VitaBit Score - explanation

General description of VitaBit Score:

VitaBit Score displays a number between 0 (= very unhealthy, too much sitting, no compensation) and 100 (= perfectly healthy sit – stand – walk – pattern) depending how healthy the user is performing at this moment of the day.

The score was developed by our health psychologist, involving scientific information about daily sitting patterns in relation to energy expenditure and health risks (e.g., Diabetes Type II, Lower Back Pain).

Foto_20-06-17_14_04_10.png



Background information

We know that employees with sitting occupations sit on average 11 hours per work day, which cause health issues both the long run as well short term. Therefore, our goal is to help people to change their activity patterns by guiding them through the basic behavioral change process (monitoring the behavior, setting goals, achieving goals/overcoming hurdles, building the habit).

The time of overall sitting per se, is actually not the factor that matters the most for our health. Long sitting periods should be frequently interrupted by healthier options such as standing or walking.  Additionally, the length of single sitting periods should not be too long. Combining all relevant factors (you can find the list below) by a mathematic algorithm, our VitaBit Score indicates the current calculated healthiness of a sit-stand-walk pattern having been performed up to this moment of a day.

Target

“Monitoring” and “building healthy habits” are the main steps of the behavioral change process being targeted by the VitaBit Score.

If a user sees (monitors) a very low score (e.g., 25), he will know, that he is not well performing and, thus, will do more steps or interrupt his sitting behavior more often throughout the day.

People will feel "rewarded" by high scores and will continue performing similarly. Healthy habits can be established when the user continuously behaves in this healthy manner.

Technical information

We use real time analysis of the sit/stand/walk data, were we make use machine learning algorithms to output the VitaBitScore to the user.

The steps to generate the VitaBitScore were

  1. Calculating the VitaBitScores of activity data (daily sit-stand-walk patterns) by the help of an own developed scoring procedure
  2. Delivering input data (activity data) and output data (corresponding calculated VitaBitScores) as well as important parameters to be considered (again, see features-list below) to the machine learning program.
  3. The weights being returned are used as VitaBitScore calculation on the VitaBit application on the user device

Features (x - values):

ScaledNrOfStepsBucket

Steps scaled to 16h of use. If, for example, there was 2 hour of tracked data, the number of steps is multiplied by 8. (16/2)

ScaledNrOfSedentaryBoutsLongerThanThirtyMinutes

For each continuous sitting time, calculate the number of half hours and round this value down.

This value is the sum of the number of half hours bouts from all continues sitting times. This value is scaled in a similar way as the steps bucket only scaled to 8 hours instead of 16h

AverageSedentaryBoutDuration

Average duration of continues sitting times where the minimal time is 30 minutes. (value can never be lower than 30)

SedentaryBoutCorrection

The longer the sitting periods, the worse the score should get. Sometimes, longer periods result in less periods resulting in a lower score because less of the longer periods fit in a day. This variable considers the variable AverageSedentaryBoutDuration.

Have more questions? Submit a request

0 Comments

Please sign in to leave a comment.