Prevalence and Factors Associated with Illicit Drug and High-Risk Alcohol Use among Adolescents Living in Urban Slums of Kampala, Uganda

Background; Illicit drug and high risk alcohol use among adolescents leads to poor health outcomes. We enrolled adolescents from urban slums in Kampala, Uganda, to assess baseline prevalence, and factors associated with illicit drug and high-risk alcohol consumption. Methods; We conducted a cross-sectional study using data collected in a cohort that enrolled 14–19-year-old male and female participants from 25-March-2019 to 30-March 2020. Data was collected on social-demographics, sexual behavior and reproductive health using interviewer administered questionnaires. The main outcomes were illicit drug use and high-risk alcohol use. Data on alcohol use was collected using the Alcohol Use Disorder Identification Test (AUDIT); results were dichotomized. Factors associated with each outcome were analyzed using multivariable logistic regression. Results; We enrolled 490 participants (60.6% female) with median age 18 (IQR 17–18) years, 91.0% had less than secondary education, 48.4% had their sexual debut before 15years, 47.1% reported paid sex in the past 3 months and 24.7% had a sexually transmitted infection (chlamydia, gonorrhea and/ or active syphilis) at enrolment. The prevalence of illicit drug use was 34.9% while 16.1% were screened as high-risk alcohol users. Illicit drug use was associated with being male (aOR 9.62; 95% CI 5.74–16.11), being married (aOR 2.24; 95%CI 1.07–4.68) and having ≥10 paying sexual partners in the past 3 months (aOR 3.13; 95%CI 1.40–6.98). High risk alcohol use was associated with reporting sex work as the main job (aOR 3.19; 95%CI 1.02–9.94) and having experienced physical (aOR 1.96 95%CI 1.01–3.81) or emotional violence (aOR 2.08; 95%CI 1.14–3.82) from sexual partners. Conclusion: Illicit drug and high-risk alcohol use are prevalent among adolescents involved in high risk sexual behavior and living in urban slums of Kampala. Comprehensive interventions that target substance use among this group of young people are needed and should include measures against intimate partner violence.


Introduction
proportionately more but in absolute terms the mortality rate was greatest in low and 60 middle income countries (LMICs) where data availability was more limited due to 61 underestimation of illicit drug use (2). DALY rates of illicit drug use are estimated to be 62 2.5 times higher in Sub-Saharan Africa (SSA) than other regions (3). Alcohol and illicit 63 drug use are a major threat to education, health and the economy (1, 2, 4-6), contributing 64 to 5.1% morbidity and 5.3% mortality worldwide. In 2015, 450,000 people died as a result 65 of illicit drug use; 37.3% of deaths were a direct result of drug use disorders (4, 7). 66 UNODC report that according to the demographic surveys it is projected that by 2030, the 67 number of people using illicit drugs will have risen by 11% globally and 40% in SSA due 68 to its rapidly growing and young population (1). The predictors of illicit drug use among . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint adolescent's world over include male gender, peer pressure, low education, psychological 70 factors and increasing poverty levels (8-10). Illicit drug and alcohol use contribute to 71 crime, disease, physical and mental incapacitation (1, 5, 11) in both adolescents and 72 adults, however the impact on adolescents with developing minds and bodies may be 73 more profound. illicit drugs may peak in late adolescence (18-19 years) and among young people aged 81 20-24 years (7). Illicit drug and alcohol use increase with age (11, 13) and health 82 behaviors adopted during adolescence such as use of substances have implications that 83 can persist throughout the life course (14). The WHO 2018 global status report on alcohol 84 and health estimates current alcohol use among 15-17 year-olds in SSA at 21.4% (4). 85 Furthermore, a systematic review of studies done in SSA found that the overall 86 prevalence of any substance use among adolescents was 41.6% (15). Reports show that 87 the risk of illicit drug and alcohol use among adolescents in SSA is increased by key 88 drivers such as: transactional sex, availability of disposable income, poverty, gender 89 inequalities and poor work/ living environments (16,17). Negative effects of illicit drug 90 and alcohol use on adolescents are associated with increased risky behaviors that 91 increase the likelihood of HIV infection, and addictive disorders that commonly result in . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. in urban slums where use of these substances may be more prevalent.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023.  119 We conducted a cross sectional study using data collected in a cohort that recruited  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint project. Adolescents were initially identified by project field workers with support from 137 community mobilisers and later invited to the GHWP clinic for screening and enrolment.

138
Participants enrolled in the study also mobilized peers using a snowball approach.   is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint Simple advice on reduction of low risk and hazardous drinking (1-7 and 8-15 AUDIT 160 scores, respectively) was given, brief counseling and continued monitoring was provided 161 for high risk/harmful alcohol use (16)(17)(18)(19) AUDIT scores), diagnostic evaluation for alcohol 162 dependence (≥20 AUDIT scores) and referral was recommended as in the WHO AUDIT  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint Factors associated with illicit drug use in the past 3 months and high-risk alcohol use at 204 baseline were determined for each outcome and each independent variable using logistic 205 regression. Bivariate/unadjusted logistic regression analysis was conducted and 206 variables with p-value < 0.15 were considered for the adjusted, multivariable analysis.

207
Multivariable logistic regression was used to identify the factors that are independently 208 associated with illicit drugs and high-risk alcohol use. Each outcome was considered 209 separately, with its own model. In the multivariable analysis, variables were kept in the 210 model if removing them significantly affected model fit. Variables that had a P-value <0.05 211 were considered significantly associated with illicit drug and high-risk alcohol use. In both 212 models, we considered age at enrolment and sex as priori confounders that were included 213 in the multivariable models regardless of its unadjusted p-values.   222 We included 490 participants in our analysis; 297 (60.6%) females and 193 males. The 223 median age of participants was 18 (IQR 17-18) years, 50.8% were ≥18years, 91.0% had 224 less than secondary education and 97.3% were out of school at the time of enrollment.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023.

Baseline characteristics associated with illicit drug.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023.  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023.  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023.

Discussion
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint Our study found a high prevalence of illicit drug and high-risk alcohol use among 14-19-284 year old adolescents, and this is similar to other studies done among at-risk adolescents 285 in similar vulnerable situations (15,20). This is perhaps not surprising, given that the 286 source GHWP clinic provides services to participants from Kampala slums that are 287 characterized by a high prevalence of crime, illicit drug and alcohol use (25); this 288 environment exposed adolescents to use these substances. However, our findings differ 289 from a study done by Berhane et al in nine communities in SSA that reported very low 290 prevalence of alcohol and illicit drug use among both school and non-school going 291 adolescents 10-19-years (14), which the authors suggest may be attributed to under 292 reporting. We found that adolescents who most abused drugs were taking care of 293 themselves with no or under employment and this is in line with the recent warning from 294 UNODC that increasing drug use rates are becoming more pronounced in countries with 295 low levels of income (1).

296
Illicit drug use in our study was associated with being male and is consistent with similar 297 studies done in Uganda, SSA and other parts of the world (10, 26, 27). Male participants 298 in our cohort were married young, some had more than10 paying sexual partners in the 299 last three months before enrolment and some experienced IPV from their sexual partners, 300 they therefore likely resorted to use of drugs to mainly forget their problems caused by 301 adolescent marriages, experiences of IPV from sexual partners (28, 29) and also to feel 302 good. Indeed, we observed in this study that more than half of male participants who used 303 drugs reported using drugs as a coping mechanism to mainly forget their problems and 304 to feel good (21, 30). We further found that marijuana and khat were the most used drugs 305 by adolescents (Figure1) which is similar to studies done in Uganda and SSA (10, 11, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint 21). These two drugs are locally available and cheap to purchase which makes it more 307 accessible and affordable to adolescents with low levels of income compared to other 308 drugs (10). 309 We found that illicit drug use was associated to being married which differs from another 310 study done in Uganda in fishing communities that portrayed association of illicit drug use Our findings also indicated that participants who reported engaging in sex work as their 318 main job were 3-times more likely to be screened as high risk alcohol users compared to 319 those who reported other main jobs, and having 10 or more paying partners was also 320 associated with high risk alcohol use (8). Almost three quarters of our enrolled females 321 reported engaging in paid sex in the past 3 months and most had low or no education. 322 Sex work has been associated with alcohol use because women consume alcoholic 323 drinks that clients buy for them (33), sex work sometimes occurs in bars and similar 324 venues that sell alcohol (16,33) and women report that taking alcohol gives them 325 confidence to deal with many clients (20). Some also started sex work as their first job so 326 took alcohol to cope with the stress of their job and other stressful issues linked to sex 327 work e.g. discrimination, criminalization and also to cope with the cold (33, 34). The length 328 of time involved in sex work also contributes to long-term alcohol use which can progress . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint to high risk alcohol use over time (20, 33). Others likely abused alcohol due to peer 330 influence around them given similar age group and involvement in similar high risk 331 behavior such as sex work. Previous research has found that adolescents who perceived 332 their friends as drug users, socialized with drug using peers and used alcohol for coping 333 or fun were at increased risk of illicit drug, alcohol use and misuse (35). A much broader 334 understanding of adolescent peers and their behavior may be one of the factors in curbing 335 illicit and high-risk alcohol use among adolescents living in urban slums. 336 We further found that adolescents who experienced physical and emotional violence from 337 sexual partners were more likely to be high risk alcohol users (20, 36). Most of our study 338 participants engaged in paid sex, and it has been shown that when compared to older 339 women, young women who engage in paid sex or sex work are more vulnerable to 340 intimate partner violence from clients (37, 38). Younger women may also struggle with 341 negotiation skills e.g., condom use, money paid per sex act, and this disadvantage may

Limitations
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint Our study used a cross-sectional study design which limits our ability to infer causality in 352 the associations seen between the dependent and independent outcomes. Our study was 353 done among a unique population of adolescents that also included emancipated minors 354 and findings may not be generalizable to adolescents in different settings and those who 355 need parental/ guardian consent.  Comprehensive interventions that reduce illicit drug and high-risk alcohol use are needed 361 among adolescents living in urban slums to prevent negative long-term health 362 consequences that may persist into adulthood.

364
The authors have no conflict of interest to declare.

365
Availability of data and materials 366 The datasets used and analyzed during the current study are available from the 367 corresponding author. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.20.23292973 doi: medRxiv preprint