Can technology predict falls in older adults?
30th October 2017 · 0 Comments
By Randy Rieland
Contributing Writer
(Special from Next Avenue/New America Media) – The prospect of aging can conjure up a multitude of horrors — a mind stolen by dementia, a body debilitated by illness, a soul crushed by social isolation. For most, fear of falling would be well down the list.
But falls are, in fact, one of the more common and consequential risks older adults face. The statistics, compiled by the Centers for Disease Control and Prevention, are both eye opening and alarming.
One out of four Americans 65 or older falls at least once every year. Every 11 seconds, an older adult in the United States is treated in an emergency room for a fall; every 19 minutes, one dies from a fall. By 2020, the financial cost related to falls by older adults in the U.S. is expected to top $67 billion per year.
Determining Early Signs
It’s not surprising that an increasing amount of research is focusing on ways to predict if, and even when, a person is likely to fall. Much of that effort is built around using emerging technologies — from infrared depth sensors to brain imaging to virtual reality.
“Technology allows you to monitor people in their homes in a way you couldn’t have in the past,” said Marjorie Skubic, director of the University of Missouri’s Center for Eldercare and Rehabilitation Technology.
She’s been refining the use of sensors and motion-capture technology to study older adults in their homes, potentially helping them age in place. “We’ve found that once sensors have been in a place for three or four weeks, people completely forget about them. And that’s what we want — to capture their normal activity in their homes.”
Gait Watching
While Skubic’s research has focused broadly on how sensors can help detect early signs of physical and cognitive decline, a recent study zeroed in on finding a more precise correlation between a person’s walking gait and his or her likelihood of falling.
Using sensor measurements of walking speeds and stride length of residents at TigerPlace, a retirement community in Columbia, Mo., researchers found a clear connection between a slowing pace and the risk of falling.
In fact, analysis of the multi-terabyte-sized set of data, gathered over 10 years, showed that people whose gait slowed by five centimeters per second within a week had an 86 percent probability of falling during the next three weeks. That was four times more likely than someone whose walking speed hadn’t changed.
The researchers also determined that the shortening of a person’s stride indicated a possible fall in the near future, albeit not as clearly as decreasing speed.
When the sensor system detects notable changes in a person’s gait, it sends an alert to the caregiver so she or he can take steps to help prevent a fall.
Invasion of Privacy?
What about privacy concerns? Aren’t people anxious about having their every step recorded?
The key, says Skubic, is that the system reflects each person as only a silhouette, instead of a clear image captured by a conventional camera. “It’s not streaming video,” she noted. “You can’t tell what someone’s wearing or if their hair is made up. You just get this shape, but you can get a lot of information from that shape.”
At first, Skubic thought she might need to blur the silhouettes to have people feel more comfortable. It turned out that wasn’t necessary. Rather than worrying mainly about invasion of privacy, she found, “People said they could see that a crisp silhouette could be easier to interpret, in terms of looking for something related to fall risk. They were OK with that.”
Skubic noted several benefits of a sensor-based warning system. For starters, it does not require people to wear or interact directly with a device. She cited research that found older people were less likely to engage with technology if they weren’t feeling well — a time when captured data can be most helpful.
Perhaps more importantly, the sensor system enables monitoring to occur over time. “The big difference is that we’re looking at the average in-home gait speed,” Skubic said. “We and other researchers have found that your typical at-home gait pattern is different than if you’re in a lab and someone says, ‘Now walk across the room.’ People walk differently in their own homes.”
Another study found TigerPlace residents whose homes had sensors were able to age in place there 1.7 years longer than those in a control group without sensors.
A testament to Skubic’s belief in the sensor system was her decision to install one in the home of her aging parents, who live several states away in South Dakota.
“My mother is 93. My father just turned 96. And they want to stay in their own home,” she said. Skubic added, “I can see how it can help me help them. I installed it on my mother’s birthday in January. It was my birthday present to her.”
Brain Work
Researchers at the Albert Einstein College of Medicine in New York have taken a different approach in using technology to predict falls: They’re looking into people’s brains.
Specifically, they tracked the brain activity of a group of 166 high-functioning adults with an average age of 75 as they performed various activities — walking, talking and then walking while talking. They found that those who needed the front part of their brains to work harder while multi-tasking — reciting every other letter of the alphabet while walking — were more likely to have a fall in the next few years.
A lead researcher for the study, Joseph Verghese, MD, explained that cognitively impaired people tend to fall at a much higher rate than those with more normal cognition. The challenge was to see if there was a way to determine which people in the second group might have a higher risk of falling.
“When you look at them in the community, they’re walking around doing their activities without any impairment,” he said. “You really need to stress them to reveal the abnormality that would predict falls.”
The high-functioning people did slow down a bit when walking and talking, but that’s pretty typical, and Verghese said it wasn’t enough to help predict falls.
But, he continued, “When we measured their brain activity, it appeared they were trying to compensate really hard, using their brain function to maintain their physical performance.” He added, “It wasn’t something you could see. But you could measure it.” The brain activity was signaling its stress.
More Brain Activity, More Falls
Through follow-ups with test subjects every few months over the next four years, the researchers found that 71 of the 166 had fallen, some more than once. And those who had registered more brain activity while walking and talking were more likely to be in that group. In fact, each incremental increase in brain activity resulted in a 32 percent increased risk of falls.
The goal, said Verghese, is to be able to use this approach to detect if a person has a higher risk of falls before any physical signs appear. “Most of the research has been on identifying impairments that lead to falls. Less attention has been paid to abnormal biology or brain abnormalities that might do that,” he said.
“But what if you could step back to an earlier point in time and treat this as a biological syndrome that leads to clinical impairments like poor balance or worse gait, which then lead to falls? The idea would be to catch this early,” he said.
Verghese said the next phase of the research will look at the activity levels of other parts of the brain during the walk/talk test to see what role they may play in how people perform.
Randy Rieland wrote this article for the PBS Next Avenue website supported by a journalism fellowship from New America Media, Gerontological Society of America and AARP. He often writes about aging and technology for the Innovations blog on Smithsonian.com.
This article originally published in the October 30, 2017 print edition of The Louisiana Weekly newspaper.
