🧠 neuroscience

Interoceptive autonomic regulation in typical development and autism spectrum disorder: A computational model integrating multiple physiological systems

This study employs a computational model to demonstrate that while typically developing individuals exhibit differentiated sympathetic-parasympathetic coordination during postural challenges, individuals with autism spectrum disorder show convergent autonomic activity patterns, with respiration identified as a key regulatory lever for cardiovascular stabilization.

Original authors: Li, R., Liu, H., Nagai, Y.

Published 2026-02-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Li, R., Liu, H., Nagai, Y.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: The Body's "Internal Dashboard"

Imagine your body has a complex dashboard inside your chest, constantly monitoring your heart rate (HR) and blood pressure (BP). This dashboard is run by a control center called the Autonomic Nervous System (ANS). Think of the ANS as a two-person management team:

The Gas Pedal (Sympathetic System - SNS): This speeds things up when you need energy (like running from a bear).
The Brake Pedal (Parasympathetic System - PSNS): This slows things down to help you rest and digest.

In a healthy, "Typically Developing" (TD) person, these two pedals work in perfect harmony. When you stand up quickly, the Gas Pedal presses down, and the Brake Pedal lifts up immediately to keep your blood pressure stable.

The Problem: In people with Autism Spectrum Disorder (ASD), this dashboard often acts strangely. Sometimes the gas is stuck, sometimes the brake is stuck, or sometimes both are pressed at the same time. Scientists have seen this in real life, but they haven't had a good way to simulate exactly how these two systems are misfiring to cause the symptoms.

The Solution: A "Virtual Body" Computer Model

The authors of this paper built a computer simulation (a "virtual body") to understand what's happening inside. They didn't just look at the heart; they built a model that connects:

The Heart & Blood Vessels (The Engine)
The Lungs (The Air Supply)
The Nervous System (The Control Center)

They used this model to run a virtual test called the Head-Up Tilt (HUT). Imagine lying flat on a table, and then the table suddenly tilts you upright. This is a stress test for your body's balance system.

What They Discovered: The "Dance" of the Pedals

The researchers tested three different ways the Gas and Brake pedals could interact:

The Perfect Dance (Coupled Reciprocal): When one goes up, the other goes down. (This is what healthy bodies do).
The Team Huddle (Coupled Nonreciprocal): Both pedals move in the same direction (both up or both down).
The Solo Act (Uncoupled): One pedal moves while the other stays frozen.

The Findings:

Typical People (TD): When the table tilted, their virtual bodies switched into the "Perfect Dance" mode. The Gas pedal pressed hard, and the Brake pedal let go. This kept their blood pressure stable.
Autistic People (ASD): Their virtual bodies didn't switch modes as easily. Instead of a clear "Gas on, Brake off" signal, their systems showed a confused mix. The model suggested that in ASD, the "Brake" (Parasympathetic system) wasn't letting go as much as it should, even when the body needed to speed up. It's like trying to drive up a hill with your foot still on the brake.

Why does this matter?
Because the "Brake" wasn't releasing, the body struggled to adjust to the change in position. This creates a state of instability. The brain receives confusing signals about what the body is doing, which might explain why people with autism sometimes feel overwhelmed, anxious, or have trouble processing their own emotions (interoception).

The Secret Weapon: Breathing

The second part of the study asked: Can we fix this virtual dashboard?

They tested a simple intervention: Breathing.

Normal Breathing: Didn't change much.
Deep, Slow Breathing: This was a game-changer.

When they simulated deep breathing, the "Brake" pedal became much more effective at lowering blood pressure, even when the "Gas" pedal was stuck in the "on" position.

The Analogy: Imagine your blood pressure is a bucket of water that is overflowing (too high).

Normal breathing is like using a small cup to scoop water out. It helps a little.
Deep breathing is like opening a large drain valve. It clears the water out much faster and more efficiently, stabilizing the system even when the "Gas" is still running hot.

The Takeaway

It's a Coordination Issue, Not Just a "High Heart Rate" Issue: The study suggests that autism isn't just about having a fast heart rate; it's about the timing and coordination between the "Gas" and "Brake" systems. They aren't talking to each other correctly.
Breathing is a Remote Control: The model proves that deep, slow breathing isn't just "relaxing"; it physically changes how the nervous system regulates blood pressure. It acts as a lever to help the body stabilize itself.
A New Tool for Understanding: This computer model gives scientists a way to "tweak" the settings in a virtual body to see what causes specific symptoms. This could lead to better, personalized therapies for people with autism, helping them regulate their bodies and, consequently, their minds.

In short: The paper uses a computer to show that in autism, the body's internal "gas and brake" systems are out of sync. However, by using deep breathing, we can help retune that system and bring the body back to a stable, calm state.

1. Problem Statement

Autism Spectrum Disorder (ASD) is frequently associated with atypical interoceptive processing (the perception of internal bodily states), which contributes to sensory atypicalities and emotional dysregulation. While empirical studies have reported inconsistent findings regarding autonomic nervous system (ANS) dysfunction in ASD—ranging from sympathetic dominance to parasympathetic deficits or normal baseline activity—there is a lack of a unified, mechanistic framework to explain these discrepancies.

The Gap: Existing computational models of interoception often focus on single physiological systems or lack the resolution to characterize how specific ANS coordination patterns (control modes) and branch-specific activities (Sympathetic [SNS] vs. Parasympathetic [PSNS]) differ between Typically Developing (TD) and ASD populations.
The Goal: To develop a closed-loop computational model integrating cardiovascular, respiratory, and autonomic systems to quantitatively estimate latent ANS regulatory structures and compare them between TD and ASD individuals under physiological stress.

2. Methodology

The authors developed a closed-loop multiscale computational model that integrates three subsystems:

Cardiovascular System (CVS): A hemodynamic model based on Navier-Stokes approximations and lumped-parameter circuits to simulate Heart Rate (HR) and Blood Pressure (BP) waveforms. It includes arterial/venous compliance, vascular resistance, and cardiac elastance.
Respiratory System (RS): Modeled as a source of Intrathoracic Pressure (ITP) fluctuations that modulate venous return and BP waveforms. It simulates normal and deep breathing patterns.
Autonomic Nervous System (ANS): The core innovation of the study. The ANS is formalized using two distinct dimensions:
- Autonomic Control Modes: Three structural patterns of SNS-PSNS coordination defined by Berntson et al.:
  - Coupled Reciprocal: SNS and PSNS change in opposite directions (antagonistic).
  - Coupled Nonreciprocal: SNS and PSNS change in the same direction (co-activation or co-inhibition).
  - Uncoupled: One branch varies independently of the other.
- Relative Activity: Weighting parameters ( $\alpha$ for SNS, $\beta$ for PSNS) that scale the gain of each branch's output.

Experimental Protocol:

Stimulus: The model simulated a Head-Up Tilt (HUT) test, a standard orthostatic challenge used to assess autonomic regulation.
Data Integration: Empirical HR and BP data from TD and ASD groups (pre- and post-HUT) were used to initialize afferent inputs and define target baselines.
Simulation Strategy: The model generated 3D response surfaces mapping HR and BP outcomes across the full spectrum of SNS-PSNS activity combinations ($0$ to $1.0$) for each control mode. These surfaces were compared against empirical data ranges to identify the underlying regulatory states.
Respiratory Intervention: A second experiment evaluated the impact of deep respiration (low frequency, high amplitude) vs. normal respiration on Mean Arterial Pressure (MAP) recovery during the coupled reciprocal mode.

3. Key Contributions

Novel Computational Framework: The study introduces a model that decouples structural control modes from branch-specific activity levels, allowing for a more granular analysis of autonomic regulation than previous baroreflex-only models.
Mechanistic Characterization of ASD: It provides a state-space characterization of autonomic coordination, moving beyond descriptive statistics to infer latent regulatory mechanisms.
Unified Baseline Approach: The authors demonstrate the critical importance of using a unified regulatory target (baseline) for both groups. Using group-specific baselines (e.g., setting the ASD target to their elevated resting HR) masks autonomic imbalances, whereas a unified target reveals the underlying dysregulation.
Respiratory Modulation: The model identifies deep respiration as a specific regulatory lever that can enhance parasympathetic efficacy in lowering blood pressure, particularly under conditions of high sympathetic tone.

4. Key Results

A. Autonomic Regulation Differences (TD vs. ASD)

TD Individuals: Exhibited differentiated SNS-PSNS coordination patterns depending on the control mode. Under the coupled reciprocal mode (the most physiologically expected response to HUT), TD data clustered in regions of high SNS and low PSNS activity, consistent with a compensatory increase in HR and BP stabilization.
ASD Individuals: Showed convergence of relative SNS-PSNS activity across all three control modes.
- Empirical ASD data extended toward regions of relatively higher PSNS weighting compared to TD, even under postural stress.
- This suggests a blunted parasympathetic reactivity or an inability to effectively suppress PSNS activity during stress, leading to reduced flexibility in autonomic control strategies.
- The lack of differentiation across control modes implies ANS instability, where the system fails to recruit a specific, stable regulatory strategy in response to external stimuli.

B. Mechanisms of Regulation

Direction of Change: The model revealed that the direction of HR/BP change is determined by the difference between the afferent signal (current state) and the modulation signal ( $\sigma$ ), which is a weighted sum of SNS and PSNS outputs.
Baseline Sensitivity: Simulations confirmed that if the baseline target is set to the ASD group's elevated resting HR, the model incorrectly predicts "normal" regulation. Using a unified baseline (TD resting HR) correctly exposes the ASD group's initial state of elevated SNS drive and reduced PSNS capacity.

C. Respiratory Effects on Recovery

Deep Respiration: In the coupled reciprocal mode, introducing deep respiration (9 bpm) significantly enhanced the reduction of Mean Arterial Pressure (MAP) during recovery compared to normal respiration or no respiration.
Synergy: Under conditions of high SNS activity (simulating stress), deep respiration amplified the effectiveness of PSNS modulation, resulting in a steeper decline in MAP. This supports the hypothesis that controlled breathing can act as a "regulatory lever" to stabilize cardiovascular dynamics in high-stress states.

5. Significance and Implications

Theoretical Insight: The findings support the "predictive coding" framework of interoception, suggesting that the instability and reduced flexibility of ANS control in ASD may lead to elevated interoceptive prediction errors. This physiological noise could underlie the cognitive and emotional processing difficulties observed in ASD (e.g., reduced confidence in bodily states, anxiety).
Clinical Application: The model provides a quantitative tool to infer latent autonomic states from observable phenotypes (HR/BP). It suggests that respiratory interventions (specifically deep or resonance-frequency breathing) could be effective non-pharmacological strategies to restore cardiovascular stability in individuals with ASD who exhibit autonomic dysregulation.
Future Directions: The framework lays the groundwork for personalized physiological interventions and highlights the need to integrate multimodal interoceptive signals (electrodermal, gastrointestinal) to fully capture the complexity of neurodivergent autonomic regulation.

In summary, this paper successfully bridges the gap between empirical inconsistencies in ASD autonomic research and mechanistic theory by using a computational model to reveal that ASD is characterized not just by static imbalances, but by a reduced flexibility in switching autonomic control modes and a blunted ability to modulate parasympathetic activity under stress.