The study aimed to estimate the prevalence of alcohol use and associated factors using the Health Belief Model. The findings from the survey revealed significant levels of alcohol consumption and exposure to alcohol advertising and promotion among the ABYM. More than 1 in 3 had ever consumed alcohol, 1 in 10 had consumed alcohol 30 days preceding the interview, almost half (1244, 49.8%) had seen “A lot” of alcohol advertisements on TV in the 30 days preceding the survey and more than 1 in 10 owned an item with an alcohol brand logo on it. These findings are consistent with those in a survey among 457 male and female out-of-school youth in slums of Kampala City in which more than 1 in 3 reported problem drinking and drunkenness and 62.1% had been exposed to alcohol advertisement and promotion . ABYM who possessed items with an alcohol brand logo on it had significantly higher odds of consuming alcohol. Alcohol advertisement, promotion and sponsorship have a direct effect on early initiation of use, frequency and quantity of alcohol consumption and may explain the heavy burden of alcohol use among adults in Uganda . Indeed, a cohort study on the impact of alcohol marketing on youth drinking behaviour demonstrated that alcohol marketing at baseline was predictive of both the initiation of alcohol consumption and frequency of drinking . There is a need for the establishment of a legal framework that will protect the youth from the influence of the alcohol industry through strategies such as prohibition of sale to minors, prohibition of alcohol in education institutions, limiting density of outlets, limiting alcohol promotion and sponsorship among others. An assessment by WHO established that banning alcohol advertising to reduce alcohol consumption is cost effective and can have incremental benefits when coupled with taxation .
Adolescent boys and young men who were out of school had higher odds of consuming alcohol compared with their in-school counterparts. Swahn and colleagues have reported similar findings of high alcohol consumption among youth who are out of school in Kampala City [31, 34]. A similar assessment among 132,600 youth noted higher levels of alcohol and illicit drug use among those that dropped out of school compared with those who were in school . Although the design of the current study precludes conclusions on the direction of causality between schooling status and alcohol use, a previous study showed that youth who are out of school are at a significantly higher risk of adverse health outcomes such alcohol use because of the influence of the disadvantaged living environment  that facilitates risky behaviour. Policy makers, youth leaders, parents, teachers, school administrators need to be informed about the problem of alcohol use among ABYM in general and those who drop out of school in particular to develop innovative ways of creating alcohol free neighbourhoods while keeping boys in school.
The study has also shown that compared with ABYM whose parents/ guardians and siblings did not drink alcohol, ABYM who had parents/ guardians or siblings that consume alcohol had significantly higher odds of consuming alcohol. According to the Health Belief Model, engagement in behaviour is predicted by risk perception, perceived benefit of engagement in the behaviour, self-efficacy, perceived barriers and cues to action . Cues to action refer to the internal and external enablers that facilitate engagement in certain behaviour. In this case, alcohol use by siblings and parents/ guardians is a cue that inadvertently certifies and inculcates positive attitudes towards the behaviour to the ABYM at an age at which they are vulnerable to being influenced by what they experience in their environment. Familial alcohol problems have been previously documented to be significant predictors of alcohol use among youth . Strategies for control and prevention of alcohol use among ABYM should adopt a whole of society approach by recognizing the role of the family, school and the community at large in the prevention of alcohol use in this age group.
The Health Belief Model also posits that self-efficacy or an individual’s confidence in their ability to engage in a certain behaviour is an important predictor for engagement in that behaviour . Compared with ABYM who were certain that they could limit their alcohol consumption, those who were less certain had higher odds of consuming alcohol. Deficiencies in confidence to limit alcohol use could also be explained by the effects of the marketing strategies by the alcohol industry where almost half of our respondents (1244, 49.8%) had seen “A lot” of alcohol advertisements on TV in the 30 days preceding the survey and more than 1 in 10 owned an item with an alcohol brand logo on it. A prospective study conducted among men and women on treatment for alcohol dependence found that higher self-efficacy scores were associated with better outcomes such as lower likelihood of consuming alcohol . There is a need to enact legislation that protects young people from exposure to deceptive marketing strategies that glorify alcohol use. This also calls for the availability of alcohol dependence treatment services that incorporate aspects of building confidence among the youth such as addressing mental health challenges like depression, and enhancing social support .
The major limitation of the current study was the attempt to make causal inferences in a design where both predictor and outcome variables were assessed simultaneously. The other limitation was that data about the dependent variable on alcohol use were obtained using self-reports which introduces bias. However, we collected data from a significant sample size and used validated tools to obtain information. Both these factors lend credence to the interpretation of the findings and conclusions reached.