Unobtrusive Perceived Sleep Quality Monitoring in the Wild

Introduction Military Veterans tend to experience chronic pain and sleep problems of higher intensity and prevalence than the general population, and may be more likely to self-medicate with substances such as alcohol. Research suggests that sleep and pain are interrelated. However, few studies have examined bidirectional links between daily sleep quality and duration with pain among Veterans, especially those with alcohol use disorders. Furthermore, it has not been studied whether daily alcohol consumption (whether they drank or not) moderates everyday associations between sleep and pain among Veterans with alcohol use disorders. Methods 114 Veterans with alcohol use disorder (82.5% men: mean age = 38.86 [range 21–67]) completed 14 days of diaries assessing their sleep quality, sleep duration, pain, and drinking quantity. Multilevel modeling examined the effects of daily variability in sleep (within-person effects) and average levels of sleep (between-person effects) on pain the following day. To examine the effects of pain on sleep, multilevel modeling examined the effects of daily variability in pain and average levels of pain on sleep. Results Regarding bidirectional links of sleep quality and pain, greater daily (b=−.86, se=.41, p=.04) and average (b=−15.51, se=3.28, p<.001) sleep quality were associated with lower next-day pain. Higher average pain was associated with poorer sleep quality (b=−.01, se=.002, p< .001), but daily pain was not associated with daily sleep quality. Regarding bidirectional links of sleep duration and pain, there was no daily association between sleep duration and pain, but greater average sleep duration was associated with lower pain levels (b=−4.84, se=1.60, p=.003). Daily drinking did not moderate these associations. Conclusion Both sleep quality and sleep duration are closely associated with pain among Veterans with alcohol use disorder. Fluctuations of daily sleep quality affected Veterans’ pain experience in everyday life regardless of alcohol consumption; in contrast, fluctuation of daily pain did not affect daily sleep quality, implying that Veterans with alcohol use disorder are sensitive to poor sleep quality. The results support that sleep quality may buffer daily pain experiences among this patient population. Support (if any) 5K23AA026895

This study focuses on the development of a neural network model to predict perceived sleep quality using data from wearable devices. We collected various physiological metrics from 18 participants over four weeks, including heart rate, physical activity, and both device-measured and self-reported sleep quality. The primary objective was to correlate wearable device data with subjective sleep quality perceptions. Our approach used data processing, feature engineering, and optimizing a Multi-Layer Perceptron classifier. Despite comprehensive data analysis and model experimentation, the predictive accuracy for perceived sleep quality was moderate (59%), highlighting the complexities in accurately quantifying subjective sleep experiences through wearable data. Applying a tolerance of 1 grade (on a scale from 1-5), increased accuracy to 92%. More in-depth analysis is required to fully comprehend how wearables and artificial intelligence might assist in understanding sleep behavior.

IntroductionSleep is critical to the health, wellbeing and performance of military personnel during basic training. This two-part study evaluated sleep-wake patterns and sleep disturbances in junior soldiers (JS) and infantry recruits in Autumn 2021 (study 1), and non-infantry recruits in spring 2022 (study 2).MethodsDuring studies 1 and 2, validated wearable technology combined with a sleep diary was used to quantify sleep-wake indices, sleep disturbances and perceptions of sleep quality. Sleep diary data was analysed descriptively. A series of repeated-measures ANOVAs examined differences in objective sleep-wake indices. Correlation analysis determined associations between time in bed (TIB) and total sleep time (TST).ResultsSignificant (p < 0.05) differences in most sleep-wake indices were observed between weeks of basic training for all cohorts. Strong positive correlations between TIB and TST were observed for each cohort across basic training (r = 0.681 – 0.970, p < 0.001), with longer TST associated with greater TIB. The mean±SD sleep duration (hours and mins [hm]) for JS (06:22 ± 00:27hm), non-infantry (05:41 ± 00:47hm) and infantry (05:46 ± 00:34hm) recruits across basic training was consistently below national recommendations. The mean±SD bed and wake times for JS (bedtime: 23:01 ± 00:32hm; awake: 05:34 ± 00:10hm), non-infantry (bedtime: 23:38 ± 01:09hm; awake: 04:47 ± 00:58hm), and infantry (bedtime: 23:13 ± 00:29hm; awake: 05:38 ± 00:26hm) recruits varied across weeks of basic training, with over 80% reporting “fairly bad” or “very bad” sleep quality and frequent periods of “dozing off” during daytime activity. The most commonly reported sleep disturbing factors identified during basic training involved: late-night military admin (e.g., ironing, boot cleaning, kit set up etc), early morning wake times, extraneous noise, light and hot room temperatures within the primary sleeping environment, bed/mattress discomfort, muscle soreness and feelings of stress and anxiety.Discussion/ConclusionOur findings contribute to the existing evidence that long-term sleep loss is pervasive during initial military training programmes. The average sleep durations indicate chronic and unrecoverable sleep loss which would be expected to significantly impair physical and cognitive military performance, and increase the risk of injury, illness and attrition rates during basic training. Changes in the design and scheduling of basic training programmes to enable, at the least, minimum sleep recommendations to be met, and to improve sleep hygiene in the primary sleeping environment are warranted.

Introduction: Assessing preoperative cardiorespiratory fitness (CRF) is essential for determining surgical risk and outcomes. The advent of wrist-worn devices equipped with heart rate and activity monitoring presents new opportunities for CRF evaluation. Traditionally, CRF has been measured using submaximal exercise tests, such as the 6-minute walk test (6MWT). Monitoring CRF in everyday activities may offer valuable insights into patients' preoperative fitness levels. This study explores the potential of predicting 6MWT distances using a data-driven physiological CRF model derived from wearable devices in elderly patients in their free-living environments. Hypothesis: Machine-learning algorithms can accurately estimate CRF using data from fitness trackers. Methods: We examined heart rate and activity data from Fitbit devices worn by 65 older adult patients undergoing major noncardiac surgery. Data was collected over a week in free-living conditions. We aimed to utilize this wearable technology to forecast CRF in this group. A machine-learning ensemble regression model was employed to predict CRF, with 6MWT outcomes as the reference. The Shapley Feature Attribution method was applied to understand the contribution of specific features derived from wearable data to the model, aiding in personalized fitness predictions. Results: Higher CRF was associated with increased levels of moderate to vigorous physical activity (MVPA), maximal activity energy expenditure (aEEmax), 4-minute heart rate recovery (HRR), and non-linear heart rate variability (HRV) features. These measures positively correlated with improved 6MWT outcomes. Our regression models, including random forest and linear regression, showed strong predictive capabilities, with coefficients of determination (R^2) of 0.91 and 0.81, respectively. Shapley Feature Attribution analysis revealed that higher levels of MVPA, aEEmax, HRR240, and non-linear HRV dynamics are reliable indicators of better CRF test performance. Conclusions: Integrating wearable activity and heart rate data offers a robust framework for preoperative CRF assessment, enhancing surgical risk evaluation and patient outcomes in everyday conditions.

Introduction Research on the link between sleep quality and depression in the postpartum period has focused primarily on mothers. Although fathers also experience poorer postpartum sleep and are at risk of developing depressive symptoms, they remain understudied. To date, the limited research focusing on paternal sleep and depression has relied on subjective measures of sleep, without objective verification. The current study implemented a multi-measure approach using subjective and objective indices to explore the relationship between sleep and depressive symptoms in fathers at 6 months postpartum. Methods Fifty-four healthy fathers participated in this cross-sectional study. Paternal sleep was assessed for 2 weeks utilizing: 1) a self-report daily sleep diary, 2) a self-report perceived sleep quality rating, and 3) actigraphy. Subjective indices via the sleep diary measured participants’ perception of their total nocturnal sleep duration and total number of awakenings (self-reported sleep duration and fragmentation). Perceived sleep quality ratings measured participants’ perceptions of how well they thought they slept. Objective sleep variables measured through actigraphy included: total nocturnal sleep duration, number of awakenings, sleep efficiency, and wake after sleep onset (WASO). Paternal depressive symptoms were assessed with the Center for Epidemiologic Studies – Depression Scale (CES-D). Results Regression analyses showed that subjective sleep variables (measured by the sleep diary) and objective sleep variables (measured by actigraphy) did not significantly predict postpartum depressive symptoms in fathers (p &amp;gt; .05). However, self-reported perceived sleep quality significantly predicted postpartum depressive symptom severity in fathers (R2 = .172, p = .034). Conclusion These findings advance our understanding of the link between sleep and depression in fathers. The results highlight the important role of fathers’ perceptions of sleep quality, rather than the actual quality or quantity of their sleep (measured through the sleep diary or actigraphy), in the development of postpartum depressive symptoms. The multi-measure approach to sleep implemented in this study expanded our knowledge about how different facets of sleep relate to depression. These findings have important implications for the development of clinical interventions targeting paternal sleep and mood in the months following childbirth. Support (if any) Social-Science and Humanities Research Council (SSHRC) and Fonds de recherche du Québec – Santé (FRQS)

BackgroundAcute aortic dissection is a life-threatening condition, and malperfusion significantly exacerbates the prognosis of patients diagnosed with type A Acute aortic dissection (ATAAD). Current risk assessment tools often fail to consider the impact of nutritional and inflammatory factors, limiting their predictive accuracy. The aim of this study was to develop a machine learning model that integrates nutritional and inflammatory indices to predict 1-year mortality in ATAAD patients with malperfusion.MethodsThis retrospective study included 433 ATAAD patients with malperfusion from Henan Provincial Chest Hospital (August 2020 to June 2023). Four machine learning models—logistic regression, XGBoost, random forest, and deep neural network—were developed to predict 1-year mortality using inflammatory and nutritional laboratory values, indices, and other clinical variables. Model training employed stratified 5-fold cross-validation and SMOTE for imbalanced data. The area under the receiver operating characteristic (ROC AUC) and other performance metrics were used to evaluate model efficacy, while SHAP values were computed to interpret feature importance.ResultsAmong 433 ATAAD patients with malperfusion, the random forest model used inflammatory and nutritional laboratory values to achieve the highest discrimination (AUC = 0.8242, 95% CI 0.7095–0.9219), while the XGBoost model performed best with inflammatory and nutritional indices (AUC = 0.7334, 95% CI 0.6115–0.8488). Calibration curves and Brier scores indicated good agreement between predicted and observed outcomes. Decision curve analysis demonstrated consistent net benefit for random forest and XGBoost models across clinically relevant threshold probabilities. Feature importance and SHAP analyses identified albumin, platelet count, total cholesterol, and C-reactive protein as consistently influential predictors.ConclusionNutritional and inflammatory factors significantly contribute to the 1-year mortality risk of ATAAD patients with malperfusion. Machine learning models that incorporate these factors, particularly random forest and XGBoost, can effectively stratify patient risk and support clinical decision-making. These findings underscore the importance of a comprehensive approach to risk assessment that includes metabolic and inflammatory markers to enhance patient outcomes and guide personalized interventions.

Sleep disturbance is a core transdiagnostic factor in mental health, intricately linked to the onset and maintenance of depression. Objectives: To identify the underlying clinical dimensions of sleep disturbance in university students and explore their behavioral and cognitive predictors. Methods: This cross-sectional study of 151 students from the University of the Punjab was conducted using a comprehensive sleep questionnaire. Data were analyzed using exploratory factor analysis to identify latent constructs of sleep disturbance, with group differences examined via Mann-Whitney U and Kruskal-Wallis tests. Regression analyses identified key behavioral and cognitive predictors of adverse sleep outcomes. Results: Factor analysis revealed eight distinct clinical dimensions of sleep disturbance: Daytime Impairment, Clinical Sleep Disorders, Perceived Sleep Quality, Sleep Hygiene, Sleep Fragmentation, Pre-sleep Stimulation, Nocturnal Hyperarousal, and Autonomic Symptoms. Clinically significant group differences emerged: females reported greater daytime functional impairment (p=0.009), while males endorsed poorer perceived sleep quality (p=0.027). Lower family income was uniquely associated with increased sleep fragmentation (p=0.034). Critically, regression models identified pre-bed screen use as a significant predictor of reduced total sleep duration (p=0.007) and nocturnal cognitive hyperarousal (worry) as a predictor of prolonged sleep onset latency (p=0.036). Conclusions: Sleep disturbances in university students comprise multiple, co-occurring dimensions rather than a single deficit. Personalized strategies combining stimulus control and cognitive techniques can directly address the core drivers of sleep dysfunction in this population.

Gully erosion is a widespread environmental danger, threatening global socio-economic stability and sustainable development. This study comprehensively applied seven machine learning (ML) models including SVM, KNN, RF, XGBoost, ANN, DT, and LR, and evaluated gully erosion susceptibility in the Tensift catchment and predict it within the Haouz plain, Morocco. To ensure the reliability of the findings, the study employed a robust combination of gully erosion inventory, sentinel images, and Digital Surface Model. Eighteen predictors, encompassing topographical, geomorphological, environmental, and hydrological factors, were selected after multicollinearity analyses. The gully erosion susceptibility of the study revealed that approximately 28.18% of the Tensift catchment is at a very high risk of erosion. Furthermore, 15.13% and 31.28% of the catchment are categorized as low and very low respectively. These findings extend to the Haouz plain, where 7.84% of the surface area are very highly risking erosion, while 18.25% and 55.18% are characterized as low and very low risk areas. To gauge the performance of the ML models, an array of metrics including specificity, precision, sensitivity, and accuracy were employed. The study highlights XGBoost and KNN as the most promising models, achieving AUC ROC values of 0.96 and 0.93 in the test phase. The remaining models namely RF (AUC ROC = 0.89), LR (AUC ROC = 0.80), SVM (AUC ROC = 0.81), DT (AUC ROC = 0.86), and ANN (AUC ROC = 0.78), also displayed commendable performance. The novelty of this research is its innovative approach to combat gully erosion through cutting edge ML models, offering practical solutions for watershed conservation, sustainable management, and the prevention of land degradation. These insights are invaluable for addressing the challenges posed by gully erosion within the region, and beyond its geographical boundaries and can be used for defining appropriate mitigation strategies at local to national scale.

The association between health behaviors and sleep patterns in youths has been tested using cross-sectional data, and prospective evidence is needed. The aim of this study was to investigate the cross-sectional and prospective associations between behavioral variables and sleep quality and duration. A secondary analysis was conducted using data that were collected in March and December 2006 from a random sample of high school students (aged 14-24 years) from Recife and Florianopolis, Brazil, for the Saude na Boa intervention. Data on the perception of sleep quality and duration and lifestyle variables were obtained through self-reported questionnaires. Data were analyzed using crude and adjusted binary logistic regressions. The percentages of students reporting poor sleep quality and insufficient sleep duration were 45.7% and 76.7%, respectively, considering the cross-sectional data, and 45.8% and 77.5%, respectively, considering the longitudinal data. There was a cross-sectional association between lower physical activity (OR = 0.74, 95% CI: 0.55, 0.99) and higher snack consumption (OR = 1.67, 95% CI: 1.18, 2.36) and negative sleep quality. Excessive TV watching (OR = 0.48, 95% CI: 0.30, 0.75) and higher levels of soft drink consumption (OR = 1.84, 95% CI: 1.19, 2.84) were associated with insufficient sleep duration. However, no behavioral variables were associated with sleep quality and duration in the prospective analysis. The prevalence of self-reported poor sleep quality and duration were similar in the cross-sectional and longitudinal analyses. Some behaviors were associated with sleep patterns in the cross-sectional analyses, but these findings were not confirmed in the prospective analyses.

ABSTRACTBackgroundCoronary artery plaque rupture (PR) is closely associated with immune‐inflammatory responses. The systemic inflammatory index (SII) and the systemic inflammatory response index (SIRI) have shown potential in predicting the occurrence of PR.ObjectiveThis study aims to establish a machine learning (ML) model that integrates baseline patient characteristics, SII, and SIRI to predict PR. The goal is to identify high‐risk PR patients before intravascular imaging examinations.MethodsWe included 337 patients with acute coronary syndrome who underwent emergency percutaneous coronary intervention and coronary optical coherence tomography (OCT) at the Affiliated Hospital of Zunyi Medical University, China, from May 2023 to October 2023. PR was determined by OCT images. Through manual feature selection, nine features, including SII and SIRI, were included, and an ML model was built using the XGBoost algorithm. Model performance was evaluated using receiver operating characteristic curves and calibration curves. SHAP values were used to assess the contribution of each feature to the model.ResultsThe ML model demonstrated a higher area under the curve value (AUC = 0.81) compared to using SII or SIRI alone for prediction. The ML model also showed good calibration. SHAP values revealed that the top three features in the ML model were SII, LDL‐C, and SIRI.ConclusionThe immuno‐inflammatory index, which integrates comprehensive clinical characteristics, can predict the occurrence of PR. However, large‐scale, multicenter studies are needed to confirm the generalizability of the predictive model.

The preoperative classification of pleomorphic adenomas (PMA) and Warthin tumors (WT) in the parotid gland plays an essential role in determining therapeutic strategies. This study aims to develop and validate an ultrasound-based ensemble machine learning (USEML) model, employing nonradiative and noninvasive features to differentiate PMA from WT. A total of 203 patients with histologically confirmed PMA or WT who underwent parotidectomy from two centers were enrolled. Clinical factors, ultrasound (US) features, and radiomic features were extracted to develop three types of machine learning model: clinical models, US models, and USEML models. The diagnostic performance of the USEML model, as well as that of physicians based on experience, was evaluated and validated using receiver operating characteristic (ROC) curves in internal and external validation cohorts. DeLong's test was used for comparisons of AUCs. SHAP values were also utilized to explain the classification model. The USEML model achieved the highest AUC of 0.891 (95% CI, 0.774-0.961), surpassing the AUCs of both the US (0.847; 95% CI, 0.720-0.932) and clinical (0.814; 95% CI, 0.682-0.908) models. The USEML model also outperformed physicians in both internal and external validation datasets (both p < 0.05). The sensitivity, specificity, negative predictive value, and positive predictive value of the USEML model and physician experience were 89.3%/75.0%, 87.5%/54.2%, 87.5%/65.6%, and 89.3%/65.0%, respectively. The USEML model, incorporating clinical factors, ultrasound factors, and radiomic features, demonstrated efficient performance in distinguishing PMA from WT in the parotid gland. This study developed a machine learning model for preoperative diagnosis of pleomorphic adenoma and Warthin tumor in the parotid gland based on clinical, ultrasound, and radiomic features. Furthermore, it outperformed physicians in an external validation dataset, indicating its potential for clinical application. • Differentiating pleomorphic adenoma (PMA) and Warthin tumor (WT) affects management decisions and is currently done by invasive biopsy. • Integration of US-radiomic, clinical, and ultrasound findings in a machine learning model results in improved diagnostic accuracy. • The ultrasound-based ensemble machine learning (USEML) model consistently outperforms physicians, suggesting its potential applicability in clinical settings.

An empirical assessment of ML models for 5G network intrusion detection: A data leakage-free approach

Objectives: This study examines the relationship between nutrition, lifestyle habits, and perceived sleep quality in a cross-sectional analysis of 22,181 Spanish adults. Methods: Data were collected between August 2020 and November 2021 using the Nutritional and Social Healthy Habits (NutSo-HH) questionnaire, which assessed variables such as sleep duration, self-perceived restfulness, dietary patterns, and physical activity. Results: Findings indicate that 48.9% of participants sleep 7-8 h per night, while 8.6% sleep less than 6 h. Approximately 50% report frequently feeling rested, whereas 45.4% seldom or sometimes feel rested. Non-parametric Mann-Whitney and Kruskal-Wallis tests with Dwass-Steel-Critchlow-Fligner (DSCF) correction revealed that perceived sleep quality had an average score of 3.39 on a 0-5 scale, with significant differences based on socio-demographic and lifestyle factors (p < 0.001 for sex, age, education, income, and living in a family). Participants with sufficient sleep reported a lower BMI, a higher nutritional index, and more weekly physical activity. A network analysis demonstrated strong clustering between sleep variables and eating behaviors. Although causality cannot be established in this observational study, the results suggest that better sleep is associated with the lower consumption of sugary drinks and ultra-processed foods, as well as improved body image and mental health. Conclusions: These findings highlight the interconnectedness of sleep, nutrition, and lifestyle habits, suggesting that targeted interventions in any of these areas could positively influence the others, ultimately improving overall health outcomes.

Machine learning (ML) models are widely used to predict wildfire occurrence and susceptibility (Brys et al., 2025). However, while these models excel at prediction, they often fail to provide insights into their inner workings or uncover the causal pathways driving wildfires. This study addresses this limitation by extending ML models beyond prediction to explore the drivers and causal pathways underlying wildfire occurrence. Our primary aim is to identify meaningful, interpretable patterns from wildfire data.We developed a novel multi-stage clustering methodology inspired by Cooper et al. (2021) and Cohen et al. (2024). This approach integrates feature attribution (SHAP values), dimensionality reduction (UMAP), hierarchical clustering (HDBSCAN), and causal discovery methods: PC and FCI (Spirtes et al., 2001), and DirectLiNGAM (Shimizu et al., 2011). The causal methods were enhanced with prior background knowledge to derive meaningful insights. We used datasets from Italy (Cilli et al., 2022) and the Netherlands.A central feature of our methodology is the use of SHAP values to define subgroups and derive causal pathways. SHAP values reduce noise in the feature space while preserving critical information for clustering. By reducing multidimensional SHAP values to two dimensions with UMAP, we improved clustering performance and interpretability. The resulting clusters were described using concise, non-overlapping decision rules based on the original variables, eliminating the need for manual filtering commonly required in clustering raw feature space. The identified clusters revealed specific relationships between wildfire drivers and occurrence. For each cluster, we applied advanced causal discovery techniques to derive probable causal pathways, aligning the findings with the knowledge of stakeholders and domain experts. These actionable and interpretable explanations offer practical utility.Findings from the case studies demonstrate that supervised clustering effectively characterizes wildfire occurrence by linking it to influencing factors. Furthermore, the approach provides valuable insights into cluster-specific causal pathways. The methodology translates complex relationships into simple causal logic, offering stakeholders and domain experts the necessary context to understand the model's behavior.&amp;#160;Brys, C., La Red Mart&amp;#237;nez, D.L. &amp;amp; Marinelli, M. Machine learning methods for wildfire risk assessment. Earth Science Informatics 18, 148 (2025). https://doi.org/10.1007/s12145-024-01690-zCilli, R., Elia, M., D&amp;#8217;Este, M., Giannico, V., Amoroso, N., Sanesi, G., Lombardi, A., Pantaleo, E., Monaco, A., Tangaro, S., Bellotti, R. &amp;amp; Lafortezza, R. (2022). Explainable artificial intelligence (XAI) detects wildfire occurrence in the Mediterranean countries of Southern Europe. Scientific Reports 12, 16349. https://doi.org/10.1038/s41598-022-20347-9Cohen, J., Huan, X. &amp;amp; Ni, J. (2024). Shapley-based explainable AI for clustering applications in fault diagnosis and prognosis. Journal of Intelligent Manufacturing, 35, 4071-4086. https://doi.org/10.1007/s10845-024-02468-2Cooper, A., Doyle, O. &amp;amp; Bourke, A. (2021). Supervised clustering for subgroup discovery: An application to covid-19 symptomatology. Communications in Computer and Information Science, 1525, 408&amp;#8211;422. https://doi.org/10.1007/978-3-030-93733-1_29Shimizu, S., Inazumi, T., Sogawa, Y., Hyv&amp;#228;rinen, A., Kawahara, Y., Washio, T., Hoyer, P. O., &amp;amp; Bollen, K. (2011). DirectLiNGAM: A direct method for learning a linear non-Gaussian structural equation model. The Journal of Machine Learning Research, 12, 1225&amp;#8211;1248. https://doi.org/10.48550/arXiv.1101.2489Spirtes, P., Glymour, C. &amp;amp; Scheines, R. (2001). Causation, Prediction, and Search. Second Edition. MIT Press. https://doi.org/10.7551/mitpress/1754.001.0001&amp;#160;

AimTo investigate whether pain, sleep duration, insomnia, sleepiness, work-related factors, anxiety, and depression associate with excessive fatigue in nurses.BackgroundFatigue among nurses is a problem in the context of ongoing nursing shortages. While myriad factors are associated with fatigue not all relationships are understood. Prior studies have not examined excessive fatigue in the context of pain, sleep, mental health, and work factors in a working population to determine if associations between excessive fatigue and each of these factors remain when adjusting for each other.MethodsA cross-sectional questionnaire study among 1,335 Norwegian nurses. The questionnaire included measures for fatigue (Chalder Fatigue Questionnaire, score ≥4 categorized as excessive fatigue), pain, sleep duration, insomnia (Bergen Insomnia Scale), daytime sleepiness (Epworth Sleepiness Scale), anxiety and depression (Hospital Anxiety and Depression Scale), and work-related factors. Associations between the exposure variables and excessive fatigue were analyzed using chi-square tests and logistic regression analyses.ResultsIn the fully adjusted model, significant associations were found between excessive fatigue and pain severity scores for arms/wrists/hands (adjusted OR (aOR) = 1.09, CI = 1.02–1.17), hips/legs/knees/feet (aOR = 1.11, CI = 1.05–1.18), and headaches/migraines (aOR = 1.16, CI = 1.07–1.27), sleep duration of <6 hours (aOR = 2.02, CI = 1.08–3.77), and total symptom scores for insomnia (aOR = 1.05, CI = 1.03–1.08), sleepiness (aOR = 1.11, CI = 1.06–1.17), anxiety (aOR = 1.09, CI = 1.03–1.16), and depression (aOR = 1.24, CI = 1.16–1.33). The musculoskeletal complaint-severity index score (aOR = 1.27, CI = 1.13–1.42) was associated with excessive fatigue in a separate model adjusted for all variables and demographics. Excessive fatigue was also associated with shift work disorder (OR = 2.25, CI = 1.76–2.89) in a model adjusted for demographics. We found no associations with shift work, number of night shifts and number of quick returns (<11 hours between shifts) in the fully adjusted model.ConclusionExcessive fatigue was associated with pain, sleep- and mental health-factors in a fully adjusted model.