Cut from the same cloth: neurobiological continuity between childhood and adult ADHD

The growing recognition of adult attention‐deficit/hyperactivity disorder (ADHD) has been accompanied by discussions over the strength of its evidence base, that in turn feed into debates over the “reality” – the ontological status – of ADHD itself. Against this context, it is helpful to consider in detail the literature on the neurobiological foundations of adult ADHD, expanding upon the points already raised by Cortese et al ¹ in their excellent paper. It is particularly salient to ascertain the extent to which the neural features tied to childhood ADHD extend into adulthood among those who retain the diagnosis, and the genetic variants underpinning childhood ADHD overlap with those observed in adult ADHD.

To what extent are the neural differences seen in adult ADHD carried forward from childhood? This question is most directly answered by combining clinical and neuroimaging observations acquired in tandem from childhood into adulthood. There are only a handful of such studies, reflecting the challenges of longitudinal work conducted over decades, compounded by instability in the measurement of neural features arising from rapid advances in neuroimaging technologies.

One longitudinal in vivo structural magnetic resonance imaging (MRI) study, conducted over 20 years on the same magnetic resonance scanner, found that young adults whose ADHD symptoms had persisted from childhood showed neural differences that also persisted in a relatively “fixed” manner from childhood ² . These neuroanatomic features localized to the cognitive control and default mode networks, with changes to the thickness of the cortex in the posterior cingulate, right inferior parietal, and dorsolateral prefrontal regions. By contrast, those whose childhood ADHD symptoms largely resolved by adulthood showed an accompanying significant convergence toward typical cortical dimensions, rectifying early anomalies.

A similar theme emerged in a longitudinal study of cerebellar anatomy, albeit over a much shorter adolescent time window. While some midline cerebellar regions (superior vermis) showed fixed differences, atypical features of the cerebellar hemispheres persisted only among those who had persisting symptoms. Finally, a study spanning late childhood into early adolescence showed that persisting atypical microstructure of thalamic, striatal, and long association white matter tracts was found among those youth whose ADHD symptoms persisted ³ . In short, longitudinal neuroimaging studies point to subtle anatomic and white matter microstructural differences in children with ADHD that are often carried forward into adulthood if core symptoms persist.

Given the sparsity of longitudinal imaging data, other studies have focused on adults who have been followed clinically since childhood, thus enhancing diagnostic certainty, but who have had neuroimaging for the first time in adulthood. Such studies allow not only diagnostic comparisons but also contrasts of adults whose childhood ADHD has persisted against those whose childhood ADHD has remitted ⁴ . They find that those with adult ADHD (i.e., with the persistent form of ADHD) show atypical anatomic features, ranging from a thinner cortex and decreased thalamic grey matter density, to atypical microstructure of the white matter tracts within attention control (e.g., the inferior fronto‐occipital fasciculus) and reward processing (e.g., the uncinate fasciculus) networks. Atypical features in adults with ADHD symptoms persisting from childhood are also found for the brain's intrinsic functional architecture, mapped by both resting state functional MRI and resting state magnetoencephalography, and for brain activity during ADHD‐related cognitive processes, such as response inhibition. It seems likely that these neural features, which resemble those reported in studies of childhood ADHD, are carried forward from childhood in tandem with symptom persistence, while they are not present in adults whose childhood ADHD has remitted.

Counter to this evidence for neural differences in adult ADHD, several meta‐analytic studies report either minimal or no diagnostic differences ⁵ . The meta‐analytic null findings may stem from the reliance on mostly small, underpowered cross‐sectional studies. Faced with the limitations of meta‐analyses, some initiatives such as the ENIGMA consortium have taken a mega‐analytic approach. This involves the use of individual‐level imaging data acquired from multiple cohorts, which provides impressive sample sizes that can be analyzed using methods that account for “site of acquisition” effects. Additionally, many mega‐analytic studies process the “raw” imaging data on uniform pipelines and can thus employ consistent quality control standards. The use of individual‐level data also allows individual‐level confounds to be controlled, including medication history and co‐occurring conditions.

What do these mega‐analytic studies find? Considering neuroanatomy, the ENIGMA consortium reported that ADHD diagnostic differences are most marked in childhood, and present only at trend level, if at all, in adults, though the limited number of adults does not allow definitive conclusions ⁶ . Mega‐analytic studies of brain's functional architecture in children with ADHD find significant but small differences in the connectivity between the default mode network and task‐positive networks, and within the brain's cortico‐striatal information processing loops ⁷ . It will be fascinating to see if similar differences are present in forthcoming well‐powered mega‐analyses of adult brain function.

Looking to the future, the neurobiological understanding of adult ADHD will be transformed by rapid technological advances, such as imaging at ultra‐high field strengths (currently of 7 Tesla). Among its many advantages, high field strength imaging allows the precise quantification of key neurotransmitters, both inhibitory (such as GABA) and excitatory (such as glutamate). It is noteworthy that early in vivo imaging studies suggest an altered balance between GABA and glutamate levels in ADHD, and this finding is complemented by genetic studies that also point to these neurotransmitters ⁸ .

Turning to genomics, there is a compelling case for an overlap between the genetic features underpinning childhood and adult ADHD. The common genetic variants that explain part of the high heritability of adult ADHD are very similar to those found in childhood ADHD, with a genetic correlation around 0.8 ⁹ . Indeed, polygenic scores, that reflect genome‐wide measures of common variants tied to ADHD, are higher among those with ADHD that persists into adulthood compared to childhood‐limited forms. Furthermore, longitudinal twin studies show that genetic factors account for most of the adolescent change in hyperactive‐impulsive symptoms and around half of the change in inattention, with much more modest contributions from environmental factors. In short, the high heritability of adult ADHD is partly explained by common genetic variation that is shared with childhood ADHD. The next step will be to quantify the role of rarer forms of genetic variation, such as copy number variants and deleterious point mutations.

Due partly to collaborative efforts, brain‐based and genomic models of adult ADHD are being rigorously tested. Several neural features robustly tied to childhood ADHD extend into adulthood when symptoms persist, and the genetic variants that underpin childhood ADHD overlap considerably with those seen in adult ADHD. This neurobiological continuity provides an important evidence base to inform both scientific thinking and the public understanding of adult ADHD.