BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 25, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Exploring the concept of Problematic khat use in the Gurage community, South-central Ethiopia: A qualitative study

Journal: BMJ Open ManuscriptFor ID peerbmjopen-2020-037907 review only Article Type: Original research

Date Submitted by the 27-Feb-2020 Author:

Complete List of Authors: Awoke, Mihretu; Addis Ababa University College of Health Sciences, Psychiatry; college of natural and social sciences, Addis Ababa science and Technology University , Department of social sciences Fekadu, Abebaw; Centre for Innovative Drug Development and Therapeutic Trials in Africa (CDT-Africa); Department of Psychiatry, School of Medicine, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia Habtamu, Kassahun; School of Psychology, College of Education and Behavioral Studies, Addis Ababa University, Addis Ababa, Ethiopia NHUNZVI, CLEMENT; University of Zimbabwe, Rehabilitation Norton, Sam; King's College London, Psychology Department, Insitute of Psychiatry Teferra, Solomon; Addis Ababa University College of Health Sciences,

School of Medicine, Department of Psychiatry; Harvard T.H. Chan School http://bmjopen.bmj.com/ of Public Health, Boston, USA

MENTAL HEALTH, Substance misuse < PSYCHIATRY, QUALITATIVE Keywords: RESEARCH

on September 25, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 39 BMJ Open

1 2

3 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from 4 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on September 25, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

http://bmjopen.bmj.com/ 33 34 35 1 Department of Psychiatry, School of Medicine, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia 36 37 2 Department of social sciences, college of natural and social sciences, Addis Ababa science and Technology University 38 39 3Centre for Innovative Drug Development and Therapeutic Trials for Africa (CDT Africa), College of Health Sciences, Addis Ababa University, 40 41 Addis Ababa, Ethiopia on September 25, 2021 by guest. Protected copyright. 42 43 4School of Psychology, College of Education and Behavioral Studies, Addis Ababa University, Addis Ababa, Ethiopia 44 45 5Department of Rehabilitation, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe 46 47 6Psychology Department, Institute of Psychiatry, King’s College London, London, UK 48 49 7Harvard T.H. Chan School of Public Health, Boston, USA 50 51 52 53 54 55 56 Abstract 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Objective: This study aimed at exploring how problematic khat use is characterized in the 4 5 6 Gurage community, South-central Ethiopia. 7 8 Design: qualitative study. 9 10 Setting: Gurage community, South-central Ethiopia 11 12 Participants: We conducted in-depth interviews with 14 khat users and 5 non-khat users, and 3 13 14 15 focus-group discussions with khat users. 16 For peer review only 17 Methods: All participants were selected purposively based on their exposure to khat or khat use. 18 19 We used interview guide to explore perception of participants about khat use and problematic 20 21 22 khat use. We analyzed the data thematically using Open Code software version 4.03. We used 23 24 iterative data collection and analysis, triangulation of methods and respondent validation to 25 26 ensure scientific rigour. 27 28 29 Findings: We identified three major themes: sociocultural khat use, khat suse (khat addiction), 30 31 and negative consequences of khat use. Sociocultural khat use included a broad range of contexts 32 33 and patterns including use of khat for functional, social, cultural and religious reasons. Khat http://bmjopen.bmj.com/ 34 35 addiction was mainly explained in terms of associated khat withdrawal experiences, including 36 37 38 harara/craving, and inability to quit. We identified mental health, sexual life, physical health, 39 40 social and financial related negative consequences of khat use. The local idiom Jezba was used 41 on September 25, 2021 by guest. Protected copyright. 42 to label subgroup of individuals with khat suse (khat addiction). 43 44 45 Conclusion: The study has identified what constitutes normative and problematic khat use in the 46 47 Gurage community in South-central Ethiopia. Problematic khat use is broad concepts which 48 49 include frequency, reasons, contexts, negative consequences and addiction of khat use. Insights 50 51 52 generated can be used to inform future studies on development of tools to measure problematic 53 54 khat use. 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Keywords: khat use; normative khat use; problematic khat use; qualitative study; Ethiopia 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Introduction 4 5 6 Khat, the psychoactive stimulant, is very common and its use is growing in East Africa and the 7 8 9 Arabian Peninsula[1]. Khat use has also been reported in Western countries, such as the United 10 11 Kingdom [2] and Australia [3], primarily among Ethiopian, Somalian, Yemeni and Kenyan 12 13 diaspora communities [4]. The prevalence of use was estimated to be 67.9% in Yemen, 59% in 14 15 Somalia [5, 6], and 16%-50% in Ethiopia [7-9]. Many people chew khat for its stimulating 16 For peer review only 17 18 effects, to gain concentration and energy during work [10-12]. 19 20 21 Khat use has been embedded in the culture and social life of East Africa and the Arabian 22 23 Peninsula especially among Muslim- dominant societies [13, 14]. Some studies have reported the 24 25 association between khat use and being Muslim [9, 15], but the position of the Islam religion 26 27 28 regarding khat use remains unclear [16]. In the literature, especially from Yemen and Saudi 29 30 Arabia, the following three outstanding themes of discourse have been raised with regard to khat 31 32 use: halal (permissible), makruh (disliked or discouraged) or haram (forbidden). For example, in 33 http://bmjopen.bmj.com/ 34 35 Ethiopia, many Muslims chew khat when they go to pilgrimage centers and while doing rituals 36 37 such as singing, prayer-du'a, blessing and other activities [17]. Muslim women also use khat 38 39 when they are gathered for prayers directed on women in labor: a social ritual called Fatimaye, 40 on September 25, 2021 by guest. Protected copyright. 41 invoking the name of Fatima, the daughter of Prophet Muhammad [18]. The perceived social 42 43 44 uses of khat were mainly for social gathering like weddings and funerals [19]. In the Yemeni 45 46 society, khat use was associated with important social occasions to meet other people and 47 48 exchange of ideas [20]. 49 50 51 Although it has important cultural routes and functions, khat use has significant health harms. 52 53 54 The physical health problems include increased body temperature, loss of appetite, gastritis, 55 56 haemorrhoids, insomnia, and oral health problems [21-24], and even hypertension [25] and other 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 cardiovascular dysfunctions [20, 26, 27]. The mental health impacts include psychotic 4 5 6 symptoms [28-30] and depressive symptoms [31]. However, the mental health impacts remain 7 8 inconclusive because other studies have revealed negative findings [1, 32]. 9 10 11 Across the literature, conceptualization of problematic, sociocultural and recreational khat use 12 13 has been an important research gap. Problematic khat use, rather than khat use per se, is usually 14 15 the interest of the public, researchers and policymakers. Nevertheless, only few previous studies 16 For peer review only 17 18 were conducted on problematic khat use [33]. The inconclusive reports about the different harms of 19 20 khat use could also be due to poor definition of problematic khat use. 21 22 23 The Diagnostic Statistical Manuel(DSM-5) definition of stimulant use disorders [34] could have 24 25 more important clinical utility (for severe cases) than screening individuals with problematic 26 27 khat use at earlier stage. Lack of screening tool for problematic khat use, especially among 28 29 30 nonclinical cases could hamper efforts to curb the problem including early identification and effective 31 32 management of positive cases. Therefore, there is a need for valid problematic khat use screening 33 http://bmjopen.bmj.com/ 34 tools which would facilitate clinical care in primary health care settings and for future research. 35 36 37 Although there are no strong validation studies, few studies used Harmful Khat Use Scale [35] 38 39 and Severity of Dependence Scale to measure the construct problematic khat use [36], but a 40 41 systematic review [33] and exploratory studies[37] suggested broader indicators of problematic on September 25, 2021 by guest. Protected copyright. 42 43 44 khat use, including amount, frequency, context and duration of khat session. Therefore, the aim 45 46 of this study was to conceptualize problematic khat use from the perspective of users and non- 47 48 users in a dominantly rural setting, Gurage, south-central Ethiopia. 49 50 51 52 53 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Methods 4 5 6 Study setting 7 8 9 The study was conducted in the Gurage zone; Southern Nations, Nationalities and Peoples 10 11 12 Region (SNNPR), Ethiopia. Ethiopian Orthodox Christianity (48.17%) and Muslim (42.31%) are 13 14 the two dominant religions in Wolketie town, Gurage’s capital [38]. Peasant farming is the main 15 16 productive occupationFor in rural peer areas while reviewpetty trading is more only common in urban areas. The area 17 18 is known for its khat production and khat use [10]. Khat might have been introduced to Gurage 19 20 21 area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 1543) army or neighboring Muslim 22 23 Wolane or Oromo ethnic groups [39]. 24 25 26 Study design 27 28 29 The study employed a qualitative study design [40] which allowed for the understanding and 30 31 description of the experiences and perspectives of people towards problematic khat use. This 32 33 http://bmjopen.bmj.com/ 34 study was guided by the standard of reporting qualitative research (SRQR) and consolidated 35 36 criteria for reporting qualitative research (COREQ)[40]. 37 38 39 Study participants and recruitment procedures 40 41 on September 25, 2021 by guest. Protected copyright. 42 We conducted face to face in-depth interviews with 14 current khat users and five non-khat 43 44 users. Twenty-one khat users participated in focus group discussions (FGD). The first FGD 45 46 consisted of six women, the second and the third FGDs had seven and eight participants 47 48 49 respectively. Participants were selected purposively based on their experience of khat use, and 50 51 we also aimed for maximum variation considering the socio-demographic characteristics of the 52 53 participants. Participants were invited by the community health workers and the first author did 54 55 56 the informed consent. All the interviews and FGDs were conducted in Amharic and tape- 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 recorded. The first author, who has good experience of interviewing khat users, did the 4 5 6 interviews and facilitated the focus group discussions assisted by a trained moderator. The 7 8 interviews took about 40 minutes and the FGDs about one hour in average. The first and the last 9 10 authors designed in-depth interview and FGD guides. The guides mainly focused on the 11 12 13 experiences and perceptions of the participants regarding khat use and problematic khat use. 14 15 Data management and analysis 16 For peer review only 17 18 The data were transcribed verbatim in Amharic, then translated into English by the first author 19 20 and experienced research assistants. All interview and FGD translations were coded 21 22 independently by the first and fourth author (AM and CN). Iterative thematic analysis [41] was 23 24 25 done simultaneously with data collection. We used computer software, open code 4.03 to 26 27 manage and analyse the data [42]. 28 29 30 Regarding to data quality management and rigour, the iterative process of data collection, data 31 32 analysis and checking unclear issues from the participants added to the quality of the study. We 33 http://bmjopen.bmj.com/ 34 35 did also data triangulation from different sources, including religious fathers, key informants for 36 37 the culture and legal officers. Participants were also diverse in terms of socio-demographic 38 39 characteristics. The field worker, the first author, has previous experience interviewing people 40 on September 25, 2021 by guest. Protected copyright. 41 about khat use. Above all, we tried to purely and openly present the ideas of the participants 42 43 44 without personal impression interference. 45 46 47 Ethical considerations 48 49 50 Ethical clearance was obtained from the Institutional Review Board (IRB) of the College of 51 52 Health Sciences, Addis Ababa University (Ref 008/18/Psy). In addition, we also obtained a letter 53 54 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 of support from the Gurage zone health department. Written informed consent was sought and 4 5 6 obtained from all participants before data collection. 7 8 9 Patient and public involvement 10 No patient involved. 11 12 13 Findings 14 15 16 Characteristics of participantsFor peer review only 17 18 19 Table 1 Socio-demographic characteristics of participants 20 21 Characteristics Interview FGD 22 23 Age 3 8 24 25 20-34 26 35-44 5 2 27 45-59 6 8 28 60 and above 5 3 29 Gender 30 31 Male 16 15 32 Female 3 6 33 Residence http://bmjopen.bmj.com/ 34 Urban 12 7 35 Rural 7 14 36 Marital status 37 38 Single 1 4 39 Married 17 15 40 Widowed or divorced 1 2 41 Education on September 25, 2021 by guest. Protected copyright. 42 Can’t read and write 3 2 43 44 Read and write only 1 2 45 Primary 2 5 46 Secondary 2 7 47 Post-secondary 11 5 48 Religion 49 Muslim 8 9 50 51 Christian 11 12 52 Employment status 53 Self-employed 6 8 54 Unemployed 4 8 55 Formally employed 8 2 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Student 1 3 4 5 Khat use status 6 Khat users 14 18 7 Non-khat users 5 3 8 9 10 Participants were diversified in several socio-demographic characteristics, including sex, age, 11 12 residence, religion, occupation and educational status (Table 1). The major themes that emerged 13 14 from the iterative thematic analysis were: (1) sociocultural khat use, (2) suse “(addictive)” khat 15 16 For peer review only 17 use, and (3) negative consequences of khat use. The second theme had two categories: 18 19 withdrawal experiences, quitting khat use and mirqanna/feeling high after khat use. The negative 20 21 consequences were categorized to mental health, sexual life, physical health, social and financial. 22 23 24 These themes are discussed with support of quotes directly from the participants. 25 26 27 Sociocultural khat use 28 29 30 The sociocultural khat use theme emerged from participants’ notions of using khat for spiritual, 31 32 social and cultural reasons. Among Muslim participants, khat use was very common and 33 http://bmjopen.bmj.com/ 34 acceptable during religious rituals including the fasting month-remedan, holiday celebration and 35 36 anniversaries for the shrine of religious forefathers and at pilgrimage center. During these 37 38 39 situations, khat was one of the presents-zihara. 40 41 on September 25, 2021 by guest. Protected copyright. 42 Many Muslim participants, especially elderly men, reported chewing khat for prayer. One 43 44 participant who was an elderly Muslim man described his use of khat as “… a prayer instrument 45 46 (duw’a)”. Some Muslim farmers who participated in this study reported chewing khat during 47 48 49 prayer as an inherited religious virtue-Ibada. Other participants described that the religious norm 50 51 of khat use for prayer was acceptable if it was in group chewing than chewing alone. A 52 53 participant making reference to the sword as a symbol for prayer alongside chewing khat shared 54 55 56 that: 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 “One sword can’t kill rather it only stabs. Thus, we usually chew being three or more” (Age late 4 5 6 50’s, male, urban) 7 8 9 In rural areas, unlike urban areas, participants shared that they chew khat and do religious 10 11 rituals in mosques on weekly venerations of saints-wali. For example, Tuesdays had special 12 13 meaning because they were a memorial for Nurahusene, a religious forefather. Gathering in 14 15 mosques on special days of the week to chew khat and pray was part of the religious practice. 16 For peer review only 17 18 “Although khat chewing doesn’t have any religious basis, we [Muslims] chew khat on the days 19 20 21 which are the shrine of our religious forefathers such as Esnel (Monday), Megergbia (Tuesday), 22 23 Gelale (Wednesday) and Sehiare (Saturday).” (Muslim religion scholar, age late 50’s) 24 25 26 Ceremonial khat use was also viewed as common practice at pilgrimage centres and during 27 28 annual Muslim religious holidays; Arafa (i.e. Id al-Adha), Id alfeter and moulid. In the study 29 30 31 setting, Aberiat, Qatbarie (Shaykh Isa Hamza), Alkeso, and Zebimola were frequently mentioned 32 33 as the most colourful wali venerating practices. Pilgrims had a religious ritual called hadra. It is http://bmjopen.bmj.com/ 34 35 a ritual by small groups of pilgrims for religious ritual and khat chewing session. They did this 36 37 being in a separate partition in the mosque compound. 38 39 40 Khat use was not seen as limited to subgroups like the elderly and Muslim men, rather it was part 41 on September 25, 2021 by guest. Protected copyright. 42 of the life of many people from different walks of life in the Gurage area. Many participants 43 44 agreed that, through time, khat was becoming acceptable and meaningful in the broader social 45 46 47 and cultural events. The participants agreed that, khat use was an important part of the Gurage 48 49 culture. This study indicated that khat chewing was shaping the day to day activities, beliefs, 50 51 tradition, and rituals of many people. Khat users, especially in rural areas, reported that khat 52 53 54 chewing was a significant part of their daily life. They frequently said that they used khat daily to 55 56 express both their mourning and happiness. 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Participants reported that khat use was acceptable and to chew khat was to conform to societal 4 5 6 norms and values but there were specific nuances in how this was done to remain acceptable. 7 8 This was mainly not to have a socially agreed problems resulting from one’s chew of khat. The 9 10 normal and accepted khat user was portrayed as “…a humble, honest, generous, and positive 11 12 thinker.” (Age in 40’s, rural, male) 13 14 15 Other definitions of normal khat use were related to patterns of use. Participants had set criteria 16 For peer review only 17 18 for who, when, how much and where to chew. For example, participants said it was part of their 19 20 culture not to sanitize khat use in the morning, but recognised khat use in the afternoon. 21 22 23 Many participants did not accept daily khat use as normal. Normal khat users were also 24 25 supposed to limit the amount of their khat consumption. Chewing in public places was also 26 27 28 acceptable and normative among men but considered abnormal among women. 29 30 31 This was shared by one participant in the focus group discussions and they highlighted: “there 32 33 are males who chew khat in cafeterias and verandas, but a woman does not chew in public http://bmjopen.bmj.com/ 34 35 places even we don’t buy and hold khat, except commercial sex workers.” (Age early 30’s, 36 37 female, urban) 38 39 40

We found much restriction of the norms among rural than urban participants. Participants from on September 25, 2021 by guest. Protected copyright. 41 42 43 rural areas emphasized that as areqi (a local illicit alcoholic drink) have been part and parcel of 44 45 the life of many Christians and the same was true for khat among Muslims. But now, khat use 46 47 had become a normal behaviour of both Muslims and Christians in both urban and rural areas. 48 49 50 Use was considered normal as long as it did not come with other problems. 51 52 53 Many participants also shared that sociocultural khat use was an accepted way of use valued in 54 55 their community. Examples of sociocultural uses that emerged from the analysis included: 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 hosting a visitor, grieving period after a funeral, work parties, as part of normal socialisation, 4 5 6 weddings and in rare cases when paying a visit to the sick. A participant highlighted as: “During 7 8 condolence after funeral, wedding ceremony…when we chew khat, it will be very easy to 9 10 socialize and easily close with others. In our home, we feel confident to host guest if we can buy 11 12 khat otherwise I don’t say it will be full hospitality.” (Age early 30’s, male, urban) 13 14 15 Despite the normative khat use reported, the participants also shared insights around what 16 For peer review only 17 constituted problematic khat use. 18 19 20 21 22 Khat suse-addiction 23 24 25 Most participants conclusively stated that khat use is an addictive behavior-suse. This was the 26 27 most commonly reported reason to continue chewing by many participants so that they could 28 29 30 accomplish their day to day activities. Different withdrawal experiences (presented in the next 31 32 section) were reported to justify the suse or addiction to khat use. At the addictive stage, those 33 http://bmjopen.bmj.com/ 34 affected are seen by keeping some khat in their pockets and chew almost all the time without 35 36 regard for place too. This included chewing at school, funeral, market, and workplace either 37 38 39 office or farm. A participant explained khat suse/addiction as; 40 41 on September 25, 2021 by guest. Protected copyright. 42 I should have khat in my bloodstream to open my eyes and to activate my body. Morning khat 43 44 use-ejebena is a sign of khat suse, isn’t it? (Age in 40’s, urban, male) 45 46 47 It was then this type of use that was viewed as causing problems to the health and social aspects 48 49 of both the user and those around him/her. 50 51 52 Participants also mentioned harara/craving as indicator of khat suse/addiction. Many said that 53 54 55 khat use is an addictive behaviour because it has a strong harara/craving. They justified that the 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 strong harara/craving even forced people to beg for money to buy khat when they did not have 4 5 6 money or they would collect the leftover khat and chew. When these behaviours start to show, 7 8 the khat chewing is then considered abnormal and problematic. 9 10 11 Those who chew because of harara/craving, unlike those who chew for prayer, chew khat 12 13 regularly, any time including in the morning, chew without ritual or ceremony and without 14 15 washing themselves. All these patterns of use are abnormal in my culture. (Age late 50’s, male, 16 For peer review only 17 18 urban) 19 20 21 Chewing khat by alms was perceived as the last and worst stage of a khat user who was called 22 23 Jezba. 24 25 26 Participants used the term Jezba to label subgroups of people with khat suse or addiction. 27 28 Jezbas are people with khat suse or addiction and characterized syndrome of the behavior. Jezba 29 30 31 syndrome indicates; poor connection with God, others and material needs, not confirming to 32 33 shared values and norms of the society. Jezba khat users prioritize khat use rather than other http://bmjopen.bmj.com/ 34 35 important aspects of life or they have limited interest for other important areas of life except 36 37 khat. They also had reduced motivation for work and other social life or total disengagement 38 39 40 when the khat is withdrawn as well as frequent work absenteeism or abandoning work. It also 41 on September 25, 2021 by guest. Protected copyright. 42 includes; deterioration in critical thinking, poor social skills/self-care, beg khat or money for 43 44 khat, external attribution, belief of inability or no interest to control khat use and related 45 46 47 behaviors, chewing for many hours of the day (more than 6 hours) and on the street or while 48 49 walking. 50 51 52 In the long term, khat will make you Jezba-ያጀዝብሀል. Jezba doesn’t value anything important 53 54 other than khat. (Age in 40's, rural, male) 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 For many participants, the use of khat against sociocultural conditions was considered as a sign 4 5 6 of khat suse/addiction. This includes chewing khat for individual reasons such as to cope with 7 8 adverse life experiences while cultural khat users chew for communal reasons; for prayer, 9 10 socialization or companionship, and functional reasons. 11 12 13 Khat addiction was also characterized in terms of chewing increased amount of khat through 14 15 time as well as chewing for longer time to get the desired effect. 16 For peer review only 17 18 When you chew too much [greater than a bundle of khat], it will cause many problems; being 19 20 21 addictive could be one. Only limited amount of khat benefits [the stimulating effect]. (Age late 22 23 50’s, male, rural) 24 25 26 In other views, some participants said that increased use of khat overtime depended on one's 27 28 mindset or psychological expectation, amount of budgeted time for chewing, emotional state or 29 30 31 group cohesion during chewing and financial capacity to buy khat. Age was also mentioned as an 32 33 important factor for heavy and problematic use of khat. Youths and early adult participants http://bmjopen.bmj.com/ 34 35 showed more addictive use of khat, unlike the elderly. 36 37 38 Withdrawal experiences of khat use 39 40 41 Other experiences reported associated with abnormal khat use included withdrawal experiences on September 25, 2021 by guest. Protected copyright. 42 43 when one was weaning off khat after heavy usage. The study found many important 44 45 46 psychological withdrawal experiences of khat suse-addiction. Feeling depressed, irritable and 47 48 aggression were the most typical and commonly reported experiences. One participant said, “I 49 50 am usually against people’s communication even aggressive when I didn’t chew khat. I 51 52 53 remember that I once threw away [smacked] my kid when she was talking about her school 54 55 affairs.” (Age late 50’s, urban, male) 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Different participants reported different withdrawal symptoms, including lack of motivation, 4 5 6 unable or lack interest to function or socialize, poor concentration and learning, unable to receive 7 8 message, and dukak (vivid unpleasant dream). Here are two dukak or vivid unpleasant dream 9 10 experiences; 11 12 13 One day my husband skipped chewing khat and went to bed. Then, he spent the night spitting. In 14 15 the morning, the bed sheet was wet. When I ask him what was wrong with him. He said; “people 16 For peer review only 17 18 were punishing me with the smoke of red paper and I had been feeling burning sensation for the 19 20 whole night”. (Age early 30’s, urban, women) 21 22 23 Another experience: 24 25 26 My experience was...ehm... usually a man would hold my hair and hung me or put me into a hole 27 28 and I would wake up in panic. (Age in 30’s, male, urban) 29 30 31 Some participants reported physical withdrawal experiences, including abdominal pain, 32 33 http://bmjopen.bmj.com/ 34 headache, being drowsy, red eyes, increased appetite and sleep, yawning, uncontrollable tears, 35 36 shivering hands, and loss of energy. Another interesting finding was an observation that to 37 38 manage some withdrawal syndromes, some khat users opted to sleeping. During fasting season, 39 40

many Muslim participants did not use khat either in the morning or for the whole day, but they on September 25, 2021 by guest. Protected copyright. 41 42 43 would usually spent the day sleeping. A Muslim participant shared his experience and said; 44 45 46 During remedan [fasting month], we don’t chew khat during the day time, but no one did a 47 48 serious work. We spend the day sleeping. If you don’t chew khat, you cannot be stimulated, 49 50 energetic..... (Age late 50’s, male, rural) 51 52 53 54 55 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Inability to control khat use 4 5 6 Quitting khat use was usually contingent on the participants’ lifestyle and pattern of khat use. For 7 8 9 participants who didn’t chew daily, quitting was perceived as an easy task, but it was very 10 11 challenging to participants without job because khat use would be an important leisure activity. 12 13 External factors; culture and pressure from others, were also reported challenges to quit from 14 15 chewing. Some thought that quitting khat use also includes quitting important social networks. 16 For peer review only 17 18 Thus, they continue chewing and consider khat use as a matter of surviving in the social system. 19 20 21 Some participants reported as they forward a rational decision to continue chewing after 22 23 evaluating the benefits and harms of khat use. Participants who perceive more benefits, than 24 25 harms from khat use, such as positive effect on motivation, work performance and socialization 26 27 28 continue chewing khat. 29 30 31 A participant stated as follows; 32 33 http://bmjopen.bmj.com/ 34 I will not quit chewing because it helps me to share information, got social support and other 35 36 things as well as to relieve stress and worries in life. (Age in 30’s, urban, male) 37 38 39 Many participants said that only a few and the fortunate could quit early, but many realize the 40 41 harms later. Many quit chewing when they lose satisfaction from chewing, usually at the end of on September 25, 2021 by guest. Protected copyright. 42 43 their life. Hence chewing until the end of life was considered an abnormal or addictive practice 44 45 46 of khat use. 47 48 49 After years, one would lose satisfaction from chewing. You would lose the passion to chew khat 50 51 and you will decide to say it is time to quit. Only the unfortunate will chew to the end of their life. 52 53 (Age late 50’s, rural, male) 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Mirqanna/feeling high after khat use 4 5 6 Many of the khat user participants liked the stimulant effect of khat, but when the stimulation 7 8 9 was too high, they reported feelings of distress. Mirqanna (induced distress) was due to chewing 10 11 excessive amounts or chewing in combination with shisha or cigarettes. This was repeatedly 12 13 reported among participants in their early phases of khat use. Insomnia induced by mirqanna was 14 15 also a perceived cause of distress. Many khat users would then resort to drinking alcohol to break 16 For peer review only 17 18 the mirqanna and induce sleep. Other khat users, experiencing mirqanna but cannot afford to 19 20 drink alcohol because of lack of money, reported continued feelings of being restless, being on 21 22 the move and confusion during. One such participant shared; 23 24 25 One day, I chewed khat for five hours. Then at night, I couldn’t sleep. I had spent the night 26 27 28 itching, feeling fever and sweating. I was very panic about my condition, but I was also feeling 29 30 fatigue and low energy to treat myself. (Age in 30’s, urban, male) 31 32 33 Additional symptoms of mirqanna reported by participants included; being absorbed to an http://bmjopen.bmj.com/ 34 35 inauthentic personal world such as considering oneself as fortunate, being extremely humble, and 36 37 considering oneself as high achiever, more like delusions. In some participants, they reported 38 39 40 lack of interest to communicate during mirqanna state, though talkativeness was also reported 41 on September 25, 2021 by guest. Protected copyright. 42 among others. Excessive fear including avoiding any exposure or engagement with others as well 43 44 as fear to make decisions was also common. Participants emphasized that the mirqanna 45 46 47 experiences were different from their experiences of intoxication due to excessive alcohol 48 49 drinking. 50 51 52 When it [mirqanna] is severe, when one chews too much, one might spend the night outside the 53 54 house, on the street, which is risky because wild animals could harm him. Sometimes alcohol 55 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 intoxication might be better than khat mirqanna. The intoxication of alcohol can be reversed 4 5 6 soon with different techniques including getting sleep, but the khat mirqanna couldn't be 7 8 reversed easily once the person is at the severe stage. (Age late 50’s, rural, male) 9 10 11 Other participants thought that severe mirqanna is like psychosis state-qezete. Mumbling alone, 12 13 confusion, spending the night on the street were perceived symptoms of psychosis induced by 14 15 khat. The following quote was from the experience of one participant. 16 For peer review only 17 18 When I am at mirqanna state, I imagine as I own a big building, big car…By the next day, I 19 20 21 realize everything was a fantasy. All were gone as cloud. (Age late 50’s, urban, male) 22 23 24 Negative consequences of khat use 25 26 27 Negative consequences of khat use that participants reported included mental health, physical 28 29 health, social and financial adverse effects. 30 31 32 Mental health related consequences 33 http://bmjopen.bmj.com/ 34 35 Participants reported that khat could lead to mental health disorders. Depression-debert and 36 37 psychosis-qezete were commonly reported mental health disorders. Participants indicated that 38 39 40 depression is associated with khat withdrawal or secondary to different crisis especially 41 on September 25, 2021 by guest. Protected copyright. 42 financial, caused by khat use. Psychosis which is known by different phrases such as qezete (the 43 44 acute form of psychosis), chereken metal, aemeron mesat, yeaemero menawet were common 45 46 among participants who chew too much amount of khat and don’t have meal before chewing. A 47 48 49 participant described this as follows; 50 51 52 …only individuals who had nutritional capacity- body fluid can resist the adverse effect of the 53 54 khat. The brain will be vulnerable when one chews without having a meal. People who chew 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 without having a protective, food, will be negatively affected by the khat. (Age in 30’s, rural, 4 5 6 male) 7 8 9 Participants reported that different behavioral symptoms of chronic khat users could be similar 10 11 with the behaviors of people with severe mental illness. They further described that among 12 13 chronic khat users, like people with severe mental disorders, there was broken down of family 14 15 and other social activities including poor self-care and dressing. In addition, they chew khat on 16 For peer review only 17 18 the street and they also chew leftover khat or beg money for khat. 19 20 21 Social related consequences of khat use 22 23 24 Another important problem of khat use was its social harms such as family chaos and 25 26 breakdowns. This was related with spending too much of the family budget on khat and 27 28 abandoning their responsibilities in the family. Negative behaviors of participants with khat use 29 30 31 such as irritability also led to family conflict. Two participants’ experiences are stated as follows; 32 33 http://bmjopen.bmj.com/ 34 I am not giving time for my kids. I just delegate my elderly daughter to take care and control her 35 36 siblings. (Age in 40’s, urban, male) 37 38 39 I have been forced by my wife to stop chewing. I don’t want to quit, but my wife is insisting me. 40 41 We have been quarrelling and separated due to this issue. Her parents came to mediate us and on September 25, 2021 by guest. Protected copyright. 42 43 asked me to quit chewing, but I told them my position that I got the khat before her and now I am 44 45 46 not interested to stop chewing…….I will not stop. Imagine? (Age in 30’s, urban, male) 47 48 49 Among urban dwellers, problem to discharge their government responsibilities were commonly 50 51 reported. Some reported their experience of being blamed by frequent work absenteeism and 52 53 work inefficiency because they lack the motivation to go to work when they don’t chew khat. 54 55 56 They usually go for khat abandoning their work. 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Khat suse/addiction sometimes led to crime. Some, especially in rural areas, usually quarrel 4 5 6 with others. The main source of reported conflict was khat theft. Participants observed that khat 7 8 users who don’t have money to buy usually theft someone's khat from his/her farm. Among 9 10 urban participants, few khat users sometimes also engaged in theft. Others sell their home 11 12 utensils to get money for khat without the consent of their family members. Theft and violence 13 14 15 were common among khat users who drink alcohol excessively. 16 For peer review only 17 18 Financial consequences of khat use 19 20 21 Participants were concerned about the price of khat. The minimum reported daily expense was 22 23 about 1.5 USD for a bundle of khat, but many also spent for alcohol, coffee, shisha and cigarette. 24 25 Some participants, at sever stage, had sold assets or home utensils and spend the money for khat. 26 27 28 In addition to the direct adverse effect of khat use on finance, participants were also concerned 29 30 31 about the amount of time they spend by chewing khat. In order to get the desired stimulation 32 33 from khat, they were supposed to stay chewing khat for long hours. Participants question how http://bmjopen.bmj.com/ 34 35 much they could earn if they didn't spend their time by chewing khat. Others who chew being in 36 37 a group usually spend more time and unable to quit the session and go for work being attracted 38 39 40 by the fun and the chat. 41 on September 25, 2021 by guest. Protected copyright. 42 43 The financial advantage from khat was reported from participants who have khat farm and sell 44 45 khat. They were relatively better in terms of financial capacity. For them, khat is an important 46 47 cash crop which allows them to cover home expenses and pay government tax easily while 48 49 50 others were under stress. 51 52 53 Physical health related consequences 54 55 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Many of the participants admitted that after khat use, it is common to experience loss of 4 5 6 appetite. Thus, weight loss and malnutrition were commonly reported. 7 8 9 Khat absorbs your body fluid and makes you dry so that a khat user is thin and latter would be 10 11 vulnerable to different diseases. (Age in 20’s, rural, male) 12 13 14 Gastrointestinal and oral health problems were frequently reported. Bad odour of the mouth, a 15 16 colour change of theFor tongue andpeer teeth, and reviewteeth spoil were among only the major complaints. 17 18 Chronic khat user participants couldn't prefer sauce and beef, which are common and valued in a 19 20 21 routine dish of the culture because of the burning sensation of their mouth and teeth damage. 22 23 Dehydration and constipation were also other major complaints. During khat withdrawal phase; 24 25 general physical pain, severe headache, burning sensation, and redness of the eyes were common 26 27 28 complaints. 29 30 31 For participants from Wolketie, chewing khat and drinking alcohol were considered as risky 32 33 behaviour for unsafe sex and thus HIV/AIDS infection. Many were also excessive alcohol users http://bmjopen.bmj.com/ 34 35 and vulnerable to all the adverse effects of excessive alcohol use. The leftover khat disposed 36 37 elsewhere in the city was also another concern for their health. Participants rarely reported 38 39 40 accidents and injuries related to khat use. Table 2 below shows major indicators for normal and 41 on September 25, 2021 by guest. Protected copyright. 42 problematic khat use. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Page 23 of 39 BMJ Open

1 2 3 Table 2; A conceptual summary of normative and problematic khat use 4 5 6 Parameters Normal khat use Problematic khat use 7 The reason of khat use -Khat use for prayer, leisure, functional and other social -Chewing to manage personal pain and distress 8 activities -Khat use is an end by itself 9 -Khat use is a means to an end -Continue khat use because of dependency, craving 10 -Continue khat use to conform to the social norms 11 Who chews khat -Healthy male adults, rarely women from cities -Khat use by women and children was perceived as problematic 12 -The communityFor doesn’t recommendpeer khat usereview by persons especially only in rural areas 13 14 with mental illness and with other critical health 15 conditions 16 Frequency of khat use -Infrequently where the maximum was three times per -Regularly 17 week - Almost daily 18 -Situation or event led khat use -Many can’t skip for a day or theirhttp://bmjopen.bmj.com/ fixed daily khat chewing 19 session 20 21 Amount of khat -Limited amount; after chewing few leaves, they could -Chew increased amount of khat compared to their friends and 22 divert their attention from the khat to their work couldn’t divert their attention except chewing 23 -Some had been chewing a lot -Long sessions such as half a day or more 24 - Short sessions, long sessions were for recreational users -Less sever problematic khat users chew while accomplishing 25 their routines on September 25, 2021 by guest. Protected copyright. 26 Other contexts of khat -Chewing after meal -Chewing even when there is no meal or usually skip meal 27 28 use -The chewing pattern is in line with the social norm -The chewing pattern is deviated from the societal norm 29 (place, time, situation) -Negative attitude from others 30 -Favorable attitude from others 31 Khat related benefits and -Perceived benefits or minor harms -Health, social and economic harms including malnutrition and 32 harms -There is normal functioning or productivity reduced body weight as well as different physical health 33 - From mild to strong level of social support complaints, poverty, family break up, separated from social 34 35 -khat is an agent of survival in the community support system and living on the street, begging khat or theft for 36 daily khat consumption or collecting left over khat, depression, 37 idleness, violating religious ritual e.gsalat-prayer 38 39 Settings of khat use -Team and ceremonial -Alone and no ceremony 40 -Sometimes people chew being alone and without ritual -If people chew together there is no team spirit 41 42 43 22 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

BMJ Open Page 24 of 39

1 2 3 -Chewing in home, mosque, or other special places in the -Chewing in home, khat cafeterias, public chewing including on 4 community the street and at work 5 6 Session and time of - Deciding the session of khat use before time usually -Session is not common and appropriate 7 chewing khat when the team have convenient situations -If there is a session, it is long and couldn’t leave the session 8 -Afternoon or rarely at night when other important personal and social affairs emerge 9 - Chewing khat in the morning, morning and afternoon, many 10 hours of the day, sometimes they also chew at night 11 Value of chewing khat - Symbolic and reality perception about khat use - Reality perception about khat use 12 - Value things Forwhich are contingent peer on khat reviewuse -Value only the khat use itself 13 14 -Khat is most valued at different sociocultural situations; -Companionship with the khat 15 hospitality, work party, mourning, wedding and spiritual 16 practise because it induces alertness, energy, 17 concentration, open discussion http://bmjopen.bmj.com/ 18 -Companionship with the group, with the prayer, 19 20 discussion 21 - There is pleasure and utility associated with khat use 22 Suse/addiction -Khat use results in harmony in the family and society - Associated harms on family and social, work 23 -There is no or minor craving -Unable to function without khat 24 -people experience well, minor and transient experiences -There is craving-harara 25 of distresses before and after khat use -Serious withdrawal experiences and distressing experiences after on September 25, 2021 by guest. Protected copyright. 26 - Quitting is easy, but there is no frequent thought or khat use /mirqanna 27 28 attempt about quitting - Quitting is difficult because of the distressing experiences when 29 the khat is withdrawn and persons usually think or attempt to quit 30 Associated behaviours - Some use, other psychoactive substances, others don’t -Khat use was in combination with smoking cigarettes, shisha and 31 drinking alcohol 32 Khat use and the broader -Careful for other important areas of life; self-care, deity -Abandonment of other important areas of life including poor self- 33 area of life -Purpose and meaning in life since there is fun, care and frequently skip a meal 34 35 socialization, social support -A problem in purpose and meaning in life as there are depression, 36 -A motivation for many areas of life and other withdrawal experiences as well as poor social support 37 -Amotivational syndrome 38 39 40 41 42 43 23 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 25 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Discussion 4 5 6 In this study, our aim was to conceptualize problematic khat us in a predominantly rural setting 7 8 9 in south-central Ethiopia. The study helped to answer the question: what constitutes problematic 10 11 khat use? Our results could inform development of screening tool to measure problematic khat 12 13 use. 14 15 16 Although there wasFor also peer sociocultural review khat use pattern, only khat addiction and negative 17 18 consequences of khat use constitute problematic khat use. The local term khat suse, semantically 19 20 21 equivalent to khat addiction which doesn’t conclusively infer to DSM-5 stimulant use disorder 22 23 definition[43]. Khat suse/khat addiction could only qualify impaired control and pharmacological 24 25 criteria among the criteria of stimulant use disorders. Khat suse/ khat addiction shares some 26 27 28 characteristics from other substances use disorders than stimulant use disorders. For example, it 29 30 has a similar functional consequences of cannabis use disorder-amotivational syndrome[43]. In 31 32 the case of khat suse, the local idiom Jezba could be conceptually related but broader. In the 33 http://bmjopen.bmj.com/ 34 35 current study setting, the use of the word, Jezba, indicates the existing stigma and it could also be 36 37 a good explanation how much sever form of problematic khat use is well recognized in the 38 39 setting since it has a negative connotation. Khat suse is also indicated by frequent yawning when 40 on September 25, 2021 by guest. Protected copyright. 41 not using the khat. This withdrawal criterion is similar with opioid withdrawal[43]. 42 43 44 Major negative consequences of khat use, which are other indicators of problematic khat use, 45 46 47 include sexual dysfunction, depression, psychosis, various oral health problems, and wastage of 48 49 time. Many of these indicators of problematic khat use are reported in several previous studies 50 51 [33, 37, 44]. Frequency and amount are important predictors of problematic alcohol and cannabis 52 53 54 use[45, 46] so that similar inference could be applied for problematic khat use. Studies also 55 56 estimate safe limits of alcohol use[47] and cannabis use[48]. This qualitative study explored that 57 58 24 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 there are normal and problematic khat use patterns, and negative consequences are more likely to 4 5 6 be experienced among problematic khat users who are using khat frequently and in large amount 7 8 although there are additional problematic khat use criteria. Therefore, future studies could 9 10 investigate “safe limit” of khat use to inform measures of khat harm reduction as it is done for 11 12 cannabis [49]. 13 14 15 The sociocultural khat use pattern was considered as normative khat use. This has limits in 16 For peer review only 17 18 terms of amount, frequency, duration of khat session and contexts of use. People chew khat 19 20 during weddings, funeral ceremonies, working, social meetings, and other cultural activities in 21 22 different settings both in Ethiopia[50, 51], and elsewhere such as in Kenya [52], Somaliland [19], 23 24 25 Somalia [53], Yemen [20], other African and Middle Eastern countries[54], and among 26 27 immigrants in the West [55-57]. Different factors such as accessibility and availability, social 28 29 accommodation and cultural acceptable could facilitate the process of drug normalization in 30 31 32 general [58] and khat use in particular[50]. 33 http://bmjopen.bmj.com/ 34 35 The study had different implications. Intervention for problematic khat use should be 36 37 systematically designed and planned. One systematic review [59] indicated community, family 38 39 and individual level interventions were acceptable and showed modest efficacy. This qualitative 40 on September 25, 2021 by guest. Protected copyright. 41 study adds to the findings of the systematic review that the sociocultural background of khat use 42 43 44 needs to be considered in designing policies and implementing interventions. Since there is 45 46 normative khat use in the current study setting as well as in other settings, mentioned above, 47 48 abstinence might not be effective in addressing problematic khat use pattern[59]. 49 50 51 Above all, problematic khat use should be an important component of substance use disorders or 52 53 54 mental health care system of the country. Practitioners should be committed to screen and offer 55 56 interventions for people with problematic khat use. Culturally adapted and psychometrically 57 58 25 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 valid screening tools for problematic khat use would be the priority to facilitate the clinical 4 5 6 practice and further research. Future studies would also play an important role to produce and 7 8 adapt evidence based interventions. The current study, aligned with previous studies, informs 9 10 policy makers to focus on khat use regulation and harm reduction strategies. 11 12 13 Strengths and limitations 14 15 16 Regarding methodologicalFor concern peer of the reviewstudy, it didn’t include only the perspectives of families of 17 18 khat users in detail, but the perspective of non-khat users in general were included. The analysis 19 20 21 of the study was also not theorized using a well-established existing model or theory, but this 22 23 allows the data to speak for itself without imposing some other framework that may poorly fit to 24 25 the data. This qualitative study, tapped into indigenous knowledge systems, was community- 26 27 28 based which make it very strong to define problematic khat use and develop an understanding of 29 30 its domains broadly. 31 32 33 Conclusion http://bmjopen.bmj.com/ 34 35 36 The study has illustrated what constitutes normative [acceptable] and problematic khat use in the 37 38 Gurage community in south-central Ethiopia. We found that problematic khat use is 39 40

characterized by patterns of use, reasons for use, contexts or norms, adverse psychological on September 25, 2021 by guest. Protected copyright. 41 42 43 reactions after use, khat suse/addiction, and khat use-related harms. The study would inform 44 45 future studies on development of tools to measure problematic khat use. The study will also be 46 47 used as a formative study for future longitudinal and intervention studies focusing on estimating 48 49 50 and addressing multidimensional impacts of problematic khat use. 51 52 53 Acknowledgements 54 55 56 57 58 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 The authors would like to acknowledge study participants and the African mental health research 4 5 6 initiative (AMARI) of DELTAS Africa Initiative. 7 8 9 Funding: 10 11 12 This work was supported through the DELTAS Africa Initiative [DEL-15-01]. The DELTAS 13 14 Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s 15 16 Alliance for AcceleratingFor Excellence peer in Science review in Africa (AESA) only and supported by the New 17 18 Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) 19 20 21 with funding from the Wellcome Trust [DEL-15-01] and the UK government. The views 22 23 expressed in this publication are those of the author(s) and not necessarily those of AAS, 24 25 NEPAD Agency, Wellcome Trust or the UK government. 26 27 28 Authors’ contributions: 29 30 31 AM ST AF conceived and designed the study. AM did the interview. AM and CN coded the 32 33 http://bmjopen.bmj.com/ 34 data. AM AF KH CN SN ST contributed to the analysis and the write-up of the manuscript. All 35 36 authors agree with the results and conclusions of the study. 37 38 39 Ethical approval and consent to participate 40 41 on September 25, 2021 by guest. Protected copyright. 42 The study was approved by the Institutional Review Board of the College of Health Sciences of 43 44 Addis Ababa University (REF. 008/18/psy). Participants took part in the study after providing 45 46 written informed consent. 47 48 49 Competing interests: 50 51 52 None declared 53 54 55 Data sharing statement: No additional unpublished data are available from this study. 56 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 References 4 5 6 1. Cox, G. and H. Rampes, Adverse effects of khat: A review. Advances in Psychiatric Treatment, 7 2003. 9(6): p. 456-463. 8 2. Kassim, S., et al., Before the ban--an exploratory study of a local khat market in East London, 9 10 U.K. Harm Reduct J, 2015. 12: p. 19. 11 3. Douglas, H. and A. Hersi, Khat and islamic legal perspectives: issues for consideration. The 12 Journal of Legal Pluralism and Unofficial Law, 2010. 42(62): p. 95-114. 13 4. EMCDDA, European Monitoring Center for Drugs and Drug Addiction. (2011). Drugs in focus. 14 ISSN 1681- 157. 211. 15 5. Numan, N., Exploration of adverse psychological symptoms in Yemeni khat users by the 16 Symptoms Checklist-90For (SCL-90).peer Addiction, review 2004. 99(1): p. only 61-65. 17 6. Elmi, A.S., The chewing of khat in Somalia. Journal of ethnopharmacology, 1983. 8(2): p. 163- 18 19 176. 20 7. Haile, D. and Y. Lakew, Khat chewing practice and associated factors among adults in Ethiopia: 21 further analysis using the 2011 demographic and health survey. PloS one, 2015. 10(6): p. 22 e0130460. 23 8. Teklie, H., et al., Prevalence of Khat chewing and associated factors in Ethiopia: Findings from 24 the 2015 national Non-communicable diseases STEPS survey. Ethiopian Journal of Health 25 Development, 2017. 31(1): p. 320-330. 26 9. Alem, A., D. Kebede, and G. Kullgren, The prevalence and socio-demographic correlates of khat 27 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica, Supplement, 1999. 99(397): p. 84- 28 29 91. 30 10. Alem, A., Kebede, D., & Kullgren, G. , The prevalence and socio-demographic correlates of khat 31 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica Supplementum, 1999. 100: p. 84- 32 91. 33 11. Adugna, F., C. Jira, and T. Molla, Khat chewing among Agaro secondary school students, Agaro, http://bmjopen.bmj.com/ 34 southwestern Ethiopia. Ethiopian Medical Journal, 1994. 32(3): p. 161-166. 35 12. al., G.e., Prevalence and Predictors of harmful Khat use among university Students in Ethiopia. . 36 Substance Abuse: Research and Treatment 2014. 8: p. 45-51. 37 38 13. Patel, S.L., R. Murray, and G. Britain, Khat use among Somalis in four English cities. 2005: 39 Citeseer. 40 14. Gebissa, E., Scourge of life or an economic lifeline? Public discourses on khat (Catha edulis) in 41 Ethiopia. Substance Use & Misuse, 2008. 43(6): p. 784-802. on September 25, 2021 by guest. Protected copyright. 42 15. Haile, D. and Y. Lakew, Khat chewing practice and associated factors among adults in Ethiopia: 43 Further analysis using the 2011 demographic and health survey. PLoS ONE, 2015. 10 (6) (no 44 pagination)(e0130460). 45 16. Hersi, M.A. and M. Abdalla, Sharī ‘a Law and the Legality of Consumption of Khat (Catha Edulis): 46 47 Views of Australian Imāms. 2013. 48 17. Geda, G.J., Pilgrimages and Syncretism: Religious transformation among the Arsi Oromo of 49 Ethiopia. 2015. 50 18. Griffioen, S., Mohammed Girma, Understanding Religion and Social Change in Ethiopia. Toward 51 a Hermeneutic of Covenant. Palgrave Macmillan, New York, 2012. 240 pages. ISBN 978-1-137- 52 269416. Philosophia Reformata, 2013. 78(2): p. 222-225. 53 19. Hansen, P., The ambiguity of khat in Somaliland. Journal of ethnopharmacology, 2010. 132(3): p. 54 590-599. 55 56 57 58 28 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 20. Al-Motarreb, A., K. Baker, and K.J. Broadley, Khat: pharmacological and medical aspects and its 4 social use in Yemen. Phytotherapy research, 2002. 16(5): p. 403-413. 5 6 21. Mihretu, A., S. Teferra, and A. Fekadu, What constitutes problematic khat use? An exploratory 7 mixed methods study in Ethiopia. Substance Abuse Treatment, Prevention, and Policy, 2017. 8 12(1): p. 17. 9 22. WHO., Assessment of Khat (Catha edulis Forsk) Geneva: WHO. . 2006. 10 23. Asgedom, S.W., E.K. Gudina, and T.A. Desse, Assessment of Blood Pressure Control among 11 Hypertensive Patients in Southwest Ethiopia. PloS one, 2016. 11(11): p. e0166432. 12 24. Al-Hadrani, A.M., Khat induced hemorrhoidal disease in Yemen. Saudi Medical Journal, 2000. 13 21(5): p. 475-477. 14 25. Hassen, K., et al., Khat as a risk factor for hypertension: A systematic review. JBI Database of 15 16 Systematic ReviewsFor and peer Implementation review Reports, 2012. 10 only(44): p. 2882-2905. 17 26. Al-Habori, M., The potential adverse effects of habitual use of Catha edulis (khat). Expert opinion 18 on drug safety, 2005. 4(6): p. 1145-1154. 19 27. Hassan, N.A., et al., The effect of Qat chewing on blood pressure and heart rate in healthy 20 volunteers. Tropical doctor, 2000. 30(2): p. 107-108. 21 28. Mikulica, J., et al., Khat Use, PTSD and Psychotic Symptoms among Somali Refugees in Nairobi-A 22 Pilot Study. 2014. 23 29. Bhui, K., et al., Mental disorders among Somali refugees. Social psychiatry and psychiatric 24 25 epidemiology, 2006. 41(5): p. 400-408. 26 30. Odenwald, M., Neuner, F., Schawer, M., Elbert, T.R., Catani, C., Lingenfelder, B., Hinkel, H., 27 Hafner, H. & Stroh, B. , Khat use as risk factor for psychotic disorders: A cross-sectional and case- 28 control study in Somalia. . BMC Medicine, 2005. 3(5). 29 31. Hassan, N.A., et al., The effect of chewing Khat leaves on human mood. Saudi medical journal, 30 2002. 23(7): p. 850-853. 31 32. Bhui, K. and N. Warfa, Trauma, khat and common psychotic symptoms among Somali 32

immigrants: A quantitative study. J Ethnopharmacol, 2010. 132(3): p. 549-553. http://bmjopen.bmj.com/ 33 33. Mihretu, A., et al., Definition and Validity of the Construct “Problematic Khat Use”: A Systematic 34 35 Review. European addiction research, 2019. 25(4): p. 161-172. 36 34. Duresso, S., Matthews, A., Ferguson, S. & Bruno,R., Is khat use disorder a valid diagnostic entity? 37 . School of Medicine, University of Tasmania, Hobart, Australia, 2015. 38 35. Gebrehanna, E., Y. Berhane, and A. Worku, Khat chewing among Ethiopian University Students-- 39 a growing concern. BMC public health, 2014. 14: p. 1198. 40 36. Kassim, S., S. Islam, and R. Croucher, Validity and reliability of a Severity of Dependence Scale for on September 25, 2021 by guest. Protected copyright. 41 khat (SDS-khat). Journal of Ethnopharmacology, 2010. 132(3): p. 570-577. 42 37. Mihretu, A., S. Teferra, and A. Fekadu, What constitutes problematic khat use? An exploratory 43 44 mixed methods study in Ethiopia. Substance Abuse Treatment, Prevention, and Policy, 2017. 12: 45 p. 17. 46 38. CSA, E., 2014 projected population Size of Towns by Sex, Region, Zone and Wereda 2015. 47 39. Crass, J. and R. Meyer. The Qabena and the Wolane: Two peoples of the Gurage region and their 48 respective histories according to their own oral traditions. in Annales d'Éthiopie. 2001. Editions 49 de la Table Ronde. 50 40. Tong, A., P. Sainsbury, and J. Craig, Consolidated criteria for reporting qualitative research 51 (COREQ): a 32-item checklist for interviews and focus groups. International journal for quality in 52 53 health care, 2007. 19(6): p. 349-357. 54 41. Braun, V. and V. Clarke, Using thematic analysis in psychology. Qualitative research in 55 psychology, 2006. 3(2): p. 77-101. 56 57 58 29 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 42. Umea° University: UMDAC and Epidemiology, D.o.P.H.a.C.M.a.U.U., Open Code version 4.03. 4 2013. 5 6 43. Association, A.P., Diagnostic and statistical manual of mental disorders (DSM-5®). 2013: 7 American Psychiatric Pub. 8 44. Duresso, S.W., et al., Is khat use disorder a valid diagnostic entity? Addiction, 2016. 111(9): p. 9 1666-1676. 10 45. Walden, N. and M. Earleywine, How high: quantity as a predictor of cannabis-related problems. 11 Harm Reduction Journal, 2008. 5(1): p. 20. 12 46. Gmel, G., J.-L. Heeb, and J. Rehm, Is frequency of drinking an indicator of problem drinking? A 13 psychometric analysis of a modified version of the alcohol use disorders identification test in 14 Switzerland. Drug and Alcohol Dependence, 2001. 64(2): p. 151-163. 15 16 47. Stockwell, T. Forand E. Single, peer Standard unitreview labelling of alcohol only containers. Alcohol: Minimising the 17 Harm: What Works, 1997: p. 85-104. 18 48. Zeisser, C., et al., A ‘standard joint’? The role of quantity in predicting cannabis-related problems. 19 Addiction Research & Theory, 2012. 20(1): p. 82-92. 20 49. Fischer, B., et al., Lower-risk cannabis use guidelines: a comprehensive update of evidence and 21 recommendations. American journal of public health, 2017. 107(8): p. e1-e12. 22 50. Gebissa, E., Leaf of Allah: khat & agricultural transformation in Harerge, Ethiopia 1875-1991. 23 2004: Ohio State University Press. 24 25 51. Alem, A., D. Kebede, and G. Kullgren, The prevalence and socio-demographic correlates of khat 26 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica Supplementum, 1999. 100(Suppl 27 397): p. 84-91. 28 52. Carrier, N., ‘Miraa is cool’: the cultural importance of miraa (khat) for Tigania and Igembe youth 29 in Kenya*. Journal of African Cultural Studies. 17(2): p. 201-218. 30 53. Basunaid, S., M. Van Dongen, and T.J. Cleophas, Khat abuse in Yemen: A population-based 31 survey. Clinical Research and Regulatory Affairs, 2008. 25(2): p. 87-92. 32

54. Manghi, R.A., et al., Khat use: lifestyle or addiction? Journal of psychoactive drugs, 2009. 41(1): http://bmjopen.bmj.com/ 33 p. 1-10. 34 35 55. Douglas, H., M. Boyle, and N. Lintzeris, The health impacts of khat: A qualitative study among 36 Somali-Australians. Medical Journal of Australia, 2011. 195(11): p. 666-669. 37 56. Patel, S.L., Attitudes to khat use within the Somali community in England. Drugs: Education, 38 Prevention & Policy, 2008. 15(1): p. 37-53. 39 57. Stevenson, M., J. Fitzgerald, and C. Banwell, Chewing as a social act: Cultural displacement and 40 khat consumption in the East African communities of Melbourne. Drug and Alcohol Review, on September 25, 2021 by guest. Protected copyright. 41 1996. 15(1): p. 73-82. 42 58. Parker, H., Normalization as a barometer: Recreational drug use and the consumption of leisure 43 44 by younger Britons. Addiction research & theory, 2005. 13(3): p. 205-215. 45 59. Ahmed, S., H. Minami, and A. Rasmussen, A Systematic Review of Treatments for Problematic 46 Khat Use. Substance use & misuse, 2019: p. 1-12. 47 48 49 50 51 52 ` 53 54 55 56 57 58 30 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 31 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Reporting checklist for qualitative study. 4 5 6 7 8 Page 9 10 Reporting Item Number 11 12 Title 13 14 #1 Exploring the concept of Problematic khat use in the Gurage 1 15 16 For peer review only 17 community, South-central Ethiopia: A qualitative study 18 19 20 21 22 Abstract 23 24 Objective: This study aimed at exploring how problematic khat use is 25 #2 3 26 27 characterized in the Gurage community, South-central Ethiopia. 28 29 30 Design: qualitative study. 31 32 33 Setting: Gurage community, South-central Ethiopia http://bmjopen.bmj.com/ 34 35 36 Participants: We conducted in-depth interviews with 14 khat users and 37 38 5 non-khat users, and 3 focus-group discussions with khat users. 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Methods: All participants were selected purposively based on their 43 44 exposure to khat or khat use. We used interview guide to explore 45 46 perception of participants about khat use and problematic khat use. We 47 48 49 analyzed the data thematically using Open Code software version 4.03. 50 51 We used iterative data collection and analysis, triangulation of methods 52 53 and respondent validation to ensure scientific rigour. 54 55 56 Findings: We identified three major themes: sociocultural khat use, 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from khat suse (khat addiction), and negative consequences of khat use. 1 2 3 Sociocultural khat use included a broad range of contexts and patterns 4 5 including use of khat for functional, social, cultural and religious 6 7 reasons. Khat addiction was mainly explained in terms of associated 8 9 khat withdrawal experiences, including harara/craving, and inability to 10 11 12 quit. We identified mental health, sexual life, physical health, social 13 14 and financial related negative consequences of khat use. The local 15 16 Foridiom Jezbapeer was used review to label subgroup only of individuals with khat suse 17 18 19 (khat addiction). 20 21 22 Conclusion: The study has identified what constitutes normative and 23 24 problematic khat use in the Gurage community in South-central 25 26 Ethiopia. Problematic khat use is broad concepts which include 27 28 29 frequency, reasons, contexts, negative consequences and addiction of 30 31 khat use. Insights generated can be used to inform future studies on 32 33 development of tools to measure problematic khat use. http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Introduction 43 44 Problem formulation #3 Across the literature, conceptualization of problematic, sociocultural 5 45 46 47 and recreational khat use has been an important research gap. 48 49 Problematic khat use, rather than khat use per se, is usually the interest 50 51 of the public, researchers and policymakers. Nevertheless, only few 52 53 54 previous studies were conducted on problematic khat use [33]. The 55 56 inconclusive reports about the different harms of khat use could also be 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from due to poor definition of problematic khat use. The Diagnostic 1 2 3 Statistical Manuel(DSM-5) definition of stimulant use disorders [34] 4 5 could have more important clinical utility (for severe cases) than 6 7 screening individuals with problematic khat use at earlier stage. Lack 8 9 of screening tool for problematic khat use, especially among nonclinical 10 11 12 cases could hamper efforts to curb the problem including early 13 14 identification and effective management of positive cases. Therefore, 15 16 Forthere is peera need for valid review problematic khat only use screening tools which 17 18 19 would facilitate clinical care in primary health care settings and for 20 21 future research. Although there are no strong validation studies, few 22 23 studies used Harmful Khat Use Scale [35] and Severity of Dependence 24 25 26 Scale to measure the construct problematic khat use [36], but a 27 28 systematic review [33] and exploratory studies[37] suggested broader 29 30 indicators of problematic khat use, including amount, frequency, 31 32

context and duration of khat session. http://bmjopen.bmj.com/ 33 34 35 36 Purpose or research #4 The aim of this study was to conceptualize problematic khat use from 5 37 question 38 the perspective of users and non-users in a dominantly rural setting, 39 40 Gurage, south-central Ethiopia. 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 Methods 47 48 Qualitative approach and #5 The study employed a qualitative study design [40] which allowed for 6 49 research paradigm 50 the understanding and description of the experiences and perspectives 51 52 53 of people towards problematic khat use. This study was guided by the 54 55 consolidated criteria for reporting qualitative research (COREQ)[40]. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Researcher characteristics #6 Participants were also diverse in terms of socio-demographic 7 and 2 and reflexivity across the 3 characteristics. The field worker, the first author, has previous 4 paper 5 6 experience interviewing people about khat use. Above all, we tried to 7 8 purely and openly present the ideas of the participants without personal 9 10 impression interference. 11 12 13 14 15 16 Context #7 ForThe study peer was conducted review in the Gurage only zone; Southern Nations, 6 17 18 Nationalities and Peoples Region (SNNPR), Ethiopia. Ethiopian 19 20 Orthodox Christianity (48.17%) and Muslim (42.31%) are the two 21 22 23 dominant religions in Wolketie town, Gurage’s capital [38]. Peasant 24 25 farming is the main productive occupation in rural areas while petty 26 27 trading is more common in urban areas. The area is known for its khat 28 29 30 production and khat use [10]. Khat might have been introduced to 31 32 Gurage area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 33 http://bmjopen.bmj.com/ 34 1543) army or neighboring Muslim Wolane or Oromo ethnic groups 35 36 37 [39]. 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Sampling strategy #8 We conducted face to face in-depth interviews with 14 current khat 6 43 44 users and five non-khat users. Twenty-one khat users participated in 45 46 47 focus group discussions (FGD). The first FGD consisted of six women, 48 49 the second and the third FGDs had seven and eight participants 50 51 respectively. Participants were selected purposively based on their 52 53 54 experience of khat use, and we also aimed for maximum variation 55 56 considering the socio-demographic characteristics of the participants. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Ethical issues pertaining to #9 Ethical clearance was obtained from the Institutional Review Board 7 2 human subjects 3 (IRB) of the College of Health Sciences, Addis Ababa University (Ref 4 5 6 008/18/Psy). In addition, we also obtained a letter of support from the 7 8 Gurage zone health department. Written informed consent was sought 9 10 and obtained from all participants before data collection. 11 12 13 14 15 16 Data collection methods #10ForParticipants peer were invited review by the community only health workers and the first 6 17 author did the informed consent. All the interviews and FGDs were 18 19 conducted in Amharic and tape-recorded. The first author, who has 20 good experience of interviewing khat users, did the interviews and 21 22 facilitated the focus group discussions assisted by a trained moderator. 23 The interviews took about 40 minutes and the FGDs about one hour in 24 25 average. The first and the last authors designed in-depth interview and 26 FGD guides. 27 28 29 Data collection #11 The guides mainly focused on the experiences and perceptions of the 6 30 instruments and participants regarding khat use and problematic khat use. 31 32 technologies 33 http://bmjopen.bmj.com/ 34 Units of study #12 Participants were diversified in several socio-demographic 8 35 36 characteristics, including sex, age, residence, religion, occupation and 37 educational status (Table 1). 38 39 Data processing #13 The data were transcribed verbatim in Amharic, then translated into 7 40 41 English by the first author and experienced research assistants. on September 25, 2021 by guest. Protected copyright. 42 43 Data analysis #14 . All interview and FGD translations were coded independently by the 7 44 45 first and fourth author (AM and CN). Iterative thematic analysis [41] 46 was done simultaneously with data collection. We used computer 47 48 software, open code 4.03 to manage and analyse the data [42]. 49 50 Techniques to enhance #15 Regarding to data quality management and rigour, the iterative process 7 51 52 trustworthiness of data collection, data analysis and checking unclear issues from the 53 participants added to the quality of the study. We did also data 54 55 triangulation from different sources, including religious fathers, key 56 informants for the culture and legal officers. 57 58 Results/findings 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Syntheses and #16 The major themes that emerged from the iterative thematic analysis 9 2 3 interpretation were: (1) sociocultural khat use, (2) suse “(addictive)” khat use, and (3) 4 negative consequences of khat use. The second theme had two 5 6 categories: withdrawal experiences, quitting khat use and 7 mirqanna/feeling high after khat use. The negative consequences were 8 9 categorized to mental health, sexual life, physical health, social and 10 financial. These themes are discussed with support of quotes directly 11 12 from the participants. 13 14 Links to empirical data #17 Withdrawal experiences of khat use 15 15 16 For peer review only 17 Other experiences reported associated with abnormal khat use included 18 19 withdrawal experiences when one was weaning off khat after heavy 20 21 22 usage. The study found many important psychological withdrawal 23 24 experiences of khat suse-addiction. Feeling depressed, irritable and 25 26 aggression were the most typical and commonly reported experiences. 27 28 29 One participant said, “I am usually against people’s communication 30 31 even aggressive when I didn’t chew khat. I remember that I once threw 32 33 away [smacked] my kid when she was talking about her school affairs.” http://bmjopen.bmj.com/ 34 35 (Age 67, urban, male) 36 37 38 39 Different participants reported different withdrawal symptoms, 40 41 including lack of motivation, unable or lack interest to function or on September 25, 2021 by guest. Protected copyright. 42 43 socialize, poor concentration and learning, unable to receive message, 44 45 and dukak (vivid unpleasant dream). Here are two dukak or vivid 46 47 48 unpleasant dream experiences; 49 50 51 One day my husband skipped chewing khat and went to bed. Then, he 52 53 spent the night spitting. In the morning, the bed sheet was wet. When I 54 55 56 ask him what was wrong with him. He said; “people were punishing me 57 58 with the smoke of red paper and I had been feeling burning sensation 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 39 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from for the whole night”. (Age 28, urban, women) 1 2 3 Another experience: 4 5 6 7 My experience was...ehm... usually a man would hold my hair and hung 8 9 me or put me into a hole and I would wake up in panic. (Age 38, male, 10 11 urban) 12 13 14 15 16 For peer review only 17 Discussion 18 19 Intergration with prior #18 In this study, our aim was to conceptualize problematic khat us in a 24 20 work, implications, 21 predominantly rural setting in south-central Ethiopia. The study helped 22 transferability and 23 24 contribution(s) to the field to answer the question: what constitutes problematic khat use? Our 25 26 results could inform development of screening tool to measure 27 28 problematic khat use. 29 30 31 Although there was also sociocultural khat use pattern, khat addiction 32 33 and negative consequences of khat use constitute problematic khat use. http://bmjopen.bmj.com/ 34 The local term khat suse, semantically equivalent to khat addiction 35 36 which doesn’t conclusively infer to DSM-5 stimulant use disorder 37 definition[43]. Khat suse/khat addiction could only qualify impaired 38 39 control and pharmacological criteria among the criteria of stimulant use 40 disorders. Khat suse/ khat addiction shares some characteristics from 41 on September 25, 2021 by guest. Protected copyright. 42 other substances use disorders than stimulant use disorders. For 43 example, it has a similar functional consequences of cannabis use 44 45 disorder-amotivational syndrome[43]. In the case of khat suse, the local 46 idiom Jezba could be conceptually related but broader. In the current 47 48 study setting, the use of the word, Jezba, indicates the existing stigma 49 and it could also be a good explanation how much sever form of 50 51 problematic khat use is well recognized in the setting since it has a 52 negative connotation. Khat suse is also indicated by frequent yawning 53 54 when not using the khat. This withdrawal criterion is similar with 55 opioid withdrawal[43]. 56 57 58 Limitations #19 Regarding methodological concern of the study, it didn’t include the 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 39 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from perspectives of families of khat users in detail, but the perspective of 1 2 3 non-khat users in general were included. The analysis of the study was 4 5 also not theorized using a well-established existing model or theory, but 6 7 this allows the data to speak for itself without imposing some other 8 9 framework that may poorly fit to the data. This qualitative study, tapped 10 11 12 into indigenous knowledge systems, was community-based which make 13 14 it very strong to define problematic khat use and develop an 15 16 Forunderstanding peer of its domainsreview broadly. only 17 18 19 20 Other 21 22 Conflicts of interest #20 None declared 28 23 24 Funding #21 This work was supported through the DELTAS Africa Initiative [DEL- 27 25 26 27 15-01]. The DELTAS Africa Initiative is an 28 29 independent funding scheme of the African Academy of Sciences 30 31 (AAS)’s Alliance for Accelerating Excellence in Science in Africa 32 http://bmjopen.bmj.com/ 33 (AESA) and supported by the New Partnership for Africa’s 34 35 36 Development Planning and Coordinating Agency (NEPAD Agency) 37 38 with funding from the Wellcome Trust [DEL-15-01] and the UK 39 40

government. The views expressed in this publication are those of the on September 25, 2021 by guest. Protected copyright. 41 42 43 author(s) and not necessarily those of AAS, NEPAD Agency, 44 45 Wellcome Trust or the UK government. 46 47 48 49 50 51 None The SRQR checklist is distributed with permission of Wolters Kluwer © 2014 by the Association of American Medical Colleges. 52 This checklist can be completed online using https://www.goodreports.org/, a tool made by the EQUATOR Network in collaboration with 53 54 Penelope.ai 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Exploring the concept of Problematic khat use in the Gurage community, South-central Ethiopia: A qualitative study

Journal: BMJ Open ManuscriptFor ID peerbmjopen-2020-037907.R1 review only Article Type: Original research

Date Submitted by the 04-May-2020 Author:

School of Medicine, Department of Psychiatry; Harvard T.H. Chan School http://bmjopen.bmj.com/ of Public Health, Boston, USA

Primary Subject Addiction Heading:

Secondary Subject Heading: Addiction, Mental health

MENTAL HEALTH, Substance misuse < PSYCHIATRY, QUALITATIVE Keywords: RESEARCH on September 25, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 41 BMJ Open

1 2

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on September 25, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

13 Correspondence author :Awoke Mihretu= [email protected] 14 15 16 Postal address:16417 For peer review only 17 18 telephone number: +251921331306 19 20 21 Abebaw Fekadu= [email protected] 22 23 24 Kassahun Habtamu= [email protected] 25 26 27 Clement Nhunzvi= [email protected] 28 29 30 Sam Norton= [email protected] 31 32 Solomon Teferra = [email protected] 33 http://bmjopen.bmj.com/ 34 35 1 Department of Psychiatry, School of Medicine, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia 36 37 2 Department of social sciences, college of natural and social sciences, Addis Ababa science and Technology University 38 39 3Centre for Innovative Drug Development and Therapeutic Trials for Africa (CDT Africa), College of Health Sciences, Addis Ababa University, 40 41 Addis Ababa, Ethiopia on September 25, 2021 by guest. Protected copyright. 42 43 4School of Psychology, College of Education and Behavioral Studies, Addis Ababa University, Addis Ababa, Ethiopia 44 45 5Department of Rehabilitation, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe 46 47 6Psychology Department, Institute of Psychiatry, King’s College London, London, UK 48 49 7Harvard T.H. Chan School of Public Health, Boston, USA 50 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Abstract 4 5 6 Objective: This study aimed at exploring how problematic khat use is characterized in the 7 8 Gurage community, South-central Ethiopia. 9 10 Design: qualitative study. 11 12 Setting: Gurage community, South-central Ethiopia 13 14 15 Participants: We conducted in-depth interviews with 14 khat users and 5 non-khat users, and 3 16 For peer review only 17 focus-group discussions with khat users. 18 19 Methods: All participants were selected purposively based on their exposure to khat or khat use. 20 21 22 We used interview guide to explore perception of participants about khat use and problematic 23 24 khat use. We analyzed the data thematically using Open Code software version 4.03. We used 25 26 iterative data collection and analysis, triangulation of methods and respondent validation to 27 28 29 ensure scientific rigour. 30 31 Findings: We identified three major themes: sociocultural khat use, khat suse (khat addiction), 32 33 and negative consequences of khat use. Sociocultural khat use included a broad range of contexts http://bmjopen.bmj.com/ 34 35 and patterns including use of khat for functional, social, cultural and religious reasons. Khat 36 37 38 addiction was mainly explained in terms of associated khat withdrawal experiences, including 39 40 harara/craving, and inability to quit. We identified mental health, sexual life, physical health, 41 on September 25, 2021 by guest. Protected copyright. 42 social and financial related negative consequences of khat use. The local idiom Jezba was used 43 44 45 to label subgroup of individuals with khat suse (khat addiction). 46 47 Conclusion: The study has identified what constitutes normative and problematic khat use in the 48 49 Gurage community in South-central Ethiopia. Problematic khat use is broad concepts which 50 51 52 include frequency, reasons, contexts, negative consequences and addiction of khat use. Insights 53 54 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 generated can be used to inform future studies on development of tools to measure problematic 4 5 6 khat use. 7 8 Keywords: khat use; normative khat use; problematic khat use; khat addiction; khat withdrawal; 9 10 qualitative study; Ethiopia 11 12 13 Strengths and limitations of the study 14 15 - Regarding methodological concern of the study, it didn’t include the perspectives of 16 For peer review only 17 18 families of khat users in detail, but the perspectives of non-khat users in general were 19 20 included. 21 22 - The analysis of the study was also not theorized using a well-established existing model 23 24 25 or theory, but this allows the data to speak for itself without imposing some other 26 27 framework that may poorly fit to the data. 28 29 - This qualitative study, tapped into indigenous knowledge systems, was community-based 30 31 32 which make it very strong to define problematic khat use and develop an understanding 33 http://bmjopen.bmj.com/ 34 of its domains broadly. 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 3 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Introduction 4 5 6 Khat, the psychoactive stimulant, is very common and its use is growing in East Africa and the 7 8 9 Arabian Peninsula(1). The ancient mode of khat use was “khat tea drinking”, “khat coffee”, “as 10 11 flavoring to local alcoholic drinks”, but the recent and common practice is chewing the fresh 12 13 leafs of khat(2). Khat use has also been reported in Western countries, such as the United 14 15 Kingdom (3) and Australia (4), primarily among Ethiopian, Somalian, Yemeni and Kenyan 16 For peer review only 17 18 diaspora communities (5). The prevalence of use was estimated to be 67.9% in Yemen, 59% in 19 20 Somalia (6, 7), and 16%-50% in Ethiopia (8-10). In Gurage zone, Butajira district, the current 21 22 prevalence of khat use was 50% and 17.4% chew daily (11). Many people chew khat for its 23 24 25 stimulating effects, to gain concentration and energy during work (12-14). 26 27 28 Khat use has been embedded in the culture and social life of East Africa and the Arabian 29 30 Peninsula especially among Muslim- dominant societies (15, 16). Some studies have reported the 31 32 association between khat use and being Muslim (10, 17), but the position of the Islam religion 33 http://bmjopen.bmj.com/ 34 35 regarding khat use remains unclear (18). In the literature, especially from Yemen and Saudi 36 37 Arabia, the following three outstanding themes of discourse have been raised with regard to khat 38 39 use: halal (permissible), makruh (disliked or discouraged) or haram (forbidden). For example, in 40 on September 25, 2021 by guest. Protected copyright. 41 Ethiopia, many Muslims chew khat when they go to pilgrimage centers and while doing rituals 42 43 44 such as singing, prayer-du'a, blessing and other activities (19). Muslim women also use khat 45 46 when they are gathered for prayers directed on women in labor: a social ritual called Fatimaye, 47 48 invoking the name of Fatima, the daughter of Prophet Muhammad (20). The perceived social 49 50 51 uses of khat were mainly for social gathering like weddings and funerals (21). In the Yemeni 52 53 society, khat use was associated with important social occasions to meet other people and 54 55 exchange of ideas (22). 56 57 58 4 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Although it has important cultural roots and functions(15, 16), khat use has significant health 4 5 6 harms(23-26), The physical health problems include increased body temperature, loss of 7 8 appetite, gastritis, haemorrhoids, insomnia, and oral health problems (23-26), and even 9 10 hypertension (27) and other cardiovascular dysfunctions (22, 28, 29). 11 12 13 The mental health impacts include psychotic symptoms (30-32) and depressive symptoms (33). 14 15 However, the mental health impacts remain inconclusive because other studies have revealed 16 For peer review only 17 18 negative findings (1, 34). 19 20 21 Across the literature, conceptualization of problematic, sociocultural and recreational khat use 22 23 has been an important research gap. Problematic khat use, rather than khat use per se, is usually 24 25 the interest of the public, researchers and policymakers. Nevertheless, only few previous studies 26 27 28 were conducted on problematic khat use (35). The inconclusive reports about the different harms of 29 30 khat use could also be due to poor definition of problematic khat use(35). 31 32 33 The Diagnostic Statistical Manuel(DSM-5) definition of stimulant use disorders (36) could have http://bmjopen.bmj.com/ 34 35 more important clinical utility (for severe cases) than screening individuals with problematic 36 37 khat use at earlier stage. Lack of screening tool for problematic khat use, especially among 38 39 40 nonclinical cases could hamper efforts to curb the problem including early identification and effective 41 on September 25, 2021 by guest. Protected copyright. 42 management of positive cases. Therefore, there is a need for valid problematic khat use screening 43 44 tools which would facilitate clinical care in primary health care settings and for future research. 45 46 47 Although there are no strong validation studies, few studies used Harmful Khat Use Scale (37) 48 49 and Severity of Dependence Scale to measure the construct problematic khat use (38), but a 50 51 systematic review (35) and exploratory studies(39) suggested broader indicators of problematic 52 53 khat use, including amount, frequency, context and duration of khat session. Therefore, the aim 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 of this study was to conceptualize problematic khat use from the perspective of users and non- 4 5 6 users in a dominantly rural setting, Gurage, south-central Ethiopia. 7 8 9 10 11 12 13 14 Methods 15 16 For peer review only 17 Study setting 18 19 20 The study was conducted in the Gurage zone; Southern Nations, Nationalities and Peoples 21 22 23 Region (SNNPR), Ethiopia. Ethiopian Orthodox Christianity (48.17%) and Muslim (42.31%) are 24 25 the two dominant religions in Wolketie town, Gurage’s capital (40). Peasant farming is the main 26 27 productive occupation in rural areas while petty trading is more common in urban areas. The area 28 29 30 is known for its khat production and khat use (12). Khat might have been introduced to Gurage 31 32 area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 1543) army or neighboring Muslim 33 http://bmjopen.bmj.com/ 34 Wolane or Oromo ethnic groups (41). 35 36 37 Study design 38 39 40 The study employed a qualitative study design (42) which allowed for the understanding and 41 on September 25, 2021 by guest. Protected copyright. 42 description of the experiences and perspectives of people towards problematic khat use. This 43 44 45 study was guided by the standard of reporting qualitative research(SRQR) and consolidated 46 47 criteria for reporting qualitative research (COREQ)(42). 48 49 50 Study participants and recruitment procedures 51 52 53 We conducted face to face in-depth interviews with 14 current khat users and five non-khat 54 55 users. Khat users are participants who use khat for the last 30 days. Twenty-one khat users 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 participated in focus group discussions (FGD). The first FGD consisted of six women, the 4 5 second and the third FGDs were males with seven and eight participants respectively. Sample 6 7 size was determined based on data saturation during data gathering and analyzing. Theoretical 8 data saturation was also informed by our previous studies(23) and other literature(43). 9 10 Participants were selected purposively based on their experience of khat use, and we also aimed 11 12 for maximum variation considering the socio-demographic characteristics of the participants. In 13 14 addition to khat users, other people who have knowledge about the history, cultural, religious 15 base of khat use and frequent contact with khat users were interviewed. Thus, 5 religious fathers, 16 For peer review only 17 police officers and culture experts were participated in the study. 18 19 20 Participants were invited by the community health workers when they had gathered for their 21 22 regular meeting and the first author did the informed consent. 23 24 Data collection procedures 25 26 27 Consented individuals were interviewed in their home, and at open places in the community. All 28 29 focus group discussions were conducted at health facilities (health posts) garden in the 30 31 community. Religious fathers, police officers and culture experts were interviewed at their 32 http://bmjopen.bmj.com/ 33 office. All the interviews and FGDs were conducted in Amharic and tape-recorded. The first 34 35 36 author, who has good experience of interviewing khat users, did the interviews and facilitated the 37 38 focus group discussions assisted by a trained moderator. The interviews took about 40 minutes 39 40 and the FGDs about one hour in average. The first and the last authors designed in-depth 41 on September 25, 2021 by guest. Protected copyright. 42 43 interview and FGD guides. We piloted the final version of the topic guides across the process of 44 45 data collection. 46 47 In-depth interview topic guides focused on khat use experiences, patterns of khat use through 48 49 50 time, reasons for use, perception about khat and khat use, criteria of problematic khat use, 51 52 behaviors after khat use and when they do not use and quitting experience from khat use. 53 54 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Focus group discussion guides were also similar focusing more on shared perceptions of normal 4 5 6 and problematic khat use, and community level harms. 7 8 Data management and analysis 9 10 11 The data were transcribed verbatim in Amharic, then translated into English by the first author 12 13 and experienced research assistants. All interview and FGD translations were coded 14 15 independently by the first and fourth author (AM and CN). Iterative thematic analysis (44) was 16 For peer review only 17 18 done simultaneously with data collection. We used computer software, open code 4.03 to 19 20 manage and analyse the data (45). 21 22 23 Regarding to data quality management and rigour, the iterative process of data collection, data 24 25 analysis and checking unclear issues from the participants added to the quality of the study. 26 Triangulation was achieved through data collection methods (IDI and FGDs) and data sources 27 28 (different groups of participants including religious fathers, key informants for the culture and 29 30 legal officers) and the team of researchers with diverse research experiences did crosschecking 31 32 the data and interpretations. The field worker, the first author, has previous experience http://bmjopen.bmj.com/ 33 interviewing people about khat use. Above all, we tried to purely and openly present the ideas of 34 35 the participants without personal impression interference. 36 37 Ethical considerations 38 39 40

Ethical clearance was obtained from the Institutional Review Board (IRB) of the College of on September 25, 2021 by guest. Protected copyright. 41 42 Health Sciences, Addis Ababa University (Ref 008/18/Psy). In addition, we also obtained a letter 43 44 of support from the Gurage zone health department. Written informed consent was sought and 45 46 obtained from all participants before data collection. For those participants who couldn’t read 47 and write, the primary author read the information sheet for them in the presence of a witness 48 49 who can read to confirm the veracity of the information, and those who agreed to participate in 50 51 the study signed on the consent form with their thumbprint. 52 53 54 55 56 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Patient and public involvement 4 5 No patient involved. 6 7 8 Findings 9 10 11 Characteristics of participants 12 13 14 Table 1 Socio-demographic characteristics of participants 15 16 Characteristics For peer reviewInterview only FGD 17 18 Age 3 8 19 20 20-34 21 22 35-44 5 2 23 24 45-59 6 8 25 26 60 and above 5 3 27 28 Gender 29 30 Male 16 15 31 32 Female 3 6 33 http://bmjopen.bmj.com/ 34 Residence 35 36 Urban 12 7 37 38 Rural 7 14 39 40 Marital status on September 25, 2021 by guest. Protected copyright. 41 Single 1 4 42 43 Married 17 15 44 45 Widowed or divorced 1 2 46 47 Education 48 49 Can’t read and write 3 2 50 51 Read and write only 1 2 52 53 Primary 2 5 54 55 Secondary 2 7 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Post-secondary 11 5 4 5 Religion 6 7 Muslim 8 9 8 9 Christian 11 12 10 11 Employment status 12 13 Self-employed 6 8 14 15 Unemployed 4 8 16 For peer review only 17 Formally employed 8 2 18 19 Student 1 3 20 21 Khat use status 22 23 Current khat users 14 21 24 25 Non-khat users 5 - 26 27 28 Participants were diverse in several socio-demographic characteristics, including sex, age, 29 30 31 residence, religion, occupation and educational status (Table 1). All non-khat users were males 32 33 and their age was in 50s years. They are religious fathers, police officers and culture experts. http://bmjopen.bmj.com/ 34 35 36 The major themes that emerged from the iterative thematic analysis were: (1) sociocultural khat 37 38 use, (2) suse “(addictive)” khat use, and (3) negative consequences of khat use. The second 39 40

theme had two categories: withdrawal experiences, quitting khat use and mirqanna/feeling high on September 25, 2021 by guest. Protected copyright. 41 42 43 after khat use. The negative consequences were categorized to mental health, sexual life, 44 45 physical health, social and financial. These themes are discussed with support of quotes directly 46 47 from the participants. 48 49 50 Sociocultural khat use 51 52 53 The sociocultural khat use theme emerged from participants’ notions of using khat for spiritual, 54 55 56 social and cultural reasons. Among Muslim participants, khat use was very common and 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 acceptable during religious rituals including the fasting month-remedan, holiday celebration and 4 5 6 anniversaries for the shrine of religious forefathers and at pilgrimage center. During these 7 8 situations, khat was one of the presents-zihara. 9 10 11 Many Muslim participants, especially elderly men, reported chewing khat for prayer. One 12 13 participant who was an elderly Muslim man described his use of khat as “… a prayer instrument 14 15 (duw’a)”. Some Muslim farmers who participated in this study reported chewing khat during 16 For peer review only 17 18 prayer as an inherited religious virtue-Ibada. Other participants described that the religious norm 19 20 of khat use for prayer was acceptable if it was in group chewing than chewing alone. A 21 22 participant making reference to the sword as a symbol for prayer alongside chewing khat shared 23 24 25 that: 26 27 28 “One sword can’t kill rather it only stabs. Thus, we usually chew being three or more” (Age late 29 30 50’s, male, urban) 31 32 33 In rural areas, unlike urban areas, participants shared that they chew khat and do religious http://bmjopen.bmj.com/ 34 35 rituals in mosques on weekly venerations of saints-wali. For example, Tuesdays had special 36 37 meaning because they were a memorial for Nurahusene, a religious forefather. Gathering in 38 39 40 mosques on special days of the week to chew khat and pray was part of the religious practice. 41 on September 25, 2021 by guest. Protected copyright. 42 43 “Although khat chewing doesn’t have any religious basis, we [Muslims] chew khat on the days 44 45 which are the shrine of our religious forefathers such as Esnel (Monday), Megergbia (Tuesday), 46 47 Gelale (Wednesday) and Sehiare (Saturday).” (Muslim religion scholar, age late 50’s) 48 49 50 Ceremonial khat use was also viewed as common practice at pilgrimage centres and during 51 52 53 annual Muslim religious holidays; Arafa (i.e. Id al-Adha), Id alfeter and moulid. In the study 54 55 setting, Aberiat, Qatbarie (Shaykh Isa Hamza), Alkeso, and Zebimola were frequently mentioned 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 as the most colourful wali venerating practices. Pilgrims had a religious ritual called hadra. It is 4 5 6 a ritual by small groups of pilgrims for religious ritual and khat chewing session. They did this 7 8 being in a separate partition in the mosque compound. 9 10 Khat use was not seen as limited to subgroups like the elderly and Muslim men, rather it was part 11 12 of the life of many people from different walks of life in the Gurage area. Many participants 13 14 15 agreed that, through time, khat was becoming acceptable and meaningful in the broader social 16 For peer review only 17 and cultural events. The participants agreed that, khat use was an important part of the Gurage 18 19 culture. This study indicated that khat chewing was shaping the day to day activities, beliefs, 20 21 22 tradition, and rituals of many people. Khat users, especially in rural areas, reported that khat 23 24 chewing was a significant part of their daily life. They frequently said that they used khat daily to 25 26 express both their mourning and happiness. 27 28 29 Participants reported that khat use was acceptable and to chew khat was to conform to societal 30 31 norms and values but there were specific nuances in how this was done to remain acceptable. 32 33 This was mainly not to have a socially agreed problems resulting from one’s chew of khat. The http://bmjopen.bmj.com/ 34 35 normal and accepted khat user was portrayed as “…a humble, honest, generous, and positive 36 37 38 thinker.” (Age in 40’s, rural, male) 39 40 41 Other definitions of normal khat use were related to patterns of use. Participants had set criteria on September 25, 2021 by guest. Protected copyright. 42 43 for who, when, how much and where to chew. For example, participants said it was part of their 44 45 culture not to chew khat in the morning, but recognised khat use in the afternoon. 46 47 48 Many participants did not accept daily khat use as normal. Normal khat users were also 49 50 51 supposed to limit the amount of their khat consumption. Chewing in public places was also 52 53 acceptable and normative among men but considered abnormal among women. 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 This was shared by one participant in the focus group discussions and they highlighted: “there 4 5 6 are males who chew khat in cafeterias and verandas, but a woman does not chew in public 7 8 places even they don’t buy and hold khat, except commercial sex workers.” (Age early 30’s, 9 10 female, urban) 11 12 13 We found much restriction of the norms among rural than urban participants. Participants from 14 15 rural areas emphasized that as areqi (a local illicit alcoholic drink) have been part and parcel of 16 For peer review only 17 18 the life of many Christians and the same was true for khat among Muslims. But now, khat use 19 20 had become a normal behaviour of both Muslims and Christians in both urban and rural areas. 21 22 Use was considered normal as long as it did not come with other problems. 23 24 25 Many participants also shared that sociocultural khat use was an accepted way of use valued in 26 27 28 their community. Examples of sociocultural uses that emerged from the analysis included: 29 30 hosting visitors, grieving period after funeral, work parties, as part of normal socialisation, 31 32 weddings and in rare cases when paying visit to the sick. A participant highlighted as: “during 33 http://bmjopen.bmj.com/ 34 35 condolence after funeral, wedding ceremony…when we chew khat, it will be very easy to 36 37 socialize and easily close with others. In our home, we feel confident to host guest if we can buy 38 39 khat otherwise I don’t say it will be full hospitality.” (Age early 30’s, male, urban) 40 41 on September 25, 2021 by guest. Protected copyright. 42 Despite the normative khat use reported, the participants also shared insights around what 43 44 constituted problematic khat use. 45 46 47 48 49 Khat suse-addiction 50 51 52 Most participants conclusively stated that khat use is an addictive behavior-suse. This was the 53 54 most commonly reported reason to continue chewing by many participants so that they could 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 accomplish their day to day activities. Different withdrawal experiences (presented in the next 4 5 6 section) were reported to justify the suse or addiction to khat use. At the addictive stage, those 7 8 affected are seen by keeping some khat in their pockets and chew almost all the time without 9 10 regard for place too. This included chewing at school, funeral, market, and workplace either 11 12 office or farm. A participant explained khat suse/addiction as; 13 14 15 I should have khat in my bloodstream to open my eyes and to activate my body. Morning khat 16 For peer review only 17 18 use-ejebena is a sign of khat suse, isn’t it? (Age in 40’s, urban, male) 19 20 21 It was then this type of use that was viewed as causing problems to the health and social aspects 22 23 of both the user and those around him/her. 24 25 26 Participants also mentioned harara/craving as indicator of khat suse/addiction. Many said that 27 28 khat use is an addictive behaviour because it has a strong harara/craving. They justified that the 29 30 31 strong harara/craving even forced people to beg for money to buy khat when they did not have 32 33 money or they would collect the leftover khat and chew. When these behaviours start to show, http://bmjopen.bmj.com/ 34 35 the khat chewing is then considered abnormal and problematic. 36 37 38 Those who chew because of harara/craving, unlike those who chew for prayer, chew khat 39 40

regularly, any time including in the morning, chew without ritual or ceremony and without on September 25, 2021 by guest. Protected copyright. 41 42 43 washing themselves. All these patterns of use are abnormal in my culture. (Age late 50’s, male, 44 45 urban) 46 47 48 Chewing khat by alms was perceived as the last and worst stage of a khat user who was called 49 50 Jezba. 51 52 53 Participants used the term Jezba to label subgroups of people with khat suse or addiction. 54 55 56 Jezbas are people with khat suse or addiction and characterized syndrome of the behavior. Jezba 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 syndrome indicates; poor connection with God, others and material needs, not confirming to 4 5 6 shared values and norms of the society. Jezba khat users prioritize khat use rather than other 7 8 important aspects of life or they have limited interest for other important areas of life except 9 10 khat. They also had reduced motivation for work and other social life or total disengagement 11 12 when the khat is withdrawn as well as frequent work absenteeism or abandoning work. It also 13 14 15 includes; deterioration in critical thinking, poor social skills/self-care, beg khat or money for 16 For peer review only 17 khat, external attribution, belief of inability or no interest to control khat use and related 18 19 behaviors, chewing for many hours of the day (more than 6 hours) and on the street or while 20 21 22 walking. 23 24 25 In the long term, khat will make you Jezba-ያጀዝብሀል. Jezba doesn’t value anything important 26 27 other than khat. (Age in 40's, rural, male) 28 29 30 For many participants, the use of khat against sociocultural conditions was considered as a sign 31 32 of khat suse/addiction. This includes chewing khat for individual reasons such as to cope with 33 http://bmjopen.bmj.com/ 34 35 adverse life experiences while cultural khat users chew for communal reasons; for prayer, 36 37 socialization or companionship, and functional reasons. 38 39 40 Khat addiction was also characterized in terms of chewing increased amount of khat through 41 on September 25, 2021 by guest. Protected copyright. 42 time as well as chewing for longer time to get the desired effect. 43 44 45 When you chew too much [greater than a bundle of khat], it will cause many problems; being 46 47 addictive could be one. Only limited amount of khat benefits [the stimulating effect]. (Age late 48 49 50 50’s, male, rural) 51 52 53 In other views, some participants said that increased use of khat overtime depended on one's 54 55 mindset or psychological expectation, amount of budgeted time for chewing, emotional state or 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 group cohesion during chewing and financial capacity to buy khat. Age was also mentioned as an 4 5 6 important factor for heavy and problematic use of khat. Youths and early adult participants 7 8 showed more addictive use of khat, unlike the elderly. 9 10 11 Withdrawal experiences of khat use 12 13 14 Other experiences reported associated with abnormal khat use included withdrawal experiences 15 16 when one was weaningFor off khat peer after heavy review usage. The study onlyfound many important 17 18 psychological withdrawal experiences of khat suse-addiction. Feeling depressed, irritable and 19 20 21 aggression were the most typical and commonly reported experiences. One participant said, “I 22 23 am usually against people’s communication even aggressive when I didn’t chew khat. I 24 25 remember that I once threw away [smacked] my kid when she was talking about her school 26 27 28 affairs.” (Age late 50’s, urban, male) 29 30 31 Different participants reported different withdrawal symptoms, including lack of motivation, 32 33 unable or lack interest to function or socialize, poor concentration and learning, unable to receive http://bmjopen.bmj.com/ 34 35 message, and dukak (vivid unpleasant dream). Here are two dukak or vivid unpleasant dream 36 37 experiences; 38 39 40

One day my husband skipped chewing khat and went to bed. Then, he spent the night spitting. In on September 25, 2021 by guest. Protected copyright. 41 42 43 the morning, the bed sheet was wet. When I ask him what was wrong with him. He said; “people 44 45 were punishing me with the smoke of red paper [forcing people to inhale the smoke of red paper 46 47 is one of harsh corporal punishments among a few in the culture] and I had been feeling burning 48 49 50 sensation for the whole night”. (Age early 30’s, urban, women) 51 52 53 Another experience: 54 55 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 My experience was...ehm... usually a man would hold my hair and hung me or put me into a hole 4 5 6 and I would wake up in panic. (Age in 30’s, male, urban) 7 8 9 Some participants reported physical withdrawal experiences, including abdominal pain, 10 11 headache, being drowsy, red eyes, increased appetite and sleep, yawning, uncontrollable tears, 12 13 shivering hands, and loss of energy. Another interesting finding was an observation that to 14 15 manage some withdrawal syndromes, some khat users opted to sleeping. During fasting season, 16 For peer review only 17 18 many Muslim participants did not use khat either in the morning or for the whole day, but they 19 20 would usually spent the day sleeping. A Muslim participant shared his experience and said; 21 22 23 During remedan [fasting month], we don’t chew khat during the day time, but no one did a 24 25 serious work. We spend the day sleeping. If you don’t chew khat, you cannot be stimulated, 26 27 28 energetic..... (Age late 50’s, male, rural) 29 30 31 Inability to control khat use 32 33 http://bmjopen.bmj.com/ 34 Quitting khat use was usually contingent on the participants’ lifestyle and pattern of khat use. For 35 36 participants who didn’t chew daily, quitting was perceived as an easy task, but it was very 37 38 challenging to participants without job because khat use would be an important leisure activity. 39 40

External factors; culture and pressure from others, were also reported challenges to quit from on September 25, 2021 by guest. Protected copyright. 41 42 43 chewing. Some thought that quitting khat use also includes quitting important social networks. 44 45 Thus, they continue chewing and consider khat use as a matter of surviving in the social system. 46 47 48 Some participants reported as they forward a rational decision to continue chewing after 49 50 evaluating the benefits and harms of khat use. Participants who perceive more benefits, than 51 52 53 harms from khat use, such as positive effect on motivation, work performance and socialization 54 55 continue chewing khat. 56 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 A participant stated as follows; 4 5 6 I will not quit chewing because it helps me to share information, got social support and other 7 8 9 things as well as to relieve stress and worries in life. (Age in 30’s, urban, male) 10 11 12 Many participants said that only a few and the fortunate could quit early, but many realize the 13 14 harms later. Many quit chewing when they lose satisfaction from chewing, usually at the end of 15 16 their life. Hence chewingFor until peer the end of lifereview was considered onlyan abnormal or addictive practice 17 18 of khat use. 19 20 21 After years, one would lose satisfaction from chewing. You would lose the passion to chew khat 22 23 24 and you will decide to say it is time to quit. Only the unfortunate will chew to the end of their life. 25 26 (Age late 50’s, rural, male) 27 28 29 Mirqanna/feeling high after khat use 30 31 32 Many of the khat user participants liked the stimulant effect of khat, but when the stimulation 33 http://bmjopen.bmj.com/ 34 was too high, they reported feelings of distress. Mirqanna (induced distress) was due to chewing 35 36 37 excessive amounts or chewing in combination with shisha or cigarettes. This was repeatedly 38 39 reported among participants in their early phases of khat use. Insomnia induced by mirqanna was 40 41 also a perceived cause of distress. Many khat users would then resort to drinking alcohol to break on September 25, 2021 by guest. Protected copyright. 42 43 the mirqanna and induce sleep. Other khat users, experiencing mirqanna but cannot afford to 44 45 46 drink alcohol because of lack of money, reported continued feelings of being restless, being on 47 48 the move and confusion during. One such participant shared; 49 50 51 One day, I chewed khat for five hours. Then at night, I couldn’t sleep. I had spent the night 52 53 itching, feeling fever and sweating. I was very panic about my condition, but I was also feeling 54 55 56 fatigue and low energy to treat myself. (Age in 30’s, urban, male) 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Additional symptoms of mirqanna reported by participants included; being absorbed to an 4 5 6 inauthentic personal world such as considering oneself as fortunate, being extremely humble, and 7 8 considering oneself as high achiever, more like delusions. In some participants, they reported 9 10 lack of interest to communicate during mirqanna state, though talkativeness was also reported 11 12 among others. Excessive fear including avoiding any exposure or engagement with others as well 13 14 15 as fear to make decisions was also common. Participants emphasized that the mirqanna 16 For peer review only 17 experiences were different from their experiences of intoxication due to excessive alcohol 18 19 drinking. 20 21 22 When it [mirqanna] is severe, when one chews too much, one might spend the night outside the 23 24 25 house, on the street, which is risky because wild animals could harm him. Sometimes alcohol 26 27 intoxication might be better than khat mirqanna. The intoxication of alcohol can be reversed 28 29 soon with different techniques including getting sleep, but the khat mirqanna couldn't be 30 31 32 reversed easily once the person is at the severe stage. (Age late 50’s, rural, male) 33 http://bmjopen.bmj.com/ 34 35 Other participants thought that severe mirqanna is like psychosis state-qezete. Mumbling alone, 36 37 confusion, spending the night on the street were perceived symptoms of psychosis induced by 38 39 khat. The following quote was from the experience of one participant. 40 41 on September 25, 2021 by guest. Protected copyright. 42 When I am at mirqanna state, I imagine as I own a big building, big car…By the next day, I 43 44 realize everything was a fantasy. All were gone as cloud. (Age late 50’s, urban, male) 45 46 47 Mental health related consequences 48 49 50 Participants reported that khat could lead to mental health disorders. Depression-debert and 51 52 53 psychosis-qezete were commonly reported mental health disorders. Participants indicated that 54 55 depression is associated with khat withdrawal or secondary to different crisis especially 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 financial, caused by khat use. Psychosis which is known by different phrases such as qezete (the 4 5 6 acute form of psychosis), chereken metal, aemeron mesat, yeaemero menawet were common 7 8 among participants who chew too much amount of khat and don’t have meal before chewing. A 9 10 participant described this as follows; 11 12 13 …only individuals who had nutritional capacity- body fluid can resist the adverse effect of the 14 15 khat. The brain will be vulnerable when one chews without having a meal. People who chew 16 For peer review only 17 18 without having a protective, food, will be negatively affected by the khat. (Age in 30’s, rural, 19 20 male) 21 22 23 Participants reported that different behavioral symptoms of chronic khat users could be similar 24 25 with the behaviors of people with severe mental illness. They further described that among 26 27 28 chronic khat users, like people with severe mental disorders, there was broken down of family 29 30 and other social activities including poor self-care and dressing. In addition, they chew khat on 31 32 the street and they also chew leftover khat or beg money for khat. 33 http://bmjopen.bmj.com/ 34 35 Social related consequences of khat use 36 37 38 Another important problem of khat use was its social harms such as family chaos and 39 40

breakdowns. This was related with spending too much of the family budget on khat and on September 25, 2021 by guest. Protected copyright. 41 42 43 abandoning their responsibilities in the family. Negative behaviors of participants with khat use 44 45 such as irritability also led to family conflict. Two participants’ experiences are stated as follows; 46 47 48 I am not giving time for my kids. I just delegate my elderly daughter to take care and control her 49 50 siblings. (Age in 40’s, urban, male) 51 52 53 I have been forced by my wife to stop chewing. I don’t want to quit, but my wife is insisting me. 54 55 56 We have been quarrelling and separated due to this issue. Her parents came to mediate us and 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 asked me to quit chewing, but I told them my position that I got the khat before her and now I am 4 5 6 not interested to stop chewing…….I will not stop. Imagine? (Age in 30’s, urban, male) 7 8 9 Among urban dwellers, problem to discharge their government responsibilities were commonly 10 11 reported. Some reported their experience of being blamed by frequent work absenteeism and 12 13 work inefficiency because they lack the motivation to go to work when they don’t chew khat. 14 15 They usually go for khat abandoning their work. 16 For peer review only 17 18 Khat suse/addiction sometimes led to crime. Some, especially in rural areas, usually quarrel 19 20 21 with others. The main source of reported conflict was khat stolen. Participants observed that khat 22 23 users who don’t have money to buy usually theft someone's khat from his/her farm. Among 24 25 urban participants, few khat users sometimes also stole others’ properties. Others sell their home 26 27 28 utensils to get money for khat without the consent of their family members. Theft and violence 29 30 were common among khat users who drink alcohol excessively. 31 32 33 Financial consequences of khat use http://bmjopen.bmj.com/ 34 35 36 Participants were concerned about the price of khat. The minimum reported daily expense was 37 38 about 1.5 USD for a bundle of khat, but many also spent for alcohol, coffee, shisha and cigarette. 39 40

Some participants, at sever stage, had sold assets or home utensils and spend the money for khat. on September 25, 2021 by guest. Protected copyright. 41 42 43 In addition to the direct adverse effect of khat use on finance, participants were also concerned 44 45 46 about the amount of time they spend by chewing khat. In order to get the desired stimulation 47 48 from khat, they were supposed to stay chewing khat for long hours; the longest was 6 and more 49 50 hours. Participants question how much they could earn if they didn't spend their time by chewing 51 52 53 khat. Others who chew being in a group usually spend more time and unable to quit the session 54 55 and go for work being attracted by the fun and the chat. 56 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 The financial advantage from khat was reported from participants who have khat farm and sell 4 5 6 khat. They were relatively better in terms of financial capacity. For them, khat is an important 7 8 cash crop which allows them to cover home expenses and pay government tax easily while 9 10 others were under stress. 11 12 13 Physical health related consequences 14 15 Many of the participants admitted that after khat use, it is common to experience loss of 16 For peer review only 17 18 appetite. Thus, weight loss and malnutrition were commonly reported. 19 20 21 Khat absorbs your body fluid and makes you dry so that a khat user is thin and latter would be 22 23 vulnerable to different diseases. (Age in 20’s, rural, male) 24 25 26 Gastrointestinal and oral health problems were frequently reported. Bad odour of the mouth, a 27 28 colour change of the tongue and teeth, and teeth spoil were among the major complaints. 29 30 31 Chronic khat user participants couldn't prefer sauce and beef, which are common and valued in a 32 33 routine dish of the culture because of the burning sensation of their mouth and teeth damage. http://bmjopen.bmj.com/ 34 35 Dehydration and constipation were also other major complaints. During khat withdrawal phase; 36 37 general physical pain, severe headache, burning sensation, and redness of the eyes were common 38 39 40 complaints. 41 on September 25, 2021 by guest. Protected copyright. 42 43 For participants from Wolketie, chewing khat and drinking alcohol were considered as risky 44 45 behaviour for unsafe sex and thus HIV/AIDS infection. Many were also excessive alcohol users 46 47 and vulnerable to all the adverse effects of excessive alcohol use. The leftover khat disposed 48 49 50 elsewhere in the city was also another concern for their health. Participants rarely reported 51 52 accidents and injuries related to khat use. Table 2 below shows major indicators for normal and 53 54 problematic khat use. 55 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Page 25 of 41 BMJ Open

1 2 3 Table 2; A conceptual summary of normative and problematic khat use 4 5 6 Parameters Normal khat use Problematic khat use 7 8 The reason of khat use -Khat use for prayer, leisure, functional and other social -Chewing to manage personal pain and distress activities 9 -Khat use is an end by itself 10 -Khat use is a means to an end 11 -Continue khat use because of dependency, craving 12 -Continue khatFor use to conform peer to the social normsreview only 13 14 15 Who chews khat -Healthy male adults, rarely women from cities -Khat use by women and children was perceived as problematic 16 especially in rural areas 17 -The community doesn’t recommend khat use by persons 18 with mental illness and with other critical health http://bmjopen.bmj.com/ 19 20 conditions 21 22 Frequency of khat use -Infrequently where the maximum was three times per -Regularly week 23 - Almost daily 24 -Situation or event led khat use 25 -Many can’t skip for a day or their fixed daily khat chewing 26 on September 25, 2021 by guest. Protected copyright. 27 session 28 29 Amount of khat -Limited amount; after chewing few leaves, they could -Chew increased amount of khat compared to their friends and 30 divert their attention from the khat to their work couldn’t divert their attention except chewing 31 32 -Some had been chewing a lot -Long sessions such as half a day or more 33 34 - Short sessions, long sessions were for recreational users -Less sever problematic khat users chew while accomplishing 35 their routines 36 37 Other contexts of khat -Chewing after meal -Chewing even when there is no meal or usually skip meal 38 use 39 -The chewing pattern is in line with the social norm -The chewing pattern is deviated from the societal norm 40 (place, time, situation) 41 42 43 24 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

BMJ Open Page 26 of 41

1 2 3 -Favorable attitude from others -Negative attitude from others 4 5 Khat related benefits and -Perceived benefits or minor harms -Health, social and economic harms including malnutrition and 6 harms reduced body weight as well as different physical health 7 -There is normal functioning or productivity 8 complaints, poverty, family break up, separated from social 9 - From mild to strong level of social support support system and living on the street, begging khat or theft for 10 daily khat consumption or collecting left over khat, depression, 11 -khat is an agent of survival in the community idleness, violating religious ritual e.gsalat-prayer 12 For peer review only 13 14 15 Settings of khat use -Group and ceremonial -Alone and no ceremony 16 17 -Sometimes people chew being alone and without ritual -If people chew together there is no sense of belongingness 18 http://bmjopen.bmj.com/ 19 -Chewing in home, mosque, or other special places in the -Chewing in home, khat cafeterias, public chewing including on 20 community the street and at work 21 22 Session and time of - Deciding the session of khat use before time usually -Session is not common and appropriate 23 chewing khat when the group have convenient situations 24 -If there is a session, it is long and couldn’t leave the session 25 26 -Afternoon or rarely at night when other important personal on September 25, 2021 by guest. Protected copyright. and social affairs emerge 27 - Chewing khat in the morning, morning and afternoon, many 28 29 hours of the day, sometimes they also chew at night 30 31 Value of chewing khat - Symbolic and reality perception about khat use - Reality perception about khat use 32 33 - Value things which are contingent on khat use -Value the khat use itself 34 -Khat is most valued at different sociocultural situations; -Companionship with the khat 35 36 hospitality, work party, mourning, wedding and spiritual 37 practise because it induces alertness, energy, 38 concentration, open discussion 39 40 -Companionship with the group, with the prayer, 41 42 43 25 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Page 27 of 41 BMJ Open

1 2 3 discussion 4 5 - There is pleasure and utility associated with khat use 6 7 8 9 10 Suse/addiction -Khat use results in harmony in the family and society - Associated harms on family and social, work 11 12 -There is no orFor minor craving peer review-Unable only to function without khat 13 14 -people experience well, minor and transient experiences -There is craving-harara 15 of distresses before and after khat use 16 -Serious withdrawal experiences and distressing experiences after 17 - Quitting is easy, but there is no frequent thought or khat use /mirqanna 18 attempt about quitting http://bmjopen.bmj.com/ 19 - Quitting is difficult because of the distressing experiences when 20 the khat is withdrawn and persons usually think or attempt to quit 21 22 Associated behaviours - Some use, other psychoactive substances, others don’t -Khat use was in combination with smoking cigarettes, shisha and 23 drinking alcohol 24 25

Khat use and the broader -Careful for other important areas of life; self-care, deity -Abandonment of other important on September 25, 2021 by guest. Protected copyright. areas of life including poor self- 26 area of life care and frequently skip a meal 27 -Purpose and meaning in life since there is fun, 28 29 socialization, social support -A problem in purpose and meaning in life as there are depression, and other withdrawal experiences as well as poor social support 30 -A motivation for many areas of life 31 32 -Amotivational syndrome 33 34 35 36 37 38 39 40 41 42 43 26 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 28 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Discussion 4 5 6 In this study, our aim was to conceptualize problematic khat us in a predominantly rural setting 7 8 9 in south-central Ethiopia. The study helped to answer the question: what constitutes problematic 10 11 khat use? Our results could inform development of screening tool to measure problematic khat 12 13 use. 14 15 16 Although there wasFor also peer sociocultural review khat use pattern, only khat addiction and negative 17 18 consequences of khat use constitute problematic khat use. The local term khat suse, semantically 19 20 21 equivalent to khat addiction which doesn’t conclusively infer to DSM-5 stimulant use disorder 22 23 definition(46). Khat suse/khat addiction could only qualify impaired control and pharmacological 24 25 criteria among the criteria of stimulant use disorders. Khat suse/ khat addiction shares some 26 27 28 characteristics from other substances use disorders than stimulant use disorders. For example, it 29 30 has a similar functional consequences of cannabis use disorder-amotivational syndrome(46). In 31 32 the case of khat suse, the local idiom Jezba could be conceptually related but broader. In the 33 http://bmjopen.bmj.com/ 34 35 current study setting, the use of the word, Jezba, indicates the existing stigma and it could also be 36 37 a good explanation how much sever form of problematic khat use is well recognized in the 38 39 setting since it has a negative connotation. Khat suse is also indicated by frequent yawning when 40 on September 25, 2021 by guest. Protected copyright. 41 not using the khat. This withdrawal criterion is similar with opioid withdrawal(46). 42 43 44 Major negative consequences of khat use, which are other indicators of problematic khat use, 45 46 47 include sexual dysfunction, depression, psychosis, various oral health problems, and wastage of 48 49 time. Many of these indicators of problematic khat use are reported in several previous studies 50 51 (35, 39, 47). In other settings, chewing khat with sugar was stated as a factor for some health 52 53 54 problems especially oral health problems(48). But in the current study the use of sugar with khat 55 56 and sugary drinks is not common in rural areas because of cost and availability. They even drink 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 coffee without sugar; sometimes they add a pinch of salt to their coffee, but drink mostly without 4 5 6 sugar. 7 8 9 Frequency and amount are important predictors of problematic alcohol and cannabis use(49, 50) 10 11 so that similar inference could be applied for problematic khat use. Studies also estimate safe 12 13 limits of alcohol use(51) and cannabis use(52). This qualitative study explored that there are 14 15 normal and problematic khat use patterns, and negative consequences are more likely to be 16 For peer review only 17 18 experienced among problematic khat users who are using khat frequently and in large amount 19 20 although there are additional problematic khat use criteria. Therefore, future studies could 21 22 investigate “safe limit” of khat use to inform measures of khat harm reduction as it is done for 23 24 25 cannabis (53). 26 27 28 The sociocultural khat use pattern was considered as normative khat use. This has limits in 29 30 terms of amount, frequency, duration of khat session and contexts of use. People chew khat 31 32 during weddings, funeral ceremonies, working, social meetings, and other cultural activities in 33 http://bmjopen.bmj.com/ 34 35 different settings both in Ethiopia(54, 55), and elsewhere such as in Kenya (56), Somaliland (21), 36 37 Somalia (57), Yemen (22), other African and Middle Eastern countries(58), and among 38 39 immigrants in the West (59-61). Different factors such as accessibility and availability, social 40 on September 25, 2021 by guest. Protected copyright. 41 accommodation and cultural acceptable could facilitate the process of drug normalization in 42 43 44 general (62) and khat use in particular(54). 45 46 47 The study had different implications. Intervention for problematic khat use should be 48 49 systematically designed and planned. One systematic review (63) indicated community, family 50 51 and individual level interventions were acceptable and showed modest efficacy. This qualitative 52 53 54 study adds to the findings of the systematic review that the sociocultural background of khat use 55 56 needs to be considered in designing policies and implementing interventions. Since there is 57 58 28 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 normative khat use in the current study setting as well as in other settings, mentioned above, 4 5 6 abstinence might not be effective in addressing problematic khat use pattern(63). 7 8 9 Above all, problematic khat use should be an important component of substance use disorders or 10 11 mental health care system of the country. Practitioners should be committed to screen and offer 12 13 interventions for people with problematic khat use. Culturally adapted and psychometrically 14 15 valid screening tools for problematic khat use would be the priority to facilitate the clinical 16 For peer review only 17 18 practice and further research. Future studies would also play an important role to produce and 19 20 adapt evidence based interventions. The current study, aligned with previous studies, informs 21 22 policy makers to focus on khat use regulation and harm reduction strategies. 23 24 25 Many of the findings about problematic khat use could be transferable. Khat addiction, withdrawal 26 27 28 experiences of khat use, inability to control khat use, mirqanna/feeling high after khat use and 29 30 impacts of khat use are transferable. Others cultural basis of khat use could be understood in the 31 32 study setting context. 33 http://bmjopen.bmj.com/ 34 35 Conclusion 36 37 38 The study has illustrated what constitutes normative [acceptable] and problematic khat use in the 39 40

Gurage community in south-central Ethiopia. We found that problematic khat use is on September 25, 2021 by guest. Protected copyright. 41 42 43 characterized by patterns of use, reasons for use, contexts or norms, adverse psychological 44 45 reactions after use, khat suse/addiction, and khat use-related harms. The study would inform 46 47 future studies on development of tools to measure problematic khat use. The study will also be 48 49 50 used as a formative study for future longitudinal and intervention studies focusing on estimating 51 52 and addressing multidimensional impacts of problematic khat use. 53 54 55 Acknowledgements 56 57 58 29 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 The authors would like to acknowledge study participants and the African mental health research 4 5 6 initiative (AMARI) of DELTAS Africa Initiative. 7 8 9 Funding: 10 11 12 This work was supported through the DELTAS Africa Initiative [DEL-15-01]. The DELTAS 13 14 Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s 15 16 Alliance for AcceleratingFor Excellence peer in Science review in Africa (AESA) only and supported by the New 17 18 Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) 19 20 21 with funding from the Wellcome Trust [DEL-15-01] and the UK government. The views 22 23 expressed in this publication are those of the author(s) and not necessarily those of AAS, 24 25 NEPAD Agency, Wellcome Trust or the UK government. 26 27 28 Authors’ contributions: 29 30 31 AM ST AF conceived and designed the study. AM did the interview. AM and CN coded the 32 33 http://bmjopen.bmj.com/ 34 data. AM AF KH CN SN ST contributed to the analysis and the write-up of the manuscript. All 35 36 authors agree with the results and conclusions of the study. 37 38 39 Ethical approval and consent to participate 40 41 on September 25, 2021 by guest. Protected copyright. 42 The study was approved by the Institutional Review Board of the College of Health Sciences of 43 44 Addis Ababa University (REF. 008/18/psy). Participants took part in the study after providing 45 46 written informed consent. 47 48 49 Competing interests: 50 51 52 None declared 53 54 55 Data sharing statement: No additional unpublished data are available from this study. 56 57 58 30 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 References 4 5 6 1. Cox G, Rampes H. Adverse effects of khat: A review. Advances in Psychiatric Treatment. 7 2003;9(6):456-63. 8 2. Krikorian AD. Kat and its use: an historical perspective. J Ethnopharmacol. 1984;12(2):115-78. 9 10 3. Kassim S, Dalsania A, Nordgren J, Klein A, Hulbert J. Before the ban--an exploratory study of a 11 local khat market in East London, U.K. Harm Reduct J. 2015;12:19. 12 4. Douglas H, Hersi A. Khat and islamic legal perspectives: issues for consideration. The Journal of 13 Legal Pluralism and Unofficial Law. 2010;42(62):95-114. 14 5. EMCDDA. European Monitoring Center for Drugs and Drug Addiction. (2011). Drugs in focus. 15 ISSN 1681- 157. 211. 16 6. Numan N. ExplorationFor ofpeer adverse psychological review symptoms only in Yemeni khat users by the 17 Symptoms Checklist-90 (SCL-90). Addiction. 2004;99(1):61-5. 18 19 7. Elmi AS. The chewing of khat in Somalia. Journal of ethnopharmacology. 1983;8(2):163-76. 20 8. Haile D, Lakew Y. Khat chewing practice and associated factors among adults in Ethiopia: further 21 analysis using the 2011 demographic and health survey. PloS one. 2015;10(6):e0130460. 22 9. Teklie H, Gonfa G, Getachew T, Defar A, Bekele A, Bekele A, et al. Prevalence of Khat chewing 23 and associated factors in Ethiopia: Findings from the 2015 national Non-communicable diseases STEPS 24 survey. Ethiopian Journal of Health Development. 2017;31(1):320-30. 25 10. Alem A, Kebede D, Kullgren G. The prevalence and socio-demographic correlates of khat 26 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica, Supplement. 1999;99(397):84-91. 27 11. Alem A, Kebede D, Kullgren G. The prevalence and socio-demographic correlates of khat 28 29 chewing in Butajira, Ethiopia. Acta Psychiatr Scand Suppl. 1999;397:84-91. 30 12. Alem A, Kebede, D., & Kullgren, G. . The prevalence and socio-demographic correlates of khat 31 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica Supplementum. 1999;100:84-91. 32 13. Adugna F, Jira C, Molla T. Khat chewing among Agaro secondary school students, Agaro, 33 southwestern Ethiopia. Ethiopian Medical Journal. 1994;32(3):161-6. http://bmjopen.bmj.com/ 34 14. al. Ge. Prevalence and Predictors of harmful Khat use among university Students in Ethiopia. . 35 Substance Abuse: Research and Treatment 2014;8:45-51. 36 15. Patel SL, Murray R, Britain G. Khat use among Somalis in four English cities: Citeseer; 2005. 37 38 16. Gebissa E. Scourge of life or an economic lifeline? Public discourses on khat (Catha edulis) in 39 Ethiopia. Substance Use & Misuse. 2008;43(6):784-802. 40 17. Haile D, Lakew Y. Khat chewing practice and associated factors among adults in Ethiopia: Further 41 analysis using the 2011 demographic and health survey. PLoS ONE. 2015;10 (6) (no on September 25, 2021 by guest. Protected copyright. 42 pagination)(e0130460). 43 18. Hersi MA, Abdalla M. Sharī ‘a Law and the Legality of Consumption of Khat (Catha Edulis): Views 44 of Australian Imāms. 2013. 45 19. Geda GJ. Pilgrimages and Syncretism: Religious transformation among the Arsi Oromo of 46 47 Ethiopia 2015. 48 20. Griffioen S. Mohammed Girma, Understanding Religion and Social Change in Ethiopia. Toward a 49 Hermeneutic of Covenant. Palgrave Macmillan, New York, 2012. 240 pages. ISBN 978-1-137-269416. 50 Philosophia Reformata. 2013;78(2):222-5. 51 21. Hansen P. The ambiguity of khat in Somaliland. Journal of ethnopharmacology. 2010;132(3):590- 52 9. 53 22. Al-Motarreb A, Baker K, Broadley KJ. Khat: pharmacological and medical aspects and its social 54 use in Yemen. Phytotherapy research. 2002;16(5):403-13. 55 56 57 58 31 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 23. Mihretu A, Teferra S, Fekadu A. What constitutes problematic khat use? An exploratory mixed 4 methods study in Ethiopia. Substance Abuse Treatment, Prevention, and Policy. 2017;12(1):17. 5 6 24. WHO. Assessment of Khat (Catha edulis Forsk) Geneva: WHO. . 2006. 7 25. Asgedom SW, Gudina EK, Desse TA. Assessment of Blood Pressure Control among Hypertensive 8 Patients in Southwest Ethiopia. PloS one. 2016;11(11):e0166432. 9 26. Al-Hadrani AM. Khat induced hemorrhoidal disease in Yemen. Saudi Medical Journal. 10 2000;21(5):475-7. 11 27. Hassen K, Abdulahi M, Dejene T, Wolde M, Sudhakar M. Khat as a risk factor for hypertension: A 12 systematic review. JBI Database of Systematic Reviews and Implementation Reports. 2012;10(44):2882- 13 905. 14 28. Al-Habori M. The potential adverse effects of habitual use of Catha edulis (khat). Expert opinion 15 16 on drug safety. 2005;4(6):1145-54.For peer review only 17 29. Hassan NA, Gunaid AA, Abdo-Rabbo AA, Abdel-Kader ZY, Al-Mansoob MA, Awad AY, et al. The 18 effect of Qat chewing on blood pressure and heart rate in healthy volunteers. Tropical doctor. 19 2000;30(2):107-8. 20 30. Mikulica J, Odenwald M, Ndetei D, Widmann M, Warsame A, al'Absi M, et al. Khat Use, PTSD 21 and Psychotic Symptoms among Somali Refugees in Nairobi-A Pilot Study. 2014. 22 31. Bhui K, Craig T, Mohamud S, Warfa N, Stansfeld SA, Thornicroft G, et al. Mental disorders among 23 Somali refugees. Social psychiatry and psychiatric epidemiology. 2006;41(5):400-8. 24 25 32. Odenwald M, Neuner, F., Schawer, M., Elbert, T.R., Catani, C., Lingenfelder, B., Hinkel, H., 26 Hafner, H. & Stroh, B. . Khat use as risk factor for psychotic disorders: A cross-sectional and case-control 27 study in Somalia. . BMC Medicine. 2005;3(5). 28 33. Hassan NA, Gunaid AA, El-Khally FM, Murray-Lyon IM. The effect of chewing Khat leaves on 29 human mood. Saudi medical journal. 2002;23(7):850-3. 30 34. Bhui K, Warfa N. Trauma, khat and common psychotic symptoms among Somali immigrants: A 31 quantitative study. J Ethnopharmacol. 2010;132(3):549-53. 32

35. Mihretu A, Nhunzvi C, Fekadu A, Norton S, Teferra S. Definition and Validity of the Construct http://bmjopen.bmj.com/ 33 “Problematic Khat Use”: A Systematic Review. European addiction research. 2019;25(4):161-72. 34 35 36. Duresso S, Matthews, A., Ferguson, S. & Bruno,R. Is khat use disorder a valid diagnostic entity? . 36 School of Medicine, University of Tasmania, Hobart, Australia. 2015. 37 37. Gebrehanna E, Berhane Y, Worku A. Khat chewing among Ethiopian University Students--a 38 growing concern. BMC public health. 2014;14:1198. 39 38. Kassim S, Islam S, Croucher R. Validity and reliability of a Severity of Dependence Scale for khat 40 (SDS-khat). Journal of Ethnopharmacology. 2010;132(3):570-7. on September 25, 2021 by guest. Protected copyright. 41 39. Mihretu A, Teferra S, Fekadu A. What constitutes problematic khat use? An exploratory mixed 42 methods study in Ethiopia. Substance Abuse Treatment, Prevention, and Policy. 2017;12:17. 43 44 40. CSA E. 2014 projected population Size of Towns by Sex, Region, Zone and Wereda 2015. 45 41. Crass J, Meyer R, editors. The Qabena and the Wolane: Two peoples of the Gurage region and 46 their respective histories according to their own oral traditions. Annales d'Éthiopie; 2001: Editions de la 47 Table Ronde. 48 42. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 49 32-item checklist for interviews and focus groups. International journal for quality in health care. 50 2007;19(6):349-57. 51 43. Patton MQ. Qualitative research & evaluation methods: Integrating theory and practice: Sage 52 53 publications; 2014. 54 44. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative research in psychology. 55 2006;3(2):77-101. 56 45. Umea° University: UMDAC and Epidemiology DoPHaCMaUU. Open Code version 4.03. 2013. 57 58 32 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 46. Association AP. Diagnostic and statistical manual of mental disorders (DSM-5®): American 4 Psychiatric Pub; 2013. 5 6 47. Duresso SW, Matthews AJ, Ferguson SG, Bruno R. Is khat use disorder a valid diagnostic entity? 7 Addiction. 2016;111(9):1666-76. 8 48. Douglas H, Boyle M, Lintzeris N. The health impacts of khat: a qualitative study among Somali- 9 Australians. Med J Aust. 2011;195(11-12):666-9. 10 49. Walden N, Earleywine M. How high: quantity as a predictor of cannabis-related problems. Harm 11 reduction journal. 2008;5(1):20. 12 50. Gmel G, Heeb J-L, Rehm J. Is frequency of drinking an indicator of problem drinking? A 13 psychometric analysis of a modified version of the alcohol use disorders identification test in 14 Switzerland. Drug and alcohol dependence. 2001;64(2):151-63. 15 16 51. Stockwell T, SingleFor E. Standard peer unit labelling review of alcohol containers. only Alcohol: Minimising the 17 Harm: What Works. 1997:85-104. 18 52. Zeisser C, Thompson K, Stockwell T, Duff C, Chow C, Vallance K, et al. A ‘standard joint’? The role 19 of quantity in predicting cannabis-related problems. Addiction Research & Theory. 2012;20(1):82-92. 20 53. Fischer B, Russell C, Sabioni P, Van Den Brink W, Le Foll B, Hall W, et al. Lower-risk cannabis use 21 guidelines: a comprehensive update of evidence and recommendations. American journal of public 22 health. 2017;107(8):e1-e12. 23 54. Gebissa E. Leaf of Allah: khat & agricultural transformation in Harerge, Ethiopia 1875-1991: Ohio 24 25 State University Press; 2004. 26 55. Alem A, Kebede D, Kullgren G. The prevalence and socio-demographic correlates of khat 27 chewing in Butajira, Ethiopia. Acta Psychiatrica Scandinavica Supplementum. 1999;100(Suppl 397):84- 28 91. 29 56. Carrier N. ‘Miraa is cool’: the cultural importance of miraa (khat) for Tigania and Igembe youth in 30 Kenya*. Journal of African Cultural Studies.17(2):201-18. 31 57. Basunaid S, Van Dongen M, Cleophas TJ. Khat abuse in Yemen: A population-based survey. 32

Clinical Research and Regulatory Affairs. 2008;25(2):87-92. http://bmjopen.bmj.com/ 33 58. Manghi RA, Broers B, Khan R, Benguettat D, Khazaal Y, Zullino DF. Khat use: lifestyle or 34 35 addiction? Journal of psychoactive drugs. 2009;41(1):1-10. 36 59. Douglas H, Boyle M, Lintzeris N. The health impacts of khat: A qualitative study among Somali- 37 Australians. Medical Journal of Australia. 2011;195(11):666-9. 38 60. Patel SL. Attitudes to khat use within the Somali community in England. Drugs: Education, 39 Prevention & Policy. 2008;15(1):37-53. 40 61. Stevenson M, Fitzgerald J, Banwell C. Chewing as a social act: Cultural displacement and khat on September 25, 2021 by guest. Protected copyright. 41 consumption in the East African communities of Melbourne. Drug and alcohol review. 1996;15(1):73-82. 42 62. Parker H. Normalization as a barometer: Recreational drug use and the consumption of leisure 43 44 by younger Britons. Addiction research & theory. 2005;13(3):205-15. 45 63. Ahmed S, Minami H, Rasmussen A. A Systematic Review of Treatments for Problematic Khat 46 Use. Substance use & misuse. 2019:1-12. 47 48 49 50 51 52 ` 53 54 55 56 57 58 33 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Reporting checklist for qualitative study. 4 5 6 7 8 Page 9 10 Reporting Item Number 11 12 Title 13 14 #1 Exploring the concept of Problematic khat use in the Gurage 1 15 16 For peer review only 17 community, South-central Ethiopia: A qualitative study 18 19 20 21 22 Abstract 23 24 Objective: This study aimed at exploring how problematic khat use is 25 #2 3 26 27 characterized in the Gurage community, South-central Ethiopia. 28 29 30 Design: qualitative study. 31 32 33 Setting: Gurage community, South-central Ethiopia http://bmjopen.bmj.com/ 34 35 36 Participants: We conducted in-depth interviews with 14 khat users and 37 38 5 non-khat users, and 3 focus-group discussions with khat users. 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Methods: All participants were selected purposively based on their 43 44 exposure to khat or khat use. We used interview guide to explore 45 46 perception of participants about khat use and problematic khat use. We 47 48 49 analyzed the data thematically using Open Code software version 4.03. 50 51 We used iterative data collection and analysis, triangulation of methods 52 53 and respondent validation to ensure scientific rigour. 54 55 56 Findings: We identified three major themes: sociocultural khat use, 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from khat suse (khat addiction), and negative consequences of khat use. 1 2 3 Sociocultural khat use included a broad range of contexts and patterns 4 5 including use of khat for functional, social, cultural and religious 6 7 reasons. Khat addiction was mainly explained in terms of associated 8 9 khat withdrawal experiences, including harara/craving, and inability to 10 11 12 quit. We identified mental health, sexual life, physical health, social 13 14 and financial related negative consequences of khat use. The local 15 16 Foridiom Jezbapeer was used review to label subgroup only of individuals with khat suse 17 18 19 (khat addiction). 20 21 22 Conclusion: The study has identified what constitutes normative and 23 24 problematic khat use in the Gurage community in South-central 25 26 Ethiopia. Problematic khat use is broad concepts which include 27 28 29 frequency, reasons, contexts, negative consequences and addiction of 30 31 khat use. Insights generated can be used to inform future studies on 32 33 development of tools to measure problematic khat use. http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Introduction 43 44 Problem formulation #3 Across the literature, conceptualization of problematic, sociocultural 5 45 46 47 and recreational khat use has been an important research gap. 48 49 Problematic khat use, rather than khat use per se, is usually the interest 50 51 of the public, researchers and policymakers. Nevertheless, only few 52 53 54 previous studies were conducted on problematic khat use [33]. The 55 56 inconclusive reports about the different harms of khat use could also be 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from due to poor definition of problematic khat use. The Diagnostic 1 2 3 Statistical Manuel(DSM-5) definition of stimulant use disorders [34] 4 5 could have more important clinical utility (for severe cases) than 6 7 screening individuals with problematic khat use at earlier stage. Lack 8 9 of screening tool for problematic khat use, especially among nonclinical 10 11 12 cases could hamper efforts to curb the problem including early 13 14 identification and effective management of positive cases. Therefore, 15 16 Forthere is peera need for valid review problematic khat only use screening tools which 17 18 19 would facilitate clinical care in primary health care settings and for 20 21 future research. Although there are no strong validation studies, few 22 23 studies used Harmful Khat Use Scale [35] and Severity of Dependence 24 25 26 Scale to measure the construct problematic khat use [36], but a 27 28 systematic review [33] and exploratory studies[37] suggested broader 29 30 indicators of problematic khat use, including amount, frequency, 31 32

context and duration of khat session. http://bmjopen.bmj.com/ 33 34 35 36 Purpose or research #4 The aim of this study was to conceptualize problematic khat use from 5 37 question 38 the perspective of users and non-users in a dominantly rural setting, 39 40 Gurage, south-central Ethiopia. 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 Methods 47 48 Qualitative approach and #5 The study employed a qualitative study design [40] which allowed for 6 49 research paradigm 50 the understanding and description of the experiences and perspectives 51 52 53 of people towards problematic khat use. This study was guided by the 54 55 consolidated criteria for reporting qualitative research (COREQ)[40]. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Researcher characteristics #6 Participants were also diverse in terms of socio-demographic 7 and 2 and reflexivity across the 3 characteristics. The field worker, the first author, has previous 4 paper 5 6 experience interviewing people about khat use. Above all, we tried to 7 8 purely and openly present the ideas of the participants without personal 9 10 impression interference. 11 12 13 14 15 16 Context #7 ForThe study peer was conducted review in the Gurage only zone; Southern Nations, 6 17 18 Nationalities and Peoples Region (SNNPR), Ethiopia. Ethiopian 19 20 Orthodox Christianity (48.17%) and Muslim (42.31%) are the two 21 22 23 dominant religions in Wolketie town, Gurage’s capital [38]. Peasant 24 25 farming is the main productive occupation in rural areas while petty 26 27 trading is more common in urban areas. The area is known for its khat 28 29 30 production and khat use [10]. Khat might have been introduced to 31 32 Gurage area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 33 http://bmjopen.bmj.com/ 34 1543) army or neighboring Muslim Wolane or Oromo ethnic groups 35 36 37 [39]. 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Sampling strategy #8 We conducted face to face in-depth interviews with 14 current khat 6 43 44 users and five non-khat users. Twenty-one khat users participated in 45 46 47 focus group discussions (FGD). The first FGD consisted of six women, 48 49 the second and the third FGDs had seven and eight participants 50 51 respectively. Participants were selected purposively based on their 52 53 54 experience of khat use, and we also aimed for maximum variation 55 56 considering the socio-demographic characteristics of the participants. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Ethical issues pertaining to #9 Ethical clearance was obtained from the Institutional Review Board 7 2 human subjects 3 (IRB) of the College of Health Sciences, Addis Ababa University (Ref 4 5 6 008/18/Psy). In addition, we also obtained a letter of support from the 7 8 Gurage zone health department. Written informed consent was sought 9 10 and obtained from all participants before data collection. 11 12 13 14 15 16 Data collection methods #10ForParticipants peer were invited review by the community only health workers and the first 6 17 author did the informed consent. All the interviews and FGDs were 18 19 conducted in Amharic and tape-recorded. The first author, who has 20 good experience of interviewing khat users, did the interviews and 21 22 facilitated the focus group discussions assisted by a trained moderator. 23 The interviews took about 40 minutes and the FGDs about one hour in 24 25 average. The first and the last authors designed in-depth interview and 26 FGD guides. 27 28 29 Data collection #11 The guides mainly focused on the experiences and perceptions of the 6 30 instruments and participants regarding khat use and problematic khat use. 31 32 technologies 33 http://bmjopen.bmj.com/ 34 Units of study #12 Participants were diversified in several socio-demographic 8 35 36 characteristics, including sex, age, residence, religion, occupation and 37 educational status (Table 1). 38 39 Data processing #13 The data were transcribed verbatim in Amharic, then translated into 7 40 41 English by the first author and experienced research assistants. on September 25, 2021 by guest. Protected copyright. 42 43 Data analysis #14 . All interview and FGD translations were coded independently by the 7 44 45 first and fourth author (AM and CN). Iterative thematic analysis [41] 46 was done simultaneously with data collection. We used computer 47 48 software, open code 4.03 to manage and analyse the data [42]. 49 50 Techniques to enhance #15 Regarding to data quality management and rigour, the iterative process 7 51 52 trustworthiness of data collection, data analysis and checking unclear issues from the 53 participants added to the quality of the study. We did also data 54 55 triangulation from different sources, including religious fathers, key 56 informants for the culture and legal officers. 57 58 Results/findings 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Syntheses and #16 The major themes that emerged from the iterative thematic analysis 9 2 3 interpretation were: (1) sociocultural khat use, (2) suse “(addictive)” khat use, and (3) 4 negative consequences of khat use. The second theme had two 5 6 categories: withdrawal experiences, quitting khat use and 7 mirqanna/feeling high after khat use. The negative consequences were 8 9 categorized to mental health, sexual life, physical health, social and 10 financial. These themes are discussed with support of quotes directly 11 12 from the participants. 13 14 Links to empirical data #17 Withdrawal experiences of khat use 15 15 16 For peer review only 17 Other experiences reported associated with abnormal khat use included 18 19 withdrawal experiences when one was weaning off khat after heavy 20 21 22 usage. The study found many important psychological withdrawal 23 24 experiences of khat suse-addiction. Feeling depressed, irritable and 25 26 aggression were the most typical and commonly reported experiences. 27 28 29 One participant said, “I am usually against people’s communication 30 31 even aggressive when I didn’t chew khat. I remember that I once threw 32 33 away [smacked] my kid when she was talking about her school affairs.” http://bmjopen.bmj.com/ 34 35 (Age 67, urban, male) 36 37 38 39 Different participants reported different withdrawal symptoms, 40 41 including lack of motivation, unable or lack interest to function or on September 25, 2021 by guest. Protected copyright. 42 43 socialize, poor concentration and learning, unable to receive message, 44 45 and dukak (vivid unpleasant dream). Here are two dukak or vivid 46 47 48 unpleasant dream experiences; 49 50 51 One day my husband skipped chewing khat and went to bed. Then, he 52 53 spent the night spitting. In the morning, the bed sheet was wet. When I 54 55 56 ask him what was wrong with him. He said; “people were punishing me 57 58 with the smoke of red paper and I had been feeling burning sensation 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from for the whole night”. (Age 28, urban, women) 1 2 3 Another experience: 4 5 6 7 My experience was...ehm... usually a man would hold my hair and hung 8 9 me or put me into a hole and I would wake up in panic. (Age 38, male, 10 11 urban) 12 13 14 15 16 For peer review only 17 Discussion 18 19 Intergration with prior #18 In this study, our aim was to conceptualize problematic khat us in a 24 20 work, implications, 21 predominantly rural setting in south-central Ethiopia. The study helped 22 transferability and 23 24 contribution(s) to the field to answer the question: what constitutes problematic khat use? Our 25 26 results could inform development of screening tool to measure 27 28 problematic khat use. 29 30 31 Although there was also sociocultural khat use pattern, khat addiction 32 33 and negative consequences of khat use constitute problematic khat use. http://bmjopen.bmj.com/ 34 The local term khat suse, semantically equivalent to khat addiction 35 36 which doesn’t conclusively infer to DSM-5 stimulant use disorder 37 definition[43]. Khat suse/khat addiction could only qualify impaired 38 39 control and pharmacological criteria among the criteria of stimulant use 40 disorders. Khat suse/ khat addiction shares some characteristics from 41 on September 25, 2021 by guest. Protected copyright. 42 other substances use disorders than stimulant use disorders. For 43 example, it has a similar functional consequences of cannabis use 44 45 disorder-amotivational syndrome[43]. In the case of khat suse, the local 46 idiom Jezba could be conceptually related but broader. In the current 47 48 study setting, the use of the word, Jezba, indicates the existing stigma 49 and it could also be a good explanation how much sever form of 50 51 problematic khat use is well recognized in the setting since it has a 52 negative connotation. Khat suse is also indicated by frequent yawning 53 54 when not using the khat. This withdrawal criterion is similar with 55 opioid withdrawal[43]. 56 57 58 Limitations #19 Regarding methodological concern of the study, it didn’t include the 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from perspectives of families of khat users in detail, but the perspective of 1 2 3 non-khat users in general were included. The analysis of the study was 4 5 also not theorized using a well-established existing model or theory, but 6 7 this allows the data to speak for itself without imposing some other 8 9 framework that may poorly fit to the data. This qualitative study, tapped 10 11 12 into indigenous knowledge systems, was community-based which make 13 14 it very strong to define problematic khat use and develop an 15 16 Forunderstanding peer of its domainsreview broadly. only 17 18 19 20 Other 21 22 Conflicts of interest #20 None declared 28 23 24 Funding #21 This work was supported through the DELTAS Africa Initiative [DEL- 27 25 26 27 15-01]. The DELTAS Africa Initiative is an 28 29 independent funding scheme of the African Academy of Sciences 30 31 (AAS)’s Alliance for Accelerating Excellence in Science in Africa 32 http://bmjopen.bmj.com/ 33 (AESA) and supported by the New Partnership for Africa’s 34 35 36 Development Planning and Coordinating Agency (NEPAD Agency) 37 38 with funding from the Wellcome Trust [DEL-15-01] and the UK 39 40

Exploring the concept of Problematic khat use in the Gurage community, South-central Ethiopia: A qualitative study

Journal: BMJ Open ManuscriptFor ID peerbmjopen-2020-037907.R2 review only Article Type: Original research

Date Submitted by the 29-Jul-2020 Author:

School of Medicine, Department of Psychiatry; Harvard T.H. Chan School http://bmjopen.bmj.com/ of Public Health, Boston, USA

Primary Subject Addiction Heading:

Secondary Subject Heading: Addiction, Mental health

MENTAL HEALTH, Substance misuse < PSYCHIATRY, QUALITATIVE Keywords: RESEARCH on September 25, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 41 BMJ Open

1 2

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on September 25, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Exploring the concept of Problematic khat use in the Gurage community, South-central 4 5 6 Ethiopia: A qualitative study 7 8 Awoke Mihretu*1,2, Abebaw Fekadu1, 3, Kassahun Habtamu4, Clement Nhunzvi5, Sam Norton6, Solomon Teferra 1, 7 9 10 11 Email address of authors: 12 13 Correspondence author :Awoke Mihretu= [email protected] 14 15 16 For peer review only 17 Postal address:16417 18 19 20 telephone number: +251921331306 21 22 23 Abebaw Fekadu= [email protected] 24 25 26 Kassahun Habtamu= [email protected] 27 28 29 Clement Nhunzvi= [email protected] 30 31 Sam Norton= 32 [email protected] 33 http://bmjopen.bmj.com/ 34 Solomon Teferra = [email protected] 35 36 1 Department of Psychiatry, School of Medicine, College of Health Sciences, Addis Ababa University, Addis Ababa, Ethiopia 37 38 2 39 Department of social sciences, college of natural and social sciences, Addis Ababa science and Technology University 40 41 3Centre for Innovative Drug Development and Therapeutic Trials for Africa (CDT Africa), College of Health Sciences, Addis Ababa University, on September 25, 2021 by guest. Protected copyright. 42 Addis Ababa, Ethiopia 43 44 4 45 School of Psychology, College of Education and Behavioral Studies, Addis Ababa University, Addis Ababa, Ethiopia 46 47 5Department of Rehabilitation, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe 48 6Psychology Department, Institute of Psychiatry, King’s College London, London, UK 49 50 7Harvard T.H. Chan School of Public Health, Boston, USA 51 52 53 54 55 56 57 58 1 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 4 5 Abstract 6 7 8 Objective: This study aimed at exploring how problematic khat use is characterized in the 9 10 Gurage community, South-central Ethiopia. 11 12 Design: qualitative study. 13 14 15 Setting: Gurage community, South-central Ethiopia 16 For peer review only 17 Participants: We conducted in-depth interviews with 14 khat users and 5 non-khat users, and 3 18 19 focus-group discussions with khat users. 20 21 22 Methods: All participants were selected purposively based on their exposure to khat or khat use. 23 24 We used interview guide to explore perception of participants about khat use and problematic 25 26 khat use. We analyzed the data thematically using Open Code software version 4.03. We used 27 28 iterative data collection and analysis, triangulation of methods and respondent validation to 29 30 31 ensure scientific rigour. 32 33 Findings: We identified three major themes: sociocultural khat use, khat suse (khat addiction), http://bmjopen.bmj.com/ 34 35 and negative consequences of khat use. Sociocultural khat use included a broad range of contexts 36 37 38 and patterns including use of khat for functional, social, cultural and religious reasons. Khat 39 40 addiction was mainly explained in terms of associated khat withdrawal experiences, including 41 on September 25, 2021 by guest. Protected copyright. 42 harara/craving, and inability to quit. We identified mental health, sexual life, physical health, 43 44 45 social and financial related negative consequences of khat use. The local idiom Jezba was used 46 47 to label subgroup of individuals with khat suse (khat addiction). 48 49 Conclusion: The study has identified what constitutes normative and problematic khat use in the 50 51 Gurage community in South-central Ethiopia. Problematic khat use is broad concept which 52 53 54 include frequency, reasons, contexts, negative consequences and addiction of khat use. Insights 55 56 57 58 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Although it has important cultural roots and functions(15, 16), khat use has significant health 4 5 6 harms(23-26), The physical health problems include increased body temperature, loss of 7 8 appetite, gastritis, haemorrhoids, insomnia, and oral health problems (23-26), and even 9 10 hypertension (27) and other cardiovascular dysfunctions (22, 28, 29). 11 12 13 The mental health impacts include psychotic symptoms (30-32) and depressive symptoms (33). 14 15 However, the mental health impacts remain inconclusive because other studies have revealed 16 For peer review only 17 18 negative findings (1, 34). 19 20 21 Across the literature, conceptualization of problematic, sociocultural and recreational khat use 22 23 has been an important research gap. Problematic khat use, rather than khat use per se, is usually 24 25 the interest of the public, researchers and policymakers. Nevertheless, only few previous studies 26 27 28 were conducted on problematic khat use (35). The inconclusive reports about the different harms of 29 30 khat use could also be due to poor definition of problematic khat use(35). 31 32 33 The Diagnostic Statistical Manual(DSM-5) definition of stimulant use disorders (36) could have http://bmjopen.bmj.com/ 34 35 more important clinical utility (for severe cases) than screening individuals with problematic 36 37 khat use at earlier stage. Lack of screening tool for problematic khat use, especially among 38 39 40 nonclinical cases could hamper efforts to curb the problem including early identification and effective 41 on September 25, 2021 by guest. Protected copyright. 42 management of positive cases. Therefore, there is a need for valid problematic khat use screening 43 44 tools which would facilitate clinical care in primary health care settings and for future research. 45 46 47 Although there are no strong validation studies, few studies used Harmful Khat Use Scale (37) 48 49 and Severity of Dependence Scale to measure the construct problematic khat use (38), but a 50 51 systematic review (35) and exploratory studies(39) suggested broader indicators of problematic 52 53 khat use, including amount, frequency, context and duration of khat session. Therefore, the aim 54 55 56 57 58 5 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 of this study was to conceptualize and define problematic khat use from the perspective of users 4 5 6 and non-users in a dominantly rural setting, Gurage, south-central Ethiopia. 7 8 9 10 11 12 Methods 13 14 Study setting 15 16 For peer review only 17 The study was conducted in the Gurage zone; Southern Nations, Nationalities and Peoples 18 19 20 Region (SNNPR), Ethiopia. Ethiopian Orthodox Christianity (48.17%) and Muslim (42.31%) are 21 22 23 the two dominant religions in Wolketie town, Gurage’s capital (40). Peasant farming is the main 24 25 26 productive occupation in rural areas while petty trading is more common in urban areas. The area 27 28 is known for its khat production and khat use (12). Khat might have been introduced to Gurage 29 30 area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 1543) army or neighboring Muslim 31 32

Wolane or Oromo ethnic groups (41). http://bmjopen.bmj.com/ 33 34 35 Study design 36 37 38 The study employed a qualitative study design (42) which allowed for the understanding and 39 40 41 description of the experiences and perspectives of people towards problematic khat use. This on September 25, 2021 by guest. Protected copyright. 42 43 study was guided by the standard of reporting qualitative research(SRQR) and consolidated 44 45 criteria for reporting qualitative research (COREQ)(42). 46 47 48 Study participants and recruitment procedures 49 50 51 We conducted face to face in-depth interviews with 14 current khat users and five non-khat 52 53 users. Khat users are participants who use khat for the last 30 days. We included non-khat users to 54 55 understand their perception about khat users and khat use behavior if it would be different from khat 56 57 58 6 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 users. Twenty-one khat users participated in focus group discussions (FGD). The first FGD 4 5 consisted of six women, the second and the third FGDs were males with seven and eight 6 7 participants respectively. Sample size was determined based on data saturation. Theoretical data 8 9 saturation (15 to 20 participants) was planned before data collection and analysis for both khat 10 users and non-khat users. We first did data collection from khat users. Then we had continued 11 12 data collection among non-khat users until the saturation of themes was determined to have 13 14 occurred. Theoretical data saturation for all types of participants was informed by our previous 15 studies(23) and other literature(43). We planned (15 to 20 participants) and recruited more 16 For peer review only 17 participants (40 participants), because the subject area is new and not well defined theoretically. 18 19 Participants were selected purposively based on their experience of khat use, and we also aimed 20 21 for maximum variation considering the socio-demographic characteristics of the participants. In 22 addition to khat users, other people who have knowledge about the history, cultural, religious 23 24 base of khat use and frequent contact with khat users were interviewed. Thus, 5 religious fathers, 25 26 police officers and culture experts were participated in the study. Participants were invited by the 27 community health workers when they had gathered for their regular meeting and the first author 28 29 did the informed consent. 30 31 32 Data collection procedures 33 http://bmjopen.bmj.com/ 34 Consented individuals were interviewed in their home, and at open places in the community. All 35 36 focus group discussions were conducted at health facilities (health posts) garden in the 37 38 39 community. Religious fathers, police officers and culture experts were interviewed at their 40 41 office. All the interviews and FGDs were conducted in Amharic and tape-recorded. The first on September 25, 2021 by guest. Protected copyright. 42 43 author, who has good experience of interviewing khat users, did the interviews and facilitated the 44 45 46 focus group discussions assisted by a trained moderator. The interviews took about 40 minutes 47 48 and the FGDs about one hour in average. The first and the last authors designed in-depth 49 50 interview and FGD guides. We piloted the final version of the topic guides across the process of 51 52 data collection. 53 54 55 56 57 58 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 In-depth interview topic guides focused on khat use experiences, patterns of khat use through 4 5 6 time, reasons for use, perception about khat and khat use, criteria of problematic khat use, 7 8 behaviors after khat use and when they do not use and quitting experience from khat use. 9 10 Focus group discussion guides were also similar focusing more on shared perceptions of normal 11 12 and problematic khat use, and community level harms. 13 14 15 Data management and analysis 16 For peer review only 17 18 The data were transcribed verbatim in Amharic; then translated into English by the first author 19 20 and experienced research assistants. All interview and FGD translations were coded 21 22 independently by the first and fourth author (AM and CN). Iterative thematic analysis (44) was 23 24 25 done simultaneously with data collection. We used computer software, open code 4.03 to 26 27 manage and analyse the data (45). 28 29 30 Regarding to data quality management and rigour, the iterative process of data collection, data 31 32 analysis and checking unclear issues from the participants added to the quality of the study. http://bmjopen.bmj.com/ 33 Triangulation was achieved through data collection methods (IDI and FGDs) and data sources 34 35 (different groups of participants including religious fathers, key informants for the culture and 36 37 legal officers) and the team of researchers with diverse research experiences did crosschecking 38 39 the data and interpretations. The IDI reports about cultural and religious basis of khat use were 40 triangulated with religious fathers’ perception. Community level impacts of problematic khat use 41 on September 25, 2021 by guest. Protected copyright. 42 were also triangulated with opinions of key informants and experts. The field worker, the first 43 44 author, has previous experience interviewing people about khat use. Above all, we tried to purely 45 and openly present the ideas of the participants without personal impression interference. 46 47 48 Ethical considerations 49 50 51 Ethical clearance was obtained from the Institutional Review Board (IRB) of the College of 52 53 Health Sciences, Addis Ababa University (Ref 008/18/Psy). In addition, we also obtained a letter 54 of support from the Gurage zone health department. Written informed consent was sought and 55 56 obtained from all participants before data collection. For those participants who couldn’t read 57 58 8 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 and write, the primary author read the information sheet for them in the presence of a witness 4 5 who can read to confirm the veracity of the information, and those who agreed to participate in 6 7 the study signed on the consent form with their thumbprint. 8 9 10 11 12 Patient and public involvement 13 14 No patient involved. 15 16 For peer review only 17 Findings 18 19 20 Characteristics of participants 21 22 23 Table 1 Socio-demographic characteristics of participants 24 25 Characteristics Interview FGD 26 27 Age 3 8 28 29 20-34 30 31 35-44 5 2 32 33 45-59 6 8 http://bmjopen.bmj.com/ 34 35 60 and above 5 3 36 37 Gender 38 39 Male 16 15 40 41 Female 3 6 on September 25, 2021 by guest. Protected copyright. 42 43 Residence 44 Urban 12 7 45 46 Rural 7 14 47 48 Marital status 49 50 Single 1 4 51 52 Married 17 15 53 54 Widowed or divorced 1 2 55 56 57 58 9 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Education 4 5 Can’t read and write 3 2 6 7 Read and write only 1 2 8 9 Primary 2 5 10 11 Secondary 2 7 12 13 Post-secondary 11 5 14 15 Religion 16 For peer review only 17 Muslim 8 9 18 19 Christian 11 12 20 21 Employment status 22 23 Self-employed 6 8 24 25 Unemployed 4 8 26 27 Formally employed 8 2 28 29 Student 1 3 30 31 Khat use status 32

Current khat users 14 21 http://bmjopen.bmj.com/ 33 34 Non-khat users 5 - 35 36 37 38 Participants were diverse in several socio-demographic characteristics, including sex, age, 39 40 residence, religion, occupation and educational status (Table 1). All non-khat users were males 41 on September 25, 2021 by guest. Protected copyright. 42 43 and their age was in 50s years. They are religious fathers, police officers and culture experts. 44 45 46 The major themes that emerged from the iterative thematic analysis were: (1) sociocultural khat 47 48 use, (2) suse “(addictive)” khat use, and (3) negative consequences of khat use. The second 49 50 theme had two categories: withdrawal experiences, quitting khat use and mirqanna/feeling high 51 52 after khat use. The negative consequences were categorized to mental health, sexual life, 53 54 55 56 57 58 10 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 physical health, social and financial. These themes are discussed with support of quotes directly 4 5 6 from the participants. 7 8 9 Sociocultural khat use 10 11 12 The sociocultural khat use theme emerged from participants’ notions of using khat for spiritual, 13 14 social and cultural reasons. Among Muslim participants, khat use was very common and 15 16 acceptable during religiousFor rituals peer including review the fasting month- onlyremedan, holiday celebration and 17 18 anniversaries for the shrine of religious forefathers and at pilgrimage center. During these 19 20 21 situations, khat was one of the presents-zihara. 22 23 24 Many Muslim participants, especially elderly men, reported chewing khat for prayer. One 25 26 participant who was an elderly Muslim man described his use of khat as “… a prayer instrument 27 28 (duw’a)”. Some Muslim farmers who participated in this study reported chewing khat during 29 30 31 prayer as an inherited religious virtue-Ibada. Other participants described that the religious norm 32 33 of khat use for prayer was acceptable if it was in group chewing than chewing alone. A http://bmjopen.bmj.com/ 34 35 participant making reference to the sword as a symbol for prayer alongside chewing khat shared 36 37 that: 38 39 40

“One sword can’t kill rather it only stabs. Thus, we usually chew being three or more” (Age late on September 25, 2021 by guest. Protected copyright. 41 42 43 50’s, male, urban) 44 45 46 In rural areas, unlike urban areas, participants shared that they chew khat and do religious 47 48 rituals in mosques on weekly venerations of saints-wali. For example, Tuesdays had special 49 50 meaning because they were a memorial for Nurahusene, a religious forefather. Gathering in 51 52 53 mosques on special days of the week to chew khat and pray was part of the religious practice. 54 55 56 57 58 11 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 “Although khat chewing doesn’t have any religious basis, we [Muslims] chew khat on the days 4 5 6 which are the shrine of our religious forefathers such as Esnel (Monday), Megergbia (Tuesday), 7 8 Gelale (Wednesday) and Sehiare (Saturday).” (Muslim religion scholar, age late 50’s) 9 10 11 Ceremonial khat use was also viewed as common practice at pilgrimage centres and during 12 13 annual Muslim religious holidays; Arafa (i.e. Id al-Adha), Id alfeter and moulid. In the study 14 15 setting, Aberiat, Qatbarie (Shaykh Isa Hamza), Alkeso, and Zebimola were frequently mentioned 16 For peer review only 17 18 as the most colourful wali venerating practices. Pilgrims had a religious ritual called hadra. It is 19 20 a ritual by small groups of pilgrims for religious ritual and khat chewing session. They did this 21 22 being in a separate partition in the mosque compound. 23 24 25 Khat use was not seen as limited to subgroups like the elderly and Muslim men, rather it was part 26 27 of the life of many people from different walks of life in the Gurage area. Many participants 28 29 agreed that, through time, khat was becoming acceptable and meaningful in the broader social 30 31 32 and cultural events. The participants agreed that, khat use was an important part of the Gurage 33 http://bmjopen.bmj.com/ 34 culture. This study indicated that khat chewing was shaping the day to day activities, beliefs, 35 36 tradition, and rituals of many people. Khat users, especially in rural areas, reported that khat 37 38 chewing was a significant part of their daily life. They frequently said that they used khat daily to 39 40 41 express both their mourning and happiness. on September 25, 2021 by guest. Protected copyright. 42 43 Participants reported that khat use was acceptable and to chew khat was to conform to societal 44 45 norms and values but there were specific nuances in how this was done to remain acceptable. 46 47 48 This was mainly not to have a socially agreed problems resulting from one’s chew of khat. The 49 50 normal and accepted khat user was portrayed as “…a humble, honest, generous, and positive 51 52 thinker.” (Age in 40’s, rural, male) 53 54 55 56 57 58 12 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Other definitions of normal khat use were related to patterns of use. Participants had set criteria 4 5 6 for who, when, how much and where to chew. For example, participants said it was part of their 7 8 culture not to chew khat in the morning, but recognised khat use in the afternoon. 9 10 11 Many participants did not accept daily khat use as normal. Normal khat users were also 12 13 supposed to limit the amount of their khat consumption. Chewing in public places was also 14 15 acceptable and normative among men but considered abnormal among women. 16 For peer review only 17 18 This was shared by one participant in the focus group discussions and they highlighted: “there 19 20 21 are males who chew khat in cafeterias and verandas, but a woman does not chew in public 22 23 places even they don’t buy and hold khat, except commercial sex workers.” (Age early 30’s, 24 25 female, urban) 26 27 28 We found much restriction of the norms among rural than urban participants. Participants from 29 30 31 rural areas emphasized that as areqi (a local illicit alcoholic drink) have been part and parcel of 32 33 the life of many Christians and the same was true for khat among Muslims. But now, khat use http://bmjopen.bmj.com/ 34 35 had become a normal behaviour of both Muslims and Christians in both urban and rural areas. 36 37 Use was considered normal as long as it did not come with other problems. 38 39 40

Many participants also shared that sociocultural khat use was an accepted way of use valued in on September 25, 2021 by guest. Protected copyright. 41 42 43 their community. Examples of sociocultural uses that emerged from the analysis included: 44 45 hosting visitors, grieving period after funeral, work parties, as part of normal socialisation, 46 47 weddings and in rare cases when paying visit to the sick. A participant highlighted as: “during 48 49 50 condolence after funeral, wedding ceremony…when we chew khat, it will be very easy to 51 52 socialize and easily close with others. In our home, we feel confident to host guest if we can buy 53 54 khat otherwise I don’t say it will be full hospitality.” (Age early 30’s, male, urban) 55 56 57 58 13 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Despite the normative khat use reported, the participants also shared insights around what 4 5 constituted problematic khat use. 6 7 8 9 10 Khat suse-addiction 11 12 13 Most participants conclusively stated that khat use is an addictive behavior-suse. This was the 14 15 most commonly reported reason to continue chewing by many participants so that they could 16 For peer review only 17 accomplish their day to day activities. Different withdrawal experiences (presented in the next 18 19 20 section) were reported to justify the suse or addiction to khat use. At the addictive stage, those 21 22 affected are seen by keeping some khat in their pockets and chew almost all the time without 23 24 regard for place too. This included chewing at school, funeral, market, and workplace either 25 26 27 office or farm. A participant explained khat suse/addiction as; 28 29 30 I should have khat in my bloodstream to open my eyes and to activate my body. Morning khat 31 32 use-ejebena is a sign of khat suse, isn’t it? (Age in 40’s, urban, male) 33 http://bmjopen.bmj.com/ 34 35 It was then this type of use that was viewed as causing problems to the health and social aspects 36 37 of both the user and those around him/her. 38 39 40 Participants also mentioned harara/craving as indicator of khat suse/addiction. Many said that 41 on September 25, 2021 by guest. Protected copyright. 42 khat use is an addictive behaviour because it has a strong harara/craving. They justified that the 43 44 45 strong harara/craving even forced people to beg for money to buy khat when they did not have 46 47 money or they would collect the leftover khat and chew. When these behaviours start to show, 48 49 the khat chewing is then considered abnormal and problematic. 50 51 52 Those who chew because of harara/craving, unlike those who chew for prayer, chew khat 53 54 55 regularly, any time including in the morning, chew without ritual or ceremony and without 56 57 58 14 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 washing themselves. All these patterns of use are abnormal in my culture. (Age late 50’s, male, 4 5 6 urban) 7 8 9 Chewing khat by alms was perceived as the last and worst stage of a khat user who was called 10 11 Jezba. 12 13 14 Participants used the term Jezba to label subgroups of people with khat suse or addiction. 15 16 Jezbas are people withFor khat susepeer or addiction review and characterized only syndrome of the behavior. Jezba 17 18 syndrome indicates; poor connection with God, others and material needs, not confirming to 19 20 21 shared values and norms of the society. Jezba khat users prioritize khat use rather than other 22 23 important aspects of life or they have limited interest for other important areas of life except 24 25 khat. They also had reduced motivation for work and other social life or total disengagement 26 27 28 when the khat is withdrawn as well as frequent work absenteeism or abandoning work. It also 29 30 includes; deterioration in critical thinking, poor social skills/self-care, beg khat or money for 31 32 khat, external attribution, belief of inability or no interest to control khat use and related 33 http://bmjopen.bmj.com/ 34 35 behaviors, chewing for many hours of the day (more than 6 hours) and on the street or while 36 37 walking. 38 39 40 In the long term, khat will make you Jezba-ያጀዝብሀል. Jezba doesn’t value anything important 41 on September 25, 2021 by guest. Protected copyright. 42 other than khat. (Age in 40's, rural, male) 43 44 45 For many participants, the use of khat against sociocultural conditions was considered as a sign 46 47 of khat suse/addiction. This includes chewing khat for individual reasons such as to cope with 48 49 50 adverse life experiences while cultural khat users chew for communal reasons; for prayer, 51 52 socialization or companionship, and functional reasons. 53 54 55 56 57 58 15 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Khat addiction was also characterized in terms of chewing increased amount of khat through 4 5 6 time as well as chewing for longer time to get the desired effect. 7 8 9 When you chew too much [greater than a bundle of khat], it will cause many problems; being 10 11 addictive could be one. Only limited amount of khat benefits [the stimulating effect]. (Age late 12 13 50’s, male, rural) 14 15 16 In other views, someFor participants peer said that review increased use ofonly khat overtime depended on one's 17 18 mindset or psychological expectation, amount of budgeted time for chewing, emotional state or 19 20 21 group cohesion during chewing and financial capacity to buy khat. Age was also mentioned as an 22 23 important factor for heavy and problematic use of khat. Youths and early adult participants 24 25 showed more addictive use of khat, unlike the elderly. 26 27 28 Withdrawal experiences of khat use 29 30 31 Other experiences reported associated with abnormal khat use included withdrawal experiences 32 33 http://bmjopen.bmj.com/ 34 when one was weaning off khat after heavy usage. The study found many important 35 36 psychological withdrawal experiences of khat suse-addiction. Feeling depressed, irritable and 37 38 aggression were the most typical and commonly reported experiences. One participant said, “I 39 40

am usually against people’s communication even aggressive when I didn’t chew khat. I on September 25, 2021 by guest. Protected copyright. 41 42 43 remember that I once threw away [smacked] my kid when she was talking about her school 44 45 affairs.” (Age late 50’s, urban, male) 46 47 48 Different participants reported different withdrawal symptoms, including lack of motivation, 49 50 unable or lack interest to function or socialize poor concentration and learning, unable to receive 51 52 53 message, and dukak (vivid unpleasant dream). Here are two dukak or vivid unpleasant dream 54 55 experiences; 56 57 58 16 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 One day my husband skipped chewing khat and went to bed. Then, he spent the night spitting. In 4 5 6 the morning, the bed sheet was wet. When I ask him what was wrong with him. He said; “people 7 8 were punishing me with the smoke of red paper [forcing people to inhale the smoke of red paper 9 10 is one of harsh corporal punishments among a few in the culture] and I had been feeling burning 11 12 sensation for the whole night”. (Age early 30’s, urban, women) 13 14 15 Another experience: 16 For peer review only 17 18 My experience was...ehm... usually a man would hold my hair and hung me or put me into a hole 19 20 21 and I would wake up in panic. (Age in 30’s, male, urban) 22 23 24 Some participants reported physical withdrawal experiences, including abdominal pain, 25 26 headache, being drowsy, red eyes, increased appetite and sleep, yawning, uncontrollable tears, 27 28 shivering hands, and loss of energy. Another interesting finding was an observation that to 29 30 31 manage some withdrawal syndromes, some khat users opted to sleeping. During fasting season, 32 33 many Muslim participants did not use khat either in the morning or for the whole day, but they http://bmjopen.bmj.com/ 34 35 would usually spent the day sleeping. A Muslim participant shared his experience and said; 36 37 38 During remedan [fasting month], we don’t chew khat during the day time, but no one did a 39 40

serious work. We spend the day sleeping. If you don’t chew khat, you cannot be stimulated, on September 25, 2021 by guest. Protected copyright. 41 42 43 energetic..... (Age late 50’s, male, rural) 44 45 46 Inability to control khat use 47 48 49 Quitting khat use was usually contingent on the participants’ lifestyle and pattern of khat use. For 50 51 participants who didn’t chew daily, quitting was perceived as an easy task, but it was very 52 53 challenging to participants without job because khat use would be an important leisure activity. 54 55 56 External factors; culture and pressure from others, were also reported challenges to quit from 57 58 17 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 chewing. Some thought that quitting khat use also includes quitting important social networks. 4 5 6 Thus, they continue chewing and consider khat use as a matter of surviving in the social system. 7 8 9 Some participants reported as they forward a rational decision to continue chewing after 10 11 evaluating the benefits and harms of khat use. Participants who perceive more benefits, than 12 13 harms from khat use, such as positive effect on motivation, work performance and socialization 14 15 continue chewing khat. 16 For peer review only 17 18 A participant stated as follows; 19 20 21 I will not quit chewing because it helps me to share information, got social support and other 22 23 24 things as well as to relieve stress and worries in life. (Age in 30’s, urban, male) 25 26 27 Many participants said that only a few and the fortunate could quit early, but many realize the 28 29 harms later. Many quit chewing when they lose satisfaction from their chewing behavior, usually 30 31 at the end of their life. Hence chewing until the end of life was considered abnormal or addictive 32 33 http://bmjopen.bmj.com/ 34 behavior. 35 36 37 After years, one would lose satisfaction from chewing. You would lose the passion to chew khat 38 39 and you will decide to say it is time to quit. Only the unfortunate will chew to the end of their life. 40 41 (Age late 50’s, rural, male) on September 25, 2021 by guest. Protected copyright. 42 43 44 Mirqanna/feeling high after khat use 45 46 47 Many of the khat user participants liked the stimulant effect of khat, but when the stimulation 48 49 was too high, they reported feelings of distress. Mirqanna (induced distress) was due to chewing 50 51 52 excessive amounts or chewing in combination with shisha or cigarettes. This was repeatedly 53 54 reported among participants in their early phases of khat use. Insomnia induced by mirqanna was 55 56 57 58 18 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 also a perceived cause of distress. Many khat users would then resort to drinking alcohol to break 4 5 6 the mirqanna and induce sleep. Other khat users, experiencing mirqanna but cannot afford to 7 8 drink alcohol because of lack of money, reported continued feelings of being restless, being on 9 10 the move and confusion during. One such participant shared; 11 12 13 One day, I chewed khat for five hours. Then at night, I couldn’t sleep. I had spent the night 14 15 itching, feeling fever and sweating. I was very panic about my condition, but I was also feeling 16 For peer review only 17 18 fatigue and low energy to treat myself. (Age in 30’s, urban, male) 19 20 21 Additional symptoms of mirqanna reported by participants included; being absorbed to an 22 23 inauthentic personal world such as considering oneself as fortunate, being extremely humble, and 24 25 considering oneself as high achiever, more like delusions. In some participants, they reported 26 27 28 lack of interest to communicate during mirqanna state, though talkativeness was also reported 29 30 among others. Excessive fear including avoiding any exposure or engagement with others as well 31 32 as fear to make decisions was also common. Participants emphasized that the mirqanna 33 http://bmjopen.bmj.com/ 34 35 experiences were different from their experiences of intoxication due to excessive alcohol 36 37 drinking. 38 39 40 When it [mirqanna] is severe, when one chews too much, one might spend the night outside the 41 on September 25, 2021 by guest. Protected copyright. 42 house, on the street, which is risky because wild animals could harm him. Sometimes alcohol 43 44 intoxication might be better than khat mirqanna. The intoxication of alcohol can be reversed 45 46 47 soon with different techniques including getting sleep, but the khat mirqanna couldn't be 48 49 reversed easily once the person is at the severe stage. (Age late 50’s, rural, male) 50 51 52 53 54 55 56 57 58 19 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Other participants thought that severe mirqanna is like psychosis state-qezete. Mumbling alone, 4 5 6 confusion, spending the night on the street were perceived symptoms of psychosis induced by 7 8 khat. The following quote was from the experience of one participant. 9 10 11 When I am at mirqanna state, I imagine as I own a big building, big car…By the next day, I 12 13 realize everything was a fantasy. All were gone as cloud. (Age late 50’s, urban, male) 14 15 16 Mental health relatedFor consequences peer review only 17 18 19 Participants reported that khat could lead to mental health disorders. Depression-debert and 20 21 psychosis-qezete were commonly reported mental health disorders. Participants indicated that 22 23 24 depression is associated with khat withdrawal or secondary to different crisis especially 25 26 financial, caused by khat use. Psychosis which is known by different phrases such as qezete (the 27 28 acute form of psychosis), chereken metal, aemeron mesat, yeaemero menawet were common 29 30 31 among participants who chew too much amount of khat and don’t have meal before chewing. A 32 33 participant described this as follows; http://bmjopen.bmj.com/ 34 35 36 …only individuals who had nutritional capacity- body fluid can resist the adverse effect of the 37 38 khat. The brain will be vulnerable when one chews without having a meal. People who chew 39 40

without having a protective, food, will be negatively affected by the khat. (Age in 30’s, rural, on September 25, 2021 by guest. Protected copyright. 41 42 43 male) 44 45 46 Participants reported that different behavioral symptoms of chronic khat users could be similar 47 48 with the behaviors of people with severe mental illness. They further described that among 49 50 chronic khat users, like people with severe mental disorders, there was broken down of family 51 52 53 and other social activities including poor self-care and dressing. In addition, they chew khat on 54 55 the street and they also chew leftover khat or beg money for khat. 56 57 58 20 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Social related consequences of khat use 4 5 6 Another important problem of khat use was its social harms such as family chaos and 7 8 9 breakdowns. This was related with spending too much of the family budget on khat and 10 11 abandoning their responsibilities in the family. Negative behaviors of participants with khat use 12 13 such as irritability also led to family conflict. Two participants’ experiences are stated as follows; 14 15 16 I am not giving timeFor for my kids. peer I just delegate review my elderly daughter only to take care and control her 17 18 siblings. (Age in 40’s, urban, male) 19 20 21 I have been forced by my wife to stop chewing. I don’t want to quit, but my wife is insisting me. 22 23 24 We have been quarrelling and separated due to this issue. Her parents came to mediate us and 25 26 asked me to quit chewing, but I told them my position that I got the khat before her and now I am 27 28 not interested to stop chewing…….I will not stop. Imagine? (Age in 30’s, urban, male) 29 30 31 Among urban dwellers, problem to discharge their government responsibilities were commonly 32 33 http://bmjopen.bmj.com/ 34 reported. Some reported their experience of being blamed by frequent work absenteeism and 35 36 work inefficiency because they lack the motivation to go to work when they don’t chew khat. 37 38 They usually go for khat abandoning their work. 39 40 41 Khat suse/addiction sometimes led to crime. Some, especially in rural areas, usually quarrel on September 25, 2021 by guest. Protected copyright. 42 43 with others. The main source of reported conflict was khat stolen. Participants observed that khat 44 45 46 users who don’t have money to buy usually theft someone's khat from his/her farm. Among 47 48 urban participants, few khat users sometimes also stole others’ properties. Others sell their home 49 50 utensils to get money for khat without the consent of their family members. Theft and violence 51 52 53 were common among khat users who drink alcohol excessively. 54 55 56 Financial consequences of khat use 57 58 21 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Participants were concerned about the price of khat. The minimum reported daily expense was 4 5 6 about 1.5 USD for a bundle of khat, but many also spent for alcohol, coffee, shisha and cigarette. 7 8 Some participants, at sever stage, had sold assets or home utensils and spend the money for khat. 9 10 11 In addition to the direct adverse effect of khat use on finance, participants were also concerned 12 13 about the amount of time they spend by chewing khat. In order to get the desired stimulation 14 15 from khat, they were supposed to stay chewing khat for long hours; the longest was 6 and more 16 For peer review only 17 18 hours. Participants question how much they could earn if they didn't spend their time by chewing 19 20 khat. Others who chew being in a group usually spend more time and unable to quit the session 21 22 and go for work being attracted by the fun and the chat. 23 24 25 The financial advantage from khat was reported from participants who have khat farm and sell 26 27 28 khat. They were relatively better in terms of financial capacity. For them, khat is an important 29 30 cash crop which allows them to cover home expenses and pay government tax easily while 31 32 others were under stress. 33 http://bmjopen.bmj.com/ 34 35 Physical health related consequences 36 37 Many of the participants admitted that after khat use, it is common to experience loss of 38 39 40 appetite. Thus, weight loss and malnutrition were commonly reported. 41 on September 25, 2021 by guest. Protected copyright. 42 43 Khat absorbs your body fluid and makes you dry so that a khat user is thin and latter would be 44 45 vulnerable to different diseases. (Age in 20’s, rural, male) 46 47 48 Gastrointestinal and oral health problems were frequently reported. Bad odour of the mouth, a 49 50 colour change of the tongue and teeth, and teeth spoil were among the major complaints. 51 52 53 Chronic khat user participants couldn't prefer sauce and beef, which are common and valued in a 54 55 routine dish of the culture because of the burning sensation of their mouth and teeth damage. 56 57 58 22 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Dehydration and constipation were also other major complaints. During khat withdrawal phase; 4 5 6 general physical pain, severe headache, burning sensation, and redness of the eyes were common 7 8 complaints. 9 10 11 For participants from Wolketie, chewing khat and drinking alcohol were considered as risky 12 13 behaviour for unsafe sex and thus HIV/AIDS infection. Many were also excessive alcohol users 14 15 and vulnerable to all the adverse effects of excessive alcohol use. The leftover khat disposed 16 For peer review only 17 18 elsewhere in the city was also another concern for their health. Participants rarely reported 19 20 accidents and injuries related to khat use. Table 2 below shows major indicators for normal and 21 22 problematic khat use. 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Page 25 of 41 BMJ Open

BMJ Open Page 26 of 41

Page 27 of 41 BMJ Open

1 2 3 discussion 4 5 - There is pleasure and utility associated with khat use 6 7 8 9 10 Suse/addiction -Khat use results in harmony in the family and society - Associated harms on family and social, work 11 12 -There is no orFor minor craving peer review-Unable only to function without khat 13 14 -people experience well, minor and transient experiences -There is craving-harara 15 of distresses before and after khat use 16 -Serious withdrawal experiences and distressing experiences after 17 - Quitting is easy, but there is no frequent thought or khat use /mirqanna 18 attempt about quitting http://bmjopen.bmj.com/ 19 - Quitting is difficult because of the distressing experiences when 20 the khat is withdrawn and persons usually think or attempt to quit 21 22 Associated behaviours - Some use other psychoactive substances, others don’t -Khat use was in combination with smoking cigarettes, shisha and 23 drinking alcohol 24 25

1 2 3 Discussion 4 5 6 In this study, our aim was to conceptualize problematic khat us in a predominantly rural setting 7 8 9 in south-central Ethiopia. The study helped to answer the question: what constitutes problematic 10 11 khat use? Our results could inform development of screening tool to measure problematic khat 12 13 use. 14 15 16 Although there wasFor also peer sociocultural review khat use pattern, only khat addiction and negative 17 18 consequences of khat use constitute problematic khat use. The local term khat suse, semantically 19 20 21 equivalent to khat addiction which doesn’t conclusively infer to DSM-5 stimulant use disorder 22 23 definition(46). Khat suse/khat addiction could only qualify impaired control and pharmacological 24 25 criteria among the criteria of stimulant use disorders. Khat suse/ khat addiction shares some 26 27 28 characteristics from other substances use disorders than stimulant use disorders. For example, it 29 30 has a similar functional consequences of cannabis use disorder-amotivational syndrome(46). In 31 32 the case of khat suse, the local idiom Jezba could be conceptually related but broader. In the 33 http://bmjopen.bmj.com/ 34 35 current study setting, the use of the word, Jezba, indicates the existing stigma and it could also be 36 37 a good explanation how much sever form of problematic khat use is well recognized in the 38 39 setting since it has a negative connotation. Khat suse is also indicated by frequent yawning when 40 on September 25, 2021 by guest. Protected copyright. 41 not using the khat. This withdrawal criterion is similar with opioid withdrawal(46). 42 43 44 Major negative consequences of khat use, which are other indicators of problematic khat use, 45 46 47 include sexual dysfunction, depression, psychosis, various oral health problems, and wastage of 48 49 time. Many of these indicators of problematic khat use are reported in several previous studies 50 51 (35, 39, 47). In other settings, chewing khat with sugar was stated as a factor for some health 52 53 54 problems especially oral health problems(48). But in the current study the use of sugar with khat 55 56 and sugary drinks is not common in rural areas because of cost and availability. They even drink 57 58 27 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 The authors would like to acknowledge study participants and the African mental health research 4 5 6 initiative (AMARI) of DELTAS Africa Initiative. 7 8 9 Funding: 10 11 12 This work was supported through the DELTAS Africa Initiative [DEL-15-01]. The DELTAS 13 14 Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)’s 15 16 Alliance for AcceleratingFor Excellence peer in Science review in Africa (AESA) only and supported by the New 17 18 Partnership for Africa’s Development Planning and Coordinating Agency (NEPAD Agency) 19 20 21 with funding from the Wellcome Trust [DEL-15-01] and the UK government. The views 22 23 expressed in this publication are those of the author(s) and not necessarily those of AAS, 24 25 NEPAD Agency, Wellcome Trust or the UK government. 26 27 28 Authors’ contributions: 29 30 31 AM ST AF conceived and designed the study. AM did the interview. AM and CN coded the 32 33 http://bmjopen.bmj.com/ 34 data. AM AF KH CN SN ST contributed to the analysis and the write-up of the manuscript. All 35 36 authors agree with the results and conclusions of the study. 37 38 39 Ethical approval and consent to participate 40 41 on September 25, 2021 by guest. Protected copyright. 42 The study was approved by the Institutional Review Board of the College of Health Sciences of 43 44 Addis Ababa University (REF. 008/18/psy). Participants took part in the study after providing 45 46 written informed consent. 47 48 49 Competing interests: 50 51 52 None declared 53 54 55 Data sharing statement: No additional unpublished data are available from this study. 56 57 58 30 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

Clinical Research and Regulatory Affairs. 2008;25(2):87-92. http://bmjopen.bmj.com/ 33 58. Manghi RA, Broers B, Khan R, Benguettat D, Khazaal Y, Zullino DF. Khat use: lifestyle or 34 35 addiction? Journal of psychoactive drugs. 2009;41(1):1-10. 36 59. Douglas H, Boyle M, Lintzeris N. The health impacts of khat: A qualitative study among Somali- 37 Australians. Medical Journal of Australia. 2011;195(11):666-9. 38 60. Patel SL. Attitudes to khat use within the Somali community in England. Drugs: Education, 39 Prevention & Policy. 2008;15(1):37-53. 40 61. Stevenson M, Fitzgerald J, Banwell C. Chewing as a social act: Cultural displacement and khat on September 25, 2021 by guest. Protected copyright. 41 consumption in the East African communities of Melbourne. Drug and alcohol review. 1996;15(1):73-82. 42 62. Parker H. Normalization as a barometer: Recreational drug use and the consumption of leisure 43 44 by younger Britons. Addiction research & theory. 2005;13(3):205-15. 45 63. Ahmed S, Minami H, Rasmussen A. A Systematic Review of Treatments for Problematic Khat 46 Use. Substance use & misuse. 2019:1-12. 47 48 49 50 51 52 53 54 55 56 57 58 33 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 2 3 Reporting checklist for qualitative study. 4 5 6 7 8 Page 9 10 Reporting Item Number 11 12 Title 13 14 #1 Exploring the concept of Problematic khat use in the Gurage 1 15 16 For peer review only 17 community, South-central Ethiopia: A qualitative study 18 19 20 21 22 Abstract 23 24 Objective: This study aimed at exploring how problematic khat use is 25 #2 3 26 27 characterized in the Gurage community, South-central Ethiopia. 28 29 30 Design: qualitative study. 31 32 33 Setting: Gurage community, South-central Ethiopia http://bmjopen.bmj.com/ 34 35 36 Participants: We conducted in-depth interviews with 14 khat users and 37 38 5 non-khat users, and 3 focus-group discussions with khat users. 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Methods: All participants were selected purposively based on their 43 44 exposure to khat or khat use. We used interview guide to explore 45 46 perception of participants about khat use and problematic khat use. We 47 48 49 analyzed the data thematically using Open Code software version 4.03. 50 51 We used iterative data collection and analysis, triangulation of methods 52 53 and respondent validation to ensure scientific rigour. 54 55 56 Findings: We identified three major themes: sociocultural khat use, 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from khat suse (khat addiction), and negative consequences of khat use. 1 2 3 Sociocultural khat use included a broad range of contexts and patterns 4 5 including use of khat for functional, social, cultural and religious 6 7 reasons. Khat addiction was mainly explained in terms of associated 8 9 khat withdrawal experiences, including harara/craving, and inability to 10 11 12 quit. We identified mental health, sexual life, physical health, social 13 14 and financial related negative consequences of khat use. The local 15 16 Foridiom Jezbapeer was used review to label subgroup only of individuals with khat suse 17 18 19 (khat addiction). 20 21 22 Conclusion: The study has identified what constitutes normative and 23 24 problematic khat use in the Gurage community in South-central 25 26 Ethiopia. Problematic khat use is broad concepts which include 27 28 29 frequency, reasons, contexts, negative consequences and addiction of 30 31 khat use. Insights generated can be used to inform future studies on 32 33 development of tools to measure problematic khat use. http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Introduction 43 44 Problem formulation #3 Across the literature, conceptualization of problematic, sociocultural 5 45 46 47 and recreational khat use has been an important research gap. 48 49 Problematic khat use, rather than khat use per se, is usually the interest 50 51 of the public, researchers and policymakers. Nevertheless, only few 52 53 54 previous studies were conducted on problematic khat use [33]. The 55 56 inconclusive reports about the different harms of khat use could also be 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from due to poor definition of problematic khat use. The Diagnostic 1 2 3 Statistical Manuel(DSM-5) definition of stimulant use disorders [34] 4 5 could have more important clinical utility (for severe cases) than 6 7 screening individuals with problematic khat use at earlier stage. Lack 8 9 of screening tool for problematic khat use, especially among nonclinical 10 11 12 cases could hamper efforts to curb the problem including early 13 14 identification and effective management of positive cases. Therefore, 15 16 Forthere is peera need for valid review problematic khat only use screening tools which 17 18 19 would facilitate clinical care in primary health care settings and for 20 21 future research. Although there are no strong validation studies, few 22 23 studies used Harmful Khat Use Scale [35] and Severity of Dependence 24 25 26 Scale to measure the construct problematic khat use [36], but a 27 28 systematic review [33] and exploratory studies[37] suggested broader 29 30 indicators of problematic khat use, including amount, frequency, 31 32

context and duration of khat session. http://bmjopen.bmj.com/ 33 34 35 36 Purpose or research #4 The aim of this study was to conceptualize problematic khat use from 5 37 question 38 the perspective of users and non-users in a dominantly rural setting, 39 40 Gurage, south-central Ethiopia. 41 on September 25, 2021 by guest. Protected copyright. 42 43 44 45 46 Methods 47 48 Qualitative approach and #5 The study employed a qualitative study design [40] which allowed for 6 49 research paradigm 50 the understanding and description of the experiences and perspectives 51 52 53 of people towards problematic khat use. This study was guided by the 54 55 consolidated criteria for reporting qualitative research (COREQ)[40]. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Researcher characteristics #6 Participants were also diverse in terms of socio-demographic 7 and 2 and reflexivity across the 3 characteristics. The field worker, the first author, has previous 4 paper 5 6 experience interviewing people about khat use. Above all, we tried to 7 8 purely and openly present the ideas of the participants without personal 9 10 impression interference. 11 12 13 14 15 16 Context #7 ForThe study peer was conducted review in the Gurage only zone; Southern Nations, 6 17 18 Nationalities and Peoples Region (SNNPR), Ethiopia. Ethiopian 19 20 Orthodox Christianity (48.17%) and Muslim (42.31%) are the two 21 22 23 dominant religions in Wolketie town, Gurage’s capital [38]. Peasant 24 25 farming is the main productive occupation in rural areas while petty 26 27 trading is more common in urban areas. The area is known for its khat 28 29 30 production and khat use [10]. Khat might have been introduced to 31 32 Gurage area by the remnants Ahmad Ibn Ibrahim al-Ghazi (1506 – 33 http://bmjopen.bmj.com/ 34 1543) army or neighboring Muslim Wolane or Oromo ethnic groups 35 36 37 [39]. 38 39 40 41 on September 25, 2021 by guest. Protected copyright. 42 Sampling strategy #8 We conducted face to face in-depth interviews with 14 current khat 6 43 44 users and five non-khat users. Twenty-one khat users participated in 45 46 47 focus group discussions (FGD). The first FGD consisted of six women, 48 49 the second and the third FGDs had seven and eight participants 50 51 respectively. Participants were selected purposively based on their 52 53 54 experience of khat use, and we also aimed for maximum variation 55 56 considering the socio-demographic characteristics of the participants. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Ethical issues pertaining to #9 Ethical clearance was obtained from the Institutional Review Board 7 2 human subjects 3 (IRB) of the College of Health Sciences, Addis Ababa University (Ref 4 5 6 008/18/Psy). In addition, we also obtained a letter of support from the 7 8 Gurage zone health department. Written informed consent was sought 9 10 and obtained from all participants before data collection. 11 12 13 14 15 16 Data collection methods #10ForParticipants peer were invited review by the community only health workers and the first 6 17 author did the informed consent. All the interviews and FGDs were 18 19 conducted in Amharic and tape-recorded. The first author, who has 20 good experience of interviewing khat users, did the interviews and 21 22 facilitated the focus group discussions assisted by a trained moderator. 23 The interviews took about 40 minutes and the FGDs about one hour in 24 25 average. The first and the last authors designed in-depth interview and 26 FGD guides. 27 28 29 Data collection #11 The guides mainly focused on the experiences and perceptions of the 6 30 instruments and participants regarding khat use and problematic khat use. 31 32 technologies 33 http://bmjopen.bmj.com/ 34 Units of study #12 Participants were diversified in several socio-demographic 8 35 36 characteristics, including sex, age, residence, religion, occupation and 37 educational status (Table 1). 38 39 Data processing #13 The data were transcribed verbatim in Amharic, then translated into 7 40 41 English by the first author and experienced research assistants. on September 25, 2021 by guest. Protected copyright. 42 43 Data analysis #14 . All interview and FGD translations were coded independently by the 7 44 45 first and fourth author (AM and CN). Iterative thematic analysis [41] 46 was done simultaneously with data collection. We used computer 47 48 software, open code 4.03 to manage and analyse the data [42]. 49 50 Techniques to enhance #15 Regarding to data quality management and rigour, the iterative process 7 51 52 trustworthiness of data collection, data analysis and checking unclear issues from the 53 participants added to the quality of the study. We did also data 54 55 triangulation from different sources, including religious fathers, key 56 informants for the culture and legal officers. 57 58 Results/findings 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from

1 Syntheses and #16 The major themes that emerged from the iterative thematic analysis 9 2 3 interpretation were: (1) sociocultural khat use, (2) suse “(addictive)” khat use, and (3) 4 negative consequences of khat use. The second theme had two 5 6 categories: withdrawal experiences, quitting khat use and 7 mirqanna/feeling high after khat use. The negative consequences were 8 9 categorized to mental health, sexual life, physical health, social and 10 financial. These themes are discussed with support of quotes directly 11 12 from the participants. 13 14 Links to empirical data #17 Withdrawal experiences of khat use 15 15 16 For peer review only 17 Other experiences reported associated with abnormal khat use included 18 19 withdrawal experiences when one was weaning off khat after heavy 20 21 22 usage. The study found many important psychological withdrawal 23 24 experiences of khat suse-addiction. Feeling depressed, irritable and 25 26 aggression were the most typical and commonly reported experiences. 27 28 29 One participant said, “I am usually against people’s communication 30 31 even aggressive when I didn’t chew khat. I remember that I once threw 32 33 away [smacked] my kid when she was talking about her school affairs.” http://bmjopen.bmj.com/ 34 35 (Age 67, urban, male) 36 37 38 39 Different participants reported different withdrawal symptoms, 40 41 including lack of motivation, unable or lack interest to function or on September 25, 2021 by guest. Protected copyright. 42 43 socialize, poor concentration and learning, unable to receive message, 44 45 and dukak (vivid unpleasant dream). Here are two dukak or vivid 46 47 48 unpleasant dream experiences; 49 50 51 One day my husband skipped chewing khat and went to bed. Then, he 52 53 spent the night spitting. In the morning, the bed sheet was wet. When I 54 55 56 ask him what was wrong with him. He said; “people were punishing me 57 58 with the smoke of red paper and I had been feeling burning sensation 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 41 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from for the whole night”. (Age 28, urban, women) 1 2 3 Another experience: 4 5 6 7 My experience was...ehm... usually a man would hold my hair and hung 8 9 me or put me into a hole and I would wake up in panic. (Age 38, male, 10 11 urban) 12 13 14 15 16 For peer review only 17 Discussion 18 19 Intergration with prior #18 In this study, our aim was to conceptualize problematic khat us in a 24 20 work, implications, 21 predominantly rural setting in south-central Ethiopia. The study helped 22 transferability and 23 24 contribution(s) to the field to answer the question: what constitutes problematic khat use? Our 25 26 results could inform development of screening tool to measure 27 28 problematic khat use. 29 30 31 Although there was also sociocultural khat use pattern, khat addiction 32 33 and negative consequences of khat use constitute problematic khat use. http://bmjopen.bmj.com/ 34 The local term khat suse, semantically equivalent to khat addiction 35 36 which doesn’t conclusively infer to DSM-5 stimulant use disorder 37 definition[43]. Khat suse/khat addiction could only qualify impaired 38 39 control and pharmacological criteria among the criteria of stimulant use 40 disorders. Khat suse/ khat addiction shares some characteristics from 41 on September 25, 2021 by guest. Protected copyright. 42 other substances use disorders than stimulant use disorders. For 43 example, it has a similar functional consequences of cannabis use 44 45 disorder-amotivational syndrome[43]. In the case of khat suse, the local 46 idiom Jezba could be conceptually related but broader. In the current 47 48 study setting, the use of the word, Jezba, indicates the existing stigma 49 and it could also be a good explanation how much sever form of 50 51 problematic khat use is well recognized in the setting since it has a 52 negative connotation. Khat suse is also indicated by frequent yawning 53 54 when not using the khat. This withdrawal criterion is similar with 55 opioid withdrawal[43]. 56 57 58 Limitations #19 Regarding methodological concern of the study, it didn’t include the 26 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 41 BMJ Open: first published as 10.1136/bmjopen-2020-037907 on 12 October 2020. Downloaded from perspectives of families of khat users in detail, but the perspective of 1 2 3 non-khat users in general were included. The analysis of the study was 4 5 also not theorized using a well-established existing model or theory, but 6 7 this allows the data to speak for itself without imposing some other 8 9 framework that may poorly fit to the data. This qualitative study, tapped 10 11 12 into indigenous knowledge systems, was community-based which make 13 14 it very strong to define problematic khat use and develop an 15 16 Forunderstanding peer of its domainsreview broadly. only 17 18 19 20 Other 21 22 Conflicts of interest #20 None declared 28 23 24 Funding #21 This work was supported through the DELTAS Africa Initiative [DEL- 27 25 26 27 15-01]. The DELTAS Africa Initiative is an 28 29 independent funding scheme of the African Academy of Sciences 30 31 (AAS)’s Alliance for Accelerating Excellence in Science in Africa 32 http://bmjopen.bmj.com/ 33 (AESA) and supported by the New Partnership for Africa’s 34 35 36 Development Planning and Coordinating Agency (NEPAD Agency) 37 38 with funding from the Wellcome Trust [DEL-15-01] and the UK 39 40