In a groundbreaking study set to redefine our understanding of the cognitive consequences of depression and its treatment, researchers have uncovered surprising insights into how various therapeutic interventions impact model-based planning — a crucial aspect of human decision-making. Model-based planning, a sophisticated cognitive process enabling humans to anticipate future outcomes by mentally simulating sequences of events, has long been suspected to be disrupted in major depressive disorder (MDD). Yet, despite widespread clinical use of antidepressants, ketamine infusions, and cutting-edge internet-delivered cognitive behavioral therapy (iCBT), effects on this form of planning were largely uncharted territory.
Published recently in Translational Psychiatry, the study led by Donegan, Hossein, Panny, and colleagues meticulously evaluated whether these commonly deployed interventions modulate model-based decision-making. Their findings are both surprising and provocative: none of the treatments, including rapid-acting ketamine, classic antidepressants, or iCBT, significantly altered model-based planning capabilities in individuals diagnosed with major depression. This challenges prevailing theories that link therapeutic efficacy directly with enhancements in goal-directed cognitive control. Instead, it suggests that the cognitive substrates of decision-making implicated in depression may remain remarkably stable despite symptomatic relief.
To contextualize these findings, it is essential to understand the distinction between model-based and model-free planning. Model-free strategies rely on habitual, experience-driven responses without flexible adaptation, whereas model-based strategies invoke an internal cognitive map to forecast the consequences of actions. Prior neuroimaging studies have implicated disrupted model-based planning in depression, positing this as a potential mechanism underlying the indecisiveness and anhedonia common in affected patients. However, this latest research injects nuance into this framework by demonstrating that effective clinical improvement does not necessarily translate into restored model-based planning performance.
The research team recruited a large cohort of depressed patients undergoing treatment with selective serotonin reuptake inhibitors (SSRIs), ketamine infusions—a novel glutamatergic agent—and those engaged in a structured iCBT program delivered via internet platforms. Through rigorous longitudinal assessment employing decision-making tasks specifically designed to dissect model-based and model-free control, the investigators quantified planning performance before and after treatment intervention. The robustness of the experimental design—including appropriate control groups, standardized clinical assessments, and comprehensive cognitive testing—underscored the reliability of the observations.
Intriguingly, the study’s results eradicate the assumption that ketamine’s rapid antidepressant effects, often linked to synaptogenesis and neuroplasticity enhancement, confer immediate cognitive gains in flexible planning domains. Despite ketamine’s potent mechanistic action at NMDA receptors and its known rapid alleviation of depressive symptoms, patients did not demonstrate enhanced abilities in simulating future action outcomes post-treatment. This outcome adds to a growing body of evidence suggesting a dissociation between mood symptom alleviation and underlying cognitive processes.
Similarly, antidepressants such as SSRIs, which primarily modulate monoaminergic neurotransmission, failed to exhibit substantive changes in model-based control after weeks of treatment. This finding may imply that the benefits of these pharmacotherapies in mood stabilization occur through pathways largely independent of the executive cognitive functions driving model-based decisions. Such independence raises critical questions about the precise cognitive targets of these treatments and how they relate to broader functional recovery in depression.
The inclusion of internet-delivered cognitive behavioral therapy broadens the scope of inquiry to non-pharmacological interventions. iCBT has revolutionized accessibility to evidence-based psychotherapy, offering scalable treatment options. However, the study revealed that even this manualized cognitive restructuring intervention did not measurably shift model-based planning. While iCBT is effective in symptom reduction, its impact might lie more in modifying maladaptive thought patterns and emotional regulation rather than recalibrating complex forward-planning faculties.
Additionally, the preservation of model-based planning despite clinical improvement raises intriguing possibilities about compensatory mechanisms. It may be that other cognitive domains or neural circuits not measured in this study mediate the benefits seen in depression treatment. Alternatively, the severities or subtypes of depression sampled might inherently retain intact goal-directed control, suggesting a need for more nuanced phenotyping when investigating cognitive remediation in mood disorders.
These insights have significant implications for both clinical practice and future research directions. For clinicians, understanding that antidepressant and ketamine treatments plus iCBT do not necessarily restore model-based planning can temper expectations around cognitive improvement following symptomatic amelioration. Moreover, it highlights unmet therapeutic needs—specifically cognitive interventions directly targeting decision-making processes might be warranted for holistic recovery in depression.
For neuroscientists and psychiatric researchers, this work opens fertile ground in exploring alternative or adjunctive therapies aimed at cognitive enhancement. Investigating pharmacological compounds or behavioral interventions that more directly target prefrontal-striatal circuits underlying model-based planning could unlock new avenues for treatment-resistant cognitive symptoms in depression. Longitudinal studies integrating neuroimaging, computational modeling, and ecological momentary assessment will be critical in dissecting the complex interplay between mood, cognition, and neurobiology.
Furthermore, the study underscores the crucial role of rigorous model-based experimental paradigms in psychiatry. By leveraging sophisticated tasks that isolate latent cognitive mechanisms, researchers can more accurately pinpoint which domains remain impaired and which remain resilient throughout the course of mental illness and treatment. This precision medicine approach promises to revolutionize psychiatric diagnostics and therapeutics.
While definitive conclusions require replication and extension across diverse populations, including varying depression severities and comorbidities, these findings constitute a pivotal step forward. They provoke a reevaluation of entrenched assumptions regarding cognitive recovery trajectories in depression and challenge the field to expand therapeutic frameworks beyond mere symptom remission toward cognitive restoration.
In sum, this landmark study illuminates the complex, and sometimes counterintuitive, landscape of cognitive function in depression treatment. Although ketamine, antidepressants, and internet-delivered CBT effectively reduce depressive symptoms, they do not seem to directly improve model-based planning capabilities. This disconnect invites a paradigm shift in how we conceptualize and measure cognitive health in mood disorders, urging a broader integration of nuanced neurocognitive assessments in clinical trials and practice.
As we enter an era where mental health treatments are increasingly personalized and neurobiologically informed, understanding the dissociations among symptomatic, cognitive, and functional recovery becomes paramount. Continuing to unravel these threads will ultimately facilitate development of targeted interventions that restore not only mood but also the complex cognitive architectures that enable adaptive, goal-directed human behavior.
Subject of Research: Cognitive effects of depression treatments on model-based planning
Article Title: Model-based planning is unaffected by ketamine, antidepressant and internet delivered cognitive behavioural therapy treatments in depression
Article References:
Donegan, K.R., Hossein, S., Panny, B.M. et al. Model-based planning is unaffected by ketamine, antidepressant and internet delivered cognitive behavioural therapy treatments in depression. Transl Psychiatry 15, 505 (2025). https://doi.org/10.1038/s41398-025-03722-8
Image Credits: AI Generated
DOI: 10.1038/s41398-025-03722-8
