Social determinants of mental health among older adolescent girls living in urban informal settlements in Kenya and Nigeria during the COVID-19 pandemic

ABSTRACT The health burden due to mental health has historically been underestimated with focus on communicable diseases and deaths and little consideration of disability and comorbidity effects of poor mental health. Recent data show increasing trends of mental health disorders as a share of global health burdens and vulnerability of adolescents. This paper aims to explore social determinants of mental health as experienced by adolescent girls, drawing attention to gendered risks during the COVID-19 pandemic. Semi-structured interviews with twenty-two adolescent girls in urban informal settlements in Kenya and Nigeria reveal unique environmental, socio-cultural, economic and educational factors that threatened their mental wellbeing. The pandemic exacerbated these determinants. An equitable recovery will require a consideration of not only disproportional mental health outcomes, but also social determinants that contribute to these outcomes. As more than half of the urban population in sub-Saharan Africa reside in informal settlements, this study has implications for youth-focused mental health interventions in these and similar settings.


Introduction
Mental health disorders represent a significant health burden in low-and middle-income countries, yet they remain understudied, undiagnosed, misdiagnosed or untreated (Freeman, 2022;GBD 2019Mental Disorders Collaborators, 2022;Jenkins et al., 2011;Sankoh et al., 2018).Recent data indicate that mental health as a share of disability, and disability as a share of health burden, is on the rise, with depressive disorders, anxiety disorders and self-harm as leading causes of disability among adolescents (Vos et al., 2020).Nonetheless, mental health falls behind competing health priorities in sub-Saharan Africa as reflected in national budgets and development aid (Freeman, 2022;Jenkins et al., 2010).Young people face high risk of developing mental health disorders and unwellness.Nearly 50 percent of all lifetime mental disorders start in teenage years with less severe disorders (which typically precede severe disorders) rarely clinically diagnosed (Kessler et al., 2007).There is need to document adolescents' experiences of mental health in order to inform investments and programme development.
Mental health research reveals that while biological and psychological factors are significant determinants of mental wellness, social determinants are also important contributors (Compton & Shim, 2015;World Health Organization, 2001).Social determinants of health are non-medical factors that influence health outcomes such as conditions in which people live and the policies shaping these conditions (Commission on Social Determinants of Health, 2008).Deriving from this definition, social determinants of mental health (SDoMH) refer to socio-economic conditions and systems that uniquely affect mental health outcomes, even where they don't affect physical health outcomes (Allen et al., 2014;Compton & Shim, 2015;World Health Organization, 2014).In sub-Saharan Africa, for instance, factors such as poverty, child labour, exposure to violence and trauma, HIV and AIDS status, orphanhood and being out of school have been identified as SDoMH (Ismayilova et al., 2016;Jörns-Presentati et al., 2021).Gender inequality is also highlighted as an additional risk factor, based on biological sex, as well as gendered social norms and attitudes (Hyman et al., 2006;Ogbonna et al., 2020).While there are mixed findings on gendered prevalence of mental disorders among adolescents (Jörns-Presentati et al., 2021;Ogbonna et al., 2020;Volley et al., 2022), determinants such as gender-based discrimination and violence, early marriages and gendered expectations particularly affect girls.Despite the growing body of research on mental health among adolescents in sub-Saharan Africa, research on associated gendered risks is limited.
Research on adolescent mental health and COVID-19 suggests higher rates of depression and anxiety symptoms among older adolescents and girls, compared to boys and younger adolescents (Benton et al., 2021).While studies identify social isolation, fear of illness and education interruptions as drivers of poor mental health, they do not consider how pre-existing social determinants, such as gender inequality, might interact with the effects of the pandemic and influences differences in mental health outcomes (Racine et al., 2021).There is also less research on adolescents in sub-Saharan Africa and living in conditions where they are at greater risk of the negative social determinants of health, compared to those living in high-income and more secure contexts.
This paper aims to explore SDoMH among adolescent girls in urban informal settlements in Kenya and Nigeria during the COVID-19 pandemic, revealing how environmental, economic, educational and socio-cultural determinants can influence mental health during crisis.As this is not meant to be a clinical study, but a contribution to literature on SDoMH, we define mental health as a continuum of mental functioning including individuals' subjective perceptions of their wellbeing (Keyes, 2005;World Health Organization, 2001).An individual's mental health is not static but can change across this continuum over time, including in response to unexpected shocks and life events (Friedman & Thomas, 2009).
COVID-19 provides a critical context through which to understand how such shocks interact with pre-existing SDoMH of adolescent girls.In Kenya and Nigeria, pandemic public health measures had negative socio-economic effects, particularly for women and youth (Murage et al., 2022;Tan et al., 2021).These effects ranged from lost employment and incomes, school and service closures, incidents of gender-based violence and early pregnancies, and reduced access to sexual and reproductive health services, and were particularly pronounced in low-income informal settlements (Murage et al., 2022;Okem et al., 2022;Osembo et al., 2022Tan et al., 2021).As more than half of the urban population in sub-Saharan Africa, and many youths, live in informal settlements, this study hence has implications for youth-focused mental health interventions in these and similar settings (UN Habitat, 2013;Zerbo et al., 2020).Our study focusses on experiences of adolescent girls because of reported adverse effects of the pandemic on women and girls, as well as a dearth of mental health research on this demographic.

Materials and methods
Data were collected through semi-structured interviews with 22 adolescent girls between ages 15 and 19 years in Kenya (Nairobi County) and Nigeria (Lagos State) between March and May 2021.Inclusion criteria were: identifying as a girl, age between 15 and 19, and living in informal settlement.The age inclusion was based on an assumption of potential similarities in cognitive, social and emotional development signified by higher level of self-identity, independence, emotional self-regulation and expression, and future orientation, etc. compared to younger adolescents (Sanders, 2013).
Research participants were purposively sampled and recruited through identified faith-based groups and non-profit organisations.While this recruitment strategy facilitated access to participants, it presented plausible bias as participants were likely to have access to some form of psychosocial support through these organisations.The sample was limited due to constraints faced in recruiting adolescent girls during the pandemic such as fear of COVID-19 infection and familyimposed restriction on movement and engagement outside the home.Consequently, as opposed to striving for data saturation, we interviewed as many participants as volunteered, noting that studies have found that nine to 17 interviews can reach saturation (Guest et al., 2006;Hennink & Kaiser, 2022).We originally conceptualised the study as a comparison, but as more similarities than differences emerged in the data, and considering the small sample in each case study (14 participants in Kenya and eight in Nigeria), this paper focuses on what respondents had in common.
Two qualitative researchers (co-authors in this paper) based in Kenya and Nigeria, respectively, collected data through face-to-face interviews with the support of experienced research assistants.Public health measures such as wearing of face masks and maintaining physical distance were followed.The two researchers and one other co-author were involved in data analysis.Interviews were sound recorded and transcribed to facilitate data analysis.Researchers conducted open-ended semistructured interviews, using a guide developed from the gender and COVID-19 matrix, which allows an examination of gender power relations as it pertains to the impacts of the pandemic.The matrix has been used to guide qualitative studies on the gendered effects of the pandemic (Morgan et al., 2022;Smith et al., 2022).Specifically, the matrix explores how six COVID-19 domains (Risk or vulnerability to COVID-19 infection; illness and treatment of COVID-19; Health systems and services; Social impacts; economic impacts; security impacts) intersect with five gender domains (access to resources; labour/roles; norms/belief; power; institutions) (Morgan et al., 2022).Our interview guide was based on the questions in those domains, as well as an early literature review and media analysis that brought to light issues likely to affect adolescent girls n.d.).Interview questions were broad e.g.'how did the pandemic affect you and your wellbeing?' with probing questions allowing an understanding of interaction with domain-related issues noted above.The semi-structured interview approach reduces researcher's bias as it allows participants to bring up topics or themes important to them (Busetto et al., 2020).Given the sensitivity of the research topic, researchers were prepared to address emotional distress e.g. by allowing participants to skip questions, pausing or ending the interview, and providing information on where participants could receive psychosocial support.
The two country studies were selected for this study because of similarities across the contexts: social e.g.gender norms and expectations; economic e.g.dominance of the informal sector and gendered division of labour; political e.g.distrust in government and security approach to COVID-19 response; and healthcare e.g.limited budgetary and resource allocation for mental health services (Murage et al., 2022;Tan et al., 2021).While 'informal settlements' can be defined differently based on context (Diab et al., 2020), in both Kenya and Nigeria, the term is used interchangeably with the term 'slums' to refer to groups of dwellings inhabited without compliance with planning and building regulations and which are characterised by crowded living arrangements, limited access to clean water and inadequate sanitation (Ministry of Land, Housing and Urban Development, 2016; UN-Habitat, 2021).Those living in these settings typically face significant economic hardship, with many relying on daily wages.Table 1 illustrates the basic demographic information of participants.Incorporating similar contexts in two different countries broadens understanding of SDoMH with implications for other similar settings.
Authors utilised a SDoMH lens to conduct thematic content analysis of interview transcripts (Nowell et al., 2017).Social determinant frameworks, though presenting a range in terms of focus, aim to understand how the circumstances in which people live, work, study, etc. shape their health outcomes.We first read the transcripts and coded for perceptions and experience of poor mental health, such as mention of stressors, feelings of depression or powerlessness.We then grouped these according to social determinants, such as economic status, access to education and security, identifying themes within the groupings.The quotes in the results reflect illustrative examples from these groupings.Within this analysis we did not use clinical measures of mental health but rather relied on self-reported perceptions of emotional and social wellbeing (Keyes, 2005).
Ethical approval for the study was secured through the AMREF Ethics and Scientific Review Committee in Kenya (ESRC P917/2020) and the Health Research Ethics Committee at the Lagos State University Teaching Hospital in Nigeria (LREC/06/10/1522).All participants were briefed about the study before the interview and written consent was obtained before beginning each interview, including consent for the interview to be sound recorded and transcribed.

Results
An analysis of the transcripts resulted in four main themes identifiable as social determinants of health through a variety of foundational models of health determinants (Compton & Shim, 2015;Dahlgren et al., 2006;Dyar et al., 2022).Our findings logically clustered around the following social determinants: environmental, economic, educational and socio-cultural determinants.These determinants directly or indirectly contributed to negative mental health outcomes of adolescent girls in this study.While in some models, social and economic determinants are clustered together, in our study, social and cultural determinants clustered together.Additionally, other models cluster educational determinants under socio-economic determinants.In this study, educational determinants were prominent and hence were more appropriately addressed singularly.
Environmental determinants pertained to living conditions of the adolescent girls including housing, water and sanitation.Economic determinants demonstrated effects of lost livelihoods and consequent reduced capacity to access necessities.Educational determinants addressed the effects of the pandemic and subsequent health emergencies on educational opportunities and aspirations.Socio-cultural determinants pertained to security factors (e.g.domestic violence), gender roles and social connection.While these determinants are distinct, they do not operate in isolation from each other.For instance, environmental factors such as inadequate housing is associated with economic determinants.Also, disproportionate distribution of unpaid care work (under sociocultural determinants) is associated with educational factors.
These determinants were associated with mental health outcomes as expressed by study participants.It is notable that while some participants were explicit on mental health effects associated with mental health disorders such as feeling 'depressed' or 'anxiety', others implicitly used words and expressions, or asked rhetorical questions, that allowed an examination of perceptions of

Environmental determinants
Adolescent girls participating in our study explained how awareness that environmental and related factorssuch as inadequate sanitation and crowded living conditions -made them particularly vulnerable to COVID-19 contributed to anxiety around infection and illness.Participants noted that they were not always able to wash their hands frequently due to limited water supply even when they knew this would reduce their chances of being infected.Hand sanitisers offered an alternative, but this was out of reach for many because of associated costs.Crowded living conditions were noted to make it impossible to social distance within the household and with neighbours (due to closeness of housing and shared amenities).Most housing entailed single-room-or two-room dwellings in rows next to each other, each accommodating one family: participants' average household size was 5.5.
The first time I heard it, I was in school, I was scared because they actually said a mere contact with someone can make you contact the disease.At that moment we were asked to vacate the school premises because the school didn't want any risk, so they were asking people to go home and take the necessary precautions … this is Lagos.You'll pass beside someone, and they would surely sneeze, someone would surely cough and use their hand to rub their face and touch you.So, I was really scared and didn't really know what to do (A08).
Participants were aware of the COVID-19 public health advisory but felt unable to follow through on them, and thus felt powerless to protect themselves.The heightened risk of COVID-19 infections due to environmental factors coupled with inability to protect themselves contributed to disillusionment regarding risk to their physical health and that of their family.This was also linked to anxiety associated with potential loss of household income: 'Sometimes I think like when this COVID, when my parents maybe get COVID and they die, how are we going to cope up, who is going to help us?' (AG006) Participants shared how fear of COVID-19 was layered on top of concerns about pre-existing health challenges.One shared the anxiety brought by COVID-19 pandemic at a time when her neighbourhood was dealing with a cholera outbreak.Due to population density, and inadequate housing and sanitation, informal settlements are more vulnerable to outbreaks of communicable diseases and shoulder a high disease burden (Oppong et al., 2015;Rebaudet et al., 2013).Additionally, some participants, or their families, also suffered from a pre-existing condition including asthma and ulcers.They shared struggles of managing these conditions even before the pandemic.Access to health care services was limited due to financial constraints, distance and delayed services in hospitals.Almost all participants narrated how seeking services in clinics or hospitals was done as a last resort.They first opted for home remedies, herbal medicine, self-medicating, or following the advice of pharmacists.The pandemic made participants afraid of seeking healthcare services as they feared they would get infected, or test positive for COVID and then be forced into isolation units.Their anxieties and behaviour around healthcare seeking were also influenced by the fear of their caregivers: My grandma won't even allow me [to go to the hospital] … she did some herbs, then I went to chemist to buy drugs, anti-malaria … She's scared because her own definition of COVID is death like that's a death sentence, that's her own definition of COVID.(A03) Participants shared that they or their family members did not go into hospitals but rather self-medicated for malaria when they experienced COVID-like symptoms.
During the lockdown, my dad was panicking he could not take him [participant's brother] to anywhere because they will be like COVID-19, so he just went to one of these chemists [pharmacies] to get normal malaria drug … .We are not the type of people that go to the hospital regularly, it is built in us, and we are scared of all those types of things (A04).Some participants in high school received health care provided by schools; this was unavailable to them during school closures.Limited access to and fear of health facilities likely reduced opportunities for mental health diagnosis and treatment for participants (Ogueji & Okoloba, 2022).

Economic determinants
Economic hardship caused by COVID-19 response policies (e.g.social distancing, curfews, lockdowns) negatively affected mental wellness of adolescent girls due to anxieties about insecurity of household incomes and inability to meet their basic needs.Some coped by taking on new responsibilities to contribute to their household income, further compromising their wellbeing.
Those living in informal settlements in Kenya and Nigeria were particularly affected by the COVID-19-related economic downturn as most residents rely on casual employment, micro-to small-enterprises and daily wages.Most participants reported that their family relied on a single person's income, including five with single mothers and two relying on their grandmothers.Evening curfews reduced the number of working hours, particularly affecting enterprises dependent on sales from wage workers coming from work in the evening.
The curfew is the one which has affected me and my family most … Now, when curfew time comes, we have to lock up [mother's business].Sometimes you close up when you haven't sold anything, so the curfew is affecting us a lot … there is a problem of balancing equations.What is being spent is more than what is being earned; it is little (A005).
Additionally, as most enterprises required close contact with clients, the social distancing measure and fear of being infected reduced business.Residents from informal settlements also faced stigma while seeking paid work: 'Her [participant's mother] potential employers find it hard to employ people from Kibera [informal settlement] because they think that they have Corona' (AG009).While governments encouraged employers to allow employees to work from home, this advisory was not applicable to most residents in informal settlements due to the nature of their work.
Participants shared how anxiety regarding household income negatively affected their mental wellness: 'I felt bad because corona made my father stop working, my mother stayed home, everything turned out badly' (AG010).This participant noted that she sought employment after completing high school to support her family but then lost her job due to the pandemic, making her 'angry'.
Families coped with ensuing economic hardships by reducing the quantity and quality of their food, drawing from savings and borrowing money to meet urgent needs.Food and housing insecurity was a source of constant anxiety among the adolescent girls.There was little room for contingency.One participant shared how her father compromised his health by not going to the hospital when sick for fear of being quarantined as having COVID-19 and consequently being unable to work.Another narrated how savings previously secured for her higher education were used to pay for her grandmother's hospital bills.'It [the COVID-19 pandemic] affected us through hunger, financially and mentally' (A01).
Participants shared feeling depressed watching their parents struggle to provide for them on a daily basis: It becomes stressful to her [participant's mother] that sometimes you find her thinking and when you ask her what she is thinking about she says she is thinking about how she will pay the loan, yet she doesn't have money.(AG018) Where parents remained employed, there was constant fear that they would soon lose their employment: 'I'm afraid that she might also be among the people they might decide to drop … we are too young to take care of ourselves' (A08).Some participants described feeling responsible to help support their families financially, with one participant highlighting the gendered aspect of such responsibility: It is not all of them [boys] that will drop [contribute] money for the family even if they have money.But we girls, we feel that all the responsibilities are on us, so we have to.It is like our job (A04) These feelings of responsibility were particularly highlights by adolescents who had completed high school.
Preoccupation with survival and support for family during this economic hardship came at a cost to the adolescent girls.One girl noted the health impacts of assisting her mother with domestic work for a client, 'The work used to make me tired that it would make me sick sometimes' (A06).Others described feeling anxious to ask for basic necessities, such as sanitary pads: 'My friend told me she is afraid of asking from her parents … they ask her if they buy pads for her, how are they going to buy food?' (AG001).Survival strategies such as borrowing and asking for favours were noted to make adolescent girls vulnerable to sexual violence and early pregnancies.Their observation of this consequence made them feel disheartened: 'We saw that a lot of girls got pregnant.Some of them were saying that they were involving themselves in sex so that they can get pads … That really hurts me' (AG016).

Educational determinants
Interruptions and changes in education delivery was a significant stressor for the adolescent girls.Following school closures in March 2020 (in Nigeria and Kenya), students were expected to continue studying, with some schools facilitating virtual learning.More than half of the research participants were in high school at the time of the interview: One had dropped out of high school while the rest had completed high school and mostly engaging in vocational or trades training.Participants, particularly those who were in high school, narrated how learning from home was impossible in many instances because of their crowded living situation.Having no private room to retreat for study, some participants noted that they could not study because of noise by smaller siblings.As houses are close to each other, noise from neighbours and children playing outside affected the student's concentration.Further, students who previously relied on public libraries as conducive learning spaces outside of school were unable to access them due to closures.Other avenues of learning such as private tutoring and discussions with peers were also curtailed.Increased domestic and care work, resulting from everyone being at home, fell disproportionately on adolescent girls leaving them with little time for learning.
While introduction of virtual learning extended learning opportunities beyond the classroom, this created additional burdens for participants.Many did not have access to devices or financial resources to facilitate virtual learning.In some cases, devices within their households were out of reach: 'My parents, they don't allow me to access the phones and stuff.They don't.… so mostly I rely on my books' (AG016).Due to financial constraints, some did not have access to school textbooks.
Actually, in our home no one has a smartphone, so I go to neighbor's and I ask them to allow me use their phone then I copy from there … Sometimes when he doesn't want me to use his phone he tells me that he doesn't have bundles [internet] even when he has.So, when that happens I go to another person (AG009).
The adolescent girls narrated how disruption of their learning and academic progress affected their mental wellbeing.Participants highlighted indignities they faced on a daily basis as they tried to access learning materials.For example, 'when you borrow a textbook from someone, they mock you by saying you don't have money' (AG007).Being unable to continue with their education at home led to feelings of despondence.
You have to remain behind because there's no way you can access what you need at that time.So, it's very discouraging.And sometimes I just think like I just let go of all these issues about studying [drop out of school] and whatever if I can't get the required materials.So, it's been hard (A006).
The adolescent girls described feeling anxious, hopeless and depressed because of not knowing when their education would resume or when their delayed examinations would take place, with one participant describing how she thought her 'life was over' (AG015).
I had fear because the schools were going to be closed and I didn't know how long it would take for us to go back … Sometimes I get depressed … if we don't get back to school or we again repeat classes and you know the days are moving and we are not growing younger.If you repeat you add years to your age and people will think you are too old to be in that class, and its only brought by Corona (AG009).
Participants were future-oriented, describing education as their path for economic mobility.Educational disruption registered as disruption of their future opportunities, dreams and ambitions: 'Educationally, it [the pandemic] has left us stranded; we just don't know what to will happen' (A018); 'if it's getting worse then we are told to go out of school, then what is going to happen to us?' (AG006) Some described how loss of hope contributed to some students dropping out of school and others getting married: 'they [other students] have gone to be married, others have dropped out at form one [first year of high school] saying their life has just ended like that' (AG015).Participants described feeling disheartened when they realised, on reopening school, that some of their classmates and friends did not return to school because they were pregnant.A participant who became pregnant during the pandemic noted how 'the girl child' was particularly affected by the school closure: At the time of the interview, she was unable to go back to school, an outcome she blamed on discrimination.
Most of my friends, some are even married which isn't meant to be.We all did JAMB [examination] together with the intentions of being in school and passing, only for me to hear that they got married.Their mum said, 'what is the point of staying at home when JAMB isn't saying anything, there's a pandemic and who knows when schools would resume'.So, their parents just decided to send them into marriage (A08).
Financial constraints, exacerbated by the pandemic, continued to impede the learning of students even after schools were reopened.Participants narrated their prior experience of being sent home because of lack of school-related fees hence missing class.One shared that she might be sent home from school soon: 'Mum has not paid school fees, but I haven't been sent away yet to bring the money' (AG006).Similar struggles affected those who had completed high school and wanted to transition to higher levels of education.Some participants who wanted to pursue university programmes, settled on doing trades or going into the labour market to supplement their parents' income: One girl noted that with time her mind got 'off education totally' (A04).This forced abandonment of educational goals was demoralising to participants.Impact on education and future educational and career opportunities was notably the most discussed issue by the adolescent girls; this frequently came up as the first issue brought up when asked generally about wellbeing.

Socio-cultural determinants
Participants indicated a heightened sense of insecurity during the pandemic, both outside and inside their homes.The adolescent girls shared feeling vulnerable to harassment and assault by older boys who sat together in various corners of their neighbourhood; these 'area boys' sometimes engaged in drugs and crime.They felt more vulnerable as these boys now frequently gathered because of school closures, lost jobs and lockdowns.Having fewer people outside during curfew hours contributed to anxiety about their personal security; participants described their reduced ability to seek help from by-passers in case of assault.'During the lockdown because everybody was indoor, even though you are to go out to go and buy something, even if you [were] raped [and] you called for help, there is no body to come and help' (A03).Another shared, 'It's been really tough … .We are planning to move away because this place is not that secure especially for a female child' (A08).
Within their homes, adolescent girls were made more vulnerable by lockdown measures if they lived with an abusive individual.One participant narrated the experience of a 15-year-old friend who was sexually assaulted on multiple occasions by an uncle, who had lost his job and hence stayed at home.This friend was reluctant to expose her uncle because no one would believe her and that it would bring 'shame to their family' (A04).Feeling powerless to assist her friend, the participant only had advice to offer: 'I told her to avoid him, when he is inside, she should go outside and vice versa.That way if he sees less of her maybe it will stop' (A04).Another participant narrating a similar incident highlighted that in such situations justice is often not afforded to adolescent girls; such cases, even when known to other members of the family, are often not reported.
Increased cases of domestic violence and increased tensions at home also affected the mental wellbeing of the adolescent girls.Participants spoke of arguments and violence between their parents, and between themselves and their parents.They noted that these incidents were provoked by frustrations due to financial constraints in meeting basic needs, parents being over-protective, increased domestic and care work for adolescent girls, and generally being together all the time at home 'jam-packed'.
When mom and dad are home you witness them disagree all the time because dad every time sees that things are not done the way he wants them done and therefore keep complaining unlike when he is at work … You know when he is at work, he doesn't spend much time at home with us and does not see all those mistakes.He sees mistakes where there are none according to us (AG009).
The only thing I am not used to is staying with my parents the whole day in the house.I am not used to that … you find sometimes we don't get along that much.So, it is very hard for me … .I never knew that my dad was someone who is harsh (AG016).
Participants illustrated the gendered nature of their experience noting that parents were more likely to direct their frustrations to girls than boys.Conflict in the home contributed to feelings of consistent anxiety, which one participant noted made it hard to 'breath' and to express her feelings (AG007).
The pressure on the parent, they sometimes take it upon the girls unlike the boys because some of the parent can't handle them [boys], so they will take almost all their pressure to the girls; they will beat the girl and stuffs like that (A03).
Another source of family tension and emotional stress emanated from increased unpaid care work assigned to them, including caring for their younger siblings, caring for those who fell sick over the course of the pandemic, and domestic chores.As all family members were at home, these tasks not only increased but were continuous throughout the day and seemed to be never-ending to participants.These roles were clearly gendered, with the girls and particularly older adolescent girls doing almost all care work at home.Adolescent girls, who worked or went for trades training, narrated waking up as early as 5 pm to carry out these responsibilities, and again doing this work at the end of their day: '[I] fetch water, wash cloths, clean the house … when I am through with that, I bathe, and I'll prepare for the market' (A02).One participant noted that she had to do these chores while sick during the pandemic because her brother was unwilling to assist.Participants shared of the pressures that came with fulfilling these roles in time and to the satisfaction of the parents.They narrated how they felt constantly under 'pressure' to fulfil these roles making them both physically and mentally exhausted.
It has been very busy because you are always on toes trying to beat the curfew and sometimes dad comes at around 6:50pm or if he is late at 7:30pm.When he arrives, he would want you give him water to bath, tea, and food should be ready by that time.It becomes hard because every time your mind is running; you have no time to rest, it becomes hard (AG018).
The only pressure I felt was the one that had to do with the cooking … Everybody was at home and I had to cook for everybody in the morning and afternoon and also make sure that everybody had something to eat before going to bed.It was like that everyday … I lack appetite to eat, because after I do the cooking and everything I get so tired and I tell myself I'm okay and want to rest (A08).
Their attempts at negotiating these roles were met by reprimands or disciplinary measures, adding to family conflict, anxiety, and stress.I use my medications daily [painkillers for recurring headaches] and I am still the one doing the house chores … this family they will just be like obirin, obirin [the girl child] are the one that is supposed to do the house chores.Even if I am like 'I am washing plates naa, let my brother fetch water', they will say 'if you reach your husband's house, is it your brother that you will come and call to fetch water for you?' (A03) Mental and emotional stresses that characterised the pandemic experiences of the adolescent girls were exacerbated by social isolation.Participants commented on their inability to visit their extended relatives in other states or counties.Others noted how their community connections through church and volunteer activities were also severed.The pandemic also limited the opportunity for peer-engagement for the adolescent girls.Some participants indicated that while their mobility was previously limited because of overprotective parents, security concerns and gendered expectations, their movements were further curtailed during the pandemic by their parents and their own fear of infection.While home visits could have offered an alternative, participants noted that this was not possible because of all family members being at home in close quarters and fear of COVID-19 infection.Due to this isolation, they shared feeling 'bad' (AG005, AG003) and 'lonely' (AG001) You feel like that person will bring you corona here, so you distance yourself from the person … .Because you don't want to go out there and get corona and then come and affect family members … they are saying I am pretending, or I am proud (AG018).
Those who had strong friendship through school connections were particularly affected by school closures, especially those who went to boarding schools.One participant feared physical displacement through demolition of their housing for a government infrastructure project; threatening to further isolate her.
The adolescent girls indicated that disconnection from the peers was the most significant social disruption of the pandemic.At a time of social re-orientation among adolescents, towards peers and away from their caregivers (Morningstar et al., 2019), some participants felt like they lost their first line of emotional support.A young mother lamented: 'It is just like something is bothering your mind and you know that if you tell your mother there is a way she will react and the person you can talk to is far away.So, it becomes difficult' (PNM012).

Discussion
The COVID-19 pandemic accentuated and exacerbated SDoMH particularly affecting young people (Meherali et al., 2021;Santomauro et al., 2021).A 2020 survey on mental health in Nigeria indicated that cases of depression were about ten times higher compared to pre-pandemic levels; one in three respondents had depression (Olibamoyo et al., 2020); negative mental health impact in Nigeria has been reported elsewhere (Ogueji et al., 2021).A survey in Kenya, in the same time period, found that almost all participating adolescents reported anxiety with about 50 percent reporting symptoms consistent with depression (Population Council, 2020).Research also suggests higher rates of depression and anxiety symptoms among older adolescents and girls, compared to boys and younger adolescents during the pandemic (Benton et al., 2021).While COVID-19-related measures, such as school closures and physical distancing, have been identified as contributing to these trends, this study uniquely considers how these measures interacted with pre-pandemic SDoMH to affect the mental health of those already experiencing multiple inequitiesadolescent girls in informal settlements.Findings shed light on how gender inequalities, often exacerbated during crisis, contribute to differences in mental health outcomes.
Recognising unique gendered risks likely to compound SDoMH (Abdi et al., 2021), our study explored experiences of adolescent girls residing in informal settlements.We found four aggregated factors which together contributed to the mental unwellness of study participants during the pandemic: environmental factors, economic factors, and educational factors, and social-cultural factors.
Environmental factors such as inadequate sanitation and crowded housing which increased risk of COVID-19 infection heightened anxiety among participants as they felt unable to protect themselves from infection.These fears were not unfounded: In Nairobi, 60 percent of all COVID-19 cases were reported in informal settlements by 2020 (Namwaya & Abdi, 2021).Furthermore, such fears can contribute to ongoing mental health threats.A study in Nigeria found that stress due to fear of contracting COVID-19 increased the likelihood of developing PTSD by over three times (Olibamoyo et al., 2020).The pandemic also added to stresses of pre-existing disease burden with lockdown, curfews and isolation mandates further limiting access to healthcare.
Economic factors characterised by lost incomes, food and housing insecurity, and debt contributed to anxiety over unmet needs.A qualitative study in Nigeria made similar conclusions on lost incomes as a determinant of mental health during the pandemic (Ogueji et al., 2021).Food insecurity has been highlighted elsewhere as a significant determinant of mental health in informal settlements (Abdi et al., 2021): By July 2020, severe food insecurity was reported in over 50 percent of households in 10 Kenyan informal settlements (Gikandi, 2020).The focus on adolescent girls in this study suggests that food security-related stress among caregivers may also lead to increased stress among dependents.Pre-occupation with survival and supporting family by taking on paid work came at a risk to adolescent girls' health and education, with some resorting to transactional sex, which has further been linked to poor mental health outcomes (Jacobson et al., 2020;Osembo et al., 2022).
As highlighted elsewhere, school closures affect adolescent's access to material resources such as food and sanitary pads, as well as psycho-social support through counselling programmes and friendships (Murage et al., 2022).Participants reported increased concern about their education due to the challenges of studying from home in crowded living conditions and increased unpaid care work.Additionally, participants felt undignified for constantly needing to borrow digital devices and books from neighbours and friends to facilitate virtual learning.This study indicates that the pandemic threatened to widen a gender gap in education access as virtual education was blind to digital divides, unpaid care limited opportunities to learn at home, and educational disruption and economic hardships contributed to early pregnancies and marriages.Uncertainty around educational continuity stood out as a significant contributor to mental unwellness as this was viewed as the avenue for social mobility, and therefore hope for the future.
Increased sense of insecurity also effected participants' mental health; having fewer people in the streets made them more vulnerable to assault.Inside the home, girls were also made vulnerable to sexual, physical and emotional abuse.Such vulnerability increased risk of mental unwellness.A study in Nigeria found a strong association between mental health outcomes and experiences of abuse, stress of staying at home, and apprehension or helplessness due to lockdown measures (Olibamoyo et al., 2020).While a pre-pandemic quantitative study in Nigeria found no statistically significant gender difference in children's exposure to violence (Asagba et al., 2022), this study indicates that parents of study participants were more likely to direct their pandemic-related frustration to girls than boys in the form of domestic violence.Participants in this study further discussed how the combination of increased conflict in the home, and lack of social connection outside of it, contributed to declining mental health.They described increased anxiety from witnessing domestic violence at home, and how gendered expectations that they would take on increased unpaid care work resulted in conflict.With fewer opportunities to engage outside of the home, girls felt isolated in processing and dealing with these conflicts.
These findings highlight how health emergencies can act as mediators or moderators of SDoMH particularly those related to gender inequality (Bernardini et al., 2021).Experiences of the adolescent girls in the study indicate that as moderators, the pandemic and subsequent public health measures might have exacerbated pre-existing social determinants such as insecurity, crowded housing and unpaid care work.As mediators, the pandemic and health measures introduced new social conditions affecting mental health such as food insecurity, education disruption and social isolation.
That we found similar experiences across countries, but in similar informal settlement contexts, both point to the transnational gendered experience of the pandemic, and opportunities to develop responses to these widespread experiences, that can then be tailored to local needs (Smith et al., 2021).Based on findings from this study, such interventions might include peer support for survivors of domestic and gender-based violence during lockdown, counselling for those whose schooling has been delayed or interrupted, and conflict resolution programmes for families.Our research further suggests that mental health intervention at the community level may be more effective than at clinical facilities because of limited mobility of adolescent girls beyond their communities: The study also revealed that hospitals and clinics are not frequented venues in seeking health care in these settings.Barriers to accessing formal health systems reduce access to mental health services, potentially further exacerbating the social and economic effects of poor mental health; this can become a vicious circle.As Freedman once observed, health systems are 'core social institutions that help define the experience of poverty' (Freedman, 2005, p. 20).Contextually relevant, accessible and equitable health care can moderate economic determinants of mental health.
To reduce effects of unpaid care work on the mental health of adolescent girls, community-based interventions might promote shared roles between boys and girls, and school curriculum could promote transformative gender norms (Mũrage et al., 2022).Such interventions can support adolescent girls in recovery following the initial years of the COVID-19 pandemic.Concurrently, there needs to be increased investment and attention to the SDoMH more broadly, such as inadequate sanitation and protection from violence, including recognition of how such determinants are shaped by gender norms and roles, as well as other inequities.To mitigate negative mental health impacts in future pandemics, preparedness efforts need not only focus on how to respond to pathogens but also how to address consequent inequities and mental health risks.
This qualitative study is limited in sample and geography.Although the sample size was sufficient in gathering common themes on identified social determinants of health, findings cannot be generalised.While some aspects are relevant to comparable contexts i.e. informal settlements in other African countries, we recognise a need for more extensive research, including quantitative studies that employ a SDoMH framework and gender lens.In addition, our analysis of gendered risks associated with mental health was based on experiences of adolescent girls; a future study could include adolescent boys to enable more comprehensive gender comparisons.This study contributes to a body of work on SDoMH and gendered risks of poor mental health during health emergencies, particularly in informal settlement settings.

Disclosure statement
No potential conflict of interest was reported by the author(s).
emotional and social wellbeing e.g.feeling 'sad', 'bad', 'lonely' and 'pressure', or 'what is going to happen to us?'.