Letters

Prevalence of autism in early 1970s may have been underestimated

EDITORRecently, concerns about an apparent increase in the prevalence of autism in the general population, and the pathophysiology behind this, have been prominent in the press. Before these concerns can be addressed we need to be sure that the prevalence has truly increased. A cohort study was thought suitable to provide an insight into whether the prevalence of autistic disorders has increased.

In the British cohort 1970 study (BCS70) only five children were identified as having autism (and one as having suspected autism) at the age of 5 in disability data files. This gives a prevalence of 6/13 135 (0.45/1000)comparable with that found in previous studies. ^{1 2}

A focus group consisting of practising consultants (from adult and paediatric disciplines) experienced in the diagnosis of autistic disorders was convened. This group identified several diagnostic features from the available data that they thought were important in making a diagnosis of autistic spectrum disorder. The aim was to identify whether there were missing or undiagnosed cases with current commonly used diagnostic features. The features identified from the original BCS70 questionnaire included:

• At age 5
  Restless
  Solitary; does things on own
  Fearful, afraid of new situations
  Fussy or over particular
• At age 10
  Does things on own; rather solitary
  Afraid of new situations
  Fussy or over particular
  Hums or makes odd noises
  Obsessional
  Requests must be met immediately

When an analogue scale was used, only the most extreme casesthat is, the top fifthwere identified. Cases in which the child had all features present at both age 5 and age 10 were then identified.

Using the methodology above, we identified 56 cases from 14 904 children studied at age 10, giving a prevalence of 3.76/1000. This suggests that these children have an autistic spectrum disorder when contemporary diagnostic features are used.

Our finding agrees with current lifetime prevalence figures suggested by Powell et al.³ Thus estimates of prevalence from the early 1970s may have seriously underestimated the prevalence at that time. Confirmation of this suggestion would require contemporary assessment of the individuals involved.

Helen Heussler, lecturer in community child health. 
honey.heussler{at}nottingham.ac.uk

Leon Polnay, professor of community paediatrics. 
Division of Child Health, School of Human Development, University of Nottingham, Nottingham NG7 2UH

Elizabeth Marder, consultant community paediatrician. 
Department of Paediatrics, Queens Medical Centre, Nottingham NG7 32UH

Penny Standen, reader in health psychology and learning disability. 
Division of Rehabilitation and Ageing, School of Community Health Sciences, University of Nottingham

Chin Lyn U, formerly data manager BCS70 Nottingham. 
Division of Psychiatry, School of Community Health Sciences, University of Nottingham, Nottingham NG3 6AA

Neville Butler, senior research fellow BCS70. 
80 Cumberland Road, Bristol BS1 6UG