Is moderate depression associated with sleep stage architecture in adolescence? Testing the stage type associations using network and transition probability approaches

Characterization of sleep efficiency transitions in family caregivers

Family caregivers are at high risk of psychological distress and low sleep efficiency resulting from their caregiving responsibilities. Although psychological symptoms are associated with sleep efficiency, there is limited knowledge about the association of psychological distress with variations in sleep efficiency. We aimed to characterize the short- and long-term patterns of caregivers' sleep efficiency using Markov chain models and compare these patterns between groups with high and low psychological symptoms (i.e., depression, anxiety, and caregiving stress). Based on 7-day actigraphy data from 33 caregivers, we categorized sleep efficiency into three states, < 75% (S1), 75-84% (S2), and ≥ 85% (S3), and developed Markov chain models. Caregivers were likely to maintain a consistent sleep efficiency state from one night to the next without returning efficiently to a normal state. On average, it took 3.6-5.1 days to return to a night of normal sleep efficiency (S3) from lower states, and the long-term probability of achieving normal sleep was 42%. We observed lower probabilities of transitioning to or remaining in a normal sleep efficiency state (S3) in the high depression and anxiety groups compared to the low symptom groups. The differences in the time required to return to a normal state were inconsistent by symptom levels. The long-term probability of achieving normal sleep efficiency was significantly lower for caregivers with high depression and anxiety compared to the low symptom groups. Caregivers' sleep efficiency appears to remain relatively consistent over time and does not show rapid recovery. Caregivers with higher levels of depression and anxiety may be more vulnerable to sustained suboptimal sleep efficiency.

A two-branch trade-off neural network for balanced scoring sleep stages on multiple cohorts

Introduction

Automatic sleep staging is a classification process with severe class imbalance and suffers from instability of scoring stage N1. Decreased accuracy in classifying stage N1 significantly impacts the staging of individuals with sleep disorders. We aim to achieve automatic sleep staging with expert-level performance in both N1 stage and overall scoring.

Methods

A neural network model combines an attention-based convolutional neural network and a classifier with two branches is developed. A transitive training strategy is employed to balance universal feature learning and contextual referencing. Parameter optimization and benchmark comparisons are conducted using a large-scale dataset, followed by evaluation on seven datasets in five cohorts.

Results

The proposed model achieves an accuracy of 88.16%, Cohen's kappa of 0.836, and MF1 score of 0.818 on the SHHS1 test set, also with comparable performance to human scorers in scoring stage N1. Incorporating multiple cohort data improves its performance. Notably, the model maintains high performance when applied to unseen datasets and patients with neurological or psychiatric disorders.

Discussion

The proposed algorithm demonstrates strong performance and generalizablility, and its direct transferability is noteworthy among similar studies on automated sleep staging. It is publicly available, which is conducive to expanding access to sleep-related analysis, especially those associated with neurological or psychiatric disorders.

Health-promoting behavior to enhance perceived meaning and control of life in chronic disease patients with role limitations and depressive symptoms: a network approach

The association between health-related role limitations in the mental and physical subdomains and clinical status (i.e., chronic disease and comorbid depressive symptoms) is mediated by health-promoting behaviors. To enhance health-promoting behaviors in adults with chronic disease, it is necessary to identify item-level associations among targets of health-related monitoring and management. Therefore, the current study used a network approach to examine associations among health-related role limitations, depressive symptoms, existential well-being, socioeconomic position, and health-promoting behavior in adults with chronic disease. A total of 535 adults (mean ± SD age = 62.9 ± 11.9 years; males, n = 231, females, n = 304) who were regularly visiting an outpatient clinic for chronic disease treatment participated in this cross-sectional study. Data on participant demographics, chronic disease diagnoses, socioeconomic status, health-related role limitations (12-item short form survey scores), depressive symptoms (patient health questionnaire-9 scores), existential well-being (scores for four items of the McGill quality of life questionnaire-Revised), and health-promoting behavior (Healthy Habits Questionnaire scores) were acquired. "Undirected regularized partial correlations" and "directional joint probability distributions" among these variables were calculated using a mixed graphical model (MGM) and directed acyclic graph (DAG). In the MGM, the most influential nodes were emotional well-being, feelings of failure, and health-related limitations affecting usual role and physical activities. According to both the MGM and DAG, the relationship between emotional well-being and feelings of failure mediated the relationships of health-related role limitations with concentration difficulty and suicidal ideation. A positive mindset was dependent on the probability distributions of suicidal ideation, controllability of life, and positive self-image. Both the meaning of life and a positive mindset had direct associations with proactive living. Specifically, proactive living was associated with a balanced diet, regular exercise, volunteering in the community, and nurturing intimacy in social interactions. The meaning and controllability of life in individuals with chronic diseases could mediate the relationships of health-promoting behavior with health-related limitations related to usual role activities, physical activities, and depressive symptoms. Thus, interventions targeting health-promoting behaviors should aim to enhance the meaning and controllability of life (as it pertains to limitations in usual role and physical activities), as well as promote proactive screening and timely psychiatric treatment of depressive symptoms including feelings of failure, concentration difficulties, and suicidal ideation.

Long-term cumulative light exposure from the natural environment and sleep: A cohort study

We analysed (A) the association of short-term as well as long-term cumulative exposure to natural light, and (B) the association of detailed temporal patterns of natural light exposure history with three indicators of sleep: sleep duration, sleep problems, and diurnal preference. Data (N = 1,962; 55% women; mean age 41.4 years) were from the prospective Young Finns Study, which we linked to daily meteorological data on each participant's neighbourhood natural light exposure using residential postal codes. Sleep outcomes were self-reported in 2011. We first examined associations of the sleep outcomes with cumulative light exposure of 5-year, 2-year, 1-year, and 2-month periods prior to the sleep assessment using linear and Poisson regression models adjusting for potential confounders. We then used a data-driven time series approach to detect clusters of participants with different light exposure histories and assessed the associations of these clusters with the sleep outcomes using linear and Poisson regression analyses. A greater cumulative light exposure over ≥1 year was associated with a shorter sleep duration (β = -0.10, 95% confidence interval [CI] -0.15 to -0.04), more sleep problems (incident rate ratio [IRR] 1.04, 95% CI 1.0-1.07) and diurnal preference towards eveningness (β = -0.09, 95% CI -0.14 to -0.03). The data-driven exposure pattern of "slowly increasing" light exposure was associated with fewer overall sleep problems (IRR 0.93, 95% CI 0.88-0.98) compared to a "recently declining" light exposure group representing the "average-exposure" group. These findings suggest that living in an area with relatively more intense light exposure for a longer period of time influences sleep.

Phenotype Network and Brain Structural Covariance Network of Major Depression

Phenotype networks enable clinicians to elucidate the patterns of coexistence and interactions among the clinical symptoms, negative cognitive styles , neurocognitive performance, and environmental factors in major depressive disorder (MDD). Results of phenotype network approach could be used in finding the target symptoms as these are tightly connected or associated with many other phenomena within the phenotype network of MDD specifically when comorbid psychiatric disorder(s) is/are present. Further, by comparing the differential patterns of phenotype networks before and after the treatment, changing or enduring patterns of associations among the clinical phenomena in MDD have been deciphered.Brain structural covariance networks describe the inter-regional co-varying patterns of brain morphologies, and overlapping findings have been reported between the brain structural covariance network and coordinated trajectories of brain development and maturation. Intra-individual brain structural covariance illustrates the degrees of similarities among the different brain regions for how much the values of brain morphological features are deviated from those of healthy controls. Inter-individual brain structural covariance reflects the degrees of concordance among the different brain regions for the inter-individual distribution of brain morphologic values. Estimation of the graph metrics for these brain structural covariance networks uncovers the organizational profile of brain morphological variations in the whole brain and the regional distribution of brain hubs.