Summary:
A year-long longitudinal study using gamified smartphone tasks shows that subtle increases in risky decision-making precede rises in alcohol consumption among people with mild to moderate alcohol use disorder (AUD). By measuring month-to-month cognitive and behavioral changes, researchers found that shifts in risk tolerance and information-gathering behavior predicted next-month drinking increases, suggesting a pathway for real-time digital monitoring and intervention.
Key Facts:
- Temporal Direction Established: Higher risk-taking in mixed gambling scenarios and reduced information sampling reliably predicted greater alcohol intake in the following month, indicating that changes in decision-making come before increased drinking.
- Non-Reciprocal Pattern: Greater alcohol consumption did not predict later increases in risky decision-making, ruling out reverse causality after adjusting for baseline drinking.
- Specific Cognitive Biomarker: Broad measures of self-control—such as working memory and motor response inhibition—did not track month-to-month drinking changes. Instead, specific shifts in risk sensitivity and evidence-gathering were the strongest predictors.
Source: Elsevier / University Hospital Würzburg
Alcohol use disorder (AUD) is a major global contributor to disability, chronic disease, and preventable death. While many cross-sectional studies have linked AUD to impaired cognitive control and elevated risk-taking, the causal sequence has been unclear: do cognitive changes drive hazardous drinking, or does alcohol exposure erode decision-making?
A new one-year longitudinal investigation published in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging clarifies that, in people with mild to moderate AUD, changes in specific decision-making processes tend to precede increases in alcohol consumption rather than follow them.
Tracking Real-Life Fluctuations with Gamified Mobile Tools
Traditional clinic assessments and retrospective questionnaires struggle to capture the dynamic, day-to-day variations in cognitive control and decision-making that accompany AUD. To address this, researchers designed monthly gamified tasks delivered via smartphone to measure real-world fluctuations over a year.
The study enrolled 603 people with mostly mild-to-moderate AUD. Participants completed monthly mobile tasks assessing working memory, response inhibition, risk-taking across gain and loss contexts, and information sampling (how much evidence they collected before deciding). They also reported alcohol consumption every two days, enabling a dense, longitudinal data set linking cognitive markers to drinking behavior.
Key predictive patterns emerged from the month-to-month data:
- Mixed-context risk-taking: When participants showed higher risk-taking in mixed gambling tasks—situations with both potential gains and losses—they tended to increase their alcohol consumption the next month.
- Loss-context sensitivity: Lower risk-taking in situations focused on avoiding losses also preceded later drinking increases, indicating context-specific shifts in risk evaluation.
- Reduced information sampling: Participants who collected less information before deciding—making quicker, less evidence-based choices—were more likely to consume more alcohol in the following month.
Co-lead investigator Hilmar Zech, PhD, summarized the findings by noting that monthly changes in specific decision-making processes forecasted subsequent drinking increases: higher mixed-context risk-taking, lower loss-context risk-taking, and decreased information sampling each preceded next-month drinking surges.
Specificity and Unidirectional Effects
Crucially, the observed relationships were unidirectional. Prior alcohol intake did not predict future deterioration in decision-making or cognitive control on monthly tests, eliminating reverse causality in this sample. Broad executive measures—working memory and motor inhibition—remained stable and did not track with month-to-month drinking changes, reinforcing that relapse vulnerability appears linked to specific computations about risk and evidence rather than generalized executive failure.
Lorenz Deserno, MD, PhD, emphasized that the temporal effect remained after accounting for previous drinking levels and that fluctuations in self-control or reward-driven behavior were not associated with drinking swings.
Implications for Real-Time, Just-In-Time Interventions
By validating smartphone-based behavioral tasks as objective, prospective predictors of real-life drinking increases, the study supports the promise of digital phenotyping in addiction care. Gamified mobile tasks can capture subtle cognitive signals that herald elevated risk, potentially enabling mechanism-based, real-time interventions.
Editor-in-Chief Cameron S. Carter, MD, noted that experimental tasks delivered by smartphone provide a more accurate window into decision-making than interviews or self-reports. With further validation, passive monitoring or micro-gamified assessments could power just-in-time adaptive interventions—alerting users or clinicians when cognitive risk markers suggest an impending binge before the first drink.
Editorial Notes:
- Article edited by a Neuroscience News editor.
- Journal paper reviewed in full by staff.
- Additional contextual information added by the editorial team.
About this AUD and Neuroscience Research:
- Media Contact: Eileen Leahy
- Source: Elsevier
- Image Credit: Image credited to Neuroscience News
- Original Research (Open Access): Biological Psychiatry: Cognitive Neuroscience and Neuroimaging (Oct 1, 2026). Title: “Fluctuations in Real-Life Drinking in Alcohol Use Disorder: A 1-Year Longitudinal Study Using Gamified Tasks of Decision Making and Cognitive Control.” Authors listed in the original publication.
- DOI: 10.1016/j.bpsc.2026.07.019
Abstract
Fluctuations in Real-Life Drinking in Alcohol Use Disorder: A 1-Year Longitudinal Study Using Gamified Tasks of Decision Making and Cognitive Control
Background
Alcohol use disorder contributes substantially to global disease burden. Cross-sectional research ties AUD to weaker cognitive control and greater risky decision-making, but the temporal relationship has been unclear: are cognitive-behavioral changes a cause or consequence of changing alcohol consumption?
Methods
This study used a set of smartphone-based gamified tasks in a year-long ecological momentary assessment of 603 participants with largely mild-to-moderate AUD. Monthly tasks measured cognitive control (working memory, response inhibition) and decision-making (risk-taking across contexts, information sampling). Participants reported alcohol intake every two days.
Results
Monthly fluctuations in specific decision-making measures predicted subsequent consumption. Higher mixed-context risk-taking, lower loss-context risk-taking, and decreased information sampling each forecasted increased drinking the next month. The directionality was specific and unidirectional—prior alcohol use did not predict later changes in task performance—and the effect persisted after correcting for autocorrelation in drinking. Fluctuations in general cognitive control and win-context risk-taking were not associated with drinking variation.
Conclusions
These findings implicate specific decision-making alterations as drivers of changing alcohol consumption in AUD. Smartphone-delivered gamified tasks may help identify periods of heightened relapse risk and support the development of timely, mechanism-based digital interventions.