What Alcohol Does to the Brain, and What the Brain Has to Do to Come Back

Somewhere around day three of not drinking, the craving arrives in a way that feels less like wanting and more like need. It’s not a preference. It doesn’t feel like preference. It arrives with a physical quality — a restlessness in the chest, a narrowing of attention, a background pull that makes everything else harder to focus on. People in early alcohol recovery describe it in terms that sound like physics: a gravity. Something that has to be actively resisted rather than simply declined.

That experience is not a moral failing. It is what a brain shaped by years of alcohol exposure does when the alcohol stops. Understanding the neuroscience behind that experience doesn’t make it easier to live through, but it does make it legible — and it explains why certain treatments work, why relapse is so common, and why the first year of recovery is neurologically as hard as it feels.

The reward system and the hijack

Alcohol affects the brain across multiple neurotransmitter systems simultaneously — which is part of what makes alcohol use disorder so neurologically complex to treat. Alcohol enhances the effects of GABA (the brain’s primary inhibitory neurotransmitter), suppresses glutamate (the primary excitatory one), and triggers surges of dopamine in the nucleus accumbens, the brain’s reward center. That combination produces the relief and euphoria that characterizes early drinking.

With repeated alcohol exposure, the brain adapts. GABA receptors downregulate; glutamate receptors upregulate. The system recalibrates around the presence of alcohol, rebalancing itself against the chemical that has been consistently present. The result is tolerance: larger quantities are needed to achieve the same effect. But it’s also dependence: when alcohol is removed, the rebalanced system swings the other way. Now GABA is suppressed and glutamate is surging with nothing to check it. That is alcohol withdrawal — neurological excitation without inhibition, in its most severe forms producing seizures and delirium tremens.

The dopamine system tells a parallel story. Alcohol initially elevates dopamine in the reward circuits. Over time, the baseline dopamine tone in those circuits drops — the brain, again, adapting to the consistent input by reducing its own production. The person now needs alcohol to feel normal. Ordinary rewards — food, social connection, activities that previously produced pleasure — register weakly against the recalibrated system. This is anhedonia, and it is one of the most reliable features of early recovery from alcohol use disorder. It is also one of the least discussed reasons why relapse happens: in the first weeks and months off alcohol, the sober brain is literally less able to feel pleasure than it was before. That’s not a character flaw. It’s a neurochemical state.

The stress axis and the craving circuit

A 2026 study published in eLife explored how alcohol interferes with the brain’s stress-response circuits — specifically, how it attenuates the activity of corticotropin-releasing factor (CRF) signaling in the extended amygdala. CRF is a key driver of the stress response, and its interaction with dopaminergic circuits in the striatum helps govern how the brain responds to threat, discomfort, and craving.

What this research adds to the picture is a mechanistic explanation for something clinicians have observed for decades: that cravings for alcohol are strongly triggered by stress. In a brain shaped by chronic alcohol use, the alcohol has been suppressing the CRF-driven stress signal in the amygdala. When alcohol is removed, that suppression lifts. Stress responses become more intense and harder to regulate. The craving is partly the brain searching for the chemical it has been using as a stress buffer — and the search is activated by exactly the kind of life situations (conflict, financial pressure, loss, difficult emotions) that are most likely to be present in early recovery.

In a brain shaped by chronic alcohol use, the alcohol has been suppressing the CRF-driven stress signal in the amygdala.

This is the biology behind what clinicians call craving-to-use pathways: the sequence from stress to urge to relapse that characterizes so many returns to drinking. It runs through the amygdala and its connections to the striatum. It is not a decision. It is a circuit.

What the gut has to do with it

Research over the past few years has established a significant role for the gut-brain axis in alcohol withdrawal, withdrawal-associated depression, and craving. The picture emerging from this line of research, synthesized in a 2023-24 NIH review, is that the immune response in the gut contributes directly to the psychological symptoms of alcohol withdrawal.

People with alcohol use disorder often have gut microbiome disruptions, leaky gut walls, and elevated inflammatory markers. During withdrawal, these inflammatory signals travel via the vagus nerve to the brain, contributing to depressive symptoms, anxiety, and heightened craving. The gut-brain connection means that alcohol withdrawal is not purely a neurological event — it’s a whole-body inflammatory state that manifests partly in the brain. This explains why some people experience prolonged psychological symptoms well past the acute withdrawal phase, in what clinicians call post-acute withdrawal syndrome (PAWS): the nervous system is still processing an inflammatory signal that originated in the gut.

The relapse brain

New research published in 2026 in the European Archives of Psychiatry and Clinical Neuroscience found that relapse in alcohol dependence is specifically associated with disrupted modular brain network organization — meaning that the normal separation of functional brain networks becomes less distinct in people who relapse. A healthy brain maintains clear boundaries between networks involved in cognitive control, emotional processing, and reward. In people who relapse following a period of abstinence, those networks become more diffuse, less well-organized, and less capable of the top-down regulation that enables someone to override an impulse.

This finding matters because it gives biological substance to what people in recovery describe as the experience of relapse: not a conscious decision so much as a collapse of the part of the brain that usually intervenes between impulse and action. The prefrontal cortex, which handles executive function and impulse control, is among the last regions to fully recover from chronic alcohol exposure — and its recovery can be disrupted by stress, sleep deprivation, and emotional overwhelm. The relapse brain is often a brain that has been working hard under adverse conditions and whose regulatory circuits have been worn down rather than rebuilt.

What coming back looks like, neurologically

The good news embedded in all of this is genuine: the brain recovers. Neuroplasticity — the same property that allowed alcohol to reshape the brain’s circuits in the first place — also allows recovery to reshape them back. The timeline is specific. In the first two weeks following acute withdrawal, initial grey matter recovery begins and small cognitive improvements emerge. By one to three months, dopamine levels begin to normalize, mood and sleep improve, and cravings decrease in intensity.

That timeline does not mean recovery is complete at three months. The prefrontal cortex recovery extends much further — years, in some cases. But it does mean that the anhedonia and the intense cravings of the first weeks are not permanent features. They are a neurological state that changes.

A 2025 randomized controlled trial found that wearable HRV (heart rate variability) biofeedback, used as an adjunct to standard treatment, was associated with reductions in negative affect, craving, and alcohol and other drug use — and specifically appeared to weaken the craving-to-use pathway. This approach works partly by training the autonomic nervous system to regulate stress responses more effectively, directly addressing the CRF/amygdala circuit that alcohol had been suppressing. It is not a cure. It is a tool that helps the brain do what it is already trying to do.

It is a tool that helps the brain do what it is already trying to do.

At UCSF, a current clinical trial is evaluating brenipatide — a novel compound — compared to placebo in people with alcohol use disorder. The NIAAA’s Medications Development Branch is actively supporting additional trials. The pharmacological landscape for alcohol use disorder remains limited relative to the scale of the problem, but it is expanding.

Alcohol killed 105,400 Americans in 2022, roughly matching drug overdose deaths in the same year. It kills quietly, without the same crisis infrastructure of overdose surveillance and naloxone distribution. The neuroscience above is the reason it’s so hard to stop, and the reason that “just don’t drink” is not a clinical plan. It is also the reason that treatment works — when the brain has support, time, and a managed environment, it rebuilds what the alcohol disrupted.

The craving that feels like physics is physics. The brain that can come back from it is the same one.

See more in the science & medicine category, or explore the alcohol substance pillar for treatment resources and further reading.