Accuracy of step count measured by physical activity monitors: The effect of gait speed and anatomical placement site

Comparison of Physical Activity Measures Derived From the Fitbit Flex and the ActiGraph GT3X+ in an Employee Population With Chronic Knee Symptoms

“…One potential confounder includes the wrist versus waist location of the monitoring devices during data collection, although this arrangement is consistent with that of similar studies that compared data from these devices (2)(3)(4)10,(13)(14)(15)(16). The study could not control for wear time of the consumer devices, and this might have impacted results.…”

Accuracy of wearable physical activity trackers in people with Parkinson’s disease

“…Accuracy of wearable activity monitors can be influenced by pattern and speed of gait, and anatomical placement of the device 11 . Waist-worn activity monitors have been shown to be more accurate than wrist-worn devices at counting steps at gait speeds ranging from slow to fast 11 . The accuracy of wrist-worn devices may be further reduced in people with PD due to reduced arm swing often experienced during gait in this population.…”

Adjusting Insulin Delivery to Activity (AIDA) clinical trial: Effects of activity‐based insulin profiles on glucose control in children with type 1 diabetes

“…To assess the underlying cause of the hypoglycemia, CGM data need to be combined with contextual insulin delivery and activity data. Common activity monitors can provide detailed information about activity, but it is not clear that the step count or heart rate data obtained from these monitors are sufficiently reliable to effect insulin dosing adjustments …”

“…Common activity monitors can provide detailed information about activity, but it is not clear that the step count or heart rate data obtained from these monitors are sufficiently reliable to effect insulin dosing adjustments. 20,21 The aims of the Adjusting Insulin Delivery to Activity (AIDA) study were therefore to determine whether the daytime step count obtained from a Fitbit activity monitor could be combined with nighttime CGM data to identify children at risk for nighttime hypoglycemia following high activity days and if so, whether appropriate basal insulin profiles for use on nights following low and high activity days could be identified from an automated weekly analysis of these data.…”

“…significant relationship between daytime step count and nighttime nadir glucose was detected in 10 of 18 subjects (P < 0.05 all; example daily profiles and regression analysis for subjects with and without activity effect providedFigure 1). Median time to detect an activity effect was 21[18][19][20][21][22][23][24][25][26][27][28][29][30][31] days (19 days for subject shownFigure 1). For subjects with activity effect (example subject shownFigure 2), algorithm-derived changes in nighttime basal rates ranged from −15.3% (net decrease in insulin delivery following high activity days) to +16.7% (net increase in insulin following low activity days).…”