Abstract
Approximately 27% of women worldwide have experienced physical and/or sexual intimate partner violence (IPV), with an estimated 10%–16% reporting such experiences within the past year. Furthermore, one in four women has been subjected to IPV before the age of 19 (Sardinha et al., 2022). In Sweden, 35% of women report having been victims of some form of interpersonal violence, and 15% report such experiences within the past year (BRÅ, 2024).
Sweden, together with other United Nations member states, has committed to the 2030 Agenda for Sustainable Development. Goal 5.2 within this agenda specifically calls for the elimination of all forms of violence against women and girls. While preventive measures, identification of risk factors, and addressing perpetrator behavior are crucial components in achieving this goal, the persistently high prevalence of IPV underscores the parallel need for comprehensive and accessible victim support services.
Despite the prevalence of IPV, only about one in five (19%) affected women seek help, with the most common sources being therapists or psychologists (BRÅ, 2024). In Sweden, support is available through approximately 200 girls’ and women's shelters (National Centre for Knowledge on Men's Violence Against Women, 2025) as well as via municipal social welfare services. Municipal social services in Sweden are legally obliged to provide victims of domestic violence with information, counseling, long-term housing solutions, parenting support, and assistance in navigating the welfare system (HSLF-FS 2025:40 (2025)). Previous research on specialized social welfare units indicates that commonly used interventions include financial and practical support (e.g., accompanying women to court proceedings), risk and safety measures, and psychoeducation and counselling (Ekström, 2018; Helmersson, 2017; Skillmark et al., 2019). Women's shelters provide services like accommodation, activities, counseling, and sustained support, including practical assistance (Ekström & Hvenmark, 2025).
A wide range of mental health problems following experiences of domestic violence has been consistently documented in the literature. Exposure to domestic violence is a strong predictor of depression, anxiety, and posttraumatic stress disorder (PTSD) (Spencer et al., 2019). Ferrari et al. (2014) examined the association between the severity of abuse and mental health outcomes in a UK sample of women exposed to domestic violence, identifying high levels of depression, anxiety, and PTSD among women seeking support services.
Similarly, a Spanish study by Vilariño et al. (2018) assessed mental health symptoms among women exposed to domestic violence using the Symptom Checklist-90 (SCL-90) and compared their scores with those of a normative sample. Women who had experienced domestic violence reported significantly higher levels of clinical symptoms across all assessed dimensions. In a Swedish study, Scheffer Lindgren and Renck (2008) examined the mental health of women after leaving violent relationships using the SCL-90. Participants completed self-report questionnaires measuring psychological distress. Their scores on the SCL-90 subscales of somatization, depression, and anxiety, as well as on the Global Severity Index (GSI), were significantly higher among participants than those of a nonclinical female population in Sweden. Moreover, scores for somatization, anxiety, and the GSI exceeded those observed in a female clinical comparison group. Treatment provided in womeńs shelters could be effective in reducing trauma symptoms and levels of PTSD (Johnson et al., 2020). Placement in a shelter can serve as a protective environment, and the sense of safety provided may contribute to improvements in PTSD symptoms (Dokkedahl et al., 2022). However, the secure environment can also entail unintended consequences. Survivors may face isolation from their social networks due to geographic distance and limited opportunities to maintain close relationships, including challenges in receiving visitors at the shelter (Stylianou et al., 2021). Exposure to stressful life events is associated with reduced self-esteem (Orth & Luciano, 2015), a pattern that is also evident among individuals subjected to domestic violence (Brunton & Dryer, 2024; Matheson et al., 2015). Experiences of domestic violence have been shown to substantially erode self-esteem and perceived self-worth (Matheson et al., 2015) and are further linked to the development of learned helplessness (Iftikhar et al., 2025). Self-esteem is positively correlated with perceived quality of life, suggesting broader implications for overall psychological well-being (Eklund et al., 2018). Although psychotherapeutic interventions for survivors of domestic violence have demonstrated clear benefits in reducing depressive symptoms, current evidence indicates that their effectiveness in improving self-esteem remains limited (Emirza & Uzun, 2024).
Prior research emphasizes the importance of therapists engaging in relationship-building. Key elements of effective support include demonstrating respect, fostering autonomy, and providing clear information, including information regarding violence. Furthermore, therapeutic competence expressed through nonverbal communication, such as body language and attentive presence, has been shown to convey care (Hargrave et al., 2024). A Swedish study evaluating four regional centers for domestic violence, which provide services to both perpetrators and victims, highlights the importance of professionals possessing specialized knowledge of violence and offering holistic support that includes practical assistance. Victims further emphasized the value of discussing different forms of violence, the process of normalization, and the dynamics of power and control. Through these discussions, participants reported gaining a deeper understanding of both the nature of violence and their own experiences (Eriksson et al., 2021).
To assess the value of interventions for women exposed to violence, studies using pre- and posttest designs are needed. However, studies examining changes in symptom burden following intervention remain limited. Even fewer studies have explored outcomes of support provided to victims of domestic violence across different service settings, such as social services and women's shelters. This study aims to address this gap in the literature.
Study Aim and Hypothesis
The aim of this study was to examine the backgrounds and forms of domestic violence experienced by the participating women, victims of domestic violence, and to assess whether support and interventions contributed to improved psychological health among victims who sought assistance from social welfare (SW) services or women's shelters (WS). In addition, the study sought to identify potential differences in outcomes based on the type of service provider the victims of domestic violence turned to for support.
The aim generated the following four hypotheses: Hypothesis 1: DV victims seeking support are exposed to severe forms of violence. Hypothesis 2: The participating women will report improved mental health at posttest. Hypothesis 3: The trauma symptoms will be reduced. Hypothesis 4: The service provided by DV specialists within social welfare and women's shelters will be experienced as satisfactory.
Method
To address the research questions, victims of domestic violence (DV) who had sought support either through specialized units within social welfare services or through women's shelters were invited to participate in the study. Six service units providing support to DV victims were approached and asked to distribute study invitations to potential participants. Of these units, three were specialized social welfare services and three were women's shelters. The participating social welfare units were located in municipalities in southern Sweden with populations ranging from 52,000 to 98,000 inhabitants, while the participating women's shelters were situated in municipalities with populations ranging from 18,000 to 98,000 inhabitants. The units offered various forms of support, such as counseling and regular treatment sessions over shorter or longer periods. In particular, women's shelters also provided a series of group sessions. At one unit, group sessions were conducted in collaboration with the social services. The units also offered practical support, for example, accompanying women to court proceedings. None of the participants stayed at a shelter where they received the support examined in this study.
The study employed a pre–post design, in which participants completed standardized assessment scales and took part in a structured interview on two occasions. Data collection commenced in February 2020 and concluded in October 2023.
Participants
Fifty-one women agreed to participate in the study. All participants had sought support following experiences of domestic violence, either from a specialized social welfare team or a women's shelter.
Procedure
At the posttreatment follow-up, conducted approximately 1 year later (M = 13.6 months, SD = 2.76), the sample comprised 44 women. Of these, 30 (57%) had sought assistance from a domestic violence unit within SW, while 21 (43%) had accessed support through WS. Difficulties in reaching participants to plan and carry out posttreatment interviews were the most common reason the interval exceeded 12 months. Seven women (5 SW and 2 WS) did not participate in the posttreatment data collection. Three indicated that they declined to take part in the posttreatment data collection. The remaining nonparticipants could not be reached, despite repeated attempts to schedule a session for data collection. Due to the COVID-19 pandemic, women could choose if they wished for meetings to be online (using Zoom) or within the facility where they received support. Nine out of 95 data collection sessions were conducted digitally.
Participants met with the researcher at the premises of the service from which they turned to for support. The participants took part in a structured interview before completing five self-assessment questionnaires. The interview covered sociodemographic information, current and past experiences of domestic violence, details about help-seeking behavior, the impact of the violence on the participants and any children, and, at the follow-up interview, the participants’ perceptions of the support they had received. Participants’ responses during the interviews were documented in a structured questionnaire and subsequently treated as quantitative data. All interviews were conducted by researchers, one, two, and five.
Abbreviations Used in the Text
In this article, the abbreviation SW refers to a specialized unit within social welfare services, and WS refers to a women's shelter. When referring to the units within the aforementioned organizations that support victims of domestic violence, the abbreviation DV unit is used. The term DV victim is used to describe a woman who has experienced domestic violence, while DV worker refers to professionals working within units, either within social welfare services or women's shelters that provide support to DV victims.
Outcome Measures
Childhood exposure to potentially traumatic events was assessed using the Linköping Youth Life Event Scale (LYLES), which evaluates 41 items related to 23 distinct types of events. Respondents indicate whether a specific event occurred during their childhood (up to the age of 18). It consists of three subscales: noninterpersonal events (nIPE), interpersonal events (IPE), and adverse life circumstances (ACE) (Nilsson et al., 2010). LYLES was applied to explore what traumatic childhood experiences the respondents have been exposed to. The reliability of the scale was Cronbach's α = .80. Test–retest reliability has previously been reported for LYLES r = .79 (Nilsson et al., 2010).
Exposure to different forms of partner violence was assessed using CTS2 (Conflict Tactic Scale 2; Straus et al., 1996) as well as the severity of violent acts, at the pretreatment data collection. CTS2 is a questionnaire aimed at assessing the nature and frequency of tactics used by partners in an intimate relationship. In this study, the analysis is confined to the frequency of respondents’ exposure to violence. The questionnaire measures five different dimensions of conflict behaviors. In this study, three dimensions are reported: psychological aggression, physical assault, and sexual coercion. The reliability of the subscales varied between Cronbach's α = .83 for the subscale psychological aggression and Cronbach's α = .90 for the subscale physical assault. Straus et al. (1996) reported Cronbach's α ranging from .79 to .95.
Mental health was assessed using the Symptom Checklist-90 (SCL-90; Derogatis, 1977; Derogatis & Fitzpatrick, 2004) on two occasions: before the intervention (pretest) and following its completion (posttest). The questionnaire asks respondents to rate their mental health status during the preceding week. The questionnaire contains 90 items identifying whether the person has been troubled by specific symptoms, and responses are given on a 5-point scale of distress from 0 (none) to 4 (extreme). The responses on 83 items form nine subscales. In this study, the reliability of the subscales varied between Cronbach's α = .71 (paranoid ideation) and Cronbach's α = .91 (obsessive–compulsive). The reliability for the total GSI was Cronbach's α = .97. Previous research has demonstrated similarly strong reliability for the SCL-90, including studies of females exposed to intimate partner violence (Karaaziz & Tansel, 2022) and adults who experienced childhood maltreatment (Barker-Collo & Read, 2011), with reported Cronbach's alpha coefficients of .97.
To assess perceived self-esteem, respondents completed the 10-item Rosenberg Self-Esteem Scale (Rosenberg, 1989) twice, at pretest and posttest. The scale consists of 10 statements rated on a 4-point Likert scale ranging from strongly agree to strongly disagree (e.g., “I feel I do not have much to be proud of”). The responses are scored 0–3, and the total score range is 0–30. The reliability was Cronbach's α = .77. The scale has been used in several studies concerning DV (for example, Brunton & Dryer, 2024) and showed good psychometric properties in a Swedish evaluation with Cronbach's α = >.90 (Eklund et al., 2018).
To assess potential reductions in trauma symptoms, the Impact of Event Scale–Revised (IES-R) was administered at pretest and posttest (Horowitz et al., 1979; Weiss & Marmar, 1997). IES-R is a 22-item self-report measure (based on DSM-IV) that assesses subjective distress caused by traumatic events. Respondents are asked to identify a specific stressful life event and then indicate how much they were distressed or bothered during the past 7 days by each difficulty listed. Items are rated on a 5-point scale ranging from 0 (not at all) to 4 (extremely). The IES-R yields a total score (ranging from 0 to 88), and subscale scores can also be calculated for the Intrusion, Avoidance, and Hyperarousal subscales. In this study, the reliability of the subscales varied between Cronbach's α = .71 (avoidance) and Cronbach's α = .84 (intrusion). The Cronbach's α for the total scale was .87. Previous research (Weiss & Marmar, 1997) has reported α = .96 for the total scale.
The Treatment Satisfaction Scale-2 (TSS-2; Clinton et al., 2004) measures treatment satisfaction and contains six items. Five items relate to the treatment: (a) how the client has been approached, (b) the structure of the sessions, (c) whether the client has been listened to and understood, (d) trust, and (e) the treatment targets. There were four alternative responses: 0 = no; 1 = yes, to some extent; 2 = yes, to a fairly high degree; and 3 = yes, to a very high degree. The sixth question asked the clients for an overall evaluation of the unit: −1 = very negative, 0 = negative, 1 = neutral, 2 = positive, and 3 = very positive. The reliability of the scale was good, Cronbach's α = .84. In a previous study (Clinton et al., 2004), the Cronbach's α = .87.
Data Analysis
Paired sample t-tests were used to evaluate potential pre- to posttreatment changes for ratio or interval-scaled measures. Effect sizes were calculated in accordance with Cohen (1988; small effect was 0.2–0.5, medium effect was 0.5–0.8, and large effect was 0.8 and higher).
Ethical Approval
This research has been approved by the Swedish Ethical Review Agency. Informed consent was obtained from all participants prior to data collection.
Results
Pretreatment Interview
The majority of the 51 participants were of Swedish origin, while 8 (16%) reported a country of origin outside Sweden (Table 1).
Sociodemographic Characteristics.
Among the eight participants with a non-Swedish background, the duration of residence in Sweden ranged from 8 to 20 years (M = 13.6, SD = 4.39). Women with a non-Swedish background were more likely to have sought support from a women's shelter (n = 5, 24%) than from social welfare services (n = 3, 10%), although this difference was not statistically significant.
Educational Background, Occupational Status, and Perceived Financial Well-Being
The distribution of educational attainment was comparable across the subgroups (Table 1). However, compared with the general population of Swedish women (Statistics Sweden [SCB], 2022), the proportion of women with a university degree was lower in this study sample (20% versus 40%). No significant differences were observed across subgroups.
The distribution of employment status was similar across the two subgroups (Table 1). However, when considering women not in full-time employment—including those in part-time work, part-time studies, unemployed, or on parental leave—the proportion was higher among women who sought support from WS compared with those who turned to SW (57% vs. 43%). This difference, however, was not statistically significant.
Respondents were asked to assess their satisfaction with their financial situation, choosing between two categories: very/fairly satisfied (classified as satisfied) and not very/not at all satisfied (classified as not satisfied). A greater proportion of respondents who sought support from WS reported being dissatisfied with their financial situation (61% not satisfied) compared with those who sought support from SW (40% not satisfied), a statistically significant difference (χ2 = 4.56, p = .033). This disparity is likely influenced by the previously noted finding that the majority of individuals seeking support from WS were either not in full-time employment or were dependent on welfare benefits for their livelihood.
Perpetrators of Domestic Violence
The majority of perpetrators of domestic violence were of Swedish origin (69%). Among the remaining individuals, 11% originated from other European countries, and 20% from countries outside Europe. Information on the country of origin was missing for 11 perpetrators. Women who sought support from a WS were significantly more likely to have been abused by a man of non-Swedish origin (47%) compared with those who sought help from SW (17%) (χ2 = 3.75, p = .05).
Children
Thirty-five women (69%) had between one and three children aged 19 years or younger living at home, resulting in a total of 66 children, evenly distributed between girls and boys, with an average of 1.89 children per mother. At the time of the first interview, the ages of the children ranged from 1 to 19 years (M = 8.9 years). An additional seven women had adult children. In most cases (71%), the perpetrator was also the father of the child or children.
Exposure to Potential Traumatic Experiences in Childhood
In line with Hypothesis 1, we expected that victims of DV seeking support would be exposed to severe forms of violence. To test this hypothesis, respondents’ experiences of violence in both childhood and adulthood were assessed. Childhood traumatic experiences were measured using the LYLES scale. No significant differences were observed between SW and WS respondents on the aggregated subscales of non-interpersonal events, interpersonal events, and adverse life circumstances (data not shown). The most frequently reported events included having a family member hospitalized (65%), the death of a close person (61%), and either experiencing bullying or the death of a family member (47% respectively).
The self-report data indicated a highly burdened group, with many participants having been exposed to multiple potentially traumatic experiences before the age of 18. Thirty-three of the respondents (65%) reported experiences of emotional, physical, or sexual abuse during childhood. Specifically, 35% reported emotional abuse during childhood, while 31% reported physical abuse by a parent. Moreover, 31% had lived with a parent who used drugs and/or alcohol, and 28% reported having a parent with a mental illness. Sixteen percent stated that they had been involuntarily separated from their parents and required to live elsewhere during childhood.
Exposure to Interpersonal Violence
During the pretreatment interview, respondents were asked whether they had been exposed to violence in prior relationships. Fifteen participants (29%) reported experiences of partner violence during adolescence or adulthood. During data collection, respondents’ current experiences of various forms of interpersonal violence were assessed using the CTS2, in line with Hypothesis 1, which posits that DV victims seeking support from either SW or WS are exposed to severe forms of violence. Respondents reported the prevalence, frequency, and severity of different types of violence.
Analysis of past-year prevalence revealed that psychological aggression was the most commonly experienced form of violence (94%), followed by physical assault (82%) and sexual coercion (61%). No significant group differences were observed.
In addition to completing the CTS2 questionnaire, participants were also asked during pretreatment interviews to indicate which forms of violence they had experienced. Five response options were provided: psychological violence, physical violence, sexual violence, material violence (damage to or destruction of possessions of personal significance), and social violence/control (restrictions on personal freedom and autonomy). The prevalence of psychological, physical, and sexual violence reported during the interviews closely corresponded with the findings obtained from the CTS2 questionnaire. Furthermore, 63% of respondents reported exposure to material violence, while 49% reported experiencing social violence/control. No significant differences between the two groups were identified.
When further examining the frequency of violent acts in accordance with CTS2, respondents who turned to WS reported higher exposure to all forms of violence, with a significant difference for physical violence (see Table 2), with a medium effect size.
Frequency of Exposure to Violent Acts (N = 49).
Note. CI = confidence interval. SD = standard deviation.
The prevalence of severe violent acts, as defined by Straus et al. (1996), is presented in Table 3. One example of an item classified as severe sexual coercion is: “My partner used threats to make me have sex.” One significant difference between the subgroups seeking support from SW and WS was identified when comparing the frequency of severe sexual coercion during the past year. Respondents in the WS group reported significantly higher levels of severe sexual coercion than those in the SW group (see Table 3).
Prevalence of Exposure of Severe Assaults (N = 49).
*p = .006.
Hypothesis 1 is supported by the extensive reports of potentially traumatic experiences during childhood, as well as by the reported exposure to violence as victims of domestic violence.
Outcomes at Follow-up
At the time of post-treatment data collection, the majority of respondents (n = 39; 89%) had left the relationship with the abusive partner. Five respondents (11%) remained in or had returned to the relationship.
In line with Hypothesis 2, it was expected that participation in the intervention would lead to improvements in mental health among clients. Mental health symptoms, as measured by the Symptom Checklist-90 (SCL-90), showed a significant reduction at posttreatment (see Table 4). Significant improvements were observed on all subscales Somatization, Obsessive-Compulsivity, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism, as well as on the GSI.
Mental Health at Pre- and Posttest (N = 43).
Note. CI = confidence interval; SD = standard deviation.
For the Depression and Anxiety subscales, as well as the GSI total score, the changes from pre- to posttest demonstrated large effect sizes, indicating substantial clinical improvement. No significant differences were found between the two subgroups (SW vs. WS).
The reported scores on Rosenberg's Self-Esteem Scale (Table 5) were slightly lower at pretreatment among women seeking support from WS, though the difference was not statistically significant. At posttreatment, the subgroup scores were more comparable, suggesting a greater improvement in self-esteem among the WS participants. For women in the WS group, as well as for the total sample, the improvement from pre- to posttest was statistically significant.
Scores on Rosenberg Self-Esteem Scale (N = 44).
Note. CI = confidence interval; SD = standard deviation.
Participants responded to the IES-R to assess symptoms of trauma. According to the criteria outlined in Assessing Psychological Trauma and PTSD (Weiss, 2004), a mean of ≥1.8 (total scale) and ≥1.89 (subscales) indicates the presence of PTSD. As shown in Table 6, participants’ mean scores exceeded this threshold across all subscales at pretest. However, self-reported symptoms were significantly reduced at posttest, with no mean scores indicating PTSD at 1-year follow-up. Changes on two subscales (intrusion and avoidance) and on the total scale demonstrated large effect sizes (Cohen's d ranging from .93 to 1.11), while one subscale (hyperarousal) showed an effect size just below the threshold for a large effect (.79).
Trauma Symptoms Pre- and Posttest (N = 44).
Note. CI = confidence interval; SD = standard deviation.
No significant differences in mean scores between women who received support from SW and those who received support from WS were observed on either the subscales or the total scale, at pre- or posttest.
Hypothesis 3 is supported by the significant reduction of trauma symptoms reported by posttest.
Hypothesis 4 predicted that the interventions offered by DV specialists within social welfare and women's shelters would be experienced as satisfactory. At posttest, women responded to six items (on a 3-point scale) indicating their experiences of being responded to at the unit, the content of treatment, being listened to, confidence in the staff, agreement on treatment goals, and general opinion concerning the unit.
The estimated frequency of meetings with their DV workers varied substantially, ranging from 4 to over 100 occasions, with a mean of 25 (SD = 18.0). The participants who turned to SW reported a mean of 21 sessions, and those who turned to WS reported a mean of 30 sessions; there was no significant difference. Due to the pandemic, meetings switch forms from physical meetings to also including, or being replaced by, digital meetings and/or phone calls (see Agevall Gross et al., 2025; Thulin et al., 2026; Table 7).
Satisfaction With Treatment (N = 43).
Note. SD = standard deviation.
Responses were given on a scale from 0 (no) to 3 (yes, to a very high degree).
Responses were given on a scale from −1 (very negative) to 3 (very positive).
Hypothesis 4 is supported, as the findings indicate that participants felt well supported, with DV workers perceived as attentive, empathetic, and fostering high levels of confidence. There were no significant differences between women receiving support at specialized units and those at women's shelters, except for the “overall impression” subscale, where women receiving support from shelters reported significantly higher satisfaction, t(42) = 1.89, p = .034.
Discussion and Applications to Practice
The study yields several noteworthy findings. The results indicate a substantial trauma burden among participants, reflected in both exposure to trauma (in childhood and adulthood) and associated symptoms. As predicted in Hypothesis 1, DV victims seeking support were exposed to severe forms of violence. Respondents reported extensive exposure to violence in both childhood and adulthood. In adulthood, psychological abuse was the most reported form of abuse. When behaviors are categorized as less and more serious violence, the majority reported being exposed to serious psychological and physical assault. However, regarding sexual violence, women who have sought support from WS appear to be exposed to significantly more severe forms of sexual coercion.
Mental health symptom levels among participants at baseline were high. Participants in the present study reported higher symptom levels, as measured by the SCL-90, across all subscales and the total scale than those previously reported for clinical groups (inpatient and outpatient populations) in Swedish studies (Fridell, 2002). Similar high symptom levels were reported in a previous Swedish study of women exposed to domestic violence (Scheffer Lindgren & Renck, 2008).
A reduction in mental health symptoms, as measured by the SCL-90, was observed in line with Hypothesis 2, with statistically significant improvements identified on all subscales. This represents a positive outcome, and the large effect sizes observed for the Depression and Anxiety subscales indicate clinically meaningful improvements in mental health. Nevertheless, posttreatment scores remained considerably higher than those observed among women in the general Swedish population (Fridell, 2002). Furthermore, participants’ posttreatment scores were more comparable to those of other clinical groups than to those of the general female population.
Most of the women had children, implying continued contact with the children's father, who in many cases was the perpetrator. Several of the women reported being in legal processes concerning custody and accommodation, at posttreatment test, increasing their levels of stress. Taken together, the transition from living with their perpetrator to a life free from violence could be protracted, without a sharp line. When forced to have continued contact, there are several arenas for continuing violence. Even if the women seek to break their trajectory and move forward, several obstacles remain, even after the separation.
The study group was likely affected by the recent incidents of violence but may also carry a preexisting psychological burden, as indicated by reports of childhood trauma and prior exposure to violence in adulthood. Prolonged exposure to such adversity may have further exacerbated mental health difficulties among some participants. Moreover, the effects of chronic stress related to violence often require considerable time to heal. Although the primary focus of the study was on outcomes of interventions, we also examined participants’ experiences of potentially traumatic events during childhood. Childhood abuse has been found to correlate with experiencing DV (Cherrier et al., 2023). The results suggested that exposure to severe adversities may represent a life-course trajectory for some of the women, beginning in childhood and continuing into adulthood. For some participants, abuse seems to be a persistent factor, just changing form or perpetrator over time. For instance, 65% of the participants reported experiences of emotional, physical, or sexual abuse during childhood. It was also common to have lived with a parent with mental illness or a parent with substance use. The extent to which these earlier experiences had previously been acknowledged or addressed in treatment remains unclear. Some participants stated that this was the first time they had disclosed such experiences, and it is unknown whether these issues were discussed in sessions with their DV workers. These findings underscore the importance of systematically assessing clients’ histories of trauma and providing appropriate therapeutic tools to address the long-term impact of such experiences.
Stressful life events, such as experiences of domestic violence, have been shown to negatively influence self-esteem (Brunton & Dryer, 2024; Matheson et al., 2015). The findings of the present study suggest that receiving support from DV workers may contribute to improvements in self-esteem. Women seeking support at a womeńs shelter reported lower levels of self-esteem (although the difference was not statistically significant), but both groups reached a similar level following treatment. As previous research has demonstrated a correlation between self-esteem and perceived quality of life (Eklund et al., 2018), these findings underscore the importance of interventions that promote self-esteem as a means of enhancing overall well-being.
Regarding Hypothesis 3, the hypothesis was supported. Before receiving support, women reported trauma symptom scores that exceeded the threshold for PTSD on all subscales. Their symptoms were significantly reduced after completing the interventions, no longer meeting the threshold for any of the subscales. There were no differences between groups, indicating the importance of support independently of service providers.
Regarding Hypothesis 4, the respondents’ answers reflected overall satisfaction with the support received. In particular, the DV workers’ ability to listen attentively and instill a sense of trust and confidence was especially appreciated. Even though both groups reported high levels of satisfaction, women who received support within a shelter context reported significantly higher levels of satisfaction in their overall evaluation of the service unit. Several factors may account for this difference. Notably, women receiving support at shelters often participated in group-based interventions, which allowed for peer interaction and the exchange of emotional and practical support with others who had similar experiences. Furthermore, staff at shelters may be better positioned to adopt a holistic approach, responding flexibly to the expressed needs of the women. Unlike social welfare professionals, shelter staff typically do not hold the formal mandate to provide needs-assessed interventions. As a result, their role may be more focused on emotional support and empowerment rather than formal case management (Stanley et al., 2025). Accessibility may also contribute to the higher levels of satisfaction reported. Women who contacted a WS did not experience any waiting time before meeting with a DV worker, whereas this was reported among women who contacted the SW units. Compared to municipal services, shelters may offer more immediate and approachable forms of assistance, which could enhance users’ experiences of responsiveness and relevance. Future research could enhance this perspective further.
Despite the difference in their overall evaluation, women from both groups reported high levels of satisfaction with the treatment received. It is noteworthy to point out the fact that data collection, to a high degree, took place during the COVID-19 outbreak. Sweden, unlike most other countries, did not have a full lockdown; however, the support changed forms. Several units experienced temporary shutdowns, and a shift from in-person to digital meetings was common (Agevall Gross et al., 2025; Thulin et al., 2026). Despite these challenges, attrition was low, which is notable given the 1-year follow-up period. A total of 86% of the women completed both the pre- and posttreatment data collection, indicating a high level of participant retention. The positive result indicates that the participating services had been able to meet the needs of the women, despite a turbulent time.
However, the study has limitations. The underrepresentation of women of non-Swedish origin raises concerns. While approximately 20% of women living in Sweden are foreign-born, this proportion was lower in the present sample. It remains unclear whether this reflects lower help-seeking rates, limited access to participating services, or recruitment barriers. We also do not know whether all eligible women were informed about the study or whether some declined participation. This issue was discussed with DV professionals during the study and highlighted the importance of ensuring that all women receiving support were informed about research opportunities. The pandemic may have hindered recruitment, as external disruptions affected both initial contact with services and the transition to digital interactions. At times, broader contextual factors may have taken precedence over presenting the study and recruiting participants.
The study design does not allow for control of potential confounding factors, such as natural recovery over time. At follow-up, participants were in varying life circumstances that may have influenced their self-reported health and symptom levels. For example, some were involved in ongoing custody disputes and legal proceedings during the final phase of data collection. Finally, for ethical reasons, it was not feasible to include a control group of women who did not receive support.
There are several implications of the results of this study that may inform social work practice in the field of domestic violence. A majority of respondents reported early life adversity in the form of potentially traumatic childhood experiences. Furthermore, approximately one-third reported having experienced intimate partner violence prior to their current violent relationship. At pretreatment, participants reported trauma symptom levels exceeding the threshold for PTSD, underscoring the severity of their psychological distress at the time of seeking support. No information was collected regarding whether or how respondents had received support in processing these earlier experiences of violence. In some cases, it emerged that these were experiences the respondents had never previously disclosed. It can therefore be assumed that some women exposed to violence may have a history of prior traumatic life experiences. It is important that services providing support to victims of violence, as well as professionals working within these services, are aware that individuals seeking help may carry a preexisting burden in the form of traumatic childhood experiences. Such experiences are likely to influence the type of support and intervention required. The results indicate that appropriate interventions can substantially reduce trauma-related symptoms. Short-term support with a predetermined number of sessions, as reported by some respondents, is unlikely to be sufficient in these cases. The results of this study indicate that appropriate interventions can substantially reduce trauma-related symptoms and contribute to recovery from PTSD.
The findings also suggest that specialized social workers and domestic violence workers at women's shelters providing services to the respondents in this study possess the necessary training and experience to deliver valuable and effective support. Maintaining this level of competence among practitioners is likely a prerequisite for effectively supporting women exposed to violence.
Footnotes
Acknowledgments
The authors would like to thank the women who participated in this study.
Funding
This research was funded by the Swedish Research Council for Health, Working Life and Welfare (2018-01327).
Forskningsrådet om Hälsa, Arbetsliv och Välfärd (grant number 2018-01327).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.