Editor’s Note: Over the next several months, Mad in America is publishing a serialized version of Peter Gøtzsche’s book, Critical Psychiatry Textbook. In this blog, he discusses the results of the MTA study on ADHD drugs and the misleading statements textbooks make about ADHD treatment. Each Monday, a new section of the book is published, and all chapters are archived here.
The large MTA trial and institutional corruption
A textbook recommended psychoeducation for adult ADHD,17:617 and the book about child and adolescent psychiatry mentioned non-pharmacological treatment first and pointed out that there can be a vicious circle where the parents scold the child a lot.19:114
This little glimpse of non-pharmacological hope vanished already on the next two pages, which said that drugs should be used in children above 6 years of age with moderately severe and severe ADHD, and that they often have quick and pretty dramatic effects. The authors noted that the effect was well documented up to 12-18 months; that the long-term effect was insufficiently elucidated; and that newer register studies suggested a reduction in number of accidents and emergency visits and a positive long-term effect on learning, marks, and schooling.19:116
It is misleading to mention highly positive effects up to 12-18 months; to pretend we do not know what happens after this; and then say that register studies, which are less reliable than randomised trials, suggest the drugs have important long-term effects. The authors presented 19 references but none of them was to the short-term results of a large study, which child and adolescent psychiatrists otherwise cite a lot.
This is the famous MTA trial sponsored by the US National Institute of Mental Health,7:148 in which 579 children were randomised to methylphenidate, behavioural therapy, both, or routine community care. The NIMH published the 14-month results in 1999.516
Many scales and outcomes were used, with no less than 19 primary outcomes, which is extraordinarily many, but the only differences between drug and behavioural therapy were that the children were less hyperactive or impulsive and paid more attention when on the drug. Combined treatment was no better than drug alone for core ADHD symptoms.
The authors considered ADHD a chronic disorder and advocated ongoing treatment, which agreed poorly with the improvement in symptoms, which, in all four groups was sometimes much larger than the differences between the treatments, e.g. for inattention and social skills.
More importantly, did the reported differences in scores translate into anything useful for the children? They didn’t, as judged by the long-term results, which the psychiatrists weren’t eager to publish. It took another eight years before the three-year results were published.517
This time, the investigators revealed their financial conflicts of interest, which were extreme. Sixteen authors listed a total of 214 drug companies, 13 per author. These relationships were mostly described as research funding, “unrestricted grants” (a euphemism for corruption6), consulting, and being on speakers’ bureaus and advisory boards. Not exactly a group of people that would be likely to give us an unbiased view of the value of methylphenidate.
After three years, the four treatment groups didn’t differ significantly for any of the numerous efficacy outcomes. The investigators partly ascribed this to the fact that many children in the two non-drug groups took drugs, diluting the treatment contrast. But they also mentioned that the results were possibly due to an age-related decline in ADHD symptoms, thereby contradicting their claim that ADHD is a chronic disorder.
This was cognitive dissonance. Most people experience this from time to time, but for many leading psychiatrists, it seems to be a chronic disorder.
A companion paper was close to impossible to interpret, as the findings were drowned in advanced statistics, but the limited relevant data the authors presented showed a lower rate of substance abuse in the behaviour therapy group than in the drug group.518 Methylphenidate didn’t protect against delinquency and substance abuse; if anything, it caused them.
A priori, one would expect stimulants to increase these risks. But a very large and long-term study about this was never published. The main investigator, Nadine Lambert, died in a car accident in 2006519 and her colleagues did not ensure that her unique research got published. Perhaps because they disliked the results.
There is an account of her study in a news release from the University of California, Berkeley.520 She presented her report to the NIH for the 1998 consensus meeting where Vonnegut couldn’t explain what ADHD is, which was the basis for the University’s news release.
Lambert conducted a 26-year study of 492 children, half of whom were diagnosed with ADHD. While nearly half of the youngsters treated with methylphenidate had become regular smokers by age 17, only 30% of those who had never been treated smoked daily. Only 2% of those who had never smoked or taken methylphenidate were dependent on cocaine as adults, compared to 40% of those who both smoked and were treated with stimulants. We cannot know to which degree confounding might explain her results, but I mention them because it is one of the biggest prospective studies ever made of this important issue. After she had presented her results in 1998, the National Institute on Drug Abuse stopped funding her work.5:306
The results in the MTA study after six and eight years were also discouraging.521 The treatment groups didn’t differ significantly for grades earned in school, arrests, psychiatric hospitalisations, or other clinically relevant outcomes. Medication use decreased by 62% after the 14-month controlled trial but adjusting for this didn’t change the results.
The follow-up papers are also difficult to grasp, as they confuse readers with unnecessarily complicated statistics, which looks like deliberate obfuscation, as it would have been much easier to simply describe the disappointing results. One of the investigators was honest, not in any of the over 100 scientific papers that the MTA study generated, but in a newspaper interview:522
“I think that we exaggerated the beneficial impact of medication … The children had a substantial decrease in their rate of growth … there were no beneficial effects – none … that information should be made very clear to parents.”
It wasn’t. It became clear in another newspaper article that the public was duped, seduced and lied to.523 A news release issued by NIMH presented the negative results in a favourable light with the title, “Improvement Following ADHD Treatment Sustained in Most Children.” Free fantasy.
It was not possible to see any difference to drug company propaganda. One of the authors on the payroll of many drug companies, Peter Jensen, said, “We were struck by the remarkable improvement in symptoms and functioning across all treatment groups.” And rather than saying that the growth of children on medication was stunted, the press release said that children who were not on medication “grew somewhat larger.”
The drug industry deceives people in the same way. When Merck found out that its arthritis drug Vioxx was deadly and caused more thromboses than another arthritis drug, naproxen, the company invented the hoax that naproxen was protective rather than Vioxx being harmful, which nonsense the New England Journal of Medicine allowed Merck, a US company, to publish.6:156
The stunting of growth methylphenidate caused was huge. After 16 years, those who consistently took their pills were 5 cm shorter than those who took very little, and there were many other harms.524 We can only speculate which permanent effects these drugs might have on the children’s developing brains but it seems likely—based on what we know about other brain active substances11—that they harm the brain.
The short-term effect is that the children sit still in class, but that effect disappears quite quickly. Short-term harms include tics, twitches, and behaviours consistent with obsessive-compulsive symptoms,506 all of which can become common.4,525 Stimulants reduce overall spontaneous mental and behavioural activity, including social interest, which leads to apathy or indifference, and many children—more than half in some studies—develop depression and compulsive, meaningless behaviors.135,526 Mental activity and interaction with other people are important for brain development, so this also suggests that the drugs may harm the brain permanently.
Animal studies have confirmed these findings,526 and my research group has documented other harms, e.g. that the drugs impair reproduction even after the animals were taken off them.527
At school, the compulsive behaviour is often misinterpreted as an improvement even though the child may just obsessively copy everything shown on the board without learning anything. The drugs used for ADHD have hallucinogenic properties,506 and some children develop mania or other psychoses.135,528 The harms of the drugs are often mistaken for a worsening of the social construct called a “neurodevelopmental disease,” which leads to additional diagnoses, e.g. depression, obsessive-compulsive disorder or bipolar—and additional drugs, leading to chronicity.526
Patients and their relatives commonly refer to depression pills as “Psychiatry’s Starter Kit.” This is because many people start their psychiatric “careers” by consulting their family doctor with some problem many of us have from time to time and leave the doctor’s office with a prescription for a depression pill, which starts a chronic course with multiple diagnoses and multiple drugs. ADHD drugs are also one of Psychiatry’s Starter Kits.
Misleading textbook information and advice
One textbook advised to continue with drugs for years, namely for as long as there is clinical effect and harms are tolerated.19:118 However, it is impossible to judge if there is any benefit in the individual case. We can only say what the randomised trials have shown and they do not lend support to treating people for years, particularly not when the harms are also taken into consideration. This section had 11 references, none of which were elucidating, and not a single one was to the MTA trial, although this was a textbook in child and adolescent psychiatry from 2019, and the 16-year results from the MTA trial were published in 2017.524
A section in another book, written by two psychologists, advised that the treatment should be broad, flexible, and long-term and should start with a series of non-pharmacological methods.20:472 But they also bought into the misleading psychiatric narrative. They claimed on the next page that drugs effectively reduce symptoms of impulsivity, hyperactivity and inattention; improve social interaction with children of the same age and with parents; may alleviate aggression; have a moderate to large effect in children aged 6-18 years; and appear to reduce the risk of subsequent drug abuse.
The authors also claimed that drugs will improve symptoms significantly in adults. They cited the MTA study but only for use in a figure about co-morbidity with overlapping Venn circles. There wasn’t a single word about the results of this trial in the whole chapter even though the book was published in 2021. The MTA paper in their literature list was 20 years old and only reported the misleading 14-month results.529
The authors provided three other references.
One was a meta-analysis of 28 placebo-controlled trials of stimulants in children with ADHD, all of which were “published in a peer-reviewed scientific journal.”530 This provides no comfort, as the peer-reviewed literature is full of unreliable research. The authors used a home-made quality score for assessing the quality of the trials, which is a method firmly recommended against.481 They gave two points to double-blind studies and one point to single-blind studies, but single-blind studies should have been excluded, and nowhere in the paper did they say which studies were single-blind. The trials were very small, with an average of only 28 patients. The authors reported a huge effect on aggression, effect size 0.84.
This looks impressive, but: How is it possible to find such a result when it is widely known that stimulants cause aggression; when the FDA warns that “anxiety, psychosis, hostility, aggression, and suicidal or homicidal ideation have also been observed;”34 when FDA trial data show that ADHD drugs cause psychosis or mania in 2-5% of people treated for one year, whereas no such cases were reported for patients on placebo;261 and when these drugs—including atomoxetine—cause hallucinations and violence?261,401
The answer is simple: Meta-analyses of published placebo-controlled trials of psychiatric drugs are highly unreliable. As already noted, it seems that all trials in ADHD are at high risk of bias and the degree of underreporting harms like aggression is pronounced in psychiatric drug trials.
The second reference was to a 2017 editorial commenting on a study of health care claims from 3 million adolescent and adult patients diagnosed with ADHD.531 The editorialist noted that the use of stimulants could sensitise people to the rewarding effects of drugs, increasing the risk of substance use disorders, and that the sensitising effects of amphetamine are well established in animal studies. On the other hand, he also noted that, by reducing the symptoms and impairments of ADHD, stimulant medications may decrease the risk for substance use disorder.
The study found a reduction of events related to substance use disorders, such as emergency department visits, during periods of treatment with ADHD medications whereas the use of SSRIs increased such events. The patients were their own controls, but they could be more motivated to reduce substance use during periods in which they were engaged with medical treatment.
The editorialist noted that a 2003 meta-analysis found a 1.9-fold reduction in risk for substance use disorders in the treated group. He didn’t provide a reference to this meta-analysis, and I was unable to find it even though I tried many search strategies on PubMed and browsed hundreds of records. However, I doubt this meta-analysis can be important because the MTA trial of 579 children and adolescents found the opposite, a higher rate of substance abuse in the drug group than in the behaviour therapy group.518 Furthermore, the MTA trial did not find that the drug reduced arrests, psychiatric hospitalisations, or other clinically relevant outcomes.521
It is therefore difficult to believe the results of research on registry data from Sweden and Denmark that reported 30-50% reductions in criminal convictions, which the editorialist mentioned.
When searching on PubMed, I found a systematic review from 2021 that appeared to have been carefully conducted.532 The authors concluded that, based on the limited evidence available, strong clinical recommendations are not justified, but that provisionally, stimulant treatment in children with ADHD “may prevent” the development of substance use disorders. “More studies are needed.” I am not so sure about this. It seems likely that ADHD drugs increase substance abuse, and, at any rate, these drugs should not be used because they are harmful.
The editorialist had numerous financial conflicts of interest and had been a speaker at drug company sponsored events. An editor wrote in the paper that he had “found no evidence of influence from these relationships.” It is funny how people always try to get away from the indisputable fact: Financial conflicts of interest corrupt, which is why the drug industry buys doctors.6,7,533
The third reference was a meta-analysis of adults from 2010.534 The authors acknowledged that all meta-analyses are limited by the quality of the studies they included, but they did not provide any assessment of the risk of bias in the individual studies. They included 7 placebo-controlled trials of short-acting stimulants (459 patients) and 5 trials of extended-release preparations (637 patients) and reported huge effect sizes, 0.96 and 0.73, respectively. They translated these effects, measured on scales, to binary data and reported that the number needed to treat to benefit one patient was only about 2-3.
As already explained, these miraculous results are fake and there cannot exist any number needed to treat for psychiatric drugs, only number needed to harm (see Chapter 8, Part Three). Furthermore, one of the world’s finest biostatisticians, Douglas Altman, who was statistical advisor for the BMJ for many years, has advised against dichotomising continuous variables.535
The first author of this meta-analysis had received consulting fees or research support from or had been on advisory boards or a speaker for companies selling ADHD drugs. These were called “potential conflicts of interest,” which is a misnomer often used to downplay the problems. Conflicts of interest cannot be potential; they are real.
About ADHD in adults, child and adolescent psychiatrist and professor Søren Dalsgaard advised in a textbook that psychoeducation should be one of the first things offered; that there is good evidence for the effect of cognitive behavioural therapy, especially when combined with drugs; and that the combination is clearly better than drugs alone.16:473
This information is strange. It starts with psychoeducation, goes on with psychotherapy, and ends by saying that the combination is better than drugs alone. Since Dalsgaard prefers psychological interventions, he should have told us if the combination is better than psychotherapy alone. Perhaps drugs are not needed?
This text is an example that psychiatrists are totally absorbed in the drug-focused paradigm. In 2015, I participated in a panel at a conference with hundreds of patients in the audience. After I had advocated for psychotherapy instead of drugs, also for patients with schizophrenia, the psychiatrist on the panel, Merete Nordentoft, remarked that drugs could not always stand alone. I turned the argument around and said that everyone should get psychotherapy and that this could not always stand alone. The audience applauded my remark. Few patients get the psychotherapy they so much want and need and many hate psychosis pills but are forced to take them.
About side effects—which was the term always used in the textbooks that never spoke of harms—Dalsgaard mentioned that, in high doses, the drugs may trigger or aggravate depressive and psychotic symptoms if the patient is predisposed toward a psychotic disorder.16:475
This is wrong. Dalsgaard provided a thinly veiled attempt at blaming the victim and not the drug, which permeates the whole of psychiatry. Depressive and psychotic symptoms may occur on usual doses and without any predisposition.
Dalsgaard’s claims were not referenced, but there was a list with 19 references of which only two were research papers related to his claims. Both were his own publications; they were observational studies; and they were not illuminating.
One article noted that 47% of children with ADHD had criminal convictions in adulthood, five times more than in the general population.536 What should we make out of that? We all know that children with this diagnosis have more problems in life than other people, but we cannot help them avoid crimes with drugs.
The other article included 208 youths with ADHD.537 The risk of substance use disorder in adulthood was 8 times higher than in the background population, and for every year older at start of treatment, the risk increased by a factor of 1.5. This suggests that kids should be medicated from birth if only doctors could identify symptoms of ADHD that early. It also means that the risk of drug abuse is 130 times higher (1.512) if a child starts treatment at age 18 rather than at age 6.
It can be calculated from the article that the background rate in the population is 0.69%. This means that 0.69% x 130 = 90% of all children with an ADHD diagnosis from age 6 will become substance abusers if they are not treated before age 18. The article did not specify the age span that provided the data for the 1.5 times annual risk increase, and I might have extrapolated too liberally by using a span of 12 years, but the study is absurd. There must have been huge confounding because children starting drug treatment late are very different to other children.
Dalsgaard did not mention the MTA trial in his book chapter, which is the best evidence we have about crime and drug abuse when children with the ADHD diagnosis grow up.
To see the list of all references cited, click here.
Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.
Mad in America has made some changes to the commenting process. You no longer need to login or create an account on our site to comment. The only information needed is your name, email and comment text. Comments made with an account prior to this change will remain visible on the site.