Skip to Main Content

An official website of the United States government

Principal Investigator
Name
Dane Van Domelen
Degrees
Ph.D.
Institution
Johns Hopkins Bloomberg School of Public Health
Position Title
Postdoctoral fellow
Email
About this CDAS Project
Study
IDATA (Learn more about this study)
Project ID
IDATA-30
Initial CDAS Request Approval
Apr 8, 2019
Title
Measurement error methods for wearable device data, treating daily measurements as imprecise draws from long-term average
Summary
Standard practice for analyzing wearable-device data is to calculate variables for each day of monitoring, average over monitoring days, and conduct inference on the averages. This approach fails to account for device-related measurement error (e.g. measured steps ≠ true number of steps) and sampling-related measurement error (e.g. 7-day average ≠ long-term average). In general, ignoring these sources of error will lead to biased estimation of associations between signal-derived variables and health outcomes.

We propose using formal measurement error methods to analyze wearable device data. The data structure lends itself to maximum likelihood estimation, treating daily measurements as (perhaps correlated) replicates from each subject’s long-term average. For example, one might assume that the device-recorded sedentary time on a particular day is not exactly equal to the true sedentary time, but rather the true sedentary time plus a mean-0 error. Further, the true sedentary time on a particular day might be assumed to reflect the long-term average plus a mean-0 random error. Alternatively, considering that activity variables are usually right-skewed and almost always non-negative, it might be more reasonable to assume multiplicative mean-1 errors, perhaps distributed lognormal or Gamma. Within a maximum likelihood framework, metrics like AIC or BIC can be helpful for specifying the error structure and other parametric assumptions.
Aims

We plan to focus primarily on cross-sectional and cohort study designs, where the goal is to estimate a covariate-adjusted association relating an activity variable to a health outcome (e.g. BMI, waist circumference, cardiometabolic variables). We will compare point and interval estimates for the naïve approach to those obtained by fitting various maximum likelihood estimators, and compare metrics including AIC and BIC. We plan to run simulation studies mimicking IDATA to evaluate the consequences of ignoring measurement error and to evaluate performance of the corrective methods (e.g. validity, numerical stability, robustness to model misspecification). We may also consider measurement error methods other than maximum likelihood.

Collaborators

Vadim Zipunnikov, PhD (Johns Hopkins Bloomberg School of Public Health)