PURPOSE:The review surveys the type of machine learning approaches currently used in the alcohol literature, reviews challenges in applying machine learning tools to alcohol data, and explores how overcoming these challenges could advance personalized medicine for alcohol use disorder (AUD). SEARCH METHODS:The authors conducted a search of publications on PubMed, ScienceDirect, and EBSCO Academic Search Premier published from 2015 to April 15, 2025, for articles that used machine learning to analyze alcohol-related outcomes. Search terms were ("drinking" OR "alcohol") AND ("machine learning" OR "deep learning" OR "predict" OR "classify") in the title or abstract. SEARCH RESULTS:The search returned 2,618 manuscripts. Keeping those that predicted alcohol-related outcomes and excluding those that merely used alcohol as a predictor for other outcomes reduced the selection to 567 manuscripts. A final manual selection resulted in 110 original peer-reviewed human research studies that primarily analyzed alcohol consumption behaviors and tested their models on data that they were not trained on. DISCUSSION AND CONCLUSIONS:Predictions focused on alcohol consumption or AUD diagnosis in cohorts with a mean age of 50 years or younger (i.e., when long-term drinking behaviors are being or have been established). Most studies confined the data-driven searches to a single modality and relied on conventional machine learning approaches, which tended to produce accurate and transparent predictions on the relatively small datasets typically collected by AUD studies. The small number of available samples was the most common limitation mentioned by the reviewed articles. Investigators also wished for machine learning models to provide insights about causality. Gaining these insights will be essential to improve diagnosis and treatment of AUD, for which the field must foster multidisciplinary research teams to build rigorous and trustworthy machine learning models and quantitative benchmarks that can capture the multifaceted nature of alcohol use and its comorbidities.
BACKGROUND:Alcohol use and alcohol use disorder (AUD) among older women (defined as age 50 and over) have increased substantially in recent years. Compared to men and younger women, older women negotiate biopsychosocial transitions that put them at greater risk of alcohol-related health conditions as they age and are less reliably screened and treated for heavy alcohol use and AUD. OBJECTIVES:This scoping review represents a critical evaluation of research methodology and findings since 2004 on biopsychosocial correlates and consequences of alcohol misuse (i.e., heavy alcohol use, defined as four or more drinks on any day or eight or more drinks per week; and/or AUD) among older women. These findings, with a focus on their implications for treatment and prevention strategies among this at-risk population, are discussed. The goal of the review is to summarize current research as well as treatment and prevention options available for older women and to identify gaps in the literature and potential for future research. ELIGIBILITY CRITERIA:2,579 research articles were reviewed for inclusion in the paper. Inclusion criteria required that a study's findings were relevant to the review's objectives, with analyses and results that had clinically relevant findings for older women with heavy alcohol use and/or AUD. SOURCES OF EVIDENCE:Authors searched PubMed, Embase, and PsycInfo for articles published between January 1, 1994, and August 5, 2024. CHARTING METHODS:Two authors developed the template for data extraction, and four authors charted data. Weekly meetings were used to establish, monitor, and maintain calibration among authors. RESULTS:127 articles met inclusion/exclusion criteria and were included in the final extraction and results summary. Evidence-based information is presented on (1) clinical presentation of older women with alcohol misuse; (2) biological correlates of alcohol use in older age; (3) psychosocial correlates and consequences, such as mood disorders, social networks, and other substance use; (4) increased risk of health conditions, such as breast cancer and depression; and (5) current state of treatment and prevention needs and efforts. CONCLUSIONS:The research reviewed here indicates a need for further age- and sex-specific research in the field of alcohol use and AUD. After age 50, men and women continue to differ in important ways in relation to the correlates, consequences, and treatment of alcohol misuse. The current literature includes significant methodological limitations and inconsistencies. Research with samples of older women who drink at heavy levels may be particularly helpful in addressing these limitations. Existing prevention and treatment interventions may be good options for older women, but more research is needed.
PURPOSE:Resilience is crucial in mitigating the risk of stress-related health issues. Although many people can adapt to adverse stress or trauma, stress exposure can increase the risk of health issues, including obesity, cardiovascular disease, and digestive illnesses. Some individuals may even develop debilitating conditions, such as post-traumatic stress disorder (PTSD). People with PTSD often struggle to adapt, sometimes turning to alcohol to cope, which can lead to alcohol use disorder (AUD), characterized by excessive alcohol-seeking and dependence. Understanding the biological underpinnings of resilience, therefore, is a key to preventing both PTSD and AUD. Recent research has uncovered the neurobiological traits that protect against the development of stress-induced alcohol dependence. Studies have shown that proactive coping and a lack of stress-related symptoms are associated with resilience. Preclinical studies, especially in rodents, have provided deeper insights into how stress impacts alcohol-seeking behaviors. Research has shown that resilience involves adaptive changes at the molecular, cellular, neural circuit, and systems levels. This review aims to integrate this research to better understand what makes people vulnerable to stress and alcohol consumption, highlighting aspects frequently overlooked in clinical models. SEARCH METHODS:This review employed systematic search strategies to achieve a comprehensive and structured assessment of the existing body of literature using the academic databases PubMed, Google Scholar, and Web of Science. Targeted keywords included "stress," "PTSD," "trauma," "alcohol," "AUD," "resilience," "vulnerability," "susceptibility," "sex difference," and "animal models"; Boolean operators (AND, OR, NOT) were used to refine the results. Exclusion criteria included research published before 1990, research that was not peer reviewed, and publications in languages other than English. Additional studies were identified by reviewing the references cited in key articles as well as by identifying subsequent studies that referenced these pivotal works. The search was carried out from April to June 2025 focusing on, but not limited to, experiments involving rodent models. SEARCH RESULTS:The search yielded a total of 347 articles. After screening, 283 articles were removed, either because they met the predefined exclusion criteria or because they were duplicates. This process resulted in 64 articles, forming the core of this comprehensive review. DISCUSSION AND CONCLUSIONS:This review summarizes an overview of recent progress in studies of PTSD and AUD, primarily focusing on the effect of resilience on post-stress alcohol intake in animal models. The findings highlight several biomarkers that may help identify individuals at risk for PTSD, AUD, or their co-occurrence, acknowledging that no single identifier can predict post-trauma outcomes. The identification of these markers is an ongoing process, yet it will be crucial for early diagnosis and risk assessment moving forward.
PURPOSE:There is an urgent need for novel and innovative advancements in alcohol research and treatment. Artificial intelligence (AI) is one such advancement that demonstrates promise, though not without important ethical concerns. This Perspective article examines current applications, ethical considerations, and future implications of integrating AI in alcohol research and treatment. SEARCH METHODS AND RESULTS:A nonsystematic narrative review of peer-reviewed, English-language studies of contemporary applications of AI (specifically, large language models and machine learning) in alcohol research and treatment was conducted in December 2024 through January 2025, using searches of the PubMed, Google Scholar, Web of Science, and PsycInfo databases. Search terms included the following: "alcohol," "alcohol use," "alcohol use disorder," "drinking," "treatment," "screening," "artificial intelligence," "chatbot," "large language models," "research subject recruitment," "research subject enrollment," "electronic health records," "support vector machine," "neural network," "black-box," "random forest," and "machine learning." The article highlights studies that discussed the role of AI in bolstering recruitment and retention for human subjects research and in the application of advanced data analytic techniques for alcohol research. Further, the article examines studies that described current and prospective uses of AI in alcohol use disorder treatment, including assessment and identification of individuals who may benefit from treatment, and as an adjunct to evidence-based treatments for alcohol use disorder. Potential ethical concerns are highlighted alongside potential considerations for the safe and ethical use of AI across these domains. CONCLUSIONS:Based on the identified primary and secondary sources, this review highlights the promises of AI in advancing both alcohol research and treatment, and how its successful implementation is supported by careful and ongoing attention to potential risks. The reviewed studies provide evidence that robust data security protocols, comprehensive informed consent, and ongoing monitoring for potential biases and harms can allow for these tools to advance scientific understanding while improving access to effective and equitable care. However, the current literature and the potential solutions suggested in this Perspective have limitations due to the pace of AI advancements that continue to accelerate at an unprecedent rate. FUTURE DIRECTIONS:Future research can advance the field by explicitly testing and evaluating the utility of these tools in real-world contexts, their risks for bias and privacy violations, and their ethical integration into alcohol research and treatment.
