The network and dimensionality structure of affective psychoses: an exploratory graph analysis approach

• Published in: Journal of affective disorders Vol. 277; pp. 182 - 191
• Main Authors: Peralta, Victor, Gil-Berrozpe, Gustavo J., Sánchez-Torres, Ana, Cuesta, Manuel J.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.12.2020
• Subjects: Affective Disorders, Psychotic - diagnosis; Affective psychoses; Cross-Sectional Studies; Dimensionality; Factor analysis; Hallucinations - diagnosis; Humans; Network analysis; Network structure; Psychiatric/Mental Health; Psychotic Disorders - diagnosis; Schizophrenia; Symptom connectivity; Network structure; Dimensionality; Factor analysis; Symptom connectivity; Network analysis; Affective psychoses
• ISSN: 0165-0327; 1573-2517; 1573-2517
• DOI: 10.1016/j.jad.2020.08.008

•Network analyses allow exploring symptom relationships across classes of affective psychoses.•Classes of affective psychoses significantly differed in network structure but not in global network connectivity.•The connectivity pattern between mood and psychotic symptoms highly depends on the specific diagnosis of affective psychosis.•Differences in symptom patterns might give rise to tailored treatment hypotheses. The dimensional symptom structure of classes of affective psychoses, and more specifically the relationships between affective and mood symptoms, has been poorly researched. Here, we examined these questions from a network analysis perspective. Using Exploratory Graph Analysis (EGA) and network centrality parameters, we examined the dimensionality and network structure of 28 mood and psychotic symptoms in subjects diagnosed with schizoaffective disorder (n=124), psychotic bipolar disorder (n=345) or psychotic depression (n=245), such as in the global sample of affective psychoses. EGA identified four dimensions in subjects with schizoaffective or bipolar disorders (depression, mania, positive and negative) and three dimensions in subjects with psychotic depression (depression, psychosis and activation). The item composition of dimensions and the most central symptoms varied substantially across diagnoses. The most central (i.e., interconnected) symptoms in schizoaffective disorder, psychotic bipolar disorder and psychotic depression were hallucinations, delusions and depressive mood, respectively. Classes of affective psychoses significantly differed in terms of network structure but not in network global strength. The cross-sectional nature of this study precludes conclusions about the causal dynamics between affective and psychotic symptoms. EGA is a powerful tool for examining the dimensionality and network structure of symptoms in affective psychoses showing that both the interconnectivity pattern between affective and psychotic symptoms and the most central symptoms vary across classes of affective psychoses. The findings outline the value of specific diagnoses in explaining the relationships between mood and affective symptoms.