# MRI Based Presymptomatic Prediction of ASD

> **NIH NIH R01** · WASHINGTON UNIVERSITY · 2021 · $2,548,810

## Abstract

ABSTRACT
The overarching goal of this proposal is to lower the age of detection in autism to early infancy, making
presymptomatic (i.e., before the emergence of ASD-specific behavioral features) intervention feasible. Infants
with an older autistic sibling have up to a 20% risk of developing autism spectrum disorder (ASD). Prospective
high familial risk (HR) infant sibling studies have shown that the defining behaviors of ASD do not emerge until
the latter part of the first year and into the second year of life. Therefore, the vast majority of affected children
are diagnosed after age 2. No behavioral markers in the first year of life have yet been identified that can
predict later ASD diagnosis with sufficient accuracy (i.e., positive predictive value: PPV ≥ 80%) to justify
presymptomatic intervention. We recently published two independent approaches that use brain imaging in the
first year of life to predict which HR infants will be diagnosed with ASD at 2 years of age. Specifically, structural
MRI (sMRI) at 6 and 12 months of age, and resting state functional connectivity MRI (fcMRI) at 6 months of
age independently predicted later ASD diagnosis in HR infants with over 80% PPV. Our preliminary data show
that a third MRI approach, using regions of CSF volume and cortical shape at 6 months of age can also
accurately predict later ASD diagnosis. If we replicate and extend these findings, we will be able to identify
individual infants at “ultra-high risk” (80% chance) of developing ASD, rather than being limited to group-level
risk (20% chance), where we do not know who will later be affected. This R01 application aims to move our
initial findings toward a clinical test for ASD in HR infants in the first year of life. Aim 1 will validate our previous
findings in a new, independent sample of HR infants, extend our methods to a new MRI platform, and examine
whether fcMRI and/or sMRI, with and without behavioral information, during the presymptomatic period in
infancy, accurately predict ASD diagnosis at 24 months of age. Aim 2 will move beyond predicting categorical
diagnosis to predicting dimensional, clinically-relevant characteristics for individual infants. Specific
dimensional targets include expressive language level, social responsiveness, initiation of joint attention, and
repetitive behavior. Validating and extending our findings on presymptomatic prediction of ASD in a new
sample, on a different MRI scanner, and with dimensional developmental characteristics are critical next steps
for moving the field forward toward (a) the development of a clinically-useful, presymptomatic test for
identifying ultra-high risk infants who would benefit from very early intervention in infancy, (b) efficient studies
of presymptomatic intervention strategies in individuals at ultra-high risk, and (c) the development of future
presymptomatic tests for use in the general (not just HR) population.

## Key facts

- **NIH application ID:** 10109147
- **Project number:** 5R01MH118362-03
- **Recipient organization:** WASHINGTON UNIVERSITY
- **Principal Investigator:** Joseph Piven
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2021
- **Award amount:** $2,548,810
- **Award type:** 5
- **Project period:** 2019-04-01 → 2024-01-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10109147

## Citation

> US National Institutes of Health, RePORTER application 10109147, MRI Based Presymptomatic Prediction of ASD (5R01MH118362-03). Retrieved via AI Analytics 2026-05-22 from https://api.ai-analytics.org/grant/nih/10109147. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*