Why the need for multiple body composition methods?

For users with a certain body type, our regular body composition algorithm may not provide the most accurate body fat percentage measurement. This is especially common for users that are very lean and more on the athletic side of the body fat percentage scale. See the section below

Available body composition methods

Prism currently provides three body composition methods listed below.

Recommendations on how to integrate the different body composition methods

Option A: single method (most common)

Choose coco as the sole body composition method. This option is recommended for most partners, especially if your userbase doesn’t include a significant amount of athletic or very lean users.

Option B: combination of Regular and Athlete modes

If your userbase is composed by a significant number of users in the athlete/very lean category or you would like to include an Athlete Mode, choose a combination of coco and extended_navy_thinboost, and name them Regular Mode and Athlete Mode, respectively. To integrate this combination of methods, you typically have two ways of implementing it:

1. Give your users the option to choose the method they want to use. Key items to consider in this implementation:
   1. One option is to set the Regular Mode as the default method on the first scan and expose the Athlete Mode option: (you want to avoid the pitfall of having many users toggling this mode on because they just want to see on average lower values):
      1. on a secondary or tertiary screen like the profile screen, or
      2. guiding the users on finding Athlete mode as an option in case they are not happy with the default body composition method value. For example, you can expose a tooltip along the lines in the results page: “Expecting a lower body fat measurement? You might want to try our Athlete Mode.”
   2. Another option is to ask the question (”Athlete” vs “Regular Mode”) directly to the user on first scan or first account setup. This option is mostly recommended if your user base has a significant amount of users in the athlete mode category. The cons are that more “Regular Mode” people will likely feel tempted to switch to “Athlete Mode”.
   3. For both options, always educate your users on what are the criteria to decide for choosing Athlete Mode (see points below).
2. Have internal logic that selects what method to enable for the user based on proxy signals (user inputs or properties) such as (without explicitly asking the user about Athlete Mode):
   1. How often does the user work out? More than 8 hours a week, they might typically want to use Athlete mode.
   2. Have they done a DXA (or a similar medical) scan and/or are they under 15% body fat percentage for males and 20% for females?
   3. Is the resting heart rate under 55 bpm?

Additional considerations for Athlete Mode

• Athletic and very lean users generally have a different body type which requires a specific body fat mass calculation algorithm and that is why we have a specific method for these users. This method can still be used even if the users are no longer part of this category, however, please note that the accuracy vs DXA will be lower when compared to Regular Mode, especially as the body fat % increases (>30% body fat).
• When Athlete mode is enabled, users on average should expect lower body fat percentages and higher lean mass measurements because the user's athletic body type is taken into account.
• All scans prior to the updates that adds more body composition methods will all stay with the legacy method tina_fit.