Abstract
Insomnia is a highly prevalent sleep disorder with strong bidirectional associations with depressive symptoms. The circadian preference for eveningness has been shown to be associated with depressive symptoms in insomnia and other mental health conditions. However, there is a lack of studies in insomnia investigating whether objective measures, such as dim light melatonin onset (DLMO) or polysomnographic (PSG) sleep, are associated with depressive symptoms. Therefore, we investigated the associations between subjective measures (questionnaires assessing anxiety, sleep quality and circadian preference, and sleep diary) and depressive symptoms and whether the addition of objective measures (DLMO, PSG parameters) would strengthen the associations with depressive symptoms. In 115 insomnia disorder patients we found that anxiety was strongly associated with depressive symptoms in a model including circadian preference, dysfunctional beliefs of sleep, and self-reported previous depressive symptoms (R(2) = 0.496, p < 0.001). The addition of sleep diary measures did not strengthen the model. We also found that the addition of objective measures (DLMO, PSG parameters) did not improve the subjective associations with depressive symptoms. Our data suggest that objective circadian markers are less important in the prediction of depressive symptoms in insomnia compared to subjective measures.