Attention Research Update

April 2000

"Helping parents, professionals and educators stay informed about new research on ADHD"

David Rabiner, Ph.D.  Senior Research Scientist, Duke University

In this issue...

An objective procedure for assessing ADHD?

Increasing use of stimulant medication treatment for preschoolers

The continuity of behavior problems and ADHD
symptoms from preschool through early adolescence

How does young adult outcome relate to childhood ADHD,
aggression, and stimulant medication use?


An ongoing concern among both parents and professionals in regards to evaluating children for ADHD is the lack of an "objective" procedure for making the diagnosis.  Currently, a careful evaluation involves combining information from multiple sources (i.e. parents, teachers, and child) using both interview and behavior rating scale data.  Depending on the particular circumstances, the use of formalized psychological tests may also be incorporated.  In the end, however, the diagnosis is based on the clinician's best judgment using all available sources of data (for a more complete discussion of the evaluation process click here.)

Although there is overwhelming support for the fact that well-trained clinicians using empirically based procedures are able to make the diagnosis of ADHD in a reliable and valid manner, many object to the fact that the diagnosis still comes down to the clinician's "judgment".  (Of course, it is important to be aware that this is no different from the diagnosis of any other type of psychiatric disorder.)  In response to this concern, there have been several efforts to develop "objective" procedures for diagnosing ADHD, the most widely used of these being the Continuous Performance Test (i.e. CPT - you'll find a brief description of this procedure below).

Recent data on the differences found in different aspects of brain structure and function in individuals with ADHD have also contributed to the hope that an "objective" diagnostic procedure used sophisticated technologies could be developed.  The problem with such technologies for diagnosis, however, is that they have not been sensitive or specific enough to be used for diagnosing individuals.  In other words, even though groups of individuals with ADHD appear different using such procedures compared to groups of individuals without ADHD, there has generally been a fair amount of overlap between the groups.  Thus, some individuals with ADHD do not look any different from "normal" subjects on these tests and some "normal" subjects produce results that are similar to those found in many ADHD individuals. Using these technologies as the basis for diagnosis would thus lead to many diagnostic errors that would "identify" false positives (i.e. diagnosing someone with ADHD who did not have it) and false negatives (i.e. failing to diagnose someone with ADHD who had the condition).

A study published in a recent issue of Neuropsychology, however, provides initial evidence of the potential utility of at least one objective procedure for diagnosing ADHD Monastra, V.J., et al. (1999). Assessing ADHD via quantitative electroencephalography, Neuropsychology, 13, 424-433).  The hypothesis underlying this study is that cortical slowing in the prefrontal region of the brain can serve as a basis for differentiating individuals with ADHD from nonclinical control groups.  (Recall the article from the March issue of ADHD RESEARCH UPDATE in which enhancing physiological arousal - and presumably cortical activity -  was associated with improved performance on several tasks by children with ADHD.  This is also consistent with the hypothesis that cortical underactivity in particular brain areas is the underlying biological basis for ADHD.)

Participants in this study were 482 individuals between the ages of 6 and 30.  Both males and females were well- represented in the sample.  Participants were classified into 3 groups (i.e. ADHD-Inattentive Type, ADHD-Combined Type, and controls) on the basis of the results of a standardized clinical interview, behavioral rating scales, and a Continuous Performance Test (i.e. CPT).  The CPT is a procedure in which individuals sit in front of a computer screen and are required to press - or not press - certain keys depending on which stimuli appear on the screen.  It is designed to evaluate both sustained attention and impulsivity, and is one of the most frequently used "objective" procedures to aid clinicians in evaluating individuals for ADHD. (For a discussion of the distinction between the inattentive and combined subtype of ADHD click here.)

Cortical activity in the prefrontal region for these participants was then assessed using a quantitative electroencephalographic procedure (QEEG) - a technique that assesses electrical activity in particular brain areas. The question of interest is the degree to which individuals who had been previously diagnosed with ADHD on the basis of more traditional diagnostic procedures would also be correctly identified as having ADHD or not, based on these QEEG data alone.


The results were impressive.  Overall, the authors report that this "diagnostic test" yielded sensitivity and specificity results of 86% and 98% respectively.  This means is that 86% of the individuals diagnosed with ADHD using the conventional diagnostic procedures were also classified as having ADHD using the QEEG procedure alone.  Similarly, 98% of those who did not have ADHD according to the interview, CPT, and rating scale data were not classified as having ADHD using the QEEG data.  Thus, there was substantial agreement between the diagnostic determinations made using these very different approaches.  Consistent with what has been found previously, individuals with ADHD were found to produce results on the QEEG that indicated cortical slowing (i.e. less activity in the brain area being evaluated).  According to the authors, these results held up across the two subtypes of ADHD and for both genders as well.

Because some researchers - notably Russell Barkley - have suggested that the inattentive subtype of ADHD is likely to represent a very different condition that the hyperactive/impulsive or combined subtypes, it is especially interesting that similar results were obtained for the inattentive and combined subtypes in this study.  This suggests that individuals with these different subtypes of ADHD may have a common neurological feature (i.e. slowed cortical activity in the prefrontal areas of the brain) in common, although additional research on this issue is necessary prior to making any such conclusion.

Summary and Implications

The results of this study would seem to provide promising initial evidence that this procedure may have important utility as an "objective" test for diagnosing ADHD.  It is important to remember, however, that one's opinion about this depends on what you think the "gold standard" for diagnosis should be.  If you believe that careful interviews of parents, teachers, and children - along with the collection of standardized behavior rating scales - provides the most complete set of data on which to make a diagnostic decision, then the procedure investigated here still failed to identify about 14% of individuals of having ADHD who "really did".

To put this another way, what should one do when information from parents and teachers - and perhaps even direct observational data - all support an ADHD diagnosis for a child, but the QEEG does not?  A prudent approach in this instance would be to first be sure to rule out possible alternative explanations for the symptoms that are being so prominently displayed.  For example, one would want to be sure that the child's symptoms were not better explained by some type of mood or anxiety disorder, reaction to a recent stressor, elevated lead levels, etc..  If all plausible alternatives are ruled out, however, than what?  Does one want to conclude that it can not be ADHD because the "objective"  QEEG data is inconsistent with this?

I don't think there is a clear-cut answer to a question like this, but this is an important issue to consider as efforts to further develop objective diagnostic procedures for ADHD proceed onward. It would have been quite interesting if the authors provided more detailed information about what seemed to be going on among these participants in their study (i.e. those who were diagnosed with ADHD using the traditional means but not the QEEG).  Nonetheless, this is a very interesting study that provides initial information on what may become a useful component of the diagnostic process.  Replications of these results in an independent sample is of course required - as the authors note - before the utility of this approach becomes clearer.

It is also important to note that there was no clinical control group in this study - i.e. participants without ADHD but with some other psychiatric diagnosis.  Thus, it is possible that the procedure they used may be effective for distinguishing between people with and without some type of psychiatric diagnosis, rather than being specific to ADHD per se.  Future studies should include a clinical control group in addition to "normal" controls so that this important issue can be disentangled.  For example, if the authors find that their QEEG procedure is just as effective in differentiating individuals with ADHD from individuals with another psychiatric diagnosis, as it was here in distinguishing people with ADHD from non-psychiatric control subjects, the results would be even more impressive.


As you may already be aware, stimulant medication has been back in the news recently in regards to the increasing frequency with which it is being prescribed for preschool children.  Hillary Clinton herself has become involved in this important issue and has expressed what I believe are some appropriate concerns about the use of stimulants and other medications with preschool-age children.

The research that has promoted this emerging debate was published last month in the Journal of the American Medical Association (Zito, J.M., et al., (2000). Trends in the prescribing of psychotropic medications to preschoolers. JAMA, 283(8), 1025-1031).  In this important study, the authors examined the prescription records from 2 state Medicaid programs as well as a large HMO.  Specifically, they looked at the use of stimulant medications, clonidine (often prescribed for treating ADHD), and antidepressants with children between the ages of 2 and 4.  Data from 3 different years - 1991, 1993, and 1995 was examined so that they could look at trends in the use of these medications with young children that may have emerged over time.

At each year, over 200,000 children in the 2 to 4-year-old age range comprised their sample. The data were examined separately within the 2 state Medicaid programs and the HMO to look at possible differential usage in these different health systems.  Because this was such a large and representative group of children, it is quite reasonable to believe that the results provide an accurate reflection of changes in medication use for preschoolers that have occurred throughout the nation during this time, although the possibility of important regional variation can not be overlooked.


The authors' report results that provide clear indication of increasing use of all 3 types of medications - stimulants, clonidine, and antidepressants - between 1991 and 1995.  Below are what appear to be the most important findings:

* Within all 3 systems, there was a significant increase in the use of each medication for preschool children between 1991 and 1995.

The rate of increase ranged from just under 200% at one site to about 300% at the others.  Percent increases in the use of clonidine were even more dramatic - as high as 2300% in one of the sites.  The increases in the use of antidepressants were more modest, but also showed considerable variability between the 3 different insurance system populations.

* Although the increase in the use of these medications was large, the absolute usage of each medication type remained relatively small.

An important point that seems to be missing from media coverage of this study so far is that the absolute usage of these medications - although it increased substantially - remained relatively modest in an absolute sense. (Of course, what constitutes "relatively modest" is a judgment call.)

For example, the actual rate of stimulant medication usage in the 1995 data ranged from about .05% of children per year at one site to about 1.2% at the highest use site.  In other words, even though the rate of prescribing to preschoolers had increased dramatically between 1991 and 1995, there remained relatively few 2-4 year olds who were receiving such prescriptions during the highest use year.  The percentage of children who were prescribed clonidine or antidepressants was far lower - well less than 1 child per 100 at all 3 sites.

In contrast, the percentage of 5-9 year old children who were prescribed stimulant medication during 1995 was much higher - representing about 6% of this population.

* There was a greater proportional increase in the number of preschool age-girls receiving stimulant medication treatment.

Once again, this varied substantially across the 3 sites. Overall, however, the ratio of boys to girls who received stimulant medication treatment between 1991 and 1995 decreased.

Summary and Implications

These data tell an important story in changes that are occurring in how the medical community is responding to the emotional and behavioral difficulties in young children that they are often asked to address.  As these data make clear, the use of medication to address these issues is becoming an increasingly frequent method used to try and help children with such problems.  Although I am not sure if anyone knows whether the trends that were identified between 1991 and 1995 have continued, it would not be surprising to me if they have, and this would mean that an even larger percentage of young children are currently being treated with stimulants and other medications. These data also suggest that this may be particularly true for young girls, although this is not completely clear and is likely to vary from region to region.

In light of these facts, it is instructive to review what is currently known about the efficacy of such medications for this age group, and what national professional associations have suggested about the appropriateness of such treatment in young children.  Regarding the use of stimulant meds such as Ritalin with preschoolers, there is considerably less data available than there is to support the use of such meds with older children.  Those studies that have been published have reported results that are positive, but the need for additional research with this age group is clearly recognized.  In addition, a number of professionals have expressed concerns about possible adverse effects of long-term treatment with stimulants that is initiated at such a young age, and the authors of this study identify the important need to carefully study this issue.

In the practice guidelines for ADHD published by the American Academy of Child and Adolescent Psychiatry in October of 1997, important cautions about the use of stimulant medication for preschool children were raised. These guidelines stressed the need to carefully consider the presence of alternative explanations of a young child's symptoms before making a diagnosis of ADHD in a preschool age child.  This, of course, is important to do at all ages but is especially critical in young children who may show behaviors that lead one to wonder about the possibility of ADHD for a variety of reasons.  In addition, such behaviors will tend to be less stable in young children than in school-age children because of the important maturational differences that characterize children of this age.

The guidelines also stressed that "stimulants should be used in this age group only in the more severe cases or when parent training and placement in a highly structured, well-staffed preschool program have been unsuccessful or are not possible."  If medications are used, the recommendation was to exercise more caution, to use lower doses, and to monitor the child's response more frequently.  In addition, it was noted that dietary interventions should be considered as an alternative to try with this age group.

Unfortunately, the data in this study do not provide any indication as to whether any such appropriate cautions were typically heeded.  Based on other data that has been presented about problems in how stimulant medication is often prescribed even among older children, it would certainly not be surprising if such important guidelines were often not incorporated into a young child's treatment.

One possible adverse consequence of all the good research that has been done to document the efficacy and - as far as we know - safety of stimulant medications for children with ADHD, is that it may be leading some physicians to be less cautious about prescribing it.  If this were the case, it could extend to the more frequent use in very young children - as an initial option rather than as something to consider after a variety of alternative interventions have been attempted.

One very positive outcome of the debate that seems to be emerging in response to this study would be a shift in the focus from is the use of medication to treat ADHD "good" or "bad", but rather, how can medication treatment be used most appropriately to produce the greatest possible benefits for children.  Careful attention to this issue would be likely to result in outcomes that would be of greater benefit to children and families.


An important issue raised by the article reviewed above concerns the meaning of both ADHD symptoms as well as other behavior problems that are often observed in preschoolers.  If such difficulties tend to be transient - that is, most preschoolers who show these problems tend to "grow out" of them as they develop and go on to make satisfactory adjustments later on - then the need for real caution in regards to the diagnosis and treatment of such problems in young children would be emphasized. If, however, these difficulties at an early age portend ongoing adjustment problems for many children, and are clearly associated with a diagnosis of ADHD as well as other behavior disorders later on, then the implications would be quite different.

Addressing these critical questions require a longitudinal study in which children are assessed for ADHD symptoms along with other difficulties at a young age, and then followed into childhood and beyond.  This type of longitudinal design is really the only way to determine what the childhood and early adolescent outcomes of ADHD symptoms and other problems that are observed in preschoolers.  As you can imagine, these studies are time-consuming, difficult, and expensive to do for a number of reasons.  They are, however, critically important.

A recent issue of the Journal of Clinical Child Psychology includes an excellent study of this key question (Pierce, E.W., Ewing, L.J., & Campbell, S.B. 1999. Diagnostic status and symptomatic behavior of hard-to-manage preschool children in middle childhood and early adolescence. Journal of Clinical Child Psychology, 28, 44-57).  This paper describes the results of 2 related investigations in which two cohorts of hard-to-manage preschoolers were followed from age 3 or 4 into middle childhood (i.e. age 9) or early adolescence (i.e. age 13).

Participants in these studies were initially recruited from a variety of sources including pediatricians office, preschool classrooms, and mother's "morning out" groups on the basis of parent complaints that their preschooler was showing hyperactive, impulsive, inattentive, noncompliant, and aggressive behavior.  Comparison children who were not seen as having these same types of problems were recruited from the same settings, and were matched as closely as possible to the hard-to-manage group on the basis of race, gender, and socioeconomic status.  Both boys and girls were included in the initial cohort whereas the second cohort included boys only.  Even in the first cohort, however, there were was not a sufficient number of children to make meaningful gender comparisons in the results that are reported below.

In cohort 1, the original sample included 46 hard-to-manage 3-year-olds and 22 comparison children.  Parents completed standardized behavior ratings on these children at ages 3, 6, and 9.  These ratings included items covering ADHD symptoms specifically and other types of disruptive behavior problems more generally.  When the children were 13, their mothers were administered a semi-structured clinical interview called the Child Assessment Schedule to assess the diagnostic status of the child.  Mothers and children also completed standardized behavior ratings at this time.  Approximately 75% of the original sample was included in this final follow up.  A similar procedure was used with the second cohort.   As noted above, this cohort included boys only, and participants were assessed initially at age 4 and only followed thru age 9.


The results from this impressive set of studies are extensive and more than can be fully summarized here.  Below are those aspects of the results that seemed most important to me.

* Significant behavior problems during preschool persist in many children.

One important question confronting parents with a difficult preschooler is whether their child's difficulties portend ongoing problems, or will be likely to diminish over time.  This study provides important data on this question.

In the first cohort, about 50% of children in the "hard-to-manage" group at age 3 were diagnosed with ADHD at the age 13 follow-up.  This compared to only 8% of children in the control group.  The hard-to-manage children were also significantly more likely to be diagnosed with ODD or CD at follow-up.  They were no more likely than comparison children, however, to be diagnosed with an internalizing disorder (e.g. depression or anxiety).  In the second cohort, similar results were obtained, although the differences at follow-up (age 9 for this cohort) were not as strong as those found with the initial group.  Why this may have been the case is unclear.

* Although continuity for early behavior problems is often found, many "hard-to-manage" preschoolers will make much more satisfactory adjustments over time.

This is the flip side of the data presented above.  As the figures noted above make clear, many hard-to-manage preschoolers were not showing sufficient symptoms to warrant any diagnosis at the follow-up evaluation.  Thus, many young children who are showing classic symptoms of ADHD will not display sufficient symptoms later on to warrant this diagnosis.

What seemed to make the difference?  According to the authors, preschoolers whose problems were still evident at school entry - roughly age 6 - were those who were much more likely to warrant a diagnosis for ADHD and/or another behavioral disorder (i.e. ODD or CD) at the last follow-up period.  In both cohorts roughly 50% of the hard-to-manage preschoolers were still regarded by their mothers as showing important problems at the time of school entry.  These were the children who were likely to still be showing important difficulties - including ADHD - at the age 13 (cohort1) or age 9 (cohort 2) follow-up.

So, overall, roughly 50% of preschool children showing high levels of behavioral difficulty will continue to show such problems at the time of school entry.  Of this group, the majority will continue to show sufficient problems to warrant a clinical diagnosis of ADHD, ODD, and/or CD in middle childhood or early adolescence.

* Symptom severity during preschool is the best predictor of which preschool children are likely to have persistent problems.

This finding was clear-cut and not surprising: among the hard-to-manage preschool group, those whose difficulties persisted to school-entry and beyond had significantly more severe problems at age 3 or 4 than those hard-to- manage children whose symptoms had diminished at school entry.  The combination of severe ADHD symptoms and oppositional behavior at a young age was the strongest predictor of persistent problems.

The important general conclusion to be reached from these data are that children with high levels of early symptoms are less likely to outgrow these problems, and once their problems persist through school entry, they are likely to become even more entrenched.

Summary and Implications

These results clearly underscore the importance of taking parental complaints/concerns about their preschooler's behavior seriously and of providing help in these situations. Even though a number of hard-to-manage preschoolers will apparently outgrow their difficulties, those displaying the more severe problems are less likely to do so in the absence of early intervention efforts.  The longer these difficulties persist, the more difficult it becomes to help a child get back on a good developmental track.

Do these data support a conclusion that the increase in medication treatment for preschoolers described in the article above is appropriate?  Not necessarily.  Instead, I believe these data argue that for many preschoolers showing behavioral difficulties, early intervention may be extremely important.  There is no reason, however, why this intervention necessarily needs to be the use of medication, particularly as the initial intervention tried.

Instead, it would seem quite reasonable to consider behavioral interventions that focus on helping parents deal with their child's challenging behavior more effectively, and to provide such consultation to the child's teacher where appropriate. Environmental factors that may be contributing to the child's difficulties also need to be carefully considered.  As noted by the American Academy of Child and Adolescent Psychiatry, dietary interventions may also be a useful approach in some preschool children and are another avenue to consider.

When such interventions have been carefully conceived and carefully executed, but the child's problems show little signs of abating, medication is another option that can be considered. As noted in the article above, however, there is currently far less evidence to support the use of medication for treating emotional and behavioral problems in this age group - both in terms of efficacy and safety.  So, when attempted, this should be done very carefully and the child's response should be monitored regularly.

Two common responses that parents often encounter when seeking advice about dealing with their difficult preschooler - "Lets try medication" or "Don't worry about it.  It is just a phase your child will grow out of" are probably not the most helpful ways for handling such a situation.  Instead, a careful assessment of the difficulties that lead to a well-conceived way to address them, and to evaluate the success of the intervention(s) being used, is likely to produce better outcomes down the road.  In most circumstances, this is most likely to be provided by an experienced child mental health professional or developmental pediatrician, as most family physicians and regular pediatricians will not have the same level of training or experience with such behavior problems in young children.


This important question was examined by researchers addressed in a study published recently in the Journal of Child and Adolescent Psycho- pharmacology (Paternite, Loney, Salisbury & Whaley, 1999, Childhood Inattention-overactivity, aggression, and stimulant medication history as predictors of young adult outcome. JCAP, 9 , 169-184).  This is one of the few studies yet conducted in which the long-term effects of stimulant medication treatment have been examined and is thus an important addition to thetreatment literature on ADHD.

These authors were interested in determining how ADHD symptoms, aggression, and a child's history of stimulant medication treatment related to outcome in young adulthood. This type of longitudinal work - i.e. following a group of children identified with ADHD over time - is critically important for understanding the long-term impact of ADHD and how medication treatment may effect children's outcomes.  This latter question is especially important because although many studies have documented the short-term benefits of medication, and more recent work has indicated that such benefits persist beyond a year (see the MTA study as an example of this), there is virtually no research to indicate that medication treatment during childhood results in meaningful improvements in young adult functioning.

Participants in this study were 121 boys who were initially diagnosed with ADHD when they were between 4 and 12 years old.  These boys were diagnosed back in the 1970s, and the diagnostic criteria in use at that time were substantially different then they are today.  In fact, in the official terminology in use at that time the boys were diagnosed with either hyperkinetic reaction of childhood (HK) or minimal brain dysfunction (MBD).  Using detailed information about specific symptoms contained in boys' medical charts, however, the authors estimated that approximately 71% of these boys would have qualified for a diagnosis of ADHD using today's diagnostic criteria.

All 121 boys were treated with stimulant medication (i.e. methylphenidate) and their treatment with medication was initiated between 1967 and 1972.  (It would have been preferable, of course, if girls were also included in this sample.  Several decades ago, however, females were probably even less likely to be recognized as having ADHD than they are today, so it is not surprising that only boys were available to follow.)

Based on information contained in these boys, medical records, the authors created scores for each boy that provided an indication of the strength of both ADHD symptoms (i.e. inattention and hyperactivity/impulsivity) and aggression.  These ratings were made based on information from parents, teachers, and the boys themselves that were included in the boys' charts.  Because prior research has clearly shown that ADHD symptoms and aggression both make independent contributions to children's long-term development, these two aspects of children's behavior were considered separately.  (The authors actually hypothesized that childhood aggression would be more strongly associated with negative outcomes in young adulthood than would ADHD symptoms.)

As noted above, all participants in this study were treated with stimulant medication at some point.  Thus, the authors were not comparing the long-term outcomes of children with ADHD according to whether or not they had received medication.  Instead, they carefully reviewed boys' medical records to document the characteristics of the medication treatment that was provided. They then examined how important aspects of medication treatment - i.e. overall response to medication, dose of medication received, and duration of medication treatment - were related to various outcomes in young adulthood. These ratings were based on information contained in the boys' charts and were made by raters who were blind to how the boys were doing in young adulthood.

The young adult assessment occurred when participants were between 21 and 23 years old.  Approximately 80% of the original sample participated in the young adult evaluation, and those who did were representative of the sample as a whole. A wide variety of  measures were collected at this assessment including information on psychiatric diagnoses, overall level of functioning in important life areas, intellectual functioning, academic achievement, and overall life circumstances (i.e. social, educational, and  employment circumstances). Within each of these broad domains, a number of different measures were included so the overall number of outcome measures considered was quite large.  This assessment thus provided a comprehensive account of how participants were faring at this point in their lives.


The basic questions of interest in this study are as follows:

1. How  do childhood levels of ADHD symptoms relate to functioning in young adulthood?

2. How does childhood level of aggressive behavior relate to functioning in young adulthood?

3. Does young adult outcome relate to a child's overall response to stimulant medication, the magnitude of the dose a child typically received, and/or the duration of medication treatment?

This last question is actually quite important as there are distressingly few long-term studies of the effect of stimulant medication treatment on important outcomes for children with ADHD.  As noted above, the authors of this study were not comparing outcomes for children with ADHD depending on whether they were treated with medication.  Instead, they were interested in how important aspects of medication treatment (i.e. overall response, dose, and duration) predicts important life outcomes over a a number of years.

Because of the large number of outcome measures collected, it is perhaps more important to look at the overall pattern of results rather than focusing on specific outcome variables.  When the results of the study are considered in this way, they are really quite clear.

Overall, childhood aggression was a stronger predictor of problems in young adulthood than was childhood ADHD.

The authors reported that aggression was uniquely related to 38% of the outcomes they considered while ADHD symptoms predicted only 10% of these outcomes.  Among the specific outcomes predicted by childhood aggression were depression, drug abuse, antisocial personality disorder, and a variety of functional impairments.  ADHD symptoms, in contrast, predicted only lower levels of employment and greater general levels of impairment.

Thus, these results clearly indicate that high levels of aggression during childhood poses significantly greater risks to a child's long-term development than do ADHD symptoms alone.

Children who had a better overall response to medication, who received higher doses, and who were treated for a longer duration, had better outcomes in a variety of different areas.

The aspects of medication treatment the authors examined - i.e. overall response, dose, and treatment duration - were not related to all of the different outcomes considered. For 11 different outcome measures, however, these aspects of medication treatment were found to be significant predictors, and in every case but one, results were in the direction of indicating significant benefits for medication.

Thus, it is reported that better overall response to medication was associated with lower depression scores and better social functioning in young adulthood.  They were more likely to be living independently of their parents, and to be either married or engaged. Similarly, children treated with higher doses were less likely to be diagnosed with alcoholism in young adulthood and had made significantly fewer suicide attempts. Finally, the longer a child was treated with medication the higher his IQ and academic achievement in reading and math was likely to be.  The only clear negative finding was that children who had a better overall response to medication treatment were less likely to graduate from high school.  This is a perplexing finding given the overall pattern of results, and may represent a chance finding that would not be replicated.  In general, however, medication treatment tended to have positive effects long after it had been discontinued.

Summary and Implications

What seems fairly clear from these results is that childhood aggression as measured in this study is associated with poorer young adult outcomes.  Also, MPH treatment is associated with better young adult outcomes.  Specifically, positive outcomes occur in psychiatric, cognitive, academic, and social domains when considered in children who respond well to early treatment and who were treated with higher doses for a greater period of time.

These results, although informative and potentially influential, should be viewed with caution in regards to what can be concluded about the long-term benefits of stimulant medication treatment.  There are several different reasons why such caution is necessary.  First, no information is provided about the families of these patients or the total number of clinical visits each child made.  It may be that more frequent visits to the doctor - which would be likely to correlate with a greater duration of medication treatment - is indicative of a more concerned family, or a family that is more invested in their child's well-being.  Conceivably, the child's improvement could be a consequence of this investment rather resulting from more extensive treatment with medication.

It is also important to keep in mind that the diagnostic criteria for ADHD in use at the time the children in this study were first diagnosed were quite different than they are today.  Thus, not every child in this study would be diagnosed with ADHD using current guidelines and the degree to which these data would apply to children diagnosed today is not completely clear. In establishing the long-term effects of medication treatment, it would also be important to compare outcomes for children who did, and did not, receive treatment with medication. That was not done in the current study, although the authors indicate that they have additional papers on using this data set in which such comparisons are described.  Finally, there were no girls included in the sample so the extent to which these findings would generalize to girls is completely unknown.

Despite these limitations, the value of this study is clear. The keys points to emphasize is that aggression during childhood appears to be a better predictor of negative long-term outcomes than ADHD symptoms alone.  This replicates findings that have been reported in prior work, and underscores the importance of carefully attending to this aspects of children's behavior.  When a child has both ADHD and aggressive behavior problems, parents and professionals need to be especially vigilant to make sure that interventions which address the aggressive behavior problems are an important part of the child's treatment.

In regards to medication treatment, these data do provide important initial support that such treatment is associated with long-term benefits.  When children have a good response to meds, receive an adequate dose, and are maintained on medication over a longer period of time, they are likely to benefit as a result.  This is not conclusive proof by any means that stimulant medication produces meaningful long-term gains, but it is certainly consistent with this hypothesis and is a step in the right direction of establishing such proof.  I look forward to future publications from these authors and will include summaries of their work in ADHD RESEARCH UPDATE as it appears.

(c) 2000 David Rabiner, Ph.D.

Information presented in Attention Research Update is for informational purposes only, and is not a substitute for professional medical advice.