Prevalence of khat chewing and its effect on academic performance in Sidama zone, Southern Ethiopia.

Background: Khat use is a well-established public health problem in Yemen, Arabian Peninsula, and Ethiopia. Along with its large scale production, the magnitude of khat use is increasing among students. Objective: This study was intended to assess the prevalence, determinants, and effect of khat use on academic performance of high school students in Sidama Ethiopia. Methodology: A cross-sectional study was conducted from April to June 2015. We used a stratified sampling technique to draw a total of 1,577students. The data was collected using self-administered questionnaires and analyzed using SPSS version 20. Descriptive statistics and logistic regression models were used to determine the prevalence, effects, and predictors of khat use. Result: The life time and current prevalence of khat use were 14.6% and 13%, respectively. Smoking cigarette (AOR=5.1, 95% C.I: 2.3-14.3), drinking alcohol (AOR=3.0, 95% C.I: 1.4-6.3), having a family growing khat (AOR=2.0, 95% C.I: 1.1-2.5), having friend chewing khat (AOR=3. 95% C.I: 2.0-4.6), were some of factors that increased the odds of students’ khat use. Student’s khat use increased the odds of student’s poor academic performance (AOR=2.1, 95% C.I: 1.1-3.9). Conclusion: The prevalence of khat use in high khat producing districts of Sidama and its contribution to poor academic performance demand prompt intervention.


Introduction
Khat is known for its adverse health effects 1,2 . The growing research evidence shows that khat is a risk factor for different cardio-vascular disorders such as hypertension, ischemic heart diseases, and stroke 3,4 . Khat is also a risk factor for poor oral health 5 , psychosis 6 , mental distress 7 , anemia 8 , poor quality of life 9 , disturbed sleep pattern [10][11][12] , and duodenal ulcer 13 .
The adverse effect of khat on reproductive health includes genotoxic and teratogenic effects on the fetus, causing low birth weight 14, 15 . Moreover, khat also affects the potency of male sexuality by affecting spermatogenesis and plasma testosterone concentration 14, 16 . Growing evidence shows that khat use is one of the risk factors for taking risky sexual behaviors which fuel the widespread effect of sexually transmitted infections (STIs)especially Human Immunodeficiency Virus (HIV) infection [17][18][19] .
Besides khat's effect on mental, physical, and reproductive health, it is one of the factors causing socio-economic problems. These include; social isolation, family breakdown, neglect of social responsibilities 20 and spending more than half of the domestic budget on khat 21 . In line with this, one study reported that khat use is associated with specific impairments in behavioral control which triggers conflict amongst human 22,23 . Moreover, khat is also reported as a risk factor for poor working memory 24-26, poor academic performance 27,28 , reduced productivity and increased daily financial expenditure 29 . Despite these harmful consequences, the cultivation and use of khat is on arise especially in Yemen, Arabian peninsula, East Africa, and Ethiopia in particular 30,31 .
In Ethiopia, khat has started to rapidly replace the precious cereal, coffee, fruits and other crops found in the highlands of Ethiopia. Farmers prefer khat due to different reasons including, it's profitability as a cash crop than others, less vulnerability to drought, and less labor power cost required for its cultivation. However, khat producing farmers in Ethiopia start consuming khat and this culture is also continued expanding to the nearby secondary schools and urban dwellers 29 .
Khat or other psycho-active substance use at an early age leads to a complicated type of substance use in the remaining adulthood period of an individual. Youths' substance use, therefore, contributes to the creation of a community with; substance use dependence, juvenile delinquencies, crimes, socio-economic and other public health problems. The Magnitude of students' khat use in Ethiopia ranges from 13.4% to 41% for lifetime and 6.3%-33.1% for current (past 30 days) 32 . Among the top students' expressed reasons for khat use are; to get energized for study, to stay awake or avoid sleep and to increase academic performance 32,33 . Factors increasing the odds of students' khat use include; male gender, having friend chewing khat, having family members chewing khat, Muslim religion, pocket money, living alone during school age, cigarette smoking, financial freedom, parents' educational levels, living in rural area, and others 32,34,35 .
As one of the highlands of Ethiopia growing khat, Sidamazone has a lion's share of khat production and marketing. Almost most of the districts in the zone are khat producing. Especially, Aleta Chuko, Dale, Shebedino, and Wendo Genet districts are among the top of the list 36 . As a result, the trend which was observed in previously khat producing regions seems to happen in newly khat producing zones 29 . Nevertheless, as per the investigators best systematic and rigorous literature search, there is no published research report revealing the effects of khat production and marketing activities particularly on high school students of Sidama-zone. Therefore, this study was conducted to establish prevalence of khat use, factors affecting students' khat use, and also the effect of students' khat use on their academic performance.

Method
This study is institution-based and cross-sectional in design. It was conducted from April to June 2015 to assess the prevalence and factors determining khat chewing among high school students of Sidama zone, Southern Ethiopia. This study was conducted in Sidama zone. Sidama zone is one of the khat producing zones of Ethiopia located at the Southern Nations Nationalities and Peoples Region (SNNPR). The capital of the region is Hawassa town located 275 km south of Addis Ababa, the capital city of Ethiopia. There are 21 districts in Sidama zone among which seven of them are top khat producing areas, according to the zonal agricultural office (personal communication). Considering the proportion of farmers growing khat, the first four high khat producing areas of the Zone were included: Shebedino District, Aleta Chuko District, Dara District and Wondo Genet District. During the study period, Sidama zone had 70,000 students enrolled in 40 high schools, of which 11 high schools were in the four selected districts. In total, there were 17.863 students enrolled in these high schools of which 8,645 were females. All districts had four high schools, except the three high schools in Wondogenet. Two in each district, a total of eight high schools, were selected by using simple random sampling technique.
Minimum sample size was determined using open EPI Software version 3.03. After calculating for all possible sample sizes, we selected the bigger sample size sufficient enough to address all research objectives. The bigger sample size obtained in this study was calculated using the reported proportions of khat chewing for urban(8.73%) and rural (4.52%) at 95% confidence level and power of 90% 37 . The study from which we took the proportion showed that, living in an urban area was an exposure variable than living in rural. Thus, the calculated minimum sample size was 1,558 high school students. Considering 6% non response, the final minimum size was determined to be 1,655 students.
The study participants were selected using stratified sampling technique. Considering the relative size of students in each school, the sample size was distributed proportionately. Specific sections in each grade were selected by using simple random sampling technique. The stratification considered sex and grade levels of students to allocate students proportionally and on the assumption that student's grade level and sex would affect student's khat chewing behavior. The total sample size was distributed proportionally to the selected grades and sections based on total number of students in each year of study according to sex. Finally, individual students fulfilling the inclusion criterion were randomly selected to be study participants. The questionnaire was prepared by reviewing relevant literatures 34,35,38 . It was customized to fit the context and translated to the local language Amharic and Afo-Sidamu so that the students could understand the questions. A pre-test was performed on 10% of the sampled subjects at Tula High School found in Tula sub-city of Hawassa town. The questionnaires were distributed by trained data collectors.
The collected data was cleaned, coded, entered and analyzed using SPSS version 20.0 statistical software. Descriptive and summary statistics were done. Both bivariate and multivariate logistic regression analysis was used to determine the association of each independent variables with the dependent variable. The outcome variable 'Students' khat use' was a categorical variable with "Yes"/"No" re+sponse which follows a binomial distribution. Thus we checked the effects of the co-variates against this outcome variable. The list of co-variates include sex, student's grade level, living status during school age, parents educational level, family substance use history, peer influence, family pressure, other substance use history, having family growing khat, participating in khat production and marketing activities, amount of monthly pocket money, and living in urban or rural area.
We also assessed the effects of factors contributing for students' academic performance. This outcome variable is a categorical variable with possible responses such as "poor academic performance" and "better academic performance". Student past semester grades were used to classify student academic performance as "Poor" if he/ she scored < 50 percent and "better" if it was ≥ 50%. Covariates such as, sex, living area (with responses "Ur-ban" and "Rural"), father's educational status, student's participation in khat production and marketing activities, and student's current khat chewing status were all computed to see their effect on student's academic status. On the outcome variables logistic regression model was fitted. Odds ratio with their 95% confidence intervals were computed to identify the presence and strength of association and statistical significance was affirmed if p < 0.05. To minimize the effect of confounding stratification was done by variables, which were assumed to have a confounding effect. In addition, multivariate analysis was used to control confounding variables. Variables that showed statistically significant association and main variables which could have an effect on the outcome were also included in the multivariate model. Those variables with P value ≤ 0.20 were all considered in the multivariate analysis step.
The operational definitions given to the most important variables were defined as follows. "lifetime khat use' defined as respondents who admitted to having ever used khat. "Current khat use" implying the proportion of students who used khat within 30 days preceeding the date of the study. 'Poor academic performance' denoted those students who scored, in their recent past semester exam, below 50% or failed in the previous semester. Ethical clearance was obtained from Institutional Review Board of College of Medicine and Health Sciences, Hawassa University. To ensure voluntary participation of each participant, a written and signed informed consent (assent) was obtained from each student. Additionally, the confidentiality of the information was assured by using anonymous questionnaires and by maintenance of the data in a safe and protected place.

Socio-demographic characteristics of study participants.
This study included 1,577 high school students out of 1,655 sampled study participants making the response rate 95.3%. The majority 479 (54.5%) of the study subjects were male and 1,290 (79.9%) of them were 15-19 years of age. The mean age of the respondents was 16.7 years (SD=2.3). One thousand seventy eight (68.4%) of the respondents were protestant Christians, followed by orthodox (11.6%). With regard to the ethnic group distribution the largest proportion belongs to Sidama (85.7%). The vast majority (86.6%) students were unmar-ried where as very few (6.7%) reported currently living in formal marital engagement (Table-1).

Magnitude & patterns of khat chewing, substance use, and sexual practices
The lifetime and current prevalence of khat use among high school students of Sidama zone were 230(14.6%) and 205(13%) respectively. Out of current khat chewers 43(21.0%) were female students. Among those students currently chewing khat, 64(31.2%) reported that they already developed an addiction to khat. But the vast majority of all students using khat, 164 (80%) reported their desire to stop chewing khat (Table-2). Among those students currently using khat, 32(18.5%) reported that they were chewing khat on a daily basis, whereas 12(5.9%) reported chewing khat 2-3 days per week. The proportion of students who chew khat once in a week was 62 (30.2%). This makes the overall proportion of students who chew khat on a regular basis 112(54.6%). The remaining 93(45.4%) of students reported that they chew khat occasionally. The reported students' average financial expense for buying khat was 16.00 birr with an amount ranging from 1.00 -100.00 birr. These students also reported that they increase the amount of their khat use during times of an increased khat production (20.0%) and also when khat prices get cheaper (17.6%). Moreover the vast majority (63.3%) of students chewed khat for greater than six months. Out of this 44.1% chewed khat for greater than one year. This study further assessed the types of other substances used by high school students of Sidama zone. The findings of this study reveal that 59(3.7%) of students reported to have ever used alcohol whereas 38(2.4%) had been smoking cigarettes in their life time. The shocking finding with respect to cigarettes is the information that almost all students who participated in this study (80%) reported being exposed to second hand cigarette smoke (Table-2). The findings of this study also depicted the relationship between substance use and students' sexual practice. In light of this 56(24.4%) reported that they have sex as a result of their khat use. The overall students' sexual practice history, which was committed outside of formal marital engagement, was 280 (17.8%). The proportion of students who commit casual and unprotected sex outside of formal marital engagement was 126(8.0%) ( Table-2).

Reasons for khat use
The reported reasons for Sidama high school students' khat use were as follows. In accordance with their rank order the first reason mentioned was "to stay awake and get energized for study" (52.4%). The remaining were; 'peer influence' (27.2%), ' to avoid stress' (8.4%), 'for getting happiness or pleasure' (7.9%), and ' due to an addiction' (4.2%).

Association of khat use with academic performance
This study also examined whether students' khat use history had an influence on students' academic performance or not. The multivariate logistic regression analysis revealed that there is a statistically significant association (AOR=2.0, 95%CI: 1.1-3.9). As per this analysis the likelihood of performing poorly or failing in academics is two times higher for those students who were using khat currently than those who were not chewing khat. Other factors found increasing the odds of students' poor academic performance include, being female (AOR=2.1, 95%CI: 1.3-3.6), living in urban area (AOR=1.7, 95% CI: 1.1-2.7), and also having a father with academic status of elementary education and below (AOR=1.8, 95% CI: 1.1-3.1) ( Table-3).

Determinants of students khat use
During the bivariate analysis, factors found to be significantly associated with alcohol use were: sex, living alone during school age, place of residence, having a family member chewing khat, having a family member smoking cigarettes, having a family growing khat, participating in khat production and marketing activities, drinking alcohol, smoking cigarettes, practicing sex, and having khat user friend/s. However, during the multiple logistic regression analysis: sex, having a family member chewing khat, having family member smoking cigarettes, having a family growing khat, drinking alcohol, smoking cigarettes, practicing sex, and having khat user friend/s were factors associated with khat chewing at statistically significant levels. Male students were 2.5 times more likely to have a khat chewing habit than females (AOR=2.5, 95% CI: 1.6-3.8) ( Table 4).
Those students who a had family history of khat use were two times more likely to have khat chewing than their counterparts (AOR=2.0, 95% CI: 1.2-3.0). Similarly, having a friend who used khat increased the odds of students chewing khat by three times more than those who didn't have a friends chewing khat (AOR = 3.1, 95% CI: 2.0-4.6). Most importantly, in contrast to students who had never smoked (AOR=5.7, 95% CI: 2.3-14.3), smoking cigarettes increased the risk of students' khat chewing by about six fold. Similarly, students' alcohol use also increased the risk of khat chewing by three fold, as compared to the students who did not drink alcohol (AOR=3.0, 95% CI: 1.4-6.3). In addition, having family growing khat was another factor that increased the risk of students' khat use by twofold. (AOR=2.0, 95% CI: 1.1-2.5). Lastly, practicing sex was another factor which increased the odds of students khat use by two fold (AOR=2.0, 95%CI: 1.3-3.0) ( Table-4).

Discussion
This study determined the prevalence, determinants and effects of khat use among high school students of Sidama zone. The study tries to prevent anticipated social desirability bias by using a pre-tested self-administered questionnaire and giving adequate privacy to students while collecting the data using an anonymous questionnaire. The overall life time and current prevalence of khat use among high school students of Sidama zone were 14.6% and 13 % respectively. In contrast to other study reports from outside of Ethiopia, the magnitude of khat use for Sidama zone high school students seems lesser. For instance the overall khat use prevalence registered in Saudi Arabia was 24% 35 . Similarly reports from Eastern part of Ethiopia documented a 25 % for life time khat use prevalence 34 . Other school, university, and community based studies conducted in Ethiopia also reported a much higher figure than the reports we found among Sidama students 32,34,38 . For instance the reported overall khat use prevalence among Agaro high school students was 64%. This is nearly five times bigger (64%) than the Sidama high school students 38 . However, our finding was similar with one research conducted on Ataye high school students of Ethiopia. The documented life time and current khat use prevalence from this nearby khat producing district called Gurage Zone, were 15.36 % and 13.25% respectively 37 . Compared to Ethiopian university based khat survey reports, the prevalence for students of Sidama seems to fall within the documented khat use prevalence range but lesser than the community based reports. For instance university based prevalence studies conducted in Ethiopia reported khat use prevalence ranging from 13.4% to 41% for lifetime, 7% to 20.3% for past 12 months, and 6.3% to 33.1% current or past 30 days 32 . Much bigger prevalence reports noticed in Ethiopia are from one school survey report and other community based surveys 7,34 .
In contrast to population based research reports of Ethiopia, khat use prevalence findings from our study seem lesser. To compare this, one can see the documented community based reports of Ethiopia which range from 31.7%-55.7% 34,39 . The differences noticed in these research reports may better be explained with the following possible rationales. These may include, study time differences, methodological differences, availability or production of khat in the studied areas, the surrounding community socio-cultural and religious context, the time period in which khat got introduced to the local community, and the working legal and policy issues in the surrounding regions. One of the limitations of crosssectional surveys is their inability to tell the pattern or trend of the subject under study 40 . In accordance with this fact, our study never tried to assess the pattern and the rate in which Sidama students' khat use raised from zero to 14%. Even though further studies are required in this area, as khat is new to the area, it is possible to understand that the prevalence increased with in a short period. If this is the case, the doubling and tripling time of the existing prevalence may be very soon.
The discovered prevalence in the area is at an alarming level and at more than sufficient level to establish khat use as one of the public health problems of the area. As most of khat chewing students' self-report depicted, most students already claimed developing an addiction to khat. They also expressed their desire to stop chewing khat. Thus, this is time for action by all concerned parties. As khat is a psycho-active substance, the reported prevalence is high enough to cause all the adverse outcomes mentioned in the introductory part of this article. It is also well known that using psycho-active substances in early age leads to using various drugs and continued use of substances in latter life.
In this study, we found only three types of substances currently used by the students. These were khat (13.0%), alcohol (3.7%), and cigarettes (2.4%). This finding is consistent with other studies. Most of the studies conducted in Ethiopia reported students using these three psycho-active substances 32,34 . Nevertheless, looking merely at the size of the cigarette prevalence report is misleading. Because nearly all students in these districts (80%) are passive cigarette smokers as a result of uncontrolled exposure to second hand tobacco smoke. This report may be taken as an indicator for the communities' exposure to second hand smoke. Ethiopia has already developed a policy and legal framework to safeguard citizens from exposure to second hand smoke. Our finding is, therefore, a confirmation that the implementation of the policies and reinforcement of the law are not to the expected level.

Factors determining students' khat use
As the multivariate logistic regression analysis denoted, being a male, having a family member chewing khat, drinking alcohol, smoking cigarettes, ever practicing sex, having a family growing khat, and having friend who chews khat were all factors increasing the odds of students khat use. Most important of all these, smoking cigarettes increased the risk of khat chewing by about six folds as compared to the students who had never smoked (AOR=6.0, 95% CI: 2.3-14.3).
These study findings are in accordance with other study reports. But unique to this study is that we considered additional factors such as having a family which grows khat, and students involvement in khat marketing activities 32,34,38 . The result showed that both factors were significant at the binary logistic regression analysis level. Especially students' involvement in khat producing and marketing activities was significant with (COR=3.6, CI: 95% 2.6-5.1). However, only 'having a family growing khat' passed the multivariate analysis. In addition, we also assessed the effects of ever practicing sex on students 'khat use. The result denoted that students' odd of khat use is two times higher for those who ever practiced sex than those who didn't. Therefore considering those modifiable risk factors among the list of determinants shown above will help in designing khat use preventive programs.

Effect of khat use on student's academic performance
Most of studies conducted in Ethiopia and other countries reported that students are using khat for its stimulant effect. Among the self reported reasons given by students; to wake up or avoid sleep, to get energized for study, and to increase academic performance are included 32,34 . In contrary to this, those khat chewing students in our study had a poor chance of performing well in their academics or in an other way had a likelihood of failing in their academics. That means scoring below 50%. Student's poor performance as a result of khat use is also in accordance with other study findings. These studies stated that khat use has an association with academic performance 27,28,41 . There are also other reports that documented the fact that other psychoactive substance use also increases the odds of students' poor academic performance or failure in academics 42 . The paradox between the emerging evidence ascertaining khat's effect for poor academic performance and the frequent motivational beliefs held by students might happen as a result of lack of information or getting distorted information either from deeply embedded myths in the community or from formal or informal khat marketing promotional agencies.
Other studies which assessed the association between khat and academic performance tested to discover if there was any association and determined the level of significance (P value) and chi squared testing 28 . But this study goes further to determine to what extent khat use affect students 'academic performance using the multiple logistic regression analysis and found that khat using students had odds of failing in their academics or performing poorly by two folds more than the non-chewing students. But further studies like comparative cross-sectional studies, case control, or other longitudinal studies can be conducted to explain more about the issue and establish either cause and effect or temporal relationship. As limitation of all cross-sectional surveys, this study didn't intend to explain beyond determining the odds of students' poor academic performance due to khat use.
This study also showed that being female, living in an urban area, and having a father with educational status of elementary and below showed students' odds of failing in their academics by about two fold more than their counterparts. This may be due to female students' work burden at domestic level, the availability of leisure centers and psycho-active substances in the urban area, and also lack of parental mentorship and guidance from their uneducated or less educated fathers respectively. Our primary research objectives were determining the prevalence and factors affecting students' khat use. As one of the limitations of this study we didn't exhaustively embrace all possible factors affecting students' academic performance. Among these factors, group of factors related to communication, learning facilities, availability of proper guidance, and presence of family related stressors are determinants reported by other study 43 . Since we also used students' self-report using the self-administered questionnaire, some students may fail to remember, that is, recall bias and/or intentionally misreport their last semester's cumulative grade point. Similarly, the prevalence we found in this research may be more than the report we found due to social-desirability bias.

Conclusion
The prevalence of khat use of high school students living in high khat producing and marketing districts of Sidama Zone is at an alarming level. Having a family growing khat, having a family member chewing khat, having a friend who chews khat, smoking cigarettes, drinking alcohol ,ever practicing sex, and being male are all factors increasing the odds of students' khat use. Students' khat use increased the odds of students' poor academic performance or failing in their academics. Similarly being female, living in an urban area & having a father with an academic status primary education and below are factors increasing students' good academic performance.

Recommendations
Khat's contribution to students' poor academic performance demand prompt intervention. Formulation of appropriate intervention which reduces the use of khat should be everyone's priority. Moreover, designing an effective cessation strategy to help students with established khat addiction will remain the responsibility of the students, the family, the health authorities, education authorities, agricultural authorities and all other stakeholders.