Attention Research Update

April 2003

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

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


(Note: Attention Research Update has received support from companies pharmaceutical companies who manufacture stimulant medications.  I do not believe this has influenced my review of the study described below, but I want you to be aware of this relationship.)

A concern frequently expressed about stimulant medication treatment for children with ADHD is that it will increase the risk of future subsequent use and/or abuse.  One reason given for this concern is that children who take stimulants are learning that it is "okay" to use drugs, and that this can lead to substance abuse later on.  It has also been suggested that the long-term use of stimulants may alter children's sensitivities to drugs with similar chemical properties (e.g. cocaine), and this could also increase the risk of illicit substance use.

In fact, in one previously published study it was reported that cocaine and nicotine abuse were higher in individuals who received stimulant medication as children. Other studies, however, have found no association between stimulant treatment and later substance use, or, have found that stimulant treatment reduced the risk of substance use/abuse in adolescence or young adulthood.

Because results from prior research have been somewhat contradictory, and the relation between medication treatment and substance use remains a concern for many parents and professionals, additional efforts are required to resolve this critical issue.  A paper published recently in Pediatrics (Wilens, T.E. et al., (2003). Does stimulant therapy for attention-deficit/hyperactivity disorder beget later substance abuse?  A meta-analytic review of the literature. Pediatrics, 111, 179-185) provides a comprehensive review of prior studies on this topic and offers a helpful synthesis of previous findings.

The authors did not collect any new data themselves.  Instead, they conducted what is referred to as a meta-analysis of previous research. In a meta-analysis, the researchers begin with a systematic effort to locate all published research on a given topic that meet a predetermined set of scientific standards. These studies are pooled and an analysis of the pooled data (i.e. a "meta-analysis) is conducted.  Thus, the totality of evidence across a number of studies is used to determine whether there is evidence to support a specific hypothesis.

As is evident in the above description, a "meta-analysis" includes far more subjects than were included in each individual study.  In addition, results of a meta-analysis reflect a composite of studies that may have used slightly different methods and procedures, and that had samples of participants with somewhat different characteristics.  Because results from a meta-analysis are based on a larger and more diverse population, they are generally considered to be more definitive than results provided by any individual investigation.

The authors conducted a systematic literature search of all available studies of children, adolescents, and adults that had information on childhood exposure to stimulant therapy and data on substance use disorder (i.e. SUD) in adolescence or adulthood.  A total of 6 studies on this issue were identified.  Included among this group were 4 studies in which children were followed into adolescence and 2 where children or adolescents were followed into young adulthood.

In all of these studies, some participants were treated with stimulant medication treatment and others were not.  This enabled the researchers to examine whether or not medication treatment with stimulant altered the risk of developing an SUD.  For example, if fewer children treated with stimulant medication developed an SUD than children who did not receive medication treatment, it would indicate that medication treatment lowered the risk of this outcome.  Conversely, if a larger percentage of  children receiving medication treatment developed an SUD, it would suggest that medication treatment increases the risk of developing substance use problems.

Before reviewing the results, it is important to recognize that in these studies, whether or not children received medication children was not determined by random assignment. Instead, this reflected the decision of each child's parents and/or health care provider.  This introduces potential complications because factors that may influence the decision to use medication (e.g. how severe the child's symptoms are) may also influence the likelihood of substance use disorders developing.

For example, when a child's ADHD symptoms are severe, it probably increases the odds that stimulant medication will be prescribed.  At the same time, it may also increase the risk for later substance use problems. As a result, it could appear that taking medication increases the risk for developing an SUD when it is actually the severity of the child's condition, and not exposure to medication treatment, that accounts for this. In an effort to control for this possibility, the authors were careful to consider whether the severity of ADHD symptoms, and the presence of other problems such as Conduct Disorder, was equivalent between medication treated and non-medication treated children at the beginning of treatment.


The authors tested 3 competing hypotheses. The first was that stimulant therapy would have no demonstrable effect on the development of SUD.  The second was that stimulants would produce a higher risk for SUD.  The third was that stimulant treatment would diminish later risk for SUD.  Any drug and/or alchohol use disorder other than cigarette smoking was considered an SUD.

Overall, the 6 studies included in the meta-analysis involved 674 children treated with medication and 360 who were not. In 4 of these 6 studies, medicated and unmedicated subjects demonstrated similar levels of symptom severity and co-occurring disorders at baseline. Information on treatments received by non-medication children was not provided.

Results of the meta-analysis based on pooled data from the 6 studies indicate that children treated with medication were about half as likely as other children to develop an SUD.  The difference in SUD rates between treated and untreated youth was greater in studies where children were followed into adolescence than into adulthood, and where treated and untreated youth showed equivalent symptom severity when the study began. In fact, in studies where medication treated and non-medication treated children showed equivalent severity prior to treatment, those who not receive medication were over 3 times more likely to develop an SUD.


Results from this meta-analysis indicate that medication treatment for children with ADHD is likely to reduce their risk of developing a SUD.

This was especially true when SUD outcomes were examined in adolescence.  The protective effects of medication treatment on the development of SUD in adulthood were still evident, but less pronounced.  The authors suggest this may have occurred because the studies they reviewed were done at a time when it was typical to discontinue medication treatment during adolescence. Thus, by the time participants reached adulthood, many had been without medication treatment for a number of year.  Perhaps the lack of medication coverage during this time increased the likelihood of a SUD developing.

It is important to recognize that although these results indicate that medication treatment appears to reduce the risk of SUD, the reasons why this occurs remain unclear.  It may be that medication treatment helped children experience greater success in school and with peers, and these factors reduced the risk of SUD.  Medication treatment may also have inhibited the development of severe conduct problems, which would also have made the use of illicit substances less likely. A variety of explanations are possible, and this study does not enable a conclusion to be drawn.

It is also important to note that these findings should not be interpreted to suggest that children whose treatment does not include medication will inevitably develop an SUD. If a child's ADHD symptoms are managed effectively by other methods (e.g. behavior therapy, neurofeedback, etc.) so that he/she is able to experience success in school, and serious behavior problems do not develop, there is no reason to assume that they are at increased risk for SUD because they are not receiving medication.

In all likelihood, the key factor related to the risk of a child with ADHD developing SUD is how well the child's ADHD symptoms are being managed and the level of success the child is experiencing in important life areas (e.g. school, peer relations).  Existing evidence suggests that for many children with ADHD, medication will be an important part of effective symptom management and helping the child experience success. Some children do not respond positively to medication, however, and others are able to succeed quite nicely with treatments other than medication.  For these children - i.e. those whose symptoms are managed effectively over the long-term via other methods - it seems unlikely that the absence of medication in their treatment program will increase their risk for developing an SUD.

In summary, results from this meta-analysis indicate that stimulant medication treatment is associated with reduced risk of SUD. This does not mean, of course, that children taking stimulants will never develop an SUD.  It does indicate, however, that the odds of this occurring are reduced when medication treatment is initiated. For parents who are considering medication treatment for their child, or who have already implemented this intervention, these findings should provide some comfort in regards to this important concern.

Information presented in Attention Research Update is for informational purposes only, and is not a substitute for professional medical advice.  Although newsletter sponsors offer products and services that I believe will be of interest to subscribers, sponsorship of Attention Research Update does not constitute a specific endorsement or guarantee of any company's product or services.