Predictive value of morphological features in patients with autism versus normal controls

Area of Science:

• Neurodevelopmental disorders
• Human genetics
• Biometrics

Background:

• Autism spectrum disorder (ASD) diagnosis relies on behavioral observation.
• Objective, quantifiable biomarkers for ASD are needed.
• Morphological variations have been anecdotally linked to neurodevelopmental conditions.

Purpose of the Study:

• To investigate the predictive value of physical (morphological) features in identifying autism spectrum disorder (ASD).
• To determine if specific patterns of morphological abnormalities can differentiate autistic children from neurotypical controls.
• To explore the utility of morphological data in developing objective diagnostic tools for ASD.

Main Methods:

• Receiver Operator Curves (ROC) analysis was used to assess the relationship between morphological features and ASD.
• Recursive Partitioning (RP) was employed to identify characteristic patterns of abnormalities.
• A cohort of 224 autistic patients and 224 matched neurotypical controls were analyzed for various morphological traits.

Main Results:

• Morphological features were found to be significantly more prevalent in individuals with ASD compared to controls.
• Specific morphological measures demonstrated high predictive accuracy for ASD identification.
• Facial asymmetry, multiple hair whorls, and a prominent forehead were identified as key differentiating features.

Conclusions:

• Morphological features possess significant predictive power for identifying autism spectrum disorder (ASD).
• Objective physical markers like facial asymmetry can aid in differentiating autistic individuals.
• Incorporating morphological data into multivariable models may enhance ASD risk prediction.