Data from this report comes from the National Ambulatory Medical Care Survey (NAMCS), an annual survey of a representative sample of U.S. office-based physicians.  According to these data, the proportion of times during which a second medication was prescribed in addition to a stimulant medication (e.g. Ritalin, Adderall, Concerta) for treating ADHD patients under age 18 increased from about 5% during 1993-94 to almost 25% during 1997-98.  This represents a five-fold increase.  The most commonly prescribed medications in combination with a stimulant were Clonidine and various antidepressants.

What is concerning about this trend is that there is virtually no safety or efficacy data to support combined pharmacotherapy for ADHD youth.  Thus, the rapid increase in this practice is occurring in the absence of any research-based support.  Although most experts regard the safety and efficacy of stimulant medications as well established, there is little evidence that adding other agents to stimulant medication treatment results in significantly better outcomes for youth with ADHD.  And, we know little about the long-term safety of combined medication treatment.

It is worth noting that in the MTA study -- the largest treatment study of ADHD ever conducted, there were very few children treated with stimulant medication where an additional medication was added to their treatment.  This was true even though many of the children in the study had co-morbid conditions such as depression or anxiety disorders for which other agents might have been used and suggests that when stimulant medication treatment is carefully conducted, the need for combining stimulants with other medications is likely to be far less frequent than the increasing use of this practice would suggest.

The authors of this report close by emphasizing, "definitive safety and efficacy data are needed to support common forms of combined pharmacotherapy among youths."  In the meantime, it is important for parents and physicians to be aware that these data are currently lacking, and that the increasingly frequent practice of combining stimulants with other medications may be necessary less often than it is recommended.

In the current psychiatric diagnostic system, three distinct subtypes of ADHD are specified.  ADHD Combined Type is used when both inattentive and hyperactive-impulsive symptoms are present.  When attention problems are prominent -- and hyperactive-impulsive symptoms are not -- the diagnosis that applies is ADHD, Predominantly Inattentive Type.  This is the "official" term for what people often refer to as ADD -- in contrast to ADHD.  In situations where the reverse is true, the term is ADHD, Predominantly Hyperactive-Impulsive subtype.  In this latter case, the term predominantly is included to reflect the fact that the individual's difficulties are predominantly in that area, although there can certainly be some difficulties with symptoms in the other area. (For information on diagnostic criteria click here.)

In order for this current method of dividing ADHD into different subtypes to be valid, there needs to be clear evidence of functional differences between individuals who have been diagnosed with the different subtypes.  Several studies have found evidence of such differences among children.  For example, children with the combined and hyperactive-impulsive subtypes are more likely than those with the inattentive subtype to have severe co-existing behavior disorders such as Oppositional Defiant Disorder (ODD) and Conduct Disorder (CD).  In contrast, those with the inattentive subtype appear more likely to have learning difficulties and problems with academic performance.

One limitation of prior research on subtype differences is that it has been conducted almost exclusively with children.  Thus, relatively little is known about subtype differences among older individuals.  Beginning to examine such differences in a young adult population was the focus of a study published recently in the Journal of Nervous and Mental Disease (Murphy, K. et al (2002) Young adults with ADHD: Subtype differences in comorbidity, educational, and clinical history, 190, 147-157.)

Participants in this study included 60 individuals diagnosed with ADHD, Combined Type (ADHD-C), 36 diagnosed with ADHD, Predominantly Inattentive Type (ADHD-I), and 64 comparison subjects without ADHD.  All participants were between the ages of 17 and 28. (Note:  The hyperactive-impulsive subtype is relatively infrequent -- especially among older individuals -- and was not examined specifically in this study).  The average age of participants was about 20, and between 15% and 30% of the participants in each group were female.  Participants were recruited from consecutive referrals from clinics specializing in child and adult ADHD and had been carefully diagnosed.

Participants in the community control group were recruited through newspaper advertisements, and were excluded if they reported any history of a major psychiatric disorder or if they had a significant number of ADHD symptoms.  Thus, they were not a random sample of community members because those with significant psychiatric problems were specifically excluded.  This is problematic, because it means that ADHD participants are being compared to a "healthier-than-normal" comparison group.  In particular, when rates of psychiatric diagnoses are examined, comparisons with the control group are not particularly informative because having a major psychiatric disorder was a specific reason for exclusion.  It is also important to note that control participants did not differ in socioeconomic status from those in the two ADHD groups.  Thus, the findings reported below cannot be attributed to the fact that comparison subjects were from more privileged backgrounds.
 

RESULTS

Education Outcomes

Participants in both ADHD groups had significantly fewer years of education than control subjects (an average of 13 years for both ADHD groups vs. 14.3 years for the control subjects).  Whereas almost 90% of control subjects were high school graduates, the graduation rates for combined and inattentive ADHD groups were 82% and 78%, respectively.  Participants in both ADHD groups were also significantly less likely to be college graduates -- about 7% of participants in each group compared to almost 25% of control subjects.  Participants in both ADHD groups were also significantly more likely to have received special education services.

Overall, participants with both subtypes of ADHD had clearly attained less successful educational outcomes than young adults without ADHD.  The educational attainment of participants in these groups, however, did not significantly differ from each other.  Thus, there was no indication that the risk of poor educational outcomes is different for individuals with the combined or inattentive subtype of ADHD.

Psychological Maladjustment

In this category, the researchers examined rates of psychiatric disorders and self-reported psychiatric symptoms among participants in the three groups.  The rates of diagnosis were based on results from a structured psychiatric interview.  As noted above, because control participants were excluded if they had a psychiatric disorder, comparisons with the control group are not particularly informative.  Instead, the comparisons of interest here are between the two ADHD groups.  Rates of psychiatric disorder for individuals with each ADHD subtype are shown below.  An asterisk (*) next to the disorder indicates that the difference in rates is statistically significant. (Note: In the list below, dysthymia is the term used for a chronic, low-grade depression).

Disorder	ADHD-C	ADHD-I
ODD*	45%	19%
CD	5%	3%
Major depression	13%	8%
Dysthmia	25%	17%
Alcohol abuse	37%	28%
Drug abuse	7%	0%
Learning disorder	38%	41%
Antisocial personality                 disorder	6%	0%

Disorder rates were consistently higher for the combined ADHD group than the inattentive group.  Except for Oppositional Defiant Disorder, these differences were not statistically significant.  This reflects the fact that the sample size was relatively small (it is more difficult to obtain statistically significant results with a smaller sample). A reasonable interpretation of these results is that rates of disorder among individuals with combined ADHD are greater than among those with inattentive ADHD.

Rates of self-reported psychological maladjustment were compared in the following areas: somatic complaints (i.e. aches and pains), obsessive-compulsive symptoms, depression, hostility/anger, anxiety, interpersonal sensitivity, and paranoid thinking.  As expected, both ADHD groups reported significantly more difficulties in all of these areas than control participants.  The combined ADHD group reported significantly more problems then the inattentive group in the areas of hostility/anger and paranoid thinking.

Antisocial, Drug, and Alcohol Histories

Rates of antisocial, drug, and alcohol-related difficulties for participants in the three groups are shown below.

Event	ADHD-C	ADHD-I	Control
Arrested	40%	19%	13%
Used illegal drug	80%	81%	52%
Considered self an alcoholic	12%	3%	3%
Considered self drug dependent	13%	10%	0%
Considered drug dependent by others	29%	14%	3%

Rates of arrest, self-reported drug and alcohol use, and perceptions by others of being drug- and alcohol-dependent were consistently higher among participants with ADHD.  And, rates for the ADHD-C group were higher in several areas than rates for young adults with inattentive ADHD.  Differences in arrest rates between the two ADHD groups were statistically significant.

History Of Mental Health Services

The final area examined was history of mental health service use.  As expected, rates of psychiatric medication use were higher among the ADHD groups than control participants (34%, 14%, and 2% for ADHD-C, ADHD-I, and control subjects, respectively).  Current or prior use of stimulant medication (e.g. Ritalin) was 28% for the ADHD-C participants and 13% for the ADHD-I participants.  This is a striking finding and indicates that, overall, the vast majority of participants with ADHD had never been treated with medication.  Thus, despite legitimate concerns about stimulant medication being inappropriately prescribed to individuals who may not need it, these findings highlight that many ADHD individuals who could legitimately benefit from medication treatment have never received it.  Rates of non-medical psychiatric treatment were 52% in the ADHD-C group, 29% in the ADHD-I group, and 22% among control subjects.
 

SUMMARY AND IMPLICATIONS

The major findings from this study are that young adults with either ADHD-C or ADHD-I subtypes are likely to be experiencing multiple domains of impairment.  When there are differences between the subtypes, it appears that those with ADHD-C are likely on the more negative end of the spectrum.

These findings highlight the degree to which ADHD -- when not properly treated -- can disrupt an individual's functioning beyond childhood and adolescence and into young adulthood.  And, an unfortunate situation implied by these findings is just how often individuals with ADHD fail to receive appropriate treatment (recall that fewer than 25% of these young adults reported ever having received stimulant medication treatment.)

Even though medication treatment is generally recommended to be only one component of an overall treatment program, it is concerning that so few of these participants had ever received a treatment of clearly demonstrated efficacy.  One can't help but wonder how many of the other difficulties experienced by young adults in this sample might have been avoided if they had been properly diagnosed and treated earlier in their lives.  This is really a critical issue because, although there is currently no known cure for ADHD, the general consensus among experts is that when core ADHD symptoms are managed effectively by carefully conducted treatment, the development of other problems such as those experienced by many participants in this study can be significantly reduced.

The authors of this study emphasize that young adults with ADHD seeking clinical services are not just a reflection of the normal population "...who are overly sensitive to ordinary difficulties with inattentiveness, as skeptics of adult ADHD in the popular media have sometimes contended."  Instead, these are individuals who are often struggling in multiple domains and who deserve to receive the best treatment and support available.

Results regarding the persistence of ADHD based on the results of these studies have been mixed.  In one study that began with 104 children, about two-thirds reported they were still troubled as adults (i.e. average age at last follow-up of 25) by at least one or more disabling core symptoms of the disorder, and one-third reported at least moderate-to-severe levels of hyperactive, impulsive, and inattentive symptoms.

In a second study conducted in Sweden, 49% of participants diagnosed with ADHD as children reported marked ADHD symptoms at age 22, compared to only 9% of control participants.  Although results from these studies indicate that many ADHD children continue to struggle with core ADHD symptoms as adults, formal diagnostic criteria were not used in either study.  Thus, it is not possible to determine the percentage of participants who continued to qualify for an ADHD diagnosis as young adults.

In the only longitudinal study in which DSM (Diagnostic and Statistical Manual of Mental Disorders) diagnostic criteria were employed (Manuzza et al., 1998), it was reported that only 8% met full diagnostic criteria at age 26.  (Note: A prior version of the DSM criteria from DSM-III was used in this study because that was the standard at the time.  You can review current diagnostic criteria here.)

These results imply that the vast majority of children with ADHD outgrow the diagnosis by early adulthood, and have led some to speculate that the adverse impact of ADHD becomes increasingly benign with advancing age.

Is it really the case that less than 10% of children diagnosed with ADHD will continue to meet diagnostic criteria in adulthood? Recently, Dr. Russell Barkely -- widely recognized as a leading authority on ADHD -- has identified two factors that may have contributed to this result.  First, Barkely notes that, although participants were diagnosed as children based largely on reports about their behavior provided by others (i.e. parents and teachers), their diagnostic status as adults was determined exclusively by their own reports.  Barkely suggests that that this switch could account for the low rates of adult ADHD reported, particularly since adults with ADHD may not provide accurate appraisals of their own behavior.

A second problem noted by Barkely is that, although the manifestations of ADHD may change over time, current diagnostic criteria are the same for adults as they are for children.  Thus, the same symptoms are used to diagnose children and adults and the same number of symptoms is required.  This could artificially reduce the likelihood that an individual would be diagnosed with ADHD in adulthood for two reasons.  First, if the manifestations of ADHD change over time, then the symptoms used to define the disorder in childhood may not include features that are more characteristic for adults.

Second, if ADHD symptoms as currently defined decline with age -- as they are known to do -- then it becomes increasingly unlikely for individuals to display the required number of symptoms as they become older.  As a result, Barkely suggests that the number of symptoms necessary for the diagnosis should vary with age. For example, suppose an individual displayed eight hyperactive symptoms at age seven and only four of these symptoms at age 25.  This individual would no longer qualify for an ADHD diagnosis based on current standards even though he might be just as "deviant" relative to same-age peers at age 25 as he was at age seven.  If this were the case, Barkely argues that the individual will appear to have "outgrown" the disorder by adulthood, "...whereas in fact they have only outgrown the criteria."

To account for this, Barkely suggests using age-adjusted criteria for the number of symptoms required rather than the fixed threshold method currently used.  Specifically, he argues that the number of symptoms required should be that which occurs in less than 2.5% of the population of individuals that age.  This number would be lower for adults than for children, but adults showing this reduced number of symptoms would still be as deviant relative to their same-age peers as are children displaying a greater number of symptoms.

How frequently does ADHD persist from childhood into adulthood when these two factors identified by Barkely are taken into account?  This issue was examined in a study published recently in the Journal of Abnormal Psychology  (Barkely, R.A. et al (2002). The persistence of AD/HD into young adulthood as a function of reporting source and definition of disorder, 111, 279-289).

Participants in this study were 158 young adults (ages 19-25) diagnosed with ADHD at an ADHD specialty clinic when they were between 4 and 12 -years old.  Eighty-one comparison subjects without ADHD from the same community were also followed into young adulthood.  Over 90% of individuals in both groups participated in the young adult evaluation, an extremely high retention rate for a study in which individuals were followed over so many years.

To determine ADHD status in young adulthood, two different methods were used.  First, participants' own reports of ADHD symptoms and ratings of their functioning in important life areas (e.g. school, work, peer relationships) were used to determine whether they met diagnostic criteria.  Determining diagnostic status based on the self-reports of young adults is the method that had been used in the study referenced above.

Second, parents of these young adults were asked to provide ratings of their child's ADHD symptoms so diagnostic status based on parent ratings could be determined.  Parent reports of their child's behavior were the basis on which the diagnosis had originally been made, and Barkely hypothesized that the continuity of ADHD from childhood to young adulthood would be much greater when parents were used to provide information on their child's behavior as an adult than when participants self-reports were the sole source of data.
 

RESULTS

How did self and parent ratings of current ADHD symptoms compare?

Young adults in the childhood ADHD group did not differ from control subjects in the number of ADHD symptoms they reported, with group averages of 2.1 and 1.5, respectively.  (Note - This is out of a total of 14 possible symptoms from DSM-III-R, which was the diagnostic system in place when the data were collected.  Currently, DSM-IV lists 18 specific symptoms of ADHD.)  In contrast, parents reported an average of 9.2 symptoms for adults in the childhood ADHD group vs. 1.7 for comparison subjects.

Based on self-report data, only 5% of young adults diagnosed with ADHD as children met current diagnostic criteria for ADHD.  When parent reports were used, however, 58% of these young adults met full diagnostic criteria for ADHD.  And, when the number of symptoms required for the diagnosis was adjusted to reflect the fact that ADHD symptoms tend to diminish with increasing age (see discussion above), this figure increased to 66%.

(Note: The authors also examined how many ADHD symptoms young adults recalled themselves as having displayed during childhood.  Adults in the childhood ADHD group recalled an average of 7.3 symptoms vs. 3.2 for the comparison group.  Based on these recollections of childhood symptoms, 47% of the childhood ADHD group would have qualified for a childhood diagnosis of ADHD.  Because 100% of these adults probably had the disorder in childhood, however, the accuracy of their recollections remains questionable.)

How valid are self and parent reports of ADHD symptoms in adulthood?

As indicated above, among young adults with childhood ADHD, self and parent reports of current ADHD symptoms differed dramatically.  Which reports were likely to be more accurate?

One way to examine this is to determine whether self or parent reports were more strongly related to how well the young adults were doing in various important life activities.   The life areas considered were: years of education, high school GPA, class rank during the last year of high school, employer-rated ADHD symptoms, employer-rated work performance, and number of arrests.  When the authors looked at how well parent and self-reports of ADHD symptoms predicted these life outcomes, they found that parent reports were superior in every case.  This provides compelling evidence that parent reports had greater validity than reports from the adults themselves.  The only outcomes predicted by self-reported ADHD symptoms were the number of ADHD symptoms reported by the employer and employer-rated work performance.  Even in these areas, however, parent reports were the stronger predictor.

The authors also examined how young adults in the childhood ADHD group were faring in these areas compared to adults in the comparison group.  In every area, they were found to be struggling: they had fewer years of education (12 years vs. 13.6 years), a lower high school GPA (1.7 vs. 2.5) and class rank (29th percentile vs. 49th percentile), were rated as showing more ADHD symptoms by their employer, had lower employer job ratings (3.2 vs. 4.2 on a 1-to-5 scale), and a greater number of arrests (.8 vs. .2).  (Note: All figures reported are group averages).  Interestingly, these adults were also significantly more likely to be living at home.
 

SUMMARY AND IMPLICATIONS

Results from this study make it clear that estimates of the persistence of ADHD into young adults varies dramatically depending on whether parents or young adults themselves are the source of information used to make diagnostic decisions.  When the information used is restricted to adults' self-reports of ADHD symptoms, it will seem as if the persistence of ADHD is very infrequent.  However, when parental reports are used, the persistence of ADHD becomes a far more frequent occurrence.

This was especially true when the diagnostic threshold for ADHD in adulthood was adjusted to reflect the fact that ADHD symptoms tend to diminish with age.  When this adjustment was made, two-thirds of individuals with ADHD continued to meet diagnostic criteria as young adults.  It should be noted that Barkely's suggestion to adjust the diagnostic criteria based on age is not something that everyone agrees with and is not the current standard in the field.  Whether this suggestion will be incorporated into the next round of diagnostic criteria for ADHD remains to be seen.

Results from this study also suggest that parent ratings of their adult children's ADHD symptoms are likely to be more accurate than young adults' ratings of their own symptoms.  Across multiple measures of educational and occupational functioning, parental reports of ADHD symptoms were better predictors of outcome than reports from the young adults.  The authors speculate that this may be the case because the diminished frontal lobe activity that is found in many individuals with ADHD may be associated with less accurate self-appraisal among ADHD adults.

It is tempting to interpret these results as indicating that adults with ADHD are typically inaccurate sources of information about their own functioning, and that it is problematic to rely on adults' self reports for making diagnostic decisions.   Although this conclusion appears to follow directly from the study's results, it is difficult to reconcile with the fact that many adults with ADHD are acutely aware of their struggles and seek out evaluation and treatment for that reason.  What could account for this apparent discrepancy?  Two possibilities come to mind.

First, it is important to remember that adults in this study were not seeking treatment, but were being evaluated as part of an ongoing research project.  It seems reasonable to hypothesize that adults who seek treatment for ADHD are quite different from those who do not, and that the former may be better able to report accurately on their symptoms and behaviors related to ADHD.  Thus, one should not interpret the results of this study to indicate that adults' reports of ADHD symptoms will invariably underestimate what is actually the case.

A second possibility has to do with the manner in which information about ADHD symptoms was obtained in this study.  Adults were simply asked to rate how often ADHD symptoms occurred, on a 1-to-5 scale ranging from "not at all" to "almost always".  Symptoms rated as occurring "frequently" -- this was the midpoint on the scale -- were counted as "present".  It thus appears that only a single question pertaining to each symptom was used to classify it as "present" or "absent".  This may have resulted in fewer symptoms being counted as "present" than if the researchers had conducted a more extensive interview and done more in-depth probing of the different symptoms of ADHD.  Thus, the method used may not have been as sensitive as it needed to be.

In spite of these two possibilities, results from this study certainly support the value of obtaining information about ADHD symptoms from multiple sources when evaluating an adult for ADHD.  This is certainly the procedure recommended when evaluating children, and this study highlights the value of this approach in adult evaluations as well.

It is also important to emphasize that, regardless of whether the adults in this study continued to meet full ADHD diagnostic criteria, those in the childhood ADHD group were clearly having difficulty relative to other adults.  These findings highlight the long-term adverse effects that ADHD can have on individuals' development, and the compelling need that many individuals have for ongoing support and treatment, even when they may no longer meet full diagnostic criteria.