Background

Alcohol use is a leading risk factor for premature mortality and life-long disability. Alcohol use disorders (AUD) have been linked to several risk gene clusters and neural circuits; changes in psychological constructs such as reward prediction, reward learning and cognitive control have been used to assess altered behaviors. Socioenvironmental factors such as isolation, intergenerational poverty and unemployment are thought to increase harmful alcohol use and AUD risk. 

However, current AUD assessments rely on self-reports of consumption and harm. Further, any AUD assessment will face difficulties with transcultural adaptation and translinguistic translation in global health settings. The growing AUD burden in low- and middle-income countries calls for the development and translation of tools allowing for systematic, comprehensive, coherent study of determinants of alcohol use. 

Project Description

This project builds on the work of previous teams that created the Alcohol Use Behavioral Phenotyping Test (AUBPT), a virtual tool that uses games and tasks to assess the user’s risk of AUD, and analyzed the test’s psychometric properties. This team will further develop and validate AUBPT adaptations in Hindi and other relevant regional Indian languages that will be deployed as an open-source free app on handheld devices in India. 

Team members will engage with AUBPT through app development, computational modeling, transcultural adaptation, data collection and statistical data analysis.

The team will continue to test and improve the app, and develop generative models specific to tasks on the AUBPT to better understand the causal relationships across behavioral paradigms and predict patient performance on tasks.

AUBPT’s adaptation into Swahili is currently underway. Next, the team will collaborate with the Association for Socially Applicable Research to translate it into Hindi and regional Indian languages and assess the cultural appropriateness of its content.

In order to continue testing the reliability and validity of AUBPT, the team will follow a multiphase data collection protocol including clinical (i.e., AUD diagnosis) and general population samples in the United States and India, later adding a small pilot of healthy samples from India and Brazil.

The team will create multivariate models, including interpretable supervised machine learning models, to better understand alcohol use phenotypes based on AUBPT parameters. Additionally, team members will conduct meta-analytic modeling of existing studies to understand the differences between patients with depression, anxiety and substance use disorders and healthy controls.

Anticipated Outputs

Open-source multilingual computer application for AUBPT; open-source validated artificial intelligence models; four research papers; data for NIH application; secondary data analyses of datasets from NIMH’s Data Archive; student presentations in international neuroscience and global health conferences

Student Opportunities

Ideally, this project team will include 2 graduate students and 8 undergraduate students from diverse backgrounds and academic and cultural experiences. All team members should have a strong interest in global mental health, an aptitude for quantitative and qualitative research methods and enthusiasm for working with global collaborators. Graduate students should have backgrounds in psychology, neuroscience, global health, public policy, interdisciplinary data science, computer science, cognitive science or biostatistics. 

Students can join any of the four subteams assigned based on skills and interests: Implementation, Content, App Development, and Computational Modeling and Analysis. 

Undergraduates with a major or minor in global health will be preferred for selection to the Implementation subteam. For the App Development subteam, students who have prior web development experience in open-source languages (Python, Java, C++, etc.) will be preferred. For the Computational Modeling and Analysis subteam, those with an interest or background in biostatistics, theoretical neuroscience and machine learning will be preferred. For the Content subteam, students with a major or minor in biology, neuroscience, cognitive science or psychology will be preferred. 

By participating in this project, students will gain hands-on experience in interdisciplinary research with human participants. Core competencies they will develop include study design and planning, working with international collaborators, scientific literature synthesis, quantitative and qualitative data collection, statistical techniques, contributing to research papers and presenting at conferences. Undergraduate students will gain first-hand experience working in translational research and an introduction to global health. Graduate students will lead the subteams, draft research papers as first/second authors and work independently on advanced data analyses. 

Paige O’Leary will serve as project manager.

In an optional summer component, 4 students will work 18-20 hours per week on data analysis and research manuscript drafting from June 10 to August 15, 2023. Some team members may have the opportunity to travel to Tanzania, Brazil or India during the summer as well.

Timing

Summer 2023 – Spring 2024

• Summer 2023 (optional): Validate content for Hindi adaptation of AUBPT; test application on various devices and operating systems
• Fall 2023: Collect data in U.S. test-retest sample; conduct simulation experiments for individual models
• Spring 2024: Translate AUBPT contents into Hindi and other selected Indian languages; analyze data from India; draft manuscripts for analyzed data; incorporate model-informed adaptations

Crediting

Academic credit available for fall and spring semesters; summer funding available

This Team in the News

Meet the Members of the 2023-2024 Student Advisory Council

See earlier related team, Gamifying Risk Identification for Alcohol Use Behaviors Across Countries and Cultures (2022-2023).