Evaluation in Children of Novel Bluetooth Bioimpedance Analysis Devices Designed for mhealth Applications

Body fat and other body composition measurements are of increasing interest in the clinical setting and to the general public. Excess adiposity in children is a growing concern with the now well‐established relationship between early onset obesity and later adult health. The aim of the current study was to evaluate in children (ages 10–17 yrs) three small portable wireless contact‐electrode bioimpedance analysis (BIA) systems designed for personal body composition analysis. Three hand‐held multifrequency devices were evaluated that included four electrode configurations (palm‐to‐palm [PP], two finger‐to‐finger [FF], and palm‐to‐knee (PK)). Impedance (Z) values derived by the devices at the Pennington Biomedical Research Center were evaluated against corresponding electrical pathways measured using a gel‐electrode reference multifrequency BIA system (InBodyS10 [IB], Seoul, South Korea) in 110 (53 M, 57 F) healthy white (34 M, 35F), African American (12 M, 17 F), and other (7 M, 3 F) children ranging widely in body mass index. Additionally, the main traditional lean soft tissue mass (LST) and fat‐predictor variable height 2 /Z was examined relative to LST evaluated with dual‐energy X‐ray absorptiometry (DXA; GE Lunar Prodigy, Madison, WI). There were strong correlations between the four prototype pathways at 50 KHz and the corresponding IB arm‐arm electrical pathway (R 2 , 0.90, 0.90, 0.96, and 0.91; p<0.0001, Table 1). Similarly, height 2 /Z (cm 2 /Ω) versus DXA LST (g) regions for the four‐prototype pathways yielded correlation coefficients (R 2 0.88, 0.88, 0.92, and 0.90) that were similar to the reference InBody system for the PP, both FF, and PK pathways (all p<0.001, Table 2). The novel evaluated BIA systems as small form factor mobile devices thus show promise as a means of quantifying a child's body composition. Support or Funding Information Funded by Samsung Research America 1Correlation (R 2 ) LST IB FF1 PP FF2 PKLSTInBody 0.91FF1 0.88 0.90PP 0.88 0.91 0.90FF2 0.92 0.96 0.94 0.92PK 0.90 0.91 0.89 0.862 Summary of Regression ModelsAnalysis Equation R 2 PIB vs. FF1 y = 1.14× + 112.0 R 2 = 0.90 P<0.0001 IB vs. PP y = 1.10× + 96.5 R 2 = 0.91 P<0.0001 IB vs. FF2 y = 1.09× + 71.0 R 2 = 0.96 P<0.0001 IB vs. PK y = 0.0016× − 1.5 R 2 = 0.90 P<0.0001 Ht 2 /Z IB vs. DXA LST y = 0.0012× − 2.8 R 2 = 0.91 P<0.0001 Ht 2 /Z FF1 vs. DXA LST y = 0.0008× + 1.2 R 2 = 0.88 P<0.0001 Ht 2 /Z PP vs. DXA LST y = 0.0009× + 1.4 R 2 = 0.88 P<0.0001 Ht 2 /Z FF2 vs. DXA LST y = 0.0010× − 0.4 R 2 = 0.92 P<0.0001 Ht 2 /Z PK vs. DXA LST y = 0.0016× − 1.5 R 2 = 0.90 P<0.0001Units: Ht, cm 2 ; Z, ohms; LST, g