Abstract
Introduction
The degree that posttraumatic stress disorder (PTSD) contributes to obstructive sleep apnea (OSA) or sleep specific quality of life (QOL) remains uncertain.
Methods
We evaluated consecutive military veterans (n=3,155) with suspected OSA using multivariable regressions to test associations between sleep and QOL measures including the apnea-hypopnea index (AHI), Pittsburgh Sleep Quality Index (PSQI) and Short-Form-12 mental component score (MCS). A mental health expert determined PTSD presence with severity measured utilizing the PTSD Checklist (PCL). Subjects were evaluated with prospectively collected questionnaires, sleep studies, and detailed electronic medical record reviews.
Results
Current-PTSD (n=1,172, 37%) were younger, more likely single, unemployed on disability, report non-white ethnicity or have current alcohol or drug dependence, report past suicide attempt, have current insomnia, restless sleep or nightmares, report lower MCS or higher Epworth Sleepiness, Fatigue Severity, Beck Depression Inventory-II (BDI) and PSQI scores than non-PTSD (n=1,880) or past-PTSD (n=103) veterans. Among current-PTSD, instigating trauma was 75% combat and 13% sexually related. In multivariable regressions, male gender (OR 4.5, p<0.001), age >65 years (OR 2.3, p<0.001), BMI ≥35 kg/m2 (OR 3.5, p<0.001), prior stroke (OR 1.8, p<0.006), current hypertension (OR 1.4, p<0.001), neck circumference >40 cm (OR 1.3, p=0.032), and non-white ethnicity (OR 1.2, p=0.034) were associated with moderate-severe OSA (AHI ≥15/h), however current (OR 0.9, p=0.06) or past-PTSD (OR 1.2, p=0.41) were not. PCL (p=0.937) was not associated with AHI. Factors most associated with lower MCS included current-PTSD (scaled standardized beta[SSB]=0.09, p=0.001), depression (SSB=0.09, p=0.001), age <50 years (SSB=0.09, p<0.001), non-white ethnicity (SSB=0.07, p=0.004), female gender (SSB=0.06, p=0.007) or single/no-partner (SSB=0.05, p=0.03). Likewise, factors most associated with a higher PSQI included depression (SSB=0.19, p<0.001), current-PTSD (SSB=0.15, p<0.001), unemployed on disability (SSB=0.14, p<0.001), non-white ethnicity (SSB=0.13, p<0.001) or age <50 years (SSB=0.10, p=0.001). Among current-PTSD, higher PSQI was associated with BDI ≥20 (SSB=0.31, p<0.001), PCL ≥50 (SSB=0.24, p<0.001) and non-white ethnicity (SSB=0.11, p=0.034), but not with moderate-severe OSA (SSB= -0.09, p=0.095).
Conclusion
In the largest PTSD sleep cohort to date, PTSD is associated with insomnia, restless sleep, poorer sleep specific QOL, and greater daytime sleepiness and fatigue, but is not associated with OSA.
Support
None