Prediction of estimated risk for bipolar disorder using machine learning and structural MRI features.

BACKGROUND: Individuals with bipolar disorder are commonly correctly diagnosed a decade after symptom onset. Machine learning techniques may aid in early recognition and reduce the disease burden. As both individuals at risk and those with a manifest disease display structural brain markers, structural magnetic resonance imaging may provide relevant classification features. METHODS: Following a pre-registered protocol, we trained linear support vector machine (SVM) to classify individuals according to their estimated risk for bipolar disorder using regional cortical thickness of help-seeking individuals from seven study sites (N = 276). We estimated the risk using three state-of-the-art assessment instruments (BPSS-P, BARS, EPIbipolar). RESULTS: For BPSS-P, SVM achieved a fair performance of Cohen's κ of 0.235 (95% CI 0.11-0.361) and a balanced accuracy of 63.1% (95% CI 55.9-70.3) in the 10-fold cross-validation. In the leave-one-site-out cross-validation, the model performed with a Cohen's κ of 0.128 (95% CI -0.069 to 0.325) and a balanced accuracy of 56.2% (95% CI 44.6-67.8). BARS and EPIbipolar could not be predicted. In post hoc analyses, regional surface area, subcortical volumes as well as hyperparameter optimization did not improve the performance. CONCLUSIONS: Individuals at risk for bipolar disorder, as assessed by BPSS-P, display brain structural alterations that can be detected using machine learning. The achieved performance is comparable to previous studies which attempted to classify patients with manifest disease and healthy controls. Unlike previous studies of bipolar risk, our multicenter design permitted a leave-one-site-out cross-validation. Whole-brain cortical thickness seems to be superior to other structural brain features.

Young people at risk for developing bipolar disorder: Two-year findings from the multicenter prospective, naturalistic Early-BipoLife study.

Early identification and intervention of individuals with an increased risk for bipolar disorder (BD) may improve the course of illness and prevent long­term consequences. Early-BipoLife, a multicenter, prospective, naturalistic study, examined risk factors of BD beyond family history in participants aged 15-35 years. At baseline, positively screened help-seeking participants (screenBD at-risk) were recruited at Early Detection Centers and in- and outpatient depression and attention-deficit/hyperactivity disorder (ADHD) settings, references (Ref) drawn from a representative cohort. Participants reported sociodemographics and medical history and were repeatedly examined regarding psychopathology and the course of risk factors. N = 1,083 screenBD at-risk and n = 172 Ref were eligible for baseline assessment. Within the first two years, n = 31 screenBD at-risk (2.9 %) and none of Ref developed a manifest BD. The cumulative transition risk was 0.0028 at the end of multistep assessment, 0.0169 at 12 and 0.0317 at 24 months (p = 0.021). The transition rate with a BD family history was 6.0 %, 4.7 % in the Early Phase Inventory for bipolar disorders (EPIbipolar), 6.6 % in the Bipolar Prodrome Interview and Symptom Scale-Prospective (BPSS-FP) and 3.2 % with extended Bipolar At-Risk - BARS criteria). In comparison to help-seeking young patients from psychosis detection services, transition rates in screenBD at-risk participants were lower. The findings of Early-BipoLife underscore the importance of considering risk factors beyond family history in order to improved early detection and interventions to prevent/ameliorate related impairment in the course of BD. Large long-term cohort studies are crucial to understand the developmental pathways and long-term course of BD, especially in people at- risk.

Machine Learning Prediction of Estimated Risk for Bipolar Disorders Using Hippocampal Subfield and Amygdala Nuclei Volumes.

The pathophysiology of bipolar disorder (BD) remains mostly unclear. Yet, a valid biomarker is necessary to improve upon the early detection of this serious disorder. Patients with manifest BD display reduced volumes of the hippocampal subfields and amygdala nuclei. In this pre-registered analysis, we used structural MRI (n = 271, 7 sites) to compare volumes of hippocampus, amygdala and their subfields/nuclei between help-seeking subjects divided into risk groups for BD as estimated by BPSS-P, BARS and EPIbipolar. We performed between-group comparisons using linear mixed effects models for all three risk assessment tools. Additionally, we aimed to differentiate the risk groups using a linear support vector machine. We found no significant volume differences between the risk groups for all limbic structures during the main analysis. However, the SVM could still classify subjects at risk according to BPSS-P criteria with a balanced accuracy of 66.90% (95% CI 59.2-74.6) for 10-fold cross-validation and 61.9% (95% CI 52.0-71.9) for leave-one-site-out. Structural alterations of the hippocampus and amygdala may not be as pronounced in young people at risk; nonetheless, machine learning can predict the estimated risk for BD above chance. This suggests that neural changes may not merely be a consequence of BD and may have prognostic clinical value.

Greater Emotional Distress Due to Social Distancing and Greater Symptom Severity during the COVID-19 Pandemic in Individuals with Bipolar Disorder: A Multicenter Study in Austria, Germany, and Denmark.

Throughout the COVID-19 pandemic, mental health of individuals with bipolar disorders (BD) is potentially more vulnerable, especially regarding COVID-19-related regulations and associated symptomatic changes. A multicentric online study was conducted in Austria, Germany, and Denmark during the COVID-19 pandemic. Overall, data from 494 participants were collected (203 individuals with BD, 291 healthy controls (HC)). Participants filled out questionnaires surveying emotional distress due to social distancing, fear of COVID-19, and the Brief Symptom Inventory-18 to assess symptom severity at four points of measurement between 2020 and 2021. General linear mixed models were calculated to determine the difference between the groups in these pandemic specific factors. Individuals with BD reported higher distress due to social distancing than HC, independently of measurement times. Fear of COVID-19 did not differ between groups; however, it was elevated in times of higher infection and mortality due to COVID-19. Individuals with BD reported higher psychiatric symptom severity than HC; however, symptom severity decreased throughout the measured time in the pandemic. Overall, individuals with BD experienced more distress due to the COVID-19 situation than HC. A supportive mental health system is thus recommended to ensure enhanced care, especially in times of strict COVID-19-related regulations.

Psychological and behavioral response on the COVID-19 pandemic in individuals with bipolar disorder: A multicenter study.

The COVID-19 pandemic affects both mentally healthy and ill individuals. Individuals with bipolar disorder (BD) constitute an especially vulnerable group. A multicentric online study was conducted in Austria, Denmark, and Germany after the first lockdown phase in 2020. In total, 117 healthy controls (HC) were matched according to age and sex to 117 individuals with BD. The survey included the Brief Symptom Inventory-18, Beck Depression Inventory-2, Pittsburgh Sleep Quality Index, and a self-constructed questionnaire assessing COVID-19 fears, emotional distress due to social distancing, lifestyle, and compliance to governmental measures. In individuals with BD, increased symptoms of depression, somatization, anxiety, distress due to social distancing, and poorer sleep quality were related to emotional distress due to social distancing. The correlation between emotional distress due to social distancing and anxiety showed 26% of shared variance in BD and 11% in HC. Negative lifestyle changes and lower compliance with COVID-19 regulatory measures were more likely to be observed in individuals with BD than in HC. These findings underscore the need for ongoing mental health support during the pandemic. Individuals with BD should be continuously supported during periods of social distancing to maintain a stable lifestyle and employ strategies to cope with COVID-19 fears.

Physical Health Profile and Associated Behavior During the COVID-19 Pandemic in Patients With Bipolar Disorder.

Background: The COVID-19 pandemic has led to an increased psychological strain on public mental health and may impact behavioral, mental, and physical health, presumably with effects on patients with severe mental disorders. This study examines pandemic-related physical and mental health and (compensatory) behavioral changes, in patients with BD as compared to healthy control individuals. Method: Physical and mental health and self-reported changes in daily structure and behavior due to the pandemic were assessed using a self-constructed questionnaire and the brief symptom inventory (BSI) in Germany, Austria, and Denmark in individuals with BD and a healthy control group. Results: The present study included 118 individuals with BD and 215 healthy controls. Individuals with BD reported statistically significant higher physical risk burden, increased weight gain, more physical comorbidities, and a decrease in physical activity and they further reported higher rates of COVID-19 testing, had more worries concerning health, and experienced more anxiety but less social distancing. Conclusion: The COVID-19 pandemic seems to have a greater impact on physical health in individuals with BD than in healthy controls. Individuals with BD appear to be having more difficulties compensating their behavior due to the pandemic which could amplify the effect of risk factors associated with poorer physical health. This highlights the necessity for optimizing and targeting the overall treatment of both mental and physical health in patients with BD during periods with far-reaching changes such as the COVID-19 pandemic. Limitations: Sampling issues and self-report forms, selectivity (missing elderly, and those lacking access or knowledge of technology).

Individuals at increased risk for development of bipolar disorder display structural alterations similar to people with manifest disease.

In psychiatry, there has been a growing focus on identifying at-risk populations. For schizophrenia, these efforts have led to the development of early recognition and intervention measures. Despite a similar disease burden, the populations at risk of bipolar disorder have not been sufficiently characterized. Within the BipoLife consortium, we used magnetic resonance imaging (MRI) data from a multicenter study to assess structural gray matter alterations in N = 263 help-seeking individuals from seven study sites. We defined the risk using the EPIbipolar assessment tool as no-risk, low-risk, and high-risk and used a region-of-interest approach (ROI) based on the results of two large-scale multicenter studies of bipolar disorder by the ENIGMA working group. We detected significant differences in the thickness of the left pars opercularis (Cohen's d = 0.47, p = 0.024) between groups. The cortex was significantly thinner in high-risk individuals compared to those in the no-risk group (p = 0.011). We detected no differences in the hippocampal volume. Exploratory analyses revealed no significant differences in other cortical or subcortical regions. The thinner cortex in help-seeking individuals at risk of bipolar disorder is in line with previous findings in patients with the established disorder and corresponds to the region of the highest effect size in the ENIGMA study of cortical alterations. Structural alterations in prefrontal cortex might be a trait marker of bipolar risk. This is the largest structural MRI study of help-seeking individuals at increased risk of bipolar disorder.

Early detection of bipolar disorders and treatment recommendations for help-seeking adolescents and young adults: Findings of the Early Detection and Intervention Center Dresden.

BACKGROUND: Early identification and intervention of individuals with risk factors for or subtle prodromal symptoms of bipolar disorders (BD) may improve the illness course and prevent adverse long-term consequences. METHODS: We examined sociodemographic, clinical and psychopathological characteristics of help-seeking adolescents and young adults who consulted the Early Detection and Intervention Center Dresden at the University of Dresden (Germany) and presented with or without pre-defined at-risk criteria for BD. The standardized diagnostic procedure for all help-seeking youth included a comprehensive psychiatric history and a structured clinical interview. When BD at-risk state was suspected, early detection instruments (EPIbipolar, BPSS-FP) were applied. Treatment recommendations were formulated in multi-professional case conferences. RESULTS: Out of 890 help-seeking persons between 05/2009 and 04/2018, 582 (65%) completed the diagnostic process. Of these, 24 (4%) had manifest BD and 125 (21%) fulfilled at-risk BD criteria (age = 23.9 ± 0.6 years, female = 62%). Of the pre-defined main risk factors, family history for BD was reported in 22% of the at-risk persons, (hypo-)mania risk state in 44%, and increasing cyclothymic mood swings with increased activity in 48%. The most common secondary risk factors were decreased psychosocial functioning (78%), lifetime diagnosis of depressive disorder (67%) and specific sleep/circadian rhythm disturbances (59%). Substance use was very common in subjects at-risk for BD (cannabis = 50%, alcohol = 33%) and highest in patients with BD (cannabis = 75%, alcohol = 40%). Psychiatric treatment history, including psychopharmacological therapy, was similar between the groups, while treatment recommendations differed, with more advice for psychotherapy and antidepressants in the at-risk group with a lifetime diagnosis of depression and more advice for specialized BD treatment including mood stabilizers in patients with BD. CONCLUSION: This analysis on the phenomenology of different BD at-risk stages suggests that early detection of individuals presenting with suggested risk factors for the development of BD is feasible in help-seeking young people. Future research should further develop/test stage-specific prevention and early targeted intervention approaches that were described in a naturalistic setting.

Improving early recognition and intervention in people at increased risk for the development of bipolar disorder: study protocol of a prospective-longitudinal, naturalistic cohort study (Early-BipoLife).

BACKGROUND: Bipolar disorders (BD) belong to the most severe mental disorders, characterized by an early onset and recurrent, severe episodes or a chronic course with poor psychosocial functioning in a proportion of patients. Many patients with BD experience substantial symptomatology months or even years before full BD manifestation. Adequate diagnosis and treatment is often delayed, which is associated with a worse outcome. This study aims to prospectively evaluate and improve early recognition and intervention strategies for persons at-risk for BD. METHODS: Early-BipoLife is a prospective-longitudinal cohort study of 1419 participants (aged 15-35 years) with at least five waves of assessment over a period of at least 2 years (baseline, 6, 12, 18 and 24 months). A research consortium of ten university and teaching hospitals across Germany conducts this study. The following risk groups (RGs) were recruited: RG I: help-seeking youth and young adults consulting early recognition centres/facilities presenting ≥ 1 of the proposed risk factors for BD, RG II: in-/outpatients with unipolar depressive syndrome, and RG III: in-/outpatients with attention-deficit/hyperactivity disorder (ADHD). The reference cohort was selected from the German representative IMAGEN cohort. Over the study period, the natural course of risk and resilience factors, early symptoms of BD and changes of symptom severity (including conversion to manifest BD) are observed. Psychometric properties of recently developed, structured instruments on potential risk factors for conversion to BD and subsyndromal symptomatology (Bipolar Prodrome Symptom Scale, Bipolar at-risk criteria, EPIbipolar) and biomarkers that potentially improve prediction are investigated. Moreover, actual treatment recommendations are monitored in the participating specialized services and compared to recently postulated clinical categorization and treatment guidance in the field of early BD. DISCUSSION: Findings from this study will contribute to an improved knowledge about the natural course of BD, from the onset of first noticeable symptoms (precursors) to fully developed BD, and about mechanisms of conversion from subthreshold to manifest BD. Moreover, these generated data will provide information for the development of evidence-based guidelines for early-targeted detection and preventive intervention for people at risk for BD.