CDC Reports a higher than expected prevalence of autism in Brick Township
Over the past several years, prevalence studies and genetic research have established that autism is one of the most "strongly genetic" of neuropsychiatric disorders. Most estimates of heritability are around 80%, which means that other (nongenetic) factors account for some cases of autism. These other factors include intrauterine viral infections, perinatal asphyxia, and possible exposure to environmental toxins. Concern about environmental toxins arises when there is an apparent increase in incidence or a geographic cluster of a disorder. A possible cluster was reported by the Associated Press on January 19, 1999 after some 40 cases of autism were diagnosed in Brick Township, an eastern New Jersey community with an overall population estimated at that time to be 71,000 residents. This suggested a prevalence three times greater than the "usual" rate, which prompted an intensive investigation by the Centers for Disease Control (CDC).
On April 18, 2000, the CDC announced that the prevalence of autistic disorder in Brick Township is 4.0 per 1000 children, ages 3 to 10 years and, after including PDD-NOS and Asperger's syndrome, 6.7 per 1000 children, ages 3 to 10 years living in the community during 1998. This estimate exceeds previous estimates of prevalence by 3 to 4 fold. The CDC investigators diagnosed autistic disorder in 36 children in a community with a population of 77,000.
Whether or not this constitutes a true cluster remains to be determined. A cluster is defined as an unexpectedly high prevalence exceeding a well-defined background rate. Other epidemiologic studies done over 10 years ago outside the U.S., using DSM-III-R criteria yielded age-specific prevalence estimates of 1 per 1000 for autistic disorder. In light of these older studies, the Brick Township data appear to constitute a true cluster unless further studies show that the prevalence, using DSM-IV criteria, are generally higher in the U.S. then previously believed. The CDC report references a recent unpublished study in England that also yields a higher prevalence than older estimates. On the other hand, if Brick Township represented a true cluster, one might expect a higher sibling recurrence rate than the one actually found, 4.9%, which is similar to previous estimates.
One surprising finding is the reversal of prevalence for autistic disorder versus other autistic spectrum disorders (ASDs, i.e., PDD-NOS and Asperger's syndrome): the ratio of autistic disorder (N=36) to other ASDs (N=24) was 1.5. Previous studies have generally shown that there are more children with PDD-NOS and Asperger's syndrome relative to autistic disorder. One explanation is that it is possible that higher functioning children with PDD-NOS and Asperger's syndrome did not manifest behaviors that rose to the level of concern and detection. In part this unexpected finding may be an artifact of the cutoff at age 10 years. Children with Asperger's syndrome are typically diagnosed at an older age. Another explanation offered by the CDC is that the diagnostic tool used for ascertainment, the Autism Diagnostic Observation Schedule (Generic), did not sufficiently discriminate between autistic disorder and other ASDs.
Another unexpected finding was the higher prevalence of ASDs (7.8 per 1000) in 3 to 5 year old children than in 6 to 10 year old children (6.1 per 1000). This reversal is puzzling since, as children get older, one would expect entry of new cases into the pool as social deficits in higher functioning children with PDD-NOS or Asperger's syndrome become increasingly apparent.
No definitive conclusion could be reached in terms of a specific environmental toxin. Investigators considered the possibility that seepage from waste dump sites may have contaminated drinking water. Robert Knowles, an environmental scientist with the Agency for Toxic Substances and Disease Registry (ATSDR), stated: "Based on the available data, current science, and what we found here in Brick Township, we found no association with chemicals in the environment and autism" after examining the municipal drinking water supply, the Metedeconk River, and the local landfill (quoted in the Philadelphia Inquirer, April 19, 2000).
Although a specific environmental toxic agent could not be identified, this type of epidemiologic investigation underscores the complexity of the potential impact of environmental toxins on health. The "environment" extends beyond drinking water and includes chemicals that are ingested, inhaled, or absorbed through the skin, apart from exposure to ionizing radiation, which is ubiquitous.
Sidebar: Estimates of the average annual radiation exposure in the U.S. vary between 120 and 360 mrem/year. Much of this exposure is from natural sources of background radiation of which one large component is inhaled radon. Radon is a gas that seeps out of the ground into homes built over rock such as granite or shale. Average medical and dental X-ray exposure is estimated to be between 39 to 50 mrem/year. A "safe" threshold of radiation dosage below which genetic damage does not occur has not been established with certainty and probably does not exist.
It is possible that the current epoch of industrialized societies has brought about a parallel epoch of increased genetic risk as the opportunities for exposure to a "chemicalized" environment have proliferated. The potential for exposure to mutagens may be high during the course of our daily lives since it has been estimated that there are about 50,000 chemicals produced by industry that have not been screened for adverse (including genetic) health effects.
Moreover, in the broader view, exposure to environmental factors can further contribute to the strong genetic influence affecting autism incidence. Since the prevalence of autism in the population cannot be sustained by affected individuals who do not reproduce, it is highly likely that new mutations occur. If an environmental toxin is identified at some future date, its effect may have been expressed through a germline cell mutation (affecting one or more genes in either the sperm or egg) before conception. Although speculative in relation to autism, a possible source is exposure of germline cells (sperm or ova) to environmental mutagenic agents, including X-rays and other forms of ionizing radiation, volatile organic solvents, insecticide residues on ingested foods, and certain medications or chemicals capable of damaging DNA.
Certain environmental toxins may also have direct effects on the embryo during the earliest stages of development. The birth defects (including autism in about 5%) caused by thalidomide are an example. Substances causing such anomalies are referred to as teratogens. The "window" of susceptibility for such damage appears to be during the earliest stages of the first trimester of pregnancy.
In addition to examining the water supply, other questions are appropriate:
- Were the fathers of the Brick Township children employed in occupations that caused exposure to fumes of organic solvents (such as paint solvents or work with epoxy or polyester resins (for example, auto body shops))?
- Did the parents have illnesses at anytime earlier in their lives that necessitated exposure to diagnostic X-irradiation?
- Were either of the parents active participants in sports with concomitant injuries entailing X-rays in which the radiation field would likely have involved the gonads?
- Was there occupational exposure to X-rays (health care workers) or other radiation (nuclear power industry)?
- Was there a history of substance or alcohol abuse by either parent?
- Did the parents ever live in an area of known high background radiation, such as radon (although radon is primarily a risk factor for lung cancer)?
Questions such as these are not readily answered since such historical data may not be recalled accurately. Another difficulty is that germline cell mutations are often "silent" and may not become apparent until several generations have elapsed. It is possible that the impression of increasing prevalence may reflect the cumulative effect of mutations occurring over the course of the preceding several generations, as exposure to DNA-damaging events has increased in our industrialized society during the past century.
A search for a potential mutagen or teratogen, in the face of myriad possibilities, is akin to looking for a needle in the haystack and may well need to extend beyond drinking water. According to the Philadelphia Inquirer (April 19, 2000), the report by the ATSDR stated that "while chemicals may have been found, neither children nor pregnant women were exposed to levels which would be likely to cause adverse health effects." This is not entirely reassuring, however, since all it takes is "one hit" at a critical "time and place" in cell metabolism, especially during DNA replication preceding cell division. Perhaps as little as a few molecules of a chemical or a few photons of ionizing radiation may result in damage to DNA in germline cells before conception. A similar vulnerability to small amounts of teratogens may also exist during the earliest stages of the developing nervous system in the embryo after conception.
The CDC study only considered the possibility of water-borne toxicants affecting parents living in Brick Township in 1998. However, in light of the high genetic component contributing to autism etiology, the "window" of exposure to mutagens, at least in parents (excluding previous generations) preceded the year of study, 1998, by at least 2 to 4 decades. Since the oldest children studied were age 10 years and were born in 1988, the "window" of germline mutation could extend as far back as 1948 (the year of birth of a parent who was 40 years old at the time of conception of their child) or 1968 (if a parent was 20 years old at the time of conception of their child in 1988). If one takes into account that the female germline cell number is "settled" at about 2 to 3 million during fetal life than the "window" of risk is relatively brief in females if one pre-supposes that the risk attributed to mutagens is highest during active cell division (see reference below). Beginning with the onset of menses at puberty, less than 400 eggs will mature over the next 30 years of potential child-bearing years. Thus the risk for germline cell mutation is relatively low in females compared to males.
For males, there is a similar narrow "window" of risk of mutation as the sperm progenitor cells are produced during embryonic development. However, at the onset of puberty in males, and for many years thereafter, active rates of cell division and DNA replication occur. (It has been estimated that a male produces 1 trillion sperm during his lifetime, thereby having many more opportunities for mutation (see reference below).) After the onset of puberty, at any time during the ensuing 10 to 30 years of reproductive potential, exposure to mutagens may have occurred. Since multiple genes are believed to be involved in autism (up to 20, according to Neil Risch and colleagues at Stanford), multiple "hits" would have been necessary, perhaps affecting both the male and female germline. Moreover, each of the parents themselves, rather than experiencing de novo mutations, may have inherited silent mutations extending back in time for generations. The final expression of a phenotype would represent the composite of both inherited and de novo mutations. In the final analysis, environmental toxins (including ionizing radiation) probably contribute to the high estimate of heritability for autism and to the apparent trend of increasing prevalence (if the latter is confirmed by further studies).
The task of investigating all of the possible environmental toxins is a formidable challenge for epidemiologic sleuths. The potential payoff in understanding birth defects in general, and mental impairments in particular, will be incalculable.
Media coverage of the CDC announcement included articles in The New York Times (April 19, 2000, page B4, by Maria Newman) and the Philadelphia Inquirer (April 19, 2000, page B1, by Shankar Vedantam).
Also see the press release by the National Alliance for Autism Research and the comments by Dr. Eric London
Full text of CDC report on Brick Township
Reference: Hawley RS Mori CA, The human genome. A user's guide. Academic Press, 1999
Return to list of commentaries and other autism links