. 2021 Apr 21;6(2):36. doi: 10.3390/jfmk6020036

Table 3.

Predictive Performance *.

	ELPD		IC
Outcome	Est.	SE	Est.	SE	R²	RMSE
Composition ^†	−146.13	5.42	292.25	10.84	0.36	45.84
∆RMR	−96.28	2.68	192.57	5.36	0.24	6.75
∆MT_KE	−107.37	3.18	214.74	6.35	0.25	10.07
∆MT_EF	−94.63	4.25	189.25	8.51	0.18	6.89
∆1RM_LP	−121.07	3.89	242.14	7.78	0.34	16.30
∆1RM_BP	−94.12	3.12	188.24	6.24	0.40	6.64
∆RTF_LP	−179.03	11.83	358.05	23.66	0.33	147.50
∆RTF_BP	−146.36	4.52	292.72	9.04	0.19	42.83

* Metrics reflect results of out of sample predictive performance using 10-fold cross validation. ^† Calculated as: (∆FFM/∆BM)*100. Abbreviations. ELPD: expected log pointwise predictive density; IC: information criterion; Est: estimate; SE: standard error; R²: Bayes R²; RMSE: root mean square error; RMR: resting metabolic rate relative to fat-free mass; MT_KE: muscle thickness of knee extensors; MT_EF: muscle thickness of elbow flexors; 1RM_LP: one-repetition maximum for leg press; 1RM_BP: one-repetition maximum for bench press; RTF_LP: repetitions to failure for leg press; RTF_BP: repetitions to failure for bench press.