Structure of mania: depressive, irritable, and psychotic clusters with different retrospectively-assessed course patterns of illness in randomized clinical trial participants

• Published in: Journal of affective disorders 2001-12, Vol.67 (1), p.123-132
• Main Authors: Swann, Alan C, Janicak, Philip L, Calabrese, Joseph R, Bowden, Charles L, Dilsaver, Steven C, Morris, David D, Petty, Frederick, Davis, John M
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Affective psychosis; Age of Onset; Aged; Anxiety; Bipolar disorder; Bipolar Disorder - psychology; Cluster analysis; Course of illness; Cross-Sectional Studies; Disease Progression; Factor analysis; Female; Humans; Male; Mania; Middle Aged; Mixed state; Mood Disorders - psychology; Multivariate statistics; Psychometrics; Randomized Controlled Trials as Topic; Retrospective Studies; Severity of Illness Index; Stress, Psychological
• ISSN: 0165-0327; 1573-2517
• DOI: 10.1016/S0165-0327(01)00447-5

Background: We investigated the structure of manic episodes by determining whether there was evidence for distinct groups of patients differing in clinical characteristics and course of illness. Methods: The subjects were 162 patients hospitalized for manic episodes who underwent comprehensive evaluations of behavior, symptoms, and history before a treatment study. Pretreatment behavior ratings (Schedule for Affective Disorders and Schizophrenia, rated by clinicians, and Affective Disorder Rating Scale, rated by nurses) entered a principal components factor analysis, followed by a cluster analysis of the subjects based on their factor scores. Members of the resulting clusters were compared with respect to clinical characteristics and history of illness. Results: The six factors were impulsivity, hyperactivity, anxious pessimism, distressed appearance, hostility, and psychosis. The four clusters were characterized as depressive, with high anxious pessimism ( n=22), delusional, with high psychosis ( n=39), classic ( n=72), and irritable, with high distressed appearance and hostility ( n=29). Depressive manics had the earliest onset of illness and the highest density of episodes/year, while irritable manics had later onset and the fewest previous episodes. Limitations: The number of subjects was smaller than ideal for multivariate analysis, subjects were limited to those able to consent and meet criteria for a randomized clinical trial, and course of illness was determined retrospectively. Conclusions: Manic episodes have a dimensional structure but appear to fall naturalistically into types that differ with respect to previous history, symptoms, and clinical characteristics. Whether these are distinct clinical subtypes will require further research.