Background
Multi‐frequency bioelectrical impedance analysis (MF‐BIA) offers enhanced body composition outcomes in a time‐efficient manner. The accuracy of stand‐up MF‐BIA compared against a four‐compartment (4C) criterion lacks evidence.
Objectives
To validate a stand‐up MF‐BIA compared to a 4C criterion for fat mass (FM), fat‐free mass (FFM) and body fat percentage (%fat).
Subjects/Methods
Eighty‐two healthy (32% men) normal‐weight (BMI: 18.5–24.9 kg/m2) young adults were measured for body composition determined from a stand‐up MF‐BIA and 4C model. Validity statistics included total error (TE) and standard error of the estimate (SEE) to examine prediction error between methods.
Results
For the total sample, prediction error was the highest for %fat (TE = 4.2%; SEE = 3.9%) followed by FM (TE = 2.4 kg; SEE = 2.2 kg) and FFM (TE = 2.4 kg; SEE = 2.2 kg). In men, %fat (TE = 2.5%; SEE = 2.2%) and FM (TE = 1.9 kg; SEE = 1.6 kg) were ideal; FFM was similar to FM (TE = 1.9 kg; SEE = 1.6 kg). In women, %fat (TE = 4.7%; SEE = 4.4%) ranged from good to fairly good, and FM was very good to excellent (TE = 2.6 kg; SEE = 2.4 kg); FFM was similar to FM (TE = 2.6 kg; SEE = 2.3 kg).
Conclusions
Stand‐up MF‐BIA may overestimate %fat and FM, and underestimate FFM compared to a 4C model. FM and FFM estimates from MF‐BIA demonstrate good agreement to a 4C model and may be a practical measure of body composition in normal‐weight adults. The highest error was seen in %fat for both sexes, with greater error in women.