Pupil Constriction Causes Activity in the Human Retina and Visual System

This study demonstrates that pupil constriction, independent of visual stimulation, triggers distinct activity in both the human retina and visual cortex, likely as a response to the sudden reduction in light exposure.

The Gist

Imagine your eye is a high-tech camera, and your pupil is the aperture (the little hole that opens and closes to control how much light gets in).

Usually, we think of the pupil as a passive gatekeeper. It just sits there, opening wide in the dark to let light in, or squinting in the sun to keep it out. We assume the camera sensor (your retina) only reacts when the outside world changes—like when a light bulb turns on or a cloud passes over the sun.

But this new research suggests the pupil is actually a bit of a troublemaker. It doesn't just let light in; it actually creates its own "fake" signals inside your eye and brain, even when nothing outside has changed.

Here is the story of what the scientists found, broken down simply:

1. The "Squint" Surprise

When you look at something bright, your pupil quickly shrinks (constricts) to protect your eye. This happens a split second after your brain sees the light.

The researchers discovered that the moment your pupil shrinks, it causes a sudden, tiny drop in the amount of light hitting the back of your eye (the retina). It's like someone suddenly pulling a curtain closed in a room that was already lit.

The Analogy: Imagine you are standing in a sunny room. Suddenly, someone pulls a heavy curtain shut. Even though the sun outside hasn't changed, the room suddenly gets darker. Your eyes react to that sudden darkness.

2. The Eye's "False Alarm"

The scientists found that your retina (the camera sensor) reacts to this sudden "curtain closing" just as if a real object had appeared or disappeared.

They recorded electrical signals from the eyes of 119 people. They found a specific electrical "blip" that happens exactly when the pupil shrinks, not when the light first hits the eye.

• The Catch: This blip is tiny, about 1/10th the size of the signal caused by a real light flash. But it's real.
• The Speed: The faster the pupil shrinks, the bigger the "blip." It's like a faster shutter closing creates a bigger "whoosh" of change.

3. The Brain Gets Confused (Briefly)

This isn't just happening in the eye. The signal travels up to the brain's visual cortex (the part that processes what you see). About 100 milliseconds after the pupil shrinks, the brain shows a tiny electrical reaction over the back of the head.

The Metaphor: It's like your brain is a security guard.

1. Real Event: Someone kicks the door (a light flash). The guard jumps.
2. Pupil Event: The guard realizes the door is closing on its own. Even though no one kicked the door, the sound of the door slamming makes the guard jump again.

4. Why Don't We See "Flickers"?

If your eye and brain are reacting to every time your pupil shrinks, why doesn't the world look like it's flickering or dimming every time you blink or look at a bright light?

The Mystery: The paper suggests your brain is incredibly smart at ignoring these signals. It knows, "Oh, that's just my pupil shrinking, not the world getting darker." It cancels out the noise so you can see a stable, steady world.

However, the researchers found a clue: If you mess with the pupil using eye drops (so the brain can't "feel" the pupil moving), people do perceive brightness changes. This suggests your brain uses a "copy of the command" (a mental note saying "I am shrinking the pupil") to tell the visual system, "Ignore this signal; it's just me."

The Big Picture

This study changes how we think about vision. We used to think the pupil just adjusts the volume of light. Now we know the pupil is an active participant that creates its own "noise" in the visual system.

In short: Your pupil isn't just a passive window; it's a door that slams shut so hard it shakes the whole house, and your brain has to work hard to pretend it didn't happen.

Technical Summary

1. Problem Statement

Traditionally, pupil size is viewed as a modulatory factor that adjusts the amount of light entering the eye, thereby influencing optical quality and retinal illumination. While it is well-established that pupil size modulates responses to external visual stimuli, it remains unclear whether the mechanical act of pupil constriction itself generates neural activity in the visual system, independent of external stimulation.

Previous research in mice (Lapanja et al.) suggested that pupil constriction triggers retinal ganglion cell (RGC) activity due to the sudden decrease in retinal light exposure. A follow-up human study suggested a perceptual dimming effect during constriction but lacked direct physiological evidence. The authors aim to determine if pupil constriction generates a distinct, measurable response in the human retina (via Electroretinogram, ERG) and visual cortex (via Electroencephalogram, EEG/ERP), separate from the response to the visual stimulus that triggered the constriction.

2. Methodology

Data Source and Participants

• Participants: A combined analysis of nine separate datasets involving 119 healthy adult human participants with normal, uncorrected vision.
• Stimuli: Participants viewed brief visual stimuli (light increments or decrements) designed to trigger a pupil light response.
• Recording:
  • Pupil Size: Recorded at 1000 Hz using an EyeLink 1000 eye tracker.
  • Retinal Activity (ERG): Recorded via four Electrooculogram (EOG) electrodes placed above and below the eyelids (referenced to mastoids). The ERG signal was defined as the average of these four channels.
  • Cortical Activity (ERP): Recorded via 26 scalp electrodes (10-20 system) at 1000 Hz.

Experimental Design & Analysis Strategy

• Key Innovation: The study leveraged trial-to-trial variability in pupil constriction latency. Because the time it takes for the pupil to constrict varies across trials, the researchers could dissociate signals locked to stimulus onset from signals locked to the time of maximum constriction velocity.
• Preprocessing:
  • Pupil data was interpolated for missing values and converted to millimeters.
  • EEG/ERG data was filtered (0.1–40 Hz), re-referenced, and cleaned of muscle artifacts and bad channels using automated tools (MNE, autoreject).
  • Trials were excluded based on blinks, extreme pupil sizes, or unrealistic constriction latencies/velocities.
  • Final Dataset: 48,509 trials (87.6% of total) were retained.
• Statistical Approach:
  • Permutation Testing: To confirm the existence of a constriction-locked component, the authors shuffled the timepoints of maximum constriction velocity within participants/datasets (2,000 permutations) to create a null distribution.
  • Comparison: The actual signal locked to constriction was compared against these permuted baselines using conservative Bonferroni correction ($p < .05$).
  • Velocity Binning: Trials were sorted into 10 bins based on maximum constriction velocity to test for a dose-response relationship.

3. Key Contributions

1. Discovery of a Constriction-Locked ERG Component: The paper identifies a previously unknown retinal response that is time-locked to the moment of maximum pupil constriction velocity, rather than to the visual stimulus onset.
2. Physiological Origin: The study provides evidence that this response originates in the retina (evidenced by its presence in eye electrodes and volume conduction to frontal scalp electrodes) and propagates to the visual cortex.
3. Mechanism Elucidation: It confirms that the mechanical act of pupil constriction (and the resulting rapid drop in retinal illumination) is sufficient to trigger neural activity in the human visual system, independent of external light changes.
4. Methodological Framework: The authors demonstrate a robust method for separating stimulus-locked and pupil-constriction-locked neural signals using trial-to-trial latency variability.

4. Results

Retinal Activity (ERG)

• Constriction-Locked Response: A distinct ERG component was observed when data was locked to the time of maximum constriction velocity.
  • Waveform: Characterized by a small initial negative deflection (in some datasets) followed by a larger positive deflection coinciding with maximum constriction velocity.
  • Amplitude: Roughly one order of magnitude smaller than the standard a- and b-waves triggered by light stimuli.
  • Velocity Dependence: The amplitude of this component scales with the maximum constriction velocity. Trials with faster constriction (more abrupt light decrease) produced larger ERG responses.
  • Source Localization: The effect was strongest at eye electrodes and radiated to frontal electrodes, confirming a retinal origin rather than a cortical or artifact origin.
• Stimulus-Locked vs. Constriction-Locked: When locked to stimulus onset, this component is invisible because it is masked by the variability in pupil latency.

Cortical Activity (ERP)

• Occipital Response: A secondary, smaller positive component emerged over occipital electrodes (O1, O2, Oz) approximately 120 ms after the constriction-locked ERG response (roughly 400 ms before the pupil reaches minimum size).
• Interpretation: This suggests the retinal signal propagates to the early visual cortex, creating a distinct event-related potential (ERP) locked to pupil constriction.

Control Analyses

• The results were robust against blinks, eye movements, and muscle artifacts.
• The constriction-locked ERG was observed even in datasets where the visual stimulus did not elicit a standard a-wave (dark stimuli on bright backgrounds), further isolating the effect to the constriction mechanism.

5. Significance and Implications

Theoretical Implications

• Brightness Constancy Challenge: The findings pose a significant question for the theory of brightness constancy. If the retina and cortex react to pupil-induced light changes as if they were external stimuli, why do humans not perceive the world as flickering or dimming every time their pupils constrict?
  • The authors suggest the visual system may use a corollary discharge (efference copy) of the motor command for pupil constriction to predict and cancel out these retinal signals, similar to mechanisms used for eye movements.
  • This is supported by contrasting studies: pharmacological dilation (which disrupts motor monitoring) leads to brightness overestimation, whereas natural pupil changes do not.

Future Directions

• Optimization of Perception: Pupil dynamics (both constriction and dilation) may not just be passive adjustments but active mechanisms that optimize the signal-to-noise ratio of visual processing.
• Arousal and Vision: The study bridges the gap between arousal-related pupil changes (linked to the locus coeruleus) and direct retinal processing, suggesting that pupil size changes themselves may be a direct route for modulating visual sensitivity.

Conclusion
This study fundamentally shifts the understanding of pupil function from a passive aperture to an active generator of neural activity. It demonstrates that pupil constriction triggers a specific, measurable cascade of activity from the retina to the visual cortex, necessitating a re-evaluation of how the brain maintains stable visual perception amidst constant changes in retinal illumination.