The lyrics of a certain huge hit were sometimes misheard as: “You can tell by the way I use my walk/I'm an worried man, no time to talk.” It's a pity that isn't what the Bee Gees actually sang, because they could have claimed vindication from scientific research 45 years later. A new paper has shown young people with anxiety tend to walk in ways that resemble the elderly, offering a potentially valuable diagnostic tool.

The fact anxiety can show up in people's movements has probably been known for as long as humanity has had a concept of the condition. Many people even diagnose their pets' mental health this way. A reliable technique that can play a part in psychological assessments and treatments is a different matter, which is what a group of students originally at Clarkson University sought.

In Sensors the team led by Dr Maggie Stark, now at Lake Erie Osteopathic College of Medicine, announce some promising progress.

As the paper notes, previous studies have found patterns in the gaits of people who have reported feelings of anxiety. “The purpose of this study was to identify individuals who report feeling anxious at that time using a combination of gait and quiet balance [machine learning],” the authors write.

Participants in the research were asked to take a standard mood test that could establish their level of anxiety, and take a two-minute walk around a curved track over firm and spongy ground, wearing nine mobility sensors. In between, they took a test to measure their balance.

The sensors provided a wealth of information about the study subjects' movements. The most useful of these features in predicting anxiety all related to people's gait, rather than the balance test. These included the angles at which they turned and more neck movement side-to-side, but less up and down, than those feeling confident. Putting a range of measures together allowed the authors to predict which of the subjects were suffering anxiety with 75 percent accuracy.

Even though most of the participants were young and relatively physically healthy, their movements resembled those of the elderly who are afraid of falling when they walk. This led to difficulties turning, and constant scanning from side-to-side to search for threats.

Test measures like this might seem superfluous – unlike pets, people can be asked if they're feeling anxious. However, the paper notes that self-reporting “can be influenced by an individual’s culture and gender differences.” In particular, stigma against mental illness prevents many people from seeking treatment when they would be likely to benefit. As with depression, easy-to-read objective measures for anxiety could be transformative for many individuals. If, as is likely, widespread application of these measures also reduced stigma, it could have profound social effects as well.

The accuracy achieved might be insufficient to make walking observations a standard test of anxiety just yet. However, it represents an impressive first effort with 88 participants, and team members are seeking to build on it with larger samples.

Staying Alive is already used in first aid education as the ideal speed to give CPR; if it helps people with anxiety get diagnosed, it could live up to its title in another way as well.