The Feasibility and Acceptability of Project POWER: A Cognitive-Behavioral Mindfulness-Infused Group-Based Mental and Sexual Health Intervention Among Young Pregnant Women in Liberia

Following 14 years of civil war in Liberia, war-exposure, gender-based violence, and extreme poverty have been identied as key challenges to the mental and sexual health of young pregnant women, and the health of their unborn children. Despite ongoing efforts to rebuild the country’s healthcare infrastructure, empirical research and interventions focused on addressing the consequences of war on the mental and sexual health of young pregnant women in Liberia are severely limited. To address these concerns, we developed Project POWER (Progressing Our Well-being, Emotions, and Relationships) (POWER), a mindfulness-infused cognitive-behavioral intervention for young adult pregnant women. This study sought to: 1) assess the feasibility and acceptability of POWER; and (2) determine the preliminary ecacy of POWER for improving mental and sexual health outcomes among Liberian war-exposed pregnant young adult women.


Introduction
Exposure to traumatic events has been consistently linked to mental and sexual health problems in postcon ict settings [1,2,3]. Between 1989 and 2003, Liberia experienced a deadly civil war that resulted in over 150,000 deaths and war-related trauma exposure in 60% of the population [4]. Human rights violations (e.g., sexual violence and torture) were common during the war and have been associated with adverse mental and sexual health outcomes for many Liberians [5], including high rates of depression and post-traumatic stress [6]. Sexual and gender-based violence (SGBV) and unsafe sexual practices (e.g., transactional sex) during and post-con ict have also been linked to increased rates of sexual and reproductive health problems [7,8], including sexually transmitted diseases (STIs), abortions, and unwanted pregnancies [9,5]. Existing health services are still recovering from the war and not equipped to address unmet mental and sexual health care service needs of women [3,10,11,12].
For pregnant young women in Liberia and other low-and middle-income countries (LMICS), the need for integrated mental and sexual health services is high. Prenatal distress-stress, anxiety, or depression during pregnancy-is signi cantly higher in LMICs than in high-income countries [13]. Chronic stress throughout pregnancy places women at risk for neonatal and obstetric complications, such as premature birth, postpartum depression, and unplanned cesarean delivery [14]. In addition to negative impacts on mothers, prenatal distress during pregnancy is associated with adverse perinatal outcomes among infants, such as low birth weight and preterm birth [15,16]. Implementing mental and sexual health interventions for pregnant women in LMICs is essential for women's health and their unborn children [17].
To address the need for integrated mental and sexual health services for young Liberian women, we developed and culturally adapted a mindfulness-infused cognitive-behavioral intervention called Progressing Our Well-being, Emotions, and Relationships (POWER) to improve young pregnant women's mental and sexual health in Liberia (funded by NIH/FIC grant number 1K01TW009660). There is strong evidence supporting the effectiveness of mindfulness-based practices-paying attention in the present moment without judgment-in reducing anxiety, depression, and general psychological distress in pregnant women [18,19]. The current study's objectives are: (a) to assess the feasibility and acceptability of POWER and (b) to evaluate the preliminary e cacy of POWER on mental and sexual health outcomes among war-exposed pregnant young adult women in Liberia.

Participant Recruitment and Procedures
Participants were recruited using purposive sampling in three catchment areas located in Montserrado county-the most populated county in Liberia-through a well-known local health clinic. Two catchments received POWER and one received the control condition. One nurse per clinic facilitated recruitment through: (1) gauging interest in the study during prenatal care appointments, (2) asking potential participants if they knew other pregnant women (i.e., snowball sampling), and (3) using existing clinic programming roasters. If the potential participant expressed interest in the program, the nurse provided their contact information to a research team member, who then scheduled an in-person screening.
Study inclusion criteria were: (a) receiving prenatal or health services through the clinic; (b) 18-25 years old; (c) residing in Monrovia, and (d) between 13 and 24 weeks of gestational age. Exclusion criteria included pregnancy-related medical problems. The gestation age between 13 and 24 weeks ensured that women were past the rst trimester, which comprises several risks and decisions regarding pregnancy. This time frame also increased the likelihood that women completed the 3-month follow-up before delivery. Those interested in participating who met inclusion criteria provided informed consent to participate in the study. Participants received the equivalent of $9 ($1 for travel, $3 for a meal, and $5 for completing assessments). Participation in the study was voluntary and con dential, and healthcare services were not affected by participation.

Study Design
This open pilot study used a two-group (POWER and control) pre-post design to explore feasibility, acceptability and preliminary e cacy of POWER among young pregnant women. Both groups completed quantitative assessments on mental and sexual health at baseline and 3-month follow-up. Participants in the POWER group completed a mixed-methods exit interview to assess experiences with the intervention (e.g., satisfaction with facilitators and implementation of adaptive coping skills).
Before conducting the open-pilot study, three Liberian experts reviewed intervention and control group materials extensively. Experts were mental health services providers or sexual health services providers (i.e., social workers, nurses, midwives). We conducted a two-week pre-pilot phase for both conditions to gather feedback on content, acceptability, and cultural relevance of intervention materials from a matched sub-sample of young pregnant women. Facilitator training was conducted for both conditions.
The PI/developer of POWER delivered a one-week facilitator trainings for both conditions. During the prepilot, participants provided feedback to facilitators regarding their facilitation skills. Facilitators participated in weekly supervision with the PI and program supervisor.

POWER Condition
POWER is a ten-session intervention delivered two times per week over ve weeks. It incorporates skills and concepts derived from mindfulness and cognitive-behavioral therapy-based interventions. The primary goals of POWER are to: (1) increasing knowledge and awareness of stress, emotional regulation, and prenatal distress; (2) promoting understanding and knowledge of sexual values and risk reduction strategies; and (3) increasing knowledge and awareness of healthy interpersonal relationships (see Table   1 for more session details). In the open trial, sessions were cofacilitated by a nurse and social worker, and each session lasted approximately 90 minutes.

Control Condition
Our team developed a general health education program for the control condition consisting of tensessions delivered two times a week over ve weeks. The general health curriculum provided information about the body and infectious disease control, hygiene, water and food safety, nutrition, substance use, mosquitos and malaria, hemorrhagic fevers, and infectious diseases. Each session lasted approximately 90 minutes. Intervention content was delivered by a nurse who was trained in the curriculum.

Measures
All study measures were culturally adapted using Beaton's [20] guidelines for cross-cultural adaptation of self-reported measures. Due to high illiteracy rates, all measures were administered verbally to the participant using a computer-assisted program by a trained female research assistant. Our team collected demographics (i.e., age, level of education, relationship status, number of children) and secondary outcome measures (i.e., mental and sexual health questions) from participants at baseline and 3-month follow-up.
Primary Outcomes: Feasibility and Acceptability of POWER We assessed the feasibility of POWER by examining the following: (1) number of potential participants screened; (2)  Program acceptability was operationalized as satisfaction with the process [21,22]. We assessed the acceptability of POWER by examining the following reports: (1) POWER sessions that were liked most and least, (2) skills learned from POWER, (3) how often participants practiced sessions outside of the group, and (4) participants satisfaction with POWER facilitators.

Secondary Outcome Measures: Mental and Sexual Health
Patient Health Questionnaire-9 (PHQ-9) [23]. The PHQ-9 assesses nine major depressive disorder symptoms, as de ned by the DSM-IV. Participants were asked how often they experienced each symptom over the last two weeks. The responses were modi ed for this study based on suggestions from the cultural adaptation process. The original range was 0 (not at all) to 3 (nearly every day). The range was modi ed from 0 (not at all) to 2 (every day) to measure severity. Higher scores on the severity scale indicated that depressive symptoms were more severe. The Cronbach's alpha was .72.
Prenatal Distress Questionnaire (PDQ). The PDQ is an 18-item measure that assesses the extent to which respondents are concerned, troubled, or worried about issues related to their pregnancy [24]. Items are rated on a 3-point Likert scale, 0 (not at all) to 2 (every day). Scores are added to create a total score.
Higher scores indicate higher levels of distress. The Cronbach's alpha was .88.
Life Problem Checklist (LPC). The LPC is a 28-item measure adapted from the General Life Stressors subscale [24]. Response options range from 1 (nothing like that) to 5 (always). Mean scores are calculated to derive a total score. Higher scores indicate a greater likelihood of stressors taking place. The Cronbach's alpha was .84.
PTSD Symptoms (Post Traumatic Stress Disorder Checklist-Civilian Version) (PCL) [26]. The PCL-C is a 17-item PTSD Checklist. Participants rated the severity of PTSD symptoms over the previous 30 days. The adapted response scale ranged from 0 (not at all) to 3 (all the time). The Cronbach's alpha was .76.
Transactional Sex (TSC). The TSC is a 9-item measure adapted from the Transactional Sex Scale [27]. The measure assessed three different behaviors: 1) staying in a relationship longer than one wants; 2) starting a new relationship; 3) engaging in sex strictly for monetary/other goods for three reasons: (1) paying for items they could not buy themselves, (2) paying for food, children's school fees, or taking care of the home, and (3) supporting children or family who depends on them nancially. The adapted response scale was dichotomized (yes = 1 or no = 0). Scores are calculating by summing items. Participants were considered to have engaged in transactional sex if they indicated they received a score above zero.

Data Analysis
Descriptives statistics were computed to assess the feasibility and acceptability of POWER. To investigate whether POWER demonstrated e cacy in improving mental and sexual health outcomes, we performed a series of four mixed with time (pre-intervention, post-intervention) as the within-subjects' factor and group (intervention, control) as the between-subjects' factor. Analyses were conducted using SPSS 25.0. Signi cance was established at an alpha level of .05, two-tailed. Cohen's d represents effect size. Only participants with complete data were included in the analyses. Twenty participants were dropped from analyses: one did not have complete pre-test data, and nineteen did not have post-test data.

Participant Characteristics
Eighty-seven women participated in the study (n = 45 in Power; n = 42 in control). Table 2 displays demographic characteristics. Participants were 18-25 years old (M = 20.72, SD = 2.24). Eight percent of the sample did not attend school, approximately 24% had a primary school education, 30% had a secondary education, and 30% had a high education or higher. The majority of the sample did not have living children (80% POWER (n = 36), 69% control (n = 29)). Only one participant indicated that she was not in a relationship. Based on a 4-point likert scale rating, the average rating for facilitation satisfaction was 2.51. Communication was identi ed as an area for facilitator improvement. Participants reported that the facilitator "could speak and make her words clearer" and "she could speak slower." Participants' feedback about their likes centered on themes of facilitator helpfulness (e.g., "she make me understand"; "she always talks to me to help me manage my stress"; "she is very encouraging people and explains the topics well").
Participants reported positive feelings and also some sadness about program completion. For example, one participant said, "I feel good because it educated me along." Other participants said, "I feel good and bad because the program empowered me by more knowledge… I feel bad because I won't be seeing the program again", and indicated a desire to continue learning (e.g., "I don't want the program to end," "I want to learn more").

Preliminary E cacy: Mental and Sexual Health Outcomes
The effect sizes for the secondary outcomes are provided in Table 3. These include Cohen's D's for time, group, and group by time estimates.

Discussion
The current study explored the feasibility, acceptability, and preliminary e cacy of an innovative mindfulness-infused cognitive-behavioral intervention for young adult pregnant women in post-con ict Liberia. POWER is the rst evidence-based mental and sexual health intervention that has been implemented in post-con ict Liberia. Study ndings support the preliminary bene ts of the intervention and the need for a larger, well-powered implementation-effectiveness trial to examine clinical effectiveness and implementation strategies that could enhance adoption and sustainment within Liberia's existing health service infrastructure.
We successfully enrolled the target number of participants, and retention rates were acceptable. Participant reported a high level of satisfaction with POWER sessions and practicing skills learned outside of the group, including meditation, muscle relaxation, deep breathing, and walking. Findings on intervention skills used most and least could inform further adaptations of POWER to best meet young pregnant Liberian women's needs. For example, we could incorporate additional mindfulness-based activities, such as pregnancy-friendly yoga poses or focusing on bodily sensations.
All Participants reported signi cant reductions in depression symptoms at 3-month follow-up. Findings are consistent with prior research showing mindfulness-based interventions improve maternal depression, stress, and emotion regulation skills [28, 29,30]. Though preliminary, ndings may also have implications for children's health because maternal depression during pregnancy and postpartum periods has been strongly associated with poor child development outcomes [31,32].
POWER and control participants reported signi cantly decreased prenatal distress and PTSD symptoms at 3-month follow-up. Though speculative, the focus on cultivating mindfulness skills may have helped participants increase their distress tolerance [19]. Given the adverse consequences of prenatal distress for maternal and child health, including adverse birth outcomes, lower likelihood of breastfeeding, and de cits in children's cognitive and behavioral function [15,13], effective services targeting a reduction of distress during pregnancy are essential. Findings suggest that the health education curriculum may also help reduce distress among young pregnant mothers and could be an alternative resource.
Participants reported signi cant reductions in frequency of transactional sex at 3-month follow-up, but there was a time by group interaction. Transactional sex decreased for those in controls only. One possible explanation is that a larger percentage of control participants reported transactional sex at baseline than did POWER participants Further adaptations and re nements to POWER may be necessary to better target reducing transactional sex, such as education about the health risks of transactional sex or strategies to secure income in healthier ways (e.g., economic empowerment).

Limitations
Study results should be viewed in light of several limitations. The sample size was small and not powered to determine clinically signi cant treatment effects or generalize to other populations. Longerterm bene ts of POWER were also not evaluated. Including a longer follow-up period in a future wellpowered trial would provide more information on whether intervention bene ts are sustained over time and whether there are intervention bene ts for early childhood development outcomes. Additionally, because participants were not randomly assigned to conditions, there is the potential for sampling and selection biases.

Conclusion
Preliminary nding of POWER support feasibility, acceptability, and mental and sexual health bene ts of the intervention for young pregnant Liberian women. Further investigation of POWER on a larger scale is needed to determine intervention effectiveness and ultimately expand the reach of evidence-based sexual and mental health services among young pregnant women in Liberia and other post-con ict settings to improve women's health and the health of future generations [17].