Currently, a diagnosis of ADHD according to the Diagnostic and Statistical Manual of Mental Disorders – 5th Edition (DSM-5TM) criteria is based on a polythetic approach, whereby none of the individual symptoms listed is considered essential for diagnosis, and, correspondingly, individuals with the same diagnosis may or may not share a certain symptom profile. For example, in children and adolescents, a diagnosis of ADHD based on the DSM-5TM criteria requires the presence of at least six out of nine listed symptoms of inattention, at least six out of nine symptoms of hyperactivity/impulsivity, or both; therefore, a referred child or adolescent presenting with ADHD symptoms, but with less than six symptoms of each of inattention and hyperactivity/impulsivity, would not receive a formal diagnosis of ADHD. For adolescents/adults (aged ≥17 years), the threshold for an ADHD diagnosis has been set at five symptoms of either inattention and/or hyperactivity/impulsivity; however, symptom onset in childhood, the pervasiveness and impairing nature of the symptoms, and the exclusion of other mental disorders that may better explain the presence of the symptoms, all represent additional criteria that must be met for a diagnosis of ADHD in individuals aged ≥17 years according to the DSM-5TM.
Due to the polythetic approach, it has been reported that the symptom threshold for a DSM-5TM ADHD diagnosis could theoretically be met by 116 200 different combinations of symptoms. Additionally, a recent study in two samples of children in Brazil demonstrated that, out of these 116 200 possible combinations of symptoms, 173 different symptom profiles were identified in the 189 children who had an ADHD diagnosis across both samples (Salum et al. 2018). Furthermore, only 16 (8.4%) of these children shared a symptom profile with another child in the study, and only four (2.3%) of the 173 symptom profiles were identified in both samples.
The authors noted that sources of heterogeneity in the ADHD phenotype extend further than this vast number of potential symptom profiles; comorbidity represents another source of heterogeneity. The source of symptom reporting also contributes to the heterogeneity of childhood ADHD; parents and teachers may report symptoms differently in the same child, for example they may report different symptoms entirely or give different weighting to individual symptoms. Developmental heterogeneity also exists, with the ADHD phenotype differing between, for example, adolescents and preschool-aged children. Interestingly, although a different number of symptoms are required to meet DSM-5TM criteria for a diagnosis of ADHD in adulthood compared with childhood/adolescence, the same approach has not been taken for preschool-aged children. The authors suggested that, since inattention and hyperactivity/impulsivity may, to some extent, be considered normal in very young children, a separate symptom threshold should be used for diagnosis of ADHD in this age group. Alternatively, a normative model approach could be used, whereby estimates of centiles of variation in neurobiological parameters (e.g. height and weight) across the population could be used to inform the establishment of different symptom thresholds required for a diagnosis of ADHD, according to different age groups.
The authors also suggested that heterogeneity is not the only barrier to improving the characterisation of ADHD phenotypes, and that to achieve this, the current diagnostic criteria for ADHD may need to be revised. They noted that, currently, all 18 symptoms listed in the DSM-5TM criteria have the same weight in the diagnostic algorithm, despite there being no evidence to support an equal contribution of each symptom in producing an ADHD phenotype. Accordingly, although the multicentre IMAGE study, which assessed the 18 listed ADHD symptoms in 1497 children and adolescents with ADHD and 291 without ADHD, demonstrated that each symptom contributed significantly and independently to the diagnosis of ADHD and predicted impairment, the 18 symptoms did not carry equal weight in terms of severity (Garcia Rosales et al. 2015). Additionally, other studies have found that the symptoms most likely to be associated with severity and impairment differ according to sample characteristics, leading the authors to emphasise that although prioritising symptoms for ADHD diagnosis is challenging, it should be a key research focus in ADHD.
Finally, the authors described an additional limitation to the current diagnostic procedure for ADHD; they noted that the current criteria for diagnosis are derived from research conducted in specialised clinical facilities, and do not necessarily reflect how diagnoses are reached in the real world. For example, the authors suggested that busy clinicians would not normally count the number of symptoms that a patient presents with, but instead may be more likely to use a prototypic approach, assessing the extent to which problems described by the referred patient fit the prototype of a given set of disorders. In accordance with this, the authors noted that the International Classification of Diseases, 11th Edition may eschew the current polythetic system of diagnosis in favour of a prototypic approach, and suggested that the development of diagnostic criteria for ADHD that are of clinical value in the real-world setting represents an important challenge for future research.
The authors concluded that the current diagnostic criteria for ADHD, based on 116 200 different possible combinations of symptoms, do not provide clinicians with everything they need to accurately diagnose ADHD across the lifespan in real-world clinical practice. The authors expressed their hope that the areas of concern raised in their article may help to inform the revision of the current ADHD criteria.
Cortese S, Rohde LA. ADHD diagnoses: are 116 200 permutations enough? Lancet Child Adolesc Health 2019; 3: 844-845.
Garcia Rosales A, Vitoratou S, Banaschewski T, et al. Are all the 18 DSM-IV and DSM-5 criteria equally useful for diagnosing ADHD and predicting comorbid conduct problems? Eur Child Adolesc Psychiatry 2015; 24: 1325-1337.
Salum GA, Gadelha A, Polanczyk GV, et al. Diagnostic operationalization and phenomenological hetero-geneity in psychiatry: the case of attention deficit hyperactivity disorder. Salud Mental 2018; 41: 251-261.