Predictors of conversion from major depressive disorder to bipolar disorder

Machine learning-based discrimination of unipolar depression and bipolar disorder with streamlined shortlist in adolescents of different ages

BACKGROUND

Variations in symptoms and indistinguishable depression episodes of unipolar depression (UD) and bipolar disorder (BD) make the discrimination difficult and time-consuming. For adolescents with high disease prevalence, an efficient diagnostic tool is important for the discrimination and treatment of BU and UD.

METHODS

This multi-center cross-sectional study involved 1587 UD and 246 BD adolescents aged 12-18. A combination of standard questionnaires and demographic information was collected for the construction of a full-item list. The unequal patient number was balanced with three data balancing algorithms, and 4 machine learning algorithms were compared for the discrimination ability of UD and BD in three age groups: all ages, 12-15 and 16-18. Random forest (RF) with the highest accuracy were used to rank the importance of features/items and construct the 25-item shortlist. A separate dataset was used for the final performance evaluation with the shortlist, and the discrimination ability for UD and BD was investigated.

RESULTS

RF performed the best for UD and BD discrimination in all 3 age groups (AUC 0.88-0.90). The most important features that differentiate UD from BD belong to Parental Bonding Instrument (PBI) and Loneliness Scale of the University of California at Los Angeles (UCLA). With RF and the 25-item shortlist, the diagnostic accuracy can still reach around 80 %, achieving 95 % of the accuracy levels obtained with all features.

CONCLUSIONS

Through machine learning algorithms, the most influencing factors for UD and BD classification were recombined and applied for rapid diagnosis. This highly feasible method holds the potential for convenient and accurate diagnosis of young patients in research and clinical practice.

Antidepressant-associated diagnostic change from major depressive to bipolar disorder

BACKGROUND

Anticipating diagnostic change from major depressive (MDD) to bipolar disorder (BD) can support better prognosis and treatment, especially of depression but is challenging and reported research results are inconsistent. We therefore assessed clinical characteristics associated with diagnostic change from MDD to BD with antidepressant treatments.

METHODS

We compared characteristics of 3212 initially MDD patients who became (hypo)manic during antidepressant treatment to those with stable MDD diagnoses as well as with cases of stable, spontaneous BD, using standard bivariate and multivariate statistics.

RESULTS

Among MDD patients, 6.69% [CI: 5.85-7.61] changed to BD, mostly type II (BD2, 76.7%). BD-converters had higher rates of familial mood disorders (74.1% vs. 57.1%) or BD (33.7% vs. 21.0%) and 2.8-years younger onset than stable MDD patients. They also had more prior depressive recurrences/year, years-of-illness, mood-stabilizer treatment, divorces, fewer children, more suicide attempts and drug-abuse, and higher intake cyclothymia, YMRS and MDQ scores. Predictors independently associated with diagnostic conversion were: more familial BD, depressions/year, unemployment, cyclothymic temperament, suicidal ideation or acts, and fewer children. BD-converters vs. spontaneous BD cases had significantly more suicide attempts, BD2 diagnoses, and affected relatives. Converting to vs. spontaneous BD1 was associated with more ADHD, more suicidal ideation or behavior, MDI course, and younger onset; converting to vs. spontaneous BD2 had more episodes/year, unemployment, ADHD, substance abuse, suicidal ideation or attempts, and more relatives with BD.

CONCLUSIONS

Few (6.69%) initially MDD subjects converted to BD, most (76.7%) to BD2. Independent predictive associations with diagnostic change included: familial BD, more depressions/year, unemployment, cyclothymic temperament, suicidal behavior and fewer children. Notably, several characteristics were stronger among those changing to BD during antidepressant treatment vs. others with spontaneous BD.

Based on white matter microstructure to early identify bipolar disorder from patients with depressive episode

OBJECTIVE

Because of similar clinical manifestations, bipolar disorder (BD) patients are often misdiagnosed as major depressive disorder (MDD). This study aimed to compare the difference between depressed patients later converting to BD and unipolar depression (UD) according to diffusion tensor imaging (DTI).

METHOD

Patients with MDD (562 participants) in depressive episode states and healthy controls (HCs) (145 participants) were recruited over 10 years. Demographic and magnetic resonance imaging (MRI) data were collected at the time of recruitment. All patients with MDD were followed up for 5 years and classified into the transfer to BD (tBD) group (83 participants) and UD group (160 participants) according to the follow-up results. DTI and functional magnetic resonance imaging at baseline were compared.

RESULTS

Common abnormalities were found in both tBD and UD groups, including left superior cerebellar peduncle (SCP.L), right anterior limb of the internal capsule (ALIC.R), right superior fronto-occipital fasciculus (SFOF.R), and right inferior fronto-occipital fasciculus (IFOF.R). The tBD showed more extensive abnormalities than the UD in the body of corpus callosum, fornix, left superior corona radiata, left posterior corona radiata, left superior longitudinal fasciculus, and left superior fronto-occipital fasciculus.

CONCLUSION

The study demonstrated the common and distinct abnormalities of tBD and UD when compared to HC. The tBD group showed more extensive disruptions of white matter integrity, which could be a potential biomarker for the early identification of BD.

Predictors of diagnostic conversion from major depression to bipolar disorder: a Swedish national longitudinal study

BACKGROUND

It is clinically important to predict the conversion of major depression (MD) to bipolar disorder (BD). Therefore, we sought to identify related conversion rates and risk factors.

METHODS

This cohort study included the Swedish population born from 1941 onward. Data were collected from Swedish population-based registers. Potential risk factors, including family genetic risk scores (FGRS), which were calculated based on the phenotypes of relatives in the extended family and not molecular data, and demographic/clinical characteristics from these registers were retrieved. Those with first MD registrations from 2006 were followed up until 2018. The conversion rate to BD and related risk factors were analyzed using Cox proportional hazards models. Additional analyses were performed for late converters and with stratification by sex.

RESULTS

The cumulative incidence of conversion was 5.84% [95% confidence interval (95% CI) 5.72-5.96] for 13 years. In the multivariable analysis, the strongest risk factors for conversion were high FGRS of BD [hazard ratio (HR) = 2.73, 95% CI 2.43-3.08], inpatient treatment settings (HR = 2.64, 95% CI 2.44-2.84), and psychotic depression (HR = 2.58, 95% CI 2.14-3.11). For late converters, the first registration of MD during the teenage years was a stronger risk factor when compared with the baseline model. When the interactions between risk factors and sex were significant, stratification by sex revealed that they were more predictive in females.

CONCLUSIONS

Family history of BD, inpatient treatment, and psychotic symptoms were the strongest predictors of conversion from MD to BD.

Sex effects on differentiating patients with major depressive disorder from bipolar disorder in depressive state: A fMRI study with follow-up

BACKGROUND

Bipolar disorder (BD) is difficult to discriminate from major depressive disorder (MDD) before the appearance of mania or hypomania. This study was designed to identify whether patients with MDD and those who converted to BD are distinguishable using dynamic amplitude low-frequency fluctuations (dALFF) and describe the sex effects on the identification of the two disorders.

METHODS

We compared the dALFF values of 35 BD patients who converted from MDD during the 2-year follow-up, 99 MDD patients, and 130 healthy controls (HCs) using two-way ANOVA. Pearson's correlation was used to compare dALFF in dysfunctional brain regions and clinical characteristics.

RESULTS

A main effect of diagnosis was discovered in the frontal and occipital gyrus. For the main effect of sex, both the left middle occipital gyrus and the medial part of the superior frontal gyrus had higher dALFF values in males compared to females. An interaction of sex and diagnosis effect was observed in the right precentral gyrus. Male MDD patients exhibited a higher dALFF value than male BD patients. Additionally, we discovered a higher dALFF value in females than in males in BD patients. WCST scores were positively associated with dALFF values in the frontal and occipital gyrus in MDD patients. Meanwhile, dALFF values in the occipital gyrus positively correlated with WCST in female MDD patients only.

LIMITATION

Most of the participants were on medication and the sample size was small.

CONCLUSIONS

Our study is the first to find the non-neglectable role of sex effects in differentiating BD and MDD at an early stage.

An Enhanced Ant Colony Optimization Mechanism for the Classification of Depressive Disorders

Bipolar disorder is marked by mood swings that alternate between mania and depression. The stages of bipolar disorder (BD), as one of the most common mental conditions, are often misdiagnosed as major depressive disorder (MDD), resulting in ineffective treatment and a poor prognosis. As a result, distinguishing MDD from BD at an earlier phase of the disease may aid in more efficient and targeted treatments. In this research, an enhanced ACO (IACO) technique biologically inspired by and following the required ant colony optimization (ACO) was utilized to minimize the number of features by deleting unrelated or redundant feature data. To distinguish MDD and BD individuals, the selected features were loaded into a support vector machine (SVM), a sophisticated mathematical technique for classification process, regression, functional estimates, and modeling operations. In respect of classifications efficiency and frequency of features extracted, the performance of the IACO method was linked to that of regular ACO, particle swarm optimization (PSO), and genetic algorithm (GA) techniques. The validation was performed using a nested cross-validation (CV) approach to produce nearly reliable estimates of classification error.

Diagnostic conversion from unipolar to bipolar affective disorder-A population-based study

OBJECTIVES

It is essential to clinically distinguish bipolar affective disorder from unipolar affective disorders. However, patients previously diagnosed with unipolar affective disorder are sometimes later diagnosed with bipolar affective disorder, known as diagnostic conversion. Here we investigated diagnostic conversion using data from a nationwide population-based register.

METHODS

We obtained claims data from 2007 to 2020 in Korea's Health Insurance Review Agency database and identified a cohort of patients who were diagnosed with unipolar depression in 2009 without prior psychiatric diseases within the previous 2 years. We studied the rate of diagnostic conversion and risk factors, especially antidepressants.

RESULTS

About 6.5% of patients underwent diagnostic conversion during the observation period. Younger age at disease onset and usage of antidepressants increased the relative risk for diagnostic conversion. Patients using serotonin-norepinephrine reuptake inhibitors (SNRI) showed more than twice the risk compared to no usage of antidepressant.

LIMITATION

First, this study was based on the population-based register data. Thus, we defined the patient cohort diagnosed with unipolar depression with strict inclusion criteria. Second, the exposure time differed between different antidepressants. Third, we estimated the relative risk for diagnostic conversion compared to no use of antidepressants. Moreover, we could not rule out the potential influence of antidepressant polypharmacy.

CONCLUSION

We confirmed diagnostic conversion in some patients and identified younger age or usage of antidepressants, especially SNRI, as risk factors. Because unipolar and bipolar affective disorders show different disease courses or prognoses and have different treatment strategies, clinicians should be mindful of diagnostic conversion.

A Predictive Model of Risk Factors for Conversion From Major Depressive Disorder to Bipolar Disorder Based on Clinical Characteristics and Circadian Rhythm Gene Polymorphisms

BACKGROUND

Bipolar disorder (BD) is easy to be misdiagnosed as major depressive disorder (MDD), which may contribute to a delay in treatment and affect prognosis. Circadian rhythm dysfunction is significantly associated with conversion from MDD to BD. So far, there has been no study that has revealed a relationship between circadian rhythm gene polymorphism and MDD-to-BD conversion. Furthermore, the prediction of MDD-to-BD conversion has not been made by integrating multidimensional data. The study combined clinical and genetic factors to establish a predictive model through machine learning (ML) for MDD-to-BD conversion.

METHOD

By following up for 5 years, 70 patients with MDD and 68 patients with BD were included in this study at last. Single nucleotide polymorphisms (SNPs) of the circadian rhythm genes were selected for detection. The R software was used to operate feature screening and establish a predictive model. The predictive model was established by logistic regression, which was performed by four evaluation methods.

RESULTS

It was found that age of onset was a risk factor for MDD-to-BD conversion. The younger the age of onset, the higher the risk of BD. Furthermore, suicide attempts and the number of hospitalizations were associated with MDD-to-BD conversion. Eleven circadian rhythm gene polymorphisms were associated with MDD-to-BD conversion by feature screening. These factors were used to establish two models, and 4 evaluation methods proved that the model with clinical characteristics and SNPs had the better predictive ability.

CONCLUSION

The risk factors for MDD-to-BD conversion have been found, and a predictive model has been established, with a specific guiding significance for clinical diagnosis.