Table 5.

Accuracy of predictive equations for resting energy estimation compared to Resting Energy Expenditure by indirect calorimetry in the validation group (n: 382).

Method for energy estimation	Absolute bias (kcal)	Relative bias (%)	p-value	RMSE	Under<90%	Accurate 90-110%	Over>110%
Predictive equations based on anthropometric measurements
New equation	10 ± 221	0.6	0.379	221	18	62	20
Harris-Benedict et al. (34)	-32 ± 219	-1.9	0.001	221	22	60	18
Mifflin-St Jeor et al. (35)	-145 ± 221	-8.5	0.001	264	39	54	7
FAO/WHO/UNU (33)	20 ± 221	1.2	0.073	222	16	55	29
Pavlidou (15)	-47 ± 221	-2.8	0.001	226	24	59	17
Schofield (36)	-64 ± 221	-3.7	0.001	230	25	60	15
Luhrmann et al. (3)	71 ± 223	4.1	0.001	234	10	53	37
Fredrix et al. (37)	92 ± 221	5.4	0.001	239	9	51	40
Predictive equations based on body composition measurements
New equation	12 ± 221	0.7	0.280	221	16	63	21
Cunningham (38)	-21 ± 225	-1.2	0.065	226	23	57	21
Mifflin-St Jeor et al. (35)	-232 ± 222	-13.6	0.001	321	60	37	3
Korth (39)	-13 ± 233	-0.8	0.263	233	22	55	23
Owen (40)	-249 ± 229	-14.6	0.001	338	63	35	3
Weigle (41)	-363 ± 223	-21.2	0.001	425	85	14	1
Batista (14)	149 ± 328	8.7	0.001	360	14	40	46

p-value: Paired t-test comparing predictive equations with the reference method (indirect calorimetry).

RMSE, Root mean square error.