Resilience to Depression Among Emerging Adults in South Africa: Insights From Digital Diaries

Abstract

Emerging adults facing chronic socioeconomic stress, especially depression, lack comprehensive research on resilience factors. This study analyzed digital diary entries (n = 338) from 57 individuals aged 18–24 in a South African township from July 2021 to April 2022. Participants highlighted relational, community, and cultural supports regardless of risk levels. Both high and low-risk groups faced challenges like financial instability, limited education, health threats, and lawlessness. However, institutional resource scarcity disproportionately affected higher-risk individuals, worsening issues like infrastructure deficits and violence exposure. Family and peer support emerged as crucial, especially for higher-risk participants. Individuals living in higher risk emphasized collective action and stranger support during infrastructure failures. These findings suggest that greater risk exposure may reinforce reliance on traditional, community-focused coping mechanisms, indicating the importance of studying differential resilience factors among young adults.

Introduction

Emerging adulthood (EA) is a developmental period between the ages of 18 and 29 that comprises both psychosocial growth and exposure to contextually specific stressors (Arnett, 2000). This is especially true for emerging adults living in contexts of chronic socioeconomic stress (Arnett & Mitra, 2020), such as townships in South Africa (Theron, 2023). For these young people, exposure to violence and crime, food insecurity, lack of access to sanitation and power, and lack of opportunity act as barriers to positive outcomes (Cieslik et al., 2022). The compound nature of these risks - that is the multiple layers of risk imposed by exposure to multiple stressors - faced by emerging adults living in these contexts intersects powerfully with both mental health risks and protective factors (Aguirre Velasco et al., 2020). However, to date few studies have considered which factors protect emerging adults against negative mental health outcomes (e.g., elevated symptoms of depression) relative to the levels of risk they report. Put differently, few studies have considered which factors might have differential (i.e., heightened or reduced) protective value for emerging adults reporting lower versus higher levels of risk (Ungar, 2017, 2018). By addressing this gap, this article aims to provide novel insights into the factors which might have differential protective impact to optimising emerging adults’ experiences of resilience.

Resilience is the capacity to respond adaptively (e.g., be mentally health) despite exposure to the significant stress of compound or acute risk (Masten, 2014). While earlier studies of resilience favoured trait- or person-focused explanations of adaptive capacity, more recent explanations discount such accounts (Bonanno & Westphal, 2024). Instead, syntheses of the many studies of youth resilience have pointed to multiple factors that are commonly reported to be resilience-enabling. Typically, these protective factors go beyond the individual to include supports from multiple systems, such as biological resources (e.g., physical health, a robust immune system, or genetic strength), psychological strengths (e.g., agency, problem-solving skills, or hopeful meaning-making), social supports (e.g., caring family, supportive peer group, or enabling neighbours), institutional and structural supports (e.g., accessible healthcare, effective transport systems, or opportunities for decent work), and environmental resources (e.g. safe and accessible green spaces or quality housing) (de Leeuw & Malcolm-Smith, 2023; Masten et al., 2021; Ungar & Theron, 2020).

These same factors, albeit with emphasis on personal and social ones, have been reported in syntheses of studies focusing on the resilience of emerging adults (Burt & Paysnick, 2012; Leung et al., 2022; Theron et al., 2023a). Emerging adulthood is recognised as a unique developmental phase characterized by numerous changes, including increased mobility, entry into career and work environment and changes in relationship status; because of these changes, depression may manifest differently during emerging adulthood compared to adolescence, and adulthood (Berry, 2004). The same is true of syntheses of studies reporting the resilience of youth (including emerging adults) in Africa (Pillay, 2023; Theron et al., 2023a; Van Breda & Theron, 2018). In these studies, social supports have multiple protective functions, including enabling affective, material/instrumental, and informational support (Betancourt & Khan, 2008). The emphasis on personal and social supports could be because of emerging adults’ preoccupation with relationships and personal identity development (Arnett, 2000).

Still, despite the ubiquitous nature of the protective factors that recur across resilience studies, their protective value can vary, particular in relation to mental health. In their qualitative study with 15 adults aged 18–25, with varying degrees of depression severity in the US, Kuwabara et al. (2007) found that emerging adulthood is marked by intricate interactions among depressive disorders, developmental processes, and social context. These elements may interact to intensify depressive symptoms or trigger episodes.

In this regard, Ungar’s (2017; 2018) Differential Impact Theory (DIT) has drawn attention to the variable protective nature of commonly occurring resilience-enabling factors, and informs this study in a limited way. DIT posits that protective factors will be more or less likely to support emerging adults in showing resilience relative to the level of stress exposure that they experience and the accessibility and/or meaningfulness of commonly reported supports. For instance, while Hopkins et al.’s (2014) study with 1021 Australian Aboriginal adolescents exposed to higher versus lower family adversity showed some commonality in the resources supporting these young persons’ positive outcomes (i.e., self-esteem and self-regulation), it also pointed to two differentially impactful resources for those reporting more family adversity. Only at higher levels of family adversity were pro-social friendships and residence in lower SES neighbourhoods (associated with readier access to supportive neighbours and other social capital) uniquely protective.

Similarly, a study with 233 emerging adults (average age 24.63) in South Africa living in socio-economically stressed communities showed that common protective factors (e.g., peer support; mental health services) had diminished protective value for those who reported higher risk for depression compared with those reporting lower risk (Theron et al., 2023b). The authors theorised that these factors might have been less impactful because they were inaccessible to the participants at higher risk and/or less meaningful. The South African government has prioritised the improvement of infrastructure in informal settlements nationally (Gibson & Rush, 2020), with the goal of removing all such settlements by 2014. Unfortunately, a combination of economic constraints and socio-political challenges meant that this goal was not achieved, leaving residents affected by the challenges of population density and associated health challenges, poor housing, inadequate access to sanitation and electricity (Nyashanu et al., 2020).

As other studies have shown, it is possible that the value of specific protective factors is relative to young people’s developmental stage. For example, protective caregivers and family supports are not as strongly associated with the resilience of young adults with experiences of maltreatment as they are with the resilience of children with maltreatment experiences (Yoon et al., 2021).

Despite acknowledgement of such variability, there is very little understanding of which resilience-enabling factors impact mental health outcomes at lower and higher levels of risk exposure (Ungar, 2018). Likewise, there is limited understanding of which resilience-enabling factors matter more, or less, for the mental health of African emerging adults in stressed environments or how consistent these factors are over time (Cockcroft et al., 2024; Theron et al., 2023b; Theron et al., 2023c). Given this variability in the protective value of commonly occurring supports, there is a need to better understand which protective factors, if any, are more likely to account for positive psychological outcomes and resistance to mental disorder (e.g., depression) when these emerging adults report higher levels of risk exposure.

The Study

The findings are based on data collected as part of a series of studies exploring the multisystemic factors and processes indicating resilience among 14–24 year-olds living in challenging environments (Ungar & Theron, 2020). For all studies relating to this programme of work, biographical data are collected at all time points – eligibility for the study reported here required participants to be between the ages of 18 and 26. Specifically, we sought to answer the following research question:

Which factors or processes do young African people who self-report minimal to mild levels of depression and higher risk exposure (vs. minimal to mild levels of depression and lower risk exposure) describe as resilience-enabling?

Data-gathering for the over-arching project took place in the township of eMbalenhle, a densely-populated township in Mpumalanga Province (South Africa) between July 2021 and April, 2022. The diary data specifically was gathered between July 2021 and November 2021. Each participant contributed diary entries for 10 weeks within that time period. Mpumalanga is a challenging environment; 63% of 15–34 year olds live below the official poverty line (StatsSA, 2017), and approximately 25% of all adults in eMbalenhle are unemployed. Previous work done within the community demonstrates that eMbalenhle typifies other chronically low-resource South African environment.

Methodology

We used a phenomenological design in order to document participants’ insights into a specific phenomenon (Tuohy et al., 2013) – in this case, the ways in which participants achieved positive psychological outcomes in the face of depression. In practice this meant (i) using a purposive sampling approach at the outset of the project and thematic analysis for the analysis stage, and (ii) maintaining a focus on the participants’ lived experiences (Shorey & Ng, 2022). Our previous research with other samples of youth from the same community (e.g., Theron et al., 2022) showed that digitally-mediated methods can lead to detailed participant accounts of the social-ecological determinants of resilience (Theron, 2023).

Historically, diary methods have been used to gain novel insights into participants’ contemporaneous experiences, opinions and emotional states (Alaszewski, 2006; Bolger et al., 2003) in the environments in which they usually operate (Elsweiler & Ruthven, 2007). They have been noted for three strengths (Jarrahi et al., 2021). First, ‘in situity’ allows researchers to explore participants’ daily lives, particularly through the lens of what they do (Flaherty, 2016). Second, ‘context specificity’ allows for a focus on the social and structural contexts and dynamics in which people live their day-to-day lives (Janssens et al., 2018). Finally, diary studies have been found to facilitate ‘longitudinality’, taking place over periods of time (Church et al., 2014).

The increasing embedding of ‘digital’ in the lives of young people globally (Benvenuti et al., 2023) and the recent proliferation of the field of ‘User Experience’ research, have prompted an increase in interest in the application of diary methods in digitally-mediated landscapes. Janssens et al. (2018), for example, reflect on the methodological implications and stages that should be considered in designing digital diary studies. Digital diaries have been found to offer new ways to understand the everyday lives of young people (Horton & Kraftl, 2006; Volpe, 2019), enabling participants to contribute in whichever media suits them best, immediately/at a time of their choosing, and with the possibility to amend or edit their contributions, placing their comfort at the heart of the process (Jarrahi et al., 2021).

We operationalised the method by asking participants to contribute their digital diary entries at least once a week for 10 weeks, between the months of July and November 2021. Research assistants who shared participants’ ethnicities, language, and age range interacted with them (digitally) to generate these data. It should be noted that there were still significant Covid-19 restrictions in place during these months (South African Government, n.d.) – for some of these periods our participants may have been attempting to learn remotely, returning to school and examinations after weeks of missed learning, navigating decisions to vaccinate or to adopt other health behaviour management strategies, or had limited access to some services (e.g., transportation). We trialled this digital diary process earlier during the pandemic focusing specifically on experiences of Covid-19 and its impacts, and found that weekly - rather than more frequent - requests optimised contribution levels and reduced attrition over the data collection period (Theron et al., 2022).

Participants were asked to contribute for 10 weeks in the data gathering period. They were contacted by research assistants weekly, usually via WhatsApp™. and were prompted with the same three idiomatic questions:

- What has made your life hard this past week?

- How did you manage that?

- Who or what supported you to manage that?

Participants were invited to respond digitally, and we provided them with funds to pay for their online data usage. We did not specify whether responses should be text, auditory, or via image; participants largely favoured text-based responses, mainly in English, but with some use of either isiZulu or isiXhosa (the research assistants spoke both, and Author 1 speaks isiZulu, which facilitated adequate checks of translations).

The Sample

As discussed above, this study was delivered as a follow up to the Resilient Youth in Stressed Environments study (RYSE; Ungar, 2011), in which participants self-reported symptoms of depression using the Beck Depression Inventory (BDI-II, Beck et al., 1996). Scored responses to the 21 items of the BDI-II allowed identification of four categories of depression symptoms: minimal, mild, moderate and severe. As in many resilience studies (Bonanno, 2004, 2021; Masten, 2014), we interpreted minimal symptoms of depression as evidence of resilience taking into account these youths’ exposure to community level risk factors. In addition to the depression measure, participants were also asked to self-report their risk levels during the RYSE survey using (i) a 10-item adapted version of the Family Adversity Scale (Labella et al., 2019) and (ii) the 10-item Perception of Neighbourhood Scale (Ruchkin et al., 2004). Total scores for these two measures were summed. Participants scoring at or below the mean were categorised as ‘lower risk’, with those scoring above the mean were identified as ‘higher risk’.

In the study reported here, participation was invited from emerging adult participants who had participated in the RYSE study, who had consented to follow-up in future studies, and reported minimal levels of depression (n = 60). Fifty-seven chose to participate, of whom 31 self-identified as female, and all self-identified as Black/African. Twenty-eight of the participants were located in the ‘Lower risk’ category, and 29 in ‘Higher risk’ (LR n = 28; HR n = 29). Characteristics of the sample are represented in Table 1.

Table 1.

Characteristics of the Sample, Segmented Following their Completion of the Family Adversity and Perception of Neighbourhood Scales, and the BDI-II in the RYSE Study.

High risk	Low to mild depression
Characterised by:	Characterised by low-mild (BDI-II):
• Significant challenges during upbringing, such as multiple losses, chronic instability/conflict, challenges to caregivers, disrupted caregiving (10 item Family Adversity Scale)	• Self-dislike
• Struggling neighbourhood, such as low safety and security, weak social cohesion, poorly-maintained environment, prevalence of social problems (10-item Perception of Neighbourhood Scale)	• Agitation (e.g. angry outbursts, irritability)
n = 29	• Loss of interest
Low risk	• Changes in sleep or appetite
Characterised by:	• Loss of energy
• Relatively stable and supportive upbringing, such as consistent presence and support from primary caregivers, minimal conflict/disruption, healthy caregivers, secure attachments (10 item Family Adversity Scale)	• Unexplained physical problems
• Thriving, safe, welcoming neighbourhood, such as high safety and security, strong social cohesion, well-maintained environment, minimal social problems (10-item Perception of Neighbourhood Scale)	• Thoughts of self harm
n = 28	n = 57

Ages ranged from 18–25 (Levine et al., 2024). In total 338 diary entries were received.

Analysis

We used a thematic analysis approach (Braun & Clarke, 2022) to identify data addressing the research question. Working iteratively, the first author and a UK-based research assistant labelled 20% of the dataset independently. On review there was a high degree of congruence between the independent analyses and the themes emerging from the RYSE study which also included a phase of qualitative data collection. Based on these results, a slightly revised version of that codebook was used for the remaining analysis carried out by Authors 1 and 2.

Ethics and positionality

The Ethics Committees of the Universities of Pretoria and Leicester provided ethical clearance [44368-dt228]. Participants provided written consent for the RYSE study, and informed verbal consent via telephone call for participation in this diary study (as we were operating in COVID-19 lockdown conditions and restrictions on in-person data gathering). Three trained research assistants (Educational Psychology Masters level students) explained the study to the participants emphasising voluntary participation, providing translation where needed, and recording the verbal consents for upload to a password-protected cloud-based folder. Participants received an e-voucher (value approximately USD$20) to cover the costs of their data for participation.

The importance of researcher transparency and continual reflection is an inherent part of our process (Levitt et al., 2018). Having worked with African young people for over 25 years we aim to maintain sensitivity to the complexity of the systems that contribute (or act as a barrier to) their resilience to significant, chronic, and structural disadvantage. As privileged adults (two living in South Africa, two living in Global North settings, one of whom was born and grew up in South Africa), we acknowledge our life experiences set us apart from our participants, which meant the research assistants’ (all of whom were from South Africa) role was fundamental in linking age and experience in facilitating data generation. The process was mentored through regular check-in and debriefing.

Findings

Following Cockcroft et al. (2024), our findings are reported in terms of (a) those stressors and protective factors that were common across both the higher risk/minimal-mild depression, and lower risk/minimal-mild depression groups, and (b) those stressors and protective factors that were distinctive to each of the two groups. Notably, the day-to-day immediacy of the digital diary method identified both stressors (e.g., transport and bereavement) and protective factors (e.g., unprompted focus on cultural values and practices) consistently in our data that are not represented widely in the literature. (Note: grammatical and typographical errors have been corrected to avoid negative stereotypes relating to African emerging adults’ literacy levels). We report our findings first by focusing on the stressors, examining each stressor’s ubiquitous and divergent features in turn, and then on the protective factors (Figures 1 and 2).

Figure 1.

Summary of divergent and ubiquitous stressors and protective factors reported in digital diaries for both the Lower Risk and Higher Risk groups.

Figure 2.

Stressors and protective factors reported by participants living in lower and higher risk contexts.

Ubiquitous Stressors

Some stressors were present for participants living in high- and participants living in low-risk circumstances.

Financial Conditions

The majority of participants reported experiences of financial hardship closely connected with chronic structural issues and inequalities that characterise the community in which they live. By ‘chronic structural issues’ we mean lack of credit and sustained investment in infrastructure, colonial and apartheid legacies, problematic government and local government policies, and corruption (Bjornlund et al., 2022). This was constant for those reporting higher and lower experiences of risk.

What made my life very bad this week was the NUMSA strike because our bread winners have got 2 weeks not going to work, and it’s no work no pay. Well I couldn’t cope because we’re approaching month end and it’s not all families that are able to save for the future to survive due to having many responsibilities. (Week 6, E36, Lower Risk (LR))

I did manage by listening to her [mother] that everything will be ok as soon as she gets back to work…at least we manage to eat and she made a plan for me not to sleep on an empty stomach. (Week 3, E601, Higher Risk (HR))

For participant E342 the challenges of balancing work and school continued three weeks later, and this challenge was echoed in other diary entries. Financial precarity meant that a balance be found.

I recently got a job in my neighbourhood and mixing the two things (school and work) has changed the things I do and it's has been hard for me. I had many CVs made and distributed them in many different places to apply for a job and I eventually I got it now I do the two things. (Week 8, E342, HR)

Participants reported experiencing financial hardships relating to employment for a number of reasons, including nepotism and corruption. These are significant issues in the South African context; in the January 2023 release of Transparency International’s ‘Corruption Perceptions Index’ South Africa had dropped to its lowest level since contributing data, falling into the ‘flawed democracies’ category (Corruption Watch South Africa, 2023).

It’s the unemployment. And corruption in the country. By just stay on hope that one day things will change. Its family and friends by encouraging me to not give up //Research assistant (RA): …When you say corruption in this country? What are you referring to? Did something happen? //E234: When I'm talking about unemployment I'm referring to nepotism; if you look for job you have to know someone in organisation to get that position it's something happened often lately (Week 4, E234, HR)

Barriers to Engaging in Education

Many of our participants contributed diary entries referring to a desire to enter or re-enter education, or relating to the stress of performing in educational settings. Our data suggests that it is the compound nature of the stressors (i.e., the combination of risks; Rashid & Gregory, 2014) experienced by our participants that make them extraordinary and increase their overall risk.

What made my week very bad was load shedding…I couldn’t cope because it’s exam time, again people are at work and then they come back to darkness. Nobody help since everyone was furious, fridges were melting, food was getting spoilt (Week 9, E36, LR)

I went for an interview on Tuesday, not knowing that I have an exam, so that was kinda hard for me...I wrote an email to my school and I was given a second opportunity. (Week 1, E114, LR)

Some of the stressors relating to re-entering education or securing training were closely linked to financial hardship - mostly for Higher Risk/Minimal-mild Depression group members:

Was not being able to go to school and register because I [did] not have enough money (Week 1, E586, HR)

Starting a training for any experience but unfortunately money is problem to me (Week 4, E156, HR)

Health of Self and/or Others

Health was an issue that cut across both Higher and Lower Risk groups. Contributions from participants on health-related matters and illness fell into three categories. First, many participants from both groups referred to what might be considered minor ailments, for example E65 in Week 1:

I didn’t want to overthink this headache thinking that I’ll die and stuff I apply a cold cloth to my headache bought myself a pack of Panado [paracetamol] and I took them not it’s not bad as it was that’s how I managed my headache thing. (Week 1, E65, HR)

While the ailments were minor, the frequency with which these were reported in the digital diaries (intended to capture quotidian experiences) suggests breadth rather than depth of stressor.

Second, a smaller number of participants referred to fear of serious illness and the implications of that illness as a stressor.

The hardest thing i faced last week was my sister being sick nothing was work she went to the doctor ad to the clinic but the was no difference it was hard because she just had a baby it was so scary to think that she was going to live that infant behind (Week 2, E470, HR)

The accident thing has got me with a cardinal in my head. Almost every day here in Embalenhle there are accidents all along the road. Mostly in the morning, which makes it even harder to go to work ‘cause you’ll never know what will happen. (Week 6, E104, LR)

Some participants referred to actual serious illness or injury.

Is being told that my auntie has been on a car accident. Right now she's in a coma with multiple injuries and the doctors didn't want us to see her, I'm really worried and kind of have struggled to keep up with everything (Week 2, E342, HR)

What made my life hard? The fact that I had to spend time in hospital because of some sickness I had. (Week 1, E35, LR)

What made my life hard this past week? It was when my child was admitted to Witbank hospital she was burn by hot water (Week 3, E412, HR)

Finally, just over half of the participants reported experiencing bereavement during the data gathering period. As these were diary entries, participants’ reflections often focused on the characteristics of the passed individual, for example:

…I lost my grandfather from my mother’s side. I managed that by crying and accepting that he’s gone. Even though it was really hard, because he was very supportive all the time. (Week 6, E461, LR)

Pervasive Lawlessness

Housebreaking (and associated animal theft) is currently the most common crime experienced by households in South Africa (StatsSA, 2023a), with ‘home robbery’ coming in as the second most commonly experienced crime (StatsSA, 2023a). In line with the wider policy landscape in South Africa (StatsSA, 2023a), a small number of our participants reported experiencing home robbery as a violent crime because they were present when the robbery occurred (in contrast to housebreaking which occurs when residents are elsewhere). Typically, these participants were from the group reporting higher risk exposure.

I was with my girlfriend 29 August 2021 she was visiting me at my home where I live…Someone tried to break in the house but unfortunately we heard footsteps and wake up and that person was hearing that we were not sleeping he broke all the bedroom windows and run away. So since then I was afraid to sleep because I thought maybe that person will come back to finish up what he started and each time I will wake up even if there is nothing wrong during night and my girlfriend was so scared that night because my community is not safe at all. (Week 4, E344, HR)

With regard to theft of personal property, the timing of this study meant that were more likely to be reports of this type of crime, as these have peaked in the South African winter months every year since 2018 (StatsSA, 2023b). This was indeed the case, in these examples for E303 (male) and E470 (female):

Friday night my cousin followed me when I was rushing on my uncle’s house; they robbed him his cellphone … I had to show my uncle where [the thieves] stay and I was nervous. … uncle’s friends drove fast to the boi and pick him up and ask him questions after they released him to go and fetch their phone where they hide it and he came back with it. Uncle he is there one who helped us to find the phone and the boy was afraid of him. (Week 5, E303, HR)

Last week went out with my girls, me and my baby sister almost got mugged; we were really drunk so some guys took advantage of that; they put knifes on our necks and took my baby sis phone; it was a traumatic experience. Some guys saw what was happening and they came to our rescue; they chased they guys and they managed to get the phone back then our big sister took us home but we are fine now we only have lil’ scratches from the knifes, but we оkау. (Week 5, E470, HR)

As is evident in the preceding excerpts, violent crime dominated the experiences of the Higher Risk/Minimal-mild Depression group, although it also had a presence (often as witness) in the Lower Risk group. The permeation of this issue was not surprising; Mpumalanga was one of only two provinces that did not see a reduction in the murder rate from 2021 – 2023 (StatsSA, 2023b). Diary entries reflected encountering this experience and associated gang violence.

What make my life hard this past week? Seeing a nyaope [drug-addicted] boy beaten to death because he stole an old refrigerator that was no longer in use…The pain was unbearable it was hard to [forget] what I saw that night. (Week 6, E485, LR)

One of my friends called me and told me that one of my friends that we completed with and used to play soccer with was stabbed on Saturday and passed away….Yes he was not that close to me but the fact that I know how to lose someone close to you just make me feel sad cause I remember how I felt the day I had that my closest friend who was like a brother to me passed away, locked myself in a room and cried and prayed. Then tried to forget about it all. (Week 7, E68, LR)

Finally, most of our participants reported stressors relating to substance misuse and related crime, either their own misuse or being impacted by others’ misuse. As South Africa’s drug challenge is close to twice the global norm (United Nations Office on Drugs and Crime, 2023) this was not a surprising finding; however it was indicative of the complex interplay between family relationships, domestic violence, and substance misuse (e.g. Muchiri & Dos Santos, 2018).

What made my life hard this past week was seeing my father drunk. My father promised that he’s no longer drinking alcohol. But I was shocked when I saw him drunk. He knows that he can’t control himself when he’s drunk. He knows that we don’t like seeing him drunk. There is no peace in the house when he is drunk and I can see how much it [hurts] my mom. (Week 6, E368, LR)

Some participants referred to their own substance use. Taking into account their challenging day-to-day lives, and the immediacy and informality enabled by the digital diary process, it was not surprising that our participants shared examples of strategies that had potentially negative longer-term outcomes for themselves or the people around them.

My life wasn’t hard this past week in fact I was wondering if I may be unlawful cause I was drunk the whole week and I was not even home. I was with friends and we managed to rent a house for a week, not just to get drunk, but to work on our projects ….It’s for business, we were advertising a new song called “Bio Plus” (Week 9, E342, HR)

Stressor Unique to the Higher-Risk Group

One stressor, however, was only present in the high-risk group. A significant proportion of emerging adults in Africa find themselves in a situation where they are not engaged in employment, education, or training (NEET) (Lars, 2022). In South Africa, one in three emerging adults experience NEET-hood (StatsSA, 2023b). Being NEET is stressful and associated with suboptimal development, psychological distress, and poverty (De Lannoy & Mudiriza, 2019; De Lannoy & Basic Package of Support Research Consortium, 2018; Garman et al., 2022).

NEET-Related Financial Stress and Poor Access to Infrastructure

Experiencing NEET-hood was often the case for our Higher Risk/Minimal-mild Depression group and meant that financial stressors manifested in slightly different ways than with their Lower Risk/Minimal-mild Depression peers (Theron et al., 2024).

Do not even ask me about the father of the child, I left him and he stopped supporting the child and I do not care. So for me to be strong again, to not sleep hungry, it is my parent. She is the one who is working and she is strained, she is supporting old men and an old woman my age, she is financially strained. You even ask God that “please give me a job”, even if I earn 3000 [Rand], just so I can at least do something. (Week 4, E392, HR)

Unemployment is very high in my community …. Everyone needs a proper job in order to survive or manage his/her needs and wants, more especially when you have grown up and no longer attending school…. So when it comes to toiletries when you don’t have them I usually get scared to ask my father to at least assist me with them because now I am a grown man… to me it's hard to be unemployed. (Week 2, E344, HR)

This theme of financial hardship was sustained throughout the data-gathering period. For example E344 continued to focus on this theme during their Week 8 contribution.

Unemployment is hard for a person like me because you suddenly feel useless sometimes and if maybe there is a person who always insult you …that thing make you feel sad and stressed. It’s very stressing to wake up in the morning, not know how will the day be and not having plans. Not doing anything in life, it’s boring, even if you go to town to look for a job but still not finding it. Thinking too much …. Being depressed.. So, to me right now I am very desperately in need of a good job so that I can be able to support myself and my family… I go to town ones or twice a week to hunt for a job but still nothing and I always hear people saying they will try and give a job but they giving me false promises (Week 8, E344, HR)

In addition to financial stressors the Higher Risk/Minimal-mild Depression group, particularly those in precarious/informal employment, were also more likely to report stressors relating to structural violence specifically in the form of poor access to infrastructure – including transport, access to electricity, water, and housing – than their Lower Risk peers.

Specifically, participants reporting higher risk, referred to chronic challenges of inadequate infrastructure, primarily transport and public services.

Taxi crisis ever since the beginning of shutdown. We have less taxis but too many people who are hired in this shutdown. So in order to be early at work you must wake up maybe around 03:00 am due to shortage of taxi…One of the taxi drivers offer us a transport...we now at least wake up early than before, others went straight to Vukanini [taxi association] to complain and they were told that they will release more taxis for shutdown people. (Week 2, E80, HR)

Ubiquitous Protective Factors

Some protective factors were present in the contributions of participants from the Higher- and Lower-Risk groups. Social resources (family, peers) were pronounced ubiquitous resources.

Family Resources

Our participants reported the importance of close family interactions as protective factors. Typically, family offered social support. For some, that support was material (e.g., financial) or instrumental (e.g., related to securing employment).

I asked my bigger sister to help me to look for a job; absolutely she did get me one [emoji]…My family [emoji] (Week 1, E438, HR)

I couldn’t do much since most of the things I do are done digitally. I had to visit Internet cafes that have generators. My parents really helped me with capital and transportation. (Week 9, E38, LR)

1. I don’t know what to use for my face since it’s summer because i have a skin problem, and i don’t have money to go to the doctor since am no longer on a medical aid, so am stressed that if I don’t get those creams and medications my skin will get worse now that the sun is too hot. 2. My dad said that if I go to school and get a letter that is saying am still studying they can put me back on the medical aid because he can’t afford to pay the doctor on his own. 3. My dad (Week 9, E453, HR)

For others the social support received was more affective (Betancourt & Khan, 2008), with emphasis on comfort and encouragement.

My parents supported me a lot especially my mom she always tell me that life is very hard [emoji] but having a loving and supportive family makes things better…My mom always tell me that no matter how hard life may be she will make sure that I'm always happy [emoji] (Week 1, E412, HR)

…who supported me was my mom for always having hope and also believing in me telling me that I will sure make it to the top. (Week 7, E275, LR)

Peer Resources

Even in the immediacy of a diary entry our participants reported that peer and wider community resources also act as protective factors. Peer resources, sometimes referred to friendships or romantic relationships, many of which manifested as both emotional and practical support.

It’s friends they help me settle some of my debt (Week 6, E234, HR)

We supported each other as friends that will never compete with anyone, coz we are here at college to study and help each other, not to compete with anyone (Week 7, E412, HR)

I had my friends helped me out and we had it all done in no time…People in my neighbourhood saw what we were doing and they support the work that we’ve been doing and they say we've done a great job it's beautiful (Week 6, E342, HR)

I had a plan of starting a carwash business and it isn’t easy cause it’s energy draining and iskolo ngapha [school is over there] I got someone who once had this kind of business before [for] advice. I was supported by my friends and family and gave me the boost I needed to overcome all the challenge … They helped me on my capital and boosted (Week 6, E238, HR)

At other times, peer support referred to working or collegiate relationships:

As a person who is around people most of the time in our work environment you get motivated and encouragement to do your best no matter what conditions and situations. (Week 1, E26, HR)

I took my eyes off my post just for a minute and a tragedy occurred…a colleague told me such tragedies have occurred before and yes this may turn out bad but I should keep my head up. (Week 9, E356, LR)

Me and my team mates we use to go to his house every day to check whether he is getting better or worse. 3 who/what supported you to manage that? My team mates and our parents by donating some little bit of money so that we can take him to hospital. (Week 4, E91, HR)

A final cluster of responses were more focused on peer resources for mutual support.

The surprise baby shower went well everything was fine my friend was very happy and the support from her baby daddy’s side was amazing. We were all happy about it…we managed to make our friend happy and welcoming her to the mother land. (Week 8, E469, LR)

Cultural Values, Faith and Spiritual Practices

For many of our participants, religious faith, spirituality and cultural practices (e.g., engaging with elders) were related protective factors. Following Mayer and Viviers (2014) we understand spirituality as a strong sense of inner connectedness with the world and culture as the interactions between actors, stories and ideas. While spirituality (e.g., Khumalo et al., 2014), religious faith (e.g., Kim et al., 2018), and cultural practice (e.g., Tubadji, 2021) have been identified as a protective factors in mental health individually, the relationships between these have been underexplored; the interconnected nature of our participants’ contributions suggests further exploration is needed.

The best way I think will help me is always pray and stay in my beliefs or attend a prophet to tell what to do on this situation. (Week 1, E156, HR)

The past few weeks have been hard …because a friend of mine is sick...my head was painful and broke at the same time when I hear that my friend is sick... because nowadays we see lot of funerals every week so I was scared for her ...my friends and family supported me through the situation I was facing but by the grace and mercy of GOD she is better … we thank God (Week 3, E537, HR)

I tried to talk to the elders I know around in my neighbourhood to come together and try to raise some sort of an awareness campaign but only around my neighbourhood so we can come together and be one as we are all human and want only one purpose which is to succeed in life and change our lives for the better (Week 4, E342, HR)

For some, cultural practices were an effective approach to tackling ill-health (Nicholas, 2023), as shown in these Week 8 and 9 diary entries from E104 (LR):

I’ve been having headaches and temperature change. At first, I thought I was Covid, fortunately my mother quickly saw what’s happening…My mother is a very traditional person so she suggested that we go to a traditional healer who confirmed everything going on and gave me traditional medication to use. (Week 8, E104, LR)

And then in Week 9:

I’ve worked out the whole ancestors situation.

Similarly, this quote from E57 (Week 5) in which the cultural belief linking hair to headaches, and headaches to mental wellbeing, is exemplified:

I was not okay bengi phethwe ikhanda [I had painful head] for a week … [I] went to the doctor he said I have a low blood / depression. My mom and i had to cut my hair //RA: Why did you have to cut your hair? //E57: Because of ikhanda and after I cut my hair I was okay (Week 5, E57, HR)

Protective factors Unique to the Higher-Risk Group

Participants living in higher-risk contexts described specific protective factors upon which they drew.

Collective Responses to Service Issues and Help from Strangers

In the Higher Risk/Minimal-mild Depression group, activation of resources were motivated by social or service issues and responded to as a collective, important in a context where intermittent and unreliable services and limited institutional supports (e.g., substance use programs) are a daily challenge. This cluster builds on previous literature focusing on the potential benefits of social support as a form of community resilience (Pfefferbaum et al., 2013; Sippel et al., 2015) which has been shown to offer some protection for those who have experienced chronic or other forms of trauma (e.g., Abramson et al., 2015).

It’s the loadshedding and theft of electricity cables in my area we couldn't cook or watch TV in these past days. We use gas to cook and get hot water…Supports each other as community…By collecting money to replace the cables that were stolen and calling the municipality to attend the issue (Week 10, E234, HR)

Water crisis are suffering us as a community. Sometimes 4 days end without even a drop of water on taps how can we cook without water worse toilet are become mess. How did we manage that? We used the river water for our toilets and bathing because we wanted to save our last clean water for cooking…As a community we went to municipality to report and we were told that there’s a broken pipe near Secunda that take water to Embalenhle but it will be fixed soon (Week 4) The parents in our community noticed that teenagers are impregnated each other in such a way like it’s a fashion so they decided to minimize the situation. (Week 8, E80, HR)

We called the street [committee] and he said if he continues abusing his wife, he will try to contact people that are working with these types of situations like social workers and help both the parents and children because all of them they really need help. Maybe if the father goes to the therapist for this drinking problem because I also have that problem here at home with my brother he is also drinking too much thanks again to my community at many people will find help and fight this drinking problem. (Week 4, E344, HR)

Alternatively, there were examples in which a protective or supportive response from a stranger was motivated by concern for an individual:

until an old man pass by and called me out and it ok you can walk with me and he was comforting me (Week 1, E470, HR)

What made my life hard this week? Was not being able to go to school and register because I did not have enough money. How did I manage that? I had to hike from home to the campus. Who supported me to manage that? Was a stranger because he helped me to get to my destination. (Week 1, E586, HR)

Assistance in Emergency

Those in the Higher Risk/Minimal-mild Depression group were more likely to report a different form of family resource – that of the presence (virtual, psychological, or physical) of a caring adult in circumstances of danger.

I would say it’s my dad cos I call him during midnight and he came and call the police and everyday he was calling me to check up on me. He really made me feel proud to have a father that would come rescue his son. (Week 1, E344, HR)

These past few days has been a bit difficult as I have no personal transport [emoji] [emoji] public transport does help but having to get to places at a certain time and going to places where public transport doesn't go or enter, is a bit difficult as I usually pass through dangerous places and I am unknowingly open of criminals plus it takes a bit longer to walk home from where I was [emoji]...Well how managed? I walked and I made sure I was alert and put away everything and anything that could attract them to me besides them cat calling me. I made sure I walked as quick as possible and NEVER took quiet passages. Who or what supported me to manage that? My mother, although she was at work I sent her my location so she knows where I am, so that made me feel a bit relaxes as I knew if I do go missing, someone will know at least where to start looking for me (Week 1, E1344, HR)

Where these family resources were not available, their absence was keenly felt in the Higher Risk/Minimal-mild Depression.

It’s that I mic [miss] my parents [emoji] not because the people am living with treat me badly no is because I never found that love from both parents because GOD took them away from me while I was still so you so ...it has been difficult to accept that they are no longer here with me...the pain doesn't go away .. (Week 2, E357, HR)

Discussion

Drawing on the digital diary data generated by 57 young Africans over a 10-week period, we address the question: Which factors or processes do young African people, who self-report minimal to mild levels of depression and higher risk exposure versus minimal to mild levels of depression and lower risk exposure, describe as resilience-enabling?

Our participants in both lower and higher risk groups reported a range of common stressors and protective factors in managing their resilience to depression. A smaller number of stressors and protective factors were present in only one of the two groups. As explained next, these reflected both the context and developmental phase of our participants.

Overall, participants mostly reported the same resources (i.e., relational, community, and cultural supports) regardless of the level of risk exposure they reported. This commonality could potentially be explained by the nature of the risks that both groups experience, but also the developmental stage of emerging adulthood. Work identities, peer, family and community resources speak to what we understand about the identity exploration and definition still occurring at this stage as emerging adults progress from adolescence towards adulthood (e.g., Ritchie et al., 2013). At the same time, these have potential to amplify the sense of ‘in-betweenness’ that many emerging adults experience at this stage (Arnett et al., 2014).

In both groups, participants were challenged by financial precarity, accessing and taking up education opportunities, health threats, and lawlessness. These stressors embody barriers to important developmental features of emerging adulthood, for example in facilitating a sense of possibility for the future. A resilience study with emerging adults in a stressed community in Eswatini (i.e., a South Africa’s neighbour) reported similar hopelessness, along with relational and community supports that mitigated that hopelessness (Gama & Theron, 2023).

The above stressors seemed most pronounced for those reporting Higher Risk/Minimal-mild Depression in our sample. For these emerging adults the challenges presented by lack of infrastructure, financial and educational insecurity, and exposure to chronic violence were more intense and more frequently encountered than their Lower Risk peers. For example, as reported in other research with a similar population of youth (Baldry et al., 2019; Mudiriza & De Lannoy, 2023), being NEET seems to perpetuate the experience of financial precarity and adds the burden of social censure.

Similarly, for those reporting more risk exposure, lawlessness was not a vicarious or petty experience. Instead, it typically included personal experiences with violence that occurred in their own home. For Higher Risk/Minimal-mild Depression participants these experiences featured regularly in their diary entries, suggesting they were not one-off instances of disadvantage but chronic, structural (specifically relating to colonial and apartheid legacies) and quotidian. Across the higher and lower risk-exposed groups, participants emphasised relational resources (particularly support from family and peers). This emphasis fits the long-held understanding that enabling, supportive relationships are critical to young people’s resilience (Masten, 2014), especially in African contexts. Systematic reviews of youth resilience in the context of sub-Saharan Africa flag relational resources as the most prominent source of resilience (Theron et al., 2023c; Van Breda & Theron, 2018), affording young people material, affective and instrumental support (Betancourt & Khan, 2008). The same was reported in a review of the resources that supported emerging adult resilience to COVID-related stressors (Theron, 2023a). The emphasis on relational supports aligns with the social orientation of emerging adulthood (Arnett, 2000).

Interestingly, participants who self-reported higher risk exposure were more likely than those reporting lower risk exposure to ascribe power (e.g., to rescue or defend) to supportive family members. They placed particular emphasis on the empowering presence of a trusted, caring relative during a crisis. While this underlines their appreciation of family supports, it prompts questions about the absence of formal social supports that are traditionally associated with rescue-like or defensive operations (e.g., law enforcement). Similarly institutional supports (e.g., public health services; legal services; welfare supports) were completely absent from participants’ accounts. In the absence of such supports, emerging adults’ reliance on atypical supports (i.e., factors that ameliorate stress in the short term but have the potential to do harm in the long term; Ungar, 2018), such as problematic substance use, can be understood as a maladaptive coping strategy that compensates for a lack of other choices.

Additionally, participants reporting higher risk exposure emphasized the importance of collective action in response to infrastructure failures and the value of protective or supportive strangers. Neither is surprising in the context of Africa, where collectivism and harmonious interdependence (also with strangers) are traditionally valued (Mugumbate & Chereni, 2020; Ngwenya et al., 2023), and reported in reviews of African studies of youth resilience (Theron et al., 2023c; Van Breda & Theron, 2018). We did, however, wonder why supportive strangers and collective action were largely absent from the accounts of resilience when participants reported lower risk exposure. While a definitive answer requires a follow-up study, it is possible that greater risk exposure attunes individuals to traditional, time-honoured ways-of-being and -doing (Wilbur & Gone, 2023).

What we draw from these reflections on our findings is that there is at play a confluence of resources and developmental progressions expected during emerging adulthood that are acted upon by the chronic disadvantage faced by all of our participants. The context remains challenging, and those challenges did not shift dramatically over the 10-week data-gathering period. The developmental processes experienced during emerging adulthood (self-focus, identity exploration, risk management, relationship-building), however, are dynamic, and are interwoven with the relative static circumstances in which our participants live. Against this backdrop, relational, community and cultural resources were prominent enablers of mental health resilience, reinforcing the understanding that resilience-supporting resources are sensitive to context (Theron & Ungar, 2023), and developmental stage (Yoon et al., 2021). Still, while accounting for the distinguishing nuances of the relational and community resources reported by the Higher Risk group (i.e., collective response/help from strangers and assistance during times of emergency), our findings caution against inattention to resources that matter more at higher vs. lower levels of risk.

Limitations, Advances and Contribution

Though our sample was significant for a qualitative study of this kind, it was based in a single township in South Africa, limiting what our results can say about other populations of emerging adults. Although we gathered weekly data over a 10-week period using a common framework, reached theoretical saturation, and as such have been able to develop a unique new dataset that can support additional analysis, it is possible that a larger or less homogenous sample might have led to (i) a wider range of stressors and resources than those we have shared here, (ii) further differentiation across the two risk groups, or (iii) deeper insights into the resources to which they have access.

While acknowledging these limitations, our findings represent an advance to our understanding of resilience to depression in conditions of chronic stress. It sheds light on the day-to-day experiences of a specific sub-group of emerging adults living in conditions of chronic disadvantage who report minimal to mild depression despite their deprivation. Specifically, our study is the first study of African youth to investigate the differentially impactful nature of commonly occurring protective factors and show that even when emerging adults live in the same community and share ethnic roots, their experiences of risk exposure will determine which relational, community and cultural resources matter most to mental health. Going forward, attention to differentially impactful protective factors, using methodologies that allow for such granular insight (Panter-Brick, 2015), will be crucial to the mental health of emerging adults living in chronically stressed communities.

We also make a distinctive contribution methodologically. The immediacy of the digital diary approach led to some factors appearing more frequently unprompted than they do in the wider literature. Specifically, explanations of the implications of transport infrastructure (or absence of such) as a stressor, and atypical coping strategies had a presence in the data that is otherwise underrepresented in the literature.

Our experience of using digital diaries invites future resilience studies to consider three needs in research design. First, the need to pilot and use innovative and interdisciplinary methods that allow emerging adults in stressed communities to share their lived experience without burdening them as some more traditional elicitation methods might (Liebenberg, 2019). Second, the need to account for the interplay between the ‘digital’ system (e.g., digital materialities, the labour required to maintain connectivity in a low-resource environment) (de Lanerolle et al., 2020) and other systems impacting on emerging adults’ resilience pathways. Finally, the need to confront and disrupt methodological assumptions rooted in Global North research contexts by making use of contemporary digital resources for data gathering.

Conclusion

Even though participants with lower and higher risk exposure all reported relational, community and cultural resources, supportive relatives, supportive strangers, and collective agency distinguished the relational, community and cultural resources reported by emerging adults with higher risk exposure but lower levels of depression. Our findings therefore suggest a complex interplay between the chronic stressors and protective factors as reported by our participants across both the Higher Risk and Lower Risk groups and their experience of depression, while also evidencing the differentially impactful resources that distinguished the accounts of emerging adults in the Higher Risk group.

This more detailed understanding of the commonly occurring resources cautions against methodologies that deliver broad or superficial understanding of the resources that support mental health (i.e., minimal to mild depression) at varying levels of risk exposure for Black African youth. In the absence of emerging adults voicing their lived experience of relational, community and cultural resources, we might have assumed that there were no differentially impactful resources for young adults living in eMbalenhle and self-reporting lower versus higher risk exposure. Instead, their digitally mediated insights confirmed the differential impact of specific resources. Their value appears to be heightened value as levels of risk exposure increase (Ungar, 2017, 2018). Even when emerging adults live in the same community and share ethnic roots, their experiences of risk exposure appear to determine which resources matter more for mental health. In the context of our study, youth at higher levels of risk exposure identify family members and supportive strangers who go to great lengths to keep emerging adults relatives safe and community members who honour traditional African values of collectivism as especially important to their resilience.

Supplemental Material

Supplemental Material - Resilience to Depression Among Emerging Adults in South Africa: Insights From Digital Diaries

Supplemental Material for Resilience to Depression Among Emerging Adults in South Africa: Insights From Digital Diaries in Diane Thembekile Levine, Linda Theron, Sadiyya Haffejee, and Michael Ungar in Emerging Adulthood