Abstract
Introduction
Insomnia Disorder (ID) elevates risk of incident anxiety disorders and vice versa. We examined whether ID and poor sleep are associated with greater self-reported anxiety in persons with Generalized Anxiety Disorder (GAD).
Methods
Twenty-one participants with GAD and ID (GAD+/ID+) having Insomnia Severity Index (ISI) scores ≥ 13 (mean 17.8, SD 3.6) and 14 with GAD but not ID (GAD+/ID-) having ISI scores ≤ 12 (mean 6.4, SD 3.4) completed 14 days of actigraphy and sleep diaries as well as a night of ambulatory polysomnography (PSG) following an acclimation night. Participants completed the Pittsburgh Sleep Quality Index (PSQI), the State-Trait Inventory for Cognitive and Somatic Anxiety (STICSA-T/C, -T/S), the Ford Insomnia Response to Stress Test (FIRST), the Penn State Worry Questionnaire (PSWQ), and the Anxiety Sensitivity Index (ASI). Differences in self-reported anxiety (STICSA, ASI, PSWQ) between GAD+/ID+ and GAD+/ID- were analyzed using t-tests. Relationships of anxiety with retrospective (PSQI, FIRST, ISI), longitudinal (actigraphy, diaries) and physiological (PSG) sleep variables were analyzed using simple regression.
Results
GAD+/ID+ versus GAD+/ID- participants showed trends toward higher anxiety on the PSWQ (p=0.075), ASI (p=0.072) and STICSA-T/S (p=0.078). PSQI scores were positively associated with STICSA-T/S, (R=0.417, p=0.018, N=32). Greater insomnia reactivity (FIRST) was associated with increased worry on the PSWQ (R=0.352, p=0.044, N=33). STICSA-T/C was negatively associated with mean diary (R= -0.440, p=0.015, N=30) and actigraph (R= -0.517, p=0.01, N=24) total sleep time (TST). Actigraph mean TST trended toward lower PSWQ (R= -0.376, p=0.058, N=26) while actigraph mean sleep efficiency (SE) trended toward lesser STICSA-T/C (R= -0.397, p=0.058). Greater REM% was associated with greater STICSA-T/C (R=0.613, p=0.0005, N=28) and STICSA-T/S (R=0.516, p=0.005), a relationship also seen in GAD+/ID+ alone (p=0.03 and 0.015 respectively, N=16). Slow Wave Sleep% (SWS%) was not associated with lesser STICSA-T/S across both groups (p=0.14) but was so in GAD+/ID+ (R= -0.539, p=0.031, N=16).
Conclusion
GAD+/ID+ versus GAD+/ID-, show greater worry, anxiety sensitivity and somatic anxiety. In GAD, shorter and poorer quality sleep measured retrospectively or averaged longitudinally, as well as greater REM%, are associated with greater somatic and cognitive anxiety. Among those with ID, greater SWS% is associated with less somatic anxiety.
Support (if any)
R21MH115279, R01MH109638
Validation of a French version of the Sleep Condition Indicator: a clinical screening tool for insomnia disorder according to DSM-5 criteria