Posttraumatic Stress Symptoms and Correlates in Women and Children From Backgrounds of Domestic Violence

Abstract

Ongoing research has continued to inform our understanding of the effects of living with domestic violence on both women and children. The majority of this research, however, has tended to focus on each population separately, with only a relatively few studies to date assessing the relationship between maternal and child emotional functioning, particularly for symptoms of posttraumatic stress (PTSS). This study was designed to investigate trauma symptomatology in mother–child dyads from backgrounds of domestic violence, where the children are able to self-report on their own symptoms. In addition, the study examined anxiety and depression as important correlates of PTSD in children. Participants were recruited by staff at two metropolitan Domestic Violence Services and interviewed by the first author using standardised PTSD scales and trauma inventories. Results found no significant relationship between trauma symptoms in the mother and those in her child. With respect to children who met the criteria for a diagnosis of posttraumatic stress disorder (PTSD), they were more likely to have higher levels of anxiety and depression when compared to children who did not meet PTSD criteria. Results suggest that the emotional responses of older children may tend to reflect their own experiences rather than being a reflection of maternal distress as seems more likely with younger children. Implications of these findings include the importance of independent assessments of older children, and that older children may profit from early therapeutic interventions in their own right.

When women who have been abused by their partners seek refuge in a shelter they usually bring their children with them. As it was historically the women who were, understandably, seen most in need of protection and support, the accompanying children ran the risk of becoming the “unacknowledged” or “silent” victims of domestic violence (Holden, 1998). The oversight of children as being potentially equally affected by their experiences was compounded by the fact that domestic violence was seen as occurring within a dyad and not within a whole family. Anderson and van Ee (2018) remind us that children are often equal victims in experiencing violence directly, for example in the form of emotional, physical, or sexual abuse from their parent perpetrator, and that there is often a fine margin where children move from being secondary victims to being more directly caught up in the family violence.

Definitions of domestic violence (Hegarty et al., 2000; Keynejad et al., 2021) describe a pattern of controlling, coercive, and threatening behaviours between current or former intimate partners, with these behaviours encompassing physical, sexual, psychological and financial aspects. Experiencing such abusive behaviours typically results in a range of physical and mental health problems, prominent amongst which is the emergence of symptoms of posttraumatic stress (PTSS). Fully diagnosable posttraumatic stress disorder (PTSD) as a consequence of living with domestic violence was first reported in women from the late 1990s (e.g., Jones et al., 2001), and consistently since (e.g., Pill et al., 2017). PTSD in children had similarly been reported from this same period (e.g., Mertin & Mohr, 2002), and subsequently (e.g., Graham-Berman et al., 2012).

Ongoing research has continued to inform our understanding of the nature of domestic violence, and of the effects on both women and children of living with such violence in the home. The majority of this research, however, has tended to focus on each population separately. The primary aim of the present study was therefore to investigate the relationship between PTSS in the mother and those in her child.

Only a relatively few studies to date have assessed the relationship between maternal and child functioning with particular reference to symptoms of posttraumatic stress (PTSS), with most of this research being conducted with populations of young children (e.g., Bogat et al., 2006; Lieberman et al., 2005; Samuelson & Cashman, 2009).

Bogat et al. (2006) interviewed 48 mothers who had experienced domestic violence during the first year of their infant’s life, and 18 (37%) met DSM-IV criteria for PTSD assessed on the PTSD Scale for Battered Women (Saunders, 1994). The authors reported finding a significant relationship between maternal and infant trauma symptoms (PTSS), but only when the mothers had experienced severe violence. The authors added that maternal depression was unrelated to whether or not infants exhibited trauma symptoms suggesting that the infants were sensitive to, and reacting to, maternal anxiety-based arousal.

A relationship between maternal PTSD and child behaviour problems had similarly been found by Lieberman et al. (2005). The participants were 85 mother–child dyads, with the children in this study ranging in age from 25 to 59 months. Inclusion criteria included that the children had been witness to incidents of domestic violence. The incidence of maternal PTSD was not reported, but the authors found a positive correlation between traumatic stress symptoms (PTSS) and maternal life stress, and a negative correlation with the quality of the mother–child relationship. The severity of child behaviour problems, however, was uniquely predicted by maternal PTSD, indicating that, while maternal life stresses may be considered as a risk factor predisposing children to behavioural problems, the authors concluded that the generative mechanism actualizing this risk was the mother’s psychological response to the stress of the domestic violence.

Maternal PTSS as a consequence of exposure to domestic violence acting as a predictor of child emotional and behavioural difficulties were further explored by Samuelson and Cashman (2009) using a sample of 30 community-based, mother–child dyads. In this study children’s ages ranged from five to 18 years. As with the Lieberman et al. (2005) study the incidence of maternal PTSD was not reported, but consistent with the above study, Samuelson and Cashman reported a clear relationship between maternal PTSS and their child’s emotional regulation difficulties, as well as a number of Internalizing and Externalizing behaviour problems as assessed on the Child Behavior Checklist (Achenbach, 1991).

The above results are seen as being consistent with the theory of Relational PTSD as proposed by Scheeringa and Zeanah (2001). These authors hypothesised that, in relation to young children at least, the mother’s emotional distress is seen as compromising the infant’s developing abilities to self-regulate their own emotions due to the close emotional relationship with the mother. In support, Levendosky and colleagues (Levendosky et al., 2013), found a high co-occurrence of maternal and child trauma symptoms in a sample of children, aged one to seven years, who had been exposed to domestic violence. In particular, maternal symptoms of arousal were associated with their child’s arousal symptoms. The authors therefore considered that their findings indicated that “young children are responding directly to their mother’s affective dysregulation in reaction to IPV as well as demonstrating similar dysregulation” (p.10).

A similar conclusion was reached in a recent study into dyadic profiles of PTSS in mothers and children led by Maria Galano (Galano et al., 2020), namely, that Relational PTSD was potentially the norm, not the exception, in children and their mothers exposed to intimate partner violence. Correlations between mothers’ and children’s PTSS were assessed in a sample of 231 mother–child dyads, with children ranging from ages four to 12 years. For all children PTSS were measured using care-giver report measures; for children four to seven years with the Posttraumatic Stress Disorder Semi-Structured Interview (Scheeringa & Zeanah, 1994), and for children older than seven years with the Traumatic Stress Symptom Child Scale (Graham-Berman & Levendosky, 1998). Both questionnaires assessed PTSS using DSM-IV-TR criteria.

For children aged four to seven years significant positive correlations with their mothers’ PTSS were found for all three symptom domains (re-experiencing, avoidance, and symptoms of hyperarousal), as well as for total PTSS. For children over seven years significant correlations were found for symptoms of hyperarousal as well as for total PTSS. A similar result was found in diagnostic rates of PTSD in mothers and children, where a significant association was found for the younger age group; that is, it was significantly more likely for a child to have a diagnosis of PTSD if the mother also had a diagnosis of PTSD. This association was not found for the older age group. Nevertheless, the authors considered results indicated that, in general, children exposed to family violence do not experience good mental health if their mothers are struggling with their own mental health.

Limitations of Research

In discussing their results Samuelson and Cashman (2009) felt that the heavy reliance on maternal reports of their child’s emotional and behavioural functioning represented a major limitation of their study as the ability of mothers to objectively report on their child’s functioning, even within a framework provided by a structured interview, at time when their own mental health is adversely affected, should not be taken for granted. Johnson and Lieberman (2007) had previously acknowledged that, while the reliance on maternal reports is common in studies investigating the effects on young children of experiencing domestic violence, they nevertheless cautioned that maternal perceptions of their children’s adjustment may be influenced by their own recent traumatic experiences.

As an illustration of the above, a study by Shemesh and colleagues (Shemesh et al., 2005), examining the comparisons of parent and child reports of trauma symptoms in a paediatric outpatient setting found low levels of correlation between the two. In a sample of 76 children and adolescents aged eight to 19 years, 14% met the full DSM-IV criteria for PTSD, however it was the children’s self-reports of their own symptoms that were significantly correlated with a diagnosis of PTSD. In contrast, the parent’s report was not significantly correlated with their child’s diagnosis. In addition, a more recent study by Meiser-Stedman et al. (2017) assessed more than 100 children between the ages of two and 10 years of age who had been involved in a road accident. Similar to the findings of Shemesh et al. (2005) these authors also found, inter alia, with respect to older children who could report on their own symptomatology, that parent–child agreement for PTSD was no better than chance.

Research with Child’s Own Self-report

A study investigating the relationship between maternal and child PTSD, where both mothers and children had been exposed to domestic violence, and using the child’s own self-report of their symptomatology, was conducted by Chemtob and Carlson (2004), using a small (N = 25) sample of mother–child dyads, with the children ranging in age from seven to 17 years (mean age = 11 years). The mothers were assessed on the Posttraumatic Diagnostic Scale (PTDS; Foa et al., 1997), and the Beck Depression Inventory (BDI; Beck et al., 1998), while the children were assessed using the Clinician-Administered PTSD Scale for Children and Adolescents (CAPS-CA; Nader et al., 1996).

The psychological measures documented high rates of diagnosed PTSD amongst the mothers (50%) and the children (40%). PTSD status in the mothers significantly predicted higher scores on the BDI but, unlike the findings in the above three studies reviewed, results in this study found no apparent relationship between the likelihood of PTSD in the mothers and the likelihood of PTSD in their child. In a related finding the authors mentioned that the mothers with PTSD tended to underestimate the distress experienced by their children that underlined, in the authors’ view, the need for independent psychological assessments of mothers and children, particularly in order to determine the most appropriate mental health interventions for each.

The Current Study

Despite the small number of studies conducted in this area, and the variation in study populations and assessment tools used, research focusing on mother–child dyads is important in continuing to remind clinicians that mothers and their children may not only be significantly affected by their experiences of living with domestic violence, but that, as domestic violence occurs within families, there is the potential for damage to the maternal-child relationship. In addition, the potential importance that such information may provide in determining the most appropriate therapeutic interventions for both the mother and her child(ren) should not be underestimated. The primary aim of this study was to investigate the relationship between PTSS in the mother and those in her child, where children are able to self-report their own symptoms. In addition, the study examined anxiety and depressive symptoms as important corelates of PTSD in children with a background of domestic violence.

Method

Participants

The sample comprised 50 mother and child dyads who were recruited through two metropolitan Domestic Violence Services in Adelaide, South Australia. At time of interview all women and children were living in DV Services sponsored community housing and receiving a range of community and therapeutic supports arranged through the Services. The mean time of separation was 13 months (SD = 12 months). The mothers had a mean age of 39 years (SD = 6.4 years). The mean length of relationship was 10.8 years (SD = 6.8 years), and for 20 women this was their first relationship. Of the 30 women who had been in previous relationships, 37% reported experiencing some form of domestic violence in the past. The majority of women identified themselves as being of Anglo-Australian descent, two women identified as Greek, and with one each as Aboriginal, Indian, Philippina, and Portuguese.

The 50 children (28 males and 22 females) were aged between eight to 16 years with a mean age of 11 years (SD = 2.6 years).

Measures

The assessment battery consisted of the following measures:

Posttraumatic Stress Diagnostic Scale-PDS (Foa, 1995)

The PDS is a 49-item self-report instrument designed to aid in the diagnoses of Posttraumatic Stress Disorder (PTSD), with the structure and content of the PDS mirroring the DSM-IV diagnostic criteria for PTSD. The 17 trauma symptoms contained in criteria B (re-experiencing), C (active avoidance and numbing of responsiveness), and D (symptoms of increased arousal) can be scored categorically as well as on a scale from 0 to 3 (where 0 means not at all or only one time, and 3 means 5 or more times per week/almost always), thus providing a measure of symptom severity. Symptom severity scores of 1-10 are rated as Mild, 11-20 are rated as Moderate, 21-35 are rated as Moderate to Severe, while scores greater than 36 are rated as Severe. A diagnosis of PTSD was made in accordance with DSM-IV guidelines; that is, at least one endorsed symptom for criterion B, three symptoms endorsed for criterion C, and two symptoms for criterion D. The PDS was normed on a sample of 248 individuals aged between 18 to 65 years. The Manual reports test-retest stability over a period of 10 to 22 days of 87%. In addition, a Cronbach alpha of 0.92 was calculated for the 17 items on which the Symptom Severity score was based. A Cronbach alpha for the present sample was calculated at 0.75.

Trauma Symptom Inventory- Second Edition (TSI-2-A; Briere, 2011)

The TSI-2 is a self-report measure of posttraumatic stress intended for use in the evaluation of acute and chronic trauma symptomology. This study used the altered version (referred to as the TSI-2-A), which is an 86-item version of the TSI-2, differing from the TSI-2 only in that it does not contain critical items pertaining to sexual abuse. The TSI-2-A contains seven clinical scales: Anxiety, Depression, Intrusive Experiences, Anger, Defensive Avoidance, Somatic Preoccupations and Dissociation. Symptom items within each scale are rated according to their frequency of occurrence, where 0 indicates the symptom is not experienced, to 3, indicating the symptom is experienced often. Scores for each scale are converted to standard T-scores for comparison to the normative population, with T-scores over 65 considered clinically significant. The inventory was normed on 836 subjects, having a mean age of 47.3 years (SD = 16.2 years). The Manual reports good internal consistency for all scales ranging from 0.74 to 0.94, and test–retest coefficients between 0.76 and 0.94. Cronbach alphas for the present sample ranged from 0.80 (Intrusive Experiences scale) to 0.89 (Dissociation scale).

The Child PTSD Symptom Scale (CPSS; Foa et al., 2001)

The CPSS was developed to assess the severity of Posttraumatic Stress Disorder (PTSD) symptoms in children exposed to trauma. The CPSS contains the same 17 PTSD symptom items contained in criteria B, C, and D of the PDS (Foa, 1995), and similar to the PDS, the symptoms can be scored categorically as well as on a scale from 0 to 3, providing the same measures of symptom severity; that is, from Mild through to Severe. As with the PDS, a diagnosis of PTSD requires meeting the same DSM-IV guidelines; that is, at last one symptom endorsed for criterion B, three symptoms for criterion C, and two symptoms for criterion D. The CPSS is thus able to provide a close comparison to the scores on the PDS. The CPSS was normed on 75 children and adolescents aged eight to 15 years who had experienced the Northridge California earthquake. Foa et al. (2001), and later Nixon et al. (2013), reported high internal consistency of 0.89, as well as good test–retest reliability of 0.84. A Cronbach alpha for the present sample was calculated at 0.83.

Trauma Symptom Checklist for Children (TSCC-A; Briere, 1996)

The Manual describes the TSCC as “a self-report measure of posttraumatic distress and related psychological symptomatology” (p.1), and is intended for use in the evaluation of children eight to 16 years of age who have experienced traumatic events. This study used the altered version (referred to as the TSCC-A) which has 44 items and is identical to the full version of the TSCC with the exception that it contains no sexual abuse items. The TSCC-A contains five clinical scales: Anxiety, Depression, Posttraumatic Stress, Anger, and Dissociation. Examples of questions assessing for trauma responses include: “Bad dreams or nightmares”, “Remembering things I don’t want remember”, and “Feeling nervous or jumpy inside”. Similar to the TSI, symptom items are rated according to their frequency of occurrence, where 0 indicates the symptom is not experienced, to 3, indicating the symptom is experienced nearly all the time. Scores for each scale are converted to standard T-scores for comparison to the normative population, with T-scores over 65 considered clinically significant. The inventory was normed on 3,008 children drawn from urban and suburban locations in Illinois, Colorado, and Minnesota. The norms were based upon four age by sex combinations: males and females of eight to 12 years, and males and females of 13 to 16 years. The Manual reports good reliability and validity, with reliability coefficients for the five clinical sub-scales ranging from 0.82 to 0.89. Cronbach alphas for the present sample ranged from 0.62 (Posttraumatic Stress scale) to 0.90 (Anger scale).

Procedure

The mothers and children were recruited by staff at the Domestic Violence Services who described the nature and purpose of the study. Only women who were no longer in crisis or in ongoing danger from their spouse/partner, and who had moved to community-based housing were deemed eligible for recruitment. Those mothers and children who consented to participate were interviewed by the first author at one of the participating Services, during a time when the mothers had attended for ongoing counselling or general support. The confidential nature of the study was explained, and that no identifying information would be collected or recorded. All interviews took place over a period of some 12 months. The mother and the child were interviewed separately. The mothers were asked for basic information about their relationship, such as the length of the relationship, time since separation, and if they had feared being killed at any stage in the relationship (a simple yes/no answer). To avoid distress the DV Services requested that no specific questions about the nature of the abuse they or their children had experienced in the relationship were to be asked, although general discussion with the mothers during the interviews indicated that all the children involved had at least witnessed violence in the home. The mothers were administered the two questionnaires (PDS and TSI-2-A) as per instructions from the respective Manuals, with the entire assessment process taking approximately 20 min to complete.

The children were similarly asked for basic information about themselves, such as their age, and if they had a fear of their mother being seriously injured/killed (also a simple yes/no question). The children were then administered the two questionnaires (CPSS and the TSCC-A). In order to maintain a uniform approach the questionnaires were administered in structured interview format with no adaptations for younger children. The entire assessment process took approximately 20 min to complete. Participants were then debriefed and thanked for their interest and involvement. No remuneration was offered to the participants. The collection and analysis of unidentified interview data received ethics approval from the University of South Australia Human Research Ethics Committee.

Analytic Plan

Recorded interview responses and assessment measures were scored in accordance with instructions in the respective Manuals (with respect to the PTSD questionnaires, in accordance with DSM-IV guidelines for criteria B, C and D as described above), coded, entered, and analysed in the statistical program SPSS Version 24. Pearson’s chi-square tests of contingencies were used to determine if there was a significant relationship between two nominal (categorical) variables. Bivariate Pearson’s product-moment correlations were used to evaluate relationships between two continuous-level variables. The significance level (α) was set to 0.05 for all tests.

Results

Mothers’ Emotional Functioning

Forty-five (90%) mothers who experienced domestic violence met the criteria for a diagnosis of PTSD as assessed by the Posttraumatic Stress Diagnostic Scale (PDS). Mothers in this sample recorded a mean Symptom Severity Score of 25.6 (SD = 7.4), indicating a Moderate to Severe presentation of PTSS, with six women recording scores greater than 35 (Severe category). Thirty-six (72%) women reported fearing being killed on at least one occasion during the relationship. A Pearson’s chi-square test of contingencies (with α = 0.05) was used to evaluate if there was an association between mother’s who met the diagnosis of PTSD and fear of being killed. Results found a significant relationship (X² (1, N = 50) = 7.45, p < 0.006), although the association between these variables was moderate in effect size (ɸ = 0.39). In addition, a significant relationship was found between the severity of mother’s PTSD symptoms and fear of being killed (r = 0.51, p < 0.001).

Children’s Emotional Functioning

Twenty-four (48%) children met criteria for diagnosis of PTSD as assessed by the Child PTSD Symptom Scale (CPSS). The mean Symptom Severity Score of 16.3 (SD = 7.8) was rated as being in the Moderate range, with 28 (56%) children reporting that they had a fear of their mother being seriously injured or killed. A chi-square test did not find a significant relationship between a child fulfilling the diagnostic criteria for PTSD and fear of the mother being seriously injured/ killed, (X²(1, N = 50) = 2.13, p = 0.144), although a significant relationship was found between the severity of child’s PTSD symptoms and fear of the mother being seriously injured/killed (r = 0.299, p = 0.035).

Table 1 reports symptoms interfering in day-to-day functioning for both mothers and their children, rated as either a 2 (2 to 4 times a week/half the time) or 3 (5 or more times a week/almost always) for criteria B, C and D on the PDS and CPSS respectively. This table indicates that there were insufficient symptoms reported in common for there to be a significant relationship for diagnosable PTSD between the mothers and children, although there were similarities in a number of reported symptoms, the most noticeable being feelings of irritability/anger, a marked diminished interest in taking part in activities, and conscious efforts to avoid thoughts and feeling associated with past violence.

Table 1.

Number and percentage of mothers and children endorsing troubling posttrauma symptoms*

Items	Mothers		Children
	n	%	n	%	p
Re-experiencing of events
1. Recurrent distressing thoughts	35	70	22	44	0.002
2.  Recurrent distressing dreams	22	44	14	28	0.055
3. Acting or feeling as if event(s) recurring	3	5	2	4	0.596
4. Feeling upset when think or hear about the event	37	74	23	46	0.001
5.  Having feelings in body when think or hear about event	24	48	5	10	0.001
Persistent avoidance and numbing
1. Efforts to avoid thoughts and feelings	36	72	32	64	0.176
2. Efforts to avoid activities, places, etc	25	50	4	8	0.001
3. Inability to recall	1	2	0	0	0.300
4. Marked diminished interest/participation	16	32	12	24	0.272
5. Feelings of detachment	23	46	7	14	0.001
6. Restricted range of affect	22	44	15	30	0.089
7. Sense of foreshortened future	28	56	13	26	0.001
Persistent feelings of increased arousal
1. Difficulty falling or staying asleep	29	58	19	38	0.020
2. Irritability or outbursts of anger	23	46	24	48	0.927
3. Difficulty concentrating	32	64	22	44	0.017
4. Hypervigilance	34	68	27	54	0.062
5. Exaggerated startle response	31	62	17	34	0.002

N = 50

^*Table reports endorsing of symptoms as a severity score of 2 (2 to 4 times a week/half time) or 3 (5 or more times a week/almost always) for Criteria B, C and D on the PDS and CPSS scales

The relationship between mother and child’s emotional functioning was further explored using a bivariate Pearson’s product-moment correlation coefficient. No significant association was found between mother and child’s severity of PTSD symptoms (r = 0.145, p = 0.314). Similar findings were also found when comparing the association of PTSD symptom criteria B, C, and D, between mother and child. The mother’s symptoms of re-experiencing (Criterion B) did not predict the child’s symptomatology (r = -0.121, p = 0.402), nor did the mother’s symptoms of arousal (Criterion D) relate to the child’s symptoms of arousal (r = 0.200, p = 0.164). Lastly, the mother’s symptoms of avoidance and numbing of responsiveness (Criterion C) were not associated with those of her child (r = 0.160, p = 0.268).

Similarly, the mother’s reported anxiety, as rated on the TSI-2-A was not associated with their child’s rating of anxiety (TSCC-A) (r = 0.105, p = 0.467), and neither did the mother’s rating of depression (TSI-2-A) predict their child’s depression (TSCC-A) (r = 0.271, p = 0.57).

Table 2 shows children who met the diagnostic criteria for PTSD were more likely to have higher mean symptom severity scores in all areas assessed when compared to children not meeting diagnostic criteria for PTSD. No significant gender differences were found, apart from Depression, where males reported higher mean scores (p = 0.006). An exploratory analysis between adolescents (12 years and older, mean age 13.7 years, n = 19) and younger children (younger than 12 years, mean age 9.1 years, n = 31), did not reveal any statistically significant differences (all p > 0.15) for all symptom severity scores (see Table 3).

Table 2.

Mean symptom severity scores for PTSD vs non-PTSD children

	PTSD (n = 24)		Non-PTSD (n = 26)
Scale	M	SD	M	SD	P
Criterion Ba	5.17	2.48	3.46	2.18	0.013
Criterion Ca	7.13	2.95	3.42	2.30	0.001
Criterion Da	8.08	3.88	5.57	3.56	0.021
Anxietyb	59.17	12.58	49.38	7.39	0.001
Depressionb	51.75	11.99	45.58	9.04	0.044
Post-Traumatic Stressb	54.71	7.32	48.81	6.81	0.005

^aMean symptoms severity scores (CPSS)

^bMean T-Scores (TSCC-A)

Table 3.

Mean symptom severity scores for Adolescents (12 years and older) vs younger children (less than 12 years of age)

	Adolescents (≥ 12 years; n = 19)		Younger Children (< 12 years; n = 31)
Scale	M	SD	M	SD	P
Criterion Ba	3.632	2.409	4.677	2.441	0.146
Criterion Ca	5.158	3.167	5.226	3.283	0.943
Criterion Da	6.000	3.873	7.258	3.881	0.271
Anxietyb	52.053	7.067	55.323	13.131	0.323
Depressionb	49.579	8.720	47.903	12.142	0.603
Post-Traumatic Stressb	50.895	6.806	52.097	8.109	0.592

^aMean symptoms severity scores (CPSS)

^bMean T-Scores (TSCC-A)

Table 4 shows the relationships between the mean symptom severity scores for the CPSS and the TSCC-A. All variables were significantly correlated with each other.

Table 4.

Pearson’s correlations for child variables displayed in Table 2

Variable	1	2	3	4	5	6
1. Criterion Ba	—
2. Criterion Ca	0.386**	—
3. Criterion Da	0.555***	0.544***	—
4. Anxietyb	0.548***	0.562***	0.781***	—
5. PTSb	0.739***	0.497***	0.538***	0.638***	—
6. Depressionb	0.407	0.538***	0.580***	0.526***	0.351*	—

^aMean symptoms severity scores (CPSS)

^bMean T-Scores (TSCC-A)

^*p < .05. **p < .01. ***p < .001

Discussion

There have been few studies investigating the relationship between trauma symptoms in mothers and their children from backgrounds of domestic violence. Levendosky et al. (2013) argued that because domestic violence is a unique interpersonal stressor, it is important to understand the child’s trauma symptoms in the context of the mother’s traumatic response. Whereas these studies have all demonstrated a relationship between maternal and child PTSS, they have tended to rely upon maternal reports of their child’s emotional and behavioural functioning, the reliability of which has been questioned (e.g., Johnson & Lieberman, 2007; Samuelson & Cashman, 2009). The present study, therefore, was designed to investigate the relationship between maternal and child PTSS where the children involved were old enough to report on their own symptoms with the use of standardized scales designed for the assessment of trauma as well as other emotional symptoms – principally those of anxiety and depression. Consistent with the Chemtob and Carlson (2004) study, the present study found no significant relationship in the incidence of PTSS between the mothers and their children, nor between reported levels of anxiety or depression.

Regardless of the fact that no relationship was found between the mother’s trauma symptoms and those of her child, results nevertheless indicated that both the women and the children in this study reported significantly elevated scores on the standardized measures used. With respect to the women, for example, results showed that 90% met diagnostic criteria for PTSD. This finding was higher than previous studies that found estimates of between 31 and 84% of similar women meeting diagnostic criteria for PTSD (Jones et al., 2001; Mertin & Mohr, 2000; Nathanson et al., 2012). There is no clear explanation for this finding except that direct comparison with other studies of the incidence of PTSD is difficult because of the diversity of populations across studies, and the lack of uniformity in the instruments used to assess the presence of PTSD. It is possible that the high incidence of women meeting criteria for PTSD in this study may be a function of the population of women currently being accepted by the Services involved. The high demand currently placed on Domestic Violence Services has resulted in these Services tending to focus on those women most seriously affected by their experiences as a first priority. The significant relationship, both between PTSD status and severity of symptoms, and a fear of being killed (reported by 72%), in this sample of women, for example, attests to the fact that, although not formally assessed, many of these women had most likely experienced significant levels of abuse and were most seriously affected by their experiences.

Forty-eight percent of the children in this study met criteria for a diagnosis of PTSD. The percentage of children who met this diagnosis was in the upper end of previous reports, which ranged between 13 and 50% (Chemtob & Carlson, 2004; Mertin & Mohr, 2002; Telman et al., 2016). While the symptom patterns of these children tended not to directly reflect the symptom patterns of their mothers, a salient finding was that the severity of the children’s symptoms was significantly related to the fear of their mother being seriously injured or killed. In addition, results confirmed that children who met diagnostic criteria for PTSD were more likely to have higher levels of more general emotional distress, such as anxiety, depression, and posttraumatic stress symptoms, as assessed on the TSCC-A, compared to those children who did not meet full criteria for PTSD. However, whether or not children met criteria for PTSD, symptom severity scores for the CPSS and the TSCC-A were all highly correlated within the child sample.

This study therefore provides further evidence of the adverse effects on both women and children—defined principally in terms of trauma symptomatology—of living with violence in the home environment. Where significant relationships found in studies with young children were explained as being consistent with the theory of Relational PTSD (Scheeringa & Zeanah, 2001), the results of the present study, together with those of Chemtob and Carlson (2004) are more suggestive of the fact that the emotional responses of older children may tend to reflect their own experiences.

No correlation was found between maternal and child trauma symptomatology suggesting that, with increasing cognitive maturity, older children may be better able to independently evaluate events occurring within their families and react in accordance with their own perceptions of threat and coping strategies. In addition, there is also the probability that relationships and activities outside the home, such as school, friendships and sport, may act as buffers against stressors in the home. Nevertheless, the very high percentage of the mothers meeting criteria for a diagnosis of PTSD and, in addition, the substantial number of children also meeting these criteria attests, once again, to the serious emotional consequences of living with domestic violence.

A particular strength of this study is that it relied on the children’s self-reports of their own symptomatology gathered within a structured interview using standardised measures. Listening directly to voices of children when conducting research into the effects of domestic violence has been viewed as crucial in gaining a better insight into their perceptions of the violence, their emotional responses to the violence, as well as their coping strategies (e.g., Noble-Carr et al., 2020), as children’s participation in research has important clinical consequences. Staf and Almqvist (2015) for example, remind us that it is the family situation that is pathological for these children, and that what might be interpreted as psychopathological in a child may sometimes be the most appropriate response available under the circumstances. Further, Bunston et al. (2017) have argued that the tendency to pathologize conditions over exploring causality runs the risk of missing important indicators of family violence which, in turn, increases the possibility of misdiagnosis and possibly inappropriate treatment options.

Limitations

There are, however, a number of limitations that must be acknowledged. First, the sample size is relatively small. Despite being double the size of the Chemtob and Carlson (2004) study sample, the results of this study must nevertheless be interpreted as provisional. In addition, due to the constraints placed on data gathering, there is an absence of information on the type or extent of violence experienced by the mothers and their children. Such experiences directly influence symptom severity and perceptions of threat to life; consequently, these results are able to provide only a broad indication of the effects of experiencing domestic violence on mother–child dyads.

Second, a cross-sectional approach was used, therefore nothing was learned about, for example, the precipitating factors contributing towards the final clinical presentations of these children. As Telman et al. (2016) pointed out, domestic violence rarely occurs in isolation, but generally occurs within a wider family context of disadvantage and dysfunction. These authors advised that in attempting to understand the variability in trauma symptoms among children exposed to domestic violence other stressful experiences, such as variations in safety and stability, financial privation, and parental mental health histories, alcohol and drug use, had to be taken into account. Thus, it is probable that the trajectory to the final clinical presentation in this study varied in each case depending upon the varying conditions in each household, meaning that it is not possible to discern from the results of this study which family variables may have contributed more significantly to the final outcome in each case. The inclusion of such variables would have placed the current results into greater focus.

Similarly, the cross-sectional approach provides no information about the ongoing progression of these children’s conditions, and what variables may be instrumental in helping to resolve, as opposed to maintaining, the current emotional difficulties experienced by these children. Previous research (e.g., Holt et al., 2008) has shown that the impact of domestic violence can be enduring, and that children and adolescents are at increased risk of developing ongoing emotional and behavioural problems, and of increased risk to exposure of other adversities in their lives. Such evidence underlines the importance of early intervention in order to reduce the risk of problems continuing on into adult life.

Conclusions

Taken together, results of research reviewed in this paper consistently report significant, moderate-to-large relationships between mothers’ and children’s PTSS, particularly in early childhood. Galano et al. (2020) observed, in relation to their study, that all mother–child dyads experienced relational PTSS, and varied only in the overall severity of symptoms across the dyad. These authors also found that maternal-child PTSS remained connected, albeit to a weaker extent, in older children. Such findings provided further evidence for these authors that parental PTSS is an important risk factor for childhood PTSS.

Results such as these also underscored the importance of considering maternal PTSS in the treatment of children’s PTSS. With respect to young children, researchers such as Bogat et al. (2006) and Levendosky et al. (2013) have advocated therapeutic interventions informed by the theory of Relational PTSD that attends to the needs of the mother, in particular providing safety and stability as a first priority, in order to reduce levels of maternal distress. Further assessment of the mother in these cases may then assist in determining what therapeutic intervention would be most beneficial for her and her child.

The present study found no significant relationship between mothers’ and children’s PTSS; a result possibly reflecting the fact that, unlike the above studies, the children tended to be older and self-reported on their own emotional functioning. This result may indicate a growing emotional independence in older children, and suggests that older children may benefit from some early trauma-informed therapeutic intervention in their own right. Anderson and van Ee (2018) cautioned, however, that positive maternal-child interactions may have been lost within the dynamics of family violence; an important factor needing to be addressed in any treatment approach. While acknowledging that little consensus regarding treatment and support for mothers and children exposed to intimate partner violence appears throughout the literature, based on their review of treatment interventions these authors nevertheless reported that interventions implementing a combination of separate and joint approaches seemed more successful in improving a wider range of outcomes.

To date, the present study, together with that of Chemtob and Carlson (2004), are the only two providing children’s self-reports using formal assessment measures for mental health outcomes. Although both studies are small and suffer from a number of limitations, they nevertheless reflect the importance Noble-Carr et al. (2020) placed on the need to design and conduct studies that allow children to report on their experiences about family violence. These authors argued that without access to such experiences it becomes more difficult for adults to understand and respond to children in ways that better allow them to recover and to build a sense of felt safety and emotional well-being. The present study adds to those that have allowed children to report on their perspectives about family violence (see, for example, Swanston et al., 2014). Given the growing knowledge on the impact of domestic violence on mother–child dyads, it is timely that further similar research now be conducted in this area. In particular, longitudinal study designs would better inform mental health professionals about the trajectory of trauma symptoms in mothers and their children (e.g., Galano et al., 2020), and provide better insights into what therapeutic approaches and service provisions may be beneficial in resolving symptoms of trauma, anxiety and depression consequent to experiencing domestic violence.

Declarations

Conflict of Interest

The authors declare no conflicts of interest.