Skip to main content
. 2020 Sep 18;11:561564. doi: 10.3389/fpsyt.2020.561564

Table 4.

Moderation analyses showing that neither sleep duration nor depression significantly moderated the association between insomnia and frequency of suicidal ideation in the replication sample (NHANES 2005–2006 cycle).

OR 95% CI p-value
Model 1
Insomnia 1.15 0.93–1.41 0.183
Sleep duration (h)
 <=4 8.22** 2.51–26.92 0.002
 5 2.16 0.63–7.38 0.195
 6 1.34 0.35–5.16 0.647
 7
 8 2.17 0.74–6.39 0.141
 9 6.39** 1.87–21.79 0.007
 >=10 1.54 0.13–18.53 0.709
Sleep duration X insomnia
 <=4 0.86 0.66–1.12 0.237
 5 1.00 0.77–1.30 0.979
 6 0.99 0.75–1.31 0.950
 7
 8 0.92 0.71–1.20 0.517
 9 0.82 0.62–1.09 0.150
 >=10 1.18 0.81–1.72 0.364
Model 2
Insomnia 1.11* 1.03–1.20 0.013
Depression 87.52*** 17.63–434.39 <0.001
Depression X insomnia 0.93 0.77–1.12 0.397

Each model includes adjustments for age, gender, race, education, poverty-to-income ratio, marital status, smoking status, alcohol consumption, and binge drinking. Model 1 is adjusted for depression and Model 2 is adjusted for sleep duration. N = 4,773. *p < 0.05, **p < 0.01, ***p < 0.001.