In 17 volunteers, LSD speeds up brain waves, not just weakens them

Seventeen healthy volunteers took part in a study that could only be read by measuring the brain to the millisecond. Each of them went through two sessions, two weeks apart. In one, they received an intravenous injection of 75 micrograms of LSD, a dose the researchers compare to roughly the weight of a grain of salt. In the other, they received a placebo, a saline solution. The recordings were made with magnetoencephalography (MEG), a technique that captures the tiny magnetic fields produced by neuronal activity, at Imperial College London, by the team of Robin Carhart-Harris.

For years, the usual reading of psychedelic effects on brain waves was that they weakened. Alpha waves, which dominate when we rest with our eyes closed, and beta waves, linked to concentration and decision-making, appeared less intense under the drug. This was generally interpreted as “desynchronization”, meaning neurons fire less in unison. But conventional methods did not always separate a real weakening of the waves from a change in their frequency, the number of oscillations per second. The team combined state-of-the-art signal-processing methods with artificial intelligence to tell the two apart.

The answer is that under LSD the waves do not just fade. They speed up. Alpha waves gained about 2 hertz and beta waves about 5 hertz, across large areas of the cortex. Jerbi compares the mistake to a radio left tuned to 99.5 FM when the station has moved to 100.5: the signal seems distorted or lost. The alpha shift also runs against certain anesthetics, which slow that rhythm, and points, in the authors’ reading, to a more excitable cortex.

The brain signal also became more complex. The researchers measured its fractal dimension, the way a pattern repeats at every scale, like each floret of a romanesco broccoli being a small romanesco. Both this measure and the diversity of patterns rose under LSD, so the signal looked more like a jagged coastline than a straight beach. The changes were strongest in regions tied to sensory perception, language, emotion and mental imagery, while the motor cortex stayed relatively unaffected. The team trained an AI algorithm to distinguish LSD sessions from placebo sessions, to test whether these changes could work as a signature.

What changes for research is a more precise way to read psychedelic states. If the speed of brain rhythms is part of the drug’s effect, studies that only track wave strength may have missed part of it, and researchers designing trials of psychedelics for depression, anxiety or post-traumatic stress disorder could add a second marker to follow. The link the authors draw to the geometric patterns many users report seeing with their eyes closed remains a hypothesis: Jerbi notes that “we measure the fractal structure of brain signals over time, not the fractal structure of the images people perceive.” The study covers 17 people and one dose, and the authors have not yet shown that the signature is specific to LSD.