Skip to main content

Table 2 Effect of aggregating heart rate, step count, and sleep data, and imputation of wear time in 15-, 30-, and 60-min increments, to classify time wearing the activity tracker device

From: Physical activity measured using wearable activity tracking devices associated with gout flares

Imputation* Heart rate minutes % increase from imputation of wear time Step count minutes % increase from imputation of wear time Sleep minutes % increase from imputation of wear time Composite of any (heart rate, step count, sleep) minutes % increase wear time above heart rate minutes and after imputation of wear time**
Raw data, no imputation 4,182,919 Referent 828,780 Referent 1,155,986 Referent 4,292,447 2.6%; referent
15-min imputation* intervals 4,354,035 + 4.1% 2,342,250 + 182.96% 1,204,320 + 4.2% 4,472,685 2.7%; 4.2%
30-min imputation intervals 4,426,740 + 5.8% 2,877,690 + 247.2% 1,255,920 + 8.6% 4,552,950 2.9%; 6.1%
60-min imputation intervals 4,522,500 + 8.1% 3,424,920 + 313.3% 1,357,500 + 17.4% 4,657,020*** 3.0%; 8.5%
  1. Heart rate minutes = minutes with heart rate data; step count minutes = minutes with steps data; sleep minutes = minutes with sleep data
  2. *Imputation to the specific interval was performed if there was any value (heart rate, steps, or sleep) in 1 min was non-missing
  3. **The first number refers to the increase in wear time related to the use of the 3 data types vs. only step count data, shown in the first column of the same row; the second number refers to the gain in wear time related to imputation compared to the first row