The recent report from the Centers for Disease Control (CDC) on increasing autism prevalence has created concern among parents, headlines in the media and enough questions to warrant additional exploration.
The findings from 2010 data show that autism now affects 1 in 68 children, up 30 percent from the last analysis, released two years ago, which reported that 1 in 88 children was affected. Should we panic? No.
Let’s take a closer look at where these numbers came from and what they really mean.
Developmental Disabilities Among Children Are Relatively Common
In this context, the word “prevalence” denotes the percentage of a population that is affected with a particular disease at a given time.
The new data show that the prevalence of autism in 8-year-olds was 1.47 children out of a hundred (1.47 percent)—based on an average from 11 sites in the U.S. To put these numbers in perspective, it’s helpful to compare them to prevalence rates of developmental disabilities overall.
In 2010, the CDC published a study on developmental disabilities showing the following prevalence rates among children ages 3 to 17 from 1997 to 2008:
- Learning disabilities (LD): 7.66 percent
- Attention deficit hyperactivity disorder (ADHD): 6.69 percent
- Other developmental delays (intellectual disabilities, borderline intelligence): 3.65 percent
- Autism spectrum disorders (ASD): 0.47 percent
It’s important to recognize that during this 11-year period, the prevalence of any developmental disabilities in children was 13.87 percent—one in eight children. Developmental disabilities, therefore, are more common than most people recognize.
As shown above, autism was the least prevalent of the “big four” conditions analyzed by the CDC in 2010 and would remain so even with the current increased autism prevalence rate of 1.47 percent.
Collection of Study Data
So how was this 1.47 percent rate calculated? Understanding the methodology of the study is critical.
Researchers looked at a population of 363,749 8-year-olds (9 percent of the U.S. population of this age group) from 11 geographical sites around the country.
They then reviewed records of 47,371 children known to various agencies whose data suggested autism. These data included billing codes, visits to specialty developmental disabilities clinics, and in some states, educational records of children in special education. Trained reviewers looked at the records and used the guidelines in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), to determine if these children met ASD criteria. The children were not actually seen by the reviewers; this was a paper review.
It is noteworthy that in the latest CDC report the prevalence at each of the study sites varied widely. New Jersey had the highest prevalence, at 2.19 percent, while Alabama had the lowest, at 0.57 percent. In the 4 states with the lowest prevalence rates, there was either no access or limited access to educational records, likely contributing to an underestimate in prevalence rates in those states.
Diagnosing Autism: A Judgment Call
As a developmental pediatrician who specializes in autism, I can say with certainty that autism cannot reliably be diagnosed by conducting a paper review. Other than for the group of children with significant autistic disorder, an autism diagnosis often involves a judgment call. Even in the hands of experienced diagnosticians, who have spent hours with a child and have reviewed the available records, there can be a lack of consensus regarding the diagnosis. One clinician may diagnose a child with ADHD, while another may diagnose the same child with pervasive developmental disorder not otherwise specified (PDD.nos, or mild autism). There can be a change in the child’s presentation over time; a child who had many autistic features at age 2 can, with intervention, be better described as having ADHD or LD at age 8.
Until a biomarker (such as a blood test or MRI finding) is found that is diagnostic for ASD, the diagnosis will remain a judgment call based on behavioral findings. It does not help matters that the DSM-IV PDD.nos definition is extremely broad: “This category should be used when there is a severe and pervasive impairment in the development of reciprocal social interaction OR verbal/nonverbal communication OR when stereotyped behavior, interests and activities are present, but the criteria are not met for Autism.”
The looseness of this definition is one of the main rationales for the changes in the diagnostic criteria for ASD reflected in the fifth edition of the DSM (known as DSM-5). It’s important to note that this just-released CDC analysis is based on old DSM-IV criteria by chart review, not on interviews with clinicians who saw the children or examined them directly.
Sometimes records come across my desk for a child I am scheduled to see and the child sounds extremely autistic—and then the child walks into my room . . . with ADHD. Sometimes quite the opposite occurs. Seeing the child is critical.
The Cognitive Dynamic
The current analysis shows the greatest increase in autism prevalence among those children with normal cognition. Forty-six percent, or almost half of this sample, is reported to have average or above-average cognition.
These children are most likely of the Asperger’s phenotype, a group whose diagnosis is at risk based on the new DSM-5 criteria. The number of children with frank autism, who are able to be diagnosed without a judgment call, has not increased dramatically—except among populations whose members have encountered obstacles to diagnosis.
Autism Diagnosis Obstacles Among Hispanics and Blacks
The largest increases in prevalence were seen not only among those with normal cognition, as discussed above, but in two other groups: Hispanic children and black children. We have no reason to think that white children are more likely to be affected with autism than other racial or ethnic groups, and yet they are 30 percent more likely to be diagnosed, according to the study. Researchers and clinicians have offered several explanations for why Hispanic and black children with autism are not identified so readily as white children. It appears that these groups encounter more obstacles to diagnosis.
Those obstacles vary. They can be geographic, such as living further from diagnostic centers, or economic: lacking insurance to cover such evaluations. They can be cultural, with reluctance for diagnosis on the part of families for any number of reasons (but in contrast to other groups whose members may seek out a diagnosis). Some data suggest that children from these minority groups are more likely to be diagnosed with something else (such as intellectual disabilities) than with autism, even when criteria for autism are met.
The methodology of the current CDC study attempted to overcome these obstacles by looking not only at children who were diagnosed but also at those who should have been diagnosed based on their records.
I also think that as this past decade has unfolded, autism has become a household word, and familiarity with the condition serves to diminish certain cultural obstacles on the part of families. This combination of effects—case ascertainment and cultural milieu—may be contributing to the “catch-up” diagnoses being made over the past decade for children from underrepresented minorities. The current study states, “The largest increases (in autism prevalence) from 2002 to 2008 were noted among Hispanic children, non-Hispanic black children and children without co-occurring intellectual disability.” In my view, the increase in prevalence is a positive sign that we are overcoming healthcare disparities.
Increasing Autism Prevalence: The Take-Home Message
With the limitations of the study now put in perspective, it’s important to establish the meaningful conclusions to be drawn from this analysis.
ASD, or significant social-communicative impairment, is a common problem, but not so common as ADHD, LD or cognitive impairment. It affects individuals at every level of intelligence and from every racial and ethnic group. While its signs are often evident by age 2, the average age of initial diagnosis remains much later: 4 years and 5 months. We need to do better at addressing obstacles to diagnosis.
Acknowledging that individuals with autism make up between 1 and 2 percent of the population can help us as a nation better address the various needs of individuals with autism as these children grow to adulthood: educational, therapeutic, employment and those associated with living arrangements. We need to continue to explore how best to support families and incorporate individuals with autism into our communities.
These latest data on prevalence highlight the importance of supporting research aimed at determining the causes of autism and finding more effective treatments.
Autism is here to stay. It has always been here, but as clinicians and as a society we’ve only just begun to look for it systematically and proactively.