Task-based functional connectivity in striato-motor-cortical system in autism: Associations with sex and executive function

This study reveals that female autistic youth exhibit reduced functional connectivity between the anterior putamen and frontal motor regions compared to non-autistic peers, and that this specific connectivity pattern significantly predicts executive function performance across all participants regardless of sex or neurotype.

The Gist

The Big Picture: A "Traffic Control" Problem in the Brain

Imagine your brain is a massive, bustling city. In this city, there are two main types of districts:

1. The "Motivation & Action" District (The Striatum/Putamen): This is like the city's central train station. It decides what is interesting, what you want to go toward, and gets the engines running.
2. The "Command Center" (The Frontal Lobe): This is the mayor's office. It handles planning, focus, and controlling your behavior (executive function).

In a healthy city, the train station and the mayor's office talk to each other constantly. The station says, "Hey, look at that interesting human motion!" and the mayor's office says, "Great, let's focus our attention on it and ignore the noise."

The Study's Goal:
Researchers wanted to see how this conversation works in autistic girls compared to non-autistic girls, and how it relates to their ability to focus and control their behavior. They were specifically looking at a previous finding that suggested the "train station" (putamen) in autistic girls might be a bit quieter or different than in others.

---

The Experiment: Watching a Movie vs. Static

The researchers put 184 kids (ages 8–17) in an MRI machine. They showed them two things:

1. Coherent Motion: A movie made of dots moving together to look like a person walking or dancing (like a stick figure animation).
2. Scrambled Motion: The same dots, but moving randomly, like static on an old TV.

They asked the kids to just watch. The goal was to see how the "train station" (putamen) talked to the "Command Center" (frontal lobe) when the brain was trying to make sense of a human moving.

---

The Three Big Discoveries

1. The "Missing Phone Call" in Autistic Girls

The Finding: When watching the human motion, non-autistic girls had a very strong, clear conversation between their left "train station" (anterior putamen) and their "Command Center" (dorsal premotor cortex). It was like a high-speed fiber-optic line.

The Analogy: Imagine a non-autistic girl's brain as a well-tuned radio. When a song (human motion) starts playing, the radio instantly connects to the speaker system, and the music fills the room clearly.

The Difference: In autistic girls, this connection was much weaker. It was like the radio was trying to connect to the speaker, but the signal was fuzzy or the volume was turned down. This suggests that the part of the brain responsible for "getting excited" about social movement isn't effectively telling the "focus" part of the brain to pay attention.

2. The "Wrong Connection" in Autistic Boys

The Finding: The researchers also looked at autistic boys. Surprisingly, they found that autistic boys had a different kind of weird connection. Their "train station" (posterior putamen) was talking to the right side of the "Command Center," which isn't a normal route for most people.

The Analogy: If the non-autistic brain is a standard map where the train station connects to the main office, the autistic boy's brain in this study looked like someone trying to take a shortcut through a back alley that usually isn't used. It's a different, perhaps less efficient, way of trying to process the same information.

3. The "Volume Knob" for Focus

The Finding: The researchers then asked: "Does this connection matter for real-life skills?" They looked at how well the kids could control their attention and behavior (Executive Function).

The Analogy: Think of the connection between the train station and the mayor's office as a volume knob for attention.

• High Volume (Strong Connection): The brain can easily turn up the focus when something interesting happens. These kids had better executive function (better focus, less impulsivity).
• Low Volume (Weak Connection): The brain struggles to turn up the focus. These kids had more trouble with executive function.

The Key Takeaway: It didn't matter if the child was autistic or not, or if they were a boy or a girl. Whoever had the strongest "volume knob" connection between the putamen and the frontal lobe had the best ability to control their attention and behavior.

---

Why This Matters

This study is like finding a specific broken wire in a complex machine.

• For Autistic Girls: It suggests that their unique challenges might stem from a specific "wiring issue" where the brain's motivation center doesn't effectively signal the focus center during social situations. This helps explain why they might struggle to pay attention in social settings even if they are smart.
• For Everyone: It shows that the strength of this specific brain connection is a universal predictor of how well a person can control their behavior.

In short: The brain needs a strong, clear line between "what I want to look at" and "how I control my attention." When that line is weak, focus becomes harder. This research helps us understand that for autistic girls, that line might be wired differently, offering a new target for understanding and supporting their unique brains.

Technical Summary

1. Problem Statement and Background

Autism Spectrum Disorder (ASD) is diagnosed significantly more often in males than females (~3.4:1), leading to concerns about ascertainment bias and the under-identification of autistic females. Previous work by the authors (Jack et al., 2021) identified that autistic females (Aut-F) exhibit reduced neural response in the dorsal striatum (specifically the putamen) during social perception tasks compared to non-autistic females (NAut-F). Additionally, Aut-F showed larger rare copy number variants (CNVs) containing genes expressed in early striatal development.

While these structural and activation differences were established, the systems-level implications remained unclear. Specifically, it was unknown how these striatal differences affected functional connectivity within the broader striato-motor-cortical system and whether these connectivity patterns could explain variability in executive function (EF), a core challenge in autism that often varies by sex.

2. Methodology

The study utilized secondary data analysis from the GENDAAR cohort, a sex-balanced sample of youth (ages 8–17).

• Participants:
  • Group Differences Analysis (Objective 1): $N=184$ (45 Aut-F, 47 Aut-M, 45 NAut-F, 47 NAut-M). Groups were matched on age, Full Scale IQ (FSIQ), sex, and head motion.
  • Individual Differences Analysis (Objective 2): $N=200$ (including all available cases with complete executive function data).
• Experimental Task: Participants underwent fMRI while passively viewing coherent (BIO) versus scrambled (SCRAM) point-light displays of human motion. The contrast of interest was BIO > SCRAM (social perception).
• Imaging Analysis:
  • Psychophysiological Interaction (PPI): Conducted in FSL to assess task-specific functional connectivity.
  • Seeds: Anterior Putamen (Pa) and Posterior Putamen (Pp), segmented using FIRST and divided based on connectivity-based parcellation.
  • Regions of Interest (ROIs): Dorsal Frontal Cortex (DFC; BA 9 & 46d), Dorsal Premotor Cortex (PMd/pre-SMA; BA 6), Primary Motor Cortex (M1), and Insula.
  • Statistical Inference: Randomise with Threshold-Free Cluster Enhancement (TFCE), 10,000 permutations, corrected $p < .05$. Covariates included age, site, FSIQ, and Social Responsiveness Scale (SRS-2) scores.
• Behavioral Measures: Executive function was assessed using the BRIEF-GEC (Behavior Rating Inventory of Executive Function, 2nd Ed. General Executive Composite).
• Modeling (Objective 2): A Best Subsets Regression approach (using leaps in R) with repeated cross-validation (10-fold, 10 repeats) was used to identify the optimal model predicting BRIEF-GEC scores. Features included PPI connectivity values, neurotype, sex, age, FSIQ, and interaction terms.

3. Key Results

Objective 1: Group Differences in Functional Connectivity

• NAut-F vs. Aut-F: Non-autistic females showed significantly greater task-specific functional connectivity between the Left Anterior Putamen (L-Pa) and the Dorsal Premotor Cortex/Pre-SMA (PMd/pre-SMA) compared to Aut-F. This suggests a disruption in the typical pathway linking the anterior striatum to attentional control regions in autistic females.
• Aut-M vs. Aut-F: Autistic males showed significantly greater connectivity between the Left Posterior Putamen (L-Pp) and the Right Dorsolateral Prefrontal Cortex (R-DFC) compared to Aut-F. This pattern was unexpected, as Pp-DFC connectivity is not typical in neurotypical populations, suggesting a potential marker of dysconnectivity specific to autistic males.
• Sex Differences: No significant neurotype-by-sex interactions were found for the Aut-M vs. NAut-M comparison in the primary Pa-DFC pathway.

Objective 2: Prediction of Executive Function

• Optimal Model: The best subsets regression identified a three-variable model that minimized prediction error (RMSE = 0.16; Adj. $R^2$ = 0.67) for BRIEF-GEC scores.
• Key Predictors:
  1. Neurotype: Autistic status was a strong predictor of executive deficits.
  2. L-Pa to L-DFC Connectivity: Greater functional connectivity between the Left Anterior Putamen and Left Dorsolateral Prefrontal Cortex predicted better executive function (lower BRIEF-GEC scores) across all participants.
  3. Interaction: An interaction term between connectivity and sex was included, though the main effect of connectivity remained robust.
• Interpretation: Stronger coupling between the anterior striatum (involved in motivation/valuation) and the prefrontal cortex (involved in attentional control) is associated with superior executive functioning, regardless of autism diagnosis.

4. Key Contributions

1. Sex-Specific Biomarkers: The study provides evidence that the striatal endophenotype in autism is sex-specific. Autistic females show a specific reduction in the Pa-PMd/pre-SMA pathway, which is distinct from the patterns seen in autistic males.
2. Systems-Level Mechanism: It moves beyond local activation deficits to demonstrate that functional connectivity within the striato-motor-cortical system is altered in autistic females, specifically affecting the integration of the anterior striatum with executive control networks.
3. Link to Behavior: The study establishes a direct link between neural connectivity (L-Pa to L-DFC) and real-world executive functioning. It suggests that the "adaptive" capacity of the brain in autism may depend on the integrity of this specific fronto-striatal loop.
4. Methodological Rigor: The use of a sex-balanced, IQ-matched cohort and advanced machine learning techniques (best subsets regression with cross-validation) strengthens the validity of the findings regarding sex differences and individual variability.

5. Significance and Implications

• Understanding the "Female Phenotype": These findings support the hypothesis that autistic females may utilize different neural strategies or suffer from specific connectivity disruptions (specifically in the anterior striatum) that contribute to their unique phenotypic presentation and potential diagnostic delays.
• Biomarker Development: The L-Pa to L-DFC connectivity metric shows promise as a potential biomarker for executive dysfunction in autism, applicable across sexes but with distinct baseline differences in females.
• Therapeutic Targets: By identifying the specific circuit (anterior putamen to prefrontal cortex) involved in attentional regulation and executive function, future interventions could target the modulation of this pathway to improve adaptive outcomes.
• Limitations & Future Directions: The study notes the lack of hormonal data (menstrual cycle phase), which is a known confounder in female neuroimaging due to estradiol's effect on striatal function. Future work should replicate these findings in independent samples and incorporate hormonal monitoring.

In summary, this paper elucidates how sex and neurotype interact to shape the functional architecture of the striato-motor-cortical system in autism, revealing that reduced connectivity between the anterior putamen and frontal executive regions is a key characteristic of autistic females and a significant predictor of executive function deficits.