Skip to main content
. 2024 Mar 19;119(5):1111–1121. doi: 10.1016/j.ajcnut.2024.02.005

TABLE 5.

Comparison of measured total energy expenditure with predicted total energy expenditure based on physical activity coefficient prediction equations (reported in Table 4)

Age class Activity category Males (n = 638)
Females (n =1019)
Regression1
Regression1
rc
LOA2
Intercept
Slope
rs
MAPE
rc
LOA2
Intercept
Slope
rs
MAPE
(%) kJ/d (%) (%) kJ/d (%)
All participants (≥65 y) Prediction based on age, weight, and height
Overall mean PA coefficient 0.63 3009 (±30.3) 3633 0.64 0.67 9.7 0.65 2362 (±30.1) 2810 0.66 0.66 10.1
Stratified by PA coefficient 0.87 1953 (±20.0) 1054 0.89 0.87 5.8 0.83 1755 (±21.8) 1275 0.85 0.83 7.1
Stratified by age category and mean PA coefficient 0.69 3019 (±30.2) 3850 0.63 0.66 10.1 0.65 2354 (±29.8) 2850 0.65 0.66 9.9
Stratified by age and PA coefficient
0.88
1918 (±19.5)
992
0.90
0.87
5.7
0.84
1736 (±21.5)
1226
0.85
0.83
6.9
Prediction based on age, weight, height, and RMR
Overall mean PA coefficent 0.68 2845 (±28.4.0) 3033 0.70 0.70 9.1 0.68 2269 (±28.7) 2521 0.69 0.69 9.8
Stratified by PA coefficient 0.96 1053 (±10.3) 226 0.99 0.96 4.0 0.97 842 (±10.6) 185 0.98 0.96 3.9
Stratified by age category and mean PA coefficient 0.68 2772 (±27.5) 2938 0.71 0.71 9.3 0.69 2263 (±28.5) 2577 0.68 0.69 9.9
Stratified by age and PA coefficient 0.97 1040 (±10.1) 197 0.98 0.97 3.8 0.97 830 (±10.5) 191 0.98 0.96 3.9

LOA, limit of agreement; MAPE, median absolute percentage error; PA, physical activity (see text and Table 4).

1

Regression = measured TEE versus predicted; rc = Lin’s concordance correlation coefficient; rs = Spearman correlation coefficient.

2

LOA = 1.96 SD limits of agreement.