“I became alright after birth”: Mechanisms driving improvements in mental health symptoms in the postpartum transition among persons at risk for HIV in South Africa

Abstract

Background:

Pregnant and postpartum persons (PPPs) are at increased risk for HIV acquisition, and depression and posttraumatic stress disorder (PTSD) negatively impact engagement in HIV prevention behaviors like pre-exposure prophylaxis (PrEP) use, thereby increasing risk.

Objectives:

The present study explored changes in mental health symptoms from pregnancy to postpartum to inform future interventions for PPP that address mental health and HIV prevention.

Design:

This analysis is part of a larger, mixed-methods study conducted in South Africa that examined antenatal mental health barriers to PrEP use, employing an explanatory sequential design.

Methods:

Participants recruited from an antenatal clinic were pregnant or postpartum, over 18, not on PrEP, and with no history of PrEP use. Participants first took a survey to assess depression and PTSD symptoms. Those with elevated symptoms during pregnancy completed qualitative interviews during pregnancy and postpartum. Data were analyzed via thematic analysis.

Results:

Of 110 survey participants, 23 completed qualitative interviews (10 pregnancy only, 6 postpartum only, 7 both). This analysis includes 13 participants who completed either both interviews or postpartum only. Three themes illustrated processes linked to symptom reduction postpartum: (1) increased feelings of empowerment (e.g., via acceptance of life circumstances), (2) improvements in emotional and tangible support (e.g., via improved communication, increased caretaking and financial support), and (3) conceptualizations of infants and motherhood as sources of joy and motivation (e.g., pride in maternal role, companionship with baby).

Conclusion:

These themes highlight processes that may contribute to reductions in depression and PTSD symptoms postpartum, which could be integrated into interventions targeting mental health and HIV prevention during pregnancy. Intervention components may include skills promoting self-efficacy, problem-solving, communication, and identifying sources of joy and meaning. Interventions leveraging these mechanisms of symptom improvement during the postpartum transition may enhance mental health during pregnancy and promote greater engagement in HIV prevention behaviors.

Introduction

HIV places a high public health burden on South Africa (SA) and disproportionately affects cisgender women of reproductive age. Risk for HIV acquisition increases during pregnancy and the postpartum period due to myriad biopsychosocial factors.^{1 –3} These include hormone changes like increased levels of estrogen and progesterone, immune system changes that reduce the ability to fight off HIV and other infections, and condomless sexual activity due to lower perceived risk.^{3 –7} Given these biological changes and relevant behavioral factors, uptake and adherence to pre-exposure prophylaxis (PrEP), or the use of antiretroviral medications to prevent HIV, among pregnant and postpartum persons (PPPs) could substantially reduce maternal and infant HIV acquisition. ² Oral PrEP, which prevents HIV infection when taken consistently, is safe to use during pregnancy, recommended for use by the WHO, and available in the public sector in SA. ⁸ Given increased risk for HIV acquisition and increased engagement with healthcare during the peripartum period, researchers have suggested that this window is a critical time to initiate and continue PrEP use.^9,10

Mental health concerns negatively affect HIV prevention behaviors, including PrEP use. Depression is associated with reduced uptake of and adherence to PrEP,^11,12 and the negative effects of depression on PrEP adherence may be even more salient for PPP.^9,10 Among serodiscordant couples in Kenya and Uganda, depression was significantly associated with low PrEP adherence in women (adjusted risk ratio (aRR) = 1.77). ¹¹ In a study based in SA, women with depressive symptoms were less likely to have higher PrEP adherence than those without symptoms (aRR = 0.79), and symptoms negatively affected PrEP adherence even after controlling for the effects of stigma (aRR = 0.74) and PrEP optimism (0.75). ¹² Proposed mechanisms driving the effect of depression on PrEP adherence include withdrawal and social isolation, leading to lower healthcare engagement and decreased self-care. ¹² Additionally, emotional symptoms like sadness, anhedonia, shame, and stress coupled with cognitive symptoms like decreased concentration and forgetfulness may lead to missed doses or delayed pickup of PrEP at clinics. ¹³ Decreased frequency of sexual intercourse related to depression may also affect the perceived need for PrEP. ¹² Given PPP’s elevated risk, for both depression and HIV,^3,14 addressing perinatal mental health is likely an important aspect of HIV prevention interventions during pregnancy and into the postpartum period.

In addition to negative impacts on PrEP use and other HIV prevention behaviors, perinatal mental health concerns may also lead to poor health outcomes in both the birthing parent and the infant. In general, longer duration of untreated depressive symptoms is associated with longer response times to treatment, higher relapse rates, and higher rates of depression-related disability.^15,16 Perinatal depression is a risk factor for adverse health outcomes such as low birth weight, preterm births, birthing challenges like preeclampsia, and shorter breastfeeding periods.^{14,16 –18} Furthermore, depression in the birthing parent may be detrimental to the child’s emotional, cognitive, and behavioral development.^19,20 Early intervention and adequate treatment of perinatal mental health may mitigate the long-term negative effects associated with untreated mental health symptoms, ultimately promoting better health and developmental outcomes. In HIV endemic settings, combined mental health and HIV prevention interventions may be an efficient approach to addressing both of these key issues, especially during the perinatal period.

Our study also explored posttraumatic stress disorder (PTSD) symptoms during the postpartum transition, due to high rates of exposure to violence and trauma in sub-Saharan Africa and in SA specifically.^21,22 Gender-based violence has profound mental health effects and significantly elevates HIV risk^{23 –25} while also increasing the likelihood of pregnancy. Intimate partner violence (IPV) is also highly prevalent in SA and may compromise women’s social, sexual, and reproductive agency in their primary relationships, which negatively affects mental health and increases HIV risk.^26,27 Among South African women, prior research has demonstrated strong links between IPV and depression during pregnancy. ²² A primary mechanism through which PTSD may negatively impact HIV prevention behaviors is avoidance; for example, PPPs have cited factors such as sleep disturbances and reluctance to interact with healthcare providers as contributing to healthcare avoidance ¹³ Like depression, PTSD is also associated with adverse health outcomes for birthing parents and infants; substance use disorders, panic disorder, depression, and pre-term delivery are all significantly associated with a PTSD diagnosis in the birthing parent. ²⁸ Exposure to trauma or violence also negatively impacts the parent–child relationship, with mothers who have experienced childhood abuse or IPV reporting less responsiveness, lower empathy, hostility, and aggression toward their children.^29,30 Regarding infant outcomes, infants born to parents who have PTSD have poorer fine motor skills and adaptive behaviors (e.g., communication, self-direction) relative to infants of parents who do not carry a PTSD diagnosis. ³¹

Depression and PTSD frequently co-occur and share many underlying mechanisms that negatively impact maternal mental health. Understanding their common underlying mechanisms, especially in the SA context, and how they change throughout pregnancy and postpartum could inform the development, adaptation, and timing of future interventions, which to our knowledge, has not been specifically explored. Changes in depression symptoms may reflect shifts that occur during the postpartum transition, including hormonal changes, breastfeeding, lack of sleep, financial burden, and relationship changes.^{32 –34} Though there is mixed evidence regarding differences in depression symptoms pre- and postpartum, several studies conducted in sub-Saharan Africa have documented decreases in depression symptoms post-delivery.^{35 –37} PTSD symptoms similarly displayed mixed trends throughout pregnancy and postpartum, with some studies demonstrating a decrease in symptoms ³⁸ and others demonstrating an increase from pregnancy to postpartum. ³⁹ Significant predictors of higher PTSD symptom severity throughout pregnancy and postpartum are childhood abuse, lower income, and lower educational attainment, ³⁸ which are all highly relevant to the South African context. A recent systematic review also identified IPV as a significant predictor of poor perinatal PTSD and depression outcomes, ⁴⁰ which is also relevant to this setting.

Our study seeks to better understand the specific mechanisms driving changes in mental health symptoms during the peripartum period in a high HIV prevalence setting like SA, where negative mental health symptoms during pregnancy have implications for HIV outcomes. Not only is it important to understand the ways in which mental health symptoms shift during this period, it is also critical to explore the processes that contribute to these shifts. This analysis explores participant-reported changes in mental health symptoms over the postpartum transition and identifies potential mechanisms that may be driving these changes. Given the public health significance of depression and PTSD from the perspectives of both maternal/infant health and HIV prevention, mapping symptom change and elucidating relevant mechanisms that drive those changes will inform future mental health and HIV prevention interventions for PPP.

Method

Parent study

The present study is a secondary analysis of a larger, mixed methods project investigating depression, posttraumatic stress, and other mental health barriers to PrEP uptake and continued use during pregnancy and the postpartum period in Cape Town, SA. ¹³ Eligible participants were aged 18 or older, HIV-negative, presenting for antenatal care, not on PrEP, and had no history of PrEP use. Pregnant persons were recruited using convenience sampling from an antenatal clinic in Cape Town, by trained research assistants, when they presented for their first antenatal appointment. All women who were presenting for the first antenatal visit were approached in person while waiting in the clinic queue. The study team guarded against selection bias as all participants were newly presenting for care, and they did not have a prior relationship established with the research assistants conducting the interviews. Participants were informed that the interviewers were trained research assistants from the University of Cape Town (UCT). They were told that the purpose of the study was to learn about what might affect their decision to use PrEP during pregnancy and while breastfeeding to learn how to support women to use PrEP during pregnancy and breastfeeding. Individuals who were interested in participating were screened for eligibility and provided written informed consent. The interviews were conducted in a private area, and it was ensured they did not lose their place in line for their visit. Interviews were audio recorded then transcribed and translated.

The study utilized an explanatory sequential design, first collecting cross-sectional survey data to determine which participants would be eligible for a qualitative interview, then conducting interviews among those with elevated depression and PTSD symptoms. Data were collected from May 2022 to December 2023. First, a total of 110 participants completed a survey, which included measures that probed attitudes toward PrEP, PrEP optimism, and likelihood of initiating PrEP during pregnancy, as well as depression, posttraumatic stress, and related psychosocial factors, like HIV stigma, relationship power, and dyadic trust. Most of the constructs measured in the survey are not relevant to the current analysis. Survey data were collected using REDCap, an online data collection platform. ⁴¹

A subset of individuals (n = 44) whose scores on self-reported measures of mental health symptoms indicated that they had likely depression and/or PTSD were invited to complete a semi-structured qualitative interview prior to delivery and a follow-up qualitative interview post-delivery. Of those who qualified, 23 participants completed either 1 or 2 interviews for a total of 30 interviews: 10 completed an interview during pregnancy only, 6 completed an interview postpartum only, and 7 were interviewed both during pregnancy and postpartum. Some participants only completed the pregnancy interview and were lost to follow up before the postpartum interview or chose not to continue with the study, and some participants gave birth before the scheduled pregnancy interview and only completed a postpartum interview. Qualitative interview data were collected until thematic saturation was reached. ⁴²

Semi-structured interview guides for the interviews were developed in accordance with guidelines specified by Miles and Huberman, ⁴³ utilizing open-ended questions to facilitate unbiased responding. The pregnancy interview probed for anticipated psychological barriers and facilitators to PrEP uptake, adherence, and persistence during pregnancy, as well as for experiences with mental health symptoms. The postpartum interview followed up on mental health symptoms reported in the initial interview and inquired about changes in symptoms postpartum, as well as any new challenges specific to the postpartum period that they perceived might impact PrEP use during this time. Additionally, both interviews explored how specific beliefs about PrEP and underlying psychological processes common to depression and PTSD (e.g., avoidance, withdrawal) may affect PrEP use during pregnancy/breastfeeding. The PrEP findings are not included in the current analysis.

Due to the severity of some reported mental health problems, study staff monitored suicide risk. Study staff were trained to address ideation, intent, means, and plan. In the event participants had active ideation, intent, means, and a plan, they were deemed ineligible for the study to prioritize their safety. Per protocol, participants received a referral for a psychiatric evaluation at the closest nearby hospital, and staff followed up to ensure an evaluation was scheduled. However, no participants required additional follow-up or further assessment for suicidality.

Measures

In the survey, participants completed measures assessing depression and PTSD to determine eligibility for the pregnancy interview. Two sets of study tools were prepared during study startup, one in English and one in isiXhosa. Participants were given the option to complete measures and interviews in their preferred language. Measures were translated to isiXhosa and back translated into English using the Brislin translation method for cross-cultural research. ⁴⁴ All measures used in the study were previously validated for use in SA,^{45 –49} and the Edinburgh Postnatal Depression Scale (EPDS) ⁵⁰ and PTSD checklist for DSM-5 (PCL-5) ⁵¹ have also been validated among pregnant/postpartum persons.^46,52

Depression was assessed via the EPDS and the South African Depression Scale (SADS). ⁴⁸ The EPDS measures depression symptoms over the past 7 days with 10 items (e.g., “I have been able to laugh and see the funny side of things,” “I have blamed myself unnecessarily when things went wrong”) rated from 0 to 3, with higher scores indicating greater depressive symptoms. The 17-item SADS, originally developed for South African populations living with HIV, measures the number of days over the past week that one has experienced depression symptoms (e.g., “Things were not going well in my life,” “I had difficulties sleeping”). Items are rated on a Likert-style scale from 0 to 3, with higher scores indicating greater number of days experiencing symptoms. An EPDS score ⩾10 or SADS score ⩾23 was used to indicate probable depression. Participants who met criteria based on the EPDS, SADS, or both were deemed eligible to complete an interview. The EPDS score requirement was initially ⩾13 per the recommended cutoff, ⁵⁰ but due to challenges in recruitment, we lowered the eligibility criteria to ⩾10 in accordance with validated use in African contexts.^{45 –47}

Posttraumatic stress was measured via the PCL-5. The PCL-5 is a 20-item measure that assesses past week PTSD symptoms (e.g., “Repeated, disturbing, and unwanted memories of the stressful experience?,” “Feeling very upset when something reminded you of the stressful experience?”) rated on a 0–4 Likert-type scale, with higher scores indicating greater PTSD symptom severity. A PCL-5 score ⩾31 was used to indicate probable PTSD.

The semi-structured interviews were conducted by trained bilingual female research assistants, and participants again had the option to complete the interviews in English or isiXhosa. Interviews lasted approximately 1 h. The person who conducted the interview translated the transcript, and another bilingual staff member reviewed the translation while listening to the audio files, utilizing the Brislin translation method. ⁴⁴ Interview transcripts were not returned to participants for comments or correction. Interviewers were trained by the study staff at UCT. Field notes were not kept and only the interview transcripts were used for data analysis.

Data analysis for the present study

In the current analysis, we analyzed the thematic content of 20 interviews: the 13 postpartum interviews plus 7 pregnancy interviews linked to the 7 participants who completed both a pregnancy and a postpartum interview. For those who only completed a postpartum interview, we analyzed content in which they described differences in their mental health symptom presentation between pregnancy and postpartum. For those who completed both a pregnancy and a postpartum interview, we compared their interviews to explore changes in symptoms and factors driving those changes. Participants who only completed an interview during pregnancy were not included in this analysis.

The data were analyzed in accordance with the principles of thematic analysis established by Braun and Clarke. ⁵³ Transcripts were coded using the Dedoose software by authors KK and JL, with assistance from MF and LG. Four transcripts were open-coded to develop a codebook. After establishing the codebook, two transcripts were double-coded by KK and JL to ensure consistency between raters. The remaining transcripts were coded independently by KK and JL. Coders met regularly to discuss discrepancies in coding. The PI, AMS, oversaw the creation of the codebook and coding process. The Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines ⁵⁴ informed the reporting of our results. After coding was complete, thematic analysis was conducted by KK. Participants did not provide feedback on the findings.

To ensure analytical rigor, the research team followed the eight “Big-Tent” criteria for qualitative research outlined by Tracy. ⁵⁵ The topic was deemed worthy due to the significant public health burden of both HIV and mental health diagnoses during pregnancy and the importance of intervening during antenatal care. ¹⁰ We achieved rich rigor through careful translation and rigorous standards in creating the codebook. To approach this research with sincerity, we recognized our limitations of cultural understanding as American coders and therefore sought active collaboration and guidance from our South African colleagues who conducted the interviews and led the study at the local site (JNG and LG). For credibility, we ensured that all team members converged on similar themes through biweekly meetings. Reliability and consistency were achieved through the coding process described above. To highlight the resonance of our findings, the authors selected meaningful text, centering direct quotes from participants and offering rich descriptions of participant experiences. Finally, we achieved meaningful coherence by accomplishing the research objectives, utilizing methods consistent with our theoretical paradigm, and effectively situating our findings within the current literature, ensuring an interconnected and cohesive presentation of our research.

Ethical considerations were paramount in this study. This study was approved by ethics institutions at both sites, Massachusetts General Hospital (ethics reference #2021P001849) and the UCT (ethics reference #231/2021). The SA research team adhered to local protocols to ensure participants fully understood the study, maintained privacy, and received care as usual without interruption by study participation. Written informed consent was obtained, and participants were compensated fairly. We prioritized relational ethics to foster respect and dignity and took an anticolonial approach by building sustainable relationships with our SA-based colleagues, who co-designed the study and protocol.

Results

Participant characteristics

On average, participants (n = 13), were 25.8 years old (SD = 3.9). The majority of the sample identified as Black South African (n = 11, 84.6%), most were unemployed (n = 8, 61.5%) and earned less than 3000 ZAR (approximately $163) per month (n = 8, 61.5%). Participants had an average of 1.8 previous pregnancies (SD = 0.8) and 1.5 children (SD = 0.9), and at the time of the interview were 2.5 months (SD = 2.0) post-delivery. During pregnancy, seven participants met criteria for probable depression (53.8%), five met criteria for both probable depression and PTSD (38.5%), and one met criteria for probable PTSD only (7.7%). Table 1 provides further detail on participant characteristics.

Table 1.

Participant characteristics (N = 13).

Characteristic	Mean	SD
Age	25.8	3.9
Gestational age (weeks)	33.0	6.6
Weeks postpartum at postpartum interview	10.4	7.7
Total pregnancies, including current	1.8	0.8
Previous live births	1.5	0.9
Mental health symptom severity
Depression (SADS)	22.0	5.3
Depression (EPDS)	9.5	3.8
PTSD (PCL-5)	28.5	14.3
	N	%
Race
Black South African	11.0	84.6
Black non-South African	2.0	15.4
Education
Through grade 10/Std 8	1.0	7.7
Through Grade 11/Std 9	6.0	46.2
Through Grade 12/Std 10	5.0	38.5
Vocational	1.0	7.7
Monthly income in ZAR ($)
0 (0 USD)	2.0	21.7
Less than 3000 (162 USD)	6.0	47.8
3001–6000 (162–324 USD)	5.0	26.1
Qualified for interview by
Depression	7.0	53.8
Posttraumatic stress	1.0	7.7
Depression and posttraumatic stress	5.0	38.5

Data reported in Table 1 were collected at the initial visit when participants completed REDCap surveys during pregnancy. PTSD: posttraumatic stress disorder; SD: standard deviation; EPDS: Edinburgh Postnatal Depression Scale; SADS: South African Depression Scale; PCL-5: PTSD checklist for DSM-5.

Symptoms experienced during pregnancy

In the initial qualitative interviews, participants described their depression and/or PTSD symptoms experienced during pregnancy. Some expressed feelings of sadness, such as: “I would feel sad and wish that I could just cry, but I would have isingqala” (age 24). The isiXhosa term “isingqala” translates to “deep sorrow”; this term has been discussed in other analyses exploring cultural manifestations of depression in SA. ⁴⁸ Other common symptoms included worthlessness (e.g., “It makes me feel like I’m useless” (age 24)), hopelessness (e.g., “I am just feeling hopeless. . .there is no progress in my life. . .there’s nothing good about me” (age 31)), anhedonia (e.g., “I won’t know how I feel, I would just feel numb” (age 20)), and loss of interest in sexual activity (e.g., “I don’t want to have sexual intercourse with him when I am sad” (age 30)). Additionally, participants reported other symptoms such as weight loss (e.g., “I lost weight” (age 28)) and sleep disturbances (e.g., “I can’t sleep at night because of it” (age 24)). Participants also mentioned experiencing repeated thoughts of suicide (e.g., “There are times where I want to kill myself” (age 21)) and engaging in suicidal behaviors (e.g., “I tried taking pills, and I tried slitting myself” (age 21)).

Mechanisms driving decreases in symptoms during the postpartum transition

Three key themes emerged with respect to mechanisms driving observed decreases in depressive and PTSD symptoms during the postpartum transition: (1) increased feelings of empowerment, (2) improvements in emotional and tangible support, and (3) conceptualizations of infants and motherhood as sources of joy and motivation.

Increased feelings of empowerment

Feelings of sadness related to loneliness and a sense of weakness that were described during pregnancy improved postpartum as a result of increased feelings of empowerment. Participants reported feeling empowered by overcoming fears of loneliness after realizing their own capabilities, standing up for themselves to their families, and having an increased sense of self-efficacy after delivering their babies. For example, one participant with elevated symptoms of depression and PTSD lacked support from her family during pregnancy and feared loneliness during childbirth and postpartum. During her postpartum interview, when she was asked about how the birth of her child affected her and her sadness, she stated:

“At least now I am relieved the most thing I feared is that I did not want to give birth lonely. Because most of the time there were no people around me. . .Then after giving birth I became alright because there is nothing more that I have to fear. So, I can do things on my own. So, I became alright after birth.” (Age 31, 9.4 weeks postpartum)

Following childbirth, she was relieved of her fear of loneliness and realized she could handle motherhood on her own. She gained a sense of empowerment (“I can do things on my own”) as she was perhaps less impacted by loneliness, which she identified as a key stressor during pregnancy.

Another participant with elevated symptoms of depression reported feeling guilt, self-blame, and a sense of weakness during pregnancy, due to a perceived lack of control over her living environment, specifically challenging family dynamics and boundary violations. A source of stress for her was a stranger living in her family home who she suspected to be engaging in illicit activities, and she noted that her brother complicitly allowed such activity. During her postpartum interview, while reflecting on her home life and how it made her feel about herself as a mother, she expressed:

“I sometimes felt like blaming myself and felt like I was weak. I used to wonder how I will be able to stand up for my child if I can’t stand up for what was going on at home. I always used to consider other people instead of standing up for myself.” (Age 18, 12.6 weeks postpartum)

Postpartum, she reported that prior to the birth of her child, she stood up to her brother and removed the stranger (“So, I asked him to leave, and he left”). Her faith empowered her to take control of the situation (“What helped me is the fact that I strongly believe in God with every challenge that I face”). Taking control over her situation may have facilitated a shift in her thinking patterns from negativity and self-blame to a sense of empowerment; she stated: “I didn’t feel alright but at the same time I was able to think positively. I used to convince myself that I can take care of my child.” After childbirth, empowerment through faith and decisive action appears to have reassured her of her capability as a mother. This may have contributed to an increased ability to tolerate distress and overall reductions in symptoms, even when she “didn’t feel alright.”

In another example, a participant with elevated symptoms of depression and PTSD demonstrated a shift in thinking from pregnancy to the postpartum period. During her pregnancy interview, when asked about how sadness affects how she feels about herself as a person, she stated:

“Sometimes I feel like I am useless, I would feel like the dumbest person in the world, I sometimes undermine myself, I’m really dumb I am just feeling hopeless. There is no progress in my life, and I continue to do the same mistakes in my life there is nothing good about me.” (Age 31, 36 weeks pregnant)

In her postpartum interview, she expressed a different sentiment:

“I see myself as a great mother. I see myself as a better one because beyond all that I manage to cope on my own. . .I didn’t have any negative or funny thoughts. . . I feel I am very empowered because I do things myself. I can say I feel that I am extra strong now. I do not know how to put it. I am really strong now even with the baby, there is nothing to worry about even if maybe something happens.” (Age 31, 9.4 weeks postpartum)

Despite experiencing feelings of hopelessness and uselessness during pregnancy, she recognized her own strength as a mother, even in the face of uncertainty. The postpartum transition may have enabled her to discover her own self-efficacy. That is, her ability to “cope on her own” made her feel empowered and challenged her previous self-view by presenting information inconsistent with her internalized negative beliefs and previous negative thought patterns.

Improvements in emotional and tangible support

Most participants described a strong relationship between their mental health symptoms and a lack of tangible (i.e., financial assistance, support with childcare) or emotional support. After childbirth, participants described an increase in both types of support; for example, family members and partners assisting with caretaking, providing financial support, and offering opportunities to talk about their emotions. For example, one participant noted postpartum that her sadness stemmed from a lack of emotional support from her mother: “I was worried about what would happen if I’m in labour and [my mother] is not there, and my mother told me that I should stop those thoughts” (age 23, 3.4 weeks postpartum). As per the participant, her mother feared that the participant’s sadness could negatively impact the baby: “My mother would tell me that I should decrease my stress levels because the more I get stressed, this is going to affect the baby. . .” She acknowledged that her mother directly demonstrated emotional support, illustrating care through thoughtfulness about the infant and being present for the delivery. During the postpartum transition, her mother provided the support she needed, which helped alleviate her depression symptoms: “There is nothing that is stressing me now,” as she felt emotionally cared for.

In another example of increased support linked to positive changes in symptoms, a participant with elevated symptoms of depression and PTSD described improvements in her relationship with her partner and increased tangible support postpartum as drivers of these changes. During her pregnancy interview, she discussed having a negative relationship with her partner, in that he “Says something that hurts me,” or was “Rude when drunk.” Later, in her postpartum interview, she noted that, “My partner as well, he is not the person he used to be when I was pregnant,” and the interviewer confirmed, “He no longer hurts you like he used to?” to which she agreed. Due to unemployment, she was concerned about being able to provide for her child. Since becoming a mom, she explained that she felt better now that her baby’s needs were being met:

“It’s just that the baby has not needed anything yet that would make me feel some type of way. Things are still going well. . . the child’s dad and boyfriend assist me. At least he can buy nappies and porridge when he gets money.” (Age 26, 10.4 weeks postpartum)

After the birth of the baby, she described that the father of her child improved his behavior and supported her. Throughout her interview, she identified key stressors amplifying her mental health symptoms were relieved with financial assistance and caretaking from her partner, as well as an increased sense of security that her baby’s needs would be met. Removal of the worry of her child’s immediate wellbeing through her partner’s support, as well as his behavioral change improving their relationship, may have contributed to improvements in her mental health.

Additionally, a participant noticed improvements in her depression postpartum when she received emotional support at her church. Reflecting on pregnancy during her postpartum interview, she stated:

“During my pregnancy and before pregnancy I was facing so many challenges like family issues. . .it was stressing me to such an extent I was bleeding as if I was going to get a miscarriage. . .Sometimes I felt lonely, I didn’t have someone close that I can talk to. Yes, I spoke to my boyfriend but there were times I hoped to speak to a family member.” (Age 18, 12.6 weeks postpartum)

The lack of emotional support from her family left her feeling lonely and stressed, to the degree that she anticipated having a miscarriage. Connecting with her spirituality and receiving support from those in her church, however, offered her comfort and a space to heal postpartum. She explained, “When I’m in church or someone speaks of the word of God, I heal quickly.” When asked about referrals to mental healthcare or other opportunities to discuss her emotional challenges, she pointed to her church group as her source of support, stating, “There’s a program I’m attending at church which helps me a lot.” When her family was unable to meet her emotional needs, she leveraged her relationships with her spiritual community to access that support, which she found to be beneficial to her mental health.

Conceptualizing motherhood and infants as sources of joy and motivation

For some participants, their mental health symptoms improved postpartum because they started to view their infants as strong sources of joy and motivation. Identifying as a mother during the postpartum period led to increased pride, enabled a sense of community with other mothers, and facilitated companionship that brought happiness. For example, during pregnancy, one participant with elevated symptoms of depression and PTSD feared that she would have anger toward her child, but she noticed that her feelings shifted postpartum when she felt a connection with her baby. In her initial interview, she acknowledged fears that her sadness would negatively impact her baby, expressing, “People say that if you always cry it might affect the child and you might get miscarriage” (age 21, 20 weeks pregnant). In response to a question about how she thought her sadness would affect caring for her baby, she stated, “I won’t have a relationship with this child, I don’t think I will. . . This is because I will always look at the baby with anger,” which may have been influenced by the manifestations of her depression (e.g., “I get angry”). While pregnant, she avoided addressing symptoms of depression and engaging with reminders of stressful events, such as deaths of family members. After the birth of her child, when asked how she had been feeling, she responded, “My baby is my friend. . .(laughing)” (9.6 weeks postpartum). Though she feared feeling anger toward her baby, she had a positive change in affect, possibly related to the companionship she felt with her child postpartum, in contrast to the negative feelings that she anticipated.

Others discussed motherhood, in general, as a source of joy and motivation. For example, one participant with elevated symptoms of depression and PTSD articulated, “I get hope that I should leave everything in the past and just take care of my child” (age 30, 9 weeks postpartum); thus, her child gave her a reason to overcome her mental health challenges. Being a mother also provided a community, which brought her joy; she explained: “I get happy when I’m at the clinic because I can talk with other mothers, and we share our experiences.” Integrating motherhood into her identity postpartum introduced a new type of joy, which allowed for connection with other mothers who share these experiences. When asked about sadness potentially impeding childcare responsibilities during her postpartum interview, another participant similarly expressed, “I haven’t felt that way, because I am always excited for my baby’s appointments” (age 18, 12.6 weeks postpartum). For this participant, caring for her child and engaging in care was associated with excitement, which may have reduced the negative effects of sadness on her caregiving. In contrast to descriptions of anhedonia during pregnancy, the responsibilities of motherhood may have been linked to joy and other forms of positive affect.

A participant with elevated symptoms of depression and PTSD who was not a first-time mother anticipated that her symptoms would improve postpartum, acknowledging a natural shift that had previously occurred after childbirth. Though she feared loneliness during pregnancy, she expressed confidence that her fear would likely resolve upon childbirth, as she had previously experienced a shift toward positive emotions after delivery. During her pregnancy interview, she stated:

“I had the support, but I needed more than that, but as long you have a strong bond with your child and the child is healthy and happy that’s fine with me. I think this only happens now because I am pregnant and I like more attention now, and when the baby is born, I don’t mind about other people. I am happy for my baby, so I don’t think there’s going to be any problem. I will be happy for my baby and myself.” (Age 31, 36 weeks pregnant)

When discussing changes that occurred post-delivery, she described,

“The way of thinking, my thoughts are otherwise now, I stopped seeking pity from others or whatsoever. Now I know what do. I tell myself that this and that is my responsibility therefore I must do it.” (Age 31, 9.4 weeks postpartum)

During pregnancy, her self-perception was negative (“I perceived myself as a weak person perhaps”), and she resented the lack of attention she received from her family. After her child was born, she was less bothered by the lack of support and was motivated to fulfill responsibilities for her baby. Even when she was feeling sad during the postpartum period, her ability to care for her child, like taking her to clinic appointments, was “not affected” because the well-being of her child was her priority. Her described shift in thinking may reflect improved emotional resilience and self-perception, with her role as a mother considered to be an important motivator in this transition.

Discussion

The present study explored the mechanisms that are likely driving decreases in mental health symptoms among postpartum women in SA with elevated symptoms of depression and PTSD. Thematic analysis revealed three potential mechanisms leading to decreases in mental health symptoms from pregnancy to postpartum: increased feelings of empowerment, improvements in emotional and tangible support, and conceptualizing infants and motherhood as sources of joy and motivation. Strategically timed interventions during pregnancy that build skills linking to these mechanisms and/or leverage improvements post-delivery may also promote positive health behaviors like uptake and adherence to PrEP during pregnancy and breastfeeding, which will decrease risk for HIV acquisition and transmission. Ultimately, these findings may inform the development and optimization of future mental health interventions to be delivered in antenatal care settings in SA.

First, participants observed increased feelings of empowerment and self-efficacy, which contributed to reductions in postpartum mental health symptoms. This finding aligns with previous research conducted by Leight et al. ⁵⁶ in Burkina Faso, which examined the relationship among empowerment, maternal stress, and depression in a sample of mothers of children aged 2–4 years. Maternal self-efficacy significantly predicted decreased stress (B = −0.602, p < 0.05) and depression symptoms (B = −0.987, p < 0.01). Similarly, empowerment was a strong predictor of reduced stress (B = −2.62, p < 0.01) and depression symptoms (B = −3.49, p < 0.01). Furthermore, in a randomized controlled trial exploring the effectiveness of a perinatal program for depression among women in Nigeria, Gureje et al. ⁵⁷ found that the problem-solving-based program including parenting skills, behavioral activation, and skills to manage depression significantly reduced depression symptoms postpartum for those who exhibited higher symptom severity during pregnancy (adjusted odds ratio = 1.12, 95% CI [0.73, 1.72]). Outside of sub-Saharan Africa, research conducted in Australia also demonstrated that increases in maternal self-efficacy were significantly correlated with improvements in depressive symptoms (B = −0.52, p < 0.01) and stress (B = −0.44, p < 0.01) up to 24 weeks post-delivery without intervention during the postpartum transition. ⁵⁸ Feelings of hopelessness and lack of motivation in depression or PTSD may decrease PrEP uptake and adherence; thus, promoting empowerment and self-efficacy could improve these symptoms and lead to improvements in PrEP use. Our analysis builds upon their findings by suggesting that postpartum transition may be a critical time for symptom change, indicating that implementing such an intervention prior to delivery might be of greatest benefit.

Among participants in this sample, emotional and tangible support, primarily from other mothers and church groups, was also instrumental in facilitating decreases in mental health symptoms during the postpartum transition. Other research underscores the importance of accounting for psychosocial determinants like social support in the development and course of perinatal depression. For example, an analysis of an intervention targeting perinatal depression in SA modeled depression trajectories as antenatal only or antenatal and postnatal; family support (OR = 0.91, 95% CI [0.86, 0.96]) or support from a partner (OR = 0.94, 95% CI [0.88, 1.00]) significantly decreased the odds of having both antenatal and postnatal depression. ³⁵ In our sample, support from other mothers and religious groups contributed to decreases in mental health symptoms, suggesting multiple potential sources and types of support. Although emotional support may lead directly to reduced symptoms in the moment, instrumental or tangible support may indirectly decrease depression among pregnant women in SA via decreases in food insecurity (OR = 2.5, 95% CI [1.21, 5.15]) and/or other stressors related to resource limitations ⁵⁹ ; the impact of these limitations on infants was a key concern in our sample. By identifying the sources and types of social support that may be most helpful for this population, perinatal mental health interventions and associated programming can make more effective use of existing resources for maximum impact. Additionally, leveraging emotional and tangible support may promote consistent PrEP use, as support networks can encourage adherence and provide practical help in obtaining and taking PrEP. ⁶⁰ Utilizing these support systems could also mitigate the impact of symptoms like avoidance or cognitive difficulties (i.e., with concentration, planning, and decision making) that are associated with both depression and PTSD. ¹⁰

The final theme that emerged was conceptualizing infants and motherhood as sources of joy and motivation. Remission of symptoms could be part of a natural shift that occurs postpartum, as discussed by Garman et al. ³⁵ Our data suggest that this natural shift could be driven, in part, by the joy participants found in their infants and through identification as a mother, which was viewed very positively in our sample. Research based in the United States further identified that high antenatal optimism toward maternity resulted in lower postnatal depression (r = −0.376, p < 0.001) and promoted mother–infant bonding (r = −0.361, p < 0.001; lower scores indicating less challenges with mother-infant bonding). ⁶¹ Furthermore, a pilot intervention in Botswana aimed at reducing HIV stigma among pregnant people living with HIV, improving treatment adherence, and enhancing quality of life emphasized the importance of “what matters most,” identified as achieving “respected motherhood.” The intervention led to significant reductions in depressive symptoms (d = −1.96; 95% CI [−2.89, −1.02]) and HIV stigma (d = −1.20; 95% CI [−1.99, −0.39]), demonstrating the potential power of harnessing positive motherhood identity as a motivating factor for overcoming stigma-related barriers to HIV treatment engagement. ⁶² Our data offer a preliminary extension of these findings to the South African context. Biological shifts contributing to natural remission of symptoms should also be noted, such as regulation of hormones like progesterone, estradiol, and cortisol. ⁶³

Based on the themes identified in this analysis, specific therapeutic strategies that are rooted in cognitive behavioral and acceptance-based therapies may help reduce perinatal depression and PTSD if integrated into antenatal care. First, acceptance-based skills may facilitate feelings of empowerment. Accepting one’s situation, such as not having the desired support from family, may lead to increased self-efficacy,^64,65 as demonstrated by the participant who felt empowered when she learned she could cope on her own. It should, however, be noted that in some cases acceptance may not be an appropriate solution, such as abusive situations where changes need to be made for safety. Problem-solving and communication skills may facilitate increases in emotional and tangible support, such that individuals may more effectively identify and communicate their needs to others ⁶⁶ ; learning and executing these skills may empower individuals to care for themselves and their infants effectively. The utility of problem-solving skills to improve depression outcomes and HIV medication adherence has been previously demonstrated in SA ⁶⁷ ; these strategies could be extended to PPP. Providing such skills during pregnancy may ease the transition to postpartum. To strengthen positive identity around motherhood, encouraging or facilitating participation in support groups or peer counseling with other mothers may allow for bonding and discussion of shared experiences. Connection with other mothers may build relationships that encourage reassurance of worth, such that they acknowledge their abilities as mothers and use that identity as a source of strength and motivation. ⁶⁸ Leveraging other community resources in the context of pregnancy, like church groups, may also provide support for those who feel motivated and strengthened by faith-based approaches, which is highly relevant within South African culture.^69,70

The findings of this analysis have implications for future mental health interventions to be delivered in antenatal care settings and associated trainings for healthcare providers. Importantly, decreasing depression and PTSD symptoms during pregnancy and into the postpartum period may have positive downstream effects on HIV prevention behaviors, including PrEP uptake and adherence. Providing skills to facilitate empowerment may increase feelings of self-worth to motivate self-care, ⁷¹ which could potentially extend to motivation to engage in health behaviors like PrEP use. Emotional and tangible support may also promote consistent PrEP use, as supportive networks can encourage adherence and provide practical help.^60,72 Viewing motherhood positively can enhance overall well-being, which may encourage increased engagement in preventive health behaviors. ⁷³ Healthcare provider trainings could emphasize the importance of addressing mental health concerns in antenatal care, which may be otherwise overlooked. Future research may seek to further elucidate the potential association between depression and/or PTSD and engagement in HIV prevention behaviors. Future research may also focus on developing interventions to address mental health barriers to PrEP uptake and adherence in PPP to potentially implement in antenatal care settings.

Limitations

Limitations of this study should be noted. First, due to the cross-sectional nature of the survey data, we cannot assume causality in the relationship between identified mechanisms and mental health symptom changes. Our findings rely on participants’ interview self-reports and the authors’ interpretation of qualitative data. Additionally, we did not quantitatively reassess mental health symptoms postpartum, so we were unable to compare pre-post scores. Next, of the 13 participants interviewed after delivering their infants, only seven completed interviews during pregnancy and postpartum. Therefore, for six participants (46%), we relied on their retrospective comparisons of current symptoms with their symptoms in pregnancy which could have been impacted by recall bias. Nonetheless, participants who only completed postpartum interviews were asked to describe their symptoms over the course of the perinatal period, so we were able to capture those changes to some degree. Furthermore, survey data were only collected at one time point, so we were unable to analyze quantitative change in symptoms to support our qualitative analysis. In addition, our study and much of the current literature on PPP in SA and other sub-Saharan countries have been conducted primarily among cisgender women; thus, results may not accurately represent the experiences of PPP who do not identify as cisgender women. To understand mental health symptom changes among transgender and gender expansive PPP in SA, these populations should be specifically recruited in future research.

Conclusion

The present study provides key insights into the mechanisms that may be driving decreases in mental health symptoms during the postpartum transition, including increased feelings of empowerment, improved emotional and tangible support, and conceptualizing motherhood and infants as sources of joy and motivation. The postpartum transition is a critical time for intervention because major biopsychosocial shifts pose new challenges in nearly all areas of a parent’s life, and these shifts could lead to differences in symptom presentation. Identifying mechanisms of change will enable the selection of treatment targets and/or appropriate strategies to integrate into mental health programming within antenatal and postnatal care settings. Such programming may not only reduce mental health symptoms but also promote positive health behaviors, including those associated with HIV prevention, that benefit both the parent and infant.