From slow to fast? Longitudinal dynamics of processing speed across CBT treatment for depression

Neurocognitive profiles in major depressive disorder (MDD) are particularly shaped by deficits in processing speed, frequently persisting beyond remission. This raises the critical question of whether such deficits constitute enduring neuropsychological scars or state-dependent impairments and therefore highlights the need for longitudinal research designs on this topic. Importantly, residual neurocognitive deficits have been identified as a vulnerability factor for relapse of MDD, underscoring the importance of understanding underlying factors and informing therapeutic strategies. Additionally, self-perceived cognitive deficits have been shown to impact disease trajectories. While the majority of existing studies rely on pre-post comparisons, the present study focuses on (I) dynamics of cognitive change at three different timepoints during cognitive behavioral therapy (CBT) in MDD and (II) aims to identify clinically meaningful predictors of recovery, including symptom severity, subjective cognitive deficits, number of prior depressive episodes and white-matter microstructure.

This study examines the longitudinal development of processing speed across multiple time points in a sample of 53 MDD outpatients undergoing CBT, alongside 53 age-matched healthy controls (HC). Assessments are completed at outpatient clinic intake, and at two follow-up time-points after 6 and 12 months. All participants complete a structured clinical interview for DSM-V (SCID-5-CV); with patients fulfilling criteria of an acute MDD diagnosis and a BDI-II score greater than 9. Different subdimensions of processing speed are measured at each timepoint using standardized neuropsychological tests (d2 Test of Attention, Trail Making Test A and B, Symbol Digit Modalities Test). Baseline assessment of symptom severity (measured by BDI-II), subjective cognitive deficits (measured by Perceived Deficits Questionnaire; PDQ-D), and number of prior depressive episodes are included as predictors of both baseline performance and rate of change.

Regarding analysis, Linear Mixed Models will be implied, which allow the estimation of change over time while accounting for inter-individual variability. Group (MDD vs. HC) and Time × Group interaction terms serve as primary fixed effects. Symptom severity, perceived cognitive deficits, and prior episode history will be entered as second-level predictors. Findings are expected to reveal distinct individual trajectories of processing speed recovery during CBT, moderated by symptom burden, subjective cognitive experience, and episode history. This might carry direct implications for the personalization of cognitive interventions within CBT and for identifying patients at risk of persistent cognitive impairment following treatment.

In a subsequent stage, this work will be extended by incorporating DTI structural connectome data acquired in parallel across all time points. This neuroimaging component will allow us to examine whether the observed neurocognitive trajectories are mirrored by corresponding changes in white matter network organization and whether baseline structural connectome predicts the rate and extent of cognitive recovery. Building on the network-level framework proposed by Gruber et al. [1], this longitudinal perspective aims to provide a neurobiological account of treatment-related cognitive change, moving beyond the cross-sectional designs that dominate the existing literature.

References

1. Gruber, M., Mauritz, M., Meinert, S., Grotegerd, D., De Lange, S. C., Grumbach, P., … & Repple, J. (2023). Cognitive performance and brain structural connectome alterations in major depressive disorder. Psychological Medicine, 53(14), 6611-6622.