Symptom-based Subtyping Of Major Depressive Disorder: A Latent Class Analysis

Background and objective:Depression is a heterogeneous syndrome with a large number of symptoms spanning emotional,physical and cognitive domains.Latent Class Analysis is flexible statistical approach used to divides patients with similar symptom characteristics into the same group to identify homogenous types,which can meet the clinical necessity for more homogenous subtypes of depression so as to raise the ability to efficiently predict who will benefit from which treatments.Our research aims to identify subtypes of depression with LCA and characterize their keynote symptoms and clinical correlates.Methods:The analysis was based on a total of 3275 inpatients and outpatients,who aged ≥18 years old and met the criteria of The Diagnostic and Statistical Manual of Mental Disorders,Fourth Edition Text Revision(DSM-IV-TR)for Major Depressive Disorder(current episode),recruited from 16 psychiatric hospitals and psychiatric departments of 16 general hospitals.Then we collected clinical data through an interview and a self-made questionnaire(see appendix 2)to evaluate depression related feeling and behavior.LCA,EFA and CFA were performed in Mplus.Results:LCA was performed in two datasets,one included 19 items consist of the diagnostic symptoms of MDD and four specifiers,the other included 41 items in which non-DSM symptoms were added on the basis of the former.The two models identified respectively five and four classes.Both the Severe and Melancholic classes were found in two models.The former had higher probabilities for almost all symptoms,while the latter characterized by high endorsement in general retarded depressive symptoms and.An Atypical class was accompanied by significant atypical and psychotic symptoms,high suicide risk and a history of the use of psychoactive substances.The Mild class contained the lowest proportion of almost all symptoms.The Moderate class,which only appeared in 19-item model,had a level between the Severe and the Mild in the nine A group symptoms,with low endorsement in symptoms for four features.A four-factor model was revealed by EFA reflecting 1)general depressive symptoms/ melancholic symptoms;2)sleep disturbance;3)atypical/agitated symptoms;4)anxious symptoms.Conclusions:Atypical and melancholic depression were found to be stable and independent,while anxious depression and psychotic depression may not be independent clinical subtypes.Furthermore,part of the diagnostic criteria of DSM-5 MDD specifiers doesn’t count much in clinical classification.Several non-DSM symptoms that show specificity in classification may used to be a supplement for current clinical subtypes and specifiers.Identification of these two subtypes may help to accurately separate homogeneous subgroups,carry out individualized treatment and management,and develop more targeted etiological studies.