Table 2 The relationship between circadian disruption measures and daily mood analyzed using linear mixed-effect modeling

From: The real-world association between digital markers of circadian disruption and mental health risks

 

Predictors

beta

z-statistics

Pr(>|z | )

95% C.I

Disruption vs next day mood

Before internship

CRCO misalignment

−0.014

−2.292

0.022

[−0.027, −0.002]

CRPO misalignment

−0.018

−5.261

< 0.001

[−0.025, −0.011]

Internal misalignment

−0.018

−6.182

< 0.001

[−0.024, −0.012]

During internship

CRCO misalignment

−0.013

−6.510

< 0.001

[−0.017, −0.009]

CRPO misalignment

−0.011

−7.441

< 0.001

[−0.015, −0.008]

Internal misalignment

−0.009

−7.255

< 0.001

[−0.012, −0.017]

Mood vs next day disruption

Before internship

Mood (vs CRCO)

−0.059

−3.958

< 0.001

[−0.088, −0.030]

Mood (vs CRPO)

−0.081

−2.995

0.003

[−0.134, −0.028]

Mood (vs Internal)

−0.118

−3.585

< 0.001

[−0.182, −0.053]

During internship

Mood (vs CRCO)

−0.053

−4.671

< 0.001

[−0.075, −0.031]

Mood (vs CRPO)

−0.082

−5.713

< 0.001

[−0.110, −0.054]

Mood (vs Internal)

−0.168

−9.450

< 0.001

[−0.202, −0.133]

  1. The association between current day circadian disruption and next day mood score before and during the internship was analyzed after accounting for the repeated measurements from the same subject and demographic and geographic information using linear mixed-effects models. The same analysis was repeated between current day mood and next day circadian disruption. Here, we report the beta estimate, its 95% confidence intervals (C.I.), and the p-value.