ABCT Bipolar SIG/Annual Meeting/2024

2024 annual meeting: Philadelphia, PA

Business at the SIG Meeting

1. Status review and planning for growth

Andrew Peckham provided details about number of members on listserv (>50!), number that responded to survey (>20!), and reviewed that there were 8 posters from 3 different research groups presented at the SIG Expo at 2024.

2. Dues

Dues historically had been $20/year for faculty and students are requested to contribute what they can. No dues were collected at the meeting.

3. Previous suggestions for improving listserv

1. Create a list: Name, email, current projects
2. When preparing talks, compile symposiums. SIG members could submit talks, and the SIG can compile symposiums and submit.
3. Sending around information, job offers, etc
4. Organizing ABCT bipolar SIG at ISBD

Next meeting is in Chiba, Japan in September 2025; poster submissions due in January

1. Build BD assessment battery matrix

Need a "water carrier" for this -- have multiple meta-analyses that would inform decisions.

1. Recommendations for core measures for a standard, including neurocognition

4. Presentation of Johnson-Youngstrom Poster Award

(None presented?)

5. Symposium submissions

If you have suggestions about themes for submission for ABCT 2025, feel free to drop them here!

• Symposium submissions

Posters Presented at SIG Expo

Meta-Analysis of the Bipolar Spectrum Diagnostic Scale’s Accuracy in Screening Bipolar Disorders

Xu, Y., Chow, G.C., Sisk, B., Youngstrom, E. A. (2024, November). Meta-Analysis of the Bipolar Spectrum Diagnostic Scale’s Accuracy in Screening Bipolar Disorders. Poster presented at the Annual Meeting of the Association for Behavioral and Cognitive Therapy, Philadelphia, PA. https://doi.org/10.17605/OSF.IO/6T5NK

Objective: Bipolar disorder (BD) often goes unrecognized or is incorrectly diagnosed in various contexts (Keramatian et al., 2023; Stiles et al., 2018). One way to address this issue is the appropriate use of regular screening tools for BD. The Bipolar Spectrum Diagnostic Scale (BSDS) is a free screening tool with good reported psychometric properties since it was first developed by Ghaemi et al. (2005). However, no published study has specifically evaluated the accuracy of the BSDS in screening BD. Firstly, there is well-documented evidence on the decreased effect sizes over publication year (Harrer et al., 2022), which also appears in the screening of BD (Youngstrom et al., 2018). Second, the effect size may be influenced by whether the validation study includes subtypes that are less severe in symptomatology or harder to accurately diagnose than BD-I (e.g., cyclothymic disorder, CYC; Youngstrom et al., 2015) and subtypes that are defined not as specific (i.e., BD not otherwise specified, BD-NOS). Finally, though risk of bias is now commonly reported in meta-analysis of diagnostic test accuracy (McGrath et al., 2019), there is relatively little evidence on how it may influence effect size. We aim to address this research gap and, at the same time, explore whether risk of bias (e.g., quality of study design), publication date, and whether the study included bipolar subtypes other than BD-I will moderate the BSDS’s performance across studies.

Method: We used search terms [(mani* OR “bipolar disorder” OR cyclothymia*) AND (“bipolar spectrum diagnostic scale” OR “BSDS”)] in PsychINFO and PubMED to retrieve target studies published since Ghaemi et al. (2005) to September 2023. For eligible studies, we coded their basic study information, recruitment and sample characteristics, control group information, and BSDS scores and performance; we then calculated their quality of study design and reporting (an indicator of risk of bias) using tools from Kowatch et al. (2005) and QUADAS-2 (Whiting et al., 2011). We utilized Hedges’ g as the primary measure of effect size, adjusting Cohen’s d for biases related to sample size. Multivariate meta-regression was conducted, and all statistical analyses were executed using R studio using the metafor package.

Results: A total of 16 studies published from 2005-2019 were included. The BSDS had a pooled effect size of g = 1.30 (95% CI: 1.13, 1.47) and the averaged study quality was moderately high (74%). There was significant heterogeneity of effect sizes (Cochran’s Q (15 df) = 56.60, p < .0001). The multivariate meta-regression model accounted for 38% of variance (Q (4 df) = 3.52, p = .47). None of the tested moderators (quality of study design and reporting, whether study sample included CYC or BD-NOS, and publication year) significantly associated with effect size.

Conclusions: The present findings show that the BSDS has a strong screening accuracy as evidenced by nearly two decades of research. Notably, its performance remains consistent regardless of the publication year, the study's risk of bias, or the type of bipolar type assessed. These results serve as strong support for BSDS's ongoing application in clinical practice for screening BD.

A Content Overlap Analysis of 7 Mania Rating Scales for Children and Adolescents

Xu, Y., Rodda, P.A., Baggs, J., Youngstrom, E.A. (2024, November). A Content Overlap Analysis of 7 Mania Rating Scales for Children and Adolescents. Poster presented at the Annual Meeting of the Association for Behavioral and Cognitive Therapy, Philadelphia, PA. https://doi.org/10.17605/OSF.IO/E93ZH

Abstract: Bipolar disorder (BD) is often misdiagnosed in the pediatric population, but appropriate rating scales may improve detection and assessment. However, it is common in research and practice to use a single scale while drawing general conclusions about mood disorders, relying on the assumption that different scales are interchangeable. To assess whether different scales are interchangeable, we aim to quantify the degree of similarity in content among common mania rating scales for children and adolescents. Two trained coders independently coded content for 7 common mania scales. We used the Jaccard similarity index (J) to estimate the degree of content overlap (0 = no overlap, 1 = full overlap). We identified 58 disparate symptoms for 130 items across the 7 scales.  The mean overlap among all rating scales was weak (J = .27). Content overlap between scales ranged from very weak to moderate (J = .06-.57). On average, each symptom appeared in 2.3 of the 7 scales; 43% of symptoms appear on only one scale, while no symptoms are present across all scales. Our findings indicate a lack of consensus on children and adolescent mania symptomatology and that generalizing results for BD based on specific scales can be problematic. Our work underscores the importance of careful scale selection by researchers and clinicians based on specific study objectives and symptoms of interest. Key words: Bipolar disorder, children and adolescents, mania, rating scales, content overlap.

Reward Sensitivity and Social/Circadian Rhythm Dysregulation: A Systematic Review and Meta-Analysis.

Smith, L. T., Walsh, R. F. L., Maddox, M. A., Bishop, O. C., & Alloy, L. B. (2024, November). Reward Sensitivity and Social/Circadian Rhythm Dysregulation: A Systematic Review and Meta-Analysis. Poster presented at the Bipolar Disorders special interest group at the annual meeting of the Association for Behavioral and Cognitive Therapies (ABCT), Philadelphia, PA.

Introduction: This review considers the association between reward sensitivity (RS) and social/circadian rhythm dysregulation. RS is an individual’s trait tendency to be motivated by and respond to rewarding stimuli. There are known associations between altered RS and bipolar disorder (BD). Another construct with ties to BD is circadian regulation. The body’s circadian rhythms control the timing of biological processes, and lower regularity is associated with adverse health outcomes. Social rhythms are daily activities, like sleep timing, that entrain biological rhythms and can be considered circadian proxies. In addition to the implications of each factor individually, RS and social/circadian rhythms are known to play a joint role in the etiology of BD. The purpose of this review is to evaluate their degree of correlation.

Method: A systematic search identified 852 articles for screening, and 219 then had full-text review. Twelve met the inclusion/exclusion criteria and were included in an overall meta-analysis. A second analysis was conducted for the relationship between RS and chronotype (N=10).

Results: The pooled effect sizes indicated a small, significant correlation (Overall r=.166, p=.041; Chronotype r=.123, p=.012). There were not enough articles that assessed neural RS to conduct a separate analysis, but the results similarly indicated a small, significant association.

Conclusions: Treatments for BD targeting one of these factors may have a simultaneous impact on the other. Researchers studying these traits should account for this association in power estimations. Future work should build on these findings by investigating a possible direction of effect and including new studies to allow for moderator analyses.

First author Noreen Xu and principal investigator Eric Youngstrom pose beside 3 posters that they authored for the Bipolar Special Interest Group at the 2024 convention for the Association of Behavioral and Cognitive Therapies.