An association between ADHD treatment and later substance use cannot, by itself, show that treatment caused or prevented that outcome. ADHD-related characteristics, co-occurring conditions, the age treatment began and the medication actually used all affect the comparison. A historical 2009 review illustrates why the phrase “stimulant treatment” can conceal substantial differences in evidence.
Berman and colleagues reviewed potential adverse effects of amphetamine treatment on brain and behavior. Their discussion of later substance use considered human observational follow-up, surveys and animal studies. It was not a randomized long-term trial assigning children to amphetamine or no treatment and then measuring adult addiction. Its account should be read as evidence available at that time, not current prescribing guidance.
What was the underlying causal question?
The question was whether exposure to a prescribed stimulant changed later substance-use risk beyond the risk associated with ADHD and related characteristics. A comparison between treated people with ADHD and people without ADHD does not isolate that treatment effect: diagnosis and treatment differ together.
Comparison with untreated people who also have ADHD addresses a closer question, but treatment may still be associated with severity, co-occurring conduct problems, access to care or family circumstances. These are possible alternative explanations requiring measurement and analysis, rather than assumptions that automatically invalidate or resolve the association.
Did all follow-up studies point in one direction?
No. The review described reports suggesting increased later risk, reports suggesting no change and reports suggesting protection. It stated that most studies it discussed, including a meta-analysis, found no effect on subsequent substance abuse or a lower risk. That historical summary does not supply a single causal estimate specific to amphetamine.
One prospective study summarized by the authors followed 112 White male participants with ADHD, initially aged 6–17 years, for ten years. Mean age at follow-up was approximately 22. The review reports no increase or decrease in the frequency of substance-use disorders associated with stimulant treatment, compared with ADHD participants not treated with stimulants.
It also reports an association with 1.6 additional years of alcohol abuse among the subgroup with alcohol abuse, p = 0.04. A subgroup finding about duration and an overall finding about the frequency of disorders answer different questions. Neither should be deleted to make the study appear uniformly protective, harmful or null.
Was the other prominent follow-up study an amphetamine study?
The review describes a different cohort of 176 methylphenidate-treated boys with ADHD, without conduct disorder, and 178 comparison participants without ADHD. Reassessment occurred in late adolescence and early adulthood, with mean ages of 18.4 and 25.3 years. Methylphenidate is a different drug from amphetamine, even though both are prescription stimulants.
The reported lifetime substance-abuse-disorder rates were 44% among ADHD participants who initiated treatment after age seven, 27% among those who initiated before age eight and 29% in the non-ADHD comparison group. These are percentages for different groups within that study, described through the review. They are not the risks attributable to amphetamine or predictions for an individual child.
The age-at-initiation association is scientifically relevant, but the treatment was not randomly started at alternative ages. Timing can be linked to other features of the child, disorder or care. Moreover, the review notes that 98% of this sample began treatment by age 11. This dataset therefore offered little basis for evaluating initiation during college or later adolescence.
Why do medication categories matter?
A review about amphetamine may discuss methylphenidate studies, mixed stimulant histories and animal exposure experiments to explore related questions. That broader context should not be mistaken for an amphetamine-only evidence base.
| Evidence type | Useful question | Boundary |
|---|---|---|
| ADHD participants with and without a treatment history | How later outcomes differed in the observed groups | Treatment selection may differ between groups |
| Methylphenidate-treated cohort | Associations involving that medication and treatment timing | Does not establish an amphetamine-specific effect |
| Non-ADHD comparison group | Outcome differences relative to people without ADHD | Diagnosis and treatment are not independently isolated |
| Animal exposure experiment | Behavioral or biological responses under an assigned regimen | Does not directly measure adult human substance-use disorders |
Do animal sensitization findings settle later human addiction risk?
The review discusses developmental differences, sensitization and later responses to drugs in animals. Those experiments can investigate mechanisms under controlled exposure conditions. Sensitization, however, describes an increased response after prior exposure; it is not interchangeable with a human diagnosis of a substance-use disorder.
The authors emphasize that exposure pattern and developmental stage can change animal findings. A result from one species, age or regimen cannot resolve the clinical association simply because the administered compound belongs to the stimulant class. Conversely, an observational human association does not identify the neural mechanism demonstrated in an animal experiment.
What can this evidence support in a client explanation?
It supports asking a more precise question than whether “stimulants lead to addiction.” Identify the medication, age at initiation, diagnosis, comparison group, follow-up interval and outcome definition. Distinguish lifetime disorder prevalence from duration among affected participants, and prescription exposure from later nonmedical use.
The review’s mixed historical findings do not justify treating a later disorder as proof that a prescription caused it. They also do not justify promising that a particular person’s treatment prevented later misuse. A case-specific causal opinion requires more than a medication history followed by an outcome.
This commentary addresses interpretation of historical research. It does not recommend starting, stopping or changing medication, and it does not establish present treatment standards.
Scientific source
Berman SM, Kuczenski R, McCracken JT, London ED. Potential adverse effects of amphetamine treatment on brain and behavior: a review. Molecular Psychiatry. 2009;14(2):123–142; published online August 12, 2008. doi:10.1038/mp.2008.90. Full-text author manuscript. Human cohort results are described through this supplied narrative review; the underlying datasets were not independently reanalyzed. The journal also published a 2010 correction to a separate psychosis-study passage, not the follow-up findings discussed here.
Leave a Reply