Using symbolic aggregate approximation (SAX) to visualize activity transitions among older adults.

摘要: The objective of this analysis was to apply symbolic aggregate approximation (SAX) time-series accelerometer data for activity pattern visualization stratified by self-reported mobility difficulty. A total 2393 (71.6  ±  7.9 years old) participants wore an on the hip (4  +  d; 10  +  h) during national health and nutrition examination survey (NHANES), a biannual series assessments US population. One minute epoch used perform SAX, which converted accelerometry time into four levels. Intelligent icons normalized transition prevalence, visual representation data, were examined among those who Mobility difficulty questions assessed various levels performing activities such as walking quarter mile, up ten steps, stooping/crouching/kneeling, between rooms same floor. Daily counts estimated across level using weighted-linear regression after adjusted demographics, lifestyle factors, medical conditions, wear time. Those reporting higher tended be older, female, less educated, not married, smokers. Additionally, with lower education, income, moderate-to-vigorous physical activity, history adverse conditions. Using SAX-derived intelligent icons, no showed high variations transitions all With difficulty, in constricted around low transitions. Among reported unable/don't do mobility-related activities, there only tiered prevalence prolonged activity. Adjusted estimates daily occurred but significant (p  <  0.01). In summary, apparent differences patterns even when traditional analyses volumetric yielded null results.