Neurocognitive deficits, psychotrauma, and inflammation shape major depressive disorder and its phenome features

Xueyan Wang, Ping Wang, Mengqi Niu, Chenkai Yangyang, Abbas F. Almulla, Chen Chen, Jin Li, Yingqian Zhang, Michael MaesmedRxiv 2026 · DOI

TL;DR

Researchers have developed a "cognitype"—a multivariate cognitive profile—that significantly improves our ability to characterize Major Depressive Disorder (MDD). By combining this cognitive data with measures of childhood trauma and systemic inflammation, the team created a model that explains up to 55% of the variance in depressive symptoms and identifies MDD patients with over 90% accuracy.

What Problem Does This Paper Address?

Major Depressive Disorder (MDD) is notoriously heterogeneous; two patients with the same diagnosis can have vastly different symptoms, ranging from insomnia and irritability to cognitive "fog" and physical pain. Traditionally, researchers have looked for single "smoking gun" biomarkers or specific cognitive deficits (like poor memory) to explain the disease. However, these individual indices often show small, inconsistent effects.

The authors argue that we should instead look at the "phenome"—the full, multidimensional expression of the disorder. The challenge is that we haven't had a reliable way to integrate cognitive performance with biological and environmental factors (like inflammation and past trauma) to see how they collectively shape a patient's specific experience of depression.

Methods

The study involved 87 patients in the acute phase of MDD and 40 healthy controls. The researchers used three primary "pillars" of data:

  1. The Cognitype: Instead of looking at individual test scores, they used Principal Component Analysis (PCA) on results from the CANTAB (Cambridge Neuropsychological Test Automated Battery). They focused on four specific tasks: Delayed Match to Sample (DMS), Rapid Visual Information Processing (RVP), One Touch Stockings of Cambridge (OTS), and the Emotion Recognition Test (ERT). The first principal component (PC1) became the "cognitype."
  2. The APP Index: A composite score of "acute-phase proteins" (like C-reactive protein, albumin, and transferrin) to measure systemic inflammatory activation.
  3. ACEs: Adverse Childhood Experiences, measured via the Childhood Trauma Questionnaire, to account for developmental risk factors.

They then used logistic and multiple regression models to see how well these three pillars could distinguish patients from controls and predict specific symptom clusters (affective, physical, vegetative, etc.).

Datasets

The study utilized a primary dataset of 127 participants (87 MDD, 40 Healthy Controls) recruited from the Psychiatric Center of Sichuan Provincial People’s Hospital in Chengdu, China. While not a massive "big data" set, it is a deeply phenotyped clinical sample where every participant underwent extensive neurocognitive testing and blood assays.

Key Insights and Ideas

  • Cognition is an Independent Pillar: One of the most important findings is that the "cognitype" isn't just a downstream side effect of inflammation or trauma. It provides independent explanatory power. This suggests that cognitive dysfunction is a core dimension of the illness itself, not just a symptom of being "inflamed" or "traumatized."
  • Efficiency Over Accuracy: The "cognitype" was heavily defined by latency (speed). Patients weren't necessarily getting the answers wrong; they were just much slower at planning and processing information. This points to a "reduced processing efficiency" rather than a total loss of function.
  • The "Disgust" Bias: The cognitive profile included a heightened sensitivity to disgust-related emotional cues. This aligns with "negative-bias" theories of depression, where the brain becomes hyper-vigilant toward aversive or self-critical stimuli.
  • NIMETOX Framework: The paper supports the emerging view of MDD as a NIMETOX (neuroimmune-metabolic-oxidative stress) disorder. It’s a systemic condition where the environment, the immune system, and the brain's "hardware" interact in a feedback loop.

Key Results

  • High Discriminatory Power: The multivariable model (Cognitype + ACEs + APP) discriminated MDD patients from healthy controls with an AUC (Area Under the Curve) of 0.907. In clinical psychology, an AUC above 0.90 is considered excellent.
  • Explaining the Variance: The combined effects of these three pillars explained 40–55% of the variance across different symptom dimensions (like melancholia, physical symptoms, and suicidal ideation).
  • Model Accuracy: The best-performing model achieved an overall accuracy of 88.4%, with a specificity of 92.5% in identifying healthy individuals.

Limitations and Future Work

The authors acknowledge several hurdles:

  • Cross-Sectional Design: Because the data was collected at a single point in time, the study cannot prove that inflammation causes cognitive deficits, or vice versa. Longitudinal studies are needed to map these relationships over time.
  • Sample Specificity: The participants were "acute-phase" patients (currently experiencing a severe episode). It remains to be seen if this "cognitype" persists during remission or if it changes as symptoms improve.
  • Missing Neuroimaging: While the authors hypothesize about specific brain circuits (like the fronto-striatal pathway), they didn't use MRI in this study. Future work integrating functional MRI with the "cognitype" could bridge the gap between behavior and brain structure.

Summary updated

Abstract

ABSTRACT Background Major depressive disorder (MDD) involves immune–metabolic dysregulation, psychosocial adversity, and multidomain cognitive disturbances, yet single cognitive indices often show small and inconsistent effects. We derived a multivariate Cambridge Neuropsychological Test Automated Battery (CANTAB)-based cognitive phenotype (“cognitype”) and tested whether it adds explanatory value beyond adverse childhood experiences (ACEs) and an acute-phase protein (APP) index in acute-phase MDD. Methods Eighty-seven acute-phase MDD patients and 40 healthy controls completed CANTAB testing; key outcomes from DMS, RVP, OTS, and ERT were summarized as a cognitype score (PC1). ACEs were assessed, and peripheral inflammatory markers were combined into an APP index. Logistic and multiple regression models tested discrimination between MDD and controls and prediction of multidimensional phenome features (affective, physio-somatic, vegetative, recurrence-related, personality, and suicidality domains). Results Individual CANTAB outcomes showed limited between-group differences after FDR correction, but multivariable models integrating cognitive measures with ACEs and APP robustly discriminated MDD from controls (AUC up to 0.907). The cognitype independently predicted multiple phenome domains when modeled alongside ACEs and APP, and their combined effects explained ∼40–55% of variance across symptom dimensions. Conclusion A data-driven cognitype derived from core CANTAB tasks captures clinically meaningful cognitive variation in acute-phase MDD and contributes significant predictive value beyond psychosocial adversity and inflammatory activation. Integrating cognition, ACEs, and inflammation improves characterization of symptom heterogeneity and supports precision approaches targeting neurocognitive–immune–environmental mechanisms.
BibTeX
@misc{wang2026neurocognitive,
  author = {Xueyan Wang and Ping Wang and Mengqi Niu and Chenkai Yangyang and Abbas F. Almulla and Chen Chen and Jin Li and Yingqian Zhang and Michael Maes},
  title = {{Neurocognitive deficits, psychotrauma, and inflammation shape major depressive disorder and its phenome features}},
  year = {2026},
  doi = {10.64898/2026.02.11.26346056},
  howpublished = {medRxiv},
}
Figure