Apathy-anhedonia is associated with pessimistic beliefs yet sharpened goal-directed planning

This study reveals that individuals with elevated apathy-anhedonia exhibit pessimistic beliefs about success yet demonstrate enhanced, rather than impaired, goal-directed planning and value discrimination when navigating risky, learned environments.

The Gist

The Big Idea: The "Pessimistic Planner"

Imagine you are standing at the bottom of a giant, winding mountain path. You want to reach the summit because there is a treasure chest at the top. However, the path is slippery, and every step you take has a small chance of you slipping and falling off the cliff.

Most people assume that if you are depressed or feeling apathetic (like you don't care or can't feel joy), you would look at that slippery mountain and say, "I'll never make it. It's too risky. I'm going to stay here and eat a sandwich instead."

This study found something surprising: That's not always true.

People with high levels of apathy and anhedonia (the "can't feel joy" symptoms of depression) did think they were less likely to make it to the top. They were pessimistic. But, when it came time to actually climb, they didn't stay behind. In fact, they climbed better, faster, and smarter than everyone else.

---

The Experiment: The "Video Game Maze"

To test this, the researchers created a video game-like experiment with 384 participants.

1. The Setup: Participants had to learn the layout of two complex, loop-shaped mazes over two days. They had to learn where the walls were and, crucially, learn that every step they took had a hidden ~10% chance of "game over" (failure).
2. The Choice: After learning the maze, they faced a choice:
   • Option A: Take a guaranteed, small reward (like a safe, boring walk).
   • Option B: Try to reach a goal further away in the maze for a bigger reward. The further away the goal, the more steps you had to take, and the higher the chance you'd fall off the cliff before getting there.
3. The Twist: Before making their choice, participants had to guess: "What are the odds I'll actually make it?"

The Three Types of People

The researchers didn't just look at "depressed" vs. "not depressed." They broke the participants down into three "personality clusters" based on their questionnaires:

1. The Worriers: People who are anxious and overthink everything.
2. The Impulsives: People who act without thinking and get distracted easily.
3. The Apathetic: People who feel low motivation, lack of interest, and an inability to feel pleasure (the core of this study).

The Surprising Results

1. The Belief vs. Action Gap

When asked, "How likely are you to succeed?" the Apathetic group gave the lowest scores. They truly believed they would fail. They were the most pessimistic.

However, when they actually played the game:

• They didn't give up. They chose the risky, high-reward paths just as often as everyone else.
• They were actually better at the game.

2. The "Sharp Knife" Effect

Here is where it gets really cool. The Apathetic group didn't just do "okay"; they did better than the others in three specific ways:

• Better Value Judgment: They were incredibly good at calculating if the reward was worth the risk. If the reward was too small for the danger, they skipped it. If the reward was huge, they went for it. They didn't get confused by emotions; they acted like a super-efficient computer.
• Better Navigation: When they chose a path, they moved through the maze faster and made fewer mistakes.
• Better Flexibility: Imagine a wall suddenly appears in the maze (a "roadblock"). The Apathetic group was the quickest to realize, "Oh, that path is blocked, I need to take the other way," and they adjusted instantly.

Why Did This Happen? (The Metaphor)

Think of decision-making like driving a car with two systems:

1. The Emotional GPS: This system screams, "I'm scared! This feels bad! I might crash!" It's loud and distracting.
2. The Logic Engine: This system calculates the map, the fuel, and the distance. It just does the math.

The researchers suggest that for people with Apathy/Anhedonia, the Emotional GPS is turned down very low.

• Because they don't feel the "fear" or the "excitement" as strongly, they aren't distracted by emotional noise.
• This allows their Logic Engine to run perfectly. They can see the map clearly, calculate the risks coldly, and execute the plan efficiently.

The Catch:
In real life, this "Logic Engine" mode is great for solving a specific puzzle (like the video game). But in the real world, we often have to create our own goals and push ourselves to start things when we don't feel like it. That is where apathy hurts. But once the goal is set and the path is clear, these individuals can execute the plan with surgical precision.

The Takeaway

This study challenges the idea that depression or apathy always means you are "broken" or "bad at planning."

• The Myth: "If you are apathetic, you can't plan or pursue goals."
• The Reality: "If you are apathetic, you might say you'll fail because you feel pessimistic, but once you start, you might actually be the most efficient, focused, and skilled planner in the room."

It suggests that sometimes, feeling a bit "numb" to the emotional ups and downs can actually help you make sharper, clearer decisions when the stakes are high and the path is known.

Technical Summary

1. Problem Statement

Decision-making research traditionally posits that subjective beliefs about uncertain outcomes drive value-based choices: pessimistic expectations should reduce goal pursuit, while optimistic ones should promote it. Depression, characterized by motivational symptoms like apathy and anhedonia, is associated with diminished expectations of success. However, it remains unclear whether these pessimistic beliefs actually translate into impaired goal-directed behavior, particularly in complex environments where risk must be learned through experience rather than explicitly described.

Existing literature presents a mixed picture regarding depression and model-based (goal-directed) control. While some studies suggest impairment, others (including large-scale transdiagnostic studies) find no effect or even enhanced control. A critical gap exists in isolating apathy–anhedonia from broader depression/anxiety dimensions and testing it in paradigms that require multi-step planning under compounding, learned uncertainty.

2. Methodology

Participants and Design

• Sample: $N = 384$ participants (online, UK-based, aged 18–38).
• Duration: A two-day experiment with a follow-up session.
• Task: A novel "Risky Goal Task" involving maze-like environments.
  • Learning Phase: Participants navigated loop-shaped mazes (16 locations) over two days. Each step carried an undisclosed ~10% probability of failure (step failure), which compounded over distance.
  • Choice Phase: Participants chose between a certain reward and a risky goal (varying in distance 1–7 steps and reward magnitude).
  • Key Features:
    • Compounding Risk: Longer paths offered higher rewards but exponentially higher cumulative failure risk.
    • Explicit Beliefs: Participants provided subjective probability ratings of goal success for specific start-goal pairs.
    • Flexibility Tests: "Roadblock" trials temporarily blocked paths to test model-based updating.
    • Memory Checks: Object location replacement tests ensured participants had learned the environmental structure.

Transdiagnostic Symptom Factors

Instead of using diagnostic categories, the authors performed a factor analysis on 184 items from various clinical questionnaires (e.g., AES, MASQ-AD, BIS, PSWQ). This yielded three orthogonal symptom factors:

1. Apathy–Anhedonia: Motivational symptoms (low positive valence, anhedonia).
2. Worry: Anxiety-related symptoms.
3. Impulsivity: Compulsivity and intrusive thought symptoms.

Computational Modeling

• A computational model was fitted to choice data to estimate:
  • Step Success Probability: A parameter representing the subjective probability of successfully completing a single step (risk-taking for goals).
  • Goal Engagement Cost: A bias toward or away from goal pursuit independent of distance.
  • Inverse Temperature ($\beta$): A softmax parameter indexing the sensitivity of choices to value differences (sharper discrimination = higher $\beta$).
• Drift-Diffusion Modeling (DDM): Used in supplementary analysis to examine evidence accumulation rates.

3. Key Contributions

1. Dissociation of Belief and Action: The study demonstrates that pessimistic explicit beliefs (subjective ratings) can coexist with preserved or even enhanced goal-directed behavior, challenging the assumption that negative expectations necessarily impair action.
2. Specificity of Apathy–Anhedonia: By isolating motivational symptoms from worry and impulsivity, the study identifies that apathy–anhedonia specifically drives this belief-behavior dissociation, whereas other symptom factors do not.
3. Enhanced Planning under Learned Risk: Contrary to the "deficit model" of depression, individuals with high apathy–anhedonia showed sharpened goal-directed performance in structured, multi-step planning tasks.
4. Methodological Advancement: The use of a multi-day, learned-risk paradigm allows for the separation of learning deficits (memory of structure) from planning deficits (utilization of that structure), a distinction often blurred in single-session tasks like the two-step task.

4. Key Results

A. Beliefs vs. Behavior

• Pessimistic Beliefs: Higher Apathy–Anhedonia scores were significantly associated with lower subjective success ratings (more pessimistic beliefs), even after controlling for memory performance and baseline mood. This bias was stable across days.
• Unchanged Risk-Taking: Despite these pessimistic beliefs, Apathy–Anhedonia scores were uncorrelated with the model-estimated step success probability (risk-taking). Participants with high apathy–anhedonia pursued risky goals at rates similar to those with low scores.
• Dissociation: While group-level beliefs and behavior were positively correlated, the relationship between apathy–anhedonia and these two measures diverged (negative correlation with beliefs, zero correlation with behavior).

B. Enhanced Goal-Directed Performance

Individuals with higher Apathy–Anhedonia scores exhibited enhanced performance across multiple metrics:

1. Value Sensitivity: Higher softmax inverse temperature ($\beta$), indicating sharper discrimination between goal values and certain alternatives.
2. Model-Based Flexibility: Greater avoidance of goals when a "roadblock" blocked the optimal path, indicating superior updating of the internal model.
3. Navigation Efficiency: Faster reaction times on the first navigation step (indicating rapid planning/decision-making) without loss of accuracy.
4. Reward Earnings: Higher total earnings during the choice phase.

C. Other Symptom Factors

• Impulsivity: Associated with poorer location memory (learning deficit) but, after controlling for memory, showed no significant relationship with goal-directed planning.
• Worry: Unrelated to memory, beliefs, or goal-directed behavior in this context.

D. Canonical Correlation Analysis

A multivariate analysis confirmed a coherent latent dimension where Apathy–Anhedonia loaded heavily on a factor characterized by pessimistic beliefs co-occurring with enhanced performance metrics (value sensitivity, flexibility, speed).

5. Significance and Implications

• Reframing Motivational Deficits: The findings challenge the view that apathy–anhedonia reflects a generalized deficit in goal-directed control. Instead, it suggests a selective alteration in how expectations are represented and explicitly reported, while the computational machinery for value-based planning remains intact or is even sharpened.
• Mechanism of "Sharpening": The authors hypothesize that reduced affective reactivity in apathy–anhedonia may attenuate transient emotional noise during planning. This could allow for a more systematic integration of reward magnitude and learned risk, leading to more precise, value-driven decisions in structured environments.
• Clinical Relevance:
  • Assessment: Self-reported pessimism may not accurately index planning competence in structured tasks.
  • Intervention: The results suggest that providing external structure (scaffolding goals, reducing initiation costs) could leverage preserved planning capacities in individuals with apathy, even if their explicit expectations remain pessimistic. This supports the efficacy of behavioral activation strategies.
• Theoretical Impact: The study highlights that the relationship between psychiatric symptoms and decision-making is highly context-dependent. Deficits observed in effort-based or self-initiated tasks may not generalize to structured, multi-step planning under learned uncertainty.

In summary, the paper provides robust evidence that pessimistic beliefs and goal-directed competence are dissociable, revealing a complex phenotype where motivational symptoms coexist with, and perhaps even facilitate, highly efficient value-based planning in structured environments.