Autism: most early interventions show limited effect

  • Sandbank M & al.
  • Psychol Bull
  • 1 ene. 2020

  • de Emily Willingham, PhD
  • Clinical Essentials
El acceso al contenido completo es sólo para profesionales sanitarios registrados. El acceso al contenido completo es sólo para profesionales sanitarios registrados.


  • Among 7 types of early intervention for autism, none had significant effect after being controlled for study design and bias in this meta-analysis and systematic review.
  • Authors evaluated 15 outcome categories for 6240 participants from 130 independent samples.

Why this matters

  • Early intervention is urged for children diagnosed with autism on the premise that applying these strategies early will offer greater benefit.
  • Authors say that despite results under most stringent analysis, naturalistic developmental behavioral intervention (NDBI) studies offer “a range of positive outcomes” for these children.
  • They also highlight developmental interventions as promising.

Key results

  • These authors evaluated 1615 effect sizes.
  • When all studies were included, 3 types of intervention showed effects: NDBI (26 studies), behavioral (27 studies), and developmental (14 studies).
  • No effects seen for sensory-based (7 studies), technology-based (10 studies), or animal-based (4 studies) interventions or TEACCH (6 studies).
  • With inclusion only of randomized controlled trials (RCTs), only NDBI and developmental interventions showed effects.
  • With inclusion only of RCTs and outcomes with no risk for detection bias, no interventions showed effects.

Study design

  • 150 reports included.
  • Average participant age, 54.21 months; 84% of each sample on average were male.
  • Funding: None disclosed.


  • The limitations are those of the included studies.