Abstract
Maternal support has been widely cited as an important predictor of children’s adjustment following disclosure of sexual abuse. However, few studies have examined these effects longitudinally. The current study examines the relationships between a multidimensional assessment of maternal support rated by both mothers and children and children’s adjustment in various domains (internalizing, externalizing, anger, depression, and posttraumatic stress disorder symptoms) concurrently and longitudinally. Participants were 118 mother-child dyads recruited from a Child Advocacy Center where children were determined through a forensic evaluation to be victims of sexual abuse. Child and mother ratings of maternal support and child adjustment were collected shortly after the forensic evaluation and at 9-month follow-up. Results were consistent with findings from past studies that maternal support is significantly related to children’s post-disclosure adjustment and extends these findings longitudinally. Additionally, the study sheds light on differential relations between dimensions of maternal support (Emotional Support, Blame/Doubt, Vengeful Arousal, and Skeptical Preoccupation) and child adjustment and suggests the importance of using both child and mother ratings of maternal support in future research.
Keywords: child sexual abuse, abuse disclosure, maternal support, longitudinal design
Maternal support following a child’s sexual abuse disclosure has been widely cited as an important predictor of post-abuse adjustment. Several aspects of maternal support, including reactions to disclosure, belief of the child’s report, and protection of the child have implications for subsequent development of psychopathology. Research has generally supported this link between maternal support and children’s adjustment, though there have been inconsistencies across studies (see Elliott & Carnes, 2001; Lovett, 2004; Ullman, 2003 for reviews). Further, defining maternal support has been problematic, with inconsistencies in measurement techniques and types of support measured. The current study utilizes a multi-dimensional abuse-specific measure to examine the relation between children’s and mothers’ perceptions of various aspects of maternal support and children’s adjustment following disclosure and at 9-month follow-up.
The Transactional Model
According to the transactional model of child development (Spaccarelli, 1994) childhood sexual abuse (CSA) is comprised of a host of potentially stressful experiences, and maternal reactions to CSA disclosure are critical in determining adjustment. If the mother does not provide adequate support, the child is at increased risk for maladaptive coping strategies such as avoidance and self-blame. However, with adequate support, the child may be more likely to seek emotional support from her and benefit from other ameliorative efforts (e.g., psychotherapy). Studies have documented the importance of social support as a general protective factor against psychopathology (e.g., Kawachi & Berkman, 2001; Moak & Agrawal, 2010). However, in the CSA literature, maternal support has been assigned special importance as a predictor of children’s adjustment and a target for intervention. Thus, though the transactional model conceptualizes maternal support as a key variable, the challenge remains to identify behavioral aspects of maternal support that are empirically related to children’s post-disclosure adjustment.
Studies of Maternal Support
In an early investigation of the relationship between maternal support and child adjustment, lack of support by a close adult predicted behavior problems such as withdrawal, suicide attempts, running away, fire-setting, and aggression (Adams-Tucker, 1986). Similarly, Conte and Schuerman (1987) found that lack of supportive relationships with siblings and the non-offending parent predicted parent-reported child psychopathology following CSA. Everson and colleagues (1989) found low levels of maternal support to be related to both childhood distress and behavior problems, while Feiring and colleagues (1998) reported that high levels of generic social support (including maternal support) were associated with less depression. Similarly, several studies found an association between mother-child relationship quality and fewer parent-rated externalizing behaviors (Esparza, 1993; Smith et al., 2010; Tremblay, 1999). In treatment outcome studies, parental support predicted better response to intervention among preschool aged children (Cohen & Mannarino, 1998, 2000). Similarly, in a study of female adolescents in a sexual trauma inpatient unit, youth-rated maternal support was unrelated to adjustment at baseline but, at time of discharge and 3-month follow-up, support was related to better self-concept and fewer depressive symptoms (Morrison & Clavenna-Valleroy, 1998).
Several studies reported deviations from this pattern of findings. Mannarino and Cohen (1996) reported no significant relations between maternal non-support (“blaming”) and children’s symptoms. Another failed to find a link between maternal support and problem behaviors in a sample of sexually abused girls (Hebert, Collin-Vezina, Daigneault, Parent, & Tremblay, 2006). A third did not find a link between maternal support and children’s self-blame or internalizing problems (Quas, Goodman, & Jones, 2003), but this study used a single child-rated item of maternal support. The first two studies relied solely on mother-rated maternal support, and all three used samples reporting high support. Mannarino and Cohen (1996) reported that mothers rarely endorsed responses that could be considered socially unacceptable. The other two samples likely consisted of highly supportive mothers, as one was seeking medical care for their children (Hebert et al., 2006) and the other was pursuing legal cases related to CSA (Quas et al., 2003), potentially leading to a ceiling effect. These findings highlight the importance of using multiple raters of maternal support and recruiting from diverse populations.
Limitations in Definitions and Measurement of Maternal Support
Examination of the literature reveals conceptual and methodological limitations requiring further research before firm conclusions can be drawn. For one, current definitions of maternal support lack precise and consistently utilized criteria. For example, maternal support has often been assessed without reference to the abuse itself, as reflected by overall level of support or general parent-child relationship quality (Conte & Schuerman, 1987; Esparza, 1993; Feiring et al., 1998). Further, many studies have used vague or overly inclusive definitions of maternal support. For example, studies on abuse-specific support have assessed one or more forms of protective action, verbal/emotional support following disclosure, and belief of the child’s report of abuse. Some studies measure all of these factors, while others assess just one or two, making comparisons across studies difficult. In addition, studies utilizing multiple indicators of maternal support have not specified which elements are most predictive of child adjustment. Thus, it is unclear whether the benefits of maternal support are due primarily to the effects of one type of support, some combination of different types of support, or the overall quality of the relationship.
Considerable variability also exists in terms of raters of maternal support. In several studies, mental health professionals or caseworkers have provided data (Everson et al., 1989; Heriot, 1996; Sirles & Franke, 1989), which is limited to what these professionals observe in their brief interactions with families. In other studies, mothers rated their own behaviors (Hebert et al., 2006; Hsu & Smith, 2000; Mannarino & Cohen, 1996). While this is an improvement in some respects over observer reports, mothers may endorse unrealistically positive and socially desirable responses. Surprisingly, there are only four studies of children’s perceptions of support, but these did not assess abuse-specific dimensions (Esparza, 1993; Feiring et al., 1998; Lovett, 1995; Tremblay et al., 1999). Another study reported adolescent ratings of maternal support but the measure lacked psychometric data (Morrison & Clavenna-Valleroy, 1998). Finally, an earlier study with the current sample found match between mother and child ratings of maternal support to be an important predictor of concurrent child outcomes but did not examine the construct of maternal support as a predictor of child outcomes (Bick, Zajac, Ralston, & Smith, 2014).
To address limitations in the operational definition and measurement of maternal support, two instruments, the Maternal Self-Report Support Questionnaire and the Maternal Support Questionnaire-Child Report, were constructed and validated (Smith et al., 2010; 2014). These measures make several important contributions to the field. First, they allow for both mother and child ratings of maternal support. Second, they offer a definition of maternal support composed of specific behaviors related to CSA disclosure. There are two subscales rated by the mother (Emotional Support, Blame/Doubt) and three by the child (Emotional Support, Skeptical Preoccupation, and Vengeful Arousal). Both measures have good internal consistency, predictive validity, and demonstrate multiple dimensions of maternal support (Smith et al., 2010; 2014). Further, they have been found to be related to but distinct from ratings of more general (not abuse-specific) maternal support, suggesting discriminate validity.
The Current Study
Though maternal support has been linked to children’s adjustment concurrently, it is unclear whether the maternal support has long term effects on children. Thus, we examine both child and mother-rated maternal support as predictors of children’s adjustment soon after abuse disclosure and nine months later. We predict that maternal emotional support will be related to lower levels of symptoms and behavioral problems (i.e., depression, anger, posttraumatic stress disorder, internalizing and externalizing behavior problems) at both time points. We also predict that high levels of maternal blame or doubt will increase risk of poor adjustment as evidenced by elevated rates of symptoms at both time points. The current study adds to previous literature in two ways: 1) examination of longer terms adjustment outcomes related to maternal support; and 2) providing an initial examination of the relationships between specific behavioral aspects of maternal support and children’s subsequent adjustment.
Method
Participants
Participants were 118 children and mothers (or female guardians) recruited from a Child Advocacy Center (CAC). Inclusion criteria were: a) children were 7–16 years old; b) children were victims of CSA involving physical contact (e.g., fondling, penetration) as determined by a forensic evaluation; c) mothers may not have been complicit or involved in the sexual abuse (i.e., “non-offending’) and; d) the initial report of CSA to authorities must either have occurred within the six weeks prior to the forensic evaluation or resulted from the evaluation itself. Exclusion criteria were psychotic behavior, significant cognitive disability or mental retardation, or inability to speak Spanish or English; no cases were excluded based on these criteria.
Participants were recruited over a 3-year period as part of first time forensic assessment following CSA. Over this period, 1254 children presented to the CAC; 957 children (76.3%) did not meet inclusion criteria for the study. Of the remaining 297 children, 46 did not consent to participate in the study, and 125 could not be contacted for recruitment or did not attend their appointment. The final sample of 120 mother-child dyads represented 41% of the eligible participants, 49% of those who consented to be recruited, and 86% of those who could be located for recruitment. Of these 120, two were excluded from the analyses due to missing data on a baseline variable of interest, leaving a final sample of 118.
Demographic characteristics were collected at the time of the first assessment and are summarized in Table 1. The majority of caregivers were biological mothers, and similar numbers of mothers who identified themselves as Caucasian and African American. The mothers ranged in age from 23 to 62 (M = 37.53, SD = 8.20) at the first assessment, and approximately half were married. Children had a mean age of 11.57 (SD = 2.69; range 7 to 17) at the first assessment, and a majority were female.
Table 1.
Demographic Variables at Time 1 and Time 2
| Time 1 (n = 118) | Time 2 (n = 57) | ||
|---|---|---|---|
| Caregiver Relationship to Child | |||
| Biological mother | 101 (85.6%) | 50 (87.7%) | |
| Other female relative | 14 (11.9%) | 5 (8.8%) | |
| Stepmother | 2 (1.7%) | 1 (1.8%) | |
| Adoptive mother | 1 (.8%) | 1 (1.8%) | |
| Caregiver Employment | |||
| Full-time | 53 (44.9%) | 20 (35.1%) | |
| Part-time | 16 (13.6%) | 11 (19.3%) | |
| Caregiver Race | |||
| Caucasian | 68 (57.6%) | 31 (54.4%) | |
| African American | 49 (45.1%) | 26 (45.6%) | |
| Native American | 1 (.8%) | 0 | |
| Caregiver Marital Status | |||
| Married | 59 (50%) | 28 (49.1%) | |
| Divorced/separated | 31 (26.3%) | 15 (26.3%) | |
| Single | 20 (16.9%) | 11 (19.3%) | |
| Living with a romantic partner | 4 (3.4%) | 2 (3.5%) | |
| Dating but not living with partner | 2 (1.7%) | 1 (1.8%) | |
| Widowed | 2 (1.7%) | 0 | |
| Caregiver Educational Attainment | |||
| Did not complete high school | 26 (22.0%) | 12 (21.1%) | |
| High school diploma | 36 (30.5%) | 17 (29.8%) | |
| Some college | 39 (33.1%) | 21 (36.9%) | |
| Advance degree | 17 (14.4%) | 7 (12.3%) | |
| Child Race | |||
| Caucasian | 66 (55.8%) | 30 (52.6%) | |
| African American | 49 (45.1%) | 27 (47.4%) | |
| Bi-racial | 1 (.8%) | 0 | |
| Child Gender | |||
| Male | 17 (14.4%) | 9 (15.8%) | |
| Female | 101 (85.6%) | 48 (84.2% | |
| Family Income less than $30,000 | 74 (63.2%) | 32 (57.1%) | |
Procedure
During the forensic evaluation, research assistants requested consent from eligible participants to contact them by phone. Trained female interviewers contacted dyads that met inclusion and exclusion criteria to invite them to participate. At the assessment appointment, staff reviewed the study’s purpose, explained procedures, and collected informed consent from the mother and assent from the child. Interviews with children and mothers were conducted in separate rooms. Staff read measures to participants who had reading difficulties. Mothers received $40 and children received $15 in coupons from local businesses to compensate for their time. All procedures were approved by the Institutional Review Boards (IRBs) of the medical center and the CAC.
Follow-up data collection occurred approximately 9 months later (M = 9.8 months, range = 5 to 17 months, SD = 2.23). A variety of techniques were used to retain participants in the study, including asking participants to provide names and contact information of family members and friends and attempting to contact dyads multiple times. Despite these efforts, 61 dyads were unavailable for the second wave of data collection.
Measures
Maternal Self-Report Support Questionnaire (MSSQ)
The MSSQ (Smith et al., 2010) is a 14-item questionnaire assessing a mother’s perceptions of her own behavior in two domains of abuse-specific support. The Emotional Support scale includes 6 items reflecting provision of emotional support through direct action (e.g., “reassured your child you would stand by him/her”). Higher scores correspond to higher levels of emotional support. The Blame/Doubt scale includes 8 items measuring the tendency to doubt the child’s disclosure or question the child’s role in the abuse (e.g., “questioned your child’s honesty about the abuse”). High scores indicate more blame or doubt of the child. Responses are made on a Likert-type scale ranging from 0 (not at all like me) to 6 (very much like me). Internal consistencies for the Emotional Support and Blame/Doubt scales were adequate with Cronbach’s α of .76 and .74, respectively.
Maternal Support Questionnaire - Child-Report (MSQ-CR)
The MSQ-CR (Smith et al., 2014) is a 21-item measure of children’s perceptions of abuse-specific maternal support. The 10-item Emotional Support scale assesses belief in the child’s disclosure and emotional support (e.g., “helps me feel better about what happened”). Higher scores indicate higher levels of perceived support. The Skeptical Preoccupation scale contains 5 items that measure questioning the child’s role in the abuse and persistently seeking information about the abuse (e.g., “tries to get more information to learn about it”). Finally, the Vengeful Arousal scale is comprised of 6 items measuring vengeful statements directed towards the perpetrator (e.g., “says she wants to harm the perpetrator”). Behaviors are rated on a Likert-type scale ranging from 0 (not at all) to 3 (very much). Internal consistencies for the Emotional Support, Skeptical Preoccupation, and Vengeful Arousal scales were adequate, with Cronbach’s α of .87, .71 and .73, respectively.
Child Behavior Checklist (CBCL) - Parent Report Form
Mothers completed the CBCL, a 113-item measure of children’s behavioral and emotional problems (Achenbach & Rescorla, 2001). T-scores for two subscales were used. The internalizing subscale measures depression/anxiety, somatic complaints, and withdrawal, and the externalizing subscale measures delinquency and aggression. Behaviors were rated on a scale of 0 (not true), 1 (somewhat true), and 2 (very true). Achenbach and Rescorla (2001) report evidence for good inter-rater and test-retest reliability as well as criterion-related, convergent, and discriminant validity.
Trauma Symptom Checklist for Children (TSCC)
The TSCC (Briere, 1996) is a child-rated measure of trauma-related symptoms comprised of 54 items that constitute six primary subscales: anger, anxiety, depression, dissociation, posttraumatic stress (PTSD), and sexual concerns. The PTSD, anger, and depression scales were utilized in the current study. A growing body of literature attests to the internal consistency and validity of the TSCC (Crouch, Smith, Ezzell, & Saunders, 1999; Lanktree et al., 2008; Sadowski & Friedrich, 2000).
Structured interviews
Structured clinical interviews adapted from protocols used in the Navy Family Study (Grasso et al., 2013) were administered to the mother and child. The mother’s interview collected information about family history, exposure to violence, and demographic information. The child’s interview collected information about abuse history and demographics.
Results
Descriptive Information
Descriptive statistics for child adjustment and maternal support variables for each time point are presented in Table 2. Descriptive information was also obtained regarding the CSA characteristics reported by the children. Fifty-nine (50.0%) of the perpetrators were family members (e.g., fathers, siblings, and other relatives) and 57 (48.3%) were non-family members (e.g., neighbors, family friends, and peers); data were missing for two dyads on this question. In regards to the mother’s relationship with the perpetrator, 19 (16.1%) were in a romantic relationship (married, dating, or co-habitating), 4 (3.4%) were ex-husbands, 35 (29.7%) were described as trusted family members or friends, 11 (9.3%) were other family members, 17 (14.4%) were acquaintances, 28 (23.7%) were strangers, and 4 (3.4%) were described as “Other.” The majority of children (55.2%) experienced multiple acts of abuse, and 42.2% reported some form of penetration (e.g., penile, digital).
Table 2.
Descriptive Statistics for Child Adjustment and Maternal Support Variables
| Time 1 | Time 2 | |||
|---|---|---|---|---|
| M | SD | M | SD | |
| CBCL Internalizing | 57.27 | 12.02 | 55.84 | 11.49 |
| CBCL Externalizing | 58.39 | 11.47 | 56.06 | 10.51 |
| TSCC PTSD | 10.43 | 6.27 | 9.03 | 6.74 |
| TSCC Anger | 7.72 | 6.08 | 7.70 | 7.08 |
| TSCC Depression | 7.52 | 5.54 | 6.83 | 5.98 |
| MSSQ Emotional Support | 39.65 | 5.08 | 39.53 | 3.90 |
| MSSQ Blame/Doubt | 12.45 | 9.50 | 11.58 | 9.58 |
| MSQC Emotional Support | 22.83 | 5.83 | 22.23 | 6.27 |
| MSQC Preoccupation | 6.59 | 4.05 | 5.81 | 3.71 |
| MSQC Vengeful Arousal | 9.73 | 4.15 | 8.37 | 3.88 |
CBCL=Child Behavior Checklist, TSCC=Trauma Symptom Checklist for Children, MSSQ=Maternal Self-Report Support Questionnaire, MSQC=Maternal Support Questionnaire-Child Report
Attrition
There was significant attrition between the first and second waves of data collection with 57 of the 118 dyads available at follow-up. Thus, it was important to determine whether dyads who participated in the second wave differed from those who did not. A dichotomous missing versus not missing variable was calculated. Using t-tests and χ2 statistics, each study variable was tested for significant differences between dyads missing and not missing at time 2. No significant differences were found on any child adjustment, maternal support, CSA characteristic, or demographic variable. Thus, participants who were unavailable at time 2 did not differ substantially from those who were available on variables of interest.
Child Adjustment: Bivariate Correlations
The primary goal of the study was to examine the role of maternal support in the adjustment of children both at the initial evaluation and at 9-month follow-up. Child demographics (gender, race, age) as well as abuse characteristics known to be predictors of maladjustment among children (perpetrator, frequency, penetration) were examined as predictors of children’s adjustment.
Table 3 shows the bivariate correlations between each child adjustment variable and the demographic, abuse characteristic, and maternal support variables. Mother-reported emotional support following CSA disclosure was related to lower levels of internalizing and externalizing problems at time 1 and lower levels of internalizing problems, PTSD, anger, and depression at time 2. Mother-reported Blame/Doubt was related to elevated levels of externalizing behaviors at time 1 but not time 2. Child-reported Emotional Support was related to lower levels of anger and depression at time 1 and showed no relations with child adjustment at time 2. Child-reported Skeptical Preoccupation was unrelated to child adjustment at either time point. Finally, children’s ratings of Vengeful Arousal were related to PTSD at time 1.
Table 3.
Bivariate Relationships between Child Adjustment and Child Demographics, Abuse Characteristics, and Maternal Support Variables
| Child Age |
Child Sex |
Child Race |
Perpetrator | Frequency | Penetration | MSSQ ES |
MSSQ Blame |
MSQC ES |
MSQC Preocc |
MSQC VA |
|
|---|---|---|---|---|---|---|---|---|---|---|---|
| CBCL Int T1 | .25** | −.05 | −.32** | .01 | .14 | .19* | −.19* | .14 | .01 | −.08 | .02 |
| CBCL Ext T1 | .25** | .11 | −.10 | −.11 | .01 | .23* | −.38*** | .28** | −.17 | −.02 | −.13 |
| TSCC PTSD T1 | −.01 | .04 | −.14 | .06 | .10 | −.05 | −.08 | −.16 | .02 | .13 | .29** |
| TSCC Anger T1 | .15 | .19* | −.09 | .00 | −.01 | −.01 | −.15 | −.03 | −.26** | −.07 | .09 |
| TSCC Depress T1 | −.02 | .19* | −.21* | .01 | .01 | −.05 | −.08 | −.16 | −.20* | .09 | .09 |
| CBCL Int T2 | .15 | −.17 | −.18 | −.23 | .18 | .34* | −.29* | .19 | .09 | −.04 | .01 |
| CBCL Ext T2 | .17 | −.17 | −.24 | −.35* | −.09 | .25 | −.15 | .12 | −.03 | −.01 | −.16 |
| TSCC PTSD T2 | .04 | .00 | .18 | .16 | .13 | .03 | −.28* | .03 | −.04 | .16 | .26* |
| TSCC Anger T2 | .15 | .06 | −.04 | −.11 | −.07 | .07 | −.25* | .09 | −.13 | −.15 | .05 |
| TSCC Dep T2 | −.04 | .16 | .05 | .05 | .11 | .10 | −.34* | .05 | −.12 | .15 | .08 |
p < .05.
p < .01.
p < .001.
Note. Variables are coded as follows: Child sex: 1 = male, 2 = female; Child race: 0 = all other races, 1 = African American; Perpetrator: 0 = family member, 1 = non-family member; Frequency: 1 = once, 2 = more than once; Penetration: 0 = no, 1 = yes. CBCL=Child Behavior Checklist, TSCC=Trauma Symptom Checklist for Children, MSSQ=Maternal Self-Report Support Questionnaire, MSQC=Maternal Support Questionnaire-Child Report
In addition to relationships between child adjustment and maternal support variables, there were multiple significant relationships between child adjustment and both child demographic and sexual abuse characteristic variables. Thus, significant relationships were controlled in all subsequent multiple regression models. Due to the relatively small sample size, only variables that were significantly related to child adjustment at the bivariate level were included in these models to conserve degrees of freedom.
Child Adjustment: Multiple Regression Analyses
Separate multiple regression models were conducted for each child adjustment outcome, controlling for relevant demographic and CSA characteristic variables. Variables were entered in three steps: (1) child demographics, (2) CSA characteristics, and (3) maternal support. When there were no significant bivariate relationships for either child demographic or CSA characteristics, these steps were eliminated.
The regression model predicting children’s internalizing problems at time 1 was significant, F(4, 97) = 6.57, p < .001. African American race predicted lower levels of externalizing problems; however, when controlling for child demographic variables, penetration during sexual abuse and maternal support were no longer significantly related to internalizing problems. The model predicting externalizing behaviors at time 1 was also significant, F(6, 95) = 5.62, p < .001. African American race and mother-rated Emotional Support predicted lower levels of externalizing problems, whereas maternal ratings of Blame/Doubt were related to increased externalizing problems. Table 4 summarizes these results.
Table 4.
Linear Regression Model Estimating the Relationships between Maternal Support and Child Adjustment at Time 1
| Variable | Step 1 β |
Step 2 β | Step 3 β | R2 | Δ R2 | |
|---|---|---|---|---|---|---|
| CBCL Internalizing | ||||||
| Step 1 | .16*** | |||||
| Child Age | .22* | .20* | .16 | |||
| Child Ethnicity | −.32** | −.32** | −.35*** | |||
| Step 2 | .19*** | .03 | ||||
| Penetration | .17 | .16 | ||||
| Step 3 | .21*** | .03 | ||||
| MSSQ Emotional Support | −.17 | |||||
| CBCL Externalizing | ||||||
| Step 1 | .06* | |||||
| Child Age | .23* | .20* | .07 | |||
| Child Ethnicity | −.08 | −.08 | −.23* | |||
| Step 2 | .10* | .04* | ||||
| Penetration | .20* | .16 | ||||
| Step 3 | .26*** | .16*** | ||||
| MSSQ Emotional Support | −.30** | |||||
| MSSQ Blame | .23* | |||||
| MSQC Emotional Support | −.07 | |||||
| TSCC PTSD | ||||||
| Step 1 | .11*** | |||||
| MSSQ Emotional Support | −.16 | |||||
| MSQC Vengeful Arousal | .33*** | |||||
| TSCC Anger | ||||||
| Step 1 | .04* | |||||
| Child Sex | .19* | .16 | ||||
| Step 2 | .09** | .06** | ||||
| MSQC Emotional Support | −.24** | |||||
| TSCC Depression | ||||||
| Step 1 | .08** | |||||
| Child Sex | .20* | .15 | ||||
| Child Ethnicity | −.21* | −.20* | ||||
| Step 2 | .14** | .06* | ||||
| MSSQ Blame | −.16 | |||||
| MSQC Emotional Support | −.25** | |||||
p < .05.
p < .01.
p < .001.
CBCL=Child Behavior Checklist, TSCC=Trauma Symptom Checklist for Children, MSSQ=Maternal Self-Report Support Questionnaire, MSQC=Maternal Support Questionnaire-Child Report
Models predicting children’s ratings of PTSD, anger, and depression at time 1 were all significant: F(2, 113) = 6.91, p < .01; F(2, 116) = 5.93, p < .01; F(4, 111) = 4.65, p < .01, respectively (Table 4). Child ratings of Vengeful Arousal were significantly related to PTSD while child ratings of Emotional Support were related to lower levels of anger. Both African American race and higher levels of child-rated Emotional Support were related to lower levels of depression at time 1.
At time 2, the model predicting children’s internalizing problems was significant F(2, 38) = 3.34, p < .05; however, neither penetration nor maternal emotional support were significantly related to internalizing problems in the final model (Table 5). As there were no bivariate relationships between mother- or child-rated maternal support variables and children’s externalizing problems at time 2, a regression model was not used to explore these relations.
Table 5.
Linear Regression Model Estimating the Relationships between Maternal Support and Child Adjustment at Time 2
| Variable | Step 1 β |
Step 2 β | R2 | Δ R2 | |
|---|---|---|---|---|---|
| CBCL Internalizing | |||||
| Step 1 | .27 | .11* | |||
| Penetration | .33* | ||||
| Step 2 | .15* | .04 | |||
| MSSQ Emotional Support | −.22 | ||||
| TSCC PTSD | |||||
| Step 1 | .14* | ||||
| MSSQ Emotional Support | −.26 | ||||
| MSQC Vengeful Arousal | .25 | ||||
| TSCC Anger | |||||
| Step 1 | .06 | ||||
| MSSQ Emotional Support | −.25 | ||||
| TSCC Depression | |||||
| Step 1 | .12* | ||||
| MSSQ Emotional Support | −.34* | ||||
p < .05.
CBCL=Child Behavior Checklist, TSCC=Trauma Symptom Checklist for Children, MSSQ=Maternal Self-Report Support Questionnaire, MSQC=Maternal Support Questionnaire-Child Report
Finally, as shown in Table 5, the regression models predicting child-rated symptoms of PTSD and depression were significant at time 2, F(2, 52) = 4.06, p < .05; F(1, 54) = 6.98, p < .05, but the model predicting anger was not, F(1, 53) = 3.40, p = .07. Mother-rated Emotional Support and child-rated Vengeful Arousal were not significantly related to PTSD, but mother-rated Emotional Support was significantly related to lower levels of depression at time 2.
Discussion
The primary aim of the study was to examine the concurrent and longitudinal relations between mother and child-rated maternal support and children’s adjustment. Results generally supported past findings of significant relationships between maternal support and children’s adjustment following CSA and provide some initial evidence that maternal support has implications for longer term adjustment as well. In addition, findings provide additional information regarding the specific types of support that are important in predicting different child adjustment outcomes.
Child ratings of maternal support were primarily related to child adjustment outcomes at time 1 and were only significantly related to child-rated outcomes (i.e., PTSD, depression, and anger). These findings suggest that children’s perceptions of maternal support are relatively specific to the time point at which they are obtained and to their self-assessment of adjustment problems, rather than those rated by their mothers. Interestingly, children’s ratings of their mother’s vengeful arousal were related to their own reports of elevated PTSD symptoms at time 1, with the relationship not significant but in the expected direction at time 2. This suggests that children whose mothers discuss seeking revenge on the perpetrator have more severe trauma-related symptoms. The direction of effects for this relationship cannot be determined by this study but could certainly be bi-directional, with children’s maladjustment fueling mothers’ feelings of vengeance and vice versa. Finally, the Preoccupation subscale of the child-reported maternal support measure was not related to any child adjustment outcome at either time point. This may suggest its limited utility in predicting child mental health.
Mothers’ ratings of maternal support followed a similar pattern, with the strongest associations seen at time 1 and with mother ratings of child adjustment. The Emotional Support scale of the MSSQ was consistently related to mother-rated child outcomes at time 1, with most relationships not significant but in the expected direction for both mother- and child-rated outcomes at time 2. One exception was the significant relationship between maternal ratings of support and lower levels of child-rated depression at time 2. This represents the one significant prospective cross-informant relationship in this study. Given the numerous models tested in these analyses, results should be interpreted with caution and will require replication in other studies. However, this finding provides preliminary evidence of a long-term protective relationship between maternal support and the development of depression following CSA disclosure.
An important contribution of the current study is the use of a multidimensional abuse-specific measure of maternal support that can be rated by both mothers and children. Assessment of multiple aspects of maternal support has proven to be important, as various maternal behaviors were shown to be differentially related to child outcomes. In addition to replicating past findings that maternal support is related to child adjustment following CSA disclosure, it provided additional information regarding which maternal behaviors might be most helpful (or hurtful) to children’s adjustment. Further, it highlights the fact that child and mother perceptions of maternal support are sometimes very different from one another, and both are associated with children’s well-being.
Results of the study must be interpreted in light of certain limitations. Despite intensive efforts to contact and retain participants, there was substantial attrition between time 1 and time 2. In some cases, we were unable to contact participants altogether, while other dyads declined participation. Importantly, there were no significant differences between dyads who participated in the time 2 assessment and those who did not on demographic or other study variables. However, since it is still possible that participants were not missing at random, replication of these results is important. Second, there was a relatively low recruitment rate relative to the pool of eligible dyads. Though we were not able to collect data on reasons for refusal, many families may experience their children’s disclosure and subsequent involvement in a variety of agencies as stressful. Thus, they may be more likely to turn down research opportunities they perceive as adding to these stressors. A third limitation is the reliance on mother and child ratings of maternal support that have not yet been shown to predict observations of maternal support following CSA disclosure. Thus, there is a possibility that these reports are biased in terms of their representations of actual maternal behavior. Finally, given the relatively small sample size and multiple measures of maternal support used in the analyses, there is some chance that results are unstable. Though the two measures of maternal support utilized in this study have multiple strengths, they were not designed to yield global scores; thus, condensing the measures into single mother-reported and child-reported measures for the sake of decreasing the number of analyses was not possible. Thus, the results represent initial findings and require replication with other samples. Despite these limitations, strengths of the study include the use of multiple dimensions of maternal support with both child and mother ratings as well as the recruitment of a sample of youth receiving services from a CAC soon after CSA disclosure.
Results have important clinical implications. First, following CSA disclosure, both children’s and mother’s perceptions of maternal support are relevant factors to assess. Second, as maternal support was related to child adjustment, both concurrently and 9-months later, it may be an important target of intervention. This may be particularly true in the case of depression, given the prospective relations found between mother-rated support at time 1 and child-rated depression at time 2. As many widely used treatments for CSA involve interventions with caregivers, targeting maternal support specifically may be a valuable tool for increasing the effectiveness of the intervention for the child. Though many therapeutic strategies aim to increase maternal support, less is known about whether support does indeed increase over the course of therapy, which aspects of maternal support are likely to change, and whether mothers’ and children’s perceptions of change in maternal support over the course of treatment are both relevant in predicting therapeutic outcomes for children. The use of multi-rater measures like those used in this study would facilitate this kind of treatment outcome research.
Acknowledgements
Preparation of this manuscript was supported by the National Institute on Drug Abuse through grant number K23DA034879 (PI: Kristyn Zajac) and by the National Institute of Mental Health through grant number T32MH018869 (PI: Dean G. Kilpatrick). The original research was supported by funding from the US Department of Health and Human Services Administration on Children and Families (PI: Daniel W. Smith).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Achenbach TM, Rescorla LA. Manual for the ASEBA school-age forms & profiles. Burlington: University of Vermont Research Center for Children, Youth, & Families; 2001. [Google Scholar]
- Adams-Tucker C. Proximate effects of sexual abuse in childhood: A report on 28 children. American Journal of Psychiatry. 1982;139:1252–1256. doi: 10.1176/ajp.139.10.1252. [DOI] [PubMed] [Google Scholar]
- Bick J, Zajac K, Ralston ME, Smith D. Convergence and divergence in reports of maternal support following childhood sexual abuse: Prevalence and associations with youth psychosocial adjustment. Child Abuse & Neglect. 2014;38(3):479–487. doi: 10.1016/j.chiabu.2013.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Briere J. Trauma Symptom Checklist for Children (TSCC) Odessa, FL: Psychological Assessment Resources; 1996. [Google Scholar]
- Cohen JA, Mannarino AP. Factors that mediate treatment outcome of sexually abused preschool children: Six and 12-month follow-up. Journal of the Academy of Child and Adolescent Psychiatry. 1998;37:44–51. doi: 10.1097/00004583-199801000-00016. [DOI] [PubMed] [Google Scholar]
- Cohen JA, Mannarino AP. Predictors of treatment outcome in sexually abused children. Child Abuse & Neglect. 2000;24:983–994. doi: 10.1016/s0145-2134(00)00153-8. [DOI] [PubMed] [Google Scholar]
- Conte JR, Schuerman JR. Factors associated with an increased impact of child sexual abuse. Child Abuse & Neglect. 1987;11:201–211. doi: 10.1016/0145-2134(87)90059-7. [DOI] [PubMed] [Google Scholar]
- Crouch JL, Smith DW, Ezzell CE, Saunders BE. Measuring reactions to sexual trauma among children: Comparing the CITES-R and the TSCC. Child Maltreatment. 1999;4:254–262. [Google Scholar]
- Elliott AN, Carnes CN. Reactions of nonoffending parents to the sexual abuse of their child: A review of the literature. Child Maltreatment. 2001;6:314–331. doi: 10.1177/1077559501006004005. [DOI] [PubMed] [Google Scholar]
- Esparza D. Maternal support and stress response in sexually abused girls ages 6–12. Issues in Mental Health Nursing. 1993;14:85–107. doi: 10.3109/01612849309006892. [DOI] [PubMed] [Google Scholar]
- Everson MD, Hunter WM, Runyon DK, Edelsohn GA, Coulter ML. Maternal support following disclosure of incest. American Journal of Orthopsychiatry. 1989;59:197–207. doi: 10.1111/j.1939-0025.1989.tb01651.x. [DOI] [PubMed] [Google Scholar]
- Feiring C, Taska LS, Lewis M. Social support and children’s and adolescents’ adaptation to sexual abuse. Journal of Interpersonal Violence. 1998;13:240–260. [Google Scholar]
- Finkelhor D, Browne A. Traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry. 1985;55:530–541. doi: 10.1111/j.1939-0025.1985.tb02703.x. [DOI] [PubMed] [Google Scholar]
- Grasso DJ, Saunders BE, Williams LM, Hanson R, Smith DW, Fitzgerald MM. Patterns of multiple victimization among maltreated children in navy families. Journal of Traumatic Stress. 2013;26:597–604. doi: 10.1002/jts.21853. [DOI] [PubMed] [Google Scholar]
- Hebert M, Collin-Vezina D, Daigneault I, Parent N, Tremblay C. Factors linked to outcomes in sexually abuse girls: A regression tree analysis. Comprehensive Psychiatry. 2006;47:443–455. doi: 10.1016/j.comppsych.2006.02.008. [DOI] [PubMed] [Google Scholar]
- Heriot J. Maternal protectiveness following the disclosure of intrafamilial child sexual abuse. Journal of Interpersonal Violence. 1996;11:181–194. [Google Scholar]
- Hsu E, Smith DW. Relations among maternal support, maternal distress, and child functioning following disclosure of incest; Paper presented at the San Diego Conference on Responding to Child Maltreatment; San Diego, CA. 2000. Jan, [Google Scholar]
- Kawachi I, Berkman LF. Social ties and mental health. Journal of Urban Health. 2001;78:458–467. doi: 10.1093/jurban/78.3.458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lanktree CB, Gilbert AM, Briere J, Taylor N, Chen K, Maida CA, Saltzman WR. Multi-informant assessment of maltreated children: Convergent and discriminant validity of the TSCC and TSCYC. Child Abuse & Neglect. 2008;32(6):621–625. doi: 10.1016/j.chiabu.2007.10.003. [DOI] [PubMed] [Google Scholar]
- Lovett BB. Child sexual abuse: The female victim’s relationship with her nonoffending mother. Child Abuse & Neglect. 1995;19:729–738. doi: 10.1016/0145-2134(95)00030-c. [DOI] [PubMed] [Google Scholar]
- Lovett BB. Child sexual abuse disclosure: Maternal response and other variables impacting the victim. Child and Adolescent Social Work Journal. 2004;21:355–371. [Google Scholar]
- Mannarino AP, Cohen JA. Family-related variables and psychological symptom formation in sexually abused girls. Journal of Child Sexual Abuse. 1996;5:105–120. [Google Scholar]
- Moak ZB, Agrawal A. The association between perceived interpersonal social support and physical and mental health: Results from the national epidemiological survey on alcohol and related conditions. Journal of Public Health. 2010;32:191–201. doi: 10.1093/pubmed/fdp093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morrison NC, Clavenna-Valleroy J. Perceptions of maternal support as related to self-concept and self-report of depression in sexually abused female adolescents. Journal of Child Sexual Abuse. 1998;7:23–40. [Google Scholar]
- Quas JA, Goodman GS, Jones DPH. Predictors of attributions of self-blame and internalizing behavior problems in sexually abused children. Journal of Child Psychology and Psychiatry. 2003;44:723–736. doi: 10.1111/1469-7610.00158. [DOI] [PubMed] [Google Scholar]
- Rosenthal S, Feiring C, Taska L. Emotional support and adjustment over a year’s time following sexual abuse discovery. Child Abuse & Neglect. 2003;27:641–661. doi: 10.1016/s0145-2134(03)00104-2. [DOI] [PubMed] [Google Scholar]
- Sadowski CM, Friedrich WN. Psychometric properties of the Trauma Symptom Checklist for Children (TSCC) with psychiatrically hospitalized adolescents. Child Maltreatment. 2000;5(4):364–372. doi: 10.1177/1077559500005004008. [DOI] [PubMed] [Google Scholar]
- Sirles EA, Franke PJ. Factors influencing mothers’ reactions to intrafamily sexual abuse. Child Abuse and Neglect. 1989;13:131–139. doi: 10.1016/0145-2134(89)90036-7. [DOI] [PubMed] [Google Scholar]
- Smith DW, Sawyer GK, Jones L, Cross T, McCart MR, Ralston EM. Mother reports of maternal support following child sexual abuse: Preliminary psychometric data on the Maternal Self-report Support Questionnaire (MSSQ) Child Abuse & Neglect. 2010;34:784–792. doi: 10.1016/j.chiabu.2010.02.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith DW, Sawyer GK, Heck NC, Zajac K, Self-Brown S, Danielson CK, Ralston EM. Psychometrics of a child report measure of maternal support following disclosure of sexual abuse. 2014 doi: 10.1080/10538712.2017.1296915. Manuscript under review. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spaccarelli S. Stress, appraisal, and coping in child sexual abuse: A theoretical and empirical review. Psychological Bulletin. 1994;116:340–362. doi: 10.1037/0033-2909.116.2.340. [DOI] [PubMed] [Google Scholar]
- Tremblay C, Hebert M, Piche C. Coping strategies and social support as mediators of consequences in CSA victims. Child Abuse and Neglect. 1999;23:929–945. doi: 10.1016/s0145-2134(99)00056-3. [DOI] [PubMed] [Google Scholar]
- Ullman SE. Social reactions to child sexual abuse disclosures: A critical review. Journal of Child Sexual Abuse. 2003;12:89–121. doi: 10.1300/J070v12n01_05. [DOI] [PubMed] [Google Scholar]