1006 Investigating the Modifying Effects of Pain and Depression on the Relation Between Sleep and Diurnal Salivary Cortisol in Temporomandibular Disorder

Abstract Introduction Approximately 60% of individuals with temporomandibular disorders (TMD) experience poor sleep, and 20% report daytime sleepiness. Sleep disruption influences hypothalamic-pituitary-adrenal (HPA) axis activity, yet its effects on diurnal cortisol (DC) rhythms remain poorly understood, as pain and depression are prevalent in TMD and can alter basal HPA functioning. We examined how sleep quality and sleepiness relate to DC patterns, and whether pain severity and depressive symptoms modify these trajectories. Methods TMD patients (N=33) from a university facial pain clinic completed the Pittsburgh Sleep Quality Index (PSQI), Epworth Sleepiness Scale (ESS), Brief Pain Inventory (BPI), and Beck Depression Inventory (BDI). Participants subsequently provided five saliva samples at 15-minute intervals after waking (cortisol awakening response; CAR) and seven samples every two hours from 10:00-22:00. Multilevel linear mixed-effects models examined the effects of PSQI and ESS on DC, with BPI pain severity and BDI as potential moderators. Likelihood ratio tests assessed whether adding interactions improved model fit. DC was modeled with natural cubic splines, and estimated marginal means plots and contrasts compared predicted cortisol across PSQI and ESS levels over time. Results Participants (Mage=33.5±10.1 years, 70% female, 64% white) reported poor sleep quality (MPSQI=7.4±3.6) and daytime sleepiness (MESS=9.2±4.1; Mdn=10.0). Contrasts indicated that higher PSQI predicted blunted cortisol during the CAR and between 18:00-22:00 (b=-0.37–-0.77, all ps<.05). Higher ESS predicted elevated cortisol during the CAR only (b=0.37–0.47, ps<.05). This association was amplified at low and mean BPI (b=0.50–1.12, ps<.05) and, in participants with higher BDI, throughout the CAR and between 20:00-22:00 (b=0.54–0.84, ps<.05). BPI and BDI did not modify PSQI effects. Conclusion These findings suggest the effects of sleep on cortisol are time- and context-dependent. Daytime sleepiness predicted elevated CAR at low–average pain, potentially reflecting a compensatory HPA response that is masked by high basal cortisol or dampened by chronic nociceptive stress at higher pain levels. Similarly, elevated depressive symptoms may amplify cortisol responses to acute sleepiness due to heightened stress sensitivity. No moderating effects were observed for PSQI, suggesting trait-like sleep disturbance may exert stable blunting effects on DC, whereas state-like sleepiness interacts with pain and mood. Support (if any) Johns Hopkins Medicine Blaustein Pain Research Fund Grant