First-onset major depression during the COVID-19 pandemic: A predictive machine learning model

Caldirola, Daniela; Daccò, Silvia; Cuniberti, Francesco; Grassi, Massimiliano; Alciati, Alessandra; Torti, Tatiana; Perna, Giampaolo

doi:10.1016/j.jad.2022.04.145

Cited by 14 publications

(8 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It is noteworthy, however, that these studies mostly used imbalanced datasets where the number of negative cases (no future depressive episode) far outnumber the positive cases. For example, in the two studies with the highest prediction performance – i.e., AUROC of 0.87 (Na et al, 2020) and 0.85 (Caldirola et al, 2022) – individuals with future depressive episodes amounted to approx. only 8% and 7% of the respective samples.…”

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

“…When considering those studies that had large sample sizes (i.e. N>500) and tested for generalisability in an independent test set, the reported AUROC values in the literature range between 0.71-0.87 (Caldirola et al, 2022; King et al, 2008; Librenza-Garcia et al, 2021; Na et al, 2020; Xu et al, 2019). It is noteworthy, however, that these studies mostly used imbalanced datasets where the number of negative cases (no future depressive episode) far outnumber the positive cases.…”

Section: Discussionmentioning

confidence: 99%

“…In the recent past, there have been numerous attempts to establish such models both in adolescents and adults, based on combinations of various data types, e.g. demographic, socioeconomic, cognitive, and clinical variables as well as motor activity (Caldirola et al, 2022; Chikersal et al, 2021; Gu et al, 2020; King et al, 2008; Librenza-Garcia et al, 2021; Lin et al, 2022; Na et al, 2020; Rocha et al, 2021; Rosellini et al, 2020; Sampson et al, 2021; van Eeden et al, 2021; Voorhees et al, 2008; Xu et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Predicting Future Depressive Episodes from Resting-State fMRI with Generative Embedding

Galioulline

Frässle

Harrison

et al. 2022

Preprint

View full text Add to dashboard Cite

After a first episode of major depressive disorder (MDD), there is substantial risk for a long-term remitting-relapsing course. Prevention and early interventions are thus critically important. Various studies have examined the feasibility of detecting at-risk individuals based on out-of-sample predictions about the future occurrence of depression. However, functional magnetic resonance imaging (MRI) has received very little attention for this purpose so far. Here, we explored the utility of generative models (i.e. different dynamic causal models, DCMs) as well as functional connectivity (FC) for predicting future episodes of depression in never-depressed adults, using a large dataset (N=906) of task-free ("resting state") fMRI data from the UK Biobank. Connectivity analyses were conducted using timeseries from pre-computed spatially independent components of different dimensionalities. Over a three year period, 50% of participants showed indications of at least one depressive episode, while the other 50% did not. Using nested cross-validation for training and a held-out test set (80/20 split), we systematically examined the combination of 8 connectivity feature sets and 17 classifiers. We found that a generative embedding procedure based on combining regression DCM (rDCM) with a support vector machine (SVM) enabled the best (and statistically significant) predictions, both on the training and the test set. However, on the test set, rDCM (62% accuracy) was only slightly superior to SVM predictions based on FC (59% accuracy). Interpreting model predictions based on SHAP (SHapley Additive exPlanations) values suggested that the most predictive connections were widely distributed and not confined to specific networks; the biological interpretability of predictions was aggravated by the use of IC timeseries. Overall, our analyses suggest (i) ways of improving future fMRI-based generative embedding approaches for the early detection of individuals at-risk for depression and that (ii) achieving accuracies of clinical utility may require combination of fMRI with other data modalities.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predicting Future Depressive Episodes from Resting-State fMRI with Generative Embedding

Galioulline

Frässle

Harrison

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…An online study from Italy found an increase in new-onset major depressive disorders in the first and second wave among the general population ( 72 ). A longitudinal survey among university students in Japan found that 11.8% had a new-onset of depressive symptoms, supporting our findings ( 73 ).…”

Section: Discussionmentioning

confidence: 99%

Increase in presentations with new-onset psychiatric disorders in a psychiatric emergency department in Berlin, Germany during the second wave of the COVID-19 pandemic – a retrospective cross-sectional study

Adam,

Moran,

Kippe

et al. 2023

Front. Psychiatry

View full text Add to dashboard Cite

IntroductionWhile numerous studies have identified an increase in symptoms of depression as well as anxiety and distress due to the COVID-19 pandemic, relatively few studies have investigated the new-onset of psychiatric diseases during the pandemic.MethodsThis study focuses on the number of psychiatric new-onset diagnoses in a psychiatric emergency department (pED) in Berlin, Germany during the second wave of the pandemic (i.e. from 09/15/2020 to 03/01/2021 = COVID-19-period) compared to pre-pandemic times (09/15/2019 to 03/01/2020 = control period). We focused on diagnostic subgroups and performed logistic regression analysis to investigate potential risk groups based on covariables such as age, gender, homelessness, attending in police custody and familial relationship.ResultsOverall, there was a 59.7% increase in new-onset psychiatric diagnoses during the COVID-19-period. Increases in the following diagnoses were observed: new-onset of substance-related and addictive disorders (+192.5%), depressive disorders (+115.8%), schizophrenia spectrum and psychotic disorders (+113.3%) and anxiety disorders (+63.6%). These diagnostic subgroups, together with attending in police custody, were found to predict pED presentations with new-onset during the COVID-19-period. Interestingly, in the group of new-onset psychiatric diseases in the COVID-19-period, higher amounts of job loss and living alone as well as a relative decrease in familial relationships were observed.DiscussionCOVID-19 infections and post-COVID-19 syndrome are unlikely to have played a substantial role in the increase of new-onset diseases in this study. Conclusion: Our findings underline the role of indirect factors in new-onset of psychiatric diseases during the pandemic and should be a caveat for future pandemic control policies.

show abstract

Lifestyle factors and other predictors of common mental disorders in diagnostic machine learning studies: A systematic review

Todd,

Orr,

Gamage

et al. 2025

Computers in Biology and Medicine

View full text Add to dashboard Cite

First-onset major depression during the COVID-19 pandemic: A predictive machine learning model

Cited by 14 publications

References 71 publications

Predicting Future Depressive Episodes from Resting-State fMRI with Generative Embedding

Predicting Future Depressive Episodes from Resting-State fMRI with Generative Embedding

Increase in presentations with new-onset psychiatric disorders in a psychiatric emergency department in Berlin, Germany during the second wave of the COVID-19 pandemic – a retrospective cross-sectional study

Lifestyle factors and other predictors of common mental disorders in diagnostic machine learning studies: A systematic review

Contact Info

Product

Resources

About