Gait analysis of dementia patients reliably identifies Alzheimer's

Gait analysis of dementia pati...
Doctors might soon be able to differentiate between dementia types in older patients by analyzing the way they walk
Doctors might soon be able to differentiate between dementia types in older patients by analyzing the way they walk
View 1 Image
Doctors might soon be able to differentiate between dementia types in older patients by analyzing the way they walk
Doctors might soon be able to differentiate between dementia types in older patients by analyzing the way they walk

The progression of a range cognitive conditions is often associated with a deterioration in motor control, and recently we’re seeing how the finer details of this physical decline can help us tell one type of dementia from another. A new study has demonstrated how this might work in distinguishing Alzheimer’s disease from other types of dementia, by focusing on one specific difference in the patient’s gait.

By observing the way people walk, scientists have long been able to draw conclusions about their cognitive and physical wellbeing, and by focusing on specific aspects of this, researchers have been able to tease out some fascinating insights. One 2019 study, for example, uncovered correlations between slow walking during middle age and accelerated biological aging.

Another example, again from 2019, explored the idea that Alzheimer’s disease could be distinguished from Lewy body dementia, the two most common forms of dementia, based only on a patient’s walking patterns. By studying the gait of 110 subjects, the researchers were able to pick out walking pattern differences that separated sufferers of the two conditions, with the most common relating to gait variability, or differences in the timing and length of each step.

An international team has now widened the scope of this field of study in new research involving 500 older adults, in what the researchers call the first systematic investigation of gait impairments across the cognitive spectrum. The group included sufferers of subjective and mild cognitive impairment, Parkinson’s disease, Alzheimer’s disease, frontotemporal dementia and Lewy body dementia, along with cognitively healthy controls.

The subjects had their cognitive performance assessed and their gaits analyzed, with the scientists focusing on pace, rhythm, variability and postural control. Of these different factors, the team found that gait variability alone was associated with a lower cognitive performance, and could be used to reliably distinguish Alzheimer’s disease from the other conditions with 70 percent accuracy.

“This is the first strong evidence showing that gait variability is an important marker for processes happening in areas of the brain that are linked to both cognitive impairment and motor control,” notes Dr. Frederico Perruccini-Faria, from the Lawson Health Research Institute and first author on the paper. “We’ve shown that high gait variability as a marker of this cognitive-cortical dysfunction can reliably identify Alzheimer’s disease compared to other neurodegenerative disorders.”

The hope is that as this type of research continues to advance and the technique becomes more accurate and reliable, it can become a valuable tool for clinicians working to diagnose dementia types in patients suffering cognitive decline. Doing so at an early stage can ensure they get the most suitable treatments and potentially limit the severity of the condition as it progresses.

“We have longstanding evidence showing that cognitive problems, such as poor memory and executive dysfunction, can be predictors of dementia,” says study author Dr. Manuel Montero-Odasso. “Now, we’re seeing that motor performance, specifically the way you walk, can help diagnose different types of neurodegenerative conditions.”

The research was published in the journal Alzheimer’s & Dementia.

Source: Lawson Health Research Institute

Given that many middle aged people get knees and hips replaced along with arthritis problems, analyzing their gate would seem like a poor method to use for diagnosing dementia.
While there maybe a correlation that has been shown through research, using this correlation as a diagnostics tool will be much harder. We've been seeing this issues with AI systems...only able to learn what it is taught...and breaks down with slight differences such as ethnicity. @Nobody: this is a good example of an 'external noise variable' that would most certainly cause false positives. SO could exercise, or stretching/yoga.
@akarp, you could have just said you agree but with my 31 years of quality control experience, I will await your six sigma report.
@Nobody, after two trials of 110 and 500 they presumably haven't set their technique in stone. Adding data for more and more variables, including prostheses, race etc. will give their system more accuracy. Which any good researcher will have planned already.
But prosthetic joints are unlikely to generate the variability in gait described as the most predictive data point.