The data collection ran between a strictly enforced timeframe.
To ensure consistency across subjects, reliability in results and prevent the test subjects undergoing any unnecessary excess stress, the validation was conducted under controlled conditions.
During the study, each subject was required to complete an hourly activity log with a record of any physical activity as well as beginning and end times of each meal.
The controlled exercise test for each subject was completed at a set time to maintain consistency between subjects.
Each subject completed a 10 minute cycle within an Aerobic Heart Rate Zone 2 to ensure energy demands were kept consistent relative to each subject.
This portion of the protocol was designed to validate the HeartKey Energy Expenditure algorithm during times of increased physical activity and thus increased heart rate (HR).
Each subject’s HR was monitored on the HeartKey app throughout the protocol by the test conductors.
The study validated the performance of the HeartKey Energy Expenditure algorithm against an energy expenditure algorithm of a commercially available device.
Comparing the performance of HeartKey algorithm and the commercial device, the importance of physiological metadata must be considered.
The commercial device showed that a a female test subject who was 30cm shorter and 40kg lighter than a male subject had a total energy expenditure of 3216kcal over 24 hours compared to a reported 1868kcals for the large individual . Conversely, the HeartKey algorithm reported 2236kcal and 2720kcal for the smaller and larger individuals, respectively.
Keeping in mind that test conditions including exercise were consistent across subjects it can be confidently stated that the commercial device is worryingly inaccurate across these two subjects when compared to the more realistic values the Heartkey algorithm reported.
Physiologically, the significantly smaller subject would almost always expend less calories within a controlled environment. Particularly when the magnitude of physiological difference is significant across gender, height and weight.
We were delighted to see the reliable performance of the HeartKey Energy Expenditure algorithm is further reinforced when considering BMR as a reliable basal rate of energy expenditure.