Skip to main content
. 2022 Feb 16;16:803349. doi: 10.3389/fnins.2022.803349

TABLE 2.

Weekday circadian alignment (Thursday DLMO-midsleep) predicting neural response (standardized score) during pre-weekend reward anticipation (Friday fMRI).

Pre-weekend reward anticipation Deviation model Tertile modela
Striatal response Core model b SE t-value p-value b SE t-value p-value
Intercept 4.15 3.22 1.29 0.216 2.23 3.6 0.62 0.544
Age −0.16 0.14 −1.14 0.272 −0.1 0.16 −0.6 0.559
Sex −0.45 0.34 −1.33 0.201 −0.21 0.4 −0.52 0.613
Racial identity −0.11 0.26 −0.41 0.691 0.11 0.29 0.37 0.716
Scan visit 0.64 0.36 1.81 0.089 0.51 0.39 1.32 0.206
DLMO-midsleep - short −1.12 0.33 −3.38 0.004c −1.28 0.46 −2.82 0.012c
DLMO-midsleep - long 0.04 0.19 0.20 0.848 −0.46 0.52 −0.89 0.389
Core model + prior night’s TST b
Thursday night TST 0.06 0.15 0.37 0.719 0.24 0.19 1.29 0.223
DLMO-midsleep - short −1.21 0.31 −3.94 0.002c −1.60 0.49 −3.29 0.006c
DLMO-midsleep - long 0.16 0.19 0.86 0.406 −0.93 0.59 −1.58 0.139
mPFC response Core model
Intercept 0.32 2.74 0.12 0.908 −0.99 2.87 −0.35 0.734
Age −0.06 0.12 −0.48 0.636 −0.01 0.13 −0.12 0.91
Sex 0.07 0.29 0.25 0.808 0.27 0.32 0.84 0.416
Racial identity 0.46 0.22 2.08 0.054 0.59 0.24 2.53 0.022
Scan visit 0.70 0.30 2.32 0.034 0.58 0.31 1.88 0.078
DLMO-midsleep - short −1.08 0.28 −3.82 0.002c −1.26 0.36 −3.46 0.003c
DLMO-midsleep - long 0.24 0.16 1.49 0.156 −0.54 0.41 −1.32 0.206
Core model + prior night’s TST b
Thursday night TST −0.22 0.13 −1.71 0.113 −0.12 0.16 −0.74 0.47
DLMO-midsleep - short −1.04 0.27 −3.87 0.002c −1.28 0.42 −3.08 0.010c
DLMO-midsleep - long 0.21 0.17 1.24 0.238 −0.50 0.50 −0.99 0.344

a“Medium” DLMO-midsleep was the referent group for Tertile Models. Thus, coefficient estimates for phase angle “DLMO-midsleep - short” and “DLMO-midsleep - long” indicate changes relative to the “medium” phase angle group, respectively.

bThe model adding prior night’s total sleep time (TST) includes all the covariates included in the Core model, but they are not listed for clarity of presentation. All covariate effects remained non-significant in subsequent models.

cOutcome survived Benjamin-Hochberg multiple comparison correction. Bolded values indicate p < 0.05.