A brain scan can make a behavioral treatment look unusually concrete. Colored patches appear to show the intervention working inside the skull, and a complicated recovery process seems to acquire a biological signature. A new randomized trial offers exactly that kind of image—but its most useful lesson is also a warning against mistaking a mechanism for a verdict.
Published online in JAMA Psychiatry on October 28, 2026, the study tested Mindfulness-Oriented Recovery Enhancement, or MORE, in people receiving inpatient treatment for opioid use disorder. The program combines mindfulness practice with reappraising drug cues and deliberately savoring ordinary rewards. After eight weekly two-hour group sessions, participants assigned to MORE showed a different pattern of prefrontal activity when regulating responses to opioid-related images and natural rewards than those assigned to psychoeducational supportive therapy.
That finding matters because opioid addiction can narrow attention and motivation around drug cues while everyday rewards lose some of their pull. But the trial does not show that a particular scan pattern diagnoses recovery, nor that mindfulness should replace medication. It asks a narrower and more interesting question: can a structured behavioral intervention change how people engage with competing rewards?
What changed—and what the trial actually measured
The randomized clinical trial enrolled adults with opioid use disorder at one inpatient rehabilitation facility in New York between November 2020 and September 2024. Both groups received standard care, including medications for opioid use disorder. MORE was therefore an adjunct, not a stand-alone alternative.
The principal analysis included 59 participants who completed a drug-cue reactivity task during functional magnetic resonance imaging: 24 in MORE and 35 in the comparison group. Participants practiced either reducing their response to drug images or amplifying their response to non-drug rewards. Compared with supportive therapy, MORE reduced the relative drug bias in several regions, including parts of the middle frontal gyrus and precuneus. Within the MORE group, larger shifts toward natural-reward responses in lateral and ventromedial prefrontal cortex were associated with lower cue-induced craving.
A second task used a naturalistic drug-related movie. Among 37 participants with usable data, MORE reduced synchronization in medial prefrontal activity compared with the control condition. At roughly four months after treatment, a subset of 38 participants also showed longer abstinence in the MORE group. The abstinence association was statistically significant, but the reported effect was small, and it was a secondary outcome.
These distinctions are important. Random assignment supports a causal interpretation of differences between the two therapies on measured outcomes. It does not prove that the brain changes caused the clinical changes. The link between altered prefrontal responses and craving was correlational within one treatment group; a formal mediation result was not reported.
Why a quieter prefrontal response is not simply “better”
Popular neuroscience often treats more activity as stronger control and less activity as impairment. Here the interpretation runs the other way: less drug-biased prefrontal recruitment may mean that regulating a cue requires less effort, or that drug images command less salience. That is plausible, but fMRI measures changes in blood oxygenation, not craving, choice, or recovery itself. The same signal can reflect multiple psychological processes.
The study’s preregistered record helps readers see what was planned, while the active comparison condition makes the result more informative than a wait-list design. Still, the evidence is constrained by attrition and small analytic subsets. Only 13 women were included in the main imaging analysis, participants came from one inpatient setting, and the groups were uneven in size. Results may not generalize to outpatient care, people earlier in treatment, or the many patients who cannot attend sixteen hours of group sessions.
The brain results also should not overshadow the clinical hierarchy of evidence. The US Substance Abuse and Mental Health Services Administration identifies buprenorphine, methadone, and naltrexone as evidence-based medications for opioid use disorder. Behavioral therapies can add support; they do not erase the mortality protection and treatment-retention evidence behind medication.
The real promise is an adjunct, not a neural stamp of approval
MORE is not resting on this scan study alone. A 2024 randomized telehealth trial involving 154 people receiving methadone found lower hazards of returning to drug use and leaving treatment when MORE was added to usual care, although half of participants in the MORE arm still returned to drug use during follow-up. A 2026 systematic review of 14 randomized trials concluded that mindfulness-based interventions appear promising for craving, opioid use, pain-related functioning, and distress, while emphasizing heterogeneity, variable study quality, and unanswered questions about durability and scale.
The new trial adds a mechanistic clue to that clinical record. It suggests recovery-oriented training may not work only by suppressing desire; it may also help restore attention to non-drug rewards. That is a richer account of behavior change than “use more self-control.”
But the scan is not a biomarker ready for the clinic, and it is not evidence that one therapy will work for a particular person. The honest conclusion is smaller: in a limited sample receiving standard treatment, an eight-week program changed both laboratory brain responses and some recovery-related outcomes. The next test is whether those gains replicate in larger, more diverse groups and persist in ordinary care. Until then, the fMRI image is best read as a research map—not a certificate of recovery.
