Predicting criminal and violent outcomes in psychiatry: a meta-analysis of diagnostic accuracy.

Although reducing criminal outcomes in individuals with mental illness have long been a priority for governments worldwide, there is still a lack of objective and highly accurate tools that can predict these events at an individual level. Predictive machine learning models may provide a unique opportunity to identify those at the highest risk of criminal activity and facilitate personalized rehabilitation strategies. Therefore, this systematic review and meta-analysis aims to describe the diagnostic accuracy of studies using machine learning techniques to predict criminal and violent outcomes in psychiatry. We performed meta-analyses using the mada, meta, and dmetatools packages in R to predict criminal and violent outcomes in psychiatric patients (n = 2428) (Registration Number: CRD42019127169) by searching PubMed, Scopus, and Web of Science for articles published in any language up to April 2022. Twenty studies were included in the systematic review. Overall, studies used single-nucleotide polymorphisms, text analysis, psychometric scales, hospital records, and resting-state regional cerebral blood flow to build predictive models. Of the studies described in the systematic review, nine were included in the present meta-analysis. The area under the curve (AUC) for predicting violent and criminal outcomes in psychiatry was 0.816 (95% Confidence Interval (CI): 70.57-88.15), with a partial AUC of 0.773, and average sensitivity of 73.33% (95% CI: 64.09-79.63), and average specificity of 72.90% (95% CI: 63.98-79.66), respectively. Furthermore, the pooled accuracy across models was 71.45% (95% CI: 60.88-83.86), with a tau squared (τ2) of 0.0424 (95% CI: 0.0184-0.1553). Based on available evidence, we suggest that prospective models include evidence-based risk factors identified in prior actuarial models. Moreover, there is a need for a greater emphasis on identifying biological features and incorporating novel variables which have not been explored in prior literature. Furthermore, available models remain preliminary, and prospective validation with independent datasets, and across cultures, will be required prior to clinical implementation. Nonetheless, predictive machine learning models hold promise in providing clinicians and researchers with actionable tools to improve how we prevent, detect, or intervene in relevant crime and violent-related outcomes in psychiatry.

Differential power of placebo across major psychiatric disorders: a preliminary meta-analysis and machine learning study.

The placebo effect across psychiatric disorders is still not well understood. In the present study, we conducted meta-analyses including meta-regression, and machine learning analyses to investigate whether the power of placebo effect depends on the types of psychiatric disorders. We included 108 clinical trials (32,035 participants) investigating pharmacological intervention effects on major depressive disorder (MDD), bipolar disorder (BD) and schizophrenia (SCZ). We developed measures based on clinical rating scales and Clinical Global Impression scores to compare placebo effects across these disorders. We performed meta-analysis including meta-regression using sample-size weighted bootstrapping techniques, and machine learning analysis to identify the disorder type included in a trial based on the placebo response. Consistently through multiple measures and analyses, we found differential placebo effects across the three disorders, and found lower placebo effect in SCZ compared to mood disorders. The differential placebo effects could also distinguish the condition involved in each trial between SCZ and mood disorders with machine learning. Our study indicates differential placebo effect across MDD, BD, and SCZ, which is important for future neurobiological studies of placebo effects across psychiatric disorders and may lead to potential therapeutic applications of placebo on disorders more responsive to placebo compared to other conditions.

Accelerated aging signatures in subjects with schizophrenia and their unaffected siblings.

Schizophrenia (SZ) is a chronic debilitating disease. Subjects with SZ have significant shorter life expectancy. Growing evidence suggests that a process of pathological accelerated aging occurs in SZ, leading to early development of severe clinical diseases and worse morbimortality. Furthermore, unaffected relatives can share certain endophenotypes with subjects with SZ. We aim to characterize accelerated aging as a possible endophenotype of schizophrenia by using a machine learning (ML) model of peripheral biomarkers to accurately differentiate subjects with SZ (n = 35), their unaffected siblings (SB, n = 36) and healthy controls (HC, n = 47). We used a random forest algorithm that included biomarkers related to aging: eotaxins CCL-11 and CCL-24; the oxidative stress markers thiobarbituric acid-reactive substances (TBARS), protein carbonyl content (PCC), glutathione peroxidase (GPx); and telomere length (TL). The ML algorithm of biomarkers was able to distinguish individuals with SZ from HC with prediction accuracy of 79.7%, SZ from SB with 62.5% accuracy and SB from HC with 75.5% accuracy. These results support the hypothesis that a pathological accelerated aging might occur in SZ, and this pathological aging could be an endophenotype of the disease, once this profile was also observed in SB, suggesting that SB might suffer from an accelerated aging in some level.

Prediction of depression cases, incidence, and chronicity in a large occupational cohort using machine learning techniques: an analysis of the ELSA-Brasil study.

Abstract. BACKGROUND: Depression is highly prevalent and marked by a chronic and recurrent course. Despite being a major cause of disability worldwide, little is known regarding the determinants of its heterogeneous course. Machine learning techniques present an opportunity to develop tools to predict diagnosis and prognosis at an individual level. METHODS: We examined baseline (2008-2010) and follow-up (2012-2014) data of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), a large occupational cohort study. We implemented an elastic net regularization analysis with a 10-fold cross-validation procedure using socioeconomic and clinical factors as predictors to distinguish at follow-up: (1) depressed from non-depressed participants, (2) participants with incident depression from those who did not develop depression, and (3) participants with chronic (persistent or recurrent) depression from those without depression. RESULTS: We assessed 15 105 and 13 922 participants at waves 1 and 2, respectively. The elastic net regularization model distinguished outcome levels in the test dataset with an area under the curve of 0.79 (95% CI 0.76-0.82), 0.71 (95% CI 0.66-0.77), 0.90 (95% CI 0.86-0.95) for analyses 1, 2, and 3, respectively. CONCLUSIONS: Diagnosis and prognosis related to depression can be predicted at an individual subject level by integrating low-cost variables, such as demographic and clinical data. Future studies should assess longer follow-up periods and combine biological predictors, such as genetics and blood biomarkers, to build more accurate tools to predict depression course.

Structural and Functional Brain Correlates of Neuroprogression in Bipolar Disorder.

Neuroprogression is associated with structural and functional brain changes that occur in parallel with cognitive and functioning impairments. There is substantial evidence showing early white matter changes, as well as trajectory-related gray matter alterations. Several structures, including prefrontal, parietal, temporal cortex, and limbic structures, seem to be altered over the course of bipolar disorder, especially associated with the number of episodes and length of the disease. An important limitation is that most of the studies used either a cross-sectional design or a short follow-up period, which may be insufficient to identify all neuroprogressive changes over time. In addition, the heterogeneity of patients with bipolar disorder is another challenge to determine which subjects will have a more pernicious trajectory. Larger studies and the use of new techniques, such as machine learning, may help to enable more discoveries and evidence on the role of neuroprogression in BD.

The Early Burden of Disability in Individuals With Mood and Other Common Mental Disorders in Ontario, Canada.

Importance: Large population-based data on the trajectory to disability after the first diagnosis of a mood disorder are lacking. Objective: To assess the time between an incident mood disorder diagnosis and the receipt of disability services during a follow-up period of as long as 20 years. Design, Setting, and Participants: This cohort study used health administrative and social service data from ICES for 1 902 792 adults aged 18 to 59 years living in Ontario, Canada. A narrow cohort of individuals who had a new diagnosis of a mood disorder between October 1, 1997, and March 31, 2007, matched by sex and age to individuals with no history of mood disorder, included 278 296 participants. A broader cohort of individuals who had a new diagnosis of other common mental disorders during the same period, matched by sex and age to individuals with no history of mental disorder diagnosis, included 1 624 496 individuals. All individuals were followed up to a maximum end date of March 31, 2017. Data analysis was conducted from November 2017 to June 2018. Exposure: Incident diagnosis of mood or common mental disorder. Main Outcomes and Measures: Disability outcomes were as follows: (1) entry into the Ontario Disability Support Program (ODSP), signifying long-term inability to work because of a disability, and (2) admission into a long-term care (LTC) residence, signifying the inability to live in independent housing. Cox proportional hazards models were used. Results: In the full cohort of 1 902 792 individuals, 278 296 participants (14.6%) were included in the mood disorder cohort (mean [SD] age, 37.5 [11.9] years; 157 386 [56.6%] women), and 1 624 496 participants (85.4%) were included in the common mental disorder cohort (mean [SD], 36.5 [11.8] years; 932 545 [57.4%] women). The incidence of ODSP initiation was greater among individuals with mood disorders than those without (51.5 per 10 000 person-years vs 25.5 per 10 000 person-years; adjusted hazard ratio [aHR], 2.03; 95% CI, 1.95-2.11) and for those with common mental disorders (45.0 per 10 000 person-years vs 27.6 per 10 000 person-years; aHR, 1.57; 95% CI, 1.55-1.60). The aHR for admission to LTC was also higher among individuals with mood disorders compared with those without (aHR, 2.20; 95% CI, 1.80-2.69) and those with common mental disorders compared with those without (aHR, 1.21; 95% CI, 1.14-1.29). Individuals with bipolar disorders had greater ODSP rates than individuals with major depressive disorders (crude rate ratio: 4.31 [95% CI, 3.56-5.17] vs 1.82 [95% CI, 1.36-2.43]). Conclusions and Relevance: This cohort study found that mood disorders were associated with elevated and early rates of disability services. Effective early intervention strategies targeting functional impairment in this population are encouraged.

Differential biomarker signatures in unipolar and bipolar depression: A machine learning approach.

OBJECTIVE: This study used machine learning techniques combined with peripheral biomarker measurements to build signatures to help differentiating (1) patients with bipolar depression from patients with unipolar depression, and (2) patients with bipolar depression or unipolar depression from healthy controls. METHODS: We assessed serum levels of interleukin-2, interleukin-4, interleukin-6, interleukin-10, tumor necrosis factor-α, interferon-γ, interleukin-17A, brain-derived neurotrophic factor, lipid peroxidation and oxidative protein damage in 54 outpatients with bipolar depression, 54 outpatients with unipolar depression and 54 healthy controls, matched by sex and age. Depressive symptoms were assessed using the Hamilton Depression Rating Scale. Variable selection was performed with recursive feature elimination with a linear support vector machine kernel, and the leave-one-out cross-validation method was used to test and validate our model. RESULTS: Bipolar vs unipolar depression classification achieved an area under the receiver operating characteristics (ROC) curve (AUC) of 0.69, with 0.62 sensitivity and 0.66 specificity using three selected biomarkers (interleukin-4, thiobarbituric acid reactive substances and interleukin-10). For the comparison of bipolar depression vs healthy controls, the model retained five variables (interleukin-6, interleukin-4, thiobarbituric acid reactive substances, carbonyl and interleukin-17A), with an AUC of 0.70, 0.62 sensitivity and 0.7 specificity. Finally, unipolar depression vs healthy controls comparison retained seven variables (interleukin-6, Carbonyl, brain-derived neurotrophic factor, interleukin-10, interleukin-17A, interleukin-4 and tumor necrosis factor-α), with an AUC of 0.74, a sensitivity of 0.68 and 0.70 specificity. CONCLUSION: Our findings show the potential of machine learning models to aid in clinical practice, leading to more objective assessment. Future studies will examine the possibility of combining peripheral blood biomarker data with other biological data to develop more accurate signatures.

Bullying and psychotic symptoms in youth with bipolar disorder.

BACKGROUND: Childhood trauma is associated with psychosis in adults with bipolar disorder (BD). Although bullying represents a widespread form of childhood trauma, no studies thus far have investigated the association of bullying and psychosis in pediatric bipolar disorder (PBD). We aim to examine the association between psychosis in PBD with bullying victimization. METHODS: We included 64 children and adolescents (age± mean= 12±3.43) outpatients with BD spectrum disorders. Psychiatric diagnoses were assessed with the semi- structured interview Schedule for Affective Disorders and Schizophrenia for School Age Children-Present and Lifetime (KSADS-PL) version with additional depression and manic symptom items derived from the Washington University in St. Louis Kiddie Schedule for Affective Disorders (WASH-U-KSADS). Bullying, demographic, and clinical variables were assessed during the clinical interview. RESULTS: A lifetime history of psychotic symptoms was associated with bullying (p = 0.002), suicidal behavior (p = 0.006), low socioeconomic status (p = 0.04), and severity of PBD (p = 0.02). Only bullying (OR = 7.3; 95%CI = 2-32) and suicidal behavior (OR = 7.6; 95%CI = 1.5-47.8) remained significant after adjustment for confounders. In a supplementary analysis, we developed a model using supervised machine learning to identify the most relevant variables that differentiated participants with psychotic symptoms, which included bullying, Clinical Global Impression-Severity scale (CGI-S), and suicidal behavior (accuracy = 75%, [p = 0.03]; sensitivity = 77.91%; specificity = 69.05%; area under the curve [AUC] = 0.86). LIMITATIONS: Small sample, cross-sectional design, and generalizability of findings beyond the outpatient clinical sample. CONCLUSIONS: Findings underscore the importance of assessing bullying in PBD participants. Future longitudinal studies with larger samples are needed to replicate our findings and determine causality.

Machine learning and big data analytics in bipolar disorder: A position paper from the International Society for Bipolar Disorders Big Data Task Force.

OBJECTIVES: The International Society for Bipolar Disorders Big Data Task Force assembled leading researchers in the field of bipolar disorder (BD), machine learning, and big data with extensive experience to evaluate the rationale of machine learning and big data analytics strategies for BD. METHOD: A task force was convened to examine and integrate findings from the scientific literature related to machine learning and big data based studies to clarify terminology and to describe challenges and potential applications in the field of BD. We also systematically searched PubMed, Embase, and Web of Science for articles published up to January 2019 that used machine learning in BD. RESULTS: The results suggested that big data analytics has the potential to provide risk calculators to aid in treatment decisions and predict clinical prognosis, including suicidality, for individual patients. This approach can advance diagnosis by enabling discovery of more relevant data-driven phenotypes, as well as by predicting transition to the disorder in high-risk unaffected subjects. We also discuss the most frequent challenges that big data analytics applications can face, such as heterogeneity, lack of external validation and replication of some studies, cost and non-stationary distribution of the data, and lack of appropriate funding. CONCLUSION: Machine learning-based studies, including atheoretical data-driven big data approaches, provide an opportunity to more accurately detect those who are at risk, parse-relevant phenotypes as well as inform treatment selection and prognosis. However, several methodological challenges need to be addressed in order to translate research findings to clinical settings.

Risk factors for suicidality in patients with panic disorder: A systematic review and meta-analysis.

BACKGROUND: Subjects with panic disorder are nearly 4 times as likely to attempt suicide as compared to subjects without this condition. METHODS: We searched the literature from Jan 1, 1960 to May, 4, 2019. Articles that reported a dichotomous sample of patients with panic disorder with and without suicidal behavior were included. OUTCOMES: Twelve studies with 1958 participants were included. Comorbid depression (k = 3, ES = 4.47 [2.63; 7.60]), depressive symptoms (k = 2, ES = 1.98 [1.26; 3.11]), older age (k = 3, ES = 1.66 [1.32; 2.10]), younger age of panic disorder onset (k = 2, ES = 0.65 [0.45; 0.94]), and history of alcohol dependence (k = 2, ES = 8.70 [1.20; 63.04]) were associated with suicide attempt in panic disorder. Depressive symptoms (k = 2, ES = 2.29 (1.60; 3.37]), anxiety symptoms (k = 2, ES = 1.90 [1.33; 2.69]), longer illness duration (k = 2, ES = 3.31 [1.90; 5.74]), comorbid depressive disorder (k = 4, ES = 3.88 [2.03; 7.41]), agoraphobia (k = 2, ES = 4.60 [1.47; 14.42]) and younger age of onset (k = 2, ES = 0.60 [0.38; 0.96]) were associated with suicidal ideation in panic disorder. INTERPRETATION: Our findings provide a framework for the development of suicide prevention strategies in this population.

Clinical differences between patients with pediatric bipolar disorder with and without a parental history of bipolar disorder.

Pediatric Bipolar Disorder (PBD) is a highly heritable condition responsible for 18% of all pediatric mental health hospitalizations. Despite the heritability of this disorder, few studies have assessed potential differences in the clinical manifestation of PBD among patients with a clear parental history of BD. Additionally, while recent studies suggest that attentional deficits are a potential endophenotypic marker of PBD, it is unclear whether heritability is a relevant contributor to these symptoms. In order to address this gap, the present study assessed 61 youth with PBD (6-17 years old), corresponding to 27 offspring of BD patients, and 31 PBD patients without a parental history of the disorder. All standardized assessments, including the K-SADS-PL-W were performed by trained child and adolescent psychiatrists. We performed a logistic multivariate model using the variables of ADHD, rapid cycling, and lifetime psychosis. Rates of ADHD comorbidity were significantly higher among PBD patients who had a parent with BD. Furthermore, PBD patients who had a parent with BD showed a trend toward significance of earlier symptom onset. PBD offspring did not show increased rates of suicide attempts, rapid cycling, or psychosis. Given these findings, it appears that PBD patients who have a parent with BD may represent a distinct endophenotype of the disorder. Future longitudinal and larger studies are required to confirm our findings.

Correction: Potential use of text classification tools as signatures of suicidal behavior: A proof-of-concept study using Virginia Woolf's personal writings.

[This corrects the article DOI: 10.1371/journal.pone.0204820.].

Potential use of text classification tools as signatures of suicidal behavior: A proof-of-concept study using Virginia Woolf's personal writings.

BACKGROUND: The present study analyzes the feasibility of text classification to predict individual suicidal behavior. Entries from Virginia Woolf's diaries and letters were used to assess whether a text classification algorithm could identify written patterns associated with suicide. METHODS: This is a text classification study. We compared 46 text entries from the two months before Virginia Woolf's suicide with 54 texts randomly selected from Virginia Woolf's work during other periods of her life. Letters and diaries were included, while books, novels, short stories, and article fragments were excluded. The data was analyzed using a Naïve-Bayes machine-learning algorithm. RESULTS: The model showed a balanced accuracy of 80.45%, sensitivity of 69%, and specificity of 91%. The Kappa statistic was 0.6, which means a good agreement, and the p-value of the model was 0.003. The area under the ROC curve (AUC) was 0.80. In other words, the model exhibited good performance when used for classifying Virginia Woolf's diaries and letters. DISCUSSION: The present study showed the feasibility of a machine-learning model coupled with text to identify individual written patterns associated with suicidal behavior. Our text signature was able to identify the period of two months preceding suicide with a high accuracy. This technique may be applied to subjects with psychiatric disorders by means of data captured from social media, e-mail, among others. The algorithm may then predict a specific outcome and enable early intervention by clinicians.

Neuron-glia Interaction as a Possible Pathophysiological Mechanism of Bipolar Disorder.

Accumulating evidence has shown the importance of glial cells in the neurobiology of bipolar disorder. Activated microglia and inflammatory cytokines have been pointed out as potential biomarkers of bipolar disorder. Indeed, recent studies have shown that bipolar disorder involves microglial activation in the hippocampus and alterations in peripheral cytokines, suggesting a potential link between neuroinflammation and peripheral toxicity. These abnormalities may also be the biological underpinnings of outcomes related to neuroprogression, such as cognitive impairment and brain changes. Additionally, astrocytes may have a role in the progression of bipolar disorder, as these cells amplify inflammatory response and maintain glutamate homeostasis, preventing excitotoxicity. The present review aims to discuss neuron-glia interactions and their role in the pathophysiology and treatment of bipolar disorder.

The impact of machine learning techniques in the study of bipolar disorder: A systematic review.

Machine learning techniques provide new methods to predict diagnosis and clinical outcomes at an individual level. We aim to review the existing literature on the use of machine learning techniques in the assessment of subjects with bipolar disorder. We systematically searched PubMed, Embase and Web of Science for articles published in any language up to January 2017. We found 757 abstracts and included 51 studies in our review. Most of the included studies used multiple levels of biological data to distinguish the diagnosis of bipolar disorder from other psychiatric disorders or healthy controls. We also found studies that assessed the prediction of clinical outcomes and studies using unsupervised machine learning to build more consistent clinical phenotypes of bipolar disorder. We concluded that given the clinical heterogeneity of samples of patients with BD, machine learning techniques may provide clinicians and researchers with important insights in fields such as diagnosis, personalized treatment and prognosis orientation.

Neuroprogression and illness trajectories in bipolar disorder.

INTRODUCTION: The longitudinal course of bipolar disorder is highly variable, and a subset of patients seems to present a progressive course associated with brain changes and functional impairment. Areas covered: We discuss the theory of neuroprogression in bipolar disorder. This concept considers the systemic stress response that occurs within mood episodes and late-stage deficits in functioning and cognition as well as neuroanatomic changes. We also discuss treatment refractoriness that may take place in some cases of bipolar disorder. We searched PubMed for articles published in any language up to June 4th, 2016. We found 315 abstracts and included 87 studies in our review. Expert commentary: We are of the opinion that the use of specific pharmacological strategies and functional remediation may be potentially useful in bipolar patients at late-stages. New analytic approaches using multimodal data hold the potential to help in identifying signatures of subgroups of patients who will develop a neuroprogressive course.