Table 3.

Comparison of BAI in different trials.

Study	Sample	Age (Years)	Device	Agreement between Measurement Methods/Bias	Main Finding
Present study	903 apparently healthy and sub-sample with overweight/obese	Mean age 21.4 ± 3.3	Tetrapolar frequency	Bland–Altman plots Male bias 9.1%, Female bias 3.2%, Total bias 6.0%	Overall, BAI overestimating BF%, in overweight subjects the BAI overestimated BF%, and obese group the BAI underestimated BF% both genders.
Geliebter et al. [10]	19 pre-bariatric surgery clinically severe obese, non-diabetic females	Mean age 32.6 ± 7.7	Tetrapolar frequency	Bland–Altman plots Bias 2.2%	BAI underestimating BF%
Bernhard et al. [11]	240 patients with severe obesity	Mean age 44.1 ± 11.1	A single-frequency	Intraclass correlation 0.74; 95% confidence interval = 0.68–0.79	The two methods were similar according to the intraclass correlation
Ezeukwu et al. [12]	30 obese females	Mean age 22.8 ± 3.3	A single-frequency	Bland–Altman plots Bias 15.0%	BAI underestimating BF%
Lemacks et al. [36]	187 overweight/obese postmenopausal females	Mean age 55.8 ± 3.3	Dual-energy X-ray	Concordance correlation coefficient ρc = 0.39	Poor agreement strength between Dual-energy X-ray (DEXA) BF% and BAI overestimating BF%
Vinknes et al. [35]	5193 middle-aged (47–49 years) and elderly (71–74 years) males and females	Mean range 47–72	Dual-energy X-ray	Bland–Altman plots Bias in subjects with lower BF% 6.0%, Bias in subjects with higher BF% 1.9%	BAI overestimated adiposity in subjects with lower BF% (particularly in males) and underestimated it in overweight and obese subjects