Researchers have identified a characteristic pattern of brain activity that distinguishes people with cocaine addiction from healthy individuals. This connectivity pattern also predicts how well a person will respond to magnetic stimulation therapy that reduces craving. The results were published in Nature Mental Health.
Cocaine addiction is compulsive, uncontrollable drug use with a high likelihood of relapse. Existing treatments do not help everyone, so doctors need objective biological markers for diagnosis and targeted approaches.
One way to find such markers is functional magnetic resonance imaging (fMRI). It measures blood flow and shows which brain areas are active at the same time. If regions are activated together at rest, this is called functional connectivity - the coordinated work of brain regions. This is a correlation, not proof of cause.
Another avenue is repetitive transcranial magnetic stimulation (rTMS). This is a non-invasive method: magnetic fields stimulate brain areas. Trials show that rTMS can reduce craving, but why it does not help everyone is unclear.
Bioengineers Kanhao Zhao and Yu Zhang of Lehigh University and colleagues set out to determine whether a reliable connectivity pattern exists in cocaine addiction and whether it predicts response to stimulation.
First, they collected fMRI data from 129 people in Mexico: 71 patients and 58 healthy individuals. Participants underwent resting-state scanning. The researchers then built connectivity maps and trained a machine learning algorithm to separate the groups. The model identified an addiction signature with 83% accuracy. Patients had increased connectivity between the visual processing and attention networks, as well as between the goal-directed behavior and sensory attention networks. At the same time, connectivity decreased between the default mode network (internal thoughts) and the limbic network (emotions).
In an independent sample - 82 patients and 81 healthy individuals from the United States and Mexico - accuracy was 70%, and the same differences were confirmed. This indicates a common feature of addiction rather than chance.
The researchers then tested whether the signature predicts treatment outcome. They analyzed data from a trial with 45 patients: 25 received active rTMS and 20 received sham stimulation. Based on pre-treatment scans, the model predicted a reduction in craving among those receiving active stimulation: certain connections in the frontal lobes were associated with relief. For the sham group, the prediction did not hold, which points to a link with the actual treatment. A comparison of scans before and after the course showed that active therapy normalized connectivity between the cognitive control and movement networks.
The authors note limitations. The study is observational: it cannot be stated that the patterns cause addiction - cocaine may alter connectivity, or pre-existing features may increase vulnerability. Experiments that influence these circuits are needed. Participants included more men, so the signature may not fully reflect the picture in women. Data were collected at different centers, and differences in scanner settings introduce error - standardized methods and diverse samples are needed.
The scientists hope that over time such patterns will become a screening tool: scanning will help a doctor decide whether magnetic stimulation is suitable for a patient or whether another method should be chosen.