Child Abuse & Neglect 32 (2008) 819–829

Contents lists available at ScienceDirect

Child Abuse & Neglect

Psychological abuse between parents: Associations with childmaltreatment from a population-based sample�

Jen Jen Changa,b,∗, Adrea D. Theodorec, Sandra L. Martinb, Desmond K. Runyanc

a Department of Community Health in Epidemiology, School of Public Health, St. Louis University, 3545 Lafayette Avenue,Ste. 300, St. Louis, MO 63104, USAb Department of Maternal and Child Health, School of Public Health, University of North Carolina, Chapel Hill, NC, USAc Departments of Social Medicine and Pediatrics, School of Medicine, University of North Carolina, Chapel Hill, NC, USA

a r t i c l e i n f o

Article history:Received 26 April 2006Received in revised form 6 November 2007Accepted 17 November 2007Available online 26 August 2008

Keywords:Child maltreatmentChild abuseDomestic violenceIntimate partner violencePsychological abuse

a b s t r a c t

Objective: This study examined the association between partner psychological abuse andchild maltreatment perpetration.Methods: This cross-sectional study examined a population-based sample of mothers withchildren aged 0–17 years in North and South Carolina (n = 1,149). Mothers were asked aboutthe occurrence of potentially neglectful or abusive behaviors toward their children by eitherthemselves or their husband/partner in the past year. Partner psychological abuse wascategorized as no psychological abuse (reference), husband perpetrates, wife perpetrates,or both perpetrate. Outcome measures for psychological and physical abuse of the childhad four categories: no abuse (reference), mother perpetrates, father/father-figure perpe-trates, or both parents perpetrate, whereas child neglect was binary. Adjusted relative riskratios (aRRRs), adjusted odds ratios, and 95% confidence intervals (CIs) were estimated withregression models. A relative risk ratio was the ratio of odds ratios derived from multinomiallogistic regression.Results: Children were at the greatest risk of maltreatment when parents psychologicallyabused each other versus no abuse: the aRRR for child psychological abuse by the motheronly was 16.13 (95% CI: 5.11, 50.92) compared to no abuse, controlling for child age, gender,Medicaid welfare, and mother’s level of education. Both parents psychologically abuse eachother versus no abuse also results in an aRRR of 14.57 (95% CI: 3.85, 55.16) for child physicalabuse by both parents compared to no abuse. When only the husband perpetrates towardthe wife, the odds of child neglect was 5.29 times as much as families with no psychologicalabuse (95% CI: 1.36, 20.62).Conclusions: Partner psychological abuse was strongly related to child maltreatment. Chil-dren experienced a substantially increased risk of maltreatment when partner psychologicalabuse was present in the homes.Practice implications: This study observed that intimate partner psychological abuse sig-nificantly increased risk of child maltreatment. Increased public awareness of partnerpsychological abuse is warranted. Primary prevention should include education about theseriousness of partner psychological abuse in families. Domestic violence and child welfareagencies must recognize the link between partner psychological abuse and child maltreat-ment and work together to develop effective screening for each of these problems.

© 2008 Elsevier Ltd. All rights reserved.

� This project was generously supported by The Duke Endowment, Inc., Charlotte, NC. The conclusions and opinions expressed in the manuscript arethose of the authors and do not necessarily represent the views of The Duke Endowment, its officers, or members of the Board of Directors.

∗ Corresponding author.

0145-2134/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.chiabu.2007.11.003

820 J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829

Introduction

Psychological abuse has been referred to as “coercive or aversive acts intended to produce emotional harm or threat ofharm. In contrast to physical abuse, these coercive behaviors are not directed toward the target’s bodily integrity, but areinstead directed at the recipient’s sense of self” (Murphy & Cascardi, 1999, p. 202). In other words, psychological abuse isoften intended to diminish another person’s self-esteem and mental well-being (O’Leary, 1999). Partner psychological abuseis an important form of abuse because many women report that it is harmful to their well-being or worse than physical abusethey suffer (Follingstad, Rutledge, Berg, Hause, & Polek, 1990; Tolman & Bhosley, 1991). Psychologically abused women hasbeen shown to have lower self-efficacy scores, higher depression scores, lower self-esteem scores, and increased risk ofposttraumatic stress disorders than non-abused women (Ovara, McLeod, & Sharpe, 1996; Street & Arias, 2001). In one studyof college women, being psychological abused by partners was predictive of illegal drug use, negative health perceptions,and cognitive impairment even after controlling for physical violence victimization (Straight, Harper, & Arias, 2003). Femalevictims of partner psychological aggression were also more likely than others to report physical health problems suchas arthritis, chronic pain, migraine, sexually transmitted disease, and stomach ulcers, after the effect of partner physicalviolence were adjusted in the data (Coker, Smith, Bethee, King, & McKeown, 2000; Street & Arias, 2001). Past research haddemonstrated that partner psychological aggression often predicted later physical aggression in an intimate relationship(Murphy & O’Leary, 1989; O’Leary, 1999).

Partner psychological abuse is not only deleterious to the health and well-being of adult victims, it also has implicationsfor the development of the children living in the same households. Population data on the average number of children perhousehold indicated that over half of female domestic violence victims lived in households with children under the age of12 (U.S. Department of Justice, 1998). According to family systems theory, negative affect—facial or behavioral expressionswhich appear angry, unhappy or disturbed—in one family relationship can spread to other family relationships (Margolin& Gordis, 2003). It had been postulated that abusive parents had more unrealistic and rigid expectations for children, andthat they perceived their children’s behavior more negatively than do outside observers (Slep & O’Leary, 2001). Often time,these parents attributed hostile intent to their children’s behavior, and easily became angered which might lead to coerciveparenting styles or abuse of their children (Slep & O’Leary, 2001). In addition, children could be harmed indirectly via thestress in the home and parents’ compromised ability to provide care (Dubowitz et al., 2001). In one study, Margolin andGordis (2003) indicated that the combination of financial and parenting stress potentiated the association between maritalaggression and abusive parenting behaviors. Past research had also suggested that exposure to negative forms of maritalconflict was linked with children’s aggressiveness, emotional distress, and risk of adjustment problems (Davies & Cummings,1998).

Co-occurrence of intimate partner violence and child maltreatment had been widely documented. The averageco-occurrence rate for intimate partner violence and physical child abuse was estimated to be about 40% (for com-plete review see Appel & Holden, 1998). However, research examining the impact of domestic violence on childoutcomes had mostly focused on partner physical violence victimization, ignoring the potential detrimental effectsof partner psychological abuse on their children (Cox, Kotch, & Everson, 2003; Dubowitz et al., 2001; McGuigan &Pratt, 2001; McGuigan, Vuchinich, & Pratt, 2000; Ross, 1996). In addition, previous studies on the impact of part-ner psychological abuse has primarily focused on perpetration by men in relationships, ignoring perpetration bywomen (Hines & Malley-Morrison, 2001) Psychological abuse has been shown to be equally or more prevalentamong women relative to men in dating and community samples (Hines & Saudino, 2003; Magdol, Moffitt, & Caspi,1997).

Prior studies were also limited by either only focusing on child physical abuse, or not specifying if the child maltreatmentwas sexual, physical, emotional, or neglect (Edleson, 1999). Besides child physical abuse, it is important to study psychologicalchild abuse and neglect. Psychological child abuse refers to acts such as ridiculing a child, threatening to harm a child, andother emotional abuse (McGuigan & Pratt, 2001). Psychological child abuse may not cause immediate physical harm to thechild but may have long-term mental health consequences that are just as damaging as physical abuse or neglect (Egeland& Erickson, 1987; Hart & Brassard, 1987). Child neglect is another important type of child maltreatment that received littleattention in the child maltreatment research. Child neglect includes lack of supervision, medical neglect, failure to providefood or clothing, inadequate shelter, desertion, abandonment, and other physical neglect (McGuigan & Pratt, 2001). Childneglect is the most frequently occurring type of child maltreatment (Department of Health and Human Services, 1988).Nevertheless, few past studies have examined the association between intimate partner violence and the risk of child neglect(Hartley, 2002).

Given the potential relationships between adult victims’ psychological adjustment and their parenting and, subsequently,their children’s adjustment, it is important to examine the impact of partner psychological abuse on children. Despite the evi-dence of the potential harm associated with partner psychological aggression, research in this area remains sparse. Findingsfrom existing studies on co-occurrence of intimate partner violence and child maltreatment had limited generalizability byusing convenience samples drawn from existing child protection or hospital case records which likely represent only thosemost extreme cases of reported child maltreatment (Edleson, 1999). Similarly, using battered women sample from sheltersmay only account for segment of the population of battered women and most likely represent a group of women that lacksufficient alternative safety resources and thus turn to a shelter for support (Edleson, 1999). More representative samplesare needed to draw stronger conclusions about the co-occurrence of intimate partner violence and child maltreatment risk.

J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829 821

Other limitations of existing literature included not being specific about the type of adult abuse that was occurring, or whowas maltreating the child—whether it was by the child’s mother, father/father-figure, or both parents (Edleson, 1999).

This study aimed to address a gap in the existing literature of the co-occurrence of intimate partner violence and childmaltreatment. A population-based sample was used to examine the association between partner psychological abuse andthree types of child maltreatment risk (i.e. physical, psychological child abuses, and neglect). We hypothesized that presenceof partner psychological abuse between parents increased the likelihood of child maltreatment.

Methods

Study sample

Data for this analysis were derived from the Carolina Safe Study, conducted from June to December of 2002 (Theodoreet al., 2005). This cross-sectional, anonymous telephone survey involved a probability sample of North and South Carolinamothers, aged 18 and older who were living with a child under the age of 18 years (n = 1,435). This survey was designed toassess the rates and correlates of child maltreatment. The present study only used data from mothers who reported livingwith a husband or partner at the time of interview (n = 1,232). Since the objective of the current study was to examine theimpact of partner psychological abuse on child maltreatment, the analysis excluded mothers who also reported being thevictim or perpetrator of physical or sexual intimate partner violence. With the exclusion of subjects with missing data onpartner psychological abuse (about 6.7% of the present study sample), it resulted in a final study sample of 1,149 mothers.More details regarding the design and implementation of the survey had been described elsewhere (Theodore et al., 2005).

Procedure

To enhance accuracy of the data, trained interviewers entered all interview responses directly into a computer using Blaise,a computer-assisted telephone interview (CATI) software package developed by Statistics Netherlands (Blaise for WindowsStatistical Software, 2002). Eligible study participants were informed that the interviews were completely anonymous with-out the phone numbers being available even to the interviewer once the survey began with the computer assisted interviewprogram purging the number dialed as soon as the interview process started. A dual-frame sampling approach was usedfor data collection. About half of the sample came from a purchased list of working numbers of households known to havechildren under the age of 18. This sampling frame was used to improve the “hit rate” of contacting households with childrenand lower survey costs. However, it emerged that relatively few poor households were included. The sample was augmentedby the addition of random digit dialing of numbers in order to include more poor households: 47% of the interviews camefrom a random digit-dialing (RDD) sample. To be eligible, a called number had to reach a household located in the Caroli-nas that had a child under the age of 18 and a mother or female guardian in the home. Once eligible, a referent child wasselected and the corresponding mother or female guardian was interviewed. If the household had more than one eligiblechild, one was randomly selected by a computer-generated algorithm. A total number of 8,262 numbers were used. Of these,4,611 numbers were ineligible (i.e. numbers not in service, business, etc.), 1,335 were of unknown eligibility (maximumcall attempts resulted in no answer), and 881 were eligible but did not complete an interview (refused, unavailable, etc.). Acombined response rate of 52% was achieved if unknown eligibility numbers are assumed eligible, but an adjusted responserate of 62% was derived if unknown eligibility numbers were presumed ineligible following the computational guidelines ofthe American Association for Public Opinion Research (AAPOR) (AAPOR, 2005).

Assessment

Interviewed mothers responded to a survey that included questions related to parenting practices used by them or bytheir spouses/partners to correct or punish the referent child’s behavior. They were also asked to respond about both theviolent behaviors of their intimate partners that were directed towards them and their own violent behaviors towards theirintimate partners in the past year. The survey was translated into Spanish and administered in that language to familieswho only spoke Spanish. The survey was translated and back translated by independent review. Seven mothers completedthe survey in Spanish. The study design and procedure were reviewed and approved by the School of Medicine InstitutionalReview Board of the University of North Carolina at Chapel Hill.

Measures

Due to the constraint of survey cost and the length of the interview, an abbreviated form of parent-child Conflict TacticsScales (PC-CTS) was employed in this study to measure parental disciplinary practices including those that would qualifyas child maltreatment (Straus, Hamby, Finkelhor, Moore, & Runyan, 1998). The original instruments asking about familyviolence and discipline were abbreviated after a pilot process to limit the length of the survey to just 20 min. Some questionsabout intimate partner violence were adapted from the original Conflict Tactics Scales (CTS) to assess the use of reasoning,verbal aggression, and physical aggression in the intimate partner relationship (Straus, 1979). The items from CTS chosen to

822 J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829

be included in the survey were based on feedback from field tests with the criteria of the meaning and clarity of the questionsand the ease of participants’ understanding of the survey instrument.

As in the original CTS, female subjects were asked to think about the problems and conflicts with the spouse/partner orwith a selected child during the previous 12 months and then respond to questions on what she or her spouse/partner didin those situations. Responses for each question ranged from once to more than 20 times. Respondents also had “never” or“don’t know” as an answer choice.

Partner psychological abuseThe questions about intimate partner violence were administered to female caregivers only if they were living with a

spouse or partner at the time of interview. The time frame for recall was 1 year. Partner psychological abuse was assessedwith two questions: “How many times in the past year have you insulted, belittled or demeaned your husband/partner;” “Howmany times in the past year have you threatened your husband/partner or someone he cares about?” Female caregivers werealso asked if those behaviors had been directed toward them by their husbands/partners. The partner psychological abuseindicator was created and those who reported experiencing partner psychological abuse at least once in the past year weregrouped according to who perpetrate in the relationship. The partner psychological abuse indicator was thus categorized asno partner psychological abuse (i.e. none in the past year or never; reference category), husband/partner perpetrates, wifeperpetrates, or both perpetrate in the relationship.

Child physical maltreatmentTwelve questions about harsh physical discipline were used to assess the frequency of parent to child physical aggression

in the past year before interview. They included shaking a child younger than 3 years of age, beating, burning, kicking,slapping a child or hitting a child with an object (e.g., belt, hairbrush, stick, or some other hard object) somewhere otherthan the buttocks. For example, female caregivers were asked if they or their husbands/partners had “beat [the child], thatis, hit over and over again with an object or fist.” A child physical abuse indicator was created to specify perpetration ofharsh physical discipline by either or both parents. As these 12 questions assessed harsh physical discipline, presence of anyof those physical disciplines warrants a concern. Therefore, reports of at least once or more in the past year in any of the12 questions indicates presence of child physical maltreatment. The indicator had four categories: no child physical abuse(reference), mother perpetrates, father/father-figure perpetrates, or both parents perpetrate.

Child verbal maltreatmentFemale caregivers reported parent to child verbal aggressive behaviors used by themselves and by their husbands/partners

toward the reference child in the last 12 months before the interview. Behaviors that we regarded as verbally abusive includedthreatening a child with a weapon, threatening to kick a child out of the house, threatening to leave or abandon a child,threatening a child age less than 9 with a boogieman (more than or equal to 3 times), cursing or swearing at a child (morethan or equal to 3 times), and calling the child a name (e.g., stupid, useless, ugly) (more than or equal to 3 times). A childpsychological abuse indicator was created based on 12 questions to specify perpetration by either or both parent. Presenceof child psychological abuse was indicated if there were reports of at least once or three times (as specified above) occurredin the past year in any one of the 12 parent-to-child verbal aggression assessed. The child psychological abuse indicatoralso had four categories: no abuse (reference category), mother perpetrates, father/father-figure perpetrates, or both parentsperpetrate.

Child neglectFour questions were used to construct a child neglect indicator. A child was defined to be “neglected” if the female

caregiver reported that there was a lack of adequate food or an injury “due to lack of supervision” occurred 3 or more timesin the last month; if a child less than or equal to 6 years old was left home alone 1 or more times in the last month; or if achild was unable to access medical care when he or she needed it, even once in a year. These questions pertained only tothings that happened to an interviewed female caregiver in caring for the reference child. An overall measure of child neglectwas obtained by a simple unduplicated count of all children who were neglected in any one of the domains listed above.

Sociodemographic characteristicsInformation was gathered from female caregivers on a range of sociodemographic characteristics including child’s age

at the time of the interview, child’s sex, race/ethnicity, mother’s education level, and Medicaid receipt status. Child’s sexand Medicaid use were coded as binary variables, male or female for child’s sex, yes or no for Medicaid use. Child’s ageand mother’s education were continuous variables. Race/ethnicity includes non-Hispanic Whites, non-Hispanic Blacks, andothers.

Data analysis

Sample weights were assigned to each subject in the study. The sample weights adjusted for the disproportionality in thesample arising from the sampling process and the sample imbalance on demographic factors such as age, race, and income(Theodore et al., 2005). Analysis was based on weighted data that accounted for the variation in selection probability to

J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829 823

Table 1Participant characteristics for North and South Carolina (n = 1149)

Characteristics Total (n = 1149) North Carolina (n = 574) South Carolina (n = 575)

n Crude % Crude % Crude %

Child ageMean (S.D.) 1146 8.8 (5.1) 9.1 (5.2) 8.5 (5.1)

Child sexMale 592 52 50 53

Respondent’s relationship to childBiological mother 1094 95 95 96Adopted/step mother 42 4 4 3Foster mother/other 12 1 1 1

Respondent’s highest grade completedLess than high school 36 3 3 4High school 284 25 25 25Some college 340 30 30 29College grad or beyond 489 43 43 42

Racial/ethnic backgroundCaucasian/white 988 87 88 85African-American/black 101 9 8 10Hispanic 17 1 2 1Native American/Indian 9 1 1 1Asian/Pacific-islander 7 1 1 1Mixed 15 1 1 2Other 4 0.4 0.5 0.2

Family income/year≤$20,000 59 5 5 6$20,001–$40,000 200 19 17 20>$40,000 818 76 77 75

Type of householdMarried with two parents 1106 96 96 97Unmarried partners 30 3 3 2Single parent 13 1 1 1Other

Public assistanceMedicaid 136 12 9 14WIC 68 6 5 7

derive population estimates. Descriptive and bivariate analyses examined the prevalence and association of psychologicalabuse and child maltreatment perpetration. Chi-Square (�2) test for categorical variables was used to examine the associationbetween psychological abuse and child maltreatment outcome measures.

Two multivariate multinomial logistic regression models were obtained separately for the risk of child physical abuseand psychological abuse. Multivariate logistic regression was also used to estimate the odds of child neglect if psychologicalabuse was present in the household. To reduce the bias in the estimation of risk, potential confounders were included inthe multivariate analysis, including child’s age, child’s sex, race/ethnicity, mother’s education level, and Medicaid use. Eachmultivariate multinomial logistic regression model estimated the likelihood—in terms of a relative risk ratio (RRR), whichresembles the odds ratios (ORs) given by logistic regression—of a child experiencing maltreatment by mother, father/father-figure or both parents (versus no abuse) when either or both parents experienced partner psychological abuse relative tono psychological abuse between parents. RRRs, ORs, and 95% confidence intervals (CIs) were calculated from the regressionparameters. The survey routines in STATA software were used in the data analyses to account for the survey’s nested samplingdesign (i.e. dual-frame sampling within each state) (Stata Corporation, 2003).

Results

The characteristics of the study participants were presented in Table 1. A total of 1,149 women were included in thisanalysis (574 from North Carolina and 575 from South Carolina). The mean age of the women’s children was approx-imately 9 years in both states. Boys and girls were fairly equally represented among the women’s children. The vastmajority of the respondents were the biological mothers of the referent child. Over 50% of the mothers interviewed hadachieved some college level education or beyond. Overall, most of the women were Caucasian (87%), with the next largestproportion being African-American (9%). About 76% of the mothers reported an annual household income of more than$40,000. Nearly the whole study sample (96%) were married women. About 12% of the study participants received assis-tance or services from Medicaid and about 6% are WIC beneficiaries. Chi-square tests show that there are no statistically

824 J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829

Table 2Prevalence of partner psychological abuse and child maltreatment (n = 1149)

Type of violence n Crude % Weighted %

Partner psychological abuseNone 636 55.4 55.5Only husband/partner perpetrates 109 9.5 7.6Only wife Perpetrates 39 3.4 3.6Both partners perpetrate 365 31.8 33.3

Child psychological abuseNone 960 83.6 85.2Only fathera perpetrates 32 2.8 2.0Only mother perpetrates 49 4.3 4.8Both parents perpetrate 108 9.4 8.0

Child physical abuseNone 1040 90.5 90.5Only father perpetrates 21 1.8 2.2Only mother perpetrates 59 5.1 4.7Both parents perpetrate 29 2.5 2.6

Child neglectNone 1077 94.2 92.4neglected 66 5.8 7.6

Note: n is the actual sample size for the combined states of North and South Carolina.a Father is father or father-figure.

significant differences among the study participants between the two states in most of the demographic characteristics(except for Medicaid use, �2 = 7.45, p = 0.006). Therefore, data from the two states were combined for further analy-sis.

In the weighted analysis, Table 2 shows that more than half of the families did not experience any partner psychologicalabuse. In families where partner psychological abuse was present, most women reported perpetration of psychologicalaggression being bi-directional (33%). When only one adult exhibits partner psychological aggression in the relationship, thehusband/partner was the aggressor twice as often as the wife (7.6% vs. 3.6%).

Regarding the prevalence of child maltreatment, the majority of families did not report neglectful or abusive behaviorsdirected toward their children involving our measures of psychological (85%) or physical (90%) aggression. Nonetheless, somemaltreatment by parents was suggested, with child psychological abuse being the most common type. In families wherethere was child psychological abuse, most often both parents were verbally aggressive toward the child. If only one parentpsychologically abused the child, mothers were twice as often the perpetrator compared to fathers (4.8% vs. 2%, respectively).For child physical abuse, about 2.6% of children were physically abused by both parents. If only one parent physically abusedthe child, mothers were almost twice as often the perpetrator compared to fathers (4.7% vs. 2.2%, respectively). And about 7.6%of children experienced child neglect as we measured it by parent report. Table 3 detailed specific acts of child psychologicalabuse, physical abuse, and neglect included in each type of child maltreatment for this study. Most acts of child maltreatmentwere relatively rare based on female caregivers’ self-report except for curing or swearing at a child more than two times inthe past year (10.6%).

Table 4 showed the weighted prevalence of co-occurrence of partner psychological abuse and child maltreatmentperpetration. Psychological aggression between parents was statistically significantly associated with perpetrating aggres-sion toward the child (child physical abuse: �2 = 71.2, p < 0.0001; child psychological abuse: �2 = 118.5, p < 0.0001; childneglect: �2 = 28.9, p = 0.015). In families without any partner psychological abuse, about 5.2% (=0.9% + 0.5% + 3.8%) of chil-dren were psychologically abused. But in families where partner psychological abuse was present, children were athigher risk of being maltreated. Specifically, if only the husband/partner was the psychological abuse aggressor towardthe wife, the child would be psychologically abused about 16% of the time, most often by both parents (8.1%) followedby fathers/father-figures perpetrating alone (5.2%), and mothers perpetrating alone (2.2%). If only the wife psycholog-ically abused her husband/partner, her child would be psychologically abused about 25% of the time, most frequentlyby both parents (13.7%) followed by her perpetrating alone (8.8%), and by her husband/partner alone (2.9%). If bothparents psychologically abused each other, the child would be psychologically abused 28% of times, most frequentlyperpetrated by both parents (13.5%) followed by mothers perpetrating alone (10.7%), and fathers perpetrating alone(4.2%).

Similar findings were observed for child physical abuse. The greatest number of children experienced physical abusewhen both parents psychologically abused each other (18.9%), followed by the wife as the psychological abuse perpetratoralone (12.3%), husband/partner as the perpetrator alone (5.8%), and neither perpetrated (3.7%). When partner psychologicalabuse and child physical abuse co-occurred, children were most frequently abused by both parents regardless of whom thepartner psychological abuse perpetrator was in the families. Mothers physically abused their children more frequently thanfathers when only she perpetrated toward the husband/partner or when both parents were psychological abuse aggressors.Regarding child neglect, it occurred the most frequently when only the husband/partner was psychologically aggressive

J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829 825

Table 3Prevalence of maternal self-reported child abuse and neglectful behaviors (n = 1149)

Variables n Crude % Weighted %

Child psychological abuseThreaten to leave or abandon child 32 3.1 3.3Threaten child with devil or other scary thingsa,b 8 0.7 0.8Curse or sworn at childb 141 12.3 10.6Threaten to kick child out of the house 40 3.8 4.1Call child names like stupid, ugly, or uselessb 26 2.3 1.9Threaten child with a knife or gun 0 0 0

Child physical abuseShake a child younger than 3 years old 3 0.3 0.5Hit child with a hard object (not on buttocks) 46 4.0 3.2Kicked child 3 0.3 0.3Slapped child on the face or the back of the head 60 5.2 5.3Burned, scalded or braded child 2 0.2 0.2Beat child with an object or fist 1 0.1 0.4

Child neglectc

Not able to provide child with food needed more than 3–5 times 2 0.2 0.1Left child younger than 7 years old home alone for more than 1 h without any adult supervision 2 0.5 0.1Child got hurt more often than 3–5 times due to lack of careful adult supervision 7 0.6 0.7Unable to make sure child saw a doctor when needed to in the past year 55 4.8 6.7

a Only to children younger than 9 years old.b Occurred more than 2 times in the last year.c Occurred in the past month.

toward his wife (28%), followed by when both parents perpetrate toward each other (8.2%), and only the wife perpetratestoward her husband/partner (2.6%).

The probability of child maltreatment in families where partner psychological abuse was present was further investigatedusing multinomial and binary logistic regression, adjusting for child’s age, child’s sex, race/ethnicity, mother’s educationlevel, and Medicaid receipt status. Table 5 shows the adjusted relative risk ratios (aRRR) and odds ratios with 95% confi-dence intervals from the multivariate analysis. For child psychological abuse, after controlling for the covariates, the oddsof child victimization were the greatest when the father/father-figure was the only perpetrator toward his wife and child(aRRR = 16.55, 95% CI: 3.50, 78.19). Children were also at significantly increased risk of being psychologically abused whenboth parents experienced psychological abuse themselves. For instance, in families where both parents experienced partnerpsychological abuse compared to no partner psychological abuse, the likelihood of a child being psychologically abused bythe mother alone was 16 times as much as families without child psychological abuse (95% CI: 5.11, 50.92). The multivariateanalyses results also indicated that child age was a significant predictor of child psychological abuse. The odds of a childbeing psychologically abused by the mother alone or both parents increased by more than 10% with each unit of increase inchild age after controlling for the other covariates. With each additional year of education mothers received, the likelihood ofchild psychological abuse by father also increased by 36% (95% CI: 1.02, 1.81), relative to families without child psychologicalabuse.

In terms of child physical abuse (Table 5), the likelihood of a child being physically abused by both parents versus nochild physical abuse at homes was the highest when both parents experienced psychological abuse (aRRR = 14.57, 95%CI: 3.85, 55.16) compared to families without partner psychological abuse, followed by wife as a partner psychologicalabuse perpetrator alone (aRRR = 13.81, 95% CI: 2.57, 74.33), and husband/partner as a partner psychological abuse per-petrator alone (aRRR = 7.08, 95% CI: 1.09, 46.20) after controlling for the covariates. The odds of a child being physicallyabused by the mother only, relative to no abuse, were the greatest when both parents experienced partner psychologi-cal abuse (aRRR = 7.06, 95% CI: 2.89, 17.25), followed by only wife perpetrating toward the husband/partner (aRRR = 3.72,95% CI: 1.10, 12.60) compared to no partner psychological abuse in the family. This study also found that being a maleversus a female reduced the likelihood of a child being physically abused by the mother by more than 60% (aRRR = 0.38,95% CI: 0.18, 0.77) compared to no child abuse in the family. On the contrary, the likelihood of a child being physicallyabused by the father/father-figure compared to no abuse was more than 6 times greater (95% CI: 1.11, 35.00) for malescompared to females. Receiving Medicaid welfare compared to non-Medicaid beneficiaries also increased the odds of achild’s being physically abused by both parents by nearly 4 folds (aRRR = 4.77, 95% CI:1.23, 18.54) relative to no abuse byparents.

When partner psychological abuse was present in families, it also increased the likelihood of child neglect. The odds ofchild neglect were more than 5 times greater when only the husband/partner psychologically abused his wife (95% CI: 1.36,20.62) versus families without partner psychological abuse. Other significant predictors of child neglect were Medicaid orwelfare use (aOR = 3.35, 95% CI: 1.49, 7.54) and higher levels of maternal education which was protective of child neglect(aOR = 0.87, 95% CI: 0.76, 0.99).

826J.J.Chang

etal./Child

Abuse

&N

eglect32

(2008)819–829

Table 4The prevalence of co-occurrence of partner psychological abuse and child maltreatment (n = 1149)

Partner psychological abuse Child psychological abuse Child physical abuse Child neglect (n = 66)

None (n = 960) Perpetrator None (n = 636) Perpetrator

Mother (n = 49) Father (n = 32) Both (n = 108) Mother (n = 39) Father (n = 109) Both (n = 365)

None 94.8 0.9 0.5 3.8 96.2 1.7 1.6 0.4 6.5Only husband/partner perpetrates 84.4 2.2 5.2 8.1 94.3 1.0 1.0 3.8 28.0Only wife perpetrates 74.6 8.8 2.9 13.7 87.7 6.4 1.4 4.5 2.6Both perpetrate 1.6 10.7 4.2 13.5 81.1 9.8 3.6 5.5 8.2

Note: data presented were all weighted percentages.

J.J.Changet

al./ChildA

buse&

Neglect

32(2008)

819–829827

Table 5Adjusted relative risk ratio (95% confidence intervals) from weighted multinomial logistic regression analysis for the association between partner psychological abuse and child maltreatment (n = 1149)

Variable Child psychological abuse, by perpetrator Child physical abuse, by perpetratora Child Neglecta‡

Mother vs. none Father vs. none Both vs. none Mother vs. none Father vs. none Both vs. none

Partner psychological abuseOnly husband/partner perpetrates 2.10 (0.19, 23.3) 16.55 (3.50, 78.19) 1.87 (0.48, 7.35) 0.57 (0.06, 4.98) 0.48 (0.04, 6.21) 7.08 (1.09, 46.20) 5.29 (1.36, 20.62)Only wife perpetrates 11.84 (3.09, 45.47) 8.00 (1.27, 50.29) 4.61 (1.79, 11.84) 3.72 (1.10, 12.60) 1.00 (0.14, 7.01) 13.81 (2.57, 74.33) 0.53 (0.17, 1.67)Both perpetratea 16.13 (5.11, 50.92) 11.69 (3.31, 41.23) 4.91 (2.50, 9.67) 7.06 (2.89, 17.25) 2.72 (0.73, 10.17) 14.57 (3.85, 55.16) 1.33 (0.59, 3.00)None 1.00 1.00 1.00 1.00 1.00 1.00 1.00

Child age 1.17 (1.06, 1.29) 1.06 (0.98, 1.15) 1.12 (1.06, 1.18) 1.04 (0.98, 1.11) 1.07 (0.99, 1.15) 1.05 (0.96, 1.13) 0.98 (0.91, 1.05)

Child sexMale 0.83 (0.36, 1.90) 1.84 (0.67, 5.05) 1.52 (0.83, 2.80) 0.38 (0.18, 0.77) 6.24 (1.11, 35.00) 1.27 (0.38, 4.26) 1.43 (0.70, 2.90)Female 1.00 1.00 1.00 1.00 1.00 1.00 1.00

Medicaid welfareYes 2.51 (0.67, 9.41) 1.44 (0.29, 7.08) 1.37 (0.60, 3.15) 0.72 (0.29, 1.77) 0.72 (0.18, 2.84) 4.77 (1.23, 18.54) 3.35 (1.49, 7.54)No 1.00 1.00 1.00 1.00 1.00 1.00 1.00

Maternal education 1.18 (0.97, 1.43) 1.36 (1.02, 1.81) 0.99 (0.86, 1.13) 0.95 (0.74, 1.21) 0.88 (0.69, 1.12) 1.07 (0.87, 1.31) 0.87 (0.76, 0.99)

For convenience, cells with statistically significant adjusted relative risk ratio or odds ratio are shown in bold.a Odds of child neglect are represented by adjusted odds ratios with 95% confidence interval from binary logistic regression model.

828 J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829

Discussion

Our study showed that partner psychological aggression in families was common. Forty five percent of women reportedoccurrence of once ore more of psychological aggression in their intimate relationships that include insulting, belit-tling, demeaning, or threatening acts in the previous 12 months before the interview. Consistent with existing literature,women were found twice as often as men to psychologically or physically abuse their children. This may be attributedto the fact that in the U.S. women are overwhelmingly the primary child caregiver. If mothers are the primary care-givers of children, they will have a much greater opportunity to exhibit abusive behavior toward their children (Appel& Holden, 1998). To the authors’ knowledge, this is the first study that examined co-occurrence of partner psychologi-cal abuse and likelihood of three different types of child maltreatment, specifying which parent instigated the parentaland child abuse. The results of this study indicated that the presence of partner psychological abuse alone in familieswithout any physical or sexual intimate partner violence also significantly increased a child’s risk of being maltreated.It was observed that the patterns of child maltreatment varied depending upon which parent was the psychologicalaggressor in the intimate relationship. Children in families with partner psychological abuse were most likely to expe-rience child psychological abuse when both parents perpetrated toward each other and child physical abuse was mostlikely to occur when fathers/father-figures psychologically abused mothers in families. In addition, children in familieswith partner psychological abuse were most likely to experience child neglect when only the husband/partner psycho-logically abused his wife/partner. In families where intimate partner aggression was present, the frustration of an abusiveintimate relationship combined with the stresses of parenthood might set the stage for child maltreatment (McKay, 1994).An abused parent could be so fearful of their partner’s response to the children’s behaviors that they might over-disciplinethem in an effort to control their behavior and protect them from what they perceived as even greater abuse (McKay,1994).

The findings of this study needed to be interpreted with caution. This study was based on only one informant report-ing the occurrence of spouse and child abuse. Without multiple sources of reports, it was not possible to assess thereliability or validity of the reports. This analysis was also based on self-report data of female caregivers with onlylive-in spouses/partners. Past studies have shown that men and women differ in their levels of reporting of violence(Edleson & Brygger, 1986; Szinovacz, 1983). However, it had been suggested that women’s reports of nonphysical, aswell as physical, abuse to be more valid than men’s reports (Edleson & Brygger, 1986). The study findings were alsolimited by the small sample size that resulted in imprecision for some point estimates. There were other unmea-sured factors such as maternal psychiatric disorders that could potentially confound the observed association betweenpartner psychological abuse and child maltreatment. There are strengths of this study that are worth noting. Thispopulation-based study sample was more representative of battered women and maltreated children than were sam-ples derived primarily from shelter population or social services agency reports which may represent only the mostextreme cases. In addition, unlike most existing studies, this study specified the types of child maltreatment and madethe differentiation of husbands/partners’ psychological aggression toward wives and wives’ aggression toward hus-bands.

Implications

Psychological intimate partner aggression is devastating to both adult victims and their children. The findings from thisstudy illuminated the adverse effects on children when partner psychological abuse occurs in families. Increased publicawareness of the consequences of partner psychological aggression is needed. Both domestic violence agencies and thechild welfare system must recognize the link between partner psychological abuse and child maltreatment. Both agenciesshould work together closely to develop more effective screening mechanisms for each of these problems. Whenever anindividual is identified as a victim of spouse abuse, his/her children should be considered at risk for child abuse (Campbell,1994). Conversely, when a child is reported for abuse, there should be an automatic assessment about the presence of intimatepartner violence in the family (Campbell, 1994). If a spouse were abusing both his/her partner and their child, the interventionshould address both forms of abuse. Intervention focusing solely on child abuse might be insufficient if, for instance, mother-to-child abuse were occurring secondary to husband-to-wife abuse (Slep & O’Leary, 2001). Conventional child maltreatmentparent training intervention focuses mostly on addressing parenting difficulties by coaching and practicing positive parentingskills in dyadic parent-child sessions (Harwood & Eyberg, 2006). Findings of the co-occurrence of intimate partner violenceand child maltreatment suggests that parent training intervention program should also address the problems faced by parentsand families such as marital conflict and IPV.

Over the past decade, recommendations have been made and sometimes mandated for healthcare providers to routinelyscreen for domestic violence (Dubowitz & King, 1995). Even though there was insufficient evidence to the effective-ness of routine screening for IPV for improved outcome for women, there was obvious potential harm of doing nothing(Coker, 2006). Asking about IPV via routine screening could potentially change the social norm of the secrecy of IPVfor victims (Coker, 2006). IPV screening also creates opportunity for healthcare professional to provide informationabout the adverse effects of IPV, safety planning, and guided referral (Coker, 2006). Routine screening by healthcareproviders for domestic violence, hence, should be considered in addressing the level of partner and child abuse in U.S.homes.

J.J. Chang et al. / Child Abuse & Neglect 32 (2008) 819–829 829

References

American Association for Public Opinion Research (AAPOR). (2005). Standard definitions: Final dispositions of case codes and outcome rates for surveys (3rded.). Lenexa, KS: AAPOR.

Appel, A. E., & Holden, G. W. (1998). Spouse and physical child abuse: A review and appraisal. Journal of Family Psychology, 12(4), 578–599.Blaise for Windows Statistical Software. (2002). Voorburg/Heerlen, Netherlands: Statistics Netherlands.Campbell, J. C. (1994). Child abuse and wife abuse: The connections. Maryland Medical Journal, 43(4), 349–350.Coker, A. L. (2006). Preventing intimate partner violence: How we will rise to this challenge. American Journal of Preventive Medicine, 30(6), 528–529.Coker, A. L., Smith, P. H., Bethee, L., King, M. R., & McKeown, R. E. (2000). Physical health consequences of physical and psychological intimate partner abuse.

Archives of Family Medicine, 9, 451–457.Cox, C. E., Kotch, J. B., & Everson, M. D. (2003). A longitudinal study of modifying influences in the relationship between domestic violence and child

maltreatment. Journal of Family Medicine, 18(1), 5–17.Davies, P. T., & Cummings, E. M. (1998). Exploring children’s emotional security as a mediator of the link between marital relations and child adjustment.

Child Development, 69, 124–139.Department of Health and Human Services. (1988). Study findings: National study of the incidence and severity of child abuse and neglect (DHHS Publication

No. ADM 20-01099). Washington, DC: Government Printing Office.Dubowitz, H., Black, M. M., Kerr, M. A., Hussey, J. M., Morrel, T. M., Everson, M. D., & Starr, R. H. (2001). Type and timing of mothers’ victimization: Effects on

mothers and children. Pediatrics, 107(4), 728–735.Dubowitz, H., & King, H. (1995). Family violence: A child-centered, family-focused approach. Pediatric Clinics of North America, 42, 153–163.Edleson, J. L. (1999). The overlap between child maltreatment and woman battering. Violence Against Women, 5(2), 134–154.Edleson, J. L., & Brygger, M. P. (1986). Gender differences in reporting of battering incidents. Family Relations, 35, 377–382.Egeland, B., & Erickson, M. (1987). Psychologically unavailable caring. In M. R. Brassard, R. Germain, & S. N. Hart (Eds.), Psychological maltreatment of children

and youth (pp. 110–120). New York: Pergamon Press.Follingstad, D. R., Rutledge, L. L., Berg, B. J., Hause, E. S., & Polek, D. S. (1990). The role of emotional abuse in physically abusive relationships. Journal of Family

Violence, 5, 107–120.Hart, S. N., & Brassard, M. R. (1987). Psychological maltreatment: Integration and summary. In M. R. Brassard, R. Germain, & S. N. Hart (Eds.), Psychological

maltreatment of children and youth (pp. 254–266). New York: Pergamon Press.Hartley, C. C. (2002). The co-occurrence of child maltreatment and domestic violence: Examining both neglect and child physical abuse. Child Maltreatment,

7(4), 349–358.Harwood, M. D., & Eyberg, S. M. (2006). Child-directed interaction: Prediction of change in impaired mother–child functioning. Journal of Abnormal Child

Psychology, 34(3), 335–347.Hines, D. A., & Malley-Morrison, K. (2001). Psychological effects of partner abuse against men: A neglected research area. Psychology of Men and Masculinity,

2, 75–85.Hines, D. A., & Saudino, K. J. (2003). Gender differences in psychological, physical, and sexual aggression among college students using the Revised Conflict

Tactics Scales. Violence and Victims, 18, 197–214.Magdol, L., Moffitt, T. E., & Caspi, S. (1997). Gender differences in partner violence in a birth cohort of 21-year-olds: Bridging the gap between clinical and

epidemiological approaches. Journal of Consulting and Clinical Psychology, 65, 68–78.McGuigan, W. M., & Pratt, C. C. (2001). The predictive impact of domestic violence on three types of child maltreatment. Child Abuse & Neglect, 25, 869–883.McGuigan, W. M., Vuchinich, S., & Pratt, C. C. (2000). Domestic violence, parents’ view of their infant, and risk for child abuse. Journal of Family Psychology,

14(4), 613–624.McKay, M. M. (1994). The link between domestic violence and child abuse: Assessment and treatment considerations. Child Welfare, 73(1), 29–39.Margolin, G., & Gordis, E. B. (2003). Co-occurrence between marital aggression and parents’ child abuse potential: The impact of cumulative stress. Violence

and Victims, 18(3), 243–258.Murphy, C. M., & Cascardi, M. (1999). Psychological abuse in marriage and dating relationships. In R. L. Hampton (Ed.), Family violence prevention and

treatment (2nd ed., pp. 198–226). Beverly Hills, CA: Sage.Murphy, C. M., & O’Leary, K. D. (1989). Psychological aggression predicts physical aggression in early marriage. Journal of Consulting and Clinical Psychology,

57, 579–582.O’Leary, K. D. (1999). Psychological abuse: A variable deserving critical attention in domestic violence. Violence and Victims, 14(1), 3–23.Ovara, T. A., McLeod, P. J., & Sharpe, D. (1996). Perception of control, depressive symptomatology and self-esteem of women in transition from abusive

relationships. Journal of Family Violence, 11, 167–186.Ross, S. M. (1996). Risk of physical abuse to children of spouse abusing parents. Child Abuse & Neglect, 20(7), 589–598.Slep, A. M., & O’Leary, S. G. (2001). Examining partner and child abuse: Are we ready for a more integrated approach to family violence? Clinical Child and

Family Psychology Review, 4(2), 87–107.Stata Corporation. (2003). Stata Statistical Software. College Station, TX.Straight, E. S., Harper, F. W., & Arias, I. (2003). The impact of partner psychological abuse on health behaviors and health status in college women. Journal of

Interpersonal Violence, 19, 1035–1054.Straus, M. A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics (CT) Scales. Journal of Marriage and the Family, 41(1), 75–88.Straus, M. A., Hamby, S. L., Finkelhor, D., Moore, D. W., & Runyan, D. (1998). Identification of child maltreatment with the Parent–Child Conflict Tactics Scales:

Development and psychometric data for a national sample of American parents. Child Abuse & Neglect, 22, 249–270.Street, A. E., & Arias, I. (2001). Psychological abuse and postraumatic stress disorder in battered women: Examining the roles of shame and guilt. Violence

and Victims, 16(1), 65–78.Szinovacz, M. E. (1983). Using couple data as a methodological tool: The case of marital violence. Journal of Marriage and the Family, 45, 633–644.Theodore, A. D., Chang, J. J., Runyan, D. K., Hunter, W. M., Bangdiwala, S. I., & Egans, R. (2005). Measuring the physical and sexual maltreatment of children

in the Carolinas. Pediatrics, 115, e331–e337.Tolman, R. M., & Bhosley, G. (1991). The outcome of participation in shelter-sponsored program for men who batter. In D. Knudsen & J. Miller (Eds.), Abused

and battered: Social and legal responses to family violence (pp. 113–122). Hawthorne, NY: Aldine de Gruyter.U.S. Department of Justice. (1998). Violence by intimates (NCJ–167237). Washington, DC: Bureau of Justice Statistics.