A patient I worked with years ago did everything the textbook asked. He finished a 30-day program, started naltrexone to blunt the reward from alcohol, and stayed sober for four months. Then a bad week at work, two nights of broken sleep, and a single skipped meal stacked up, and he was drinking again by Friday. His counselor framed it as a lapse in commitment. I saw something else. The medication was working on exactly one part of his brain while the relapse was coming from three or four others.
A review published in May 2026 in Frontiers in Pharmacology, led by researchers synthesizing molecular, circuit, and clinical evidence, makes the case plainly. Addiction cannot be explained by dysfunction in the brain's dopamine circuits alone. It reflects what the authors call maladaptive interactions across distributed neuromodulatory and glial networks. In plain language: at least six brain systems are involved, dopamine is only one of them, and the other five are the ones most treatment programs never look at.
Dopamine Was Only Ever Part of the Story
For 30 years, the dopamine model has done a lot of good. It moved addiction out of the moral gutter and into the brain, where it belongs. Dopamine surges with a drink or a hit, the reward circuit learns to chase that surge, and over time the circuit demands the substance just to feel normal. That account is real, and it is why I tell families this is not about willpower. It never was.
But the dopamine story has a gap. If addiction were only a dopamine problem, a drug that dampens dopamine reward should fix it. We have those drugs, and they help some people some of the time. They are not cures, because the craving brain is drawing on more than reward chemistry. The 2026 review reframes reward and aversion as a single process the authors call valence, in which dopamine constantly trades signals with sleep-wake chemistry, the stress axis, the body's metabolic state, and the brain's own cannabinoid system. Pull on one thread and the others move.
The Systems Your Rehab Never Tested
Start with orexin, the chemistry that governs wakefulness and arousal. When my patient lost two nights of sleep, his orexin system was already pushing his reward circuit toward a hair trigger before he ever thought about a drink. Then there is the stress system, the HPA axis, which the review ties directly to relapse vulnerability. Cortisol does not just make you feel bad. It reshapes how the reward circuit weighs a cue.
The metabolic system is the one that surprises people most. The review highlights GLP-1, the same gut hormone behind the new weight-loss drugs, as a genuine lever on craving. That is not a coincidence. Hunger, blood sugar, and reward share wiring, which is why a skipped meal can feel like a relapse trigger and why we have started seeing GLP-1 medications cut drinking in early trials. The endocannabinoid and histamine systems round out the picture, each tuning how strongly a cue lands. The authors even point to a recently mapped brainstem node, the subventricular tegmental nucleus, that appears to integrate signals from all of these inputs before they reach dopamine neurons.
None of this shows up on a standard intake form. A patient gets asked about substances, family history, and mental health, then gets a one-size protocol. Nobody checks the sleep system, the stress load, or the metabolic state that the science now says are driving the relapse.
Why One Drug Is Rarely Enough
Here is the main point I hope you take away from this. If addiction is a network problem, a single-target treatment is a single-target solution to a multi-target disease. Naltrexone speaks to the opioid and dopamine reward layer. It says nothing to a wrecked sleep cycle or a stress system stuck in overdrive. The medication is doing its job. The trouble is that the job was scoped too narrowly for a disease this distributed.
This is the reasoning behind how we structure care in our Rescue From Rehab program. We treat the reward circuit, and we also test and address the systems around it, the sleep architecture, the stress load, the metabolic markers, and the nutritional gaps that quietly keep the brain primed to relapse. It is the same logic behind targeted brain nutrition through Action Potential Supplements, which supports the metabolic and neurochemical groundwork that recovery depends on. Recovery holds when the whole network is stabilized, not when one node is quieted and the rest are left alone.
What This Means for You
If you or someone you love has relapsed after doing the work, read the relapse as information, not as failure. Ask what the program actually measured. Was sleep assessed and treated? Was the stress system addressed with more than a coping-skills handout? Were metabolic markers and nutrition ever checked? If the honest answer is no, the relapse may have come from a system that was never in the treatment plan to begin with.
This connects to something I have written about before, how the brain holds a relapse memory and a recovery memory side by side, each fighting for control. The wider the base of recovery, the more often the healthier memory wins. Stabilizing six systems instead of one tilts that contest in your favor.
The dopamine model gave us permission to call addiction what it is, a brain condition rather than a character defect. The multisystem model takes the next step. It tells us why recovery is hard, why single fixes fall short, and where the rest of the work has to happen. People in recovery deserve care built for the brain they actually have, not the simplified one a 30-day program was designed around.