Does the concept of attention-deficit hyperactivity disorder (ADHD) help advance scientific knowledge?

In psychiatry (apart from the dementias and a few other known organically based conditions), there is no such thing as diagnosis. In medicine diagnosis is the process of determining which disease or condition explains a person’s symptoms and signs. Pseudo-diagnoses like ADHD cannot explain behaviours and there are no signs; only symptoms that are descriptions (not explanations) of behaviours. Consider the following example: If I were to ask the question ‘what is ADHD?’ then it is not possible for me to answer that question by reference to a particular known pathological abnormality. Instead I will have to provide a description, in other words ADHD is the presence of the behaviours of hyperactivity, impulsivity and poor attention (plus a few extra qualifiers such as age of onset). Contrast this with asking the question ‘what is diabetes?’ if I were to answer this question in the same manner by just describing symptoms such as needing to urinate excessively, thirst and fatigue, I could be in deep trouble as a medical practitioner as there are plenty of other conditions that may initially present with a similar picture and diabetes itself may not present with these symptoms in a recognizable way. In order to answer the question ‘what is diabetes?’ I have to refer to its pathology involving abnormalities of sugar metabolism. I would then get independent (to my subjective opinion) empirical data to support or otherwise my hypothesis about what may be ‘causing’ the patient’s described experiences (such as testing the urine and/or blood for levels of glucose). In the rest of medicine therefore, my diagnosis explains and has some causal connection with the behaviours/symptoms that are described. In psychiatry what we are calling diagnosis (such as ADHD) will only describe but is unable to explain.

This means that in psychiatry we are mostly working with a system for classification that is descriptive, but not diagnostic. As a classification it can have its uses such as recognizing and validating people’s struggles, administrative, communication, and other uses such as aiding decisions around educational support. The problem of using a classification like ADHD to explain an observed set of behaviours (i.e. as a diagnosis) can be illustrated by asking another set of questions. If I was to ask ‘why’ a particular child can’t concentrate, is hyperactive and shows impulsivity and I were to answer that these behaviours are caused by ADHD, then a legitimate question to ask me is ‘how do you know that they are caused by ADHD?’ The only answer I can give to that question is that I know it’s ADHD because the child is presenting with hyperactivity, impulsivity and poor attention. In other words if we try to use a classification that can only describe in order to explain, we end up with what philosophically is known as a ‘tautology’ where we are trapped in circularity.

As ADHD is not a medical diagnosis, but a descriptive classification, we have no empirical method for defining ‘caseness’. The definition of what qualifies as a case is thus arbitrary and depends on the standards employed by the diagnoser, influenced by whatever the prevailing ideology concerning diagnosis they have been exposed to. As a result we cannot eliminate wide variation in ‘diagnostic’ practice or come to any valid conclusion about what percentage of the population is ‘over’ or ‘under’ diagnosed.

As ADHD is not a medical diagnosis it is not surprising that there has been a failure to find any specific and/or characteristic biological abnormality such as characteristic neuroanatomical, genetic or neurotransmitter abnormalities (Campo et al. Reference Campo, Fryer, Hong, Smith, Brichard, Acosta-Cabronero, Chamberlain, Tait, Izquierdo, Regenthal, Dowson, Suckling, Baron, Aigbirhio, Robbins, Sahakian and Muller2013; Timimi & Timimi, Reference Timimi and Timimi2015; Whitely, Reference Whitely2015).

Does the concept of ADHD help improve clinical outcomes?

The evidence on outcomes in mental health in general finds little clinically significant impact on outcomes comes from matching treatment models to diagnosis (Timimi et al. Reference Timimi, Tetley, Burgoine and Walker2013), a finding that extends to ADHD (Miller et al. Reference Miller, Wampold and Varhely2008). Instead extra-therapeutic factors and the therapeutic alliance are the biggest contributing factors to variance in outcome from treatment.

Most of the controversy in ADHD treatments revolves around the use of stimulant medication. The evidence does not favour the continual increases in stimulant prescribing that has been occurring in the last couple of decades in most Western societies. Reviews of ADHD pharmacotherapy studies note widespread poor methodology, publication bias, limited reliability of results, inadequate data regarding adverse events, and lack of evidence of long-term benefit (e.g. Storebø et al. Reference Storebø, Krogh, Ramstad, Moreira-Maia, Holmskov, Skoog, Nilausen, Magnusson, Zwi, Gillies, Rosendal, Groth, Rasmussen, Gauci, Kirubakaran, Forsbøl, Simonsen and Gluud2015). The most commonly cited reference in support of using stimulant medication is an American study, which concluded that in a 14 month randomised controlled trial, patients receiving medication had better outcomes compared with those receiving behaviour therapy only (MTA Cooperative Group, 1999). Follow up at 36 months of the same patients, could not find support for continuing beneficial effects of medication over behaviour therapy, regardless of initial severity (Jensen et al. Reference Jensen, Arnold, Swanson, Hinshaw, Gibbons, Marcus, Hur, Vitiello, Abikoff., Greenhill, Hetchman, Pelham, Wells, Connors, March, Eliott, Epstein, Hoagwood, Hoza, Molina, Newcorn, Severe and Wigal2007). In addition, those who used more medication during those 3 years were more likely to experience a deterioration in ADHD symptoms and had higher rates of delinquency, were significantly shorter (by an average of 4 cm) and lighter (by 3 kg) than those who had not taken medication (Molina et al. Reference Molina, Flory, Hinshaw, Swanson, Arnold, Gibbons, Marcus, Hur, Jensen, Vitiello, Abikoff., Greenhill, Hetchman, Pelham, Wells, Connors, March, Eliott, Epstein, Hoagwood, Hoza, Newcorn, Severe and Wigal2007; Swanson et al. Reference Swanson, Hinshaw, Arnold, Gibbons, Marcus, Hur, Jensen, Vitiello, Abikoff., Greenhill, Hetchman, Pelham, Wells, Connors, March, Eliott, Epstein, Hoagwood, Hoza, Molina, Newcorn, Severe and Wigal2007). Other naturalistic studies have come to similar conclusions finding that medication offers little prospect of improving long-term outcomes (e.g. Government of Western Australia, Department of Health, 2010; Currie et al. Reference Currie, Stabile and Jones2013).