The Changing Epidemiology of ADHD

From Medscape Pediatrics
An Expert Interview With Susanna Visser, MS

Laurie Scudder, DNP, NP

Editor's Note:

The results of the second administration of the 2007 National Survey of Children's Health (NSCH) have been reported in the Morbidity and Mortality Weekly Report.[1] The report documented an almost 22% increase in the percentage of children 4-17 years of age with a parent-reported diagnosis of attention-deficit/hyperactivity disorder (ADHD).
These findings illustrate the substantial impact of ADHD on both families and pediatric providers.
Laurie Scudder, DNP, NP, of Medscape sat down with Susanna Visser, MS, Lead Epidemiologist, National Center on Birth Defects and Developmental Disabilities, US Centers for Disease Control and Prevention, Atlanta, Georgia, to discuss the report.

Medscape: Ms. Visser, can you begin by summarizing the main findings of the 2007 NSCH study?

Ms. Visser: The primary goal of this particular report was to compare the most recent parent-reported survey results from a national survey conducted by the US Centers for Disease Control and Prevention (CDC) sponsored by the Maternal and Child Health Bureau at the Health Resources and Services Administration (HRSA) with an earlier survey.
The first survey was conducted in 2003-2004 and had 102,000 respondents. It was repeated in 2007-2008 with approximately 92,000 respondents. This large sample size allowed us to drill down into issues of child health in a way that we cannot in other surveys or through other mechanisms.

The specific question presented to parents was: "Has a doctor or health professional ever told you that your child has ADD [attention-deficit disorder] or ADHD?" Extrapolating from the survey responses to the total US population, it was determined that 4.4 million children in the United States between the ages of 4 and 17 years had received a diagnosis of ADHD as reported by their parents in 2003.
By 2007, that estimate had risen to 5.4 million children. That represents an increase of 1 million children with a parent-reported diagnosis of ADHD in approximately 4 years. That is a 22% increase over that time period and, obviously, was a number that was large and surprising. We knew that the rates were probably increasing because of our previous work, but this was a jump that really caught our attention.

Medscape: Can you describe the various stakeholders in this study and the reaction of these different groups?

Ms. Visser: At the CDC, our main goal through the use of these surveys is to keep a pulse on the health of the nation, and so what we have found here suggests that there is a growing need among American families as a result of this increase in children with ADHD.

A child with ADHD may have symptoms of inattention and hyperactivity or impulsivity, or both. Regardless of the symptoms, ADHD can cause significant functional impairment problems at home, at school, or with friends.
From a public health perspective, there are implications resulting from 5.4 million American children with ADHD. As most parents can appreciate, if you have a child with a behavioral disorder, it affects everything about that family's life, from the expectations of what a typical day is going to be like, through the expectations for school achievement, to relationships with family and friends. It can pose a number of very difficult challenges for those families. Healthcare providers are going to need to help the large number of families address the concerns that are brought into the clinical setting.

Medscape: What was the reaction of the healthcare community to this finding?

Ms. Visser: I think that clinicians -- psychologists and healthcare providers who are diagnosing, treating, and managing children with ADHD -- believed that, although this is a very large number, it was consistent with their clinical experience.

Although it is surprising to think that nearly 1 in 10 children have had a diagnosis of ADHD, it is consistent with the clinical experience. They are managing a lot of children with ADHD, with 2.7 million children taking medication in 2007.

Children typically don't just come in for a diagnosis. They are starting down a treatment and management path that we hope will include a behavioral component in addition to medical management, but we are unable to determine that from these data.

To answer your question, I think that healthcare providers believed that this was clinically consistent with what they have been seeing, and that the number, although quite large, does represent the burden of ADHD that they are seeing relative to other conditions in their practices.

Medscape: You also found some very interesting subgroup findings in different demographic groups. Could you share those findings with us?

Ms. Visser: These data confirmed previous population-based studies that revealed a 2:1 or even a 3:1 ratio of boys to girls in terms of diagnostic prevalence. That consistency really gives us a validity check. In this survey, we noted a 2.3:1 ratio of boys to girls for ever having been diagnosed with ADHD on the basis of parent report; that increases slightly to 2.6:1 for a current ADHD diagnosis. That has been very consistent in the literature and suggests that these data have some face validity. We also noted that the rates of ADHD increased with age, and that makes sense because parents were asked whether they had ever been told that their children had ADHD. Thus, those percentages should trend upward with age, and that was consistent.

We saw some interesting subgroup findings that were unique. The first was with respect to age. When we looked over time, although there were significant increases in ADHD prevalence in all of the 3 age groups that we looked at (4-10, 11-14, and 15-17 years), we saw that the 15- to 17-year-olds had a disproportionately greater increase when compared with younger children. This suggests that clinicians are probably seeing more older teens for diagnosis, treatment, and management of ADHD than in the past, and we do not have a clear understanding of why.

It may be that there is a decrease in stigma around ADHD, a greater acceptance of available treatment protocols, or increasing demands on students with greater expectations to go on to college. Children who were able to adapt to the difficulties presented by their ADHD symptoms at younger ages may have a more difficult time adapting at older ages, and the availability of more flexible treatment options may increase the likelihood that these children seek diagnosis and treatment. I think that that could be behind it. However, we do not know, and we are definitely looking to the clinical community to help elucidate that finding.

We also saw some interesting patterns across ethnic groups. Historically, in the United States, rates of ADHD have been lower in Latinos compared with non-Latino groups. Although these data confirmed that lower rate, we did see a large increase, about 53%, from 2003 to 2007 among Latinos. I think that is an important finding and suggests that the cultural differences in ADHD diagnosis or treatment may be lessening. That is an important finding to keep in mind.

Medscape: The other group who experienced a significant increase were children who were reported by parents to be multiracial.

Ms. Visser: That is correct. This is a finding that we have seen over time, and I am not sure what is behind this. There is limited literature about risk factors for ADHD and behavioral disorders in multiracial children. It is likely that the rates represent the coalescing of various social and economic risk factors within these families. However, it is not clear what is driving the findings among multiracial children.

Medscape: What about regional variance? I noted that the rate of parent-reported ADHD in North Carolina was 15.6% in this survey; however, in California it was only 6.2% -- which is obviously a dramatic difference.

Ms. Visser: That is correct. We saw a good bit of geographic variation in the rates of ADHD. Although there was a significant increase over time in the Northeast, Midwest, and South, there was not a comparable increase in the West. The regional differences were apparent at the state level, as you noted. We found that the state with the largest prevalence of parent-reported ADHD was North Carolina at 15.6%, representing nearly a 63% increase in ADHD prevalence from 2003 to 2007.

We found 3 other states with rates above 14%: Alabama at 14.3%; Louisiana at 14.2%; and Delaware at 14.1%. We are very interested in these states. We want to understand what is driving the rates up in specific states and reasons for the large variance across states. Some of our western states are quite low. In California, as you mentioned, only 6.2% of children are reported to have an ADHD diagnosis. It is important to us in terms of epidemiology to understand what is predicting both higher and lower rates of diagnosis within particular states.

We know that some of that can be accounted for by demographic factors. Risk for ADHD increases as income decreases -- a situation typically associated with lesser-resourced educational services, fewer support structures available to parents, and then more behavioral problems coupled with insufficient availability of appropriate resources and services. There are issues of access to care and limited resources to support a family in a way that will optimize the child's behavior. Those all coalesce together and relate to the environmental contributors to childhood ADHD.

Additionally, there is a genetic component to ADHD. In any population there will be a core group of children genetically predisposed to ADHD regardless of environmental factors.

It is important to understand that differences in state policies, practices, and things not related to the specific environmental and demographic factors within the state predict the diagnostic rate of ADHD, too.

We have been working with several states to try to understand the reasons for the state-level variation. We suspect that there may be very aggressive quality improvement practices and protocols in place in certain states with higher prevalence rates. When you screen more children for behavioral or developmental problems, you will find more symptoms. That greater rate of symptom identification will likely translate into greater rates of diagnosis. A state that has a higher rate is not necessarily in trouble. It may be that they are doing a fantastic job of assessing and screening the behavioral needs of their children, and that will translate to a higher rate of behavioral diagnoses. States that find themselves at the top of this list should not necessarily be concerned. It does underscore the importance of unpacking these findings to identify what is really driving these factors over time

Implications of the Changing Epidemiology of ADHD

Medscape: Do these data allow you to speculate about some of the reasons for the increased prevalence overall? You have noted that stigma may have lessened. Diagnostic awareness is certainly greater. Are there other reasons that can be teased out from these data that may be contributing to this increase?

Ms. Visser: Well, these cross-sectional data certainly cannot answer those questions for us, but we have a number of hypotheses. Greater awareness and better screening efforts are our top 2 hypotheses. There has been quite a lot of education. The American Academy of Pediatrics, for example, has really focused on quality improvement for pediatric practices, and the efforts around autism and ADHD have been focused on trying to standardize our approach to screening and diagnosis for behavioral issues.

There are probably other environmental and sociopolitical issues. As I mentioned, poverty is definitely associated with behavioral problems for a variety of reasons. The stresses of that environment on young people can really affect behavior and, when added to a lack of appropriate support for parents, can have a long-term effect on the child and family dynamic.

All of those factors likely play into it. We are focused on getting answers to a number of these questions. What we have recognized is that ADHD is not the exception. Although the prevalence of ADHD has been increasing since 1996, this recent jump has raised some concern. Reflecting that same trend, we have seen the rates of autism increase as well. We think that ADHD is a piece of the entire puzzle. American families are struggling with behavioral problems and developmental disabilities, and we need to be very aggressive about how we investigate these patterns.

Medscape: The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V) is proposing diagnostic criteria for ADHD that would increase the age of onset of symptoms from the current requirement that impairment be present at or before 7 years of age to an older onset of at or before 12 years of age.[2] Do these data support that proposed change?

Ms. Visser: I think that it is going to reflect what clinicians have already been doing, which is not weighting that criteria as heavily as other criteria. The new DSM-V criteria, with respect to age of onset, will mirror more closely clinical intuition, which is to attempt to document symptoms early on in childhood but not to be confined to before 7 years of age. However, I also think that we are going to need to understand and reflect on whether the DSM-V criteria are appropriate for teens and tweens. We may have to make some adaptations to those symptom criteria to make sure that we are able to capture the manner in which these symptoms are expressed in older children, particularly if this pattern of a rising prevalence during adolescence continues.

Medscape: What are some of the limitations to this study? Can you comment on whether parent report is a valid methodology for collection of these data?

Ms. Visser: I think the major limitation of these data is the lack of clinically validated diagnostic data. We asked parents whether a healthcare professional had told them that the child has ever had ADHD. We did not go back to medical records to verify that the diagnosis was made or how rigorously diagnostic criteria were applied. Therefore, we recognize that this estimate of prevalence reflects community diagnosis, which may include less rigorous diagnostic practices for ADHD diagnosis. However, what we do know is that parent-reported data are very reliable. They trend in a very smooth fashion over time. We have been using this question in our national surveys, this and others, since 1996, and the rates trend very smoothly. There is not a lot of jumping around, and that reliability is one measure of the psychometric value of the question.

When you look at the prevalence estimates for parent-reported ADHD, they track in magnitude very similarly to the parent report of other conditions. For example, Tourette syndrome comes up at about 2 per 1000 for current diagnosis.[3] Autism comes up at 11 per 1000, and ADHD comes up at 72 per 1000 for current ADHD diagnosis.[4] That really trends in a predictable way if you ask most clinicians. You want to see those rates stacking up in that fashion and having about that distance between them in magnitude, so there is some face validity to these estimates.

We are never going to be able to say whether the rates include overdiagnosed or underdiagnosed children with these cross-sectional data. There is probably some of each in this estimate, but there is no reason to think that there would be more or less of either of those groups. This is a cross-sectional estimate of the lived experience of ADHD in American families. If parents are telling us that their children either have had a diagnosis or are currently living with ADHD, then we believe that they are experiencing the impact of that disorder in their families, and that is an important public health measure of ADHD impact.

Medscape: Can you offer some insights into where you think future research should be focused on the basis of these data?

Ms. Visser: We really need to understand the driving forces behind this increase and try to determine whether rates are increasing due to quality improvement efforts in some states. We can look at population-based studies that are in the field now and match them to the state-based estimates in this survey, see whether they track well, and look for factors that are available at the community level to better determine who is being diagnosed and why.

That is going to be important not just for ADHD, but to inform the behavioral patterns that are seen in other developmental disorders that we are tracking. It is a priority for us. Of course, this is just one way to monitor the signal for ADHD in communities. Other studies, including both clinic-based and community-based epidemiologic studies, are also important because they allow for contribution of input from both medical professionals and teachers. These 3 pieces together allow us to get a better national representation, in a cost-effective manner, of the full picture of ADHD impact in America.

Medscape: I think the most important question for our readers is the implications of these data for primary care pediatric providers who are seeing these children on a day-to-day basis. Can you provide some insights?

Ms. Visser: Well, definitely the finding of increased rates in teens is important. Our data suggest that primary care providers may be managing more older teens with ADHD than in the past. Of course, teens should be more active participants in the treatment and management of their symptoms than younger children might be.

The other issue is that we noted an increase in rates of medication among girls as they got older. In boys, rates of medication use increase throughout childhood, peaking at 11- to 14-year-olds and then declining after that. The answer to the question of why we are seeing this variance in medication use is likely to come from the clinical community.

Finally, the findings of a significant increase in ADHD among Latino children is important. What we know from pharmacoepidemiologic studies is that the Latino community is less interested in medical treatments, and medication in particular, for behavioral and mental health conditions. Therefore, this may pose a different sort of challenge for both primary care providers and specialists who are treating Latino children with ADHD. The influence of race and ethnicity on the treatment plans for children will need to be better understood and considered.

Editor's Note: We would like to hear from you. Is this NSCH report consistent with your own clinical experience? How has the changing epidemiology of ADHD affected your practice? Please join the discussion.