The predictive impact of domestic violence on three types of child maltreatment

Abstract

Objective: This exploratory study investigated the co-occurrence of domestic violence and three types of child maltreatment: physical child abuse, psychological child abuse, and child neglect.

Method: A sample of 2544 at-risk mothers with first-born children participated in a home-visiting child abuse prevention program. A longitudinal design using multiple data collection methods investigated the effect of domestic violence during the first 6 months of child rearing on confirmed physical child abuse, psychological child abuse, and child neglect up to the child’s first 5 years.

Results: Logistic regressions revealed significant relationships between domestic violence and physical child abuse, psychological child abuse, and child neglect. These effects were significant beyond the significant effects of known risks of maltreatment, as measured by the Kempe Family Stress Inventory (KFSI). Domestic violence occurred in 59 (38%) of the 155 cases of confirmed maltreatment. Domestic violence preceded child maltreatment in 46 (78%) of the 59 cases of co-occurrence, as indicated by independent home observations and child protective service records.

Conclusions: The findings indicate that domestic violence during the first 6 months of child rearing is significantly related to all three types of child maltreatment up to the child’s fifth year. Domestic violence and risks factors measured on the KFSI continue to contribute to all three types of maltreatment up to the child’s fifth year. Prevention programs would be wise to provide services to at-risk families until the child is at least 5 years old. Addressing concurrent problems during treatment may enhance intervention.

Résumé

Objectif: Cette étude exploratoire porta sur la présence simultanée de la violence conjugale et de trois types de mauvais traitements—physiques, psychologiques et négligence.

Méthodes: Un échantillon de 2.544 femmes à risque élevé, mères d’un premier-né, ont participé à un program de visites à domicile conçu pour prévenir les mauvais traitements. Les chercheurs ont élaboré une évaluation longitudinale basée sur la collecte de plusieurs types de données, afin d’étudier les effets de la violence entre conjoints sur les trois types de mauvais traitements que leur enfant a vécu durant les six premiers mois de sa vie, allant jusqu’à l’âge de 5 ans.

Résultats: On note des liens importants entre la violence des conjoints et les trois types de mauvais traitements des enfants. Ces liens sont plus remarquables que ceux que démontre le Kempe Family Stress Inventory (KFSI). On note la violence entre conjoints dans 59 (38%) des 155 cas confirmés de mauvais traitements. Selon des observations indépendantes et les dossiers de la protection de l’enfance, cette violence conjugale précède les mauvais traitements dans 46 (78%) des 59 cas où coexistent les deux grands types de problèmes.

Conclusions: Les constats indiquent qu’il existe un lien important entre la violence conjugale et les mauvais traitements des enfants durant les six premiers mois de leur vie et jusqu’à l’âge de cinq ans. Les effets de la violence conjugale et des facteurs de risque tels que mesurés par le KFSI persistent jusqu’à l’âge de cinq ans. Il serait avantageux que les program de prévention qui desservent les familles à risque élevé offrent leurs services au minimum jusqu’à ce que l’enfant ait atteint ses cinq ans. L’intervention thérapeutique sera améliorée dans la mesure où elle prendra compte des problèmes résultant de la coexistence des deux grands types de problèmes.

Resumen

Objetivo: Este estudio exploratorio investigó la co-ocurrencia de la violencia doméstica y tres tipos de maltrato infantil: maltrato fı́sico, maltrato psicológico y negligencia infantil.

Método: Una muestra de 2,544 madres en riesgo con niños primerizos participaron en un programa de visitas domiciliarias de prevención de maltrato infantil. Con un diseño longitudinal y utilizando múltiples métodos de recogida de datos, se investigó el efecto de la violencia doméstica durante los primeros 6 meses de la crianza del niño en casos ratificados de maltrato fı́sico, y maltrato psicológico y en casos de negligencia infantil hasta los primeros 5 años del niño.

Resultados: Las regresiones logı́sticas revelaron relaciones significativas entre la violencia familiar y el maltrato fı́sico, el maltrato psicológico y la negligencia infantil. Estos efectos fueron significativos más allá de los efectos significativos de los riesgos de maltrato conocidos, medidos con el Kempe Family Stress Inventory (KFSI). La violencia doméstica ocurrió en 59 (38%) de los 155 casos confirmados de maltrato. La violencia doméstica precedió al maltrato infantil en 46 (78%) de los 59 casos de co-ocurrencia, como se indicó en las observaciones independientes en el domicilio y los registros de los servicios de protección infantil.

Conclusiones: Los resultados indican que la violencia doméstica durante los primeros 6 meses de la crianza infantil está significativamente relacionada con los tres tipos de maltrato infantil hasta los 5 años del niño/a. La violencia doméstica y los factores de riesgo medidos con el KFSI ejercieron su efecto en los tres tipos de maltrato infantil hasta los 5 años del niño. Los programas de prevención deberı́an de ser prudentes en dar servicios a familias en riesgo hasta que el niño tenga al menos 5 años. Identificar problemas co-ocurrentes durante el tratamiento puede mejorar la intervención.

Introduction

Two decades of empirical research provide overwhelming evidence that domestic violence and physical child abuse are linked within families. The average co-occurrence rate for violent families is estimated to be 40% (for complete review see Appel & Holden, 1998). Little is known, however, about the co-occurrence of domestic violence and other types of child maltreatment such as psychological abuse or neglect (Appel & Holden, 1998). Each type of child maltreatment is usually studied independently, and integrative research studies across child maltreatment types are rare.

To address this issue, this exploratory study empirically investigated the temporal relationship between domestic violence and three types of child maltreatment. Using a longitudinal design, this study examined the effect of domestic violence during the first 6 months of child rearing on subsequent physical child abuse, psychological child abuse, and child neglect during the child’s first 5 years.

Physical child abuse has been the predominant focus of co-occurrence studies. Definitions of abuse have ranged from mild acts to extreme acts of aggression. Data have been drawn from varied sources; both prevalence and incidence rates have been reported. Data derived from a sample of battered women revealed a 100% co-occurrence rate, based on lifetime prevalence rates (Kruttschnitt & Dorfeld, 1992). Data derived from official reports of physical child abuse demonstrated a 26% rate of co-occurrence based on 6-month incidence rates (Sternberg et al., 1993). Data from a nationally representative sample supported a co-occurrence rate of 5.6%, based on a lifetime referent period (Gelles & Straus, 1988).

These studies establish that domestic violence and physical child abuse have co-occurred in a variety of samples, using various measurement methods. Differences in study samples (shelter sample, national sample) and differences in the methods used to confirm physical child abuse and domestic violence (retrospective parental report vs. official confirmation) have combined to produce highly divergent rates of co-occurrence (5.6–100%). This variability in co-occurrence supports the need for more research.

Noticeably absent in co-occurrence research is studies of psychological child abuse. Psychological child abuse encompasses mental injury, ridiculing, threat of harm, emotional abuse, and emotional and mental neglect. These acts may not cause immediate physical harm but may cause long-term mental health problems that are just as damaging as physical abuse or neglect Egeland and Erickson 1987, Hart and Brassard 1987. According to a national incidence study psychological abuse accounts for 28% of all abuse cases, and is over five times more likely to occur in families making less than $15,000 per year compared to families with more adequate incomes (Department of Health and Human Services, 1988). Self-report surveys have shown the overall rate of psychological abuse to be as high as 63% (Vissing, Straus, Gelles & Harrop, 1991). Although psychological abuse can occur alone (Claussen & Crittenden, 1991), many researchers suggest that psychological abuse is a core component, and perhaps the most destructive factor, in all forms of child maltreatment Brassard and Gelardo 1987, Brassard et al 1987, Garbarino et al 1986, Navarre 1987.

Child neglect is another important type of child maltreatment. Child neglect includes lack of supervision, medical neglect, failure to provide food or clothing, inadequate shelter, desertion, abandonment, and other physical neglect. Regardless of its form, neglect traumatizes victims and effects their subsequent behavior. Child neglect is the most frequently confirmed type of maltreatment in the general population (US Department of Health and Human Services, 2000). Nevertheless, there is little empirical literature documenting the relationship between domestic violence and child neglect. Some research suggests that when domestic violence is present, parents may be emotionally numb and uncommunicative, and thus be less available to their children (Osofsky & Fenichel, 1994). Another study concurs that in domestic violence relationships, parents may give inordinate attention to one another and therefore ignore their children’s needs (McKay, 1994).

Past studies investigating co-occurrence have found primarily physical child abuse and have relied on clinical samples and cross sectional designs. To establish the temporal order of domestic violence and child maltreatment, studies have relied heavily upon single informants and retrospective reporting of lifetime prevalence rates. Across studies there are diverse, and frequently broad, definitions of constructs. These limitations have contributed to the disparity in reported rates of co-occurring domestic violence and physical child abuse, which range from 5.6% (Gelles & Straus, 1988) to 100% (Kruttschnitt & Dorfeld, 1992). Study samples have varied on key individual and family level variables that are associated with child maltreatment. These key variables include child age, family structure, and income. Few studies have controlled for multiple risk factors, despite research that has established the multiplicative nature of risks for child maltreatment Azar 1991, Belsky 1993.

To address these limitations, the present study used comprehensive risk assessment data to statistically control for multiple risk characteristics known to be associated with child maltreatment. To avoid single informant bias, this study used multiple data sources including interviews, in-home observations, birth records, and child protection agency records. Domestic violence was clearly defined as any behavior between adults with the intent to do physical harm, and incidents of child maltreatment were confirmed by the state’s child protection agency. Study families had similar demographic characteristics (e.g., age, ethnicity, income, education, and family structure), and were identified as being at-risk for poor child outcomes (including possible child maltreatment) at the time of their first child’s birth. Study families differed, however, on the occurrence of domestic violence. This difference allowed for meaningful investigation of the relationship between the presence of domestic violence and child maltreatment.

This study used a longitudinal design to assess the temporal order of domestic violence and the occurrence of each type of child maltreatment from the child’s birth up to 5 years of age. Children under 5 are disproportionately present in households experiencing domestic violence (Fantuzzo & Mohr, 1999). Children within this age group are also at the greatest risk for child maltreatment (Azar & Wolfe, 1989), and make up over 75% of maltreatment fatalities (US Department of Health and Human Services, 2000). Moreover, child maltreatment is usually detected before the child reaches age 5 (Leventhal, 1988).

The longitudinal design allowed for the testing of the following three hypotheses. In at-risk families, the occurrence of domestic violence within 6 months of the child’s birth would be significantly related to the confirmation of (1) physical child abuse, (2) psychological child abuse, and (3) child neglect, up to the child’s fifth year. Further, it was predicted these relationships would be significant above and beyond the effect of multiple risk factors related to child maltreatment.