0585 Depressive Symptoms Account for Self-Reported Sleep Quality And Insomnia Severity in Obstructive Sleep Apnea Patients

Abstract

Introduction: Insomnia is often co-morbid with OSA, and each disorder is associated with depression. A complex relationship between these three conditions was recently proposed. Presently, subjective sleep disturbance is analyzed as a function of PSG variables and depressive symptoms in a large, ethnically diverse sample of OSA patients. Method(s): Before the in-lab PSG, 960 patients (18-97y.o., 520 women) completed Insomnia Severity Index (ISI, normal range<8), Pittsburgh Sleep Quality Index (PSQI, normal range<6) and Center for Epidemiologic Studies Depression Scale-Revised (CESDR, normal range<16). ISI and, separately, PSQI scores were regressed on age, sex, BMI, followed by PSG variables selected based on the criterion of p<0.1 when entered alone, then by CESDR and the CESDRxAHI interaction. Result(s): Group means: AHI=27.2+/-33.2 (635 patients had OSA, AHI>=5); ISI=13.0+/-7.0; PSQI=9.6+/-4.4; CESDR=14.4+/-12.6. The final model for ISI included total sleep time (TST), REM%, sleep efficiency (SEff), wake after sleep onset (WASO), apnea-hypopnea arousal index, total arousal index, and AHI. Higher ISI was predicted by higher WASO (p=0.043, R2=0.5%) and higher CESDR (p<0.001, R2=26%). The final model for PSQI included TST, REM%, SEff, WASO, and AHI. Higher PSQI was predicted by higher WASO (p=0.029, R2=0.5%) and higher CESDR (p<0.001, R2=20%). Other PSG variables and the CESDRxAHI interaction were not significant in either model. CESDR means were not significantly different between patients with OSA and those without (14.0+/-13.0 vs. 15.2+/-11.6, respectively). Conclusion(s): Depressive symptomatology plays a significant role in self-reported insomnia severity and sleep disturbance, accounting for approximately a fifth to a quarter of the variance in the ISI and PSQI. As CESDR did not interact with AHI in this large sample, the role of depressive symptoms in subjective sleep appears to be independent of the presence or severity of OSA. Among PSG variables, only WASO had a significant contribution to subjective sleep, accounting for a small fraction of the variance in the ISI and PSQI.