A School-Based Study of Autistic Symptoms in 3-8 Year Olds in India from Parent and Teacher Report

Background: Despite housing ~18% of the world’s population, India does not yet have an estimate of prevalence of autism and autistic traits in the general population. One reason for this lacuna has been the lack of availability of translated and validated screening and diagnostic tools for autism, which we have addressed in our previous work (Chakrabarti et al., IMFAR 2011). Estimates drawn from studies in UK and USA suggest that India could have more than 2 million people with ASC, however this prevalence estimate has never been directly verified. There is hence a critical need to measure the prevalence of autism and the distribution of autistic traits in a general population sample in India.

Objectives: To measure autistic symptoms in a school-going sample (ages 3-8) drawn from the general population in Kolkata.

Methods: Schools from all socio-economic sectors (government, private and special schools) were selected from 3 boroughs in Kolkata. Teachers filled in the Social Communication Disorder Checklist (SCDC), a 10-item initial screening measure for autism (Skuse et al., 1995), in English or Hindi or Bengali. Parents/caregivers also filled in the SCDC, followed by the Social Communication Questionnaire (SCQ) – a 40-item diagnostic tool (Rutter et al., 1999) – and the Autism Quotient – child version (AQ-C), a 50-item tool quantifying the distribution of autistic traits (Auyeung et al., 2009).

Results:

SCDC: Teacher reports were obtained for 12764 children. 3.6% met the cut-off score of 9. Splitting the data by language shows a similar picture, with, 3.8% meeting cutoff on English, 5.49% on Hindi and 1.29% on Bengali versions of the tool. Parent response rate for SCDC was 50.1%. Interestingly, a much greater percentage met cut-off on SCDC parent report with 20.19% meeting cut off overall. Specifically, 21.67% met cutoff on English, 10.24% on Hindi, and  20.11 % on Bengali versions of the tool. Positive correlation between parent and teacher ratings were modest, but highly significant (r=0.103, p<0.001).

SCQ: Out of 2887 respondents 8.9% met the cut-off score of 15, overall. Specifically 9.64% met cutoff  on English, 8.26% on Hindi, and 10.09% on Bengali versions of the tool.

AQ-C: Out of 2901 respondents 9.86% met the cut-off score of 76. Specifically, 9.48% met cutoff on English, 11.25% on Hindi and 8.67% on Bengali versions of the tool.

Conclusions: This study provides preliminary data on the distribution of autistic symptoms in 12764 children in India, and is the first and largest study of its kind in India so far. The estimates reported suggest considerable over-reporting by parents in comparison to teachers, irrespective of language. This provides important directions for future epidemiological research into autism in India.