Amer. J. Orthopsychiat., 63(1), January 1993, 92-101.

Mothers of Sexually Abused Children: Trauma and Repair in Longitudinal Perspective

Carolyn Moore Newberger, Ed.D., Isabelle M. Gremy, M.D., Christine M. Waternaux. Ph.D., Eli H. Newberger, M.D.

Mothers whose children had been sexually abused reported experiencing serious psychological symptoms following disclosure of the abuse. Over a one-year period, their emotional status improved. Strong relationships between mothers’ reports of their own and their children’s symptoms were accompanied by persistent discrepancies between maternal and direct assessments of the children’s emotional states. Findings suggest that addressing maternal distress is important to the study and treatment of child sexual abuse.

Clinicians working with sexually abused children have often noted that mothers are also distressed by their children’s victimization and its disclosure. The impact on mothers of their children’s trauma is acknowledged in the formulation of Post-Traumatic Stress Disorder (PTSD) in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association [APA], 1987). PTSD is described as a condition that results not only from a threat or harm to the self, but also from “…serious threat or harm to one’s children” (p. 247). Yet mothers of sexually abused children are frequently overlooked in the formulation of treatment plans.

Although a growing literature has documented the frequently severe and enduring effects of sexual abuse on children (Finkelhor & Browne. 1985; Gelinas, 1983; Gomes-Schwartz, Horowitz, & Sauzier, 1985; Herman, Russell, & Trocki, 1986), mothers’ emotional functioning following disclosures of their children’s sexual abuse is only beginning to be studied (DeJong, 1988: Krigbaum-Rich, 1991; Regehr, 1990; Wagner, 1991; Winton, 1990).

It is important to study mothers’ responses for several reasons. First, if mothers do, indeed, suffer significantly from their children’s disclosures, they should be acknowledged as victims and given appropriate psychiatric care. Second, because of the well-documented association between parental psychopathology and children’s mental health, it is possible that maternal distress may impede children’s recovery following disclosure (Billings & Moos. 1983; Griest, Forehand, Wells. & McMahon, 1980). Children of depressed mothers have been noted to demonstrate higher levels of psychological symptomatology than children in normative samples (Cox, Puckering, Pound, & Mills, 1987; Downey & Coyne, 1990). The recovery of sexually abused children may similarly be influenced by their mothers’ emotional responses (Conte, 1985, 1987; Newberger & De Vos, 1988).

Mothers’ emotional functioning following disclosure has implications for the clinical comprehension of the abused child’s experience. Much of the information in the literature about the effects of sexual abuse on children derives from maternal reports (Friedrich, Urquiza. & Beilke, 1986). The accuracy of mothers’ representations of their children’s more general psychological states, however, has been brought into question (Achenbach. McConaughy, & Howell, 1987; Griest et al., 1980; Schaefer, Hunter, & Edgerton, 1987). For example, there is considerable evidence that mothers who are depressed report higher symptoms of depression in their children than do mothers who are not (Fergusson, Horwood, Gretton, & Shannon, 1985; Kazdin, Esveldt-Dawson, Sherick, & Colbus, 1985; Kochanska, Radke-Yarrow, Kuczynski, & Friedman, 1987; Webster-Stratton & Hammond, 1985). However, whether this difference is due to distortions in maternal perception or to actual higher levels of depression in these children is unclear (Richters & Pellegrini, 1989).

The accuracy of mothers’ reports of the emotional status of their sexually abused children has also been questioned. In a recent study of maternal support following incest, mothers who did not believe incest had occurred appeared to be less accurate reporters of their children’s symptomatology than were mothers who gave credence to their children’s reports (Everson, Hunter, Runyon, Edelsohn, & Coulter, 1989).

The course of mothers’ psychological symptomatology over the year following disclosure of their children’s sexual abuse will be examined, as will relationships between mothers’ emotional well-being and their children’s emotional states.

METHOD

Subjects

Over a one-year interval, 44 mothers and two maternal caregivers (a custodial stepmother and a custodial grandmother) of sexually abused children age six through 12 were studied. All were living with the children at the time of abuse, and all retained custody of their abused children for the duration of the study. Of the 46 subjects, 42 were retained for follow-up to the end of the year.

The mean age of the mothers and caregivers at the time of the first interview was 33; 76% (N=35) were white, 17% (N=8) were African-American, and 7% (N= 3) were Hispanic. In the sample, social status (SES) was evenly distributed across levels II through V of Hollingshead’s (1979) four-factor index. Of the 46 children, 72% (N= 33) were girls and 28% (N=13) were boys, with a mean age of 8.5 years at the time of the initial interview. No relationships were found between ethnicity and social status, gender of the child, or age of the child; between SES and gender or age of the child; or between gender and age of the child.

All the abuse cases were substantiated by the Massachusetts Department of Social Services and independently confirmed by the child or the perpetrator, sometimes both. The abuse appeared to have been severe; the majority were reported to have experienced more than one type of sexual act, with oral-genital contact and anal or vaginal penetration occurring in 76% (N=35) of the cases.

Duration of the abuse ranged from a single incident to repeated abuse over five years; 61% (N=28) of the sample was abused more than once. The mean duration of abuse was seven months. Force or threat of force was used with the majority of children; 46% (N = 21) were physically overpowered, and 22% (N = 10) were threatened. Biological fathers or father-figures (stepfathers or mothers’ male partners) were responsible for the abuse of 28% (N=13) of the children.

Procedure

The children and their mothers were recruited from the Emergency Department of Children’s Hospital in Boston and from four prosecutors’ (district attorneys’) offices in the greater Boston area (in Massachusetts, it should be noted, the law requires that all substantiated cases of child sexual abuse be referred by the Department of Social Services to the local prosecutor). Letters introducing the study were sent to potential subjects, who were later contacted by telephone to request their participation. Of the 77 families contacted, 64% agreed to participate; three consenting families were excluded from this analysis. Analyses of anonymous background data collected on all children eligible for the study indicate that children whose parents did and did not participate were comparable in age, gender, race, and SES.

Children were eligible for inclusion in the study if they were 6 through 12 years of age, had suffered sexual abuse that had been substantiated by protective services, and had no major physical or mental disabilities.

A detailed consent form was reviewed on initial contact and signed in all cases by both the child and the child’s mother. The initial interviews were conducted within two to four months of the child’s disclosure in all but a few cases. The median time from disclosure to the initial interview was nine weeks.

Three interviews were administered: at recruitment, at six months, and at 12 months following the initial interview. Mothers and children were interviewed separately in their homes by two-woman teams composed of professionals with social work, psychology, or special education backgrounds who had been trained in the administration of the measures used in this study.

Measures

Demographic and victimization variables. Demographic variables included age and gender of each child, educational and occupational background of the parents, and ethnicity of the family.

Victimization information was collected from the mother using a detailed questionnaire designed for this study. Restrictions imposed by the cooperating district attorneys for prosecutory purposes prohibited the gathering of victimization data directly from the children.

Four aspects of the victimization were examined: severity, force, duration, and the identity of the perpetrator. On a severity scale designed specifically for this study, weights were assigned to each act by an expert panel of professionals in the field. Such acts as anal or vaginal intercourse were weighted more heavily than, for example, kissing or fondling. The weighting scores for each act reported by the mother were added together to yield a severity score. Force was classified as absent, threatened, or used. Duration was defined as the number of days between the first and the last abuse incident.

The relationship of the child to the perpetrator was recorded as intrafamilial or extrafamilial. Intrafamilial perpetrators included biological fathers; father figures, such as stepfathers or mothers’ boyfriends; uncles; cousins; and siblings. Extrafamilial perpetrators were known or unknown assailants who had no familial connection to the child. With the exception of one female babysitter, all perpetrators were male.

Assessment of mothers’ symptoms. Maternal symptomatology was assessed from the Brief Symptom Inventory (BSI) Derogatis & Spencer, 1982), a widely used 53-item questionnaire that explores the presence and severity of symptoms of psychopathology. The General Symptom Index (GSI), a summary scale that incorporates the number and severity of symptoms reported, was used for this analysis. Separate symptom dimensions were also measured: these included Somatization, Obsessive Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism. Norms are available for normal respondents, and for psychiatric inpatients and outpatients (Derogatis & Spencer, 1982). For this analysis, t-scores for normal respondents were employed.

Extensive work has been reported on the validity and reliability of this measure. Test-retest and internal consistency reliability range from .80 to .90 on the GSI. The BSI has also been found to discriminate between clinical and nonclinical samples (Derogatis & Spencer, 1982).

Direct assessment of children’s symptoms. Children’s symptomatology was assessed through two child self-report measures: the Children’s Depression Inventory (CDI) (Kovacs, 1981) and the Revised Children’s Manifest Anxiety Scale (RCMAS) (Reynolds & Richmond, 1985).

The CDI is a 27-item self-report questionnaire. Each item contains three statements, such as: “I am sad once in a while.” “I am often sad.” “I am sad all the time.” The child is asked to select the sentence from each group that best describes his or her feelings during the previous two weeks. Statements are assigned a numerical value from 0 to 2. The higher the numerical value, the more clinically severe the behavior being rated. The depression score is the sum of the values assigned to the statements selected.

The RCMAS is a 37-item self-report questionnaire. The child answers yes or no to statements such as: “I worry about what is going to happen.” The anxiety score is the sum of positive responses. The RCMAS has been standardized, taking into account age, gender, and race. It also contains a lie scale to identify children who overrespond positively to the questions. The standardized t-score was used for these analyses. The validity and reliability of both the CDI and the RCMAS have been extensively studied and demonstrated (Kazden, French, & Unis, 1983; Kazden & Peni, 1982; Kovacs, 1981, 1985; Pela & Reynolds, 1982; Reynolds, 1981; Reynolds & Richmond, 1985).

Maternal reports of children’s symptoms. Children’s symptomatology was also assessed from maternal report using the Child Behavior Checklist (CBCL). The CBCL is an extensively used 159-item assessment of behavior problems in children aged 4 through 16 (Achenbach & Edelbrock, 1983). Three general problem scales were used for this study: Total Behavior, Internalizing Behavior, and Externalizing Behavior. The Internalizing dimension includes items reflecting anxiety, depression, somatization, and social withdrawal. The Externalizing dimension includes items reflecting conduct problems such as aggressive, hyperactive, and delinquent behavior. The CBCL scales have been standardized, taking into account age and gender.

Although extensive research has established the validity and reliability of the CBCL (Achenbach & Edelbrock, 1983; Achenbach et al., 1987), the accuracy of parental report measures, including the CBCL, has recently been challenged. Low correlations have been found between child and parent reports and between parents’ reports and the reports of other professionals (Achenbach et al., 1987; Everson et al., 1989). This issue is addressed below. Stability of outcome measures.

To assess the stability of the outcome measures on this sample, test-retest reliability was assessed. Over a 12-month period, the stability of the GSI of the BSI was .73, stability of the CDI was .51, and stability of the RCMAS was .47. Test-retest reliability on the CBCL over a 12-month period was .68 for behavioral scores, .66 for internalizing scores, and .78 for externalizing scores. These are acceptable levels of stability and reliability. All measures were read to respondents to assure that findings were not distorted by their reading ability. Analyses Spearman correlations and their associated p-values were used to assess the associations of demographic, victimization, maternal treatment, and child outcome variables (CDI, RCMAS, CBCL scales) with maternal symptomatology scores.

Table 1 BSI SYMPTOM DIMENSIONS: MEANS, STANDARD DEVIATIONS. AND DIFFERENCES FROM NORMS
	WAVE 1 (N=46)		WAVE 3 (N-42)		PAIRED t-TESTa
SYMPTOM DIMENSIONS	M	SD	M	SD
GSI	59.2	13.0****	53.3	11.5*	– 3.896****
Somatization	55.5	12.3***	50.8	10.4	– 2.323**
Obsessive-Compulsive	55.9	11.7***	50.4	10.6	– 2.999***
Interpersonal Sensitivity	55.9	11.2****	51.5	9.4	– 2.585**
Depression	56.9	11.6****	52.0	9.7	– 2.299**
Anxiety	58.3	10.9****	56.0	9.7****	– 1.163
Hostility	58.9	12.7****	55.2	9.5****	– 1 .956*
Phobic Anxiety	57.3	11.1****	53.1	9.4**	– 3.175***
Paranoid Ideation	61.0	9.4****	57.0	9.3****	– 2.241**
Psychoticism	60.3	6.01****	54.9	10.4****	– 2.585**
______________________________________________________________________ aPaired t-test for differences between Wave 1 and Wave 3. *P <.10; **P <.05; ***P <.01: ****P <.001. ______________________________________________________________________

Paired sample t-tests were used to compare maternal symptomatology at the time of the initial interview with the 12-month follow-up; t-tests were employed to compare sample means with population norms. Changes over time were also analyzed utilizing scores from all three data points. For each mother, the slope of recovery was calculated by fining a regression line to her scores over time (OLS [ordinary least-squares] linear regression). If there were no change, the slope of the line would be zero; t-tests were used to determine whether the slopes differed from zero. Finally, mothers were partitioned into clinical and nonclinical groups based on their GSI scores. Correlation coefficients indicating the level of agreement between mothers’ and children’s reports of children’s symptomatology in the clinical and nonclinical groups were compared, employing the Fisher transformation of correlation coefficients and associated p-values.

RESULTS

Maternal Symptomatology

Initial psychological functioning.
At the time of the first interview, the mothers in this sample reported numerous and severe symptoms of emotional distress, summarized in Table I. As can be seen in the first column on the table, the mean GSI score was significantly above the mean for the normal population. Furthermore, 50% of the mothers’ GSI scores exceeded one standard deviation above the mean. In the general population, only 16% of respondents would be expected to have such scores. When standard criteria were applied to categorize the scores as clinical or nonclinical, 55% of the mothers’ scores placed within the clinical range.

When the nine specific symptom subscales were examined separately, no particular symptom pattern characterized this sample. Rather, scores were highly and consistently elevated across all the symptom domains. Different women appear to have experienced different types of symptoms, although they commonly reported feeling their symptoms intensely.

Over 12-month period. The course of the mothers’ psychological status was measured in two ways: by comparing GSI scores at the first and at the 12-month interviews; and by calculating the slopes of changes in scores, taking into account data from all three interviews. As can be seen in TABLE I, by the 12-month interview the mean GSI score had declined significantly in comparison to the initial scores. The mean recovery slope was – 0.08. The value of this slope differs significantly from zero (t= – 3.33,p<.02), providing further support for the general pattern of improvement over time.

Table 2 INTERCORRELATIONS AMONG MOTHER-REPORTED AND CHILD SELF-REPORTED OUTCOME VARIABLES: WAVE 1 (N=46)
	MATERNAL REPORT				RCMAS ANXIETY CHILD REPORT
		CBCL
SCALE	GSIa	TOTAL BEHAVIOR	INTERNAL BEHAVIOR	EXTERNAL BEHAVIOR
CBCL
Total Behavior Internal Behavior External Behavior	.57*** .56*** .45***	(N/A) (N/A)	(N/A)
RCMASb	.21	.24	.19	.18
CDIc	.09	.03	-.02	.10	.40**
aMaternal symptoms. bChild anxiety. cChild depression (self-report). * p <.05; **P <.01; ***P <.001.

Notwithstanding these symptom reductions over the year’s study interval, the mean GSI score at the third interview continued to be somewhat more elevated than normal. Additionally, at the 12-month interview a third of the mothers’ GSI scores exceeded one standard deviation above the mean, double the rate expected in the normal population. The scores of these women also continued to be in the clinical range.

When each symptom dimension was examined separately, a significant decrease was found on every scale except Anxiety (TABLE I, last column). On four of the scales (Somatization, Obsessive-Compulsive, Interpersonal Sensitivity, and Depression) symptom scores fell to within the normal range. Symptom levels on the other scales (Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism), although lower than at the initial interview, remained significantly higher than normal (TABLE 1, center column).

Relationship to demographic and victimization variables.
No correlations were found between mothers’ GSI scores and the race or age of the child. Mothers who were poor, however, were significantly more likely to report higher levels of psychological distress at the time of the first interview (r=.39, p<.OO7). The relationship between poverty and psychological symptoms was somewhat weaker (r=.27, p =.08) at the third interview. Although at the first interview there were no significant associations between gender of the child and maternal symptoms, at the third interview a modest relationship was found. Mothers of daughters reported more psychological symptoms than did mothers of sons (r= – .33, p<.05).

Relationship to child’s victimization.
Two variables related to the child’s victimization were associated with maternal symptomatology at the initial interview: the severity of the sexual abuse (r=.34, p<.03) and, less significantly, the perpetrator’s use of force (r=.26, p=.09).

Relationship to psychotherapy.
By the third interview, 72% (33) of the original cohort had received at least one type of treatment from a mental health professional, including individual (24), family (12), group (7), or couple therapy (7). Mothers’ GSI scores did not determine whether they received treatment during the 12 month duration of the study. For the 33 women in treatment, however, GSI scores were associated with how much treatment they received, since mothers reporting more psychological distress on the first interview were given more therapy over the one-year period of the study (r= .36, p<.02). The change over time in mothers’ GSI scores was not associated with either the number of weeks in therapy or the number of treatment contacts they received, but when the different treatment modalities were examined separately participation in family therapy was found to be related to a decline in mothers’ symptoms. Their recovery was related both to the number of family therapy sessions (r= -.35, p<.05) and to the length of time families were in treatment (r= -.35, p=.05). No particular associations were found for mothers who participated in individual, group, or couples treatment.

Mothers’ GSI scores at the initial interview were, in contrast, related to the total number of treatment contacts their children subsequently received (r= .30, p<.05), and their children’s treatment was associated with declining maternal GSI scores (r= – .35, p<.05). Mothers’ improvement with their children’s treatment remained significant even when controlling for their own therapy and for their initial psychological symptom scores.

Relationship to child symptomatology.
Mothers’ reports of their own emotional symptoms on the GSI were strongly and consistently associated with their reports of their children’s emotional states on the CBCL, as summarized in TABLES 2 and 3. Strikingly, mothers’ reports of their children’s symptoms were unrelated to their children’s self-reported symptoms of anxiety and depression. Data from the first interview (see TABLE 2) show highly significant correlations between GSI and CBCL scores. In contrast, virtually no relationships were found between mothers’ GSI scores and children’s self-reports of anxiety and depression. This pattern emerged more consistently by the 12-month interview (see TABLE 3). At this time, correlations between GSI and CBCL scores were higher than at the initial interview, while correlations between mothers’ CBCL scores and children’s reports of anxiety and depression were lower. Furthermore, no relationships were found at any time between mothers’ psychopathology as reported on the GSI and children’s self-reported anxiety and depression scores. To explore the hypothesis that mothers’ psychiatric symptoms distorted their perceptions of their children’s symptoms, the mothers were stratified into clinical and nonclinical groups according to the established criteria (Derogatis & Spencer, 1982). Associations between maternal and child reports of children’s symptomatology were examined separately for each group, and correlation coefficients were compared. Maternal symptomatology did not appear to influence agreement between mothers and their children. Agreement between mothers and children remained low regardless of the symptom levels mothers reported for themselves.

Table 3 INTERCORRELATIONS AMONG MOTHER-REPORTED AND CHILD SELF-REPORTED OUTCOME VARIABLES: WAVE 3 (N=42)
	MATERNAL REPORT				RCMAS ANXIETY CHILD REPORT
		CBCL
SCALE	GSIa	TOTAL BEHAVIOR	INTERNAL BEHAVIOR	EXTERNAL BEHAVIOR
CBCL
Total Behavior Internal Behavior External Behavior	.60*** .59*** .56***	(N/A) (N/A)	(N/A)
RCMASb	.22	.16	.07	.13
CDIc	.07	-.04	-.16	.15	.49**
aMaternal symptoms. bChild anxiety. cChild depression (self-report). * p <.05; **P <.01; ***P <.001.

DISCUSSION

The mothers who were the subjects of this study suffered severe and extensive emotional distress following disclosures of the sexual abuse of their children. Although it seems likely that the abuse was the major cause of their distress, it is not known to what extent these findings may reflect antecedent psychological burdens. Arguably, children of disturbed mothers may be at greater risk for sexual abuse.

The longitudinal data, however, suggest that preexisting pathology cannot fully explain these findings. Over the course of the one-year follow-up period, the mean maternal symptom scores declined almost to normal. This decrease suggests that the women’s psychological distress derived at least in part from their children’s traumatic experience. More hopefully, it also suggests that recovery is not uncommon.

But a one-year follow-up can only begin to chart the course of recovery. Twelve months after the first interview, OSI scores of over one-third of the mothers remained at clinically significant levels. Furthermore, when the symptom subscales were examined separately, several remained significantly higher than normal. It is possible that these symptom domains may be more sensitive to this particular experience. For example, hostility and paranoia could represent reactions to an erosion of trust following a violation of one’s child.

Of particular note is the inclusion among the symptom dimensions of a constellation of items consistent with the diagnostic criteria for Post-Traumatic Stress Disorder (APA. 1987. pp. 247-251). The Phobic Anxiety scale contains items about fears and avoided situations that may reflect an attempt to avoid stimuli associated with these mothers’ trauma. Symptoms of increased arousal in the Anxiety and Hostility scales include feelings of shakiness, spells of terror, irritability, and outbursts of anger. The Psychoticism scale contains items suggesting diminished responsiveness, such as an inability to get close to other people and persistent feelings of loneliness. Another study of mothers of sexually abused preschool children found similar results (Kelley. 1990). The strength of these inferences, however, may be somewhat diminished by the lack of a control group; these mothers may differ from the “normative’. population in other ways affecting their patterns of emotional symptomatology.

It is also possible that the protective and legal processes following disclosure exert their own traumatic impact on families and contribute to mothers’ enduring feelings of anxiety, mistrust, and hostility: the systems intended to protect children and punish offenders frequently continue to operate for a year or longer, and mothers’ actions and motives may receive intense scrutiny during this period.

Despite the enduring stresses they experienced following disclosure, a general pattern of decline was evident in the women’s symptoms of emotional distress. Family therapy, in particular, appeared to contribute to their recovery. Yet only 12 of the 46 women received family therapy, and 13 received no psychotherapy of any kind. A “regression to the mean” effect appears unlikely, since data was gathered at three points in time as a means of dealing with this possibility.

Although the children of more distressed mothers received more intensive therapy, whether the mothers received therapy was not related to their reported levels of distress. This raises questions about the ability of intervening professionals and agencies to identify and acknowledge the importance of women’s emotional needs and refer them for appropriate treatment. On the other hand, when women do receive treatment, therapists appear to be sensitive to their patients’ needs, as is suggested by the significant associations between the severity of maternal symptoms and the duration and number of treatment contacts.

In this sample, the mothers’ psychological distress strongly correlated with their reports of their children’s emotional states.

These findings are consistent with those of a study of maternal psychopathology in which the children were in treatment for psychiatric disturbance or for physical abuse or neglect (Estroff et al., 1984). Both groups showed higher levels of maternal symptomatology on the BSI and significantly higher correlations between BSI scores and CBCL scores than did mothers in a control group. But as the direct assessment of children’s symptoms was not available, the findings of Estroff and colleagues are susceptible to several competing interpretations. Highly symptomatic mothers may have similarly symptomatic children or may have difficulty perceiving their children objectively; or both may be true, as is suggested by another study assessing the influence of maternal depression on the validity of the CBCL (Friedlander, Weiss. & Traylor, 1986).

Because both child self-report and maternal measures were obtained for our study, mother-child agreement on the children’s emotional states could be directly ascertained; it was consistently poor.

These data suggest that when mothers are highly distressed, they may be so affected by their emotional pain that they have difficulty separating their own feelings from those of their children. This hypothesis is supported by an emerging literature examining the effects of mothers’ psychopathology and beliefs on the accuracy of their reports (Everson et al., 1989; Kochanska, 1990). However, it should be moderated by acknowledging that, in general, correlations of psychological information provided by two sources are weak. When mothers in this study were partitioned by their levels of symptomatology, agreement between mothers and their children did not improve, and lack of agreement remained equally strong on the 12-month interview, by which time the mothers’ symptom levels had generally declined. In their meta-analysis of 11 studies comparing parental report and child self-report, Achenbach and colleagues (1987) found a mean correlation of only .25. In the two studies where the only parental informant was the mother, the correlations were even lower.

Limitations of the validity of child self report measures, especially with regard to age, may also figure in the divergence. For example, Achenbach and his colleagues (1987) also noted that correlations between parental report and child self-report measures were stronger for children aged six through 11 than for those aged 12 through 19. In contrast, other studies have found that accuracy of children’s self-reporting improves with age (Edelbrock, Costello, Duncan, Kalas, &: Conover, 1985). In the present study, the child’s age was unrelated to the strength of the mother-child agreement.

The findings of poor agreement would therefore appear to be consistent with findings in other populations. There may not be any unique distortions in maternal reporting that pertain to the special situation of child sexual abuse. Clinicians and researchers might therefore be well advised to attend to maternal distress in any and all contexts and to gather independent observations on children’s behavior and psychological functioning.

CONCLUSIONS

This study underscores the importance of addressing a woman’s psychological needs as an integral component of the treatment of her child’s sexual abuse. All too often, when their children are sexually abused, mothers appear to be blamed, if not for the abuse itself, then for their responses following the abuse (Mcintyre, 1981; Myer, 1985; Peters, 1988; Salt, Myer, Coleman, & Sauzier, 1990; Sirles & Franke, 1989; Wanenberg, 1985). Data from this study point to an alternative way of thinking about these women. Mothers, too, are traumatized by the abuse of their children. Sensitive and compassionate responses to child sexual abuse should acknowledge mothers’ emotional suffering, and the women’s reports of their children’s emotions should be taken seriously, though with corroboration by direct interviews of the children. Mothers of sexually abused children should receive a clinical evaluation and be offered psychotherapy and/or family therapy if indicated. The distress of the victimized children should be assessed directly. and their treatment needs should be considered from both maternal and child perspectives.

REFERENCES

Achenbach, T.M., & Edelbrock, C.S. (1983). Manual for the Child Behavior Checklist and Revised Behavior Profile. Burlington, VT: University of Vermont, Department of Psychiatry.

Achenbach, T.M., McConaughy, S., & Howell. C. (1987). Child/adolescent behavioral and emotional problems: Implications of cross-informant correlations for situational specificity. Psychological Bulletin, 101, 213-232.

American Psychiatric Association, (1987). Diagnostic and statistical manual of mental disorders, 3rd ed., rev. Washington, DC: Author.

Billings, A.G., & Moos, R.H. (1983). Comparison of children with depressed and nondepressed parents: A social environmental perspective. Journal of Abnormal Child Psychology, 11, 483-486.

Conte, J. (1985). The effects of sexual abuse on children: A critique and suggestions for future research. Victimology, 10, 110-130.

Conte, J. (1987, January). The effects of sexual abuse on children: Results of a research project. In Human sexual aggression: Current perspectives. Symposium conducted at a meeting of the New York Academy of Sciences, New York.

Cox, A., Puckering. A., Pound, A., & Mills, M. (1987). The impact of maternal depression in young children. Journal of Child Psychology and Psychiatry, 28. 917-928.

De Jong, A. (1988). Maternal responses to the sexual abuse of their children. Pediatrics, 81, 14-20.

Derogatis, L.R., & Spencer, M.S. (1982). Administration and procedures: BSI Manual 1. Baltimore. MD: Clinical Psychometric Research. Johns Hopkins University.

Downey, G., & Coyne, J.C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin. 108, 50-76.

Edelbrock, C., Costello, A.J., Duncan, M.K., Kalas. R., & Conover, N.C. (1985). Age differences in the reliability of the psychiatric interview of the child. Child Development, 56, 265-275.

Estroff, T., Herrera, C., Gaines, R., Shaffer, D., Gould, M., & Green, A. (1984). Maternal psychopathology and perception of child behavior in psychiatrically referred and child maltreatment families. Journal of the American Academy of Child Psychiatry, 23, 649-652.

Everson, M.D., Hunter, W.M., Runyon, D.K., Edelsohn, G.A., & Coulter, M.L. (1989). Maternal support following incest. American Journal of Orthapsychiatry, 59, 197-207.

Fergusson, D., Horwood, L., Grenon, M., & Shannon, F. (1985). Family life events, maternal depression, and maternal and teacher descriptions of child behavior. Pediatrics, 75, 30-35.

Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry. 55. 530-541.

Friedlander, S., Weiss, D., & Traylor, J. (1986), Assessing the influence of maternal depression on the validity of the Child Behavior Checklist. Journal of Abnormal Child Psychology. 14, 123-133.

Friedrich, W.N.. Urquiza, AJ., & Beilke, R.L. (1986). Behavior problems in sexually abused young children. Journal of Pediatric Psychology, 11, 47-57.

Gelinas, D. (1983). The persisting negative affects of incest. Psychiatry, 46. 312-332.

Gomes-Schwartz, B., Horowitz, J.M., & Sauzier, M. (1985). Severity of emotional distress among sexually abused pre-school, school age, and adolescent children, Hospital and Community Psychiatry, 36. 503-508.

Griest, D., Forehand, R., Wells. K., & McMahon, R. (1980). An examination of differences between non-clinic and behavior-problem clinic-referred children and their mothers. Journal of Abnormal Psychology, 89. 497-500.

Herman, J., Russell. D., & Trocki, K. (1986). Long term affects of incestuous abuse in childhood, American Journal of Psychiatry. 143. 1293-1296.

Hollingshead, A.B. (1979). Four-factor index of social status. Unpublished manuscript, Yale University, New Haven, CT.

Kazden, A.E., Esveldt-Dawson, K.E., Sherick, R.B., & Colbus, D. (1985). Assessment of overt behavior and childhood depression among psychiatrically disturbed children. Journal of Consulting and Clinical Psychology, 53. 201-210.

Kazden, A.E., French, N.H., & Unis, A.S. (1983). Assessment of childhood depression: Correspondence of child and parent ratings. Journal of the American Academy of Child Psychiatry, 22. 157-164.

Kazden, A.E., & Petti. T.A. (1982). Self-report interview measures of childhood and adolescent depression. Journal of Child Psychology and Psychiatry. 23. 437-457.

Kelley, S. (1990). Parental stress response to sexual abuse and ritualistic abuse of children in day-care centers. Nursing Research. 39. 25-29.

Kochanska, G. (1990). Maternal beliefs as long-term predictors of mother-child interaction and report. Child Development. 61. 1934-1943.

Kochanska, G., Radke-Yarrow, M.L., Kuczynski, L., & Friedman, S. (1987). Normal and affectively ill mothers’ beliefs about their children. American Journal of Orthopsychiatry, 57. 345-350.

Kovacs, M. (1981). Rating scales to assess depression in school aged children. Acta Paedopsychiatrica, 46. 305-315.

Kovacs, M. (1985). The Children’s Depression Inventory. Psychopharmacy Bulletin. 21. 995-998.

Krigbaum-Rich, R. (1991). Disclosure of child sexual abuse: A phenomenological study of the mother’s experience. Unpublished doctoral dissertation. Massachusetts School of Professional Psychology, Dedham. MA.

McIntyre, K. (1981). Role of mothers in father daughter incest: A feminist analysis. Social Work. 26. 462-466.

Myer, M. (1985). A new look at mothers of incest victims. Journal of Social Work and Human Sexuality. 3. 47-58.

Newberger, C.M., & De Vos, E. (1988). Abuse and victimization: A life-span developmental perspective. American Journal of Orthopsychiatry. 58. 505-511.

Pela, O.A., & Reynolds, C.R. (1982). Cross-cultural application of the Revised Children’s Manifest Anxiety Scale: Normative and reliability data for Nigerian primary school children. Psychological Reports, 51. 1135-1138.

Peters, S. (1988). Child sexual abuse and later psychological problems. In G. Wyatt & G. Powell (Eds.). Lasting effects of child sexual abuse (pp. 101-117). Newbury Park, CA: Sage Publications.

Regehr, C. (1990). Parental responses to extrafamilial child sexual assault. Child Abuse and Neglect. 14. 113-120.

Reynolds, C.R. (1981). Long-term stability of scores on the Revised Children’s Manifest Anxiety Scale. Perceptual and Motor Skills. 53. 702.

Reynolds, C.R., & Richmond, B.O. (1985). Revised Children’s Manifest Anxiety Scale (RCMAS). Los Angeles. CA: Western Psychological Services.

Richters, J., & Pellegrini, D. (1989). Depressed mothers’ judgments about their children: An examination of the depression-distortion hypothesis. Child Development. 60. 1068-1075.

Salt, P., Myer. M., Coleman. L., & Sauzier, M. (1990). The myth of the mother as “accomplice” to child sexual abuse. In B. Gomes-Schwartz. J. Horowitz, & A. Cardelli (Eds.). Child sexual abuse: The initial effects (pp. 109-131). Newbury Park, CA: Sage Publications.

Schaefer, E., Hunter. W., & Edgerton, M. (1987). Maternal prenatal, infancy, and concurrent predictors of maternal reports of child psychopathology. Psychiatry, 50. 320-331.

Sirles, E., & Franke, P. (1989). Factors influencing mothers’ reactions to intrafamily sexual abuse. Child Abuse and Neglect. 13. 131-139.

Wagner, W.G. (1991). Depression in mothers of sexually abused vs. mothers of nonabused children. Child Abuse and Neglect. 15. 99-104.

Wattenberg, E. (1985). In a different light: A feminist perspective on the role of mothers in father daughter incest. Child Welfare, 64. 203-211.

Webster-Stratton, C., & Hammond, M. (1985). Maternal depression and its relationships to life stress, perceptions of child behavior problems, parenting behaviors, and child conduct problems. Journal of Abnormal Child Psychology. 16. 299-315.

Winton, M. (1990). An evaluation of a support group for parents who have a sexually abused child. Child Abuse and Neglect. 14. 397-405.

Based on a paper presented at the 1991 annual meeting of the American Orthopsychiatric Association in Toronto. Research was supported by grants from the National Center on Child Abuse and Neglect (90-CA-II84). The National Institute of Justice (89-IJ-CX-OO.34), and NIMH (1T.32MH 8265.09). Authors are at: Departments of Psychiatry (C. Newberger). Maternal and Child Health (Gremv), Biostatistics (Waternaux), and Pediatrics (E. Newberger), Harvard University. Boston, Mass.