Drug use and its associated factors among in-school adolescents in Harari region of eastern Ethiopia

ABSTRACT

This study aimed to determine the magnitude of current drug use and its associated factors among randomly selected 3227 in-school adolescents in eastern Ethiopia using a multistage sampling technique. Data were collected using questionnaires adapted from the Global School-Based Health Survey. The magnitude of current drug use among in-school adolescents was 5.67% (95% CI: 4.92–6.52). Rural residence (AOR = 2.25, 95% CI: 1.43; 3.54), social media use (AOR = 2.53, 95% CI: 1.75; 3.65), being bullied by others at school (AOR = 8.34, 95% CI: 5.67; 12.56), having a father who used drugs (AOR = 2.20, 95% CI: 1.38; 3.39), and having low self-esteem (AOR = 5.83, 95% CI: 2.77; 12.27) were significantly associated with increased odds of current drug use. The magnitude of current drug use is high. The findings highlight that interventions should focus on rural dwellers, social media users, bully victims, adolescents whose fathers use drugs, and adolescents with low self-esteem.

KEYWORDS:

• Current drug use
• associated factors
• adolescents
• Ethiopia

Introduction

Drug abuse is a global problem; in 2016, 5.6% of the population aged 15–64 years used drugs at least once globally (United Nations Office on Drugs Crime, 2018). The global burden of disease study showed increased burden due to drug abuse among adolescents and young adults; compared to older people, usage of drugs among younger people has increased (Degenhardt et al., 2016). Untreated drug use disorders increase morbidity and mortality risks for individuals (i.e., impairment in personal, family, social, educational, occupational, or other important areas of functioning) and lead to lost productivity, premature mortality, increased healthcare expenditure, costs related to criminal justice, social welfare, and other social consequences (World Health Organization, 2023).

Substance use disorders are more likely to kill younger people (Ritchie & Roser, 2019), and such users chose cannabis (Holm et al., 2014). Drug use begins during the adolescent period, that is, a critical age of drug initiation, and young people aged 18–25  are the maximum drug users (Bartel et al., 2015; United Nations Office on Drugs and Crime, 2018). Youths are vulnerable groups who engage in the use of substances such as alcohol, tobacco, or marijuana either experimentally or recreationally (Bartel et al., 2015; Jones et al., 2012). Adolescents’ poor physical health outcomes, depression, anxiety, serious injuries, poor academic performance, and risky behaviours are attributable to substance use (Aboagye et al., 2022; Aboagye, Seidu, Adu, et al., 2021; Aboagye, Seidu, Hagan, et al., 2021; Asante & Kugbey, 2019; Degenhardt et al., 2007; Groce, 2018; Macleod et al., 2004).

The burden attributable to substance use increases substantially in adolescents and young adults, according to a 2013 global burden of disease study, and 14% of the total health burden is due to alcohol and illicit substance use (Degenhardt et al., 2016). There were an increased number of drug use disorders’ cases from 1990 to 2017 globally (Pan et al., 2020). Young people aged 20–24 years have the highest alcohol misuse risk factors (7%) for disability-adjusted life years (DALYs), and drug use accounted for 2.7% (Mokdad et al., 2016). Substance use is a major public health issue among in-school adolescents in sub-Saharan Africa (SSA). The SSA study from 2012 to 2017 indicated prevalence rates of 11.3%, 2.6%, and 2% for current alcohol use, lifetime amphetamine use, and current marijuana use, respectively, among school-going adolescents (Kugbey, 2023). Vulnerability to drug use among adolescents may be due to their strong inclination towards curiosity, experimentation, and susceptibility to peer pressure, poor self-worth, and rebellion against authority (Degenhardt et al., 2016).

Critical thinking, learning crucial cognitive skills, and a successful transition to adulthood are impaired by abusive drugs (Crews et al., 2007), thereby leading to a higher rate of mental illness and reduced well-being. Some of the risk factors that predispose adolescents to drug use are the presence of early mental and behavioural health problems, poorly equipped schools, peer pressure, poverty, high impulsivity, emotional regulation impairment, maltreatment or negative upbringing, poor family structure, poor parental supervision and relationships, poor parental education, uncontrolled pocket money, the presence of substance-using family members, a lack of opportunities, isolation, gender, and accessibility to drugs (Nawi et al., 2021; Somani & Meghani, 2016). However, religiosity, peer factors, self-esteem, self-control, parental monitoring, academic competence, school connectedness, social phobia, strong neighbourhood attachment, and anti-drug use policies are protective factors (Drabble et al., 2016; Goliath & Pretorius, 2016; Guerrero et al., 2016; Nawi et al., 2021; Nguyen & Newhill, 2016; Schinke et al., 2016).

A few studies in Ethiopia have collectively identified substance use (Birhanu et al., 2014; Dida et al., 2014; Hirpa et al., 2021; Kassaye et al., 1999; Seid et al., 2021), with varying proportions ranging from 26.5% in Kolfe-Keraniyo preparatory school (Seid et al., 2021) to 47.9% in Woreta high school adolescents (Birhanu et al., 2014). Most of these studies have assessed the use of alcohol, khat, shisha, tobacco, and cannabis. Similarly, according to a pooled analysis in Ethiopia, about 41.5% of high school students use any substance currently (Roba et al., 2021). Hence, this research requests ‘what is the current drug use and associated factors among adolescents in school?’ particularly in Harari, Eastern Ethiopia, where there is scarce information on current drug use and its associated factors in this population.

The findings will shed light on the recent trends of associated factors and the magnitude of current drug use and provide insight into the main focus factors for the prevention and control activities of the programme. Therefore, this study aimed to determine the magnitude of current drug use and its associated factors among in-school adolescents in the Harari region of eastern Ethiopia. It fills gaps in previous studies and adds to the existing knowledge and benefits in Ethiopia, where the national response to drugs is to reduce harm, law enforcement responses, drug demand reduction, coordination mechanisms, and participation of civil society. Sustainable Development Goal Target 3.5 sets strengthening the government's commitments to prevent and treat substance abuse (World Health Organization, 2023).

Methods and materials

Study setting and design

This school-based cross-sectional study was conducted in the Harari regional state, located 526 km from Addis Ababa in eastern Ethiopia. Unlike most other regions in Ethiopia, the majority (54.2%) of people in the region live in urban areas (Central Statistical Authority, CSA, 2007). The region’s capital is the ancient ‘city of Harar’, a predominantly Muslim city where people specialize in trading. It is also believed to be the holiest city for Islam because of its rich collection of important Islamic monuments, notably 82 mosques and 102 shrines (Harari BoFED, 2010). Trade is the main source of revenue for people in the region. Psychoactive substances, such as khat (Catha edulis), tobacco, and coffee, constitute a substantial share of trading activities. Linked to this, there is a high prevalence of khat chewing in the community, where about a quarter of young people do it (Central Statistical Authority, CSA, 2007). Most of the rural population of the region depends on rainfall and small-scale farming. Khat is the dominant cash crop in most of the rural sub-districts of the Harari region (Ahmed, 2008; Assefa, 2018; Harari BoFED, 2010). During the study period, there were 112 schools in the region, of which 85, the target population for our study, had both primary (7^th through 8^th grades) and secondary (9^th through 12^th grades) level students (Harari Education Office, 2018). Twenty-three schools, urban and rural, both public and private, were included in this study. Data were collected from 24 November to 31 December 2020.

Population and sampling

The source population included all in-school adolescents in Harari Regional State, whereas adolescents in randomly selected schools during the study period constituted the study population. The sample size was calculated using OpenEpi Stat software with the assumption that the prevalence of adolescent drug use in Ethiopia is 50% (since there are no studies conducted on drug use in Ethiopia), the degree of precision is 5%, a 95% confidence level, a design effect of 2, and a non-response rate of 15%. The total sample size was 826. However, the data used in this study were obtained from a comprehensive survey of the mental health of 3227 in-school adolescents (Hunduma et al., 2022).

A multistage sampling technique using simple random sampling was used to select the schools and study participants. In the first stage, schools were stratified into urban and rural areas. During the study period, 35 schools were in rural areas and 50 were in urban areas. The schools in urban areas were stratified into public and private schools. There were 22 public schools in the urban area and 28 private schools. Seventeen public schools (10 rural and seven urban) and eight private schools were randomly selected for this study. In the second stage, students from randomly selected schools were stratified into grade levels (grades 7–12). From these grades, sections were proportionally selected based on the number of sections for each grade. Finally, all students in the selected sections were considered for this study.

Data collection

A structured and standardized self-administered validated questionnaire adapted from the Global School-Based Health Survey (GSHS) was used to collect the data (World Health Organization, 2019). The questionnaire included socio-demographic characteristics, psychosocial factors, behavioural factors, and adolescent current drug use status. Data were collected from adolescent students at their schools. An appropriate setting (room) was facilitated for the students if their sections were unwilling to complete the questionnaire. The participants were given an orientation regarding the study and how to fill out the questionnaire to maintain data quality. Two data collectors were assigned per session to facilitate and guide respondents appropriately.

Variables and measurements

Dependent variables

Current drug use was the outcome variable.

Independent variables

Socioeconomic status (i.e., age, sex, place of residence, adolescent’s marital status, father's educational status, mother's educational status, and parent marital status), psychosocial (i.e., self-esteem and bullying), and behavioural (i.e., social media use, father’s drug use, and mother’s drug use) factors were independent variables in this study.

Operational definition

Drug use: it was considered drug use in this study if cannabis, marijuana, shisha, and recreational drugs such as cocaine and heroin were used. This scale assesses adolescents’ drug use and frequency in the previous 12 months (World Health Organization, 2019).

Current drug use: it was defined as the use of drugs such as cannabis, marijuana, shisha, and recreational drugs such as cocaine and heroin daily, one to two times per week, or one to three times per week in the 30 days preceding the interview, and coded as ‘1’ if adolescents reported using any of these drugs, and otherwise, it is considered ‘0’ (World Health Organization, 2019).

Self-esteem: defined as the “judgment one makes about their self-concept” or “attitude one holds toward themselves as an object” that measured via assessing a subject’s attitude about themselves as a ‘thing’. The 10-item Rosenberg Self-Esteem Scale is used to measure global self-esteem, which consists of statements related to feelings of self-worth and self-acceptance (Gray-Little et al., 1997). This 10-item scale ranges from 0 (strongly agree) to 3 (strongly disagree). The sum scores for all 10 items range from 0–30, with higher scores indicating higher self-esteem. Respondents with a total score >25 were classified as having ‘high self-esteem’; scores between 15 and 25 were within the ‘normal self-esteem’ range; and scores below 15 suggested ‘low self-esteem’.

Father substance use: it was measured as the self-reported use of any substance (i.e., alcohol, tobacco products, khat, and others) by the respondent’s father one or more times per week in the past 30 days''; otherwise, it was considered ‘0’.

Mother substance use: it was measured as the self-reported use of any substance (i.e., alcohol, tobacco products, khat, and others) by the respondent’s mother that has been used one or more times per week in the past 30 days''; otherwise, it was considered ‘0’.

Social media use: Respondents were asked to report the number of hours they spend on social media or messaging sites or apps on the internet, such as Facebook, Twitter, and WhatsApp, on a normal weekday. Then, the status of ‘Social media use’ was dichotomized as ‘0 to less than 3 hours per day’ or ‘3 or more hours per day’.

Being bullied at school: To assess whether a respondent has been bullied at school, they were asked the question, ‘How often have you been bullied at school in the past couple of months?’. If the answer was more than once a week, then the respondent was categorized as ‘1’ for being bullied at school. Otherwise, the respondent was categorized as ‘0’.

Data quality control

For all data collectors and supervisors, training was undertaken for 5 days regarding collecting the data. All data collection tools were pretested and piloted for 5% of the sample size with similar school adolescent students, and appropriate modifications were made to make them consistent and clear before using them for actual data collection. The data collection process was closely supervised daily by trained supervisors and principal investigators. Data editors were assigned to check for missing data and inconsistencies for further cleaning before entry. Finally, the completed data were double-entered by different data entry clerks for validation and reduction of errors due to entry.

Data analysis

The data were double-entered, validated, and cleaned using EpiData 3.1, and analysed using STATA 14.1. A bivariable logistic regression analysis was performed to determine the association between each independent variable and the outcome variables. All variables in the bivariate logistic regression analysis with an a p-value of less than 0.20 (Heinze & Dunkler, 2017) were entered into the multivariable binary logistic regression model. Statistical significance was set at p < 0.05, and adjusted ORs (AORs) with a 95% confidence interval (CI) were calculated. The results are presented in tables, figures, and charts using frequency and summary statistics, such as mean and percentage, to describe the study participants concerning relative variables and compare them with previous study results.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki —ethical principles for medical research involving human subjects. Ethical clearance was obtained from the Institutional Health Research Ethics Review Committee (IHRERC) of the College of Health and Medical Sciences, Haramaya University (ref. No. IHRERC/149/2019). Written informed and signed voluntary consent from one of the parents or guardians and school principals and written voluntary assent were obtained from participants between 13 and 17 years of age. Participants aged 18 years and older provided their consent. To ensure participants’ confidentiality, personal identifiers were not included in the written questionnaires. All collected data was anonymized, stored on a personal computer, and protected with a password. Both participants and their parents were informed that the information gathered would be disseminated to assist in knowledge generation only.

Results

Sociodemographic characteristics

The present study was conducted among 3227 in-school adolescents, with a 97% response rate. The mean age of the respondents was 15.69 (±1.79 standard deviation) years. The majority of respondents, 2,706 (83.85%), were from urban areas, and 2,302 (71.34%) lived with their biological parents. More than half of the participants 1,670 (51.75%) were girls; 1,540 (50.82%) were from primary schools; and 1,749 were Muslims (54.2%). More than two-thirds of participants (67%) attended public schools. A total of 1,622 (50.27%) mothers and 1,488 (46.12%) fathers did not attend formal education, whereas 1307 (40.5%) fathers and 773 (23.96%) mothers did (Table 1).

Table 1. Socio-demographic characteristics of the respondents in the Harari regional state, eastern Ethiopia, 2020 (N = 3227).

Variables	Categories	Frequency	Per cent
Sex	Male	1557	48.30
	Female	1670	51.70
Age	≤13 years	472	14.63
	14–16 years	1,579	48.93
	17–19 years	1,176	36.44
Residence	Urban	2,706	83.85
	Rural	521	16.15
Marital status	Never married	2,948	91.35
	Married or living together	194	6.01
	Divorced/separated/widowed	85	2.63
Religions	Muslim	1,749	54.20
	Cristian	1,478	45.80
School type	Public	2,162	67.00
	Private	1,065	33.00
Father’s educational status	No education	1,488	46.11
	Primary	331	10.26
	Secondary	489	15.15
	More than secondary	919	28.48
Mother’s educational status	No education	1,622	50.26
	Primary	434	13.45
	Secondary	492	15.25
	More than secondary	679	21.04
Parental marital status	Living together	2,453	76.01
	Living Separated	240	7.44
	Divorced/Separated/widowed	534	16.55

Self-reported magnitude of adolescent current drug use

In this study, the magnitude of current drug use (cannabis, marijuana, shisha, cocaine, and heroin) among in-school adolescents was 5.67% (95% CI: 4.92–6.52). Of these, 119 (65.03%) were male. Approximately 61 (1.89%) participants were daily drug users, 37 (1.15%) used drugs once or twice weekly, and 33 (1.02%) used drugs two-three times weekly. Among current drug users, 92 (50.55%) reported financial hardships, 57 (31.15%) had health problems, 73 (39.89%) fought with their friends, and 62 (33.70%) were detained by law enforcement bodies in the past 12 months as a result of their heavy drug use habits (Figure 1).

Figure 1. Proportion of problems due to drug use among school adolescents in eastern Ethiopia, 2020 (N = 183).

Factors associated with current drug use

In bivariate analysis, sex, age, residence, religion, social media use status, father’s educational status, frequency of being bullied by others, levels of self-esteem, fathers’ drug use, and mothers’ drug use were found to be associated with current drug use. However, only residence, social media use status, school type, being bullied by others, levels of self-esteem, and father drug use were significantly associated with current drug use in multivariate analysis.

The odds of drug use were 2.25 times higher among rural residents compared to their urban counterparts (AOR = 2.25, 95% CI: 1.43; 3, 54). Drug use among students was significantly associated with social media use. Adolescents who spend three or more hours daily on social media were 2.53 times more likely to use drugs than those who did not (AOR = 2.53, 95% CI: 1.75; 3.65). Similarly, being a private school attendant increased the likelihood of drug use by 2.69 times (AOR = 2.69, 95% CI: 1.72; 4.11). Similarly, this study revealed that the odds of drug use were 8.34 times higher among those who were bullied by others at school twice or more per week than among those who were not bullied (AOR = 8.34, 95% CI: 5.67; 12.56). Likewise, respondents whose fathers use drugs were 2.20 times more likely to use drugs than respondents whose fathers do not have a drug use history (AOR = 2.20, 95% CI: 1.38; 3.39). This study revealed that adolescents’ self-esteem is associated with their drug use status. Accordingly, the odds of drug use were 5.83 times higher among those with low self-esteem than those with high self-esteem (AOR = 5.83, 95% CI: 2.77; 12.27) (Table 2).

Table 2. Factors associated with current drug use among in-school adolescents in the Harari regional state, eastern Ethiopia, 2020 (N = 3227).

Variables	Categories	Drug use		COR (95% CI)	AOR (95% CI)
		Yes	No
Sex	Male	119	1,438	2.07 (1.52; 2.84)	1.42 (0.99; 2.03)
	Female	64	1,606	1	1
Age	≤13 years	33	439	1	1
	14–16 years	75	1,504	0.66 (0.43; 1.01)	1.01 (0.60; 1.72)
	17–19 years	75	1,101	0.91 (0.59; 1.38)	1.34 (0.78; 2.28)
Religion	Muslim	141	1,608	1	1
	Christians	42	1,436	0.33 (0.23; 0.47)	0.65 (0.43; 0.99)
Residence	Urban	109	2,597	1	1
	Rural	74	447	3.94 (2.88; 5.38)	2.20 (1.43; 3.39)
Social media use status	Less than 3 hours per day	57	2,134	1	1
	3 hours and above per day	126	910	5.18 (3.75; 7.15)	2.53 (1.75; 3.65)
Father’s educational status	No education	103	1,519	1.89 (1.27; 2.82)	1.02 (0.55; 1.91)
	Primary	29	405	2.21 (1.31; 3.75)	1.80 (0.87; 3.72)
	Secondary	21	471	1.22 (0.71; 2.12)	1.01(0.50; 2.01)
	Above secondary	30	649	1	1
Mother’s educational status	No education	101	1,387	1.46 (0.96; 2.22)	1.01 (0.54; 1.85)
	Primary	26	305	1.54 (0.91; 2.62)	1.06 (0.53; 2.16)
	Secondary	22	467	0.96 (0.54; 1.70)	1.05 (0.53; 2.11)
	Above secondary	34	885	1	1
School type	Governmental	127	2,035	1	1
	Private	56	1,009	0.88 (0.64; 1.22)	2.69 (1.72; 4.21)
Being bullied at school more than once per week	No	47	2,570	1	1
	Yes	136	474	15.68 (11.10; 22.17)	8.34 (5.67; 12.26)
Levels of self-steams	Low	49	295	13.78 (6.89; 27.56)	5.83 (2.77; 12.27)
	Normal	124	1,919	5.36 (2.80; 10.26)	3.23 (1.64; 6.37)
	High	10	830	1	1
Father’s drug use	No	120	2,722	1	1
	Yes	63	322	4.43 (3.20; 6.14)	2.34 (1.47; 3.70)
Mother’s drug use	No	146	2,915	1	1
	Yes	37	129	5.72 (3.83; 8.55)	1.44 (0.81; 2.55)

Bold, significant association at p < 0.05; COR, crude odds ratio; AOR, adjusted odds ratio; CI, confidence interval; 1, reference.

Discussion

This study found that more than 1 in 20 in-school adolescents in the Harari region currently use drugs, which is common among adolescents who reside in rural settings, spend three or more hours on social media, attend private schools, have low self-esteem, have been bullied at least once per week, and whose fathers have also used the drugs.

In our study, the prevalence of current drug use was similar to that measured in the study by Dida et al. (2014) (5.6%) (Dida et al., 2014) in the Bale zone, Oromia Regional State, Southeast Ethiopia, and higher than that measured in the study by Birhanu et al. (2014) (4.1%) (Birhanu et al., 2014) in Woreta town, northwest Ethiopia. The widespread use of khat chewing due to its superior economic advantages compared to all other crops (Dessie & Kinlund, 2008) is associated with a higher prevalence of ever shisha smoking in Ethiopia (Hirpa et al., 2021). Further, the use of khat, marijuana, and cigarettes was significantly associated with the use of shisha (Hirpa et al., 2021); high school students who chew khat also use tobacco products, including shisha (Kassim et al., 2015; Reda et al., 2012). The probability that adolescents may be exposed to multiple substances is high when they are exposed to one substance. Therefore, awareness creation on the health consequences, reducing access and availability and reinforcing regulations of these drugs are very important to reduce current use among adolescents.

A higher prevalence of drug use was reported in Rwanda (26.1%) (Omotehinwa et al., 2018) where private university students participated, and Sudan (13.4%) (Othman et al., 2019) where high school students participated. The fact that these studies were conducted in capital cities, where shisha smoking practices might be higher, could explain their higher prevalence. The lifetime prevalence of substance use in Kolfe-Keraniyo, Addis Ababa, and Woreta towns, northwest Ethiopia, was 26.5% (Seid et al., 2021) and 65.4% (Birhanu et al., 2014), respectively. The high prevalence of current and lifetime drug use may be due to the inclusion of all types of drugs in the outcome variables (i.e. alcohol, smoking, etc.). These data suggest that the prevalence of drug use among in-school adolescents in Ethiopia is high.

Rural resident adolescents have increased odds of current drug use compared with their urban counterparts. The findings of this study are consistent with recent studies suggesting that students living on farms were exposed to greater numbers of risk factors across multiple domains than those living in towns (Rhew et al., 2011). This might be due to familial risk factors such as low parental education, negligence, poor supervision, uncontrolled pocket money among rural, and high paternal awareness of drug abuse among urban (Nawi et al., 2021). Further, the likelihood of substance use goes down with increasing education and wealth (Saikia & Debbarma, 2020), as there are significant inequalities between rural and urban in Ethiopia (Tigre, 2020). The finding suggests that prevention of adolescent drug use in rural settings requires outreach to rural-dwelling youths. However, studies are inconclusive in this regard and tailored interventions should target adolescents regardless of place of residence.

Adolescents who spend three or more hours per day on social media have increased odds of current drug use. Similarly, Boniel-Nissim et al. (2022) identified that School-aged Children who were problematic users of social media showed the highest level of substance use (Boniel-Nissim et al., 2022). This could be attributable to the use of social networking sites more than 2 hours independently associated with increased odds of poor self-rating mental health and experiences of high levels of psychological distress and suicidal ideation (Sampasa-Kanyinga & Lewis, 2015). Further, celebrities and others on social media glorify substance use rampantly, and it is used as a strategy for selling drugs with hashtags facilitating the process of pairing buyers with sellers (Yang & Luo, 2017). Social media platforms that are fully accessible to teens are being used by industries as marketing strategies (Barry et al., 2016). Teens are exposed to cannabis industry advertising through social media, and exposure to substance use imagery is associated with subsequent onset of use (Dal Cin et al., 2012). As a result, the need to balance technology use from an early age is growing; the online environment is meeting adolescent needs due to an ever-growing reliance on media for information, work, shopping, leisure, and communication (Throuvala et al., 2019). It is crucial to limit either social media substance use advocacy or adolescent use of social media.

Adolescents who attend private schools have increased odds of drug use compared with those who attend government schools. High school students in Addis Ababa and Butajira, who attended their education in private schools, had increased odds of cigarette and cannabis consumption compared to their counterparts attending education in government schools; relaxation and entertainment were the common reasons why private school students took drugs (Kassaye et al., 1999). A study indicated substance use was inversely associated with close contact with parents and parenteral supervision (Horta et al., 2014). These adolescents need parents’ and/or teachers’ supervision on their use of mobile phones, finance, and media; they need to be told ‘use of substance to be seen as better among their peers’ is wrong assumptions.

Furthermore, adolescents who were bullied more than once per week had increased odds of current drug use compared with those who were not bullied. Similarly, bullying predicted substance use among school-going adolescents in eight sub-Saharan African countries (Kugbey, 2023), a Survey in Denmark indicated that bully victims used medicine for pain and psychological problems more often than their non-bullied counterparts (Due et al., 2007). The study also indicated that bully victims were likely to use alcohol, marijuana, and lean/krokodil (Hong et al., 2022). Parents and teachers should closely supervise students. Routine assessment of bullying and its associated modifying factors is a potential avenue for improving students' well-being in the education system (Mutiso et al., 2019).

Low self-esteem predicted drug use among adolescents in the Harari region. Consistent with this study, negative self-image and lower goal-setting, problem-solving, and self-efficacy were associated with increased odds of past-month drug use (Schinke et al., 2016). There were increased odds of recent drinking among youths with low self-esteem (Bartsch et al., 2017). Peer and school, and body image self-esteem among male and female, respectively, were predictors of alcohol use in China (Wu et al., 2014). However, higher self-esteem among adolescents was associated with lower levels of risk behaviours (Hamme Peterson et al., 2010). Thus, youth substance use can be reduced by targeting self-esteem and fostering adult support systems (Pederson et al., 2022). In order to prevent drugs use, these adolescents should be identified, receive counselling to improve their self-esteem and those with low self-esteem should get support.

The odds of drug use increased among adolescents whose fathers used the drug compared with their counterparts. Similarly, a study in a similar setting indicated that the availability of someone with similar habits in the family predicted khat chewing (Reda et al., 2012), and the odds of substance use were high among preparatory school students whose family members consume alcohol and chew khat in Kolfe-Keraniyo, Addis Ababa (Seid et al., 2021). Birhanu et al. also reported increased odds of substance use among high school students whose siblings, family, and friends use substances (Birhanu et al., 2014). This study’s findings are also consistent with recent studies suggesting that parental smoking and parental permissiveness to drink alcohol were associated with the risk of cannabis and illicit drug use among secondary school adolescents in Nigeria (Mehanović et al., 2020). The odds of substance use were increased among adolescent boys whose family members also indulged in substance use in India (Srivastava et al., 2021). This highlights family members should stop their substance use or create awareness among their kids that they should not be seen as role model when substance use is concerned.

The strengths of our study include the use of validated questionnaires adapted from the Global School-Based Health Survey (GSHS) and the inclusion of both urban and rural in-school adolescents from public and private schools. However, the study has the following limitations: (1) the study was conducted only in one region of Ethiopia, which may not be generalizable to other regions; (2) we did not establish a cause–effect relationship because of the cross-sectional nature of the study design and did not include out-of-school adolescents; and (3) data collection during the COVID-19 pandemic, students’ tendency to report acceptable behaviours, and failure to remember some important responses.

Conclusion

The present study indicates the magnitude of current drugs use among in-school adolescents is high in the Harari regional state compared to other studies. This underscores the benefit of recognizing drug use as a public health issue in the study setting. This study highlights that factors such as place of residence, spending three or more hours on social media, attending private schools, being bullied by others, having low self-esteem, and having fathers who use drugs influence the current uptake of drugs among in-school adolescents in the Harari region. It may be necessary to develop anti-bullying programmes in schools to reduce bullying and its negative consequences. Furthermore, the findings of this study can be used as crucial input for the cessation of drug use in the study area. Since the present study was conducted among in-school adolescents, future studies should explore the prevalence of drug use among out-of-school youths and identify additional factors such as family support, supervision, peer factors, academic performance, and presence of drug retailers around school, contributing to current and lifetime drug use.

Author contributions

All authors made significant contributions to the work reported, whether in the conception, study design, execution, acquisition of data, analysis, and interpretation, or in all these areas. All took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; agreed on the journal to which the article has been submitted; and agreed to be accountable for all aspects of the work.

Acknowledgments

The authors thank the Haramaya University for funding this study. We would also like to thank the Harari Regional State Education Bureau, which involved school directors and teachers, participants, and data collectors, for meticulously facilitating and conducting this work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All the data used in this study are included in this article. Additional data are available from the corresponding authors upon request.

Additional information

Funding

This study was funded by the Haramaya University Scientific Research Grant number (HURG-2020-02-01-92) after defending the study proposal. The funder has no role in the design of the study; collection, analysis, or interpretation of data; or in writing the manuscript.