How we spend the 24-hour day - our activities while we are awake and the nature of sleep when we set out to rest - are linked to trajectories of health and wellbeing among older adults. However, there are big gaps in our knowledge about variation in sleep and physical activity within and between countries due to the complexity of measuring them. The HRS family of studies - like most general cohort studies - have measured physical activity and sleep with short sets of survey questions. Studies that have compared questions like these to objective measures find poor agreement, low correlations and systematic biases, all of which would undermine the utility of the data. To some extent the problems with survey questions are common across these two domains: they call on individuals to recall, categorize, and summarize complicated and variable behaviors that they do not ordinarily think about in quantitative terms. The number of objective ways to measure sleep and physical activity have increased in recent years, but the gold standards for each are lab-based, expensive and impractical in a national cohort study: doubly labeled water to measure energy expenditure and polysomnography to measure sleep. There have been rapid technologic advances in actigraphs, personal monitors that measure movement and use the pattern and intensity of movement to characterize sleep (when worn on the wrist) and physical activity (typically worn at the waist). Their practicality has been demonstrated in population-based studies using omnidirectional accelerometers. Newer devices include triaxial accelerometers, which record and report movements in three dimensions and are likely to be better for characterizing activity when awake. Actigraph data are downloaded and analyzed offline, but there is no standard for deriving metrics about physical activity and sleep from triaxial accelerometers. Manufacturers' algorithms are often opaque to the researcher and may be based on limited validation. There is a need for developing valid metrics from the rich data recorded without depending on manufacturer software. The overall aims of this project are (1) to develop a protocol and analytic programs to measure physical activity and sleep in the HRS family of studies using a single triaxial accelerometer worn at the wrist and (2) to identify a setof survey questions that will acceptably distinguish categories of sleep and physical activity when actigraphy is infeasible.
Our aims will be accomplished through three overlapping phases of pilot studies, to be carried out in the U.S. and Israel (in Hebrew, Arabic, Russian) through the Cohen Institute for Public Opinion Research, a SHARE site. The pilots will (1) develop a bank of candidate questions and cognitively test questions in four languages, (2) compare data from omnidirectional and triaxial accelerometers concurrently worn at the wrist and waist to develop metrics for the triaxial wrist instrument, and (3) compare survey responses to candidate questions with metrics derived from 7 days of wearing a triaxial wrist actigraph, in a purposive sample of 500 older adults, collected in Israel and the U.S.
The quality and quantity of physical activity and sleep are important to health at older ages, but it is very difficult to collect valid data about them in surveys. Here we propose a series of pilot studies in the US and in Israel to develop a better way to collect physical activity and sleep data using an objective measure, a wrist- worn triaxial accelerometer, and we will also test and validate new questions to use when actigraphy is not feasible. Collecting valid and objective data about physical activity and sleep in the HRS family of studies would permit comparisons of how these key health behaviors vary within and between countries and how they relate over time to health and well-being.
| Kocherginsky, Masha; Huisingh-Scheetz, Megan; Dale, William et al. (2017) Measuring Physical Activity with Hip Accelerometry among U.S. Older Adults: How Many Days Are Enough? PLoS One 12:e0170082 |
