Back

Age-related impacts of dementia status and subtype on accelerometry measures in older adults from the National Health and Aging Trends Study

Brown, K. J.; Shutes-David, A.; Wilson, K.; Payne, S.; Jankowski, A.; Seto, E.; Tsuang, D.

2024-05-09 epidemiology
10.1101/2024.05.09.24307083 medRxiv
Show abstract

It is likely that age, dementia status, and dementia subtype all contribute to differences in activity levels among older adults and that accelerometer devices can help us better understand these relationships. To that end, this study analyzes accelerometry data from individuals with dementia (n=104) and individuals without dementia (n=634) who participated in the National Health and Aging Trends Study (NHATS), as well as from individuals with dementia with Lewy bodies (DLB; n=12) and Alzheimers disease (AD; n=8) who participated in a small pilot study. In the NHATS cohort, individuals with dementia had lower daily activity counts (t = -5.449, p < 0.001), a shorter active bout length (t = -5.283, p < 0.001), and a longer resting bout length (t = 2.210, p = 0.032) at the mean age of 79 than individuals without dementia at that same age. Differences in data collection and processing prevented direct comparisons between the cohorts, but parallel analyses revealed no statistically significant differences between AD and DLB across these three measures in the smaller cohort. Studies with larger samples of subtyped individuals with dementia will be necessary to detect clinically meaningful differences, but the patterns observed in the small cohort suggest that individuals with DLB are less active, exhibit a shorter active bout length, and exhibit a longer resting bout length than individuals with AD. Taken together, these findings highlight the potential of accelerometers to gather objective activity data that could be used to aid in the early identification of dementia and, potentially, the differentiation of dementia subtypes.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.