Home-Based Activity Changes Associated with MCI
Jeffrey A. Kaye, Hiroko H. Dodge, Nora Mattek, Daniel Austin, Stuart Hagler,
Teresa Buracchio, Michael Pavel, Tamara Hayes
Oregon Health & Science University
Support: NIH AG08017; AG024059; AG024978; AG 023014; K01AG023014
1. Kaye JA, Hayes TL, Zitzelberger TA, et al. Deploying wide-scale in-home assessment technology.
Technology and Aging, A. Mihailidis, J. Boger, H. Kautz, and L. Normie, Eds., IOS Press, 2008, 21:19-
2. Hayes T, Abendroth F, Adami A, et al. Unobtrusive assessment of activity patterns associated with mild
cognitive impairment. Alzheimer's & Dementia 4(6): 395-405, 2008
3. Hagler S, Austin D, Hayes TL, Kaye J, Pavel M. Unobtrusive and Ubiquitous In-Home Monitoring: A
Methodology for Continuous Assessment of Gait Velocity in Elders. IEEE Transactions on Biomedical
To determine if variability in motor function assessed in the
home environment characterizes persons with MCI.
Figure 2: Continuous home acquired walking speed is significantly
different (p < .01) in naMCI (non-amnestic MCI) compared to intact
subjects. Single stop-watch measure does not distinguish groups.
Changes in motor function precede cognitive decline up to a
decade before symptoms appear, a conclusion primarily
derived from brief clinical measures of motor function such as
walking speed. We hypothesized that if these measures were
predictive of MCI, before they declined in absolute
magnitude, there would be a period where the measure
would first show increased variability.
In 113 non-demented ISAAC (Intelligent Systems for
Assessing Aging Change) cohort (Kaye, 2008) seniors living
independently (mean age 84; CDR ≤ 0.5) multiple daily
walking episodes were unobtrusively recorded as subjects
traversed a line of passive infra-red motion sensors placed
strategically in their home (figure 1) for a mean of 319 ±
127 days. Daily walking speeds (Hayes, 2008; Hagler,
2009) and the variance in these measures over time were
calculated and compared to conventional single visit stop-
watch derived speed recordings in subjects with and without
MCI. Trajectory analysis using the coefficient of variation
(COV) in weekly walking speeds was applied to assess
differences in variability over time among subjects with and
Figure 1: [A] A home sensor line in place; [B] Schematic of a person
walking through a sensor line containing four sensors with their fields of
view shown. Sources: [A] Julie Keefe, the New Times, Nov 7,
2009; [B] Hagler et al. 2009.
•Continuous unobtrusive home monitoring may
identify activity changes (walking speed and
variability) that are early markers of cognitive decline.
•Home-based continuous assessment metrics for
discerning subtle early change may provide new
measures of early change not currently accessible
through conventional methodologies
Stopwatch Derived Walking Speed
Mean of Daily Median In-home Walking Speed
Figure 3: Three trajectories best described walking speed COV. MCI
subjects were more likely to be in the high variability at baseline and
increasing over time group (Blue trajectory; OR = 1.41; p = 0.012).
Disclosure: Drs. Kaye, Pavel and Hayes have received support from Intel for their
research; Dr. Hayes holds stock and/or stock options in Intel.