We will conduct a two-arm, parallel-group randomized controlled trial (RCT) comparing a 6-week mobile app (Boozebuster) with an active educational website control condition. Participants in the Boozebuster condition will receive access to the self-guided mobile app; participants in the control condition will receive access to an educational website containing information about the effects and consequences of alcohol use on health. The Scientific and Ethical Review Committee of the Faculty of Behavioral and Movement Sciences of the Vrije Universiteit Amsterdam has approved the study protocol, informational letter, and informed consent process. The primary and secondary outcomes in addition to the exact timepoints at which each measure will be applied is presented in Table 1. The study is registered with the Netherlands Trial Registry (NL8828).

Young adults will be eligible for the RCT if they meet the following criteria: (1) aged between 18 and 30 years, (2) willingness to develop healthy lifestyle behaviors and low-risk drinking habits, (3) provision of informed consent digitally, (4) proficiency in reading and writing in Dutch, (5) access to an Android or IOS device with connection to the internet, and (6) possession of an email address. We did not apply a minimum drinking severity inclusion criterion. Our reasoning for not screening drinking severity is based on the fact that the Dutch drinking guidelines recommend an alcohol consumption level of zero or no more than 1 alcoholic drink per day.

We will make use of an open recruitment strategy, specifically, by means of advertisements on social media and health-related websites and via posters on educational facilities. These advertisements contain a link to our study information website [28]. Interested respondents can apply by leaving their email address, and they will be consequently contacted via email by the research team. The participants will be invited to complete a baseline assessment (T0), 6-week postbaseline assessment (T1), and 3-month postbaseline assessment (T2) via the computerized Castor Electronic Data Capture (EDC) system [29]. All participants will be offered an incentive of €10 (US $12.2) for completing the T1 assessment and an additional €10 (US $12.2) incentive for completing the T2 assessment.

First, participants will be screened, asked to fill out the baseline assessment (T0), and then randomized into either the mobile intervention or the educational website group. The allocation sequence will be automatically generated via the computerized Castor EDC system by using a 1:1 allocation ratio with random block sizes of 4 and 6, which also ensures allocation sequence concealment. Participants will be informed about the outcome of the randomization via email by a member of our research team. Initial stratification will be into two groups according to gender (ie, male and female). Within these two groups, participants will be further stratified into two groups categorized by adherence or nonadherence to the Dutch drinking guidelines. Nonadherence will be defined as the consumption of more than 7 drinks per week or at least one binge-drinking session (≥5 drinks for male participants and ≥4 drinks for female participants) in the past 30 days. This will create four strata. Because of the nature of our study design, blinding of group allocation for participants is not possible. Figure 1 provides an overview of the trial flow.

Previous meta-analyses assessing drinking reduction in young adults compared to assessment only, attention-matched, or active controls [17,18] found small effect sizes of about g=0.25. Although we expect the Boozebuster intervention to outperform previous interventions in reducing the quantity and frequency of alcohol consumption in comparison to the control condition at the posttest assessments, we decided to base our sample size calculation on this conservative effect estimate. These estimations would result with α=.05 and 1–β=.80 in a sample size of 253 per study arm (ie, N=506).

Boozebuster is a mobile app aimed at developing low-risk drinking habits in young adults according to the Dutch drinking guidelines. It has been developed by Vrije Universiteit Amsterdam in collaboration with the Institute for Systems and Computer Engineering, Technology and Science [30]. Boozebuster is based on the Moodbuster platform, which is being used for a variety of research projects funded by the European Union [31,32].

Boozebuster focuses on the importance of the individuals’ motivation, self-efficacy, and environmental constraints with regard to drinking. The behavioral change techniques used include PNF [18], motivational interviewing [33], cognitive behavioral therapy [34], goal setting [35], self-monitoring [36], protective behavioral strategies [37], and mindfulness [38].

Boozebuster contains a total of seven modules and was developed with the following rationale in mind: the PNF module serves as the first assessment of young adults’ drinking patterns and compares it with peer norms and the official Dutch drinking guidelines [39]. This procedure creates awareness among the individuals about their alcohol consumption in relation to their peers and has shown to be an effective approach in initiating drinking-related changes [40]. The incorporated motivational interviewing module motivates the individuals to change their alcohol-related behaviors by assisting them in building confidence in their abilities for achieving their goals. In addition, Boozebuster includes a variety of lifestyle support modules, such as relaxation, improving sleep, drinking diary, gratitude diary, and an emergency button specifically designed to provide recommendations for acute cravings. These modules provide individuals with tools to deal with cravings, peer pressure, and stress that might otherwise interfere with their drinking-related goals.

Participants in the Boozebuster group will be able to monitor their daily alcohol consumption via ecological momentary assessments (EMA) [41] in terms of the amount of standard drinks consumed, their mood via daily mood ratings on a scale of 1 (low) to 10 (high), and their sleep quality via daily sleep quality ratings on a scale of 1 (low) to 10 (high). Participants will also be able to monitor their progress on the three abovementioned key behaviors (ie, drinking, mood, and sleep) via visual feedback (ie, visualization of their scores) through the mobile app.

Participants are free to use Boozebuster during the 6-week study duration as frequently as desired. The reason we opted for a 6-week period is because it is the recommended timeframe in which alcohol-related changes can occur and be measured [42]. Hence, participants will be stimulated and motivated to engage with our intervention via the daily reminders of the EMA function during the 6 weeks of study participation. Figure 2 provides an overview of the Boozebuster mobile app and its modules.

Participants in the control condition will receive access to an educational website containing information on the effects and consequences of alcohol use on health. Only participants allocated to the control condition will have access to this website.

Given that applied outcome measures in the alcohol intervention literature vary considerably we decided to follow the advice and included key outcomes based on the recommendation of the Outcome Reporting in Brief Intervention Trials on Alcohol (ORBITAL) framework [43].

After the data collection period, we will clean our data and assess it for accuracy and completeness. Missing data will be handled using either multiple regression imputation techniques or full information maximum likelihood estimation, depending on the amount of missing data and whether it is missing at random or not. All analyses for the primary and secondary outcomes will use linear mixed modeling regression analyses with the three time points nested within participants. For continuous outcomes, linear models will be used for normally distributed data and negative binomial models, for left-skewed data. For dichotomous outcomes, such as whether or not participants had a binge-drinking episode in the past 30 days, binary logistic regression will be used. All primary analyses will be conducted in line with the intention-to-treat (ITT) principle (ie, including all participants that are randomized). Sensitivity analyses will be conducted in line with the per-protocol principle (ie, intervention completers) and with study completers only. All statistical tests use a significance level α=.05. Time will be used as predictor and baseline measures will be included as covariates if appropriate. All analyses will be conducted using the R Statistical Package. Cohen d values will be used as the measure of effect size. Effect sizes of 0.8 are assumed to be large, effect sizes of 0.5 are assumed to be moderate, and effect sizes of 0.2 are assumed to be small [69]. In addition, several intervention effect moderators that might moderate drinking levels based on findings of previous studies will be investigated. In addition, potential moderators such as gender [70], readiness to change [71], engagement [72], and impulsivity [73] will be explored.

The majority of digital interventions aimed at reducing drinking levels in young adults were conducted in university settings in the United States or Australia, were delivered via a computer or web browser, and consisted of single-session interventions showing small effect sizes ranging from g=0.18 to g=0.29 [5,14-18]. Mobile apps with the purpose of managing alcohol consumption in young adults seem promising, but the evidence is still inconclusive given the mixed results [13].

To potentially enhance the effectiveness of such interventions, we decided to develop a self-guided mobile app supporting low-risk drinking habits among young adults from the general population, while incorporating relevant healthy lifestyle-related support components that may increase and prolong the effectiveness of our intervention but retain the acceptability and reach of single-session interventions [22-24].

Therefore, the present study aims at improving the existing knowledge on the effectiveness of mobile apps on alcohol reduction in young adults by employing lifestyle-related support components (ie, sleep, mood, and perceived stress). If our intervention proves to be effective, it could become an inexpensive and scalable public health intervention to improve drinking habits in young adults.

Our study has a number of strengths such as the delivery of our intervention via the use of a mobile app, which seems to be a feasible and desirable delivery method for our target group [74,75]. Second, this study used an open recruitment strategy, which will enhance the generalizability of our results compared to studies that recruit only in university settings. Given the large sample size we aim to recruit, our study will provide a reliable estimate regarding the effectiveness of our mobile intervention.

Nevertheless, there are some limitations that are worthwhile mentioning. We are measuring all outcomes via self-report measures, and we have decided against using lengthy and costly diagnostic interviews in order to reach out to a large sample group and due to the preventive nature of our intervention. However, based on the literature, we know that self-report measures are reliable for actual drinking levels in our target population [76]; furthermore, we expect that the EMA function that we included for self-monitoring purposes will augment and improve our assessments by adding a dynamic dimension to it [41]. We expect a potential large dropout rate at posttest and 3-month follow-up assessments as commonly observed in web-based interventions with an open recruitment strategy [77]. To minimize drop-out, automated reminders for completing questionnaires will be sent via email to our participants and participants will be offered financial incentives for completing the follow-up assessments.

Providing young adults with a mobile app that promotes low-risk drinking habits while also providing lifestyle-related support components might be a promising approach to increase the reach, uptake, and effectiveness of behavioral change interventions targeting alcohol reduction in this population. We expect that the use of our mobile app will reduce drinking outcomes significantly compared to the use of the educational website. In addition, findings of the trial will be informative for future researchers developing interventions for young adults with regard to the reduction of alcohol consumption and improvement of associated lifestyle behaviors.