Efficacy of very early interventions on neurodevelopmental outcomes for infants and toddlers at increased likelihood of or diagnosed with autism: A systematic review and meta-analysis.
Andrea McGlade, Koa Whittingham, Jacqui Barfoot and 2 others
PMID 37036800WHAT IT FOUND
Very early interventions for infants and toddlers at risk of or diagnosed with autism showed little to no difference in autism symptoms, cognition, or language by age 3.
A small adaptive-skills gain was found but rated very low certainty.
Key findings
01Blinded clinician-assessed outcomes showed no significant difference between intervention and control groups for autism symptomatology (MD = -0.08, p = 0.75), cognition (SMD = 0.05, p = 0.68), receptive language (SMD = 0.04, p = 0.74), or expressive language (SMD = 0.06, p = 0.45) by 3 years of age.
02A small improvement in the VABS-2 adaptive behavior composite (SMD 0.36, p = 0.009) and daily living skills (MD 3.64, p = 0.04) was statistically significant, but all contributing studies were rated high risk of bias due to detection bias and the evidence was rated very low certainty.
03EEG and eye-tracking findings were inconsistent across studies, with some measures showing between-group differences at one time point but not another, and all rated high risk of bias except the Gap-Overlap task, which was rated low risk of bias.
STILL TO COME
How it was doneWhat they foundWhat it means for OTsWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only 12 unique RCTs contributed data, and most were conducted in the USA in university or hospital settings, so the findings may not generalise to community or low-resource contexts. Most participating families identified as White and mothers had post-secondary education; no studies included socioeconomically vulnerable families. Follow-up beyond 39 months of age was rare (only two studies reported outcomes after that point), so longer-term effects are unknown. All adaptive-skills studies were rated high risk of bias due to detection bias, meaning the one significant finding in that domain is the least trustworthy result in the review. Neurocognitive biomarker studies had high missing data rates, inconsistent protocols, and only one published a pre-specified analysis plan, making it impossible to pool or draw firm conclusions. The standardised assessments used (ADOS CSS, MSEL, VABS-2) may lack sensitivity to detect small changes in this age group, and the assessment process itself may be harder for autistic children than for typically developing children.
Declared interests
Funded by a University of Queensland Graduate School Scholarship, a Children's Hospital Foundation Top-Up Grant, and an NHMRC Research Fellowship. The funders had no role in design, analysis, writing, or submission. Four of the authors (McGlade, Whittingham, Barfoot, Boyd) are key investigators in the ENACT trial, a very early support program for infants at high likelihood of autism, and two of them (Whittingham, McGlade) developed the ENACT program. They are reviewing the field while developing a competing intervention.
The easy way to misread this
Do not read the small adaptive-skills gain (p = 0.009) as evidence that early interventions work for this population. All four contributing studies were rated high risk of bias due to detection bias, the GRADE certainty is very low, and the primary outcomes (autism symptoms, cognition, receptive and expressive language) all showed no significant difference.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →