New animal research suggests that, for some individuals, stopping drinking may temporarily prime the brain for relapse rather than immediately protect against it. Scientists studying mice have identified patterns of brain activity during abstinence that appear to predict later compulsive alcohol use.
The findings point to a possible biological mechanism behind relapse in alcohol use disorder and suggest potential new ways to identify those at greatest risk. Although the research was conducted in mice, it builds on growing evidence that certain brain circuits may be reshaped during periods of sobriety.
How the abstinence study worked
Researchers gave mice long-term voluntary access to alcohol before imposing a period of forced abstinence. After this break, the animals were allowed to drink again, but researchers made the alcohol increasingly bitter by adding quinine, a compound that would normally discourage consumption.
A subset of mice continued drinking despite the intense bitterness, a behavior known as aversion-resistant alcohol intake. These animals not only ignored the unpleasant taste but also consumed larger quantities than mice that had not experienced forced abstinence.
This behavior resembles one hallmark of addiction: continuing to use a substance despite negative consequences. The results suggest that abstinence may, in some individuals, create a temporary state in which the drive to drink becomes more compulsive.
A key stress region in the brain
The researchers focused on a small but influential brain structure called the bed nucleus of the stria terminalis, or BNST. This region is involved in stress, anxiety and negative emotional states that can contribute to addictive behaviors.
By monitoring cells in the BNST, researchers found that abstinent mice showed heightened activity when they returned to an environment where alcohol had previously been available. The animals attempted to drink from the spout even when it contained only water, and these attempts coincided with increased BNST activity.
Mice that developed a preference for very bitter alcohol following abstinence showed more than twice the BNST activity of abstinent mice that did not develop compulsive drinking behavior. Importantly, this heightened activity appeared before the animals were given access to the bitter alcohol again.
What this could mean for people
The early increase in BNST activity suggests that this brain signal could potentially serve as a warning marker for relapse risk. If similar patterns are eventually confirmed in humans, clinicians could potentially identify people in early sobriety with elevated BNST activity and provide more intensive support.
Importantly, the findings do not suggest that quitting alcohol is harmful overall. Long-term abstinence remains strongly associated with better physical and mental health, particularly among people with alcohol use disorder or heavy drinking patterns.
Instead, the research suggests that the early stages of sobriety may involve biological stress that makes relapse more likely for some individuals. Recognizing and addressing this vulnerable period could improve outcomes for people attempting to stop drinking.
Alcohol’s widening public health toll
Alcohol misuse remains one of the most serious and underestimated public health challenges in the United States. Federal data indicate that alcohol-linked deaths have risen substantially over the past two decades, outpacing many other substance-related harms.
Recent analyses indicate that alcohol-related mortality is now several times higher than deaths attributed to opioids. Yet public concern and resources devoted to alcohol-related problems often remain behind the attention directed toward the opioid crisis.
More than 80% of Americans aged 12 and older consume alcohol at some point in their lives. Roughly 10% develop alcohol use disorder, meaning close to 30 million people will meet criteria for the condition and may potentially require treatment.
Despite several medications approved by the U.S. Food and Drug Administration, diagnosis and treatment rates remain low. Estimates suggest that the prevalence of alcohol use disorder in the United States has approximately doubled since the late 1990s, while access to care has not kept pace.
Rethinking treatment and relapse
Most current treatment models continue to emphasize complete abstinence, even as harm reduction approaches gain ground in the treatment of other substance use disorders. The new findings raise the possibility that a rigid focus on abstinence without strong relapse support may leave some patients particularly vulnerable.
Better tools for predicting who is most likely to relapse could allow clinicians to develop more personalized treatment strategies. People with high biological or psychological risk might benefit from more intensive follow-up, medication-assisted treatment or closer monitoring during early sobriety.
Researchers involved in the mouse study suggest that identifying brain-based markers of relapse susceptibility represents one promising direction. Such biological markers could eventually complement existing clinical assessments based on self-reported experiences and observable behavior.
Unanswered questions for brain science
Scientists still do not fully understand how the BNST influences alcohol-related behavior. It remains unclear which specific populations of neurons within this region are responsible for the increased activity observed during abstinence and subsequent compulsive drinking.
New neuroscience tools allow researchers to selectively activate or silence particular populations of neurons in animal models. Using these techniques, the team plans to investigate whether directly altering BNST activity can reduce harmful alcohol consumption in mice.
In parallel, psychiatrist Jennifer Blackford and colleagues are studying BNST function in people with alcohol use disorder during early abstinence. If similar patterns of brain activity emerge in humans, researchers could then evaluate BNST-based measures in clinical trials as potential screening tools or treatment targets.
Translating findings from animal experiments into human treatments remains a complex, multi-year process. Still, the study adds to growing evidence that brain circuits involved in stress and anxiety may play an important role in understanding why relapse is so common — and how it might eventually be prevented more effectively.
