Exploring the Association Between Residual Mood Symptoms and Self-Reported Side Effects in the Euthymic Phase of Bipolar Disorders: A Cross-Sectional Network Analysis

Abstract

Introduction: Bipolar disorders (BD) are characterized by mood symptoms that can worsen medication side effects. We aimed to study the association between residual mood signs and self-reported side effects in the euthymic phase of BD. Methods: We assessed residual mood signs using the Montgomery-Asberg Depression Rating scale (MADRS) and Young Mania Rating scale (YMRS) and self-reported side effects using the Patient-Rated Inventory of Side Effects (PRISE-M) for 880 males and 1369 females with BD. We conducted a network analysis to test the associations between 52 items of the three scales for males and females separately. We then identified clusters of nodes that fit the networks well. Results: We report only positive associations between residual mood signs and side effects. An elevated mood (YMRS) in females and increased energy (YMRS) in males were central nodes, strongly influencing the development of additional mood symptoms and side effects. Furthermore, we identified three clusters of nodes in both sexes: (1) a "mood cluster", including most YMRS and MADRS items and the PRISE-M items evaluating sedation, sleep, and restlessness, (2) a cluster of nonsexual side effects (mostly PRISE-M items), and (3) a cluster of sexual side effects. In both sexes, we identified bridge nodes that may favor the communication between mood and side effects, namely palpitations (PRISE-M) and agitation (PRISE-M). Conclusions: The results justify the particular attention of practitioners to monitor elevated moods or increased energy to try to reduce self-reported side effects and other residual mood symptoms in the euthymic phase of BD. Our findings suggest that clinicians could consider patient-reported loss of energy, difficulty in falling asleep, and restlessness as mood symptoms rather than medications' side effects. Palpitations and agitation may contribute to the development of additional mood symptoms or somatic complaints.

Keywords: adverse drug effects; bipolar disorders; euthymic phase; mood symptoms; network analysis.

Figure 1

Distribution of the scores in each item of the three questionnaires for males and females. (a) Females (n = 1369) and (b) males (n = 880).

Figure 2

Graphical model of the male and female networks of 52 items of the MADRS, YMRS, and PRISE-M. (a) Females and (b) males. All represented edges have positive weights. Edge width is proportional to edge weight. Of note, a graph is a 2D representation of the network; the relative positions of the nodes and the edge length are arbitrary values to make the graph readable. The colors of the nodes correspond to the three original questionnaires. Surrounded areas correspond to the clusters reported by the community-detection algorithm. Circles around the nodes represent the predictability, available in Table S3.

Figure 3

Strength of each node of the network. (a) Females and (b) males. Each line corresponds to one item of the YMRS, MADRS, or PRISE-M questionnaire. Strength is reported as raw scores. Centrality indices were obtained using the centrality plot function of the package bootnet (Supporting Information).