Table 3.
Relationships of polygenic score for depression with overall sleep duration, and the onset of short-sleep and long-sleep during an average 8-year follow-up.
| Models | Sleep duration | Short-sleepc | Long-sleepc | ||||||
|---|---|---|---|---|---|---|---|---|---|
| β (SE) | 95% CI | p | RRR (SE) | 95% CI | p | RRR (SE) | 95% CI | p | |
| Polygenic score for depression | |||||||||
| Model 1: Unadjusted modela, b | −0.001 (0.002) | −0.005–0.002 | 0.452 | 1.043 (0.044) | 0.960–1.133 | 0.324 | 0.972 (0.065) | 0.854–1.108 | 0.675 |
| Model 2: Model 1 + age, age2, sex, and 10 PCs | −0.002 (0.002) | −0.005–0.002 | 0.407 | 1.055 (0.045) | 0.970–1.148 | 0.212 | 0.966 (0.065) | 0.846–1.103 | 0.607 |
PCs principal components, β standardised regression coefficient, RRR relative risk ratios, SE standard error, CI confidence interval, p significance value. Alpha values have been adjusted to account for multiple testing.
aBaseline caseness of outcomes were omitted from analyses.
bSleep duration squared was included in sleep duration models to account for non-linearity.
cBaseline comparison was optimal sleep.