Chronbiologically-based sub-groups in bipolar I disorder: A latent profile analysis

Robert Gonzalez; Alok Dwivedi; Jamie Zeitzer; Trisha Suppes; Mauricio Tohen; Angelica Forero; Andres Alvarado

doi:10.1016/j.jad.2021.12.010

Chronbiologically-based sub-groups in bipolar I disorder: A latent profile analysis

J Affect Disord. 2022 Feb 15:299:691-697. doi: 10.1016/j.jad.2021.12.010. Epub 2021 Dec 5.

Authors

Robert Gonzalez¹, Alok Dwivedi², Jamie Zeitzer³, Trisha Suppes³, Mauricio Tohen⁴, Angelica Forero⁵, Andres Alvarado²

Affiliations

¹ Department of Psychiatry and Behavioral Health, Penn State Health Milton S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA. Electronic address: rgonzalez4@pennstatehealth.psu.edu.
² Department of Biomedical Sciences, Texas Tech University Health Sciences Center El Paso, 5001 El Paso Drive, El Paso, TX 79905, USA.
³ VA Palo Alto Health Care System, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, 3801 Miranda Avenue, Palo Alto, CA 94304, USA.
⁴ Department of Psychiatry and Behavioral Sciences, University of New Mexico Medical School, 1 University of New Mexico, Albuquerque, NM 87131, USA.
⁵ Department of Psychiatry, Texas Tech University Health Sciences Center El Paso, 5001 El Paso Drive, El Paso, TX 79905, USA.

PMID: 34879259
DOI: 10.1016/j.jad.2021.12.010

Abstract

Background: Bipolar disorder presents with significant phenotypic heterogeneity. The aim of this study was to investigate whether bipolar disorder, type I (BDI) subjects could be meaningfully classified into homogeneous groups according to activity, sleep, and circadian characteristics using latent profile analysis (LPA). We hypothesized that distinct BDI sub-groups would be identified based primarily on circadian-associated markers.

Materials and methods: 105 individuals with BDI were included in the study. Seventeen activity, sleep, and circadian characteristics were assessed via actigraphy and clinical assessments. LPA was conducted to stratify our sample into homogenous sub-groups. Differences between groups on demographic, clinical, activity, sleep, and circadian characteristics were explored.

Results: Two distinct groups were identified, a High Chronobiological Disturbance group (HCD) (56%, N = 59) and a Low Chronobiological Disturbance group (LCD) (41%; N = 46). Circadian variables were the defining characteristics in sub-group determination. Large effect sizes and magnitudes of association were noted in circadian variables between HCD and LCD sub-groups. Several circadian rhythm variables accounted for a large percentage of the variance between HCD and LCD sub-groups. No differences were noted between sub-groups on demographic characteristics and the psychiatric medications currently in use. Mood state did not significantly impact sub-group differences.

Limitations: The protocol was cross-sectional in design. Longitudinal studies are required to determine the stability of the identified sub-groups.

Conclusion: LPA was able to identify sub-groups in BDI with circadian variables being the most distinguishing factors in determining sub-group class membership. Future research should explore the role that circadian characteristics can play in defining sub-phenotypes of bipolar disorder.

Keywords: Actigraphy; Bipolar disorder; Chronobiological; Circadian rhythm; Latent profile analysis; Phenotype.

Publication types

Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't

MeSH terms

Actigraphy
Bipolar Disorder*
Circadian Rhythm
Cross-Sectional Studies
Humans
Sleep

Grants and funding

T32 MH067543/MH/NIMH NIH HHS/United States