Unsupervised Machine Learning to Identify Depressive Subtypes / 대한의료정보학회지

Objectives@#This study evaluated an unsupervised machine learning method, latent Dirichlet allocation (LDA), as a method for identifying subtypes of depression within symptom data. @*Methods@#Data from 18,314 depressed patients were used to create LDA models. The outcomes included future emergency presentations, crisis events, and behavioral problems. One model was chosen for further analysis based upon its potential as a clinically meaningful construct. The associations between patient groups created with the final LDA model and outcomes were tested. These steps were repeated with a commonly-used latent variable model to provide additional context to the LDA results. @*Results@#Five subtypes were identified using the final LDA model. Prior to the outcome analysis, the subtypes were labeled based upon the symptom distributions they produced: psychotic, severe, mild, agitated, and anergic-apathetic. The patient groups largely aligned with the outcome data. For example, the psychotic and severe subgroups were more likely to have emergency presentations (odds ratio [OR] = 1.29; 95% confidence interval [CI], 1.17–1.43 and OR = 1.16; 95% CI, 1.05–1.29, respectively), whereas these outcomes were less likely in the mild subgroup (OR = 0.86; 95% CI, 0.78–0.94). We found that the LDA subtypes were characterized by clusters of unique symptoms. This contrasted with the latent variable model subtypes, which were largely stratified by severity. @*Conclusions@#This study suggests that LDA can surface clinically meaningful, qualitative subtypes. Future work could be incorporated into studies concerning the biological bases of depression, thereby contributing to the development of new psychiatric therapeutics.

Associations of Serum Serotonin Levels with 12-week and 12-month Remission in Patients with Depressive Disorders

Objective@#To investigate associations between baseline serum serotonin levels and short- and long-term treatment outcomes in outpatients with depressive disorders in a naturalistic one-year prospective study design. @*Methods@#Patients were recruited at a University hospital in South Korea from March 2012 to April 2017. At baseline, blood samples were obtained from 1,094 patients who received initial antidepressant monotherapy (Step 1). After the Step 1 treatment, further treatment steps (at least Steps 2−4) could be administered every 3 weeks during the acute treatment phase (3, 6, 9, and 12 weeks; n = 1,086), and every 3 months during the continuation treatment phase (6, 9, and 12 months; n = 884). In cases showing an insufficient response or intolerable side effects, patients were asked to choose whether to remain at the current step or enter the next treatment step, with alternative strategies including switching, augmentation, combination, and a mixture of these approaches. Remission was defined as a Hamilton Depression Rating Scale score of ≤ 7. @*Results@#The remission group had significantly higher baseline serum serotonin levels among patients who received Step 1 monotherapy in both acute and continuation treatment phases. These associations remained significant after adjustment for relevant covariates. No associations were found with any other treatment steps. @*Conclusion@#Baseline serum serotonin levels may be used as a biomarker for predicting short- and long-term treatment outcomes in antidepressant monotherapy-treated patients with depressive disorders in a real-world clinical setting.

Telephone based Interventions for Psychological Problems in Hospital Isolated Patients with COVID-19

Objective@#The COVID-19 is overwhelming health care systems globally. Hospital isolation may generate considerable psychological stress. However, there has been scarce evidence on psychological interventions for these patients due to maintain staff safety. We investigated the feasibility and effectiveness of telephone based interventions for psychological problems in hospital isolated patients with COVID-19. @*Methods@#Psychiatrists visited the ward where the patients were hospitalized and interventions were given by using a ward telephone for 30 minutes. All patients were approached to receive a two-week psychological intervention program and/or pharmacotherapy whenever needed. Psychological problems were assessed at baseline, one, and two weeks. For the assessment of anxiety and depressive symptoms, the Hospital Anxiety and Depression Scale was administered to patients once a week. Insomnia severity index and Beck Depression Inventory 9 item were checked weekly to assess insomnia and suicide idea. @*Results@#Of 33 enrolled, clinically meaningful psychological symptoms were found in 6 (18%) patients for anxiety; 13 (39%) for depression; 10 (30%) for insomnia; and 3 (9%) for suicidal ideation. In 9 patients (27%), psychotropic medications were prescribed to manage anxiety, agitation, depressed mood, insomnia, impulsivity, and suicide idea.Compared to baseline, significant improvements were found in anxiety, depression, and suicidal ideation at one week.There were no statistical differences between the values evaluated at baseline and at two weeks. @*Conclusion@#Our report at least indicates potential usefulness of telephone based interventions in hospital isolated patients with COVID-19, and will hopefully form the basis for future randomized clinical trials.

Prevalence and Correlates of Problematic Internet Experiences and Computer-Using Time: A Two-Year Longitudinal Study in Korean School Children

OBJECTIVE: To measure the prevalence of and factors associated with online inappropriate sexual exposure, cyber-bullying victimisation, and computer-using time in early adolescence. METHODS: A two-year, prospective school survey was performed with 1,173 children aged 13 at baseline. Data collected included demographic factors, bullying experience, depression, anxiety, coping strategies, self-esteem, psychopathology, attention-deficit hyperactivity disorder symptoms, and school performance. These factors were investigated in relation to problematic Internet experiences and computer-using time at age 15. RESULTS: The prevalence of online inappropriate sexual exposure, cyber-bullying victimisation, academic-purpose computer overuse, and game-purpose computer overuse was 31.6%, 19.2%, 8.5%, and 21.8%, respectively, at age 15. Having older siblings, more weekly pocket money, depressive symptoms, anxiety symptoms, and passive coping strategy were associated with reported online sexual harassment. Male gender, depressive symptoms, and anxiety symptoms were associated with reported cyber-bullying victimisation. Female gender was associated with academic-purpose computer overuse, while male gender, lower academic level, increased height, and having older siblings were associated with game-purpose computer-overuse. CONCLUSION: Different environmental and psychological factors predicted different aspects of problematic Internet experiences and computer-using time. This knowledge is important for framing public health interventions to educate adolescents about, and prevent, internet-derived problems.

Thyroid Stimulating Hormone, Cognitive Impairment and Depression in an Older Korean Population

OBJECTIVE: Associations of thyroid dysfunction with cognitive impairment and depression in late-life have been described but remain controversial. This study aimed to investigate the associations of serum thyroid stimulating hormone (TSH) levels with cognitive impairment and depression after controlling for potential confounding factors. METHODS: The sample consisted of 495 community residents aged 65 or over in whom serum TSH had been assayed. Cognitive impairment was defined using the Community Screening Interview for Dementia, and depression was diagnosed using the Geriatric Mental State schedule. Age, gender, education, smoking history, physical activity, blood pressure, diabetes, and serum total cholesterol and albumin were included as covariates. RESULTS: There was a significant association between lower (hyperthyroid) serum TSH levels (<0.5 mIU/L) and cognitive impairment after adjustment [odds ratio 7.12 (95% confidence interval 1.35-37.5)]. However, no association was found between TSH levels and depression. CONCLUSION: Based on TSH levels, hyperthyroidism but not hypothyroidism was associated with cognitive impairment in this sample, and we found no evidence for an association of either with depression.