Prevalence of autism in early 1970s may have been underestimated
BMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7313.633 (Published 15 September 2001) Cite this as: BMJ 2001;323:633
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
The only thing more perplexing than the contents of this letter was
that the BMJ chose to print it. In an environment of real and genuine
concern over the rising incidence of autism, you owe your readers better
information than anecdotal musings from a group of interested parties.
Would you have printed a letter from a "focus group" of parents who
believed MMR caused their child's autism? There is no substantive
difference between these two forms of "research."
This potential abuse of such methods would normally not be of great
concern, but is critically important today because of a controversial and
unsupported current hypothesis(1). Fombonne and others have argued that
despite an apparent recent increases in the rate of autism prevalence in
many countries, the actual incidence of this devastating disorder has
remained constant. The impression of an increase, this hypothesis goes,
has been created only because of greater rates of detection in recent
years. There is therefore, proponents of this theory argue, no reason to
be concerned over possible environmental causes of autism that would
necessarily underlie an abrupt increase in incidence rates.
Despite the prominence of its proponents, there is no hard evidence
to support this hypothesis. In fact, the only studies of any depth that
have explored the question have overwhelmingly failed to find the
hypothesized "hidden horde" of autistic children from the 1970s, or for
that matter, from any other period. Burd et al (2) conducted a prevalence
study of autism in North Dakota. They found a low prevalence rate of 3.26
per 10,000 among a cohort of children born between 1967 and 1983. A twelve
year follow up survey of the same cohort (3) showed that the original
prevalence study had found 98% of the autistic children in the study area.
One child with autism had been missed.
In a similar study (4), Nylander and Gillberg screened adult
psychiatric outpatients for evidence of undiagnosed autistic spectrum
disorders. Unlike the North Dakota studies, this population had never gone
through an earlier screening for autism (which presumably would have
encouraged proper rates of initial diagnosis). The authors, based on the
disease model advanced in (1) hypothesized that they would find high rates
of undiagnosed adults with autism.
The screening procedure did in fact unearth 19 autistic spectrum
adults who had not received a prior diagnosis. However, the prevalence
rate for this group was only 2.7 per 10,000, a finding that provides
little support for the hidden horde hypothesis. The authors note this
point reluctantly, claiming that the observed prevalence rates "should be
regarded as an absolute minimum." They provide no guidance, though, for
further research and offer no other candidate populations in which to
search for the disabled victims of what has been widely considered a life-
long disorder.
For some reason, many scientists and health professionals find the
data regarding sharp increases in autism rates to be an uncomfortable
finding. A few have developed extravagant and unsupported theories as an
expression of their discomfort. These theories deserve sharp critical
examination because the simple interpretation of the existing data
supports the conclusion that autism has increased quite strongly. Indeed,
in light of the consequences of complacency, it would seem the burden of
proof should lie with those who seek to dismiss decades of epidemiological
research as flawed.
Yet we are now offered "focus groups" as a new research tool. In the
business world, we call this technique "mother-in-law" research, so I must
confess surprise that the BMJ would dignify such efforts. Good science
demands that we face the real data, no matter how inconvenient the
implications may be.
Sincerely,
Mark F. Blaxill
Cambridge MA
References
1. Pediatrics 2001 Feb;107(2):411-2. Is there an epidemic of autism?
Fombonne E.
2. J Am Acad Child Adolesc Psychiatry 1987 Sep;26(5):700-3. A prevalence
study of pervasive developmental disorders in North Dakota. Burd L, Fisher
W, Kerbeshian J.
3. Int J Circumpolar Health 2000 Jan;59(1):74-86. A prevalence methodology
for mental illness and developmental disorders in rural and frontier
settings. Burd L, Kerbeshian J, Klug MG, McCulloch K.
4. Acta Psychiatr Scand 2001 Jun;103(6):428-34. Screening for autism
spectrum disorders in adult psychiatric out-patients: a preliminary
report. Nylander L, Gillberg C.
Competing interests: No competing interests
Prevalence of autism in the past may have been underestimated
Autistic Spectrum Disorders (ASDs) cover a range of disorders
characterized by impairments in social interactions, pattern of
communication and restricted, repetitive behaviours. We have previously
demonstrated a 4-fold rise in the incidence of diagnosed ASDs in pre-
school children from 1991&1992 to 1995&1996 in the West
Midlands(1). Heussler et al(2)have reported reviewed data from the 1970s
suggesting that by contempory standards the prevelance of ASDs in the
1970s was probably as high as it now appears.
We undertook a time trend analysis of the West Midlands data set to
see if the observed increase in incidence was compatible with a new
biological aetiology or with improved diagnosis of essentially the same
condition.
Clinical data on the 148 children ascertained in the original study
(1) was collected from the medical notes in a systematic way using a
proforma. Complete data was available for analysis on 134. Data in the
following domains were recorded: family history of ASD, developmental,
neurological, other medical and psychiatric disorder; obstetric and past
medical history; personal history of epilepsy and other neurological and
medical disorder, learning difficulties and other developmental problems;
abnormalities on physical examination including visual and hearing
deficits. Linear by linear association tests were used to test for time-
trends in the data.
There were no significant trends found between any of the
characteristics examined.
To conclude there was no evidence for a change in the prevalence of
associated clinical features in children with ASD over time. We failed to
find an indication that the apparent rise in incidence was due to the
emergence of a single new cause of autism. This result is consistent,
however, with improved identification and diagnosis of ASDs in the
population over time.
William Whitehouse, Senior Lecturer in Paediatric Neurology,
University of Nottingham
Mark Abbas, Medical Student, University of Birmingham
Judy Powell, Lecturer in Public Health & Epidemiology, University
of Birmingham
1. Powell et al. Changes in the incidence of childhood autism and
other autistic spectrum disorders in preschool children from two areas of
the West Midlands, UK. Developmental Medicine & Child Neurology
2000:42:624-628.
2. Heussler et al. Prevalence of autism in early 1970s may have been
underestimated. British Medical Journal 2001:323:633.
Competing interests: No competing interests