A cross-cultural interpersonal model of adolescent depression: A qualitative study in rural Nepal

Most cross-cultural qualitative research on depression has been descriptive, documenting symptoms and explanatory models. There is a lack of qualitative research testing theoretical models of depression. The interpersonal model conceptualises grief, interpersonal disputes, role transitions and social isolation as the context in which depression develops and is the basis of interpersonal therapy (IPT), which is increasingly used in cross-cultural settings to treat depression. We aimed to qualitatively evaluate to what extent the interpersonal model can explain adolescent depression in Nepal. Data were collected between December 2018 and April 2019 and comprised transcripts from 126 participants: 25 semi-structured interviews with depressed adolescents aged 13–18; four focus group discussions with adolescents (N = 38), four with parents/caregivers (N = 39), and two with teachers (N = 17); and seven semi-structured interviews with health and non-governmental organisation workers. We coded data using an analytical framework comprising deductive codes representing key concepts from the interpersonal model of depression and IPT, including principles, techniques and strategies. Participants mainly related depression to interpersonal problem areas of grief, dispute, role transition and social isolation. Interpersonal disputes were common, and for many adolescents this involved parental physical and emotional abuse. Although role transitions were common few adolescents grieved loss of the prior role. Distress related to social isolation was evident despite close physical proximity and extensive social interaction with family and community members. Adolescents described coping strategies that were similar to strategies central to IPT, e.g. identifying helpful and unhelpful relationships and generating options and ways of managing problems. In conclusion, interpersonal problems are relevant to this population and conceptualisations align with core principles of the interpersonal model of depression. The findings highlight the importance of addressing abuse and maltreatment in depression aetiology. They also inform future cultural adaptations of IPT in Nepal and beyond, including the opportunity to integrate local coping strategies.

One theoretical model of depression that has been informative for the development and adaptation of interventions is the interpersonal model (Klerman et al., 1984). It uses a diathesis-stress model of psychopathology and incorporates relational theory and findings from research on stress and social support (Lipsitz, 2013). It draws on the ideas of psychiatrists, Adolf Meyer and Harry Stack Sullivan, working in the early and mid-twentieth century, and on early attachment theory and the importance of a secure and trusting infant-caregiver bond (Mufson et al., 2004a). Contrary to Freudian psychoanalysis and its focus on the "intrapsychic", Meyer viewed mental illness as an individual's maladaptation to their social environment. Sullivan's theory of emotions posited that mental disorders result from, and are perpetuated by, ineffective interpersonal communication. He suggested that an individual's actions should be understood in terms of their past and current social context.
The interpersonal model identifies four problem areas that can both trigger and maintain depression: interpersonal disputes, role transitions, grief, and interpersonal deficits linked to loneliness and social isolation (Weissman et al., 2018). Problems arise due to external, developmental and interpersonal factors that increase interpersonal stress, reduce social support, and create emotional difficulties. The relationship between interpersonal problems and depressive symptoms is reciprocal: emotional difficulties act to precipitate and maintain depression and in turn depression worsens interpersonal problems.
The interpersonal model has shaped the development of IPT, a diagnosis-targeted, time-limited psychological treatment developed in the 1970s to treat depression among adults (Weissman et al., 2000). IPT defines depression as a treatable medical condition. It focuses on links between mood and life events, and addressing difficulties related to the four interpersonal problem areas.
IPT has been adapted for treatment of disorders such as bipolar disorder, bulimia nervosa and posttraumatic stress disorder, and for use among adolescents, the elderly and people with HIV (Lipsitz, 2013). It has been implemented in diverse cultural contexts such as China (Gao et al., 2010), Uganda (Bolton et al., 2007;Bolton et al., 2003;Mutamba et al., 2018), Ethiopia (Asrat et al., 2020) and Hispanic populations in the US (Markowitz et al., 2009). The recent popularity of IPT J o u r n a l P r e -p r o o f coincided with its inclusion in the World Health Organization (WHO) mental health Gap Action Programme (mhGAP) guide for the management of depression, and with the publication of the WHO group IPT intervention manual for low-resource, community settings (World Health Organization, 2016;World Health Organization & Columbia University, 2016). A systematic review of psychological interventions in low-and middle-income countries (LMICs) showed that interpersonal treatment elements across a range of therapies have the strongest effect size (Singla et al., 2017). Several mechanisms of change have been postulated, including reduced interpersonal stress, increased social support, and improved interpersonal skills and emotional processing (Lipsitz, 2013;Toth et al., 2013).
A systematic review of psychotherapy studies for adolescent depression globally showed IPT to be comparably effective to CBT, with some advantages in long-term effect and lower attrition (Zhou et al., 2015). IPT has shown evidence for efficacy and effectiveness in studies of depression among adolescents in both high-and LMICs (Bolton et al., 2007;Mufson et al., 2004b). The success of IPT in LMICs raises the question of whether the interpersonal model of depression is consistent with the experience, expression, and trajectory of depression in non-Western cultural groups. As opposed to cognitive-behavioral theories of depression, which focus on individual cognitive distortions and maladaptive behavioral patterns, a focus on interpersonal relations may have more cross-cultural universality. The goal of our study was to employ qualitative methods to evaluate how adolescent depression could be conceptualised using the interpersonal model in a non-Western youth population in rural Nepal. The study was part of a larger research programme to adapt and test the feasibility of IPT in this setting (Anonymous, date of publication).

Study setting
The study was undertaken as part of a broader project to explore the feasibility of group-based IPT for adolescents with depression in Nepal, in partnership with the non-governmental organisation Transcultural Psychosocial Organization (TPO) Nepal. Adolescents aged 10-19 account for 23% of the Nepali population (Ministry of Health, 2016). We focused on adolescents because depression J o u r n a l P r e -p r o o f is one of the leading causes of Disability Adjusted Life Years in this age group, and because adolescents in Nepal are at high risk of depression due to recent and historical trauma (two major earthquakes in 2015 and a 10-year civil war between 1996 and 2006) on a background of socioeconomic deprivation.
The setting was Sindhupalchowk, a rural mountainous district with a population of around 288,000.
Sindhupalchowk was severely affected by the earthquakes in 2015: 2,071 people were killed and many lost their homes and livelihoods (Government of Nepal). A study at this time found that 40% of adolescents met criteria for depression (Silwal et al., 2018). The district population is stratified into various caste/ethnic groups. Tamang ethnic groups comprise the largest group (34%) (UN Women, 2016). Brahman, Chhetri and Newar groups are the most socioeconomically privileged groups and account for 10%, 18% and 11% of the population respectively, whilst Dalit groups (7%) are the least privileged (UN Women, 2016). The main religions are Hinduism (59%) and Buddhism (38%) (Central Bureau of Statistics, 2011). Agriculture is the main source of income, though remittances from labour migrants are increasingly important (Ministry of Labour and Employment, 2016). Literacy rates among females and males are 52% and 68% respectively. Net school enrolment is 95% at primary, 78% at lower secondary and 43% at secondary level (UN Women, 2016). Kohrt and Harper (2008) outline a Nepali ethnopsychology which identifies five elements of the self: man (heart-mind), dimaag (brain-mind), jiu (physical body), saato (spirit) and ijjat (social status) (Kohrt & Harper, 2008). The heart-mind and brain-mind are vital mental health concepts (Kohrt et al., 2012). The heart-mind is the site of emotion and memory, used to refer to feelings and tension (an English term widely used to refer to stress and mild psychological distress) (Kohrt & Harper, 2008;Kohrt et al., 2016;Kohrt et al., 2007). Being too emotional and having too much activity in the heart-mind can lead to physical and psychological illness. The term manko samasyaa (heart-mind problem) can be used to communicate feelings of anxiety and sadness or depression (udas-chinta) without incurring social stigma. Research suggests that heart-mind is J o u r n a l P r e -p r o o f understood by adolescents across genders and caste/ethnic groups in Nepal (Karki et al., 2009). The brain-mind functions in parallel with the heart-mind and represents the logical decision-making mind, controlling thinking and behaviour. Someone who acts in a socially inappropriate way (e.g. against gender or caste norms) is deemed to have a brain-mind problem (Kohrt et al., 2012). The terms paagal or baulaahaa (crazy, mad or psychotic) are used to describe someone with an extreme brain-mind problem. Individuals who are paagal are stigmatised; they and their family members are often socio-economically marginalised (Kohrt & Harper, 2008). Kohrt et al (2012) used adult case studies to explore how Nepali ethnopsychology could be applied to an interpersonal framework (Kohrt et al., 2012). They suggested that heart-mind problems and behavioural control through the brain-mind are widely understood in the context of interpersonal relationships.

Data collection
We explored interpersonal models of adolescent depression in Sindhupalchowk. Data were collected between December 2018 and April 2019 and comprised 42 transcripts: 25 semistructured interviews (SSIs) with adolescents aged 13-18 with depression (mean duration 41 minutes; range 25 to 67 min); four focus group discussions (FGDs) with adolescents aged 11-19 (N=38), four with parents/caregivers (N=39), and two with teachers (79 minutes; 40 to 114 minutes) (N=17); six SSIs with health workers (auxiliary health workers, health assistant, midwife, doctor) and one with a representative from a local non-governmental organisation (43 minutes; 17 to 65 minutes). The Project Coordinator (male, MA Sociology) and Senior Research Associate (female, MA Sociology) conducted the SSIs and FGDs with two Nepali research assistants (RAs).
The Project Coordinator and Senior Research Associate are experienced Nepali qualitative mental health researchers and conducted comprehensive training for the RAs prior to data collection.
School-going adolescents were recruited through government secondary schools. Out of school adolescents were identified through community stakeholders. We identified adolescents for the SSIs using a vignette of adolescent depression developed by Nepali and international clinicians, incorporating both ICD-11 and locally relevant symptoms (Jordans et al., 2015). RAs read the J o u r n a l P r e -p r o o f vignette to adolescents and invited those who felt they had similar symptoms for screening with the Depression Self Rating Scale (DSRS) and the Functional Impairment Tool (Birleson, 1981;Kohrt et al., 2011). Both tools have been validated in Nepal and we used established cut-off scores of ≥14 for the DSRS and ≥4 for functional impairment to identify adolescents with depression. Among those who screened positive we purposively sampled a range of caste/ethnic groups, in-and out of school and married and unmarried adolescents, across the full age range. We interviewed one adolescent who did not undergo screening but was interviewed on the basis of a diagnosis of depression made by our psychosocial counsellor. For FGDs we purposively sampled teachers, parents and adolescents across genders, caste/ethnicity and age group. Among health workers and the NGO worker we sampled a range of expertise. RAs conducted SSIs and FGDs in Nepali in a private space in a school, health post, or other community location.
We developed and piloted topic guides for SSIs and FGDs. The first part of the topic guides involved open questions and probes to explore participants' concepts and experiences of heartmind problems as a proxy for depression, including effects on feelings and behaviour, coping, and help-seeking. We asked generally about perceived triggers of heart-mind problems, then probed specifically for grief, disagreements, loneliness and shyness, and life changes. The second part of the topic guide included questions to understand participants' perceptions of a group psychological intervention for adolescents with depression.
King's College London  and Nepal Health Research Council (637/2018) provided ethical approval for the research. We explained the study (goals, reasons for doing the research) and distributed information sheets. We sought informed consent for all research participants as well as consent from a parent or caregiver for participants aged 17 and younger.

Data analysis
SSIs and FGDs were audio-recorded, transcribed immediately into Nepali and translated into English for analysis. The quality of the data was regularly checked by the site lead by reading the Nepali transcripts whilst listening to the audio files. A standardised Nepali glossary was used to J o u r n a l P r e -p r o o f translate mental health terms (Acharya et al., 2017). The Framework Method was used to manage and analyse the data (Gale et al., 2013). We developed an analytical framework of deductive codes to reduce and summarise the data. The framework codified key concepts from Klerman and Weissman's original interpersonal theory of depression, was developed through a review of the literature on interpersonal models of depression and refined and approved by an IPT master trainer (Table 1) (Klerman et al., 1984). Codes related to interpersonal problems of grief, dispute, role transition and social isolation conceptualised as depressive "triggers"; core principles of the interpersonal model (e.g. depression is a medical illness; depression is curable); and generalised techniques and specific strategies related to each of the interpersonal problems that are used in IPT to alleviate depression and improve interpersonal relations. Codes related to the same concept were grouped together. Codes and categories helped to abstract the data, moving from individual accounts to a more general impression of the data. The first and second authors used the analytical framework to code three transcripts in parallel to test and refine the framework and check consistency in application. We divided the remaining transcripts and independently applied the analytical framework to code the rest of the data. We charted the coded data into a matrix developed using Microsoft Excel, which helped us to identify patterns and connections between and within codes and categories and across participants. We wrote notes to document the analysis, explore codes and categories, help guard against selectivity in use of the data, and examine our own roles, possible biases, and potential influence on the research. The first author combined the matrices, identified patterns under each of the codes and returned to the transcripts as a whole to address any gaps in the findings related to context and process. Findings were checked for relevance and trustworthiness by an IPT master trainer, IPT trainers working in Sindhupalchowk, and researchers who conducted the SSIs and FGDs.
Quality of the data collection, analysis and reporting was done in accordance with the COREQ standards (Tong et al., 2007). Supplemental Table 1 includes the COREQ item checklist.

J o u r n a l P r e -p r o o f
The total sample comprised 126 participants representing the full range of caste/ethnic groups. Tables 2 to 4. Adolescents included in and out of school, and married and unmarried adolescents, from older and younger age groups.

Stressors and triggers
Participants used a variety of terms to describe symptoms of depression, including the English words tension and depression, man ko samasya (heart-mind problem), chinta (anxiety), manasik chintan (problem with mental thinking), aalchi (lazy), pagal (mad), rogi (frail). We asked depressed adolescents why they were experiencing these symptoms; interpersonal explanations were the first attribution for 19 out of 25 adolescents. Adolescents, parents, teachers and health workers attributed depressive symptoms to the death of a family member, having a poor financial situation, disability, love, quarrels in the family or with girlfriends or boyfriends, being beaten by family members, and being teased for belonging to a low caste. In 22 interviews with depressed adolescents the coding process revealed the presence of stressors related to the four interpersonal problem areas -dispute, grief, role transition and social isolation. In three interviews data were insufficient to identify a specific stressor or trigger. In five interviews coding revealed that adolescents linked their symptoms to more than one problem area. For example, a 14-year-old male linked his problems to a dispute with teachers involving physical abuse that forced him to leave his private boarding school and transfer to a local government school. He was finding the transition to a new school difficult in terms of adjusting to a lower quality of education and socialising with new classmates.

Disputes
Thirteen depressed adolescents linked their problem to a dispute. Of these, ten described a dispute with their parents. Adolescents attributed disputes to being given an unfair amount of household work, making mistakes with their work, and parents being overprotective and refusing to let them play away from home and spend time with their friends. Parents and teachers attributed disputes to parents putting too much pressure on their children to study, disapproving of their J o u r n a l P r e -p r o o f children having boyfriends or girlfriends from another caste/ethnic group, and refusing to buy their children clothes or the resources they needed for school.
Two adolescents described how disputes were exacerbated when their mother or father drank alcohol.

My father only scolds and shouts. He comes home drunk and beats [my] mother and when I go to
separate them he calls me things like "randi" [bitch/whore] and so on. I get angry thereafter and there will always be a quarrel. Female, aged 14 Disputes often involved physical abuse: adolescents described being beaten or hit with a piece of wood by their parents. Some adolescents stayed away from home temporarily or permanently to avoid the violence.
Adolescents also described having disputes with their siblings, friends, girlfriends and teachers. A 16-year-old female described being pagal (mad), having severe headaches, pain and suicidal ideation, which she attributed to a dispute with a friend. She first visited a dhami-jhankri (traditional healer) who told her that she was possessed by evil spirits. She tried worshipping to help her recover but when there was no relief her father took her to the local hospital, and later to a mental institute where she received medication and "got better".
Two adolescents linked their heart-mind problem to disputes that did not involve them directly. A 14-year-old female said that seeing her father beat up her mother made her angry, lose interest in food and stop talking with people. A 15-year-old male, who lived with his mother, brother and sister in rented accommodation, said that he was "tensed" because their landlord accused his mother of not paying rent and threatened to throw them out.

J o u r n a l P r e -p r o o f
Twelve adolescents with depressive symptoms described problems linked to a role transition.
Seven described having to assume more household and/or paid work because one or both of their parents were unwell, had migrated, moved out of the family home, or were unable to earn sufficient funds to meet the family's basic needs. Adolescents whose parents were unwell related their tension to concerns about their parents' survival and having to abandon their plans to study in order to earn money. They worried about how to pay for their parents' medical care, as well as clothes, food and school fees for their younger siblings. A 16-year-old female described how her father and mother were unwell and neither were able to work. She had taken on agricultural work to contribute to the family's income. She explained how she would have to give up her education in order to ensure her elder and younger brothers could continue theirs. Three adolescents described role transitions related to a physical disability or chronic health condition. They were struggling to come to terms with functional impairment that affected their friendships and study. For example, a 15-year-old female described the impact of losing the ability to walk at age seven: Previously our school was located up a hill and we had to climb […]  If we go with her we will not reach home until evening." When they said such things to me I used to get hurt. […] I used to think, when will I be able to walk like them? I can never be like them.
Parents and adolescents described early marriage as a role transition and as a potential cause of heart-mind problems among adolescents: If they get married at such a playful age of 15-19, before 20, not the age of caring or supporting the children. After the child is born, they have to provide food and clothes, educate them, and protect and support them. Due to not being the appropriate age for doing all these things they get more stress and hence they can commit suicide. Father J o u r n a l P r e -p r o o f

Social isolation
Two adolescents with depression linked their symptoms to social isolation. A 17-year-old male belonging to the Dalit caste group, described feeling "sad belonging to a low caste" and uncomfortable with adolescents from other castes. His sense of social isolation was exacerbated as a result of living away from his family in rented accommodation in order to attend school. A male aged 13 was "tensed" when his friends excluded him because of his skin condition: Other adolescents described situations of isolating themselves, for example staying alone, being absent from school, and not sharing their problems with family or friends, however from the interviews it was unclear whether depression caused the isolation or vice-versa.

Grief
I found a case [having heart-mind problems] who was a student. […] His mother had just committed suicide and due to his father's weak condition he was having a hard time studying. He used to manage money for his study by buying barley from the market and selling the alcohol that he made from it. After that I discussed with the other teachers and we gave him a scholarship […].
I told him that he should study and we would help him. Teacher Adolescents, parents and health workers identified the death of a family member, especially a parent, as a cause of depression. They described how bereaved adolescents faced financial hardship and were forced to discontinue their education and take on paid work and caring responsibilities, suggesting an overlap of grief and role transition problem areas.

J o u r n a l P r e -p r o o f
Only one of the adolescents interviewed, a 16-year-old female, described a personal experience of grief. Her brother died during the earthquake and she attributed the thoughts she had been having in her heart-mind to the fact that she was missing her brother. Teachers described loss of a parent as a cause of heart-mind problems among adolescents, and included the death, migration and separation of parents in their definition of "loss". The parallels between death and migration were evident in one 15-year-old girl's account of how she missed her mother who had migrated abroad when she was seven. Observing mothers and daughters in the community made her miss her mother even more.

Principles
IPT frames depression as a curable medical condition and not the person's fault. We found substantial evidence in support of a medical model of depression among adolescents who related the cause of depressive symptoms to "our mind or disease" and suggested visiting the doctor for treatment. They described physical symptoms of depression such as headaches and pain, and how emotions could have pathophysiological effects, for example anger affecting blood pressure.
A health worker suggested depressive symptoms could be inherited: What I found in two or three cases was [that the] mother of two daughters hanged herself due to "depression", a daughter hung herself at the age of 12, and the other daughter hung herself after one year. I did a post-mortem of both. After seeing the sequence I found that the symptoms are hereditary. A person can have a hereditary suicidal tendency for which we can do nothing.
In contrast, teachers and fathers described how some individuals perceived depression as a chhopne rog (illness category suggesting being grabbed or caught by a spirit) due to "witch or devil effect" that could be treated by dhami-jhankris.
Evidence that participants perceived depression as a curable condition was inconclusive. Teachers said that adolescents with depression were locally referred to as yo bigriyo, khattam nai bhayo or padhai khattam nai bhayo (meaning spoiled), as well as yo ta ava sakkiyo, siddhiyo (meaning J o u r n a l P r e -p r o o f finished), implying an irrevocable condition. However, they also felt that some underlying problems could be solved.
If the problem occurred due to poverty we cannot give them employment. We cannot possibly handle their financial problems in their houses. However, if the problem occurred due to lack of education and awareness, if they are ready for counselling, then [help is possible]. Teacher Parents, teachers and adolescents perceived depression as preventable by avoiding circumstances that might lead to stress and "tension" such as falling in love, or as controllable through demonstrating strength of character.

Mother
The notion of a reciprocal relationship between depression and interpersonal problems is central to IPT. As described above, adolescents, teachers, parents and health workers attributed depressive symptoms to grief, disputes, social isolation and role transitions. However, few described how depression exacerbated interpersonal conditions and when they did it was mainly related to the effects of symptoms on daily functioning.

Before I studied well and I used to be ranked first, second and third [in my class]. [Now] I feel like I don't have strength to think. […] I don't feel like doing anything and I also don't do anything I used to do before. I don't cook food or perform work like I used to do before. I only sleep in my room, don't speak to anyone and don't do homework. Female, aged 14
A 14-year-old male linked his heart-mind problem to being beaten and verbally abused by his teacher. Despite enrolling at a new school, he explained how whenever he felt "tensed and J o u r n a l P r e -p r o o f overburdened" he wanted to give up his studies. A 15-year-old female described how her parents wished she was a boy and at times this made her feel like leaving her studies or committing suicide. At other times she felt like studying hard or earning in order to prove herself.
IPT is predicated on attachment theory. The relevance of attachment problems was evident in adolescents' depression history.

My life has been miserable since early childhood. My parents used to drink a lot. They never cared how our lives would be. […] My mother used to beat me up while drunk and used to accuse me of having relations with my father. My father didn't hurt me as much as my mother did. Female, aged 18
Two adolescents recounted severe physical and verbal abuse perpetrated by their parents throughout their childhood. An 18-year-old female described seeking sanctuary with her grandparents and a 16-year-old male stayed with his uncle and aunt, however the extent to which these alternative caregivers were able to compensate for the lack of parental attachment was unclear.
Teachers and parents suggested that the absence of parents could lead to depression. Mothers described a unique openness between a mother and child, owing to the fact that the "mother gives birth", and highlighted the importance of this relationship in childhood. Adolescents described seeking support and reassurance from their mothers when under stress. A 16-year-old female who lost her eyesight and her brother described how her mother listened to and reassured her. She stated that her "life wouldn't matter" in the absence of her mother.
Mothers and teachers also described the importance of fathers.  […]. It seems like his mind has altered or something is wrong with his heart-mind. Mother Table 5 summarises evidence from interviews with adolescents concerning their use of techniques and strategies relevant to IPT to alleviate depressive symptoms. All but three adolescents linked their mood to an interpersonal event. Most identified a helpful or unhelpful relationship in the context of their depression (Figure 1). Adolescents described how siblings listened to their problems, gave advice, advocated to their parents, provided food and clothes, and sponsored their studies. They also described unhelpful relationships with fathers, mothers and teachers who scolded, cursed and physically abused them.

Techniques and strategies
Adolescents provided examples of how their mood had been affected by communicating with someone, for example dukhkha lagne (feeling sad) and rish uthne (feeling angry) when they were scolded, and man halungo hune (feeling lighter) and fresh hune (feeling fresher) when someone listened to their problems.
Twelve adolescents described how they had generated options to manage their life problems. Some had identified ways to manage their mood in the short term, including listening to music when they were tensed, or calling their siblings when they felt sad or suicidal. Others identified long-term approaches. A 13-year-old male worked as a labourer to help reduce his family's debt and a 15-year-old male worked to pay for his own health care.
We found no evidence that adolescents were given the sick role on account of their depressive symptoms. Many adolescents described inflexible daily routines: parents depended on their children's earnings and relied on them to do the household work and care for younger siblings.
Only one female participant who suffered with chronic breathing difficulties described how her sisters and mother took on her household responsibilities when she was not feeling well, however this was related to her respiratory rather than depressive symptoms.

J o u r n a l P r e -p r o o f
We searched for evidence that adolescents were implementing IPT strategies for grief, disputes, role transitions and social isolation. Among the two who linked their depressive symptoms to isolation, neither implied that they were trying to change their habits to promote social engagement. A 15-year-old female whose brother was killed in the earthquake did not describe using any IPT grief strategies. Instead she tried to manage her grief by staying alone in a peaceful place and sharing her thoughts: Among those who linked their depressive symptoms to a dispute, we found evidence that they had attempted to understand both sides of the argument. A 14-year-old male left school when he was severely beaten by his teacher but acknowledged that the teacher "might have had some problem with herself as well".
Identifying the stage of the dispute -negotiating, stuck, or wanting to end the relationship -is an IPT strategy to guide conflict resolution, and adolescents described disputes at each of these stages. An 18-year-old female described being stuck in a dispute with her husband: although she had moved out of his house due to physical abuse, she did not want to divorce him whilst her son was "too small".
Of the 12 adolescents who described role transitions, only one described mourning the loss of their old role. A 15-year-old female developed difficulty walking and her mother had migrated outside Nepal. The adolescent recalled how, before these events she played in school and visited the community. Afterwards, she found it difficult to keep up with her friends and felt overburdened by the household work she was expected to do in her mother's absence. None of the adolescents J o u r n a l P r e -p r o o f suggested they had weighed up the pros and cons of their new role, but three described how they had learned to manage it. For example, because both of her parents were ill, a 16-year-old female participant had assumed new financial and caring responsibilities which she managed by finding paid work and helping her parents to attend health facilities.

DISCUSSION
To the best of our knowledge, this study is the first to evaluate how depression could be conceptualised using the interpersonal model in a non-Western population. We found evidence that depression among adolescents in Nepal maps onto all four interpersonal problem areas of grief, dispute, role transition and social isolation. There was alignment with core IPT principles, for example some participants conceptualised depression as a medical illness. This appears to be a relatively new understanding of depression in Nepal, emerging from younger generations and the medical profession. Whilst participants intimated that depression was preventable, it was unclear whether they thought depression could be cured. We found evidence that adolescents were using some techniques and strategies common to IPT to improve depression (e.g. identifying helpful and unhelpful relationships and generating options and ways of managing problems). We found less evidence for the relevance of others, especially IPT strategies to address grief, role transitions and social isolation.
The triangulation of data from SSIs and FGDs and diversity of participants in terms of caste/ethnicity, age, role, and depressive status are strengths of this study. However, the study is limited in its focus on one district in Nepal. Data collection was not designed to explicitly elicit interpersonal models of depression. This was advantageous because participants were not prompted or influenced in describing their experiences of depression, however it also gave rise to gaps in our data. Although we probed about interpersonal triggers of depression we did not ask specifically about the use of IPT techniques and strategies and the extent to which they were or were not successful.

J o u r n a l P r e -p r o o f
Our exclusive focus on the interpersonal model of depression could be a limitation in light of research suggesting Nepali ethnopsychology may also be compatible with CBT and dialectical behaviour therapy (DBT) models (Kohrt et al., 2012). However, it is worth noting that interpersonal explanations were the first attributions for depression for the majority of adolescents. There was almost no mention of one's thought patterns as responsible for heart-mind problems. This is consistent with Nepali ethnopsychology in which emotions, experiences, and memories are generated in the heart-mind, then processed in the brain-mind where cognition occurs. Therefore, changing one's experiences and relationships to impact emotions and thinking may be more culturally consistent than a focus on changing thoughts as the first step in alleviating distress.
Regarding DBT, the focus on interpersonal relationships in that therapy may be why it has cultural appeal (Ramaiya et al., 2017). Other theories of depression are related to social comparison, social capital, and cultural consonance (Dengah, 2014;Gibbons, 1999;Song, 2015). These models are built upon subjective comparisons between one's status and others, or they examine how one's model of success or identity compares with others in the group. In adolescence the social comparison model may be particularly relevant because it is a crucial time for identity formation (Erikson, 1968). We found that adolescents in Nepal made many comparisons to others and related these to their depressive symptoms. For example, an adolescent from the Dalit caste group compared himself to peers from higher castes; an adolescent with a disability wished she was able to walk like her friends. Understanding whether the social comparison model of depression is a better fit than the interpersonal model could be the focus of future research.
We found no evidence that adolescents with depression were given the sick role. In settings where mental disorders are highly stigmatised, individuals labelled as depressed may be socially excluded leading to worsening of their symptoms. Furthermore, adolescents who self-label as depressed may experience higher levels of self-stigma and depression (Moses, 2009). According to the idiom of distress hypothesis, in collectivist societies such as Nepal somatisation can be a way to indirectly communicate distress and obtain support whilst preserving interpersonal relationships (Keyes & Ryff, 2003). Idioms of distress could therefore be an alternative to the sick role in adaptations of IPT in Nepal.

J o u r n a l P r e -p r o o f
Few adolescents reported using IPT strategies for grief, disputes, role transitions and social isolation. This could be because: we did not probe for these strategies and hence they were not reported; because adolescents were not trying to address their problems; or because adolescents were using strategies aligned with other psychological models. For example, one adolescent described managing her grief by sharing her thoughts and feelings with her mother in order to leave the past behind. This approach may fit with cognitive behavioural models where the patient creates and shares their trauma narrative with a parent and converts their relationship with the deceased from interaction to memory (Cohen et al., 2006). Some adolescents described managing their tension by being alone and peaceful. DBT, with its emphasis on distress tolerance and mindfulness, may therefore be relevant in this setting (Kohrt et al., 2012).
In Nepal migration is common and often involves separation of parents from their families for several years at a time. The mental health and developmental impacts of migration on left-behind children have been well documented, however long term effects on attachment are less clear (Fellmeth G & Rose-Clarke K et al., 2018). Migration was described by some participants in our study in terms of the "loss" of a parent. One adolescent with depression described feelings similar to grief when her mother migrated. In an adaptation of IPT for adult Hispanic immigrants in the US, migration was categorised as a role transition (Markowitz et al., 2009). Left-behind children may also experience a role transition involving changes in family structure due to the absence of a parent (Mufson et al., 2004a).
We only identified two adolescents who attributed their depressive symptoms to social isolation although it is possible that others experienced social isolation causing or owing to their depression.
In a context like Nepal (pre-COVID19) physical isolation is unlikely as adolescents often live in large extended families, however social and emotional isolation are major stressors. Disputes and behaviour within the family, interrupted education, and lack of opportunities to socialise outside the home could isolate adolescents from their peers at a time where peer to peer relationships play a crucial role in development (Blakemore, 2018). Furthermore, owing to discrimination on the basis J o u r n a l P r e -p r o o f of caste or ethnicity there are circumstances where families may be socially isolated within a community (Kiang et al., 2020).
Several adolescents described interpersonal problems that cut across different problem areas.
Many attributed their depressive symptoms to poverty which could be categorised under multiple problem areas. For example poverty could manifest as a role transition where an individual experiences the change from financially secure to insecure, or as a dispute where an individual compares their financial situation to others' (indirect dispute) or experiences discrimination related to their low financial status (direct dispute). Participants described poverty as the mediating factor through which the death of a parent can lead to depression. In a South African adaptation of IPT, stress related to not being able to provide basic necessities was perceived to be directly related to depression and poverty was framed as a separate problem area (Petersen et al., 2012). Poverty has alternatively been conceptualised as a risk factor for depression where the acute trigger is either a role transition or dispute ). An alternative model of depression among adolescents in Nepal might therefore situate interpersonal problems within a broader socioecological framework in which macro-level risk factors (poverty, economic migration, attitudes about interpersonal violence) shape interpersonal relationships.
Our findings highlight the importance of moving beyond simply categorising idiom differences in symptoms and description, to exploring how different theoretical models of depression fit with local experiences. This will help to inform cultural adaptations of successful interventions that go beyond contextualisation to include deeper changes to core content, increasing the likelihood of fit . We identified several co-existing explanatory models of depression including conceptualisation aligned with a medical model among adolescents and health workers, and more spiritual models among their parents and teachers. Future adaptations of IPT must be relevant to these models and ensure families, teachers and adolescents have a shared understanding of the goals and methods of IPT to reduce barriers to treatment. We suggest emphasising psychoeducation to raise awareness that depression is curable, which could help to give hope and reduce potential stigma towards those participating in IPT. We also recommend ethnographic J o u r n a l P r e -p r o o f research to better understand the social context in which some IPT strategies and techniques are used to reduce depression, whilst others are not. Group-based rather than individual IPT, where group members suggest approaches to address interpersonal problems, will help to ensure therapy is locally relevant and acceptable. Disputes interlinked with physical violence and child maltreatment were common. Adaptations of IPT must therefore make safety management an integral part of therapy. Narratives of poverty, and parental alcohol abuse and mental ill health were also common and highlight the need to integrate IPT into a broader package of care that links with child protection, educational support and social security schemes, and ensures parallel capacity building in adult mental health care.
In conclusion, we describe the first cross-cultural interpersonal model of depression among adolescents in Nepal. We demonstrate the relevance of interpersonal problems and provide evidence for alignment with core IPT principles. Our findings inform future cultural adaptations of IPT in this setting and beyond, including the need to integrate psychoeducation and local approaches to managing grief, disputes, role transitions and social isolation, and to link with health and social services. Our findings also highlight the importance of moving beyond categorisation and description of idioms of distress to explore how different theoretical models of depression fit with local experiences.

Data sharing
Due to the sensitive nature of the topic of this study and difficulties fully anonymising the transcripts data are not suitable for sharing.

Role of the funding source
The study was co-funded by the UK Medical Research Council, the National Institute for Health Research and the Department for International Development (study reference: MR/R020434/1). The funders had no role in the collection, analysis and interpretation of data; in the writing of the articles; and in the decision to submit it for publication.

Author contributions
BK and KR-C conceptualised the study. All authors were involved in developing the analytical framework. PBK and JM supervised and participated in data collection. KR-C and EH conducted the analysis. KR-C wrote the first draft of the paper. All authors reviewed the paper and approved the final version for submission.
J o u r n a l P r e -p r o o f Reducing the burden of current expectations on a depressed patient Generating options and ways of managing life problems Finding options to address problems will help the person to feel less hopeless and helpless.

Grief strategies
Becoming stronger in order to carry grief Finding options to address problems will help the patient to feel less hopeless and helpless. Linking relationship and circumstances to the grieving process Understanding how the nature of the patient's relationship with the deceased, and the circumstances surrounding death, can affect the grieving process Scheduling time to mourn Setting aside regular time to mourn in order to remain functional at other times Dispute strategies Understanding both sides of the argument Trying to understand both sides of a disagreement Identifying the phase of the dispute Analysing the phase of the dispute (whether both parties are still negotiating, or are stuck in the dispute, or whether one or both parties want to end the relationship) Role transition strategies