caregiver, respectively, using the revised conflict tactics scale.29 Neglect was assessed with a five-item scale developed for child welfare studies.34 Parental criminality was assessed through questions about whether the respondent’s parents were involved in criminal activities, arrested, or sent to prison. Sexual abuse was assessed with questions developed for the baseline National Comorbidity Survey20,35 about rape, sexual assault and molestation. Economic adversity was defined as having received welfare or not having a working parent as head of the household. Information on the timing of parental death, divorce, separation, serious physical illness, interparental violence and sexual abuse was used to determine whether these childhood adversities began before the initiation of dating relationships. Only those childhood adversities that began before the initiation of dating were examined as predictors of PDV.
Analysis procedures
Prevalence of PDV was estimated as the proportion of adults with at least one dating relationship before the age of 21 years who reported moderate or severe PDV. Sociodemographic correlates of PDV were examined in multivariate logistic regression models that included respondent sex, age, race-ethnicity (white, Hispanic, black and other, including Asian, Pacific Islander and Native American), nativity (two US-born parents versus one or more foreign-born parents, as no significant differences were found among respondents with one or no US-born parents), and parent education (highest level of education attained by either parent). Age of dating initiation was also examined as a predictor, as people who begin dating early have a longer period of potential exposure to dating relationships thus are at greater risk of PDV.
Clustering of the 12 childhood adversities was examined in a factor analysis of tetrachoric correlations. A series of logistic regression models was then estimated that examined the joint predictive effects of these childhood adversities on PDV, controlling for sociodemographic factors and age at initiation of dating. The first model examined the association of each individual childhood adversity with PDV (ie, one equation for each childhood adversity). The second model examined the additive multivariate effects of all 12 childhood adversities (ie, all 12 childhood adversities in one model). The third model examined the predictive effects of the number of childhood adversities (separate dummy predictor variables for exactly one, two, three, four, five, six and seven or more childhood adversities). The fourth model examined the joint predictive effects of the 12 types of childhood adversities in addition to the number of childhood adversities (separate dummy predictor variables for exactly two to seven or more childhood adversities). In the fourth model, coefficients for the individual childhood adversities represent distinct associations of each childhood adversities with PDV when that childhood adversities occurs in isolation. The coefficients for the number of childhood adversities in this model can be interpreted as interactions among the childhood adversities constrained to be constant for all combinations involving the same number of co-occurring childhood adversities.36
Associations between childhood adversities and PDV are presented as OR. CI and statistical tests were calculated using the Taylor series linearisation method as implemented in the SUDAAN software package to account for the complex sample design of the NCS-R.37 Statistical significance was assessed consistently using two-sided 0.05 level tests.
Population attributable risk proportions (PARP)—the proportion of observed PDV that would not have occurred in the absence of childhood adversities (assuming that the OR represent causal effects of childhood adversities on PDV)—were estimated from the final logistic regression model by calculating the difference in predicted prevalence of PDV between the actual sample and a counterfactual sample in which all childhood adversities have been eliminated.
RESULTS
The prevalence of PDV and sociodemographic correlates in the US household population
Among respondents who dated before the age of 21 years, 16.0% reported either perpetration of or victimisation by PDV before the age of 21 years. The reported prevalence of PDV victimisation (13.9%) was higher than the prevalence of perpetration (8.1%). Women reported more physical violence victimisation (16.3% vs 11.2%) than men, with 9.2% of women reporting severe violence victimisation compared with 4.4% of men. A similar pattern was found for reports of physical violence perpetration in dating relationships (10.6% women vs 5.3% men).
In multivariate models including all sociodemographic predictors, PDV was higher among respondents who were women, young, had parents with low educational attainment, and had both US-born parents (table 1). Non-Hispanic black individuals reported a higher prevalence of PDV perpetration than non-Hispanic white individuals. Early age at dating initiation also predicted higher odds of PDV.
Table 1.
Associations of PDV* (victimisation, perpetration, either) with demographic characteristics and age at initiation of dating†
| Ever a victim of dating violence (n=874, 13.9%)‡ OR (95% CI) |
Ever a perpetrator of dating violence (n=549, 8.1%) OR (95% CI) |
Either victim or perpetrator of dating violence (n=1010, 16.0%) OR (95% CI) |
|
|---|---|---|---|
| Gender | |||
| Female | 1.0 | 1.0 | 1.0 |
| Male | 0.6 (0.5 to 0.7) | 0.4 (0.3 to 0.5) | 0.5 (0.4 to 0.6) |
| χ2(1)=40.4, p<0.001 | χ2(1)=75.13, p<0.001 | χ2(1)=81.03, p<0.001 | |
| Age, years | |||
| 21–32 | 1.0 | 1.0 | 1.0 |
| 33–43 | 0.7 (0.5 to 0.9) | 0.6 (0.5 to 0.8) | 0.7 (0.6 to 0.8) |
| 44–55 | 0.4 (0.3 to 0.5) | 0.5 (0.4 to 0.7) | 0.5 (0.4 to 0.6) |
| 56+ | 0.2 (0.2 to 0.3) | 0.2 (0.2 to 0.3) | 0.2 (0.2 to 0.3) |
| χ2(3)=104.5, p<0.001 | χ2(3)=8745, p<0.001 | χ2(3)=118.64, p<0.001 | |
| Race-ethnicity | |||
| Hispanic | 1.1 (0.8 to 1.4) | 1.5 (1.1 to 2.1) | 1.2 (0.9 to 1.6) |
| Black | 1.4 (1.3 to 1.9) | 2.7 (2.0 to 3.8) | 1.6 (1.2 to 2.1) |
| Other | 1.5 (0.8 to 2.8) | 1.8 (1.0 to 3.4) | 1.6 (0.9 to 2.7) |
| White | 1.0 | 1.0 | 1.0 |
| χ2(3)=6.7, p=0.081 | χ2(3)=40.30, p<0.001 | χ2(3)=11.95, p=0.008 | |
| Nativity | |||
| <2 USB parent | 1.0 | 1.0 | 1.0 |
| 2 USB parents | 1.6 (1.2 to 2.3) | 1.4 (1.0 to 1.9) | 1.6 (1.1 to 2.2) |
| χ2(1)=8.0, p=0.005 | χ2(1)=5.04, p=0.025 | χ2(1)=8.00, p=0.005 | |
| Parent education | |||
| <HS | 1.0 | 1.0 | 1.0 |
| HS | 0.7 (0.6 to 0.9) | 0.9 (0.7 to 1.3) | 0.8 (0.6 to 1.0) |
| Some college | 0.8 (0.6 to 1.1) | 0.8 (0.6 to 1.1) | 0.8 (0.6 to 1.1) |
| College graduate | 0.6 (0.5 to 0.9) | 0.5 (0.4 to 0.7) | 0.7 (0.5 to 0.9) |
| χ2(3)=9.3, p=0.026 | χ2(3)=17.47, p=0.001 | χ2(3)=7.59, p=0.055 | |
| Age of first date, years | |||
| 12 or under | 1.0 | 1.0 | 1.0 |
| 13–15 | 0.5 (0.4 to 0.8) | 0.4 (0.3 to 0.7) | 0.5 (0.4 to 0.7) |
| 16–17 | 0.3 (0.2 to 0.5) | 0.3 (0.2 to 0.5) | 0.3 (0.2 to 0.5) |
| 18 or over | 0.2 (0.2 to 0.3) | 0.2 (0.1 to 0.4) | 0.2 (0.2 to 0.3) |
| χ2(1)=71.3, p<0.001 | χ2(1)=42.88, p<0.001 | χ2(1)=69.30, p<0.001 | |
Physical dating violence (PDV) defined as ever having been a victim or perpetrator of moderate or severe physical violence in a dating relationship that began before the age of 21 years.
OR estimated in logistic regression models including all listed covariates.
Percentages are weighted, numbers are actual counts.
HS, high school; USB, US born.
Prevalence and co-occurrence of childhood adversities
Approximately half (52.9%) of NCS-R respondents reported having experienced at least one childhood adversity during their childhood, with the prevalence of individual childhood adversities ranging from a low of 4.2% (serious physical illness) to a high of 16.0% (parental divorce), numbers that are consistent with other nationally representative studies (table 2).38 Among respondents with each adversity, the proportion reporting only one adversity ranges from 48.3% (death of a parent) to 5.1% (neglect).
Table 2.
Prevalence of childhood adversities and proportion of people with each childhood adversity who have multiple adversities*
| Any adversity N (%) |
Death of a parent N (%) |
Parent divorce N (%) |
Other parent loss N (%) |
Parent mental illness N (%) |
Parent substance N (%) |
Parent criminal N (%) |
Interparent violence N (%) |
Physical illness N (%) |
Physical abuse N (%) |
Sexual abuse N (%) |
Neglect N (%) |
Economic adversity N (%) |
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Prevalence of each adversity in the total sample | 2679 (52.9) | 484 (9.5) | 813 (16) | 321 (6.3) | 511 (10.1) | 449 (8.9) | 334 (6.6) | 694 (13.7) | 212 (4.2) | 434 (8.6) | 527 (10.4) | 296 (5.8) | 502 (9.9) |
| Proportion with† | |||||||||||||
| One adversity | 1324 (49.4) | 234 (48.3) | 297 (36.5) | 64 (19.9) | 127 (24.9) | 69 (15.4) | 53 (15.9) | 82 (11.8) | 87 (41.0) | 58 (13.4) | 155 (29.4) | 15 (5.1) | 83 (16.5) |
| Two adversities | 628 (23.4) | 126 (26) | 200 (24.6) | 94 (29.3) | 106 (20.7) | 86 (19.2) | 59 (17.7) | 172 (24.8) | 48 (22.6) | 64 (14.7) | 113 (21.4) | 38 (12.8) | 149 (29.7) |
| Three adversities | 324 (12.1) | 60 (12.4) | 129 (15.9) | 56 (17.4) | 86 (16.8) | 84 (18.7) | 57 (17.1) | 147 (21.2) | 35 (16.5) | 83 (19.1) | 93 (17.6) | 44 (14.9) | 98 (19.5) |
| Four adversities | 178 (6.6) | 21 (4.3) | 77 (9.5) | 29 (9) | 68 (13.3) | 82 (18.3) | 62 (18.6) | 102 (14.7) | 16 (7.5) | 75 (17.3) | 53 (10.1) | 60 (20.3) | 66 (13.1) |
| Five adversities | 123 (4.6) | 18 (3.7) | 61 (7.5) | 29 (9) | 62 (12.1) | 57 (12.7) | 44 (13.2) | 99 (14.3) | 14 (6.6) | 72 (16.6) | 50 (9.5) | 60 (20.3) | 51 (10.2) |
| Six adversities | 51 (1.9) | 11 (2.3) | 22 (2.7) | 23 (7.2) | 28 (5.5) | 31 (6.9) | 23 (6.9) | 45 (6.5) | 6 (2.8) | 36 (8.3) | 26 (4.9) | 37 (12.5) | 20 (4) |
| Seven or more adversities | 51 (1.9) | 13 (2.7) | 27 (3.3) | 26 (8.1) | 35 (6.8) | 40 (8.9) | 35 (10.5) | 47 (6.8) | 6 (2.8) | 45 (10.4) | 37 (7) | 43 (14.5) | 34 (6.8) |
Sample consists of National Comorbidity Survey Replication part II respondents aged 21 years or older with at least one dating relationship before the age of 21 years.
Percentages represent proportions of people with the adversity who have the corresponding number of total adversities before age at first date.
Associations of childhood adversities with PDV
Associations of each individual childhood adversity with PDV, controlling for sociodemographics and age at initiation of dating, were consistently positive (table 3, model 1). Of the 36 OR estimated, 35 (97.2%) were greater than 1 and 28 (77.8%) were statistically significant at the 0.05 level. These OR were attenuated in the multivariate additive model that includes 12 childhood adversities simultaneously (model 2), in which 32 of the 36 coefficients (88.9%) were greater than 1 and 10 (27.8%) were statistically significant. A greater number of childhood adversities were positively and significantly associated with PDV victimisation (33.3% of OR significant, with a range of 1.5–2.2) than PDV perpetration (16.7%, OR significant, with a range of 1.4–1.7).
Table 3.
Four models of the association of 12 childhood adversities with PDV*
| Model 1† | Model 2‡ | Model 3§ | Model 4¶ | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Victimisation OR (95% CI) |
Perpetration OR (95% CI) |
Either OR (95% CI) |
Victimisation OR (95% CI) |
Perpetration OR (95% CI) |
Either OR (95% CI) |
Victimisation OR (95% CI) |
Perpetration OR (95% CI) |
Either OR (95% CI) |
Victimisation OR (95% CI) |
Perpetration OR (95% CI) |
Either OR (95% CI) |
|
| Parent died | 1.2 (0.9 to 1.7) | 1.0 (0.6 to 1.6) | 1.0 (0.7 to 1.5) | 1.1 (0.8 to 1.6) | 0.9 (0.5 to 1.5) | 1.0 (0.7 to 1.4) | 1.3 (1.0 to 2.0) | 1.1 (0.7 to 1.8) | 1.2 (0.8 to 1.8) | |||
| Parent divorce | 1.2 (1.0 to 1.5) | 1.3 (1.0 to 1.7) | 1.3 (1.1 to 1.6) | 1.0 (0.8 to 1.2) | 1.0 (0.8 to 1.3) | 1.1 (0.8 to 1.3) | 1.2 (0.9 to 1.7) | 1.3 (0.9 to 1.8) | 1.3 (1.0 to 1.8) | |||
| Other parent loss | 2.4 (1.7 to 3.3) | 2.0 (1.3 to 3.0) | 2.1 (1.5 to 2.9) | 1.7 (1.1 to 2.6) | 1.4 (0.8 to 2.5) | 1.5 (0.9 to 2.3) | 2.2 (1.4 to 3.6) | 1.9 (1.1 to 3.3) | 2.0 (1.2 to 3.2) | |||
| Parent mental illness | 2.3 (1.6 to 3.4) | 2.4 (1.7 to 3.2) | 2.5 (1.8 to 3.4) | 1.6 (1.1 to 2.3) | 1.7 (1.2 to 2.5) | 1.7 (1.2 to 2.4) | 2.0 (1.4 to 3.1) | 2.2 (1.3 to 3.7) | 2.2 (1.5 to 3.4) | |||
| Parent substance | 2.0 (1.5 to 2.7) | 1.8 (1.3 to 2.3) | 2.0 (1.5 to 2.6) | 1.2 (0.8 to 1.6) | 1.0 (0.7 to 1.5) | 1.2 (0.8 to 1.6) | 1.5 (1.1 to 2.2) | 1.4 (0.8 to 2.2) | 1.6 (1.1 to 2.2) | |||
| Parent criminal | 2.0 (1.4 to 2.8) | 1.8 (1.3 to 2.6) | 2.1 (1.5 to 2.9) | 1.2 (0.9 to 1.78) | 1.1 (0.8 to 1.7) | 1.3 (0.9 to 1.8) | 1.7 (1.2 to 2.6) | 1.6 (0.9 to 2.8) | 1.8 (1.2 to 2.8) | |||
| Parental violence | 2.4 (1.8 to 3.2) | 2.2 (1.6 to 2.9) | 2.3 (1.8 to 2.9) | 1.5 (1.1 to 2.0) | 1.4 (1.0 to 1.9) | 1.4 (1.1 to 1.8) | 1.9 (1.4 to 2.6) | 1.9 (1.2 to 2.8) | 1.9 (1.4 to 2.7) | |||
| Physical abuse | 2.6 (2.0 to 3.3) | 2.4 (1.8 to 3.3) | 2.6 (2.1 to 3.3) | 1.3 (1.0 to 1.7) | 1.4 (0.9 to 2.0) | 1.4 (1.0 to 1.8) | 1.8 (1.3 to 2.7) | 2.0 (1.2 to 3.2) | 2.0 (1.4 to 2.9) | |||
| Sexual abuse | 3.0 (2.3 to 3.9) | 2.0 (1.5 to 2.8) | 2.7 (2.2 to 3.5) | 2.2 (1.8 to 2.8) | 1.4 (1.0 to 2.1) | 2.0 (1.6 to 2.5) | 2.9 (2.0 to 4.0) | 1.9 (1.1 to 3.3) | 2.7 (2.0 to 3.7) | |||
| Neglect | 3.0 (2.0 to 4.3) | 2.7 (2.0 to 3.6) | 2.9 (2.1 to 4.0) | 1.4 (2.0 to 2.0) | 1.3 (0.9 to 2.0) | 1.4 (1.0 to 1.9) | 2.1 (1.4 to 3.1) | 2.1 (1.2 to 3.6) | 2.1 (1.4 to 3.3) | |||
| Physical illness | 1.3 (0.8 to 2.3) | 1.3 (0.7 to 2.5) | 1.4 (0.9 to 2.2) | 1.1 (0.7 to 2.0) | 1.2 (0.6 to 2.2) | 1.2 (0.8 to 1.9) | 1.5 (0.8 to 2.5) | 1.5 (0.7 to 3.2) | 1.6 (0.9 to 2.6) | |||
| Economic adversity | 1.4 (2.0 to 1.9) | 2.0 (1.3 to 3.0) | 1.5 (1.1 to 2.2) | 0.9 (0.7 to 1.3) | 1.5 (0.9 to 2.4) | 1.1 (0.7 to 1.6) | 1.3 (0.8 to 1.9) | 2.0 (1.1 to 3.8) | 1.5 (0.9 to 2.4) | |||
| χ2(12)=147.7, p<0.001 | χ2(12)=215.7, p<0.001 | χ2(12)=149.5, p<0.001 | χ2(12)=90.7, p<0.001 | χ2(12)=41.3, p<0.001 | χ2(12)=67.2, p<0.001 | |||||||
| No of adversities | ||||||||||||
| 1 | 1.7 (1.3 to 2.2) | 1.7 (1.2 to 2.4) | 1.8 (1.4 to 2.2) | NA | NA | NA | ||||||
| 2 | 2.3 (1.7 to 3.2) | 2.0 (1.2 to 3.4) | 2.3 (1.6 to 3.3) | 0.8 (0.5 to 1.2) | 0.7 (0.4 to 1.2) | 0.8 (0.5 to 1.2) | ||||||
| 3 | 3.1 (2.2 to 4.2) | 3.3 (2.3 to 4.7) | 3.1 (2.4 to 4.1) | 0.6 (0.3 to 1.1) | 0.6 (0.3 to 1.6) | 0.5 (0.3 to 1.0) | ||||||
| 4 | 3.5 (2.5 to 4.8) | 3.6 (2.5 to 5.2) | 3.4 (2.5 to 4.6) | 0.4 (0.2 to 0.9) | 0.4 (0.1 to 1.5) | 0.3 (0.1 to 0.8) | ||||||
| 5 | 3.9 (2.3 to 6.7) | 3.1 (1.8 to 5.3) | 4.2 (2.6 to 6.8) | 0.2 (0.1 to 0.7) | 0.2 (0 to 0.9) | 0.2 (0.1 to 0.6) | ||||||
| 6 | 9.1 (3.5 to 23.4) | 4.6 (2.1 to 9.0) | 8.2 (3.2 to 20.9) | 0.3 (0.1 to 1.3) | 0.2 (0 to 1.0) | 0.2 (0 to 1.1) | ||||||
| 7 or more | 5.9 (3.4 to 10.3) | 6.5 (4.0 to 10.8) | 6.0 (3.6 to 10.2) | 0.1 (0 to 0.4) | 0.1 (0.01 to 1.00) | 0.1 (0 to 0.3) | ||||||
| χ2(7)=117.1, p<0.001 | χ2(7)=106.0, p<0.001 | χ2(7)=137.2, p<0.001 | χ2(6)=18.5, p=0.005 | χ2(6)=21.3, p=0.0016 | χ2(6)=20.0, p=0.0027 | |||||||
OR estimated in logistic regression models adjusted for age, gender, parental education, nativity, race/ethnicity and age at first date. Sample includes National Comorbidity Survey Replication part II respondents aged 21 years or older with at least one dating relationship before the age of 21 years. Values in bold are significant at the p=0.05 level.
Model 1 is estimated with one adversity at a time in addition to the controls noted in the previous footnote.
Model 2 is estimated with all 12 adversities in addition to the controls noted in the first footnote.
Model 3 is estimated with dummy predictors for number of adversities, without any information about the types of adversities. The same controls used in earlier models were also included.
Model 4 is estimated with dummy predictors for number of adversities as well as information about the types of adversities. The same controls used in earlier models were also included.
PDV, physical dating violence.
Model 3 includes only the number of childhood adversities. A single categorical variable indicating the total number of childhood adversities was used based on the factor analysis of the tetrachoric correlations among all 12 childhood adversities, which found a single factor with an eigenvalue greater than 1 (unrotated eigenvalues for first and second factors=3.9, 0.6, respectively). This model found a generally increasing odds of PDV associated with an increasing number of childhood adversities, with OR ranging from a low of 1.8 for respondents with exactly one childhood adversity (compared with respondents who had no childhood adversities) to highs of 6.0–8.2 for respondents with six or more childhood adversities. OR between the number of childhood adversities and the risk of victimisation versus perpetration were of similar magnitude.
In the final multivariate model that included indicators for both the type and number of childhood adversities (model 4), all 36 OR associated with type were greater than 1, with 23 (63.9%) statistically significant (in the range 1.5–2.8). In this model, coefficients for the individual childhood adversities represent distinct associations of each childhood adversity with PDV when that childhood adversity occurs in isolation. Six childhood adversities were significantly associated with PDV in all three PDV outcomes: other parent loss, parent mental illness, interparental violence, physical abuse, sexual abuse and neglect. An additional statistical test was performed to examine the hypothesis that the distinct associations of all 12 childhood adversities are identical in magnitude. After accounting for having any single childhood adversity, the type of childhood adversity remained significantly associated with victimisation (χ2(11)=59.98, p<0.001), perpetration (χ2(11)=29.94, p=0.002) and any PDV (χ2 (11)=37.82, p<0.001).
OR associated with the number of childhood adversities in the model including both type and number of childhood adversities (model 4) show that the predictive effects of co-occurring childhood adversities are sub-additive. That is, each additional childhood adversity is associated with a smaller incremental increase in risk. This means, for example, that the OR for PDV victimisation associated with parental mental illness is 2.0 for a person with only one childhood adversity, but only 1.2 (2.0*0.6=1.2) for a person with three childhood adversities. Statistical interactions between gender and each individual childhood adversity were tested to assess for variation by gender in the association between childhood adversities and PDV. Of the 36 interactions, none were statistically significant (results available upon request).
Population attributable risk proportions
PARP were calculated using the model with indicators for both the type and number of childhood adversities (model 4). Across outcomes (victimisation, perpetration, either), PARP for individual childhood adversities ranged from 1.6% to 15.6% (table 4). Physical illness and parental death were associated with the smallest PARP. Parent mental illness, interparental violence and childhood sexual abuse were consistently associated with PARP greater than 10%. Physical abuse and economic adversity were associated with PARP greater than 10% in predicting PDV perpetration. The PARP for all childhood adversities combined ranged from 53.4% to 56.5%, suggesting that childhood adversities accounted for approximately half of all PDV cases.
Table 4.
Attributable risk for physical dating violence due to 12 childhood adversities*
| Victimisation | Perpetration | Either | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Average predicted probability of PDV |
Sum of probabilities of PDV |
% Change in probability of PDV |
Average predicted probability of PDV |
Sum of probabilities of PDV |
% Change in probability of PDV |
Average predicted probability of PDV |
Sum of Probabilities of PDV |
% Change in Probability of PDV |
|
| Actual sample | 0.162 | 827.99 | 0.095 | 487.48 | 0.185 | 946.56 | |||
| Adversity type† | |||||||||
| Parent died | 0.158 | 807.82 | 2.4% | 0.094 | 479.60 | 1.6% | 0.182 | 931.48 | 1.6% |
| Parent divorce | 0.155 | 794.48 | 4.0% | 0.090 | 457.71 | 6.1% | 0.175 | 896.80 | 5.3% |
| Other parent loss | 0.152 | 774.83 | 6.4% | 0.089 | 452.80 | 7.1% | 0.176 | 897.51 | 5.2% |
| Parent mental illness | 0.146 | 745.33 | 10.2% | 0.084 | 427.21 | 12.6% | 0.166 | 846.60 | 10.7% |
| Parent substance | 0.153 | 782.84 | 5.5% | 0.090 | 462.28 | 5.2% | 0.175 | 895.59 | 5.4% |
| Parent criminal | 0.154 | 788.14 | 4.8% | 0.090 | 460.52 | 5.5% | 0.176 | 898.92 | 5.0% |
| Interparental violence | 0.142 | 723.20 | 12.7% | 0.081 | 416.52 | 14.6% | 0.164 | 837.00 | 11.6% |
| Physical abuse | 0.149 | 763.64 | 8.2% | 0.086 | 438.88 | 10.5% | 0.170 | 870.08 | 8.5% |
| Sexual abuse | 0.137 | 699.18 | 15.6% | 0.084 | 431.44 | 11.5% | 0.160 | 815.44 | 13.9% |
| Neglect | 0.152 | 775.95 | 6.3% | 0.088 | 451.75 | 7.3% | 0.174 | 890.76 | 5.9% |
| Physical illness | 0.159 | 814.39 | 1.6% | 0.093 | 476.93 | 2.2% | 0.182 | 929.10 | 1.8% |
| Economic adversity | 0.157 | 800.36 | 4.0% | 0.085 | 433.89 | 12.4% | 0.176 | 900.07 | 5.7% |
| All 12 childhood adversities | 0.075 | 383.39 | 53.7% | 0.041 | 211.92 | 56.5% | 0.086 | 440.72 | 53.4% |
Predicted probabilities of adolescent PDV estimated in logistic regression models with controls for age, gender, nativity, parental education, race/ethnicity and age at first date. Results for the actual sample reflect the actual distribution of all exposures in the dataset. Attributable risk proportions were estimated by calculating the predicted probability of dating violence using the same logistic regression model and a modified dataset in which the values for each exposure were set to 0=‘no exposure’.
Figures in each row show the estimated risk of PDV when the listed adversity is removed from the sample.
PDV, physical dating violence.
In order to examine the clustering of PDV among people at high risk, respondents were ranked into deciles according to their predicted probability of any PDV based on the model with indicators for both the type and number of childhood adversities. The actual prevalence of PDV (either perpetration or victimisation) ranged from 2.1% to 45.3% between the lowest and highest risk deciles (see supplementaray table 1, available online only); 46.4% of PDV cases were in the top two deciles of risk and 57.9% were in the top three deciles of risk.
DISCUSSION
Our findings confirm that PDV in adolescence is common in the US population and is positively associated with a broad range of childhood adversities including, but not limited to, interparental violence. Previous studies have found that the relationship between interparental violence and later relationship violence is attenuated or null following adjustment for co-occurring adversities in multivariate additive models.17,23,39 Our results suggest that those models do not adequately represent the association between childhood adversities and PDV. Due to clustering of childhood adversities, bivariate associations of individual childhood adversity with PDV are artificially inflated (table 3, model 1) while additive multivariate associations are artificially attenuated (table 3, model 2). A model that simply counts childhood adversities (table 3, model 3) offers no test of the influence of specific types of childhood adversity. The model developed in this study (table 3, model 4) allows for variation in the distinct effects of individual childhood adversity and for non-linear changes in the incremental effects of multiple childhood adversities. In this model, interparental violence re-emerges as a strong predictor of PDV, along with a broader range of childhood adversities, including sexual abuse and parent mental illness.40,41
The subadditive interactions between multiple co-occurring childhood adversities means that the magnitude of association between each individual childhood adversity and PDV is reduced by a fixed proportion from its estimated distinct effect when it co-occurs with other childhood adversities. The extent of reduction increases in magnitude with additional childhood adversities, in a non-linear fashion, so that PDV risk reaches a ceiling beyond which additional childhood adversities are not associated with further increases in risk. An implication of this pattern is that preventing one childhood adversity in a person with multiple childhood adversities may have a minimal impact on reducing the risk of PDV. One limitation of model 4 is the simplifying assumption that the interaction among multiple childhood adversities is diffuse,36 ie, that the interaction depends only on the number of co-occurring childhood adversities rather than the specific co-occurring childhood adversities. Estimates of diffuse interactions may be affected by the patterns of co-occurrence of childhood adversities in this population and thus the specific parameter estimates may not be generalisable across populations in which these patterns differ. However, the pattern of subadditivity is likely to be similar across populations, as evidenced by recent cross-national analyses of associations between childhood adversities and psychiatric disorders employing this model.42
PARP estimated from this model suggest that the 12 childhood adversities together account for more than half of PDV cases in this population. While not an exhaustive list of childhood familial adversities, this dataset offers a broad assessment of childhood adversities in a nationally representative sample. Realistically, the estimated associations of childhood adversities with PDV are unlikely to represent purely causal effects, due to common causes as well as other unmeasured factors; however, the analysis provides valuable information for prevention efforts. First, the large magnitude of the PARP for all 12 childhood adversities taken together suggests that attention to childhood adversities, whether as risk markers or as causes of PDV, can help prospectively identify a large proportion of youth at risk of PDV. Second, the finding that a diverse group of childhood adversities accounts for comparable proportions of PDV suggests that prevention programmes that address pathways specific to particular adversities may have limited impact. Programmes that aim to address multiple adversities or common mediating pathways have greater potential for reducing the occurrence of PDV. For instance, programmes providing social support targeted to at-risk adolescents may address a range of vulnerabilities with common origins in dysfunctional childhood family environments. Specific targeted intervention strategies to address children exposed to multiple adverse childhood experiences (such as trauma-informed cognitive behavioural therapy and parent–child interaction therapy) may be viable strategies for reducing the risk of physical violence in subsequent adolescent relationships in addition to universal PDV prevention programmes. The importance of addressing multiple childhood adversities through prevention efforts is further underlined by the finding that nearly half (46.36%) of all PDV in the population is concentrated among people in the top 20% of PDV risk as estimated in this model.
Results should be considered in the light of several limitations. First, the assessment of PDV in this survey does not include sexual violence nor emotional abuse.3,43 Future studies should investigate whether the patterns identified here apply to adolescent dating violence more generally. Second, epidemiological surveys have limited capacity to differentiate between PDV victimisation and perpetration; most respondents report both, and perpetration is likely to be underreported.7,30,44,45 The results reported here, similar across the three outcomes, should be interpreted in terms of the risk of being in a dating relationship in which physical violence occurs. Third, data on both childhood adversities and PDV are retrospective, and their association may be affected by recall bias, including family history. Our findings are similar to those of prospective studies, when similar analytical models are compared, but the potential for recall bias remains a concern. Fourth, reporting of PDV may differ between men and women, and this dataset does not include information about contexts or motivations for PDV. Statistical interactions between gender and each individual childhood adversity were tested, but none of these interactions was statistically significant (results available on request). Therefore, childhood adversities are associated with elevated PDV risk for both men and women, but whether this association occurs through similar pathways cannot be discerned from these data.
The findings add support to prevention strategies that address the shared impact of multiple childhood adversities reflective of dysfunctional family environments. In addition to primary prevention efforts that focus on strengthening families, the findings suggest the utility of testing targeted PDV interventions for adolescents with known exposure to multiple childhood adversities.
Supplementary Material
What is already known on this subject.
-
►
Physical violence in dating relationships is common and is associated with adverse childhood experiences.
-
►
This study examined the joint predictive effects of a broad range of childhood adversities on PDV, including witnessing interparental violence, in order to examine their implications for PDV prevention.
-
►
Identifying subgroups at increased risk of PDV may be useful for developing targeted interventions to supplement universal prevention efforts.
What this study adds.
-
►
Physical violence in dating relationships is positively associated with a broad range of childhood adversities including, but not limited to, childhood sexual abuse, interparental violence and parent mental illness.
-
►
The findings add support to prevention strategies that address the shared impact of multiple childhood adversities reflective of dysfunctional family environments.
Acknowledgments
Funding This study was supported by grants from Building Interdisciplinary Research Careers in Women’s Health to EM (BIRCWH, K12 HD051958–National Institute of Child Health and Human Development (NICHD); Office of Research on Women’s Health (ORWH); Office of Dietary Supplements (ODS); National Institute of Aging (NIA)) and National Institute of Mental Health KO1 MH66057 to JB. Other funders: The National Comorbidity Survey Replication (NCS-R) is supported by the National Institute of Mental Health (NIMH; U01–MH60220) with supplemental support from the National Institute on Drug Abuse (NIDA), the Substance Abuse and Mental Health Services Administration (SAMHSA), the Robert Wood Johnson Foundation (RWJF; Grant 044780), and the John W. Alden Trust. Collaborating NCS-R investigators include Ronald C. Kessler (Principal Investigator, Harvard Medical School), Kathleen Merikangas (Co-Principal Investigator, NIMH), James Anthony (Michigan State University), William Eaton (The Johns Hopkins University), Meyer Glantz (NIDA), Doreen Koretz (Harvard University), Jane McLeod (Indiana University), Mark Olfson (New York State Psychiatric Institute, College of Physicians and Surgeons of Columbia University), Harold Pincus (University of Pittsburgh), Greg Simon (Group Health Cooperative), Michael Von Korff (Group Health Cooperative), Philip S. Wang (NIMH), Kenneth Wells (UCLA), Elaine Wethington (Cornell University), and Hans-Ulrich Wittchen (Max Planck Institute of Psychiatry; Technical University of Dresden). The views and opinions expressed in this report are those of the authors and should not be construed to represent the views of any of the sponsoring organizations, agencies, or US Government. A complete list of NCS publications and the full text of all NCS-R instruments can be found at http://www.hcp.med.harvard.edu/ncs. Send correspondence to ncs@hcp.med.harvard.edu. The NCS-R is carried out in conjunction with the World Health Organization World Mental Health (WMH) Survey Initiative. We thank the staff of the WMH Data Collection and Data Analysis Coordination Centres for assistance with instrumentation, fieldwork, and consultation on data analysis. These activities were supported by the National Institute of Mental Health (R01 MH070884), the John D. and Catherine T. MacArthur Foundation, the Pfizer Foundation, the US Public Health Service (R13–MH066849, R01–MH069864, and R01 DA016558), the Fogarty International Center (FIRCA R03-TW006481), the Pan American Health Organization, Eli Lilly and Company, Ortho-McNeil Pharmaceutical, Inc., GlaxoSmithKline, and Bristol-Myers Squibb. A complete list of WMH publications can be found at http://www.hcp.med.harvard.edu/wmh/.
Footnotes
Competing interests RCK, the senior author, has been a consultant for AstraZeneca, Analysis Group, Bristol-Myers Squibb, Cerner-Galt Associates, Eli Lilly and Company, GlaxoSmithKline Inc., HealthCore Inc., Health Dialog, Integrated Benefits Institute, John Snow Inc., Kaiser Permanente, Matria Inc., Mensante, Merck and Co, Inc., Ortho-McNeil Janssen Scientific Affairs, Pfizer Inc., Primary Care Network, Research Triangle Institute, Sanofi-Aventis Groupe, Shire US Inc., SRA International, Inc., Takeda Global Research and Development, Transcept Pharmaceuticals Inc. and Wyeth-Ayerst; has served on advisory boards for Appliance Computing II, Eli Lilly and Company, Mindsite, Ortho-McNeil Janssen Scientific Affairs and Wyeth-Ayerst, and has had research support for his epidemiological studies from Analysis Group Inc., Bristol-Myers Squibb, Eli Lilly and Company, EPI-Q, GlaxoSmithKline, Johnson and Johnson Pharmaceuticals, Ortho-McNeil Janssen Scientific Affairs., Pfizer Inc., Sanofi-Aventis Groupe and Shire US, Inc. Of note, the National Comorbidity Survey Replication (NCSR), on which this manuscript is based, is a public use dataset; there is no apparent conflict of interest relevant to the subject presented in this manuscript.
Ethics approval This study was conducted with the approval of the Harvard Medical School and University of Michigan (for National Comorbidity Survey Replication primary data collection).
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
- 1.Howard D, Wang M, Yan F. Psychosocial factors associated with reports of physical dating violence among U.S. adolescent females. Adolescence. 2007;42:311–324. [PubMed] [Google Scholar]
- 2.Howard D, Wang M, Yan F. Psychosocial factors associated with reports of physical dating violence victimization among U.S. adolescent males. Adolescence. 2008;43:449–460. [PubMed] [Google Scholar]
- 3.Halpern CT, Oslak SG, Young ML, et al. Partner violence among adolescents in opposite-sex romantic relationships: findings from the National Longitudinal Study of Adolescent Health. Am J Public Health. 2001;91:1679–1685. doi: 10.2105/ajph.91.10.1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Magdol L, Moffitt TE, Caspi A, et al. Developmental antecedents of partner abuse: a prospective-longitudinal study. J Abnorm Psychol. 1998;107:375–389. doi: 10.1037//0021-843x.107.3.375. [DOI] [PubMed] [Google Scholar]
- 5.Fang XM, Corso PS. Child maltreatment, youth violence, and intimate partner violence–developmental relationships. Am J Prev Med. 2007;33:281–290. doi: 10.1016/j.amepre.2007.06.003. [DOI] [PubMed] [Google Scholar]
- 6.Wekerle C, Wolfe DA, Hawkins DL, et al. Childhood maltreatment, posttraumatic stress symptomatology, and adolescent dating violence: considering the value of adolescent perceptions of abuse and a trauma mediational model. Dev Psychopathol. 2001;13:847–871. [PubMed] [Google Scholar]
- 7.Wolfe DA, Scott K, Wekerle C, et al. Child maltreatment: risk of adjustment problems and dating violence in adolescence. J Am Acad Child Adolesc Psychiatry. 2001;40:282–289. doi: 10.1097/00004583-200103000-00007. [DOI] [PubMed] [Google Scholar]
- 8.Wolfe DA, Wekerle C, Scott K, et al. Predicting abuse in adolescent dating relationships over 1 year: the role of child maltreatment and trauma. J Abnorm Psychol. 2004;113:406–415. doi: 10.1037/0021-843X.113.3.406. [DOI] [PubMed] [Google Scholar]
- 9.Doumas D, Margolin G, John RS. The intergenerational transmission of aggression across 3 generations. J Fam Violence. 1994;9:157–175. [Google Scholar]
- 10.Straus M, Gelles R. In: Physical violence in American families: risk factors and adaptations to violence in 8,145 families. Smith C, editor. New Brunswick, NJ: Transaction Press; 1990. [Google Scholar]
- 11.Straus MA. Children as witnesses to marital violence: a risk factor for lifelong problems among a nationally representative sample of American men and women. In: Schwarz DF, editor. Children and violence: report of the twenty-third roundtable. Columbus, OH: Ross Laboratories; 1992. pp. 599–602. [Google Scholar]
- 12.Ehrensaft MK, Cohen P, Brown J, et al. Intergenerational transmission of partner violence: a 20-year prospective study. J Consult Clin Psychol. 2003;71:741–753. doi: 10.1037/0022-006x.71.4.741. [DOI] [PubMed] [Google Scholar]
- 13.Ehrensaft MK, Moffitt TE, Caspi A. Clinically abusive relationships in an unselected birth cohort: men’s and women”s participation and developmental antecedents. J Abnorm Psychol. 2004;113:258–270. doi: 10.1037/0021-843X.113.2.258. [DOI] [PubMed] [Google Scholar]
- 14.Hotaling G, Sugarman D. An analysis of risk markers in husband to wife violence: the current state of knowledge. Violence Vict. 1986;1:101–124. [PubMed] [Google Scholar]
- 15.Herrenkohl T, Mason W, Kosterman R, et al. Pathways from physical childhood abuse to partner violence in young adulthood. Violence Vict. 2004;19:123–136. doi: 10.1891/vivi.19.2.123.64099. [DOI] [PubMed] [Google Scholar]
- 16.Simons RL, Wu CI, Johnson C, et al. A test of various perspecitves on the intergenerational transmission of domestic violence. Criminology. 1995;33:141–172. [Google Scholar]
- 17.Stith SM, Rosen KH, Middleton KA, et al. The intergenerational transmission of spouse abuse: a meta-analysis. J Marriage Fam. 2000;62:640–654. [Google Scholar]
- 18.Gil-Gonzalez D, Vives-Cases C, Ruiz MT, et al. Childhood experiences of violence in perpetrators as a risk factor of intimate partner violence: a systematic review. J Public Health. 2008;30:14–22. doi: 10.1093/pubmed/fdm071. [DOI] [PubMed] [Google Scholar]
- 19.Dong M, Anda R, Felitti V, et al. The interrelatedness of multiple forms of childhood abuse, neglect, and household dysfunction. Child Abuse Negl. 2004;28:771–784. doi: 10.1016/j.chiabu.2004.01.008. [DOI] [PubMed] [Google Scholar]
- 20.Kessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychol Med. 1997;27:1101–1119. doi: 10.1017/s0033291797005588. [DOI] [PubMed] [Google Scholar]
- 21.Finkelhor D, Hotaling G, Lewis IA, et al. Sexual abuse in a national survey of adult men and women–prevalance, characteristics, and risk-factors. Child Abuse Negl. 1990;14:19–28. doi: 10.1016/0145-2134(90)90077-7. [DOI] [PubMed] [Google Scholar]
- 22.Whitfield C, Anda R, Dube S, et al. Violent childhood experiences and the risk of intimate partner violence in adults: assessment in a large health maintenance organization. J Interpers Violence. 2003;18:166–185. [Google Scholar]
- 23.Fergusson DM, Boden JM, Horwood LJ. Examining the intergenerational transmission of violence in a New Zealand birth cohort. Child Abuse Negl. 2006;30:89–108. doi: 10.1016/j.chiabu.2005.10.006. [DOI] [PubMed] [Google Scholar]
- 24.Wolfe DA, Crooks C, Jaffe P, et al. A school-based program to prevent adolescent dating violence: a cluster randomized trial. Arch Pediatr Adolesc Med. 2009;163:692–699. doi: 10.1001/archpediatrics.2009.69. [DOI] [PubMed] [Google Scholar]
- 25.Kessler R, Merikangas K. The National Comorbidity Survey Replication (NCS-R): background and aims. Int J Methods Psychiatr Res. 2004;13:60–68. doi: 10.1002/mpr.166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Breslau J, Kendler KS, Su M, et al. Lifetime risk and persistence of psychiatric disorders across ethnic groups in the United States. Psychol Med. 2005;35:317–327. doi: 10.1017/s0033291704003514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kessler RC, Berglund P, Chiu WT, et al. The US National Comorbidity Survey Replication (NCS-R): design and field procedures. Int J Methods Psychiatr Res. 2004;13:69–92. doi: 10.1002/mpr.167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kessler R, McGonagle K, Zhao S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51:8–19. doi: 10.1001/archpsyc.1994.03950010008002. [DOI] [PubMed] [Google Scholar]
- 29.Straus MA, Hamby SL, BoneyMcCoy S, et al. The revised Conflict Tactics Scales (CTS2)–development and preliminary psychometric data. J Fam Issues. 1996;17:283–316. [Google Scholar]
- 30.Jackson SM. Issues in the dating violence research: a review of the literature. Aggress Violent Behav. 1999;4:233–247. [Google Scholar]
- 31.Kendler K, Silberg J, Neale M, et al. The family history method: whose psychiatric history is measured? Am J Psychiatry. 1991;148:1501–1504. doi: 10.1176/ajp.148.11.1501. [DOI] [PubMed] [Google Scholar]
- 32.Endicott J, Andreasen N, Spitzer RL. Family history research diagnostic criteria. New York, NY: Biometrics Research, New York State Psychiatric Institute; 1978. [Google Scholar]
- 33.Andreasen NC, Endicott J, Spitzer RL, et al. The family history method using diagnostic criteria. Reliability and validity. Arch Gen Psychiatry. 1977;34:1229–1235. doi: 10.1001/archpsyc.1977.01770220111013. [DOI] [PubMed] [Google Scholar]
- 34.Courtney ME, Piliavin I, Grogan-Kaylor A, et al. Foster youth transitions to adulthood: a longitudinal view of youth leaving care. Madison, WI: Institute for Research on Poverty; 1998. [PubMed] [Google Scholar]
- 35.Molnar BE, Buka SL, Kessler RC. Child sexual abuse and subsequent psychopathology: results from the National Comorbidity Survey. Am J Public Health. 2001;91:753–760. doi: 10.2105/ajph.91.5.753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Gustafson P, Kazi A, Levy A. Extending logistic regression to model diffuse interactions. Stat Med. 2005;24:2089–2104. doi: 10.1002/sim.2093. [DOI] [PubMed] [Google Scholar]
- 37.Research Triangle Institute. SUDAAN. Professional software for data analysis, version 8.1. Research Triangle Park, NC: Research Triangle Institute; 2002. [Google Scholar]
- 38.Turner HA, Finkelhor D, Ormrod R. The effect of lifetime victimization on the mental health of children and adolescents. Soc Sci Med. 2006;62:13–27. doi: 10.1016/j.socscimed.2005.05.030. [DOI] [PubMed] [Google Scholar]
- 39.Stith SM, Smith DB, Penn CE, et al. Intimate partner physical abuse perpetration and victimization risk factors: a meta-analytic review. Aggress Violent Behav. 2004;10:65–98. [Google Scholar]
- 40.McLaughlin KA, Green JG, Gruber MJ, et al. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication ii associations with persistence of DSM-IV disorders. Arch Gen Psychiatry. 2010;67:124–132. doi: 10.1001/archgenpsychiatry.2009.187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Green JG, McLaughlin KA, Berglund PA, et al. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication i associations with first onset of DSM-IV disorders. Arch Gen Psychiatry. 2010;67:113–123. doi: 10.1001/archgenpsychiatry.2009.186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kessler RC, McLaughlin KA, Green JG, et al. Childhood adversities and adult psychopathology in the World Health Organization World Mental Health Surveys. Br J Psychiatry. 2010;197:378–385. doi: 10.1192/bjp.bp.110.080499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Silverman JG, Raj A, Mucci LA, et al. Dating violence against adolescent girls and associated substance use, unhealthy weight control, sexual risk behavior, pregnancy, and suicidality. JAMA. 2001;286:572–579. doi: 10.1001/jama.286.5.572. [DOI] [PubMed] [Google Scholar]
- 44.Hamby SL. Measuring gender differences in partner violence: implications from research on other forms of violent and socially undesirable behavior. Sex Roles. 2005;52:725–742. [Google Scholar]
- 45.Kaukinen C. The help-seeking decisions of violent crime victims–an examination of the direct and conditional effects of gender and the victim–offender relationship. J Interpers Violence. 2002;17:432–456. doi: 10.1177/0886260504268000. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.