How songs may help reduce pain—and even treat stroke symptoms

In the 2014 documentary Alive Inside, Henry, a 94-year-old nursing home resident with dementia, sits hunched in his wheelchair, his head lowered and hands clasped. Then, a social worker slides headphones over Henry’s ears and plays a 1930s Cab Calloway song from his youth.

Henry opens his eyes. He lifts his head, begins swaying to the music, and soon starts singing and speaking animatedly about his memories.

Scientists are still trying to explain moments like Henry’s. Familiar music can sometimes draw out memories and emotional responses in people with dementia, while rhythmic auditory cues can improve aspects of walking in some people with Parkinson’s disease or stroke. Now researchers are moving beyond the broad observation that music can help. They are trying to identify which of its components—rhythm, tempo, familiarity, melody, or precisely timed sensory pulses—produce measurable effects in the brain and body.

The evidence varies widely. But together, the research suggests that music is not a single treatment but a collection of biological signals—and that its therapeutic potential may depend on matching the right element of sound to the right brain.

How the brain finds the beat

Before music can move a body or stir a memory, sound must first become a neural signal. Vibrations in the air strike the eardrum and move three tiny bones in the middle ear. The last of these, the stapes, transfers that motion through the oval window into the fluid-filled inner ear. There, sensory hair cells in the cochlea convert the vibrations into electrical signals that travel along the auditory nerve through the brainstem and thalamus before reaching the auditory cortex.

But hearing music involves far more than the auditory cortex alone. It also recruits networks involved in movement, prediction, memory, emotion, and reward.

Edward Large, a theoretical neuroscientist and physicist who directs the Music Dynamics Laboratory at the University of Connecticut, studies how the brain detects and anticipates rhythm. He helped develop the neural resonance theory (NRT), which proposes that the human brain physically synchronizes its own neural pulses to an external musical beat.

(How music is used to heal the sick in Appalachia.)

Initially, Large says he was skeptical that music therapy had a measurable scientific basis. The turning point came when music therapist Concetta Tomaino—who co-founded the Institute for Music and Neurologic Function with late neurologist Oliver Sacks—invited Large to observe and participate in a 30-minute patient drum circle. As the group began to play, Large watched previously withdrawn patients begin tapping their feet, swaying in time, and responding to one another. To him, the scene offered a vivid example of how an external rhythm could organize attention and movement.

Large’s laboratory later tested a related idea: whether the brain merely follows a beat or actively constructs one. They created 30-second sequences of complex, syncopated drum patterns in which listeners could perceive a steady pulse even though the sound contained no acoustic energy at that frequency. As participants listened, the researchers used magnetoencephalography (MEG) to measure the tiny magnetic fields produced by neural activity.

The researchers detected activity at the missing pulse frequency in the auditory cortex.

“They weren't just remembering the rhythm,” explains Large. “They were physically generating it internally.”

Can rhythm change the body?

The effects of rhythm may also depend on how closely it matches the listener. In a recent study, Caroline Palmer and her team at McGill University first measured each participant’s spontaneous production rate—the natural tempo at which the person tapped a familiar melody. The researchers then played music at that rate or at tempos 15 percent faster or slower.

Participants reported the least pain when the music matched their natural tempo. The study used experimentally induced heat pain in healthy adults, so it does not show that personalized music can replace pain medication. But it suggests that music’s effects may depend not only on what a person hears, but on how its timing relates to the body’s own rhythms.

Other researchers are investigating a more experimental possibility: whether precisely timed sensory stimulation can influence patterns of neural activity disrupted by disease.

One focus is gamma activity, a range of fast neural oscillations involved in processes including perception, attention, and memory. Studies have found that gamma activity can be disrupted in Alzheimer’s disease, prompting researchers to ask whether externally delivered stimulation might reinforce those rhythms.

In a 2019 study, neuroscientist Li-Huei Tsai and colleagues at Massachusetts Institute of Technology exposed mice genetically engineered to develop Alzheimer’s-like amyloid buildup to tones pulsing 40 times per second for one hour a day over seven days. The stimulation increased 40-hertz neural activity in the auditory cortex and hippocampus and reduced several measures of amyloid burden in both regions. The mice also performed better on tests of spatial and recognition memory.

(They sing to remember: The power of memory choirs.)

When the researchers paired the tones with flickering light at the same frequency, the synchronized activity reached additional areas, including the medial prefrontal cortex, and produced broader reductions in amyloid pathology. Researchers are now testing whether 40-hertz stimulation can reliably engage the same neural activity in humans and, more importantly, whether doing so can meaningfully slow cognitive or functional decline.

Seeking to translate the underlying science into a usable therapy, Large co-founded Oscillo Biosciences to develop a wearable system called SynchronyGamma. The system pairs music with precisely synchronized visual stimulation, rather than asking users to listen to repetitive laboratory tones alone.

Whether that more personalized format improves adherence, changes neural activity, or produces clinical benefits has not yet been established. That is the larger challenge facing the field: turning an intriguing biological effect into an intervention that measurably helps patients.

Turning music into a prescription

Some of the most clinically developed applications are already moving beyond the laboratory and into rehabilitation. Johns Hopkins neurologist Alexander Pantelyat is working toward what he describes as a personalized, data-driven “acoustic prescription.” He has collaborated with researchers at Boston University and the digital therapeutics company MedRhythms to test a closed-loop rhythmic auditory stimulation system.

The resulting technology—listed as an FDA Class II prescription device for stroke (InTandem®) and Parkinson’s disease (Movive®)—uses shoe-worn sensors to stream real-time gait data to software that adjusts the music’s tempo and rhythmic cues in response. If a user falls out of step, the system can slow the tempo or add a more pronounced beat to help restore the walking rhythm. (Disclosure: Pantelyat serves on MedRhythms’ scientific advisory board.)

(Your brain shrinks after 40. Learning a musical instrument can reverse it.)

In a 2025 randomized controlled trial, 44 individuals with Parkinson’s disease were assigned either to walk with the autonomous music system or to complete a brisk-walking program without rhythmic cues. After six weeks, those using the system had added an average of 3,343 daily steps from baseline, compared with 172 steps in the control group. They also spent more time walking at moderate intensity and showed less stride-to-stride variability, a measure associated with steadier gait. Those gains in daily steps and moderate-intensity walking were no longer evident after participants stopped using the device.

The stroke evidence is also encouraging. In a randomized trial of 87 adults with chronic walking impairments after stroke, participants who used InTandem for five weeks improved their walking speed more than those following an active walking program. Forty percent of InTandem users achieved a clinically meaningful improvement, compared with 13 percent of the control group.

Future studies aim to evaluate whether this auditory entrainment can ultimately reduce fall risk. “If we can reduce falls by 10 to 20 percent, that’s a major clinical milestone,” Pantelyat says.

But researchers believe adjusting tempo is only the beginning. “Music is not a single treatment—it is a collection of therapeutic components,” says Kyurim Kang, a neuroscientist and board-certified music therapist at the Johns Hopkins Center for Music and Medicine.

The next challenge is to determine which components—rhythm, tempo, familiarity, melody, or emotional meaning—matter most for a particular symptom and patient, bringing researchers closer to the individualized “acoustic prescription” Pantelyat envisions.