Self-Compassion, Health, and Empowerment: A Pilot Randomized Controlled Trial for Chinese Immigrant Women Experiencing Intimate Partner Violence

Abstract

Chinese immigrant survivors of men’s violence experience both significant mental health impacts from abuse and barriers to formal services. Therefore, we examined the preliminary efficacy of an innovative mobile-based empowerment-based intervention (self-compassion, health, and empowerment; SHE) that specifically focuses on abused Chinese immigrant women in the US. This pilot study used a two-arm randomized controlled design with repeated measures. A convenience sample (N = 50) of Chinese immigrant women who experienced past year intimate partner violence (IPV) were recruited online and randomly assigned to the intervention or control group (25 per group). We assessed IPV exposure, safety behaviors, depressive symptoms, anxiety symptoms, post-traumatic stress disorder (PTSD) symptoms, and self-compassion at baseline, post-intervention, and 8-week follow-up. Of 95 eligible participants, 50 (52.6%) agreed to participate and completed baseline data collection; intervention completion rate was 64%. We found a significant group and time interaction for self-judgment (a self-compassion component), with a significant reduction seen in the intervention group compared to the control group. Despite no other significant group differences observed over time, the intervention group showed consistent trends toward improvements in most outcome measures, including specific types of IPV (i.e., negotiation, psychological aggression, and sexual coercion), depressive and PTSD symptoms, self-compassion, and certain components of self-compassion (i.e., isolation and over-identification) when compared to the control group. Our findings suggest that the SHE intervention shows promise in improving the mental health well-being of Chinese immigrant survivors. However, a fully powered randomized controlled trial is warranted to determine its efficacy. Our intervention has the potential to be translated in the Chinese immigrant populations with the necessary organizational support.

Introduction

Intimate partner violence (IPV) is a serious social and public health issue, due to its high prevalence, with a lifetime prevalence of 33% in the U.S. (Smith et al., 2018), and its significant physical and mental health consequences such as injury, chronic pain, and depression (Campbell, 2002; Miller & McCaw, 2019). Since the late 1960s, Asian American women have begun to advocate for their civil rights as a racial minority and as women. However, progress has been slow, hindered by various social and cultural oppression (Chow, 1992). Furthermore, the anti-domestic violence movement has predominately centered around White Americans, providing limited benefits to Asian American women (Wang, 1996). Specifically, legal and social service responses to IPV have primarily focused on the needs of White women, leaving the needs of abused Asian American women unaddressed. Resources to support abused Asian American women, especially those meeting their multicultural and multilingual needs, are scarce and concentrated in major urban areas. Moreover, given the diversity within the Asian Americans, including many ethnic subgroups such as Chinese, Korean, Indian, and Vietnamese, each with distinct languages and immigration experiences, their specific needs for support remain largely unmet (Lee, 2014).

As the largest Asian ethnic subgroup, Chinese immigrant women have been marginalized and underserved in IPV support services, despite a high prevalence of over 20% within the past year (Li et al., 2020). They have very limited access to support from formal services such as women’s shelters and law enforcement due to various sociocultural barriers. For example, many Chinese immigrant women have limited English proficiency, are often financially or legally dependent on their partners, and tend to stay at home, leading to their isolation from mainstream American society. This isolation results in low awareness of available resources and support. Furthermore, there is a lack of culturally appropriate services, particularly in areas with a low Chinese population density (Li et al., 2021). Therefore, it is crucial to develop culturally appropriate interventions that incorporate innovative strategies to overcome these barriers and enhance their acceptability and accessibility for this underserved group of women.

Some existing IPV interventions are grounded in Dutton’s empowerment theory (Dutton, 1992) which focuses on increasing IPV victims’ autonomy and ability to decision making and problem solving by providing IPV information, resources, safety planning, and support around the violence (Trabold et al., 2020). The model’s components include (1) protection, a focus on increasing the woman's safety; (2) enhanced choice making and problem solving in decisions about the relationship; and (3) healing of post-traumatic reactions. Guided by Dutton’s empowerment model, both the structured IPV intervention developed by Parker et al. (1999), the Domestic Violence Enhanced Home Visitation Program (DOVE) intervention (Sharps et al., 2016), and the advocacy intervention in Hong Kong (Tiwari et al., 2010) have been shown to significantly increase abused women’s safety behaviors and reduce re-victimization.

Safety behaviors and IPV exposures are important outcomes for IPV interventions; however, these interventions mainly focused on empowering abused women to ensure safety and make decisions about their relationship with less attention to improving victims’ psychological and cognitive functioning. This represents a significant gap for abused immigrant Chinese women, given IPV’s substantial and well-documented effects on their mental health, for example, depression, anxiety, and post-traumatic stress disorder (PTSD; Chan et al., 2010; Xu et al., 2020). An understudied but potentially related and important component of mental health among IPV survivors is self-compassion, that is, acknowledging suffering, failure, and inadequacies as part of the common human experience and offering oneself warmth, kindness, and empathy instead of being self-critical and judgmental toward oneself (Neff, 2003). Evidence consistently suggests that depression, anxiety, and PTSD are associated with decreased self-compassion (MacBeth & Gumley, 2012; Winders et al., 2020). However, due to the deep influence of the Chinese cultural values of criticism and self-criticism, Chinese immigrant women often blame themselves for the violence that happens to them, which can lead to feelings of failure and inadequacies (Li et al., 2022).

Building self-compassion has been recognized as a promising psychological approach for alleviating guilt, shame, and self-blame and enhancing mental health well-being. The Mindful Self-Compassion (MSC) program developed by Neff and Germer (2013) provides participants with a variety of tools and teach them loving-kindness skills. The MSC program has been shown to significantly increase participants’ self-compassion and decrease depression and anxiety (Neff & Germer, 2013). The MCS program has also been shown to improve psychological well-being specifically among Chinese participants in China (Finlay-Jones et al., 2018).

Similarly, interventions based on relaxation techniques such as deep breathing, progressive muscle relaxation, and visualization have been shown effective in mitigating depression, anxiety, and PTSD (Klainin-Yobas et al., 2015; Niles et al., 2018). Furthermore, it is simpler, easier to implement, and more socially acceptable than psychotherapy. Taken together, incorporating self-compassion and relaxation techniques into IPV interventions appears promising in reducing the significant impact of IPV on survivors’ mental health, an important consideration for Chinese immigrant survivors.

Finally, addressing the challenges posed by the wide geographical dispersal of the Chinese population in the U.S., as well as potential language and cultural barriers, requires innovative approaches to ensure accessibility and acceptability of culturally appropriate interventions. The use of telephone-based and mobile app-based delivery methods can facilitate access to this hard-to-reach population. Mobile health is a promising, ideal approach to deliver targeted care and support to the hard-to-reach population and to overcome barriers (e.g., stigma, shame, privacy concern) in seeking help (Silva et al., 2015) and has been safely and widely used with IPV survivors (Emezue et al., 2022).

Therefore, our intervention, Self-Compassion, Health, and Empowerment (SHE), adapted the structured IPV intervention from the DOVE study and additionally incorporated self-compassion and relaxation techniques for Chinese immigrant women experiencing IPV. The study aimed to test the preliminary efficacy of our SHE intervention in reducing IPV and improving the mental health well-being of Chinese immigrant women in the U.S. who have experienced IPV. We hypothesized that participants in the intervention group would report less frequent IPV and demonstrate improved mental health outcomes compared to the control group.

Methods

Study Design

This pilot study used a two-arm randomized controlled design with repeated measures, and it was conducted from March 2022 to March 2023. Outcome measures were assessed at baseline, post-intervention, and 8-week follow-up. The study was registered in ClinicalTrials.gov (ID: NCT05011552).

Sample and Recruitment

To be eligible for this study, participants had to be female, aged 18 or older, self-identify as Chinese currently residing in the U.S., in an intimate relationship at the time of enrollment, and with current experience of IPV (i.e., within the past year). Exclusion criteria were self-reported substance use, suicidality, or ongoing treatment for severe mental illness. Based on prior research that used an empowerment-based intervention for abused Chinese women in Hong Kong (Tiwari et al., 2010), the estimated effect size (Cohen’s d) for between-group differences in depressive symptoms at 9 months post-intervention was about 0.2. Following the stepped rules of thumb for pilot trial sample size per treatment arm (Whitehead et al., 2016), 20 participants per arm would provide a power of 80% to detect an effect size of 0.1–0.3 with a two-tailed Type I error of .05. Considering an anticipated dropout rate of up to 20%, we recruited and randomized 25 participants per arm.

A convenience sample of participants were recruited online from March 2022 to October 2022 through platforms such as WeChat, Prolific, and the Collaborative Approach for Asian Americans & Pacific Islanders Research & Education (CARE) registry. WeChat is the most widely used social media application among the Chinese population, while Prolific is a platform providing a large pool of potential participants. The CARE registry is a research registry specifically established to enhance the participation and representation of Asian American and Pacific Islander individuals in research studies. The majority of participants were recruited through WeChat Advertisements.

This study followed established safety procedures for recruitment and retention of community-based IPV survivors, to reduce the risk of breach of confidentiality and an abuser discovering they were participating in the study, which might result in further abuse (Sabri et al., 2023; Sullivan & Cain, 2004). For safety reasons, the study flyer was carefully designed to avoid implying that the study was recruiting Chinese women experiencing IPV. It described the study as a “Chinese immigrant women’s health care study” and provided a link or QR code leading to a survey for eligibility screening.

Randomization and Blinding

Potential participants completed a brief online survey to confirm their eligibility. Eligible participants were then directed to a page that provided brief study information and asked to provide their preferred safe contact information, preferred time for contact, and any special instructions for the research team to follow when contacting them if interested in participation. Alternatively, potential participants could opt to contact the research team directly via email or phone instead of entering their contact information on the page, offering additional options for safety. Potential participants were contacted and scheduled for a telephone call at a safe and convenient time chosen by the participant.

Participants were randomized in a 1:1 ratio to either the intervention or control group using blocked randomization. This was achieved by generating a list of random permutations using the website Randomization.com (https://www.randomization.com; Kim & Shin, 2014). During the initial phone call, separate informed consent forms were read for participants in the two groups, and verbal consent (employing a waiver of documentation of informed consent process) was obtained to ensure participant safety and confidentiality. In this study, only participants were blinded to the group allocation. After the collection of follow-up data, a debriefing session was conducted over the phone with the participants to disclose their group allocation and provide them with an opportunity to withdraw their data if desired. None of the participants opted to withdraw their data after being informed of their group allocation. The study was approved by the Institutional Review Board of the University of Texas at Austin (ID: STUDY00001645).

Measures

IPV screen

The Abuse Assessment Screen (AAS) was used to screen for IPV (Soeken et al., 1998). Participants who reported emotional, physical, and/or sexual abuse by an intimate partner within the past year were considered to have screened positive for IPV. The Chinese version of the AAS, which has been validated among Chinese women in Hong Kong, demonstrated satisfactory measurement accuracy, with a specificity of at least 89% for emotional, physical, and sexual abuse and a sensitivity ranging from 36% to 66% (Tiwari et al., 2007).

The frequency of IPV

The Revised Conflict Tactics Scales (CTS2) were used to measure the frequency of violent behaviors used by the partner (Straus et al., 1996). It consists of 27 items and 5 subscales, including negotiation, psychological aggression, physical assault, injury, and sexual coercion. The Chinese version of CTS2 has been found to have good reliability (Cronbach’s α: 0.84) and validity in China (Zhang et al., 2014).

Depressive symptoms

The Patient Health Questionnaire-9 (PHQ-9) was used to assess depressive symptoms (Kroenke et al., 2001). The PHQ-9 includes nine items that inquire about the frequency of depressed mood in the past 2 weeks. The four responses are “not at all,” “several days,” “more than half the days,” and “nearly every day.” The total score was computed by summing up individual item scores, ranging 0–27. Higher total scores indicate more severe depressive symptoms. In the study, a PHQ-9 score of 10 or higher was considered as indicative of the presence of depression. The Chinese version has shown good reliability (Cronbach’s α: 0.85; Test-retest coefficient: 0.87) and validity among college students in China (Zhang et al., 2013).

Anxiety symptoms

The Generalized Anxiety Disorder-7 (GAD-7) was used to measure anxiety symptoms experienced in the past 2 weeks (Spitzer et al., 2006). Respondents rated the frequency of experiencing each of the seven anxiety symptoms on a 4-point Likert-type scale (0=not at all, 1=several days, 2=more than half the days, 3=nearly every day). The total score was computed by summing up individual item scores, ranging 0–21. Higher total scores suggest more severe anxiety symptoms. A score of 10 or greater on the GAD-7 was used in this study to indicate the presence of anxiety. The Chinese version of GAD-7 has been shown to have good reliability (Cronbach’s α: 0.87; Test-retest coefficient: 0.82) and validity (Li et al., 2014).

PTSD symptoms

The PTSD Checklist for DSM-5 (PCL-5) was used to assess PTSD symptoms in response to a stressful experience during the past month (Weathers et al., 2013). It comprises of 20 items that correspond to the DSM-5 symptom criteria for PTSD. Responses are rated on a 5-point Likert-type scale, ranging from 0 (not at all) to 4 (extremely). The total symptom severity scores range 0—80, with higher scores indicating more severe PTSD symptoms. A cutoff score of 31 was used in the study to indicate probable PTSD. The Chinese version of PCL-5 has shown good reliability (Cronbach’s α: 0.91) and convergent and discriminant validity (Cheng et al., 2020).

Safety behaviors

Safety behaviors were measured using 27 items adapted from the IPV Strategies Index (Goodman et al., 2003). Respondents were asked whether they had engaged in a list of behaviors to protect their safety (e.g., seeking help from formal or informal network, legal assistance, safety planning). As not all 27 items were applicable to each participant, the adjusted total number of safety behaviors were computed in the study. Specifically, it was obtained by dividing the number of behaviors performed by the number of applicable behaviors and then multiplying by 27. The range of scores was 0–27, with higher values suggesting more safety behaviors taken by respondents.

Self-compassion

The Self-Compassion Scale-Short Form (SCS-SF) comprises 12 items and 6 subscales, including self-kindness, common humanity, mindfulness, self-judgment, isolation, and over-identification. Items are rated on a 5-point response scale ranging from 1 (almost never) to 5 (almost always). To calculate each subscale score, the negative subscale items were first reverse-scored, and then the subscale items were averaged. Higher subscale scores indicate higher levels of self-kindness, common humanity, and mindfulness, while suggesting lower levels of self-judgment, isolation, and over-identification. The SCS-SF showed adequate reliability (Cronbach’s α: 0.87) and a near-perfect correlation with the original SCS (Raes et al., 2011). Similarly, the Chinese version of the scale also demonstrated acceptable reliability (Cronbach’s α: 0.84; Test-retest coefficient: 0.89) and validity (Chen et al., 2011).

Sociodemographic characteristics

Participants were asked to self-report age, education, employment, annual personal income, marital status, number of children, religion, nativity, and immigration status.

Data Collection Methods

Sociodemographic information was collected at baseline. Outcome variables (the frequency of IPV, depressive symptoms, anxiety symptoms, PTSD symptoms, safety behaviors, and self-compassion) were measured at baseline, immediately after the 7-week intervention, and 8 weeks post-intervention. Baseline data was collected from the participants after obtaining their consent. Survey questionnaires are available in either English or Chinese. Qualtrics, a secure online data collection service, was used as the survey platform. Women were sent survey links via safe email addresses or text messages. Participants were compensated with a $25 electronic gift card for each of the three completed assessments, totaling $75 over approximately four months. This compensation aligns with recommendations for payment in clinical research studies (Grady, 2005). The 8-week follow-up period was determined based on existing research on mindfulness-based interventions for victims of interpersonal violence. In these similar studies, follow-up periods have varied widely, ranging from 4-, 8-, to 24 weeks post-intervention (Gallegos et al., 2015; Kimbrough, et al., 2010). Notably, a study involving a sample of victims of childhood sexual abuse administered an 8-week mindfulness-based stress reduction intervention. This study reported significant reductions in depression, anxiety, and PTSD symptoms at 4-, 8-, and 24-weeks post-intervention compared to baseline (Kimbrough, et al., 2010). Given the earlier report, we determined that an 8-week follow-up period would be a pragmatic choice to allow for capturing the potential sustained effects of our intervention while operating within our resource limitations.

Study Treatments

The SHE intervention

Guided by Dutton’s empowerment model, the SHE intervention consisted of seven weekly sessions. The first session was delivered over the phone with the woman, using a brochure adapted from the DOVE study (Sharps et al., 2016). The brochure describes four major areas: a) IPV information, b) Danger Assessment, c) Safety Planning, and d) Resources. The brochure was discussed with the woman in an interactive manner so that the participant was encouraged to share her experiences and choose her options as she proceeded. At the end of the first session, the interventionist provided an overview of the following sessions to ensure that the woman had a clear understanding of what to expect. The first session lasted 30–40 minutes. Sessions 2–7 focused on providing mental health self-care resources, including deep breathing, progressive muscle relaxation, visualization, self-compassion, mindfulness, and loving-kindness. These six sessions were delivered through a WeChat mini-program specifically designed for this intervention. WeChat mini-program is a lightweight micro-app hosted on WeChat. Each session included a video introduction on one of the six topics, along with audio instructions on exercises. The video durations ranged from 5 to 20 minutes, and the audio durations varied between 4 and 20 minutes.

At the beginning of each session, the woman was asked to review the video and then practice the exercises presented in that session at their preferred time and location every day for one week. Safe reminder messages, referring to the study as the “Chinese Immigrant Women’s Health Care Study,” were sent one day prior to the start of each session. In the reminder messages, we also asked about how many times they practiced that previous week’s exercises. Practicing the exercises at least four times a week was considered the completion of one session. This criterion aligns with a previous mobile-based mindfulness intervention that similarly defined the completion of one session as a minimum of four days of practice per week (Zhang et al., 2023). The intervention was provided based on the participant's preference, either in English or Chinese.

The control group

Women in the control group were provided with the same brochure as the intervention group, as well as mental health care resources and information. These resources were delivered to them on a weekly basis, either in English or Chinese, via a secure email address they had provided. It is important to note that unlike the intervention group, the control group did not participate in a phone call to discuss the brochure or have access to the WeChat mini-program.

Data Analysis

Descriptive statistics (i.e., means and standard deviations, percentages) were computed for all variables. Mixed analysis of variance (ANOVA) was conducted to test group differences in outcome measures over time from baseline to immediately after the intervention and 8 weeks post-intervention. Cohen’s d was calculated as a measure of effect size. We used the intention-to-treat analysis with the last observation carried forward method. The sensitivity analysis was also conducted, which included only participants who completed all the assessments. All statistical analyses were performed using SPSS Version 26. Statistical significance was considered at p < .05, two-tailed.

Results

Participant Flow

A total of 965 Chinese immigrant women were approached and screened for eligibility. Out of the 95 eligible women who were invited to participate, 51 agreed to participate in the study. However, one woman did not start the baseline assessment, resulting in a final sample of 50 participants (52.6% response rate), with 25 in each group (SHE and control). In the SHE intervention group, a total of 19 participants completed the assessment immediately after the intervention, while 18 completed the assessment at 8 weeks after the intervention. In the control group, 20 participants completed the assessment both immediately and at 8 weeks after the intervention. The overall retention rate was 76% at 8 weeks post-intervention. The participant flow chart is shown in Figure 1.

Figure 1.

Participant Flow Diagram.

Sample Characteristics

Baseline sociodemographic characteristics of the participants are displayed in Table 1. Among the 50 participants, the average age was 36 years. Sixty-four percent of the participants had a bachelor’s degree or above and were employed, while over one-third had no personal income. The majority of participants were married or currently partnered, 84% had at least one child, and 72% had no religious beliefs. About 70% were either U.S. citizens or permanent residents. No significant group differences were found in sociodemographic factors and outcome measures at baseline between the SHE and control groups (p < .05). Two participants in the intervention group and one participant in the control group completed the study in the English language.

Table 1.

Baseline Sample Characteristics.

Variables	Total sample (N = 50)	SHE group (n = 25)	Control group (n = 25)
Sociodemographic characteristics
Age, M (SD)	36.16 (9.89)	35.24 (10.42)	37.13 (9.44)
Education, n (%)
Less than a bachelor’s degree	18 (36.0)	10 (40.0)	8 (32.0)
Bachelor’s degree or above	32 (64.0)	15 (60.0)	17 (68.0)
Employment, n (%)
Unemployed	18 (36.0)	11 (44.0)	7 (28.0)
Employed	32 (64.0)	14 (56.0)	18 (72.0)
Annual personal income, n (%)
None	17 (34.7)	10 (40.0)	7 (29.2)
≤$40,000	15 (30.6)	8 (32.0)	7 (29.2)
>$40,000	17 (34.7)	7 (28.0)	10 (41.7)
Marital status, n (%)
Married	45 (90.0)	23 (92.0)	22 (88.0)
Not married, but having an intimate partner	3 (6.0)	1 (4.0)	2 (8.0)
Divorced	2 (4.0)	1 (4.0)	1 (4.0)
Number of children, n (%)
0	8 (16.0)	4 (16.0)	4 (16.0)
≥1	42 (84.0)	21 (84.0)	21 (84.0)
Religion, n (%)
No	36 (72.0)	18 (72.0)	18 (72.0)
Yes	14 (28.0)	7 (28.0)	7 (28.0)
Foreign born, n (%)
No	1 (2.0)	1 (4.0)	0 (0)
Yes	49 (98.0)	24 (96.0)	25 (100)
Immigration status, n (%)
U.S. citizenship or permanent residency	34 (69.4)	15 (62.5)	19 (76.0)
Non-immigration visa	15 (30.6)	9 (37.5)	6 (24.0)
Outcome variables
IPV, M (SD)
Negotiation	42.76 (38.21)	46.12 (42.75)	39.25 (33.40)
Psychological aggression	23.31 (33.28)	26.36 (37.06)	20.00 (29.08)
Physical assault	3.46 (8.38)	4.91 (11.05)	2.00 (4.13)
Injury	1.10 (4.27)	0.92 (3.34)	1.29 (5.10)
Sexual coercion	1.68 (3.69)	1.48 (3.53)	1.91 (3.94)
Depressive symptoms, M (SD)	8.46 (6.22)	9.12 (6.06)	7.80 (6.44)
Depression, n (%)
Yes	18 (36.0)	11 (44.0)	7 (28.0)
No	32 (64.0)	14 (56.0)	18 (72.0)
Anxiety symptoms, M (SD)	6.82 (5.50)	6.52 (4.98)	7.12 (6.07)
Anxiety, n (%)
Yes	11 (22.0)	6 (24.0)	5 (20.0)
No	39 (78.0)	19 (76.0)	20 (80.0)
PTSD symptoms, M (SD)	20.68 (17.56)	20.72 (17.13)	20.64 (18.34)
PTSD, n (%)
Yes	11 (22.0)	6 (24.0)	5 (20.0)
No	39 (78.0)	19 (76.0)	20 (80.0)
Number of safety behaviors, M (SD)	6.21 (7.84)	6.46 (9.05)	5.97 (6.59)
Self-compassion, M (SD)	3.26 (0.63)	3.27 (0.65)	3.24 (0.63)
Self-kindness	3.44 (0.90)	3.40 (0.97)	3.48 (0.84)
Common humanity	3.52 (1.11)	3.56 (1.00)	3.48 (1.22)
Mindfulness	3.42 (1.11)	3.52 (0.98)	3.32 (1.23)
Self-judgment	3.48 (0.84)	3.52 (0.78)	3.44 (0.91)
Isolation	2.89 (1.13)	2.90 (1.01)	2.88 (1.25)
Over-identification	2.78 (1.04)	2.74 (0.95)	2.82 (1.14)

Note. SHE = self-compassion, health, and empowerment; IPV = intimate partner violence; PTSD = post-traumatic stress disorder.

Among the entire sample, at baseline more than 1 in 3 (36%) reported clinically significant depressive symptoms (i.e., a PHQ-9 score ≥10); more than 1 in 5 (22%) reported clinically significant symptoms of anxiety (GAD-7 score ≥10) and 22% reported clinically significant PTSD symptoms (PCL-5 score ≥31).

Group Differences in Primary and Secondary Outcomes Across Time

As seen in Table 2, the mixed ANOVA results showed that there were no significant group effects for any of the outcome variables (p > .05). However, the main effect of time was significant for psychological aggression (F(2, 92) = 3.50, p = .043), depressive symptoms (F(2, 96) = 5.16, p = .010), anxiety symptoms (F(2, 96) = 3.77, p = .027), PTSD symptoms (F(2, 96) = 8.29, p < .001), and self-compassion (F(2, 96) = 3.36, p = .048) at different time points. Notably, a significant group and time interaction was observed for self-judgment (F(2, 96) = 3.84, p = .025). Specifically, self-judgment was significantly improved in the SHE intervention group compared to the control group immediately following the intervention (Mean Difference = 0.72, 95% CI (0.21, 1.23), p = .006), followed by no significant group difference at the 8-week follow-up, as shown in Figure 2.

Table 2.

Group Differences across Time on Outcome Measures.

Outcome Variables	Mean Difference (95% CI)	p	Cohen’s d	Group Effect		Time Effect		Group × Time Effect
				F	p	F	P	F	p
IPV
Negotiation				1.05	.311	1.59	.209	0.35	.704
T1	6.87 [−15.24, 28.98]	0.535	N/A
T2	11.60 [−8.93, 32.13]	.262	0.32 [−0.24, 0.88]
T3	14.64 [−10.34, 39.62]	.244	0.33 [−0.22, 0.89]
Psychological aggression				0	.987	3.50	.043*	1.70	.194
T1	6.36 [−13.11, 25.83]	0.514	N/A
T2	0.44 [−16.30, 17.18]	.958	0.015 [−0.54, 0.57]
T3	−4.96 [−19.28, 9.36]	.489	−0.20 [−0.75, 0.36]
Physical assault				0.65	.426	0.66	.520	0.91	.407
T1	2.91 [−2.05, 7.87]	.243	N/A
T2	−0.31 [−6.03, 5.41]	.914	−0.031 [−0.59, 0.52]
T3	0.41 [−5.69, 6.52]	.892	0.039 [−0.52, 0.59]
Injury				0.05	.824	1.10	.326	0.21	.761
T1	−0.38 [−2.88, 2.13]	.765	N/A
T2	0.62 [−1.98, 3.22]	.635	0.14 [−0.42, 0.69]
T3	0.21 [−0.21, 0.63]	.313	0.29 [−0.27, 0.84]
Sexual coercion				0.38	.543	0.68	.509	0.24	.789
T1	−0.43 [−2.62, 1.76]	.695	N/A
T2	−0.35 [−4.12, 3.43]	.855	−0.053 [−0.61, 0.50]
T3	−1.24 [−5.08, 2.60]	.519	−0.18 [−0.74, 0.37]
Depressive symptoms				0.05	.827	5.16	.010*	1.61	.209
T1	1.32 [−2.23, 4.87]	.459	N/A
T2	0.04 [−3.29, 3.37]	.981	0.0068 [−0.55, 0.56]
T3	−0.32 [−3.51, 2.87]	.841	−0.057 [−0.61, 0.50]
Anxiety symptoms				0.04	.853	3.77	.027*	0.31	.735
T1	−0.60 [−3.76, 2.56]	.704	N/A
T2	−0.28 [−3.24, 2.68]	.850	−0.054 [−0.61, 0.50]
T3	0.08 [−2.94, 3.10]	.958	0.015 [−0.54, 0.57]
PTSD symptoms				0.20	.655	8.29	<.001***	0.92	.401
T1	0.08 [−10.01, 10.17]	.987	N/A
T2	−3.12 [−12.65, 6.41]	.514	−0.19 [−0.74, 0.37]
T3	−3.12 [−12.63, 6.39]	.512	−0.19 [−0.74, 0.37]
Number of safety behaviors				0.08	.784	0.58	.506	0.66	.473
T1	0.48 [−4.02, 4.99]	.830	N/A
T2	−1.86 [−5.29, 1.58]	.283	−0.31 [−0.87, 0.25]
T3	−0.48 [−4.96, 4.01]	.832	−0.061 [−0.62, 0.49]
Self-compassion				0.91	.344	3.36	.048*	1.94	.157
T1	0.04 [−0.33, 0.40]	.840	N/A
T2	0.29 [−0.08, 0.65]	.119	0.45 [−0.11, 1.01]
T3	0.13 [−0.20, 0.46]	.422	0.24 [−0.32, 0.80]
Self-kindness				0.30	.587	0.64	.497	0.05	.917
T1	−0.08 [−0.59, 0.43]	.756	N/A
T2	−0.12 [−0.64, 0.40]	.645	−0.13 [−0.69, 0.42]
T3	−0.16 [−0.71, 0.39]	.559	−0.17 [−0.72, 0.39]
Common humanity				0.19	.667	2.16	.121	0.14	.871
T1	0.80 [−0.56, 0.72]	.801	N/A
T2	0.06 [−0.58, 0.70]	.851	0.054 [−0.50, 0.61]
T3	0.20 [−0.40, 0.80]	.504	0.19 [−0.37, 0.75]
Mindfulness				0.27	.606	1.14	.325	0.30	.743
T1	0.20 [−0.43, 0.83]	.529	N/A
T2	0.20 [−0.41, 0.81]	.516	0.19 [−0.37, 0.74]
T3	0.04 [−0.61, 0.69]	.902	0.035 [−0.52, 0.59]
Self-judgment				2.97	.092	0.50	.609	3.84	.025*
T1	0.08 [−0.40, 0.56]	.740	N/A
T2	0.72 [0.21, 1.23]	.006**	0.81 [0.24, 1.39]
T3	0.30 [−0.23, 0.83]	.264	0.32 [−0.24, 0.88]
Isolation				1.05	.311	2.01	.140	2.27	.109
T1	0.02 [−0.63, 0.67]	.951	N/A
T2	0.52 [−0.08, 1.12]	.088	0.49 [−0.069, 1.06]
T3	0.30 [−.30, 0.90]	.316	0.29 [−0.27, 0.84]
Over-identification				0.60	.649	0.38	.685	1.38	.255
T1	−0.08 [−0.68, 0.52]	.788	N/A
T2	0.34 [−0.30, 0.98]	.290	0.30 [−0.25, 0.86]
T3	0.12 [−0.53, 0.77]	.712	0.11 [−0.45, 0.66]

Note. SHE = self-compassion, health, and empowerment; IPV = intimate partner violence; PTSD = post-traumatic stress disorder; T1 = baseline; T2 = immediately after the intervention; T3 = 8 weeks after the intervention.

Figure 2.

Changes in self-judgment over time in the SHE and control groups.

SHE = Self-Compassion, Health, and Empowerment; T1 = Baseline; T2 = Immediately after the intervention; T3 = 8 Weeks after the intervention.

Although failing to reach statistical significance, the SHE intervention group showed consistent trends toward improvements in most of the outcome measures, including negotiation, psychological aggression, sexual coercion, depressive symptoms, PTSD symptoms, self-compassion, isolation, and over-identification when compared to the control group. Importantly, our sensitivity analysis, which only included cases with complete data, yielded similar results to the intention-to-treat analysis, thus not reported here.

Intervention Adherence

Of the 25 participants in the SHE intervention group, five were lost to follow-up during the intervention delivery stage. Among the remaining 20 participants, one withdrew from the study after the first session, while one completed the first session only. A further one participant completed the first five sessions. The remaining 16 participants completed the entire intervention, resulting in an intervention completion rate of 64%.

Discussion

This is a pilot intervention study conducted in the U.S. that specifically focuses on Chinese immigrant women who have experienced IPV. Using a randomized controlled group design, the study examined the preliminary efficacy of our SHE intervention in reducing IPV and improving mental health well-being among Chinese immigrant women who have experienced IPV. The majority of participants adhered to the intervention protocol, and the study also provides preliminary evidence supporting the efficacy of our SHE intervention, laying a foundation for future research.

Our study reported a response rate of eligible survivors who chose to enroll close to that of the DOVE study (60.5%), which involved a sample of pregnant women experiencing IPV (Sharps et al., 2016). The satisfactory response rate and intervention completion rate suggest that our intervention using an mHealth delivery approach is feasible and acceptable to Chinese immigrant IPV survivors. It offers a promising way for this underserved and isolated group of women to access necessary support. A fully powered RCT is needed to demonstrate the effectiveness of our intervention before considering its broader implementation in Chinese immigrant women with various social and immigration status.

We found a significant decrease in self-judgment, which is one of the components of self-compassion, in our SHE intervention group as compared to the control group. This finding is consistent with a previous study supporting the benefit of self-compassion training for improving self-compassion, depression, and anxiety in a Chinese community sample (Finlay-Jones et al., 2018). This may be a particularly salient finding in the specific context of Chinese culture, where criticism and self-criticism are highly encouraged and valued, and individuals are more likely to use self-criticism as a way of striving for being a good self (Chen et al., 2018). As influenced by this cultural value of criticism and self-criticism, Chinese immigrant women may criticize or judge themselves for the violence that happened to them. Consequently, the self-compassion component of our intervention proved particularly beneficial for Chinese immigrant women with a tendency of self-criticism in dealing with an experience of IPV.

Participants in both the intervention and control groups demonstrated significant reductions in psychological aggression and depressive, anxiety, and PTSD symptoms, along with improved self-compassion levels. This suggests that simply offering written materials with information about IPV and mental health resources would be beneficial to those who suffer from IPV. In addition, notable trends toward improvement were observed in our SHE intervention group in comparison to the control group across most outcome measures, including specific types of IPV (i.e., negotiation, psychological aggression, and sexual coercion), depressive symptoms, PTSD symptoms, self-compassion, and certain components of self-compassion (i.e., isolation and over-identification). These findings provide encouraging indications of the potential efficacy of our SHE intervention in addressing IPV and improving the mental health well-being of Chinese immigrant IPV survivors. However, a future RCT with an adequate sample size is warranted to assess whether these trends toward improvement are statistically significant.

There were several limitations in this study. First, as this was a pilot study with a small sample size, the study may have lacked sufficient statistical power to detect significant effects. Although we observed trends toward improvement in certain outcome measures in our SHE intervention group compared to the control group, the statistical significance of these trends remains to be confirmed. It is possible that these group differences would have reached significance in a larger sample of women. Second, our participants were recruited online, mostly via WeChat, which may have introduced potential selection bias by inadvertently excluding older women or women with low digital literacy who might be less inclined to use social media. Consequently, our sample may not be representative of the broader population of Chinese immigrant women residing in the US. However, using the online recruitment approach, we were able to reach Chinese immigrant women across different regions of the U.S. efficiently and achieve the target sample size within six months. More importantly, we successfully enrolled a significant proportion of participants who had no income (34.7%) and were unemployed (36%) by advertising on WeChat. Overall, the online recruitment method has proved to be a feasible and effective way to reach Chinese immigrant women in the U.S. Third, we did not collect data on participants’ sexuality, number of marriages, place of residence (rural or urban areas), and their experiences of seeking family or institutional support. Particularly, participants’ utilization of family or institutional support may have confounded the effects of our intervention. Lastly, it is important to note that participants self-reported their daily completion of the intervention, which may have resulted in inaccurate measurement of adherence to the intervention protocol. Future research may consider using alternative intervention delivery approach that incorporates backend recording of daily practice duration to improve the measurement accuracy of intervention adherence.

Despite these limitations, this study provides important preliminary evidence on the potential efficacy of an empowerment-based intervention for Chinese immigrant women experiencing IPV. If demonstrated effective in a future RCT with a more robust sample size, our intervention has the potential to be disseminated in the Chinese populations with various immigration status in the U.S. Collaborating with the Chinese organizations and seeking the support of influential community leaders could promote the dissemination of our intervention. For example, outreach events could be conducted in partnership with Chinese organizations or community leaders, including workshops, information sessions, or community events. Chinese social media platforms such as WeChat could also be utilized to disseminate information about the intervention by creating engaging content in Chinese such as articles, videos, and infographics. In addition, our intervention could be adapted and translated for other immigrant groups of women to help them better cope with IPV and mitigate its mental health impacts.

In conclusion, our SHE intervention, comprising a brochure-based phone conversation providing IPV information, resources, and safety planning, along with six weekly relaxation and mindfulness self-compassion exercises, demonstrates promising potential in reducing IPV and improving the mental health well-being of Chinese immigrant survivors. These findings lay a strong foundation for a future large-scale RCT to further validate its efficacy.

Author Biographies

Yang Li, PhD, RN, is an Assistant Professor in the School of Nursing, the University of Texas at Austin. Her research focuses on the effects of lifetime trauma, including adverse childhood experiences and intimate partner violence, on women’s health, with a particular focus on mental health and childbearing outcomes.

Hyekyun Rhee, PhD, RN, FAAN, is a Professor in the School of Nursing, the University of Texas at Austin. She has established a robust research program focused specifically on asthma in adolescents based on several large clinical trials supported by the NIH. Specifically, her focus has been on developing and evaluating innovative asthma self-management interventions that capitalize on unique developmental opportunities presented by adolescents and cutting-edge technologies that can ameliorate the burdens of the disease.

Linda F. C. Bullock, PhD, RN, FAAN, is a Professor Emerita in the Sinclair School of Nursing at the University of Missouri. Her research has focused on domestic violence in a population of pregnant women and has included collaborating with nurses nationally and internationally and with members of other disciplines on the MU campus. Specifically, she has investigated the prevalence of intimate partner violence, its consequences to mother and fetus, and intervention strategies to improve health outcomes in both the woman and child.

Brigid McCaw, MD, MPH, MS, FACP, is a Senior Clinical Advisor at the Center to Advance Trauma-Informed Health Care of University of California. She led the implementation of a comprehensive, coordinated approach for improving screening, identification, and services for family violence. She guides the national Kaiser Permanente efforts in this area, impacting 10 million members. Her leadership, research, and publications focus on developing a health systems response to family violence, adverse childhood experiences, and trauma-informed care.

Tina Bloom, PhD, MPH, RN, is an Associate Professor in the School of Nursing at the Notre Dame of Maryland University. Her research focused on developing and testing effective violence interventions that decrease exposure to intimate partner violence as well as decrease the negative impact that violence exposure has on their health.