Background: Previous studies showed that individuals with schizophrenia demonstrate widespread cognitive deficits. However, our understanding of how cognitive deficits relate to demographic variables, clinical presentation, and structural pathologies is limited. Here we combined harmonized cognitive data (n=970) for eight domains (language, processing speed, vigilance, working memory, verbal memory, non-verbal memory, motor function, and executive function) with previously harmonized diffusion MR, demographic, and clinical data. We investigate cognitive deficits across the lifespan and their association with white matter microstructure.
Methods: Data were harmonized utilizing T-scores for cognitive and an established harmonization method for imaging data. We applied ANCOVAs to compare eight cognitive domains between patients and healthy individuals, testing for the effects of age and sex. We calculated whole-brain fractional anisotropy (FA) and utilized regression-mediation analyses to model the association between diagnosis, FA, cognitive deficits, and processing speed.
Results: ANCOVAs demonstrated significant group differences for all cognitive domains (p<.006) with medium to large effect sizes (d=.48-1.17). Age affected vigilance and sex working memory deficits. We did not find evidence for a “cognitive deficit subtype,” almost all patients demonstrated cognitive deficits in one or more domains. FA mediated the association between diagnosis and language, processing speed, working memory, and non-verbal memory. Processing speed partially explained the influence of diagnosis and FA on cognitive deficits.
Conclusions: Our study highlights the critical role of cognitive deficits in schizophrenia, as they are present across all domains and patients, and almost all are independent of age and sex. We further show that white matter microstructure partially explains the association between schizophrenia and cognitive deficits (directly and via processing speed). The findings illustrate the need for alternative treatment strategies (including neuroprotective medication and cognitive training).
Welcome to our poster on “cognitive deficits and their association with white matter microstructure in schizophrenia.”
My name is Johanna Seitz-Holland. I am a postdoctoral research fellow at the Psychiatry Neuroimaging Laboratory at Brigham and Women’s Hospital.
Today, I am presenting work from a multisite harmonization study. Before I start, I would like to thank all collaborators for sharing their knowledge and data.
And now to the poster:
As you know, schizophrenia is a devastating mental disorder.
While the disorder’s trajectory presents with high inter-individual heterogeneity, it is well established that, as a group, individuals with schizophrenia demonstrate cognitive deficits. Those deficits are among the best predictors of real-world functioning and treatment response.
However, studies investigating the interplay between cognitive deficits, clinical presentation, and neuropathology are sparse and limited by sample size.
To overcome the limitation of small sample sizes, the field moves towards combining data from multiple centers. In that line, we have been working on developing standardization and harmonization methods. We have previously utilized our newly developed harmonization method for diffusion-weighted MRI data to explore white matter trajectories in healthy individuals and patients with schizophrenia. Next, we standardized clinical data to examine the association between white matter microstructure and the clinical trajectory of schizophrenia, a work that was recently published in Molecular Psychiatry.
For the present study, we also harmonized the cognitive data of the same multisite, large dataset.
We used T-scores for harmonization and clustered the cognitive data into eight domains, displayed in the poster’s top left.
On the graph at the right top of the poster, each column represents an individual. T-scores below the mean of 50 are presented in red, T-scores above the mean are shown in green. You can already see that most patients with schizophrenia display T-scores below the average in one or several cognitive domains.
We utilized these T-scores for three principal analyses:
First, we compared cognitive performance between patients and healthy individuals, also testing for the effects of age and sex.
Second, we tested for the association between cognitive performance and symptom severity, and medication.
Last, we applied regression mediation analyses to explore the relationship between a diagnosis of schizophrenia, white matter, and cognitive performance.
As you can see in the bottom left, patients demonstrated worse cognitive performance than healthy individuals for all cognitive domains, with processing speed being the most affected domain.
We observed an age effect for vigilance only and a sex effect for working memory only.
Worse processing speed, vigilance, working memory, and verbal memory were significantly correlated with more severe positive symptoms.
The association with medication was mixed- with more medication being associated with better performance in some and worse performance in other domains.
The mediation models are displayed in the bottom right. Results demonstrate that whole-brain white matter structure partially mediates the association between a diagnosis of schizophrenia and cognitive performance. Additionally, processing speed mediates the association between a diagnosis of schizophrenia, cognition, and white matter.
Taken together, our study highlights the critical role of cognitive deficits in schizophrenia, as they are present across all domains and patients, mostly independent of age and sex.
White matter microstructure partially explains the association between schizophrenia and cognitive deficits (directly and via processing speed).
Thus, our findings illustrate the need for alternative treatment strategies, including cognitive training and neuroprotective medication.
With that, I would like to thank you for listening. If you have any questions, please contact me or the PI of the presented study, Prof. Marek Kubicki, at the Psychiatry Neuroimaging Laboratory.
Live Zoom Session – April 21st
Johanna Seitz-Holland, MD, Ph.D., Joanne D. Wojcik, Ph.D., Suheyla Cetin-Karayumak, Ph.D., Amanda Lyall, Ph.D., Ofer Pasternak, Ph.D., Yogesh Rathi, Ph.D., Godfrey Pearlson, MD, Carol Tamminga, MD, John A. Sweeney, Ph.D., Brett A. Clementz, Ph.D., David A. Schretlen, Ph.D., Petra Verena Viher, Ph.D., Katharina Stegmayer, MD, Sebastian Walther, MD, Jungsun Lee, MD, Tim Crow, MD, Ph.D., Anthony James, MD, Aristotle Voineskos, MD, Ph.D., Robert W. Buchanan, MD, Philip R. Szeszko, Ph.D, Anil K. Malhotra, MD, Sinead Kelly, Ph.D., Martha E. Shenton, Ph.D., Matcheri Keshavan, MD, Raquelle Mesholam-Gately, Ph.D., Marek Kubicki, MD, Ph.D.