Smartwatch-based bioimpedance analysis for body composition estimation: precision and agreement with a 4-compartment...

Area of Science:

• Biomedical Engineering
• Health Informatics
• Wearable Technology

Background:

• Smartwatches are increasingly used for health monitoring, offering accessible body composition estimations.
• Bioimpedance analysis (BIA) is a common method for assessing body composition.
• Evaluating the accuracy of novel BIA methods is crucial for reliable health tracking.

Purpose of the Study:

• To assess the precision and agreement of smartwatch-based bioimpedance analysis (SW-BIA) and multifrequency bioimpedance analysis (MFBIA).
• To compare SW-BIA and MFBIA against a criterion 4-compartment (4C) model for body composition assessment.
• To determine the accuracy of these methods for estimating total body water (TBW) and body fat.

Main Methods:

• 186 participants (114 females) underwent body composition assessments using a 4C model, SW-BIA, and MFBIA.
• Total body water (TBW) values from devices were compared with bioimpedance spectroscopy.
• Precision and agreement analyses were conducted between methods and the 4C model.

Main Results:

• Precision was adequate but slightly lower for SW-BIA compared to other methods.
• Neither SW-BIA nor MFBIA demonstrated equivalence with the 4C model for overall body composition.
• SW-BIA overestimated body fat, while MFBIA underestimated it; MFBIA showed equivalence for TBW, but SW-BIA did not.

Conclusions:

• Smartwatch-based body composition estimates may have significant errors compared to clinical measures.
• While SW-BIA and MFBIA are not equivalent to the 4C model, MFBIA showed better agreement for TBW.
• Further research is needed to validate smartwatch accuracy for monitoring body composition changes over time, especially considering higher errors in males.