Table 2.

Comparison of Actigraphy and PSG Sleep Parameters

Sleep parameter	PSG	Actigraphy	Mean bias	Absolute bias	T statistic	Effect size	P-value
	Mean ±SD	Mean ±SD	Mean ±SD	Mean ±SD
GGIR: Actigraphy with sleep diary (n = 835)
TST, hours	5.95 ± 0.96	6.54 ± 0.84	0.6 ± 0.86	0.78 ± 0.70	19.99	0.69	<0.001
SOL, hours	0.29 ± 0.33	0.05 ± 0.17	−0.24 ± 0.35	0.27 ± 0.32	−20.09	−0.7	<0.001
WASO, hours	1.25 ± 0.71	0.9 ± 0.61	−0.35 ± 0.74	0.60 ± 0.55	−13.72	−0.47	<0.001
SE, %	78.0 ± 11.51	87.21 ± 9.45	9.21 ± 12.1	11.54 ± 9.9	22.00	0.76	<0.001
GGIR: Actigraphy without sleep diary (n = 832)
TST, hours	5.95 ± 0.95	6.4 ± 1.21	0.44 ± 1.15	0.91 ± 0.84	11.03	0.38	<0.001
WASO, hours	1.24 ± 0.7	0.77 ± 0.54	−0.47 ± 0.77	0.66 ± 0.61	−17.46	−0.61	<0.001
SE, %	78.09 ± 11.36	88.67 ± 7.48	10.58 ± 11.86	12.18 ± 10.21	25.72	0.89	<0.001
ActiLife: Actigraphy with sleep diary (n = 831)
TST, hours	5.95 ± 0.96	6.73 ± 0.76	0.78 ± 0.84	0.87 ± 0.75	27.00	0.94	<0.001
SOL, hours	0.29 ± 0.33	0.07 ± 0.08	−0.23 ± 0.34	0.25 ± 0.33	−19.47	−0.68	<0.001
WASO, hours	1.25 ± 0.71	0.73 ± 0.5	−0.52 ± 0.71	0.65 ± 0.59	−21.23	−0.74	<0.001
SE, %	77.99 ± 11.54	89.54 ± 6.66	11.55 ± 11.02	12.34 ± 10.13	30.21	1.05	<0.001

Comparisons were made using paired t-tests.

The sample of actigraphy files assessed in GGIR without a sleep diary was reduced by three participants due to TST being calculated as <2 hours, as per exclusion criteria (n = 832).

The sample of actigraphy files processed in ActiLife was reduced by four participants as there was an error with their data being formatted into 60-second epochs (sample, n= 831).