Table 2 Model Estimated Effects of Polysomnography in Actigraphy Measured Sleep

Sleep Quality (SQ)

Z-skew/kurt (10.8 / -2.8). Transformation did not change the findings

 G1: No PSG (patients)

-7.4 _{-18.1 | 3.3}

38.1 (28.5)

- 3.7 _{-14.4 | 7.9\0}

- 6.4 _{-14.7 | 1.9}

 G2: PSG (patients)

-16.6^*** _{-27.2 | -5.9}

57.9 (28.3)

- 30.6^*** _{-41.3 | -1.9}

- 28.0^*** _{-38.3 | -19.7}

 G3: PSG (controls)

-25.9^*** _{-37.3 | -14.6}

45.4 (22.8)

- 25.8^*** _{-37.2 | -14.5}

- 27.4^*** _{-36.1 | -18.8}

 Between G1-G2-G3 contrasts ^BC

G3 < G1^**

G2 > G1^***

G2 > G3^*

G2 < G1^***

G3 < G1^***

G2 < G1^***

G3 < G1^***

Sleep Onset Latency (SOL)

Z-skew/kurt (78.4 / 366.5), which a ln(√SOL) transformation remedied (1.1 / -2.5)

 G1: No PSG (patients)

14.7 (29.1)

- 3.6 _{-20.9 | 13.4} ^CI

- 0.4 _{-2.2 | 2.0} ^CI

6,7 ^*

 G2: PSG (patients)

20.3 (28.8)

- 6.7^*** _{-11.4 | -2.0} ^CI

- 4.1^** _{-7.6 | -0.6} ^CI

 G3: PSG (controls)

16.0 (13.6)

- 8.9^* _{-17.9 | 0.1} ^CI

- 6.7^** _{-12.4 | -1.0} ^CI

 Between G1-G2-G3 contrasts ^BC

G2 > G1^*

G2 < G1^*

G2 < G1^*

G3 < G1^*

Wake After Sleep Onset (WASO)

Z-skew/kurt (13.0 / 15.0), which a ln(WASO) transformation remedied (-0.4 / -0.2)

 G1: No PSG (patients)

36.5 (23.3)

3.9 _{-5.9 | 13.7} ^CI

1.4 _{-3.9 | 6.8} ^CI

 G2: PSG (patients)

34.6 (23.4)

0.6 _{-8.2 | 9.4} ^CI

2.8 _{-3.7 | 9.3} ^CI

 G3: PSG (controls)

25.1 (14.3)

3.6^* _{-0.8 | 7.9} ^CI

5.1^*** _{1.3 | 8.9} ^CI

 Between G1-G2-G3 contrasts ^BC

G1 > G3^**

G2 > G3^**

G3 > G1^*

Total Sleep Time (TST)

Z-skew/kurt (2.2 / 15.7). Transformation did not change the findings

 G1: No PSG (patients)

6h 40m (88m)

6m _{-40 | 29}

-8m _{-35 | 18}

-26m^**

 G2: PSG (patients)

6h 40m (85m)

15m _{-20 | 50}

2m _{-25 | 29}

 G3: PSG (controls)

6h 27m (73m)

9m _{-28 | 46}

11m _{-17 | 40}

-19m^**qt

 Between G1-G2-G3 contrasts ^BC

Sleep Efficiency (SE %)

Z-skew/kurt (-26.1 / 80.0). Due to the severy skew, a ln(SE) transformation with additional kurtosis adjustment was performed even though it did not change the findings (1.2 / 2.3)

 G1: No PSG (patients)

86.3 (8.8)

-0.3 _{-2.5 | 3.1}

-0.1 _{-2.2 | 2.1}

 G2: PSG (patients)

86.0 (7.5)

1.2 _{-1.7 | 4.0}

0.7 _{-1.5 | 2.9}

 G3: PSG (controls)

88.5 (5.7)

1.0 _{-2.0 | 4.1}

0.8 _{-1.6 | 3.1}

 Between G1-G2-G3 contrasts ^BC

G3 > G1^*

G3 > G2^*