Abstract
Background: Problematic khat use, not khat use per se, is a public health and social concern for the public, researchers, and policy makers. However, the construct problematic khat use is not well-established and not fully recognized in the modern definition of substance use disorders including Diagnostic Statistic Manual (DSM-5) and international classification of diseases (ICD-10), although DSM-5 included it in the “stimulant use disorder” category. Existing scoping reviews have focused on khat use, which could not allow the differentiation of khat use from problematic khat use. Thus, the aim of this systematic review was to define and validate the construct problematic khat use. Methods: This systematic review was reported following the PRISMA guidelines. We searched all English language studies without publication date restriction from 5 databases; PubMed, EMBASE, psychINFO, SocINDEX, and Google scholar. All studies that defined, explored, evaluated, or measured the construct problematic khat use were included. Adapted data extraction tool and criteria for quality evaluation were employed. We presented the results in tables and thematic synthesis of the major findings. Result: Overall, 30 qualitative and cross-sectional design studies were included. Associated harms with khat use, an increased amount used, increased frequency of use, and withdrawal experiences were indicators of problematic khat use. Using khat on an average of 3 or more times per week and using other psychoactive substances during and after khat use were frequently used to define problematic khat use. The most frequently reported withdrawal symptoms were depressed mood, irritability, fatigue, lack of motivation, increased sleep, and appetite. The existing measures (severity of dependence scale and DSM-5) of problematic khat use had psychometrically acceptable properties in terms of construct, criterion, and convergent validity, but they are poor in terms of other domains of validity including content, conceptual, and semantic validity. Conclusion: Problematic khat use constitutes, but is not limited to, harms, increased use over time, and frequent engagement in other psychoactive substances misuse. Khat use is different from problematic khat use since it is occasional and used for prayer, social, and functional reasons. Strong empirical studies that could establish thresholds for patterns of problematic khat use and a culturally suitable problematic khat use measures that follows a bottom-up approach of scale development are warranted.
Introduction
Khat (Catha edulis) is a plant with psychoactive properties, and the leaves and shoots are chewed for this effect. Khat is considered a “natural amphetamine” containing amphetamine-like stimulant substances such as cathinone and cathine [1]. These 2 substances that are under international control, the main one being cathinone, but the plant khat is not [2]. In the past, khat use was exclusively embedded in culture and practiced as a social custom, among subgroups, especially in Ethiopia, Yemen, Somalia, and Kenya [3]. For example, elder Muslim men in certain ethnic groups or regions chewed khat for concentration during studying religious writings and to stay alert during night prayer. Recently, khat use became popular among other groups of people, and they used it to increase concentration and performance during trading, farming, academic activities, and for socialization and leisure activities [4]. Factional khat use could also include chewing khat to cope with traumatic experiences [5].
Estimates of prevalence range between about one-sixth and one-half of adults using khat, with variation across countries and groups within countries [2, 4, 6]. The studies reported male sex and Muslim religion as important associated factors of khat use. There were not only negative physical, psychiatric, and socioeconomic outcomes of khat use [7, 8] but also harms. Physical health impacts such as myocardial infarction and other cardiovascular disorders, hepatoxicity, hypertension, upper gastrointestinal tract (gastritis, enteritis), hemorrhoids, and impaired male sexual function [2, 8, 9] were reported to be associated with khat use. Association between khat use and severe mental illness are controversial [10‒12]. Studies reported association of heavy khat use with psychotic symptoms in Somalia and among Somali refugees in Kenya and Europe [10, 13, 14]. Among special populations such as individuals with PTSD experiences, the more khat consumption was significantly associated with paranoid ideation [5].
The reported relationship between socioeconomic issues and khat use was also contradictory. Khat users reported increased functioning and socialization, believed to be beneficial, and relationship problems between children and parents, and serious financial problems, as negative consequences [8, 15].
There has been a steady increase in the study of khat use, but not much work done on defining problematic khat use, only limited to a few attempts recently [16‒18]. According to reports from Ethiopia, the prevalence of problematic khat use was estimated to be 20% in the general population [19] and 80% among khat users [20]. A study from the United Kingdom also showed 31% of dependency among khat users of Yemeni origin in the United Kingdom as measured by diagnostic criteria of the American Psychiatric Association Diagnostic Statistic Manual (DSM IV, 1994) [21]. Fifty one percent of khat users in Yemen [17] and 52.2% in Saud Arabia [22] were also estimated to use khat in a problematic pattern. In all these studies, psychological dependency was used as a chief marker of problematic khat use.
While risky/hazardous and problematic/harmful use are clearly defined for alcohol and cannabis [23‒25], there is no official equivalent definition for problematic khat use. Although khat is one of the oldest psychoactive substances used by humans, little is known about which khat use pattern would lead to the above negative consequences. There are scoping and critical reviews on khat use and associated harms [8, 26], but there is a dearth of evidence on the definition and indicators of problematic khat use.
The existing reviews were not systematic, but important to generate a hypothesis that problematic khat use is a valid construct. The reviews [27] define “problematic khat use” as excessive use pattern that is associated to negative health consequences. We still need to establish threshold for excess use pattern and other (if any) additional criteria or domains of problematic khat use should be systematically reviewed. Systematically defining problematic khat use will help the development of public health prevention and clinical intervention guidelines. Therefore, the current systematic review aimed to define and validate the construct “problematic khat use” from the existing literature and whether further studies are needed to address the issue.
Methodology
Search Strategy
The search was conducted in all English language studies without publication date restriction till June 10, 2018. We included databases such as PubMed, EMBASE, PsychINFO, SocINDEX, and Google scholar. The following search terms were used: (khat OR qat OR miraa OR jaad OR qaad OR kat OR Kafta OR ghat OR chat OR tschat OR “Abessinischer Tee” OR “Abyssinian Tea” OR “Somali Tea” OR “Arabian Tea” OR “natural Amphetamine” OR Catha [MeSH Terms]). A Google scholar search was performed using the term “khat.” We searched for title and abstract only, and backward searching of all included papers was also conducted to find additional relevant studies. Our systematic review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) on August 26, 2017, and was last updated on June 28, 2018 (registration number [CRD42017064183]).
Study Selection
Eligibility
Studies that defined, explored, evaluated, or measured the construct “problematic khat use” among human khat users were included. Studies reporting lifetime and current prevalence of khat use without indicating the problematic pattern of use were not included in this review, since khat use per se was not the concern of interest of the community, especially in countries where cultural use is common such as in Eastern Africa and Yemen [7, 28]. From previous review, many studies didn’t indicate the dose – effect relationship of khat use and harms especially on mental illness [29]. The current systematic review focused on studies that indicated dose – effect relationship of khat use and associated harms or studies with clear indication of problematic khat use.
All qualitative, quantitative, and mixed methods studies were included. The qualitative studies were helpful in understanding how and why khat chewing was considered or experienced as problematic. The quantitative studies were essential in defining problematic khat use and establishing its indicators. Studies on psychometric properties of assessment tools were also used to examine the validity of the construct “problematic khat use,” dimensionality, item wording, and thresholds. Studies that focused on pure cathinone alone were excluded because cathinone is a well-established drug of abuse, which is more harmful than the plant khat that carries low concentration of cathinone.
Study Selection and Data Extraction
The first 2 researchers independently screened the articles, did the data extraction, and risk of bias assessment. Emerging discrepancies were resolved collectively. Data were extracted using adapted data extraction tool [30, 31]. The major themes extracted were definitions, concepts, terminologies, indicators, prevalence, and associated factors of problematic khat use. From the psychometric studies, we extracted any reliability and validity characteristics, thresholds, name of the tool, and participants.
Assessment of Risk of Bias and Quality of the Studies
We used Terwee et al. [32] criteria to assess the quality of psychometric studies. The quality assessment tool focuses on 9 major areas of psychometric properties [32]. The instrument includes content validity, internal consistency, criterion validity, construct validity, agreement, reliability, responsiveness, floor and ceiling effect, and interpretability. The authors assigned measures a positive (+), intermediate (?), negative (–) rating, or a rating of (0) when there was no information regarding the relevant criteria. For the cross-sectional survey and qualitative studies, we administered the Assess the Qquality of Cross-sectional Studies tools [33] and the Joanna Briggs Institute Critical Appraisal Tool for qualitative studies [34]. The major data extracted are presented in tables. Mixed method studies are presented under qualitative studies to make the interpretation of the findings from qualitative and quantitative studies more clear. We found it difficult to report a diagnostic status of the quality of the studies. Studies, using Delphi panel, suggest that quality numerical scales can be problematic because the outputs from assessment checklists are not linear and it is also difficult to sum up or weight making them unpredictable at assessing study quality [33, 35]. We synthesized and presented the findings in a narrative way, following thematic analysis.
Results
Of 9,677 studies identified by the search, 30 publications were included in the synthesis (Fig. 1). Many of the excluded studies were focused on cathinone and/or khat use per se without reporting anything about problematic khat use. The studies included in the systematic review were classified in to 3 main categories: (a) qualitative and mixed methods studies that suggested subjective criteria of problematic khat use, (b) cross-sectional studies with indicators of problematic khat use, and (c) psychometric studies that reported validity and reliability of the different tools used to measure problematic khat use.
Regarding geographical area, many of the studies were conducted among Eastern Africa immigrant populations in Australia, USA, UK, Kenya, and Sweden (n = 13) and others from Ethiopia (n = 10), Saudi Arabia (n = 3), Somalia (n = 3), and Yemen (n = 1).
Qualitative Studies
Using Joanna Briggs Institute critical appraisal tool for qualitative studies, all of the above articles included more than 50% of the items of the tool; the major limitation of the studies was comprehensiveness. Many of the studies inadequately defined the idea of problematic khat use. Few of the studies [36‒40] had clear objectives and predesigned interview or FGD topic guides about the different dimensions of problematic khat use.
From the above studies, we found perceived indicators of both acceptable and problematic khat use. There were also intricate issues of religion (Muslim) and social and symbolic values [41, 42]. Chewing for religious, functional, leisure, and social reasons [6, 37, 38] and chewing khat for perceived health benefits [43] were indicators of culturally acceptable khat use. On the other hand, withdrawal experiences of khat use [37, 38], after chewing distressful experiences [37], perceived addictive nature of khat [43, 44], and associated harms [6, 36, 37, 40, 41, 43, 45], were reported indicators of problematic khat use. Other studies [6, 37, 45] also reported impairment of control over khat use as indicator of problematic khat use. The use of khat with other psychoactive substances was also mentioned as a pattern of use deviant from the “normal” khat chewing culture and it was an indicator of problematic khat use [36, 46]. The post chewing experiences of khat use, mirqaan, were reported to be both distressing [37] and pleasurable [41] depending on the amount of khat.
Quantitative Studies
All the 11 quantitative studies used cross-sectional study designs. Regarding the quality of the articles, many studies, 7 out of 11 studies, didn’t use adequate sample size, which was <100 participants (Table 2). Thus, they did not use appropriate statistical analysis, poorly addressed confounding variables, and didn’t identify subgroups with objective criteria. The remaining 2 studies with big sample size did not correctly measure the outcome variable, problematic khat use, using a valid instrument [39, 47].
Three studies [13, 22, 47, 48] reported prevalence and indicators of psychological dependence in khat use as measured by Severity of Dependence Scale (SDS) except one study [47] that employed DAST. Others [39] employed structured questionnaires to measure withdrawal experiences associated with problematic khat use or they [49, 50] reported subjective experiences of addiction without formal measurement.
Psychological and physical symptoms [22, 51, 52], withdrawal experiences [37, 39, 53], distressing experiences after khat use [49], and impairment of control or stopping [54] were reported indicators of problematic khat use. Nencini et al. [52] reported that reduced respiratory and body temperature was observed among customary khat users, similar to amphetamine abuse. The specific withdrawal symptoms reported were yawning, craving, mood disturbance (feeling depressed, irritability), feeling fatigue and dizziness, hot feeling, slight tremor of the tongue and body, increased appetite, sleep disturbance, headache, and blurring vision [37‒39, 43] as shown in Tables 1 and 2.
Increased frequency of use, the minimum was once in a month [47], ≥8 days in a month in another study [49], and the maximum was daily use [16], was an important indicator of problematic khat use. People who chewed for >5 h per session [22], or about more than 12 h per week [47], were also perceived as a sign of being problematic khat user. Another study indicated that chewing more than 24 h without interruption as a sign of Hazardous khat use or bingekhat use [13].It was very difficult to have a standard measure for khat, but 2 studies [16, 17] found more than 1.5 or 1.25, and another study [10] reported >2 bundles of khat per day as an indicator of problematic khat use. Early age of starting khat chewing at around 18.5 years [10, 47] was also reported as an important predictor of problematic khat use. Odenwald et al. [10] showed that people with psychotic symptoms had started to chew khat about 8.6 years earlier in life than matched controls.
Psychometric Studies
We assessed and observed poor quality in terms of applying the recommended methodology of tool development and validation across the studies. About 50 and 67% of the studies were poor in examining construct and content validity, respectively. Problematic khat use could not be comparably distributed among other psychoactive substance users and nonusers as well as across gender, but the studies didn’t report subgroup analysis such as differential item functioning. One study [55] suggested gender by age interaction for problematic khat use. Many of the studies reported that the construct psychological dependence is valid and applicable for problematic khat use [17, 21, 55, 56]. Few studies applied constructs-khat use disorder[16] and harmful khat use[19]to denote problematic khat use.
SDS has been frequently used to screen psychological dependence in problematic khat use. There was a discrepancy in the psychometric properties of SDS. Two studies [17, 56] suggested SDS is a unidimensional scale, but another study [55] found 2 factors where items 3 and 4 could be collapsed into one domain desire to stop. Kassim et al. [17] recommended ≥6 points cutoff, whereas Duresso et al. [56] established ≥3 using DSM-5 stimulant use disorder as a gold standard and ≥5 cutoff for another study [55].
Kassim et al. [21] and Duresso et al. [16] suggested that problematic khat users endorsed many criteria of DSM-IV substance use dependence and DSM-5 stimulant use disorder criteria. Duresso et al. [16] found a unifactorial model to both DSM-IV dependence and DSM-5 substance use disorder items for problematic khat use, but Kassim et al. [21] found 2 factors: behavioral and withdrawal symptoms of DSM-IV for problematic khat use. Harmful khat use measure [19]was with 11 items and ≥11 was the cutoff point. Exploratory factor analysis revealed 2 significant dimensions of the scale, but confirmatory factor analysis was not reported. The process of designing the scale was assembling items from existing measures of other psychoactive substances use. The study lacks conceptual validity as the process was not informed by qualitative study. The detail characteristics and psychometric properties of the individual studies is presented in Table 3.
Discussion
This systematic review is very important to answer the question: what constitutes problematic khat use? This will have important clinical and policy implication. The current review includes studies with different study designs that allowed us to have a comprehensive understanding about the validity of the construct problematic khat use from different dimensions: theoretical and empirical perspectives.
Compared to other psychoactive substances misuse, little progress is made on defining different levels of “problematic khat use” in particular and “khat use-related harms” in general. Institutions such as National Institute on Alcohol Abuse and Alcoholism, European Monitoring Centre for Drugs and Drug Addiction, and other scientists with extensive research backgrounds on substance misuse clearly defined risky, hazardous, harmful, heavy, and substance use disorders, especially for alcohol and cannabis [25, 57, 58]. There are more than 44 instruments for cannabis [59] and alcohol [60] misuse, which are screening or diagnostic tools to be used at different settings. Therefore, there are important lessons that could be adopted for studies on problematic khat use.
Indicators of problematic khat use, listed in box 1, were consistent with other psychoactive substances misuse. Patterns of use, dependency, tolerance, withdrawal experiences, and negative consequences were the main indicators across different problematic substances use [58, 59, 61, 62]. Absence of controlled studies for dependence, tolerance, and withdrawal symptoms of problematic khat use is a major limitation of evidence across the individual studies in this systematic review. Cannabis withdrawal was recently recognized as a criterion for cannabis use disorder in the DSM-5 using evidence from longitudinal studies [23, 58]. The same could be done to establish khat withdrawal symptoms. Regarding harms, 2 very strong studies, using Delphi procedure, classified khat among the lowest risk list of drugs [61, 62]. The studies were actually from the Western countries where the problematic status of khat could not be comparable to other countries such as Ethiopia and Yemen where khat chewing is widespread.
Increased khat use over time, which was indicated by amount, frequency, and duration of session, was very important to define problematic khat use. Recent studies on definition of substance use disorders recommend the use of heavy substance use over time for defining substance use disorders for different reasons [63]. Heavy use over time clearly explains both the social life and neurobiological changes associated with current status of the substance use on the individual. In addition, heavy use over time would avoid the cultural specificity and sensitivity of the domains of measures such as loss of control and social consequences associated with the substance. Therefore, substances that are embedded to the culture such as khat use, heavy khat use over time as suggested by the empirical studies could be used as an important indicator of problematic khat use.
Additional indicator of problematic khat use that is different from other substances misuse definition was using other psychoactive substances with or after khat chewing. This doesn’t only mean that the individual will be affected by the adverse consequences of the poly psychoactive substances, but it was also reported that using other psychoactive substances with khat is against the cultural definition of khat use. Khat use can be a gateway drug for other psychoactive substances misuse [46], which could be an evidence for indirect contribution of khat for morbidity and mortality.
The use of DSM-IV [21] and DSM-5 [16] criteria for amphetamine-like substance dependence and stimulant use disorders respectively for problematic khat use were also important for clinical diagnosis but not for screening of problematic khat use by nonmental health professionals. The studies suggested that the 2 measures are psychometrically acceptable to apply for problematic khat use, but their focus was mainly on structural validity. The studies didn’t properly conceptualize problematic khat use, and they were not informed by qualitative studies. The recommended emerging area of research on tool development would be exploratory and theory driven [64]. Cochrane handbook of a systematic review also recommended the use of qualitative studies in the development of measurements [31]. SDS is the most widely used measure of problematic khat use, but SDS only measures a narrow concept of psychological dependence. Its psychometric properties, dimensionality, and threshold were not also yet well-established. Severity of dependency scale was also a marker of psychological dependence [65], which didn’t include other content areas of problematic khat use. The validity of SDS was also in question to screen problematic cannabis use in the general adolescent population [66].
The studies didn’t consider subgroup analysis across important variables, which suggest significant variation in problematic khat use. The psychometric studies from Ethiopia [16, 19] were also mainly focused on college students, which could affect acceptability and appropriateness for the general population, especially the rural population. Another limitation of the current systematic review could be the inability to report pooled quantitative results because of the heterogamous nature of studies.
Conclusion
Comparative studies are necessary to establish symptoms of dependence and withdrawal experiences of problematic khat use. Strong empirical studies that could establish thresholds for amount, duration, and frequency for problematic khat use are warranted. Meanwhile, novel psychometric methods such as item response theory could be employed to identify important indicators of problematic khat use such as frequency, dose, and duration of khat use session. Measuring khat alkaloid and establishing standard amount of khat could be important for different reasons, but this might not facilitate epidemiological studies. Therefore, in order to facilitate early detection and intervention, culturally suitable problematic khat use screening tools with entails behavioral criteria are also important. Future studies should also clearly define different types of problematic khat use such as dependency, harmful, hazardous, or heavy khat use. Future studies are also important to explore additional indicators of problematic khat use (if any).
Finally, problematic khat use constituted, but didn’t limit to, harms, increased use over time, and frequent engagement in concurrent use of other psychoactive substances which was different from occasional khat chewing used for prayer, social, and functional reasons.
Acknowledgment
The authors would like to acknowledge the African mental health research initiative of DELTAS Africa Initiative for supporting the first 2 authors through African mental health research initiative PhD fellowship.
Statement of Ethics
Since this is a review paper, guidelines for human studies and animal welfare regulations are not directly applicable, but this study was reviewed, and ethical approval was secured from institutional Review Board of College of Health Sciences, Addis Ababa University under protocol number 008/18/psy.
Disclosure Statement
The authors declare that they have no competing interests.
Funding Source
This work was supported through the DELTAS Africa Initiative (DEL-15-01). The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences Alliance for Accelerating Excellence in Science in Africa and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency with funding from the Wellcome Trust (DEL-15-01) and the UK government. The views expressed in this publication are those of the author(s) and not necessarily those of African Academy of Sciences, New Partnership for Africa’s Development Planning and Coordinating Agency, Wellcome Trust, or the UK government.
Author Contributions
A.M. and C.N.: contributed in the searching, screening, and risk of bias assessment. A.F., S.N., and S.T.: contributed on protocol design and manuscript write-up.