"You sound like Tom Cruise" a child psychiatrist told me after a lecture. I wished he had said that I looked like Tom Cruise, but oh well.... Since then, I've continued to talk to medical audiences, and with my patients, about what I think are some underappreciated risks with amphetamine stimulants. As a new blog writer, I knew I could not avoid writing about this topic too though I hoped to delay the moment. It didn't take long.

In my last blog post, discussing the pharmaceutical industry and psychiatry, I intimated some concerns about the safety of amphetamines and the concept of ADHD. In ensuing comments, readers requested elaboration. I do so with hesitation, because I expect it will be almost impossible to say anything on this topic without offending someone or without being misunderstood. But since the academic career has its heroic obligations (credit: William James), here goes: (Final caution: I'm sure I'll need more posts to present this fully, so for now I'll focus on amphetamines, later on the ADHD concept)

Recent studies indicate that prescription drugs are the most commonly abused agents in the world. Among such agents are amphetamine stimulants. Since amphetamines do not have a medical withdrawal syndrome, nor fatal overdose risk, many physicians feel comfortable prescribing them (mainly for their primary indication of attention deficit hyperactivity disorder (ADHD) in children and adults). Outside of somewhat disreputable lay circles, the wisdom and safety of these developments are little questioned within medical settings. Yet it may be that amphetamine stimulants are being overprescribed and contribute to the epidemic of prescription drug abuse. It may further be the case that this prescription is especially harmful in children and young adults due to long-term neurobiological deterioration, a slow process which may be overlooked in lieu of the absence of short-term safety risks.

In about a year of lecturing to medical audiences on this topic, I am surprised that they are surprised when informed of the rather extensive animal research literature demonstrating long-term neurobiological risks of amphetamine stimulants. Although the extension of animal data (primarily rats) to humans is not always valid, the potential relevance of such data is widely accepted. Especially if exposed in adolescence or early adulthood, young rodents experience a pattern of neurobiological harm that is consistent with other drugs of abuse (like cocaine) and inconsistent with non-controlled substance prescription psychotropic drugs (like antidepressants or lithium or antipsychotics). This pattern includes the following: decreased functioning of dopaminergic pathways in adulthood, decreased hippocampal size with notable atrophy in long-term follow-up, and increased corticosteroid response to stress. (Translation into English: In rats, amphetamines cause neurons to become smaller and sometimes die, making certain parts of the brain that involve memory smaller. Thus, paradoxically, drugs given for cognitive problems can cause cognitive problems.) Lithium, in contrast, increases hippocampal size over time, decreases long-term depressive and anxiety behaviors, and normalizes the exaggerated corticosteroid response to stress seen in animal models of depression. (Translation: Lithium, which most people view negatively, keeps the brain alive longer and prevents neurons from dying. Another paradox: A drug that is seen as toxic in fact may improve cognition over time). There are not, in contrast, safety studies with five years or longer follow-up that demonstrate that such evidence of neurobiological harm is absent in humans given amphetamines. It is also interesting that in rat models, early exposure to amphetamines leads to increased depressive and anxiety behaviors in adulthood; the commonly discussed "comorbidity" of ADHD with mood and anxiety disorders could perhaps be reconsidered as possible stimulant-induced worsening of mood or anxiety symptoms.

That is my concern. (I published some references in a recent review paper in the journal Current Psychiatry, June 2007 issue, here. If I have missed relevant research studies that counter what I describe above, I would be glad to know of them.) Here are some of the printable responses I have heard: 1. "You cannot conclude from rat studies that these drugs hurt humans." True. But we draw such conclusions with many other drugs; in fact, such effects have led to other drugs never making it to the US market (due to rejection by FDA). In contrast, antidepressants have beneficial effects in rat brain studies, which many have touted in support of their use. (Their benefits in primates and humans is much less established). 2. "These effects may exist in animals, but they are far outweighed by the clinical benefits shown in human studies." This may be somewhat true in childhood ADHD, but even those studies are mainly short-term; whether the long-term benefits of amphetamines outweigh their long-term risks has not been shown because these kinds of neurobiological studies have not been conducted in humans with long-term exposure. Further, other non-amphetatamine drugs (like bupropion), which have no animal evidence of neurobiological harm, have been shown to have benefits in childhood ADHD. Lastly, the efficacy literature of amphetamines in adult ADHD, with which the possibility of harm might be outweighed, is much more limited than in children. 3. "Ritalin (methylphenidate) and its cognates (Adderall etc) is not an amphetamine and thus should not have these risks." Wrong. These studies were conducted with methylphenidate and show it to have similar risks to other amphetamines. (Standard pharmacy texts class it as "amphetamine-like").

I do not conclude that every child should come off ritalin, nor that Scientology has it right, nor that Tom Cruise should write a textbook of psychiatry. I do think that our profession has tended to ignore some biological realities. George Orwell once said that truth becomes untruth if uttered by your enemy: we need to stop suspecting all who critique amphetamines or ADHD simply because some may do so irresponsibly. I also think that this whole class of controlled substance has been too "grandfathered": if drugs with these effects in animal studies were proposed today, they would be highly unlikely to make it to the marketplace. Instead, since doctors have been using amphetamines since the 1930s - they are the earliest psychotropic drugs of the modern era (called "mood stabilizers" back then) - we are comfortable with them, despite their weaknesses. In my view, we should be more cautious in using these agents, trying non-drug interventions for ADHD first, and then using amphetamines mainly short-term. Even if amphetamines were as effective as many claim, these long-term safety concerns cannot be dismissed without further study. And for those who would rather keep using them until they are proven harmful, some lessons in medical history may help, showing how that kind of attitude has led to major debacles in the past (the best example is bleeding or leeching, used for about two millennia; more recently, one could cite thalidomide; or even more recently, though with more caveats too, estrogenic hormone replacement therapy). One cannot presume drugs are safe; one should presume they are harmful until their safety is proven. In the meantime, the history of medicine suggests caution as the wisest course.