Accumulation of evidence during decision making in OCD patients

Yilin Chen, Ying Liu, Zhen Wang, Tianming Yang, Qing FanFrontiers in Psychiatry 2022 · DOI

TL;DR

This study reveals that individuals with Obsessive-Compulsive Disorder (OCD) exhibit a less efficient process of transforming sensory information into evidence during decision-making compared to healthy controls. While they achieve similar accuracy by accumulating more evidence, this comes at the cost of significantly longer reaction times, particularly in challenging decision scenarios.

What Problem Does This Paper Address?

Decision-making is a complex cognitive process that often involves accumulating evidence from various sources before reaching a conclusion. Previous research on individuals with Obsessive-Compulsive Disorder (OCD) has yielded mixed results regarding their decision-making processes. Some studies suggest deficits in accuracy or speed, while others point to differences in how evidence is gathered. This inconsistency highlights the need for a more nuanced understanding of how OCD affects specific stages of decision-making, such as the transformation of sensory input into usable evidence and the subsequent accumulation of this evidence. The authors aim to disentangle these processes to pinpoint the precise nature of the decision-making alterations in OCD.

Methods

The researchers employed a probabilistic reasoning task designed to isolate the evidence accumulation component of decision-making. Participants were presented with a sequence of visual stimuli (arrows of varying directions and contrasts) that provided probabilistic evidence for one of two choices (left or right). The key innovation here is that the stimuli's visual features (direction and contrast) directly corresponded to their evidence weight, reducing the need for arbitrary learning and making the task more suitable for clinical populations.

To analyze the data, they used several approaches:

  • Behavioral Analyses: This included modeling non-decision stimulus numbers (stimuli presented too late to influence the decision) and fitting psychometric curves to understand the relationship between accumulated evidence and choice probability.
  • Logistic Regression: Used to estimate subjective weights assigned to different stimulus types and to model the influence of trial history on choices.
  • Drift-Diffusion Model (DDM): A standard computational model for decision-making was applied to infer underlying parameters like evidence accumulation rate (drift rate), decision threshold (bound height), and non-decision time. They specifically used a leaky drifting diffusion model with collapsing bounds to capture the dynamic nature of evidence accumulation.

Datasets

The study utilized behavioral data collected from 28 healthy controls and 33 individuals diagnosed with OCD at the Shanghai Mental Health Center. After excluding participants whose accuracy fell outside a 95% confidence interval, the final analysis included 26 healthy controls and 32 OCD patients. No new benchmark datasets were introduced; the study relied on data collected specifically for this research.

Key Insights and Ideas

  • Sensory-to-Evidence Transformation Deficit: The core finding is that OCD patients are less efficient at converting raw sensory information (the visual stimuli) into meaningful evidence for decision-making. This suggests a problem in how the brain processes and interprets incoming data, rather than a fundamental inability to perceive.
  • Evidence Accumulation as Compensation: To achieve similar accuracy to controls, OCD patients accumulate more evidence. This indicates they are aware of their reduced efficiency and compensate by gathering additional information, effectively "working harder" to reach a decision.
  • Difficulty-Dependent Delays: The increased reaction times in OCD patients are most pronounced in difficult trials, where evidence is weak and ambiguous. This suggests that the compensatory strategy of accumulating more evidence is particularly crucial and time-consuming when faced with uncertainty.
  • Prefrontal Cortex Involvement: The findings align with previous research implicating the orbitofrontal cortex in processing evidence. The study suggests that an altered interplay within prefrontal and cortico-basal ganglia circuits may underlie these decision-making deficits in OCD.

Key Results

  • Accuracy: OCD patients and controls achieved comparable accuracy in the task.
  • Reaction Time: OCD patients took significantly longer to make decisions, especially in difficult trials.
  • Evidence Accumulation: OCD patients accumulated more evidence (higher sum weight) than controls to achieve similar accuracy.
  • Drift-Diffusion Model Parameters:
    • Evidence Sensitivity (Drift Rate): Significantly lower in the OCD group (OCD: 17.03 ± 1.17 vs. Control: 20.91 ± 1.45). This is the primary indicator of less efficient evidence processing.
  • Decision Bounds (Initial Bound Height & Collapsing Rate): Comparable between groups, suggesting the decision threshold itself is not the primary issue.
  • Non-Decision Time: Similar between groups, indicating that basic sensory processing and motor execution speeds are not the source of the delay.

Limitations and Future Work

  • Symptom Severity Correlations: The study found weak and inconsistent correlations between decision-making parameters and OCD symptom severity (Y-BOCS, HAM-A, HAM-D). The authors suggest this might be due to a limited range of symptom severity in their cohort or the complex nature of the relationship. Future work should investigate these links with larger, more diverse samples.
  • Complex Mechanisms: The authors acknowledge that the relationship between altered cognitive functions and OCD symptoms is likely complex and may involve intricate mechanisms not captured by simple correlations. Further research is needed to unravel these complexities.
  • Prefrontal Circuitry: While the findings point to prefrontal cortex involvement, direct neural measurements would be necessary to confirm these hypotheses. Future studies could combine behavioral tasks with neuroimaging techniques.
  • Metacognition: The paper touches upon metacognition and confidence estimation, suggesting these processes might be influenced by the observed decision-making deficits. Exploring how OCD patients estimate their confidence in decisions could be a fruitful avenue for future research.

Summary updated

Abstract

Decision-making often entails the accumulation of evidence. Previous studies suggested that people with obsessive-compulsive disorder (OCD) process decision-making differently from healthy controls. Both their compulsive behavior and obsessive thoughts may influence the evidence accumulation process, yet the previous studies disagreed on the reason. To address this question, we employed a probabilistic reasoning task in which subjects made two alternative forced choices by viewing a series of visual stimuli. These stimuli carried probabilistic information toward the choices. While the OCD patients achieved similar accuracy to the control, they took longer time and accumulated more evidence, especially in difficult trials in which the evidence strength was low. We further modeled the subjects' decision making as a leaky drifting diffusion process toward two collapsing bounds. The control group showed a higher drifting rate than the OCD group, indicating that the OCD group was less sensitive to evidence. Together, these results demonstrated that the OCD patients were less efficient than the control at transforming sensory information into evidence. However, their evidence accumulation was comparable to the healthy control, and they compensated for their decision-making accuracy with longer reaction times.
BibTeX
@inproceedings{chen2022accumulation,
  author = {Yilin Chen and Ying Liu and Zhen Wang and Tianming Yang and Qing Fan},
  title = {{Accumulation of evidence during decision making in OCD patients}},
  booktitle = {Frontiers in Psychiatry},
  year = {2022},
  doi = {10.3389/fpsyt.2022.980905},
}
Figure