Bitter Pill

This article first appeared in the January/February 2005 issue.

Nadine Lambert doesn’t look the part of a radical firebrand, bent on undermining the pharmaceutical industry and sticking it to the power elite of the American child-psychiatry establishment. A pleasant, thoughtful, gray-haired academic, she’s been a research psychologist in the department of education at the University of California at Berkeley since 1973, and now heads the department’s School Psychology program. Lambert never planned on becoming the enfant terrible of the world of attention deficit/hyperactivity disorder (AD/HD) treatment. Her work in the field began quietly and methodically in the 1970s, when she became involved in a larger effort to develop special-education services for school-aged children. The original purpose of what became her life’s work was a straightforward prevalence study to determine the relative number of handicapped and/or hyperactive children in San Francisco’s East Bay area. From the mid-1970s to 1990, she regularly published scientific articles describing an ambitious research project that anticipated following these children to adulthood.

Then, in 1998, this dignified, middle-aged professor of educational psychology suddenly became the center of an academic firestorm. At the prestigious National Institutes of Health Consensus Conference on AD/HD, she announced the results of a study suggesting that use of Ritalin, one of the most routinely prescribed drugs for children in America, might contribute to later drug abuse. Her study of nearly 400 children with AD/HD showed that by the time children treated with Ritalin reached their mid-twenties, they had double the rates of cocaine abuse and cigarette smoking as young adults who hadn’t taken it in childhood.

If Lambert had lobbed a grenade into her audience of child-psychiatry researchers, her report couldn’t have been more explosive. For proponents of Ritalin and other psychiatric drugs for children–most of the attendees at the meeting–Lambert’s findings had nightmarish implications. What if outraged parents now began to flood their offices, demanding to have their children taken off these medications and given alternative treatment? Perhaps even worse, if Lambert was right about Ritalin’s dark potential, it meant that the child-psychiatry establishment was wrong. Seriously wrong. Wrong all along. For the AD/HD researchers who’d long put their faith in Ritalin–even bet their careers on it–this was truly scary stuff.

Their fears were well-founded. The media picked up on the story and loudly questioned, as it had several times before, whether Ritalin was being overprescribed for children with AD/HD. But the tight-knit community of child-psychiatry researchers and academics wasn’t about to roll over in defeat. Instead, within a year, child psychiatrist Joseph Biederman of Harvard Medical School, one of the country’s most influential AD/HD researchers and a leading advocate for Ritalin use, announced the findings of a rebuttal study. His new research showed that kids prescribed Ritalin actually had an 85 percent lower rate of substance abuse four years later. The study further claimed that the medicated children were actually “protected” from engaging in later drug abuse.

Two subsequent studies by leading researchers Timothy Wilens, a long-time member of the Biederman team, and Russell Barkley, a professor of psychology at the University of South Carolina in Charleston and the world’s leading theorist and arbiter of AD/HD, also took aim at the Lambert study. These studies differed in design from Lambert’s and, as we’ll see, also had numerous limitations. Nevertheless, the media obligingly switched course, now extensively touting the “protective benefits” of Ritalin. The pharmaceutical companies that manufacture Ritalin and similar stimulants reprinted the Biederman and Wilens papers and sent them to every pediatrician and child psychiatrist in the country.

Since this imbroglio, the National Institute of Drug Abuse (NIDA) has refused funding to Nadine Lambert, either to conduct further follow-up on her subjects or to analyze her current data more closely. Recently, a government official privately told her that her latest grant proposals have been harshly criticized during the peer-review process. Since the publication of her original study results, no major U.S. journal has been willing to publish either her new research or a fresh analysis of old research to rebut her critics. She’s been dismissed in print by at least one prominent critic as an “outlier on the far end of a bell-shaped curve”–polite academese for a loose cannon, whose research is so fringy and unsubstantial that it can safely be ignored. But is it more appropriate to see her as the canary in the coal mine, sounding a warning about the unacknowledged dangers of one of the most widely prescribed drugs for children in America?

Do All Children Have AD/HD?

No one knows precisely how many children in the United States have AD/HD. As with all psychiatric conditions, there’s no definitive test for AD/HD–no blood test or brain scan or even standardized psychological assessment that can unequivocally determine whether a particular child is or isn’t affected. Much of the controversy roiling around AD/HD stems from the problem of diagnosis. In its milder forms, the disorder’s symptoms of inattention, distractibility, impulsivity, and hyperactivity can look very much like the normal behavior of an active child.

Whatever the problems in diagnosis, Ritalin, along with newer sibling stim ulants, have become the overwhelm ing treatment of choice for children who are “diagnosed” with AD/HD. Currently, nearly one in ten 11-year-old American boys takes some kind of stimulant medication–a classification that included Dexedrine or amphetamine and now includes drugs like Ritalin, Adderall, and Concerta. All told, the United States uses 80 percent of the world’s stimulants. Critics of this trend point out that no other country addresses the behavior and school performance of children with such strong emphasis on psychiatric diagnosis and drug treatment.

But why not use these stimulants, when they seem to work wonders? This is the standard response of Ritalin’s champions, who include legions of grateful parents and teachers across the country. Often, within minutes of taking the first dose, hyperactive Johnny metamorphoses into a different child–steady, focused, and compliant. Skills such as handwriting and following directions improve instantly. Indeed, thousands of studies, mostly of school-aged boys, prove the effectiveness of stimulants in helping children perform better at school and at home.

Ever since the first published report on the use of stimulants by kids in 1937, medical pundits have argued that drugs such as Ritalin operate differently in the brains of hyperactive kids as compared to normal kids–because the AD/HD brain is different. What else could explain how a stimulant could actually calm a child, rather than rev him up even more? It’s easy to understand why the assertion that Ritalin calms only hyperactive kids carries so much weight in popular discourse.

But this claim is a myth. In fact, Ritalin and other stimulants improve anybody’s ability to focus and pay attention to boring and difficult tasks. Studies from the National Institute of Mental Health in the 1970s proved that low-dose stimulants have the same effect on all children (and adults as well), whether or not they’ve been diagnosed with AD/HD. The hyperactive child who slows down on Ritalin appears calmer to parents and teachers, demonstrating that the Ritalin is working. Because the drug works so well, they conclude that the child must have AD/HD. But given Ritalin’s universal effects on all children, by this logic it would follow that all children have AD/HD!

Understanding Ritalin’s Effects

Ritalin may help everyone focus, but does it truly calm everyone? In fact, the “calming” effect of the drug on hyperactive kids is actually the result of their becoming more methodical in their performance, which, in turn, moderates their activity level. Indeed, normally active children also become less active when they’re given the same low doses of stimulants as AD/HD kids get, though the activity slowdown in normal kids is less dramatic. Similarly, when AD/HD children take higher doses of stimulant medications, they react just like other children: both groups become overactive and agitated. Both normal kids and AD/HD kids complain that they feel “nervous” or “weird” on higher doses of stimulants.

Still, Ritalin undeniably helps hyperactive children. Surely, this pharmaceutical leg-up gives these youngsters a much better chance at achieving success as they grow into adolescence and adulthood. But, here’s the rub. After nearly seven decades of prescribing Ritalin, Adderall, and similar stimulants for millions of children, we still don’t know whether these drugs boost kids’ chances for success later in life. Given the critical importance of this issue, why are we still so steeped in ignorance?

Part of the reason lies in the daunting challenge of systematically tracking  children for a decade or more. In addition, since stimulants were introduced as a treatment for hyperactivity nearly seventy years ago, diagnostic categories for hyperactivity have changed dramatically. So it isn’t clear that earlier studies of the impact of stimulants on hyperactivity examined the same population that today’s researchers would be observing. Furthermore, studies done in the ’60s and ’70s were “naturalistic”–the children were separated into treated and control groups, based upon their families’ choice of treatments–and not randomized, so they’re vulnerable to being challenged by today’s more stringent criteria.

Finally, while virtually all children taking stimulants in the ’70s and ’80s stopped taking them at around age 13, when hyperactivity tends to fade on its own, current standards of care mandate that many adolescents continue to take these meds through high school, and even into college. Indeed, the only new long-term-prospective randomized study in the United States lasted just 14 months, because doctors felt ethically bound not to deny AD/HD children access to stimulants required by the American standard of care. So the disappointing outcomes of earlier studies are easily dismissed by medication proponents, who maintain that outmoded definitions of AD/HD, changing standards of care, and lack of randomization render the older studies useless.

Still, a handful of AD/HD treatment studies were published in the late 1980s and early 1990s, tracking children prospectively from school age into young adulthood. These studies also showed disappointing results for Ritalin: children treated with stimulants did no better (in fact, they did slightly worse) than AD/HD kids who hadn’t taken Ritalin. In this older research, all three groups of AD/HD kids studied–those getting only medication, those getting only family counseling and special education, and those getting a combined treatment of drug and nondrug interventions–did far worse than normal kids, as gauged by rates of high-school graduation, delinquency, and drug use. The most rigorous study, done by psychiatrist James Satterfield and his wife Breena Satterfield, a social worker, concluded that the medication-only children did the poorest and the combined-treatment group did the best at 10- and 15-year follow-ups.

This isn’t to say that stimulants do no good whatsoever. Based on my 27 years of practice, I personally suspect that stimulants can help a kid get through the most difficult years when he’s in mandatory schooling. But I believe that long-term outcomes are far more affected by learning problems and emotional problems that have roots in family and community factors.

Researchers and practitioners alike have long been concerned that Ritalin use in childhood could lead to later drug abuse. Lambert didn’t invent this issue to annoy the psychiatric cognoscenti or frighten the general public. Ritalin is a very powerful agent, classified by the U.S. Drug Enforcement Administration as a Schedule II drug, the strictest category of potentially abusable drugs that doctors are legally allowed to prescribe. Ritalin is a stimulant that’s similar in molecular structure to “speed” (amphetamine), “crank” (methamphetamine) and crack cocaine–all drugs with devastating addictive potential. Laboratory animals, when given the choice of pressing a lever that sends a pellet of food into the cage or one that delivers a methylphenidate (Ritalin and Concerta) or amphetamine (Adderall and Dexedrine), quickly learn to choose the drug lever, ultimately starving to death and exhibiting bizarre behavior along the way.

Studies in normal adult volunteers taking one oral dose of Ritalin show very early signs of addiction when given a second dose of the drug several weeks later. Measures of eye-blink and heart rates–subtle markers of addiction–rise in adult humans who take the second dose of medication. As disturbing as these studies are, they haven’t generated similar research on kids who take stimulants. No one had quite so clearly and bluntly linked this basic science of stimulant addiction to the real world of hyperactive kids and teens until Lambert announced her findings at the 1998 NIH Consensus Conference on AD/HD.

Why, then, has Lambert’s study been so roundly dismissed? Her critics contend that her work is simply not up to snuff scientifically, pointing to her study’s lack of randomization and the questionable validity of her control groups. Even though her naturalistic approach was state of the art for that time, there’s no question that Lambert’s study subjects weren’t randomized into drug and nondrug treatment groups. Thus, despite her prodigious efforts to document the similarities between the two groups, it’s quite possible that the sicker and more severely affected kids received Ritalin, while children with milder AD/HD didn’t. If so, it’s logical to conclude that the kids who received the drug might have gone on to have drug-abuse and other problems later in life simply because they were more troubled to begin with. This well-known study effect, called the “severity bias,” could have tilted the scales toward negative outcomes for the drug-treated group. This failure to randomize has also plagued other older studies showing poorer outcomes for young people who took Ritalin as kids.

Lambert’s findings have also been criticized for failing to take into account another childhood problem–conduct disorder. This disorder involves a pattern of delinquent behavior such as stealing, cheating, vandalism, physical violence, truancy, trouble with the law, and cruelty to animals. In his published critique, Russell Barkley suggests that the kids who took Ritalin in Lambert’s study and showed higher levels of substance abuse later on actually suffered from conduct disorder in addition to hyperactivity. In other words, the medicated kids abused drugs later on not because the Ritalin had sensitized them to other drugs, but because their conduct disorder (which made it more likely that Ritalin would be prescribed for them in the first place) made them more susceptible to substance abuse. Lambert retorts that Barkley’s reasoning is circular in that kids who abuse drugs invariably also have conduct disorder, regardless of what originally contributed to their drug use.

While Lambert’s study is certainly imperfect, what’s been ignored in most discussions of her work is that the studies cited as rebuttals to her conclusions are susceptible to the same kinds of criticisms. Both Biederman’s study and Wilens’s metanalysis were retrospective investigations that were also marred by nonrandomization and inadequate controls. Moreover, Biederman’s conclusions were based on subjects still in their teens who were still taking medication, while Lambert tracked her subjects’ drug use into their mid-twenties. Even Barkley’s study wasn’t randomized.

In fact, a dirty little secret of psychology research is that most studies are plagued with methodological difficulties, because it’s still notoriously difficult to institute unimpeachable controls. “The field is a quagmire,” observes longtime AD/HD research psychologist Stephen Hinshaw, also of the University of California at Berkeley. “No study can be absolutely free of the possibility of bias or some other unknown variable in the assignment of controls.”

So if the studies that are intended to put the definitive kibosh on Lambert’s work are similarly flawed, why can’t she get the support she needs to conduct further research on the issues raised? After all, her research was considered vital enough to be funded for 20 years by the NIDA and the California Tobacco-Related Disease Research Program. In addition, one would think that her conclusions are dramatic and disquieting enough to prompt other researchers to reexamine them, utilizing more careful controls and randomization procedures. Why is this unlikely to happen, at least in the United States?

The answer, I believe, has much to do with the politics of American psychiatry and the influence of the multibillion-dollar psychopharmacology industry on scientific debates within the field. Joseph Biederman and his colleagues at the Pediatric Psychopharmacology Clinic of the Massachusetts General Hospital, along with Russell Barkley and other psychiatric specialists in AD/HD, are at the center of a seismic shift in the direction of American psychiatry. During the last 20 years, the field has largely reversed course in its thinking about children’s behavioral problems, from assigning causation to the child’s family and environment to assigning causation to the child’s presumably malfunctioning brain. Certainly, once AD/HD children begin taking stimulants, the immediate and often global improvement in their behavior reinforces this brain-based line of reasoning, though it be logically flawed. Aspirin improves a headache, but no one says that a headache is caused by an “aspirin-deficiency.”

This erroneous belief in biological causality is hugely profitable for the pharmaceutical industry, which now earns more than a billion dollars a year from sales of Ritalin, Adderall, and other AD/HD drugs. The runaway popularity of stimulants depends, in turn, upon the benediction they receive from leading academic experts, who receive major research funding from more than a half dozen of the world’s largest pharmaceutical companies. These days, virtually all leading psychiatry researchers are biologically oriented scientists, who accept funding from pharmaceutical companies. Indeed, drug companies now supply 60 percent of all funding for biomedical research, and that percentage is undoubtedly much higher in the field of psychiatry.

Still, many doctors, both in and out of academic medicine, are troubled by the depth of the clinical research/drug industry connection. In 2001, Marcia Angell, then editor of the New England Journal of Medicine, wrote a stinging editorial entitled “Is Academic Medicine for Sale?” Charging that academic medical institutions are “growing increasingly beholden to industry,” she pointed to the potentially corrupting ties binding medical research institutions to drug companies. “There is now considerable evidence that researchers with ties to drug companies are indeed more likely to report results that are favorable to the products of those companies than researchers without such ties,” she wrote. “When the boundaries between industry and academic medicine become as blurred as they now are, the business goals of industry influence the mission of the medical schools in multiple ways.”

From Therapy to Psychopharmacology

The truth is that we don’t yet know whether Ritalin use makes kids more likely to abuse drugs later on, whether it protects them from later substance abuse, or whether it does neither. What we have are two sets of data, both significantly flawed. One set is accorded great weight and significance by the scientific powers-that-be, while the other is energetically trashed. The data in line with the most powerful economic interests and most influential academic voices are extolled as good science–even “truth”–while the other data are consigned to the dust heap of scientific irrelevance and the author placed under the research equivalent of house arrest.

We may never learn for certain whether Ritalin contributes to, or protects against, later drug abuse–at least not in any study from the United States. Ritalin, Adderall, and similar drugs have become such accepted treatments for AD/HD that it’s now considered unethical to withhold them in any randomized, double-blind trial. “Treatment with stimulants has become the standard of care in this country,” says University of California at Berkeley psychologist Hinshaw. “Running a study that withheld stimulants for a long time–the 10 or 15 years needed to check for later drug abuse–would be medically unethical in the United States.”

But in Western Europe, where doctors don’t assign an AD/HD diagnosis to as wide a swath of behaviors as do American physicians and are less apt to use stimulants to treat the disorder, there may still be opportunities to study the question. “The Germans are doing such a study right now,” says Hinshaw, “but it’ll take years, or we may never know.”

Meanwhile, Lambert, who’s now spent nearly 30 years of her life following and analyzing her group of AD/HD kids, tries to stay focused and productive. She continues to seek funding sources, but from philanthropic organizations rather than institutional medicine. She also continues to teach at Berkeley and to write rebuttals to the criticisms of her work, hoping to find a larger forum. “I’m just trying to get answers–it’s what keeps me going,” she says. But without funds to pursue her long-term AD/HD project, she doubts that she can continue the work. “I’m now aiming for a foreign journal like the Canadian Journal of Psychiatry, ” she says. “I doubt that I can get a fair review in this country.”

Even though I think Lambert’s study has real merit and should be extended, I personally believe that neither the biologically sensitizing nor the protective potential of Ritalin is as important an influence on future drug abuse as environmental factors. Having prescribed Ritalin for 25 years–and having seen its short-term benefits a thousand times–my own observation suggests that a child’s family and neighborhood are more important than the use of any medication in predicting whether a child will later abuse drugs. If I believed that Ritalin had a strong sensitizing impact on drug abuse, I’d never have prescribed it to most of the children I’ve treated with the drug.

Ironically though, the focus on Ritalin as hero or villain suggests just how far the debate has shifted over the years from the influence of the external environment on kids’ lives to the internal environment of their individual brain chemistry. As a field, our attention has become increasingly diverted from the roles of family conflict, community breakdown, and poverty as the overriding factors leading to substance abuse in young people. As passions for and against drug interventions grow, they tend to drown out discussion of nonpharmaceutical interventions for children–family therapy, behavioral interventions, school-based and community-based programs–that have a proven record of effectiveness in decreasing drug abuse. These days, it isn’t just the power and money of Big Pharma setting the agenda: it’s also our own professional culture, which seems more and more enthralled by the biochemical fix.