โก Quick Summary
This study developed and validated machine learning models to screen for suicidal ideation and depression in individuals with subthreshold insomnia, utilizing data from a Slovenian community sample of 2,989 participants. The models demonstrated strong predictive performance, with AUROC values of 0.78 and 0.79 for suicidal ideation and depression, respectively, indicating their potential for early detection in a high-risk population.
๐ Key Details
- ๐ Dataset: 2,989 participants from a Slovenian nationwide community sample
- ๐งฉ Features used: Socio-demographics, life satisfaction, behavioral changes, and coping strategies
- โ๏ธ Technology: Logistic regression models for predicting SI and depression
- ๐ Performance: AUROC for SI: 0.78 (insomnia group), 0.80 (non-insomnia group); AUROC for depression: 0.79 (insomnia group), 0.82 (non-insomnia group)
๐ Key Takeaways
- ๐ก Insomnia is a significant risk factor for both depression and suicidality.
- ๐ค Machine learning models can effectively screen for suicidal ideation and depression using indirect predictors.
- ๐ Robust performance of models indicates they are effective regardless of insomnia presence.
- ๐ ๏ธ Early detection tools can lead to timely interventions in high-risk populations.
- ๐ Study highlights the importance of addressing sleep complaints in healthcare settings.
- ๐ Study conducted in Slovenia, showcasing a nationwide approach.
- ๐ Future research could expand on these findings to enhance mental health screening.
๐ Background
Insomnia is increasingly recognized as a critical factor contributing to mental health issues, particularly depression and suicidal ideation. Despite its prevalence, many individuals with sleep complaints remain undiagnosed, leading to a pressing need for effective screening methods. This study aims to bridge that gap by leveraging machine learning to identify at-risk individuals based on indirect indicators.
๐๏ธ Study
The research involved a comprehensive analysis of data collected from a nationwide online questionnaire in Slovenia, encompassing a sample of 2,989 individuals. The study focused on developing logistic regression models to predict suicidal ideation (measured by the SIDAS) and moderate-to-severe depression (measured by the DASS-21) using various indirect predictors, including socio-demographic factors and coping strategies.
๐ Results
The models exhibited strong predictive capabilities, with AUROC values of 0.78 for the suicidal ideation model and 0.79 for the depression model in the insomnia group. In the non-insomnia group, the AUROC values were 0.80 and 0.82, respectively. These results indicate that the models are robust and can effectively identify individuals at risk for both conditions, regardless of insomnia status.
๐ Impact and Implications
The findings from this study underscore the potential of machine learning as a transformative tool in mental health screening. By focusing on individuals with sleep complaints, healthcare providers can utilize these models for early detection, potentially reducing the morbidity and mortality associated with suicidal ideation and depression. This approach not only enhances the efficiency of mental health interventions but also promotes a proactive stance in addressing these critical issues.
๐ฎ Conclusion
This study highlights the promising role of machine learning in the early detection of suicidal ideation and depression among individuals with subthreshold insomnia. The robust performance of the developed models suggests they could serve as valuable tools in clinical settings, facilitating timely interventions and improving mental health outcomes. Continued research in this area is essential to refine these models and expand their applicability across diverse populations.
๐ฌ Your comments
What are your thoughts on the use of machine learning for mental health screening? We would love to hear your insights! ๐ฌ Share your comments below or connect with us on social media:
Prediction of suicidal ideation and depression in the general population with subthreshold insomnia using machine learning models.
Abstract
BACKGROUND: Insomnia is a significant independent risk factor for depression and suicidality. However, these conditions often go undetected, particularly in individuals presenting with sleep complaints. This study aimed to develop and validate machine learning (ML) models for the indirect screening of suicidal ideation (SI) and depression and to specifically evaluate their performance in a population reporting at least subthreshold insomnia.
METHODS: Data were obtained from a Slovenian nationwide community sample (Nโ=โ2,989) via an online questionnaire. Logistic regression models were developed to predict SI (measured by SIDAS) and moderate-to-severe depression (measured by DASS-21) via indirect predictors, including socio-demographics, life satisfaction, behavioral changes, and 14 coping strategies from the Brief COPE inventory. The model performance was tested on a validation sample, which was stratified into groups with (Insomnia Severity Index [ISI] scoreโโฅโ8; nโ=โ917) and without (ISIโ<โ8; nโ=โ819) insomnia symptoms.
RESULTS: The models demonstrated strong and consistent predictive performance across both groups. The area under the receiver operating characteristic curve (AUROC) for the SI model was 0.78 in the insomnia group and 0.80 in the non-insomnia group. For the depression model, the AUROCs were 0.79 and 0.82, respectively. The minimal difference in performance indicates that the models are robust and equally effective regardless of the presence of insomnia.
CONCLUSION: Our findings demonstrate that ML models using indirect questions can effectively screen for both suicidality and depression simultaneously. The models' robust performance in individuals with insomnia highlights their potential as feasible, ethical, and efficient tools for early detection. Given that sleep complaints are a common reason for seeking healthcare, this approach offers a critical opportunity for timely intervention in a high-risk population, potentially reducing preventable morbidity and mortality associated with suicide and depression.
Author: [‘Prelog PR’, ‘Matiฤ T’, ‘Pregelj P’, ‘Sadikov A’]
Journal: BMC Psychiatry
Citation: Prelog PR, et al. Prediction of suicidal ideation and depression in the general population with subthreshold insomnia using machine learning models. Prediction of suicidal ideation and depression in the general population with subthreshold insomnia using machine learning models. 2025; 25:1003. doi: 10.1186/s12888-025-07451-6