New Predictive Equations for Resting Energy Expenditure in Normal to Overweight and Obese Population

Table 9

Accuracy of the new equations vs resting metabolic rate measured by indirect calorimetry and that calculated by some predictive equations in population used for external validation (n = 48; men: 50%).

Tools	RMR (kcal/day)	Bias (kcal)	Percent of bias (%)	Underestimation (%)	Overestimation (%)	Accurate estimation (%)	r
Measured RMR	1737.52 ± 378.49	—	—	—	—	—	—
AA_1	1770.15 ± 296.80	32.63 ± 219.84	2.00 ± 11.72	14.58	25.00	60.42	0.815
AA_FFM	1770.94 ± 313.25	33.41 ± 219.60	1.90 ± 11.49	14.58	22.92	62.50	0.815
HB 1919	1704.65 ± 292.18	32.87 ± 227.50	1.81 ± 12.62	16.67	22.92	60.42	0.800
Bernstein	1362.08 ± 218.08	375.44 ± 248.63	27.32 ± 17.12	87.50	0.00	12.50	0.781
Owen	1537.92 ± 263.66	199.60 ± 224.21	12.86 ± 13.29	58.33	4.17	37.50	0.814
Mifflin	1597.84 ± 256.53	139.68 ± 229.47	8.51 ± 13.34	35.42	4.17	60.42	0.805

AA_1 and AA_FMM are the newly developed equations. Bias = RMR measured by Quark RMR-REE predicted by equations; percent of bias = [(RMR measured by Quark RMR-REE predicted by equations) ∗ 100]/REE predicted by equations; accurate estimation = percentage of all subjects whose REE was within 90% to 110% of measured RMR by Quark RMR; underestimation = percentage of all subjects whose REE was less than 90% of measured RMR by Quark RMR; overestimation = percentage of all subjects whose REE was more than 110% of measured RMR by Quark RMR; r, Pearson’s correlation coefficient between RMR measured by Quark RMR and REE predicted by equations.Significant ().