Table 2. Overview of regression models for predicting PAEE (MJ day⁻¹), stratified by amount of accelerometer information used.

Group	Model input	>7 days			1–3 days			1–3 days, participants with >7 days
		N	SE	R2	N	SE	R2	N	SE	R2
Non-	Acc0	48	0.97	0.22***	65	0.99	0.25***	48	0.98	0.21***
pregnant	Acc1	48	0.99	0.19***	65	1.01	0.22***	48	1.00	0.17***
women	Acc2	48	0.98	0.21***	65	1.00	0.24***	48	0.98	0.20***
	Weight	48	1.09	0.01 ns	65	1.13	0.03 ns	48	1.09	0.01 ns
	Acc2+Weight	48	0.92	0.31***	65	0.95	0.31***	48	0.94	0.27***
	Acc2+Body side § (with interaction)	48	0.99	0.18*	65	1.01	0.23*	48	1.00	0.18*
Pregnant	Acc0	26	0.76	0.08 ns	30	0.73	0.08 ns	26	0.74	0.13*
women	Acc1	26	0.75	0.09 ns	30	0.72	0.10 ns	26	0.73	0.15*
	Acc2	26	0.75	0.09 ns	30	0.71	0.11*	26	0.73	0.15*
	Weight	26	0.81	−0.04 ns	30	0.76	−0.02 ns	26	0.81	−0.04 ns
	Acc2+Weight	26	0.77	0.05 ns	30	0.72	0.09 ns	26	0.74	0.12 ns
	Acc2+Body side (with interaction)	26	0.71	0.19 ns	30	0.65	0.27*	26	0.66	0.30*

[SE: Residual standard error; ***: p<.001; **: p<.01; *: p<.05; ns: not significant; § no significant contribution from Body side (values based on forced inclusion); Acc: acceleration, where Acc₀: not based on non-wear detection and succeeding imputation; Acc₁: average acceleration (g) where non-wear time was imputed by the average of all wear-time data for that participant; Acc₂: average acceleration (g) where non-wear time was imputed by all wear-time data at similar time of the day for that participant; Body side, dominant wrist vs. non-dominant wrist].