A Bayesian network analysis of sleep quality, anxiety, and depression symptoms in Peruvian Adults

It has been suggested that individuals with sleep disorders tend to experience concurrent mental health disorders, such as anxiety and depression. Therefore, this study aimed to address this gap by utilizing Bayesian network analysis to explore the potential causal relationships between sleep quality, anxiety, and depressive symptoms in a sample of 451 Peruvian adults. The network structures for sleep quality, depression, and anxiety were estimated using the Jenkins Sleep Scale, Patient Health Questionnaire-2, and General Anxiety Disorder-2, respectively. The causal relationships between symptoms were estimated using Bayesian networks from a directed acyclic graph (DAG) model. Nighttime Awakenings and Anhedonia play significant and distinct roles in the symptom network dynamics. Nighttime Awakenings showed directional probabilities of four symptoms: Nervousness, Difficulty Falling, Stay Asleep, and Depressed Mood. Anhedonia also showed directional probabilities toward three symptoms: Tiredness on Awakening, Uncontrollable Worry, and Depressed Mood. Meanwhile, although Nervousness does not have outgoing arrows to other symptoms, it shows conditional dependence with Uncontrollable Worry, Depressed Mood, and Nighttime Awakenings. The findings suggest adopting a comprehensive approach to the treatment of sleep disorders, anxiety, and depression, considering the interconnections among various symptoms and addressing not only the core symptoms but also those that function as mediators or bridges within the symptom network.