5 - (SYM 65) Examining Changes in Insomnia and Their Associations with Exercise and Psychiatric Symptoms for Individuals with Primary Anxiety Disorders in an Aerobic Exercise Intervention Trial

Background: Exercise can improve anxiety and sleep. However, the impact of exercise interventions on insomnia in anxiety disorders remains unclear, including time course of changes in sleep and psychiatric symptoms. In this study, low-active participants with primary anxiety disorders and high anxiety sensitivity (≥23 on Anxiety Sensitivity Index; ASI) participated in an 8-week exercise intervention. We hypothesized: 1) insomnia would improve over time; 2) insomnia improvement would be associated with more exercise; and 3) insomnia improvement would be associated with greater improvement in AS, anxiety, and depression.



Method: 52 participants (76.9% female, Mean age±SD=33.1±11.3) were administered structured diagnostic interviews for anxiety diagnosis and completed assessments at Weeks 1 (baseline), 4, and 8 (endpoint). Assessments were: Insomnia Severity Index (ISI), ASI, Hamilton Anxiety Rating Scale (HAM-A), and Montgomery Asberg Depression Rating Scale (MADRS); latter calculated without sleep items. Exercise was assessed by clinician-rated 7-Day Physical Activity Recall. Analyses were: 1) mixed effects model of ISI change; 2) correlation of Week 8 ISI and exercise; and 3) linear regressions examining associations between changes in ISI and psychiatric symptoms, controlling for Week 1 ISI and symptom measure.



Results: ISI improved over time (F(1,2)=5.8, p=.005, d=0.44; Mean(SE) Week 1: 11.2(.85); Week 8: 8.6(.80)), including for clinically elevated (≥15) Week 1 ISI (n=16; F(1,2)=14.7, p< .001, d=1.2; Week 1: 18.5(.80); Week 8: 12.6(1.7)). Week 8 (r=-.04, p=.77) exercise minutes and ISI were not associated. Week 1 to 8 ISI improvement was associated with Week 1 to 8 improvement in ASI (B=0.19, p=.003), HAM-A (B=0.42, p< .001), and MADRS (B=0.28, p=.004). Early ISI improvement (Week 1 to 4) was not associated with Week 1 to 8 change (all p >.18). However, early ASI (B=0.23, p=.015) and MADRS (B=0.24, p=.048) improvement were associated with Week 1 to 8 ISI improvement.



Conclusion: As hypothesized, participants with anxiety disorders demonstrated significant improvements in insomnia; those with clinical sleep disturbances improved to subthreshold by post-intervention. Early reduction in AS and depression predicted later insomnia changes but not vice versa, suggesting an exercise intervention may lead to symptom improvement which results in sleep changes. Surprisingly, total exercise minutes were not associated with sleep changes, suggesting need for further investigation, including into objective exercise and sleep.