Print Close

Longitudinal Brain Age in First-Episode Mania Youth Treated with Lithium or Quetiapine

Poster No:

508

Submission Type:

Abstract Submission

Authors:

Laura Han¹, Niousha Dehestani², Chao Suo³, Michael Berk⁴, Lianne Schmaal⁵

Institutions:

¹University of Melbourne, Melbourne, VIC, ²Deakin University, Centre for Social and Early Emotional Development, School of Psychology, Faculty, Melbourne, Australia, ³Turner Institute for Brain and Mental Health, School of Psychological Science and Monash Biomedical, Melbourne, Australia, ⁴Deakin University, IMPACT, The Institute for Mental and Physical Health and Clinical Translation, Sc, Melbourne, Australia, ⁵Centre for Youth Mental Health, The University of Melbourne, Parkville, VIC, Australia., Melbourne, Australia

First Author:

Laura Han
University of Melbourne
Melbourne, VIC

Co-Author(s):

Niousha Dehestani
Deakin University, Centre for Social and Early Emotional Development, School of Psychology, Faculty
Melbourne, Australia

Chao Suo
Turner Institute for Brain and Mental Health, School of Psychological Science and Monash Biomedical
Melbourne, Australia

Michael Berk
Deakin University, IMPACT, The Institute for Mental and Physical Health and Clinical Translation, Sc
Melbourne, Australia

Lianne Schmaal
Centre for Youth Mental Health, The University of Melbourne, Parkville, VIC, Australia.
Melbourne, Australia

Introduction:

Bipolar disorder is increasingly viewed as a disorder involving deviations from typical brain development [1]. Treatment of the disorder may involve pharmacological therapy with lithium or quetiapine. However, it is unclear if these agents have neuroprotective effects, especially in early stages of bipolar and schizoaffective disorders. If we can identify interventions with neuroprotective properties during the early stages of illness onset, i.e., after an initial first-episode of mania (FEM), we can potentially limit aberrations in neurodevelopmental trajectories. With these knowledge gaps in mind, we examined whether an age-related multivariate measure of brain structure (i.e., the brain age gap or BAG): a) differs in young individuals after a FEM compared to controls at baseline, b) improves following treatment, and c) is differentially affected by lithium or quetiapine. Finally, we explored whether improvements in clinical symptoms demonstrate parallel improvements in BAG, regardless of treatment.

Methods:

Patients were randomized to lithium (n=21) or quetiapine (n=18) monotherapy [2]. T1-weighted scans were acquired at baseline, 3 months (patients only) and 12 months. Brain age predictions for controls (n=29) and patients (15-26 years) were derived using a deep learning model trained on one of the largest and most diverse assembled datasets to date (N=53,542; https://github.com/estenhl/pyment-public) [3]. To examine test-retest reliability of the predictions generated by the model, we evaluated the intraclass correlation coefficients (ICCs) between baseline and 12-month follow-up brain age and BAG in controls. To examine changes in BAG in response to treatment over time, we performed linear mixed models with BAG as outcome and treatment group (quetiapine, lithium), time (baseline, 3 months, 12 months), age, and sex as fixed effects, while estimating random effects for patient ID. To investigate potential differential effects of quetiapine versus lithium treatment, a treatment by time interaction was included. To explore whether changes in BAG were associated with changes in clinical measures in patients (regardless of treatment), we computed repeated measures correlations (rmcorr package in R) [4].

Results:

A higher baseline BAG was found in FEM patients compared to controls (+1.86 year, p=0.04; Cohen's d=0.52 [SE=0.25], CI 95% [0.03 to 1.01]). Test-retest reliability was high for both brain age predictions: ICC=0.86, p<.0001 [95% CI: 0.72 - 0.93] and BAG: ICC=0.83, p<.0001 [0.67 - 0.91]. No significant effects of time or treatment group, nor any interaction between the two, were observed throughout the course of the study (Figure 1). Collapsed across treatment groups, significant longitudinal correlations were found between BAG and bipolar depression severity (rrm(51) = -0.30, 95% CI [-0.54, 0.16], p = 0.02) [5] and quality of life (QLS Total: rrm(49) = 0.32, 95% CI [-0.08, 0.52], p = 0.02) [6] in patients (Figure 2).

Supporting Image: brainpad_per_treatment.png

·Figure 1. Longitudinal brain age gap by treatment group.

·Figure 2. Longitudinal correlations between the brain age gap and clinical scores.

Conclusions:

This is the first longitudinal study to characterize BAG following a FEM in young individuals receiving lithium or quetiapine treatment. At baseline, individuals showed older appearing brains compared to controls. However, BAG remained stable and administration of lithium or quetiapine did not change BAG during the first year following a FEM, regardless of treatment group. Longitudinal correlations between BAG and bipolar depressive symptoms, as well as quality of life were found, suggesting compensatory mechanisms. To understand whether potential reversible effects become apparent later in the trajectory of bipolar and schizoaffective disorders, a more extended follow-up period with a larger sample is warranted.

Disorders of the Nervous System:

Psychiatric (eg. Depression, Anxiety, Schizophrenia) ¹

Lifespan Development:

Early life, Adolescence, Aging ²

Modeling and Analysis Methods:

Multivariate Approaches

Novel Imaging Acquisition Methods:

Anatomical MRI

Keywords:

Aging

Computational Neuroscience

DISORDERS

Machine Learning

MRI

Psychiatric Disorders

STRUCTURAL MRI

Therapy

^1|2Indicates the priority used for review

Provide references using author date format

1. Kloiber, S. et al. Neurodevelopmental pathways in bipolar disorder. Neurosci. Biobehav. Rev. 112, 213–226 (2020).
2. Berk, M. et al. Quetiapine v. lithium in the maintenance phase following a first episode of mania: randomised controlled trial. Br. J. Psychiatry 210, 413–421 (2017).
3. Leonardsen, E. H. et al. Deep neural networks learn general and clinically relevant representations of the ageing brain. Neuroimage 256, 119210 (2022).
4. Bakdash, J. Z. & Marusich, L. R. Repeated Measures Correlation. Front. Psychol. 8, 456 (2017).
5. Berk, M. et al. The Bipolar Depression Rating Scale (BDRS): its development, validation and utility. Bipolar Disord. 9, 571–579 (2007).
6. Heinrichs, D. W., Hanlon, T. E. & Carpenter, W. T., Jr. The Quality of Life Scale: an instrument for rating the schizophrenic deficit syndrome. Schizophr. Bull. 10, 388–398 (1984).