Subtle changes in driving habits, computer usage and medication routines could yield early clues to the development of Alzheimer's disease.

A researcher at the Minneapolis VA Medical Center is studying all three by using sensors to monitor volunteers and changes in their daily activities.

While Alzheimer's is an incurable brain disorder, early detection could allow people to receive support services or therapies to slow its onset, said Adriana Seelye, a VA neuropsychologist leading the research.

"These kind of subtle cues are not picked up early on when interventions could be put in place," she said. "A lot of times, people don't come to our attention until there is a crisis."

Seelye's prior research showed that changes in computer usage and mouse movements could predict mild cognitive impairment (MCI), a precursor to Alzheimer's. Another study found seniors with MCI were more conservative and predictable drivers.

Now she wants to assess which behavior changes are most predictive, or whether combinations of changes offer the most accurate clues. She received money from the VA for one study, and this week won a grant from the National Institutes of Health for a second. She is recruiting 130 senior volunteers who don't have dementia. Volunteers in the NIH study will be tracked for four years.

Wrist-worn fitness trackers will measure sleep and movement. In-car computer data will reveal changes over time in average speed, highway usage and right vs. left turn decisions.

Other researchers have examined changes in speech and voice, among other characteristics. Early detection methods are needed, Seelye said. The number of Minnesotans with Alzheimer's could grow from 91,000 in 2016 to 120,000 in 2025, according to the Alzheimer's Association.

Alzheimer's symptoms "develop very slowly," Seelye said. "This makes it very difficult for us as clinicians."