In-home Sensors Spot Dementia Signs In Elderly

The study results, presented last week at the 10th International Conference on Alzheimer's Disease and Related Disorders in Madrid, show that continuous, unobtrusive monitoring of in-home activity may be a reliable way of assessing changes in motor behaviors that may occur along with changes in memory.

"To see a trend over time, you need multiple measures - good days and bad days - and it often takes years to see that trend in a clinic setting," said Tamara Hayes, Ph.D., assistant professor of biomedical engineering at OHSU's OGI School of Science & Engineering, and the study's lead author. She noted that most clinic visits by elders are spaced over months or even years, and their memory and motor skills performances are evaluated in a small number of tests completed in a limited amount of time.

"In contrast, we're looking continuously at elders' activity in their own homes," Hayes said. "Since we're measuring a person's activity many times over a short period, we can understand their normal variability and identify trends. If there's a change over a period, you can see it quickly. "

Mild cognitive impairment is a known risk factor for dementia, a neurological disorder most commonly caused by Alzheimer's disease. Changes in clinical measures of activity, such as walking and finger-tapping speeds, have been shown to occur at about the same time as memory changes leading to dementia. By detecting subtle activity changes over time in the natural setting of an elder's home, researchers hope to more effectively identify when elders are starting to have trouble.

Hayes and colleagues in the OHSU Oregon Center for Aging & Technology - ORCATECH - used an unobtrusive "activity assessment system" to continuously measure in-home walking speed and overall activity of seven healthy elders and seven elders with mild cognitive impairment. The sensors, which remained in the elders' apartments at a Portland retirement community for up to 60 consecutive weeks, picked up more than 108,000 hours of activity data.

Each system included wall-mounted motion sensors placed in every room, door-mounted magnetic contact sensors, and wireless transceivers that sent the data to a computer in the home. The scientists then determined the number of times the sensors fired each minute during 24-hour blocks of time to create an overall activity level score for each participant.

Researchers found "the impaired subjects had more variability in their walking times than the healthy elderly group," and this variability was greater in the afternoon than in the morning, according to the study. This is consistent with earlier studies suggesting that variability in motor measures may predict later-onset dementia.

In addition, the elders with mild cognitive impairment were more variable in their activity during the day compared to their healthy counterparts. This inconsistency was "the most striking difference between the groups. Even with as few subjects as we had, the groups were clearly separated by the variability of these measures," Hayes noted. Differences were detectable after only four weeks of monitoring.

Such differences often are less obvious, if not imperceptible, during a typical clinical examination. In fact, many elders' desire to perform well during doctor visits make them walk faster than their normal daily paces, masking changes that may be clinically relevant, researchers say.

Hayes and her colleagues predict that as healthy elders begin to develop cognitive impairment, they become more variable day to day or even hour to hour, and that "this variability may provide an early marker that can be used to predict the later onset of dementia in a single individual." They have already begun a larger study to test that theory.

Much of the hardware used in the study is simple, widely available and inexpensive. Plus, it's getting smaller and smaller, so it can be deployed in elders' homes much less obtrusively. However, "the thing a lot of people don't realize is that although the hardware appears to be the hard part, the data management is the hard part," Hayes said. "Really, the bulk of our technology research is focused on algorithm development" and the ability to mine and manage the data these devices collect.

Study co-author Misha Pavel, Ph.D., professor of biomedical engineering and computer science and electrical engineering at OHSU's OGI School of Science & Engineering, says studies on systems and technology such as those being developed by ORCATECH are "a critical precursor of changing reactive health care to proactive care for individuals."

A more proactive health care approach may be essential to managing the rising cost of health care for aging baby boomers. And "being able to stay at home and maintain independence represents a significant improvement in quality of life of many elders who like to remain as long as possible in their familiar environments," Pavel noted.

The ability to continuously and discretely collect activity data over extended periods of time has given researchers a new, more accurate way of "identifying clinical changes that are difficult to get your arms around in the typically designed practice we have nowadays," said ORCATECH's director, Jeffrey Kaye, M.D., OHSU professor of neurology and biomedical engineering.

ORCATECH is an interdisciplinary center at OHSU established in 2004 as a National Institute on Aging Roybal Center for Aging & Technology to study and develop technology that assists in the assessment and remediation of elders in their home environments. Its goal is to help elders retain independence by discretely collecting data that may indicate health changes long before quality of life is affected.

The study was funded by grants from the National Institute on Aging and Intel Corp.