Music That Follows Your Body: What the Research Shows So Far

Music that speeds up when your heart does, or steps in time with your feet, is an appealing idea. It is also an idea with evidence behind it, some of it surprising. We covered the experience of this in adaptive music for fitness and wellness. Here we look at what researchers have actually measured.

Short version: the early results are encouraging, the studies are small, and the researchers themselves are careful. That care is worth borrowing.

A trial with 54 trail runners

In 2020, Duncan Williams, Bruno Fazenda, Victoria Williamson and György Fazekas published a study in the journal Sensors. They tested a system that generated music in real time from a runner's data, using a wearable sensor that combined heart-rate sensing and GPS. Tempo, melody and timbre were modulated by the sensor input.

Fifty-four trail runners, 27 women and 27 men, were randomly assigned to one of two conditions. In one, the music followed cadence. In the other, it followed heart rate. The researchers measured pace, duration, distance, average heart rate and perceived exertion on the Borg scale.

What they found

The abstract reports that cadence-synchronous music improved performance and decreased perceived effort in male runners. For female runners, cadence-synchronous music improved performance too, but heart-rate-synchronous music significantly reduced perceived effort, and women in that group ran longest of all.

Average heart rate did not differ significantly between conditions. In other words, the music seemed to change how hard the run felt and how long people kept going, not their measured heart rate.

The limits the authors admit

The authors list their own caveats, and they are substantial. The sample was small. Fourteen of the women reported identical distances in one condition, which raises questions about data quality. Running data was self-reported on an honor system, there were no follow-up interviews, and only one version of the music generator was tested.

So the result is a signal worth following up, not a rule. A gender difference in this one trial does not prove men and women respond differently in general.

A different use: walking after Parkinson's

Rhythm can also help people who find walking hard. In 2021, Cochen De Cock, Dotov, Damm and colleagues reported on BeatWalk, a smartphone app that uses ankle sensors to align music tempo with a person's steps, for people with Parkinson's disease.

Users followed about 78.8 percent of the prescribed time. Falls did not increase and pain decreased. After four weeks, gait improved even without music: cadence rose by about three steps per minute, and velocity and stride length improved.

This study was uncontrolled and open-label, with no comparison group, and the paper notes that a randomized controlled trial was underway. Encouraging, then, but not proof.

The broader picture from exercise psychology

A wider review gives context. In 2012, Costas Karageorghis and David-Lee Priest reviewed music in the exercise domain in the International Review of Sport and Exercise Psychology. They reported that during endurance activities, suitable music has been shown to enhance affect, reduce ratings of perceived exertion, improve energy efficiency and lead to increased work output.

Note the word "suitable." The review also points to tempo, timing, synchronization and exercise intensity as factors that can change the response. A system that follows the body is one way to keep music suitable as the body changes.

What this means for anyone designing responsive music

A few practical cautions follow from the research. Responsiveness should serve the listener's goal, not show off the sensor. The same adaptation may not suit everyone. And claims should be modest until larger trials exist.

For music makers, there is a design point too. To follow a body, music must be able to change tempo and texture without falling apart, which depends on how it was built. Our piece on designing a song so it can adapt explains why that groundwork matters.

Questions to ask when you read a claim

News stories about music and exercise often skip the details that matter. A few questions help you weigh any study you meet.

How many people took part, and how were they chosen? The trail-runner study had 54 volunteers from a local club, so its findings may not describe everyone. Was there a comparison group? The BeatWalk study had none, which the authors acknowledge.

What was actually measured? In the trail study, perceived effort and distance changed, but average heart rate did not. That is different from saying the music made people fitter.

And was the data checked, or reported by participants? Self-reported runs are weaker evidence than timed ones.

Why one size will not fit everyone

The results hint at individual differences, and the Karageorghis and Priest review makes the same point by listing tempo, timing and exercise intensity as things that can change the response. People differ in taste, fitness, mood and what they are trying to do.

For a designer, that means giving the listener a way to adjust. A system that follows the body but lets you choose the style, the voice or the pace respects those differences. A system that dictates them does not.

This is also why research on responsive music keeps returning to personalization. The body supplies the signal, but the person still decides what they want to hear.

If you are curious about trying responsive music yourself, treat it as an experiment. Notice whether it changes how long you keep going or how hard the effort feels, and keep what works for you.

Sources

• Sensors (via PubMed Central): On Performance and Perceived Effort in Trail Runners Using Sensor Control to Generate Biosynchronous Music (Williams et al., 2020)
• Frontiers in Psychology: BeatWalk, Personalized Music-Based Gait Rehabilitation in Parkinson's Disease (Cochen De Cock et al., 2021)
• Brunel University repository: Music in the exercise domain, a review and synthesis (Karageorghis and Priest, 2012)