Table 3.

Accuracy of 8-week models for predicting 4-month and 12-month weight losses.

Model	4-month prediction				12-month prediction
	R 2	ρ	mAUCa	R 2		ρ	mAUC
Baseline variables (ie, age and treatment assignment)	0.16	0.41	0.65	0.06		0.25	0.58
App-use (PC1b)+app-calories (PC2c)+self-weighing frequency	0.31	0.58	0.74	0.07		0.29	0.57
Baseline variables+app-use (PC1)+app-calories (PC2)+self-weighing frequency	0.30	0.58	0.73	0.08		0.29	0.58
4-month weight loss	N/Ad	N/A	N/A	0.31		0.52	0.66

^amAUC: multi-class area under the receiver operating characteristic curve.

^bPC₁: principal component 1.

^cPC₂: principal component 2.

^dN/A: not applicable.