Causes and consequences of unawareness (anosognosia) of tool-action errors after left-hemisphere stroke

This study demonstrates that unawareness of tool-action errors (anosognosia of apraxia) in left-hemisphere stroke survivors arises from degraded action knowledge, which impairs error detection and subsequently reduces the likelihood of attempting to correct those errors.

The Gist

The Big Picture: The "Blind" Mistake-Maker

Imagine you are trying to bake a cake. You know the recipe, you have the ingredients, but when you mix the batter, you accidentally put in salt instead of sugar. You take a bite, and it tastes terrible.

• Most people would immediately think, "Oh no, I messed up! I need to fix this."
• Some people with a specific brain injury (Left Hemisphere Stroke) might take that salty bite, shrug, and say, "This tastes perfect. I did everything right." They made a huge mistake, but their brain refuses to tell them they are wrong.

This paper is about that specific condition called Unawareness of Apraxia (UA). "Apraxia" is the medical term for the inability to perform learned movements (like using a hammer or scissors) correctly. "Unawareness" means the person doesn't know they are doing it wrong.

The researchers wanted to answer two big questions:

1. Why does this happen? (What is broken in the brain?)
2. What happens because of it? (Does it stop them from fixing their mistakes?)

---

The Theory: The "Internal GPS"

To understand the study, imagine your brain has a GPS system for every tool you use.

1. The Map (Action Knowledge): This is your brain's memory of what a movement should look and feel like. It's the mental image of "how to hold a hammer" or "how to swing scissors."
2. The Drive (Motor Command): Your brain sends instructions to your hand to move.
3. The Feedback (Sensors): Your eyes and muscles tell your brain, "Hey, the hammer is actually moving like this."
4. The Comparison (The Error Detector): Your brain compares the Map (what you wanted to do) with the Drive (what you actually did).
   • If they match: Great!
   • If they don't match: ALARM! The brain says, "Wait, that's wrong!" and tries to fix it.

The Researchers' Hypothesis:
They thought that in people with UA, the Map (Action Knowledge) is blurry or missing. If you don't have a clear picture of what "using a hammer correctly" looks like, your brain can't tell that your hand is doing it wrong. Without a clear map, there is no alarm bell, so the person never realizes they made a mistake.

---

What They Did (The Experiment)

The team studied 56 people who had strokes on the left side of their brains. They asked them to:

1. Pantomime Tool Use: Show how to use a hammer, scissors, or comb without the actual object.
2. Self-Check: Immediately after, ask, "Did you feel like you made any mistakes?"
3. Test the "Map": Show them videos of people using tools and ask, "Is this the right way to use it?" (This tests their Action Knowledge).
4. Watch for Fixes: Did the person notice a mistake while doing it and try to correct their hand movement?

They split the stroke survivors into three groups:

• Group A: No stroke symptoms (Healthy).
• Group B: Had trouble using tools, but knew they were having trouble (Aware).
• Group C: Had trouble using tools, but insisted they were doing it fine (Unaware/UA).

---

The Results: The "Blurry Map" Theory Proven

Here is what they found, translated into plain English:

1. It's not about how bad the stroke was.
The "Unaware" group and the "Aware" group made the exact same number of mistakes. They also had similar levels of general brain damage. So, being unaware isn't just because the stroke was "worse." It's something specific to how they process the idea of the movement.

2. The "Unaware" group had a broken "Map."
When tested on their knowledge of how tools should be used (the Action Knowledge task), the "Unaware" group scored significantly lower than everyone else.

• The Metaphor: Imagine trying to navigate a city without a map. You might drive in circles, but you won't know you're lost because you don't know what the destination looks like. The "Unaware" group had a blurry or missing map of how tools work.

3. The "Unaware" group never tried to fix their mistakes.
This was the most important finding.

• The "Aware" group would make a mistake, realize it, and try to adjust their hand to fix it.
• The "Unaware" group made the mistake, kept going, and never tried to correct it.
• The Metaphor: If you are driving with a broken GPS, you might take a wrong turn. If you have a working GPS, you see the red line, hear the voice say "Recalculating," and turn the wheel. The "Unaware" group didn't have the GPS voice, so they just kept driving down the wrong road.

---

Why This Matters (The Takeaway)

This study changes how we might help people recover after a stroke.

• Old Thinking: "If they can't move their hand right, we just need to practice moving the hand more."
• New Thinking: "If they don't know they are wrong, practicing the movement won't help. We first need to fix their mental map (Action Knowledge)."

If a patient doesn't realize they are making errors, they won't try to correct them. Rehabilitation needs to focus on helping them rebuild that internal "GPS" of how tools should look and feel. Once they can see the difference between "right" and "wrong" in their mind, they might finally start fixing their movements in the real world.

In short: You can't fix a mistake if you don't know you made one. For these patients, the problem isn't just the hand; it's the brain's ability to recognize the error.

Technical Summary

1. Problem Statement

Limb apraxia is a neurological disorder characterized by errors in pantomimed tool-use gestures and actual tool use, occurring frequently after left-hemisphere cerebral vascular accidents (LCVA). A significant subset of apraxic patients exhibits Unawareness of Apraxia (UA), also known as anosognosia of apraxia. These individuals perform tool-use gestures incorrectly but genuinely believe their performance is accurate.

Despite the potential impact of UA on rehabilitation outcomes and functional independence, its prevalence, underlying mechanisms, and consequences remain poorly understood. Specifically, it is unclear whether UA is merely a byproduct of severe apraxia or a distinct deficit, and how it affects a patient's ability to self-correct errors.

2. Methodology

Participants:

• Sample: 56 chronic LCVA survivors (stroke >6 months ago) and 16 neurotypical controls.
• Inclusion/Exclusion: Participants were right-handed, had no history of psychiatric disorders or TBI, and possessed sufficient language comprehension (WAB comprehension score $\ge$ 4).
• Grouping: Based on a Gesture-to-Sight (GTS) task cutoff (2 SD below neurotypical mean), participants were classified into:
  1. Non-Apraxic ($n=35$)
  2. Apraxic-Aware ($n=11$): Performed poorly on GTS but acknowledged errors.
  3. Unaware of Apraxia (UA) ($n=10$): Performed poorly on GTS but denied making errors.

Tasks:

1. Gesture-to-Sight (GTS) Task: Participants viewed 39 photographs of common tools and pantomimed their use with the ipsilesional left hand. Performance was scored on spatiotemporal dimensions (hand posture, arm posture, amplitude/timing).
   • Error Correction Coding: Trials were specifically coded for attempts to self-correct errors during the gesture.
2. Apraxia Awareness Question: Immediately post-GTS, participants were asked if they felt they had difficulty showing how to use the objects despite knowing their purpose.
3. Gesture Recognition (GR) Task: A verbal-action matching task where participants identified the correct gesture video from a foil (containing spatiotemporal or semantic errors). This measures the integrity of action knowledge (stored movement goals).
4. NIH Stroke Scale (NIHSS): Used to control for overall stroke severity.

Statistical Analysis:

• Linear Mixed Models (LMM) and General Linear Mixed Models (GLMM) were used to compare groups (Non-Apraxic, Apraxic-Aware, UA) on GTS accuracy, GR accuracy, and error correction rates.
• Pearson correlations assessed the relationship between action knowledge (GR) and error correction attempts.

3. Key Contributions & Hypotheses

The study tested a mechanistic model of performance awareness (Fig. 1 in the paper) proposing that:

1. Action Knowledge as a Goal: Stored knowledge of how an action "looks and feels" serves as the motor goal.
2. Comparator Mechanism: Sensory feedback is compared against this goal to generate an error signal.
3. Hypothesis 1 (Cause): UA arises because degraded action knowledge leads to a poorly specified motor goal, resulting in a weak or noisy error signal that fails to trigger awareness.
4. Hypothesis 2 (Consequence): Due to the lack of an error signal, individuals with UA will exhibit significantly fewer attempts to correct their errors compared to aware apraxics.

4. Results

• Prevalence: Nearly half (47.6%) of the apraxic participants exhibited UA.
• Severity vs. Awareness:
  • There was no significant difference in apraxia severity (GTS scores) between the Apraxic-Aware and UA groups.
  • There was no significant difference in overall stroke severity (NIHSS) between the UA and Apraxic-Aware groups.
  • Conclusion: UA is not simply a function of the severity of the motor deficit or general stroke impairment.
• Action Knowledge Deficit:
  • The UA group showed significantly impaired action knowledge (lower GR accuracy) compared to both the Apraxic-Aware and Non-Apraxic groups.
  • The Apraxic-Aware group performed similarly to Non-Apraxic controls on the GR task.
  • Conclusion: UA is specifically linked to a degradation of stored movement goals (action knowledge).
• Error Correction:
  • The UA group made significantly fewer error correction attempts than the Apraxic-Aware group.
  • A strong positive correlation ($\rho = 0.60, p = 0.002$) was found between action knowledge (GR accuracy) and the proportion of trials with error corrections.
  • Conclusion: Poor action knowledge directly predicts a failure to detect and correct errors.

5. Significance and Implications

Theoretical Significance:

• Mechanism of Anosognosia: The study provides evidence that UA in apraxia is driven by a specific deficit in action knowledge (the "what" and "how" of the movement) rather than a general failure of the praxis system or sensory feedback loops alone.
• Differentiation of Subtypes: The findings support a distinction between Ideational Apraxia (deficits in action knowledge/conceptualization, associated with UA) and Ideomotor Apraxia (deficits in transforming intact knowledge into motor commands, often associated with preserved awareness).
• Error Monitoring: It establishes that error detection relies on a robust comparison between a stored goal and executed movement; without the goal, the comparator cannot function.

Clinical Significance:

• Rehabilitation Strategy: Standard apraxia therapy often focuses on motor practice. The findings suggest that for patients with UA, rehabilitation must first target the restoration of action knowledge (e.g., through gesture recognition training) to re-establish the internal "goal" necessary for self-monitoring.
• Assessment: Clinicians should systematically assess UA, as it is a distinct syndrome affecting nearly 50% of apraxic patients. Identifying UA is crucial for managing caregiver expectations and tailoring interventions, as these patients are unlikely to self-correct without external feedback.
• Prognosis: The lack of error correction attempts in UA patients suggests a barrier to functional recovery that may require specific compensatory strategies.

In summary, this paper demonstrates that unawareness of apraxia is a specific cognitive deficit rooted in degraded action knowledge, which subsequently disables the brain's error-correction mechanisms, distinguishing it from general motor severity.