Where do neurodevelopmental conditions fit in transdiagnostic psychiatric frameworks? Incorporating a new neurodevelopmental spectrum

Studies that identified a neurodevelopmental factor

A factor explicitly called “neurodevelopmental” was first delineated alongside other psychiatric dimensions in a study using exploratory factor analysis on items of the Child Behavior Checklist (CBCL) in almost 10,000 children from the Adolescent Brain Cognitive Development (ABCD) study²⁵. Results showed a transdiagnostic structure including a neurodevelopmental factor along with externalizing, internalizing, somatoform and detachment factors. This neurodevelopmental factor included features of ADHD (e.g., inattention, hyperactivity), aspects of motor disorders (e.g., poor coordination, twitching), features conceptually overlapping between ASD and obsessive-compulsive disorder (OCD) (e.g., repetitive thoughts/obsessions and behaviors/compulsions, strange ideas), and poor school performance. A narrower inattentive neurodevelopmental factor, mainly defined by ADHD features, was also delineated using Adult-Self Report (ASR) items in adult parents of ABCD participants.

Interestingly, these neurodevelopmental factors first emerging in three-factor models were highly correlated with both broad internalizing and externalizing factors in the higher-level two-factor models. This finding suggests that the newly emerging factor is not a mere partition of a higher-order externalizing spectrum/dimension, but rather has associations with both internalizing and externalizing problems.

The neurodevelopmental factor was replicated in six subsequent studies. Using an equivalent analytic approach to the aforementioned ABCD study²⁵, a neurodevelopmental factor capturing inattention, hyperactivity/impulsivity, executive function problems, and learning difficulties was identified in a transdiagnostic sample enriched for problems with attention and learning⁴². This factor emerged along with internalizing and social maladjustment factors. A study on the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort at ages 7 and 13 years showed superior fit of correlated factor and bifactor models including a neurodevelopmental factor (ADHD and ASD scales) alongside emotional and behavioral factors relative to models only including two emotional and behavioral/neurodevelopmental factors⁴³.

Further, two studies of several hundreds of thousands children and adults from the Swedish population tested a neurodevelopmental factor (defined by diagnoses) within bifactor structures also including other specific factors (e.g., externalizing, internalizing) and a general psychopathology (p) factor^{44, 45}. One of these studies also modeled an ADHD factor separately and showed significantly stronger phenotypic and genetic associations of this factor with the neurodevelopmental factor than with externalizing and internalizing factors when controlling for p, congruent with the inclusion of ADHD in a neurodevelopmental spectrum (rather than in the externalizing spectrum).

Another recent study of Swedish adult twins identified a neurodevelopmental factor, capturing features of ADHD (inattention and disorganization) and ASD (social difficulties), along with internalizing, substance and impulsivity factors⁴⁶. Finally, a recent study exploring the data-driven hierarchical structure of the full set of DSM-5 clinical features in separate primary and hold-out adult population-based samples revealed a neurodevelopmental factor along with other major transdiagnostic dimensions⁴⁷. This broad neurodevelopmental factor included features of ASD, ADHD, and learning, communication, language and speech difficulties.

Findings from these studies are also in line with two additional studies that modeled a neurodevelopmental factor using various neurodevelopmental features and traits through confirmatory factor analysis^{48, 49}. However, as these studies did not include other psychiatric features alongside neurodevelopmental indicators, they cannot be used to determine the placement of the neurodevelopmental factor within a broader transdiagnostic structure.

Other evidence supporting the delineation of a neurodevelopmental spectrum

Several other lines of research, despite not explicitly using the term “neurodevelopmental”, support the integration of the proposed spectrum into transdiagnostic dimensional models. First, an exploratory factor analysis⁵⁰ and subsequent confirmatory factor analysis⁵¹ of the CBCL and a questionnaire assessing ASD and related developmental problems (Children's Social Behavior Questionnaire), in an epidemiological child and adolescent cohort, found “attention and orientation” and ASD factors, together with internalizing, externalizing and p factors. Of note, ADHD features did not cluster with externalizing, but rather with orientation features of ASD, and several ASD scales had significant loadings from both the attention and orientation factor and the ASD factor. Moreover, item-level analyses suggested that ASD and ADHD features clustered together in a model including an ASD/ADHD factor and internalizing and externalizing factors, and only differentiated into two separate factors in models with at least eight narrower factors⁵². This indicates that some features of ASD and ADHD may form a broad spectrum that may be further differentiated in more specific factors⁵².

Second, studies incorporating both personality and mental health measures have identified factors similar to the neurodevelopmental spectrum. A recent exploratory factor analysis of items from ADHD, ASD, and questionnaires measuring normal-range personality and psychopathology in adults, found that inattention, hyperactivity/impulsivity, and pragmatic language difficulties clustered together, whereas aloofness loaded onto detachment factors^{53, 54}. Another study in children and adolescents found that items of distractibility, organization, achievement, hyperactivity, and intelligence loaded together on an “organization” factor and separately from items belonging to positive personality, behavior problems, and internalizing factors⁵⁵.

Third, several previous studies using bifactor models in samples of children, adolescents and adults showed that indicators of ASD or ADHD (especially inattention) were solely or predominantly captured by the p factor, with minor or non-significant loadings onto the externalizing factor^56-64. This suggests that ASD and ADHD may be better captured by a different factor not modeled in these studies, such as a neurodevelopmental factor. Similarly, factor analytic studies using measures from the Achenbach System of Empirically Based Assessment (ASEBA) in dozens of samples from preschool to old adult age, and across societies and raters, indicate that the attention problems scale is not subsumed within the externalizing scale (aggregating rule breaking and aggression) or the internalizing scale (anxious/depressed, withdrawn/depressed, somatic complaints)^{21, 65-71}. This is consistent with studies identifying a factor capturing features of ADHD separately from externalizing factors^72-80.

Summary of the structural evidence

Overall, the studies reviewed here provide accumulating evidence supporting the robust factor structure of the neurodevelopmental spectrum alongside other transdiagnostic spectra/dimensions, providing a strong argument for the inclusion of this spectrum in transdiagnostic dimensional frameworks. Available studies indicate that features of ADHD, ASD, intellectual disability, and learning, communication and motor disorders can be conceptualized within this dimension (see Table 1).

The studies reviewed here suggest the inclusion of ADHD features, particularly inattention, in the proposed neurodevelopmental spectrum, with impulsivity being captured by both neurodevelopmental and externalizing spectra in some studies^{25, 42}. This is inconsistent with studies that modeled ADHD exclusively as an indicator of externalizing factors^81-89, which informed current transdiagnostic frameworks typically placing ADHD under a (disinhibited) externalizing spectrum^{14, 18}.

However, it should be noted that, unlike recent studies identifying a neurodevelopmental or neurodevelopmental-consistent factor, the studies placing ADHD under externalizing factors did not include features of other neurodevelopmental conditions. Further, these studies only examined diagnoses or scale total scores, resulting in only one or two ADHD indicators – an insufficient number for the emergence of a separate neurodevelopmental/ADHD factor. The use of diagnoses in this context may also be particularly problematic, because children with ADHD with high hyperactivity-impulsivity are more likely to receive an ADHD diagnosis, especially when comorbid with oppositional defiant disorder (ODD) or conduct disorder (CD), compared to children only showing inattentive features, who are often misdiagnosed or underdiagnosed⁹⁰. This artificially inflates the covariation between ADHD and ODD/CD and may result in an incorrect placement within the externalizing spectrum.

Crucially, before the DSM-5 and ICD-11, ADHD and ASD could not be diagnosed together in the same individual, and, since ASD is typically diagnosed earlier in development, an additional ADHD diagnosis was not possible⁹¹. Studies including ADHD diagnoses prior to the latest editions of those diagnostic manuals, therefore, likely missed a sizeable portion of the ADHD population with predominant neurodevelopmental and autistic features.

A few limitations of the available literature should be noted. Most of the studies used samples with wide age ranges and cross-sectional designs, precluding the examination of developmental effects. This limitation is common to many studies on the structure of psychiatric conditions, but especially problematic for delineating a neurodevelopmental spectrum with its roots in early development. Another limitation is that the majority of studies focused on ADHD and to a lesser extent ASD, while only a minority included learning, motor, intellectual and communication difficulties (see Table 1). While the few available studies consistently support the inclusion of these latter conditions in a neurodevelopmental spectrum, future studies with broad assessments of neurodevelopmental features are warranted.

Genetics

Neurodevelopmental conditions run in families, with higher rates in family members of affected individuals⁹². Twin studies provide strong evidence that neurodevelopmental conditions are highly heritable (with an average meta-analytic heritability of 0.66, ranging from 0.62 for specific learning disorders to 0.86 for intellectual disabilities).

Further, these conditions share a substantial degree of genetic influences with one another (the meta-analytic genetic correlation across neurodevelopmental conditions is 0.36, ranging from 0.07 for ADHD-specific learning disorders to 0.90 for ADHD-motor disorders)⁹², consistent with a neurodevelopmental spectrum, and show a more modest genetic overlap with other psychiatric conditions (e.g., 0.10 to 0.31 with schizophrenia, but also 0.62 with disruptive disorders)^{1, 93-96}.

Notably, the strong shared genetic influences between ASD, ADHD, the broader autism phenotype, intellectual disabilities; coordination, speech and language problems; and other social problems^{95, 97-104} can be captured by a latent genetic neurodevelopmental factor¹⁰⁵. Recent meta-analyses and reviews of twin and family studies indicate that up to ~80% of the phenotypic correlation between ASD, ADHD and other neurodevelopmental conditions, as well as conditions that also typically emerge earlier in development, such as OCD, can be explained by shared genetic influences^{92, 93, 106}.

Consistent with findings from twin studies, recent genome-wide association studies (GWAS) have also found evidence of considerable genetic overlap between neurodevelopmental conditions at the level of common single nucleotide polymorphisms (SNPs)^107-111. In the largest GWAS conducted on ASD to date (18,000 cases and 28,000 controls), one of the most robust genetic correlations was between ASD and ADHD (r_g=.36)¹⁰⁸.

To model the genetic covariance structure among multiple conditions/traits (including ASD and ADHD) using GWAS summary statistics, factor analytic methods such as genomic structural equation modeling (GenomicSEM) have been developed¹¹². Recent findings using this approach have shed new light on the genetic basis of a neurodevelopmental factor that appears to include ASD, ADHD and Tourette's syndrome^{109, 113-115}. Further, findings from GenomicSEM studies have thus far consistently converged on the existence of a neurodevelopmental factor that is genetically separable from other factors (i.e., internalizing, externalizing, psychosis^{114, 115}).

Several of the genes associated with ADHD¹⁰⁷ and ASD¹⁰⁸ are also more generally involved in neurodevelopmental processes. Regarding ADHD, FOXP2 is involved in synapse formation and speech and learning development^{108, 116-118}; SORCS3 encodes a brain-expressed transmembrane receptor important for neuronal development and plasticity^{119, 120}; SEMA6D appears to play a role in neuronal wiring during embryonic brain development^{120, 121}; and ST3GAL3 and MEF2C have been associated with severe intellectual disability^{120, 122-124}. Similarly, regarding ASD, strong enrichment was observed for regulatory elements in the brain regions involved in prenatal neurodevelopment¹⁰⁸. This is in line with studies of genes (e.g., SHANK3, MEF2C) that disrupt early synaptic function in genetic syndromes such as the Rett syndrome and the Phelan-McDermid syndrome, which often show intellectual disability, autistic features, and ADHD features¹²⁵.

Building on GWAS discoveries, polygenic risk scores seem to show robust associations across neurodevelopmental features, rather than just for individual neurodevelopmental conditions^{126, 127}. For example, in a large cohort study of children, an ADHD polygenic risk score was associated with inattention and hyperactivity but also language difficulties, and an ASD polygenic risk score was associated with inattention, hyperactivity, language and motor difficulties, whereas a schizophrenia polygenic risk score did not show significant associations with neurodevelopmental features¹²⁸. Both ADHD and ASD polygenic risk scores further predicted a neurodevelopmental spectrum based on exploratory factor analysis delineated in the ABCD study¹²⁹.

It is important to note that genomic studies of most psychiatric conditions are still in their infancy. As such, the genetic architecture of neurodevelopmental and other psychiatric conditions may change as sample sizes increase and new genomic data become available. Although GenomicSEM research has now been conducted with ASD and ADHD, other conditions that are hypothesized to be included in the neurodevelopmental spectrum, such as motor disorders and intellectual disabilities, have yet to be considered, due to the absence of GWAS of these phenotypes. It is critical that these omissions are addressed in future genomic studies of the neurodevelopmental spectrum.

Environmental risk factors

While the above-reviewed research indicates relatively large genetic influences for neurodevelopmental conditions and their covariation, meta-analytic evidence also demonstrates modest to moderate environmental influences⁹². Associations have been found of various forms of pre- and peri-natal stress^130-134, preterm birth^{135, 136} and environmental toxicant exposure (e.g., lead)^137-141 with neurodevelopmental conditions, including ASD, ADHD, intellectual disabilities, and learning and motor difficulties, as well as associated cognitive profiles. This pattern is consistent with an underlying dimension of liability to all neurodevelopmental conditions influenced by the aforementioned environmental exposures.

However, there are two important caveats when interpreting this evidence. First, it remains unclear whether and how specific these associations are to clinical features of the proposed neurodevelopmental spectrum relative to features captured by other spectra/dimensions^{133, 142}. Shared environmental risk factors, similar to genetics, may contribute to the co-occurrence between neurodevelopmental conditions and other psychiatric conditions. Yet, it is also possible that the same environmental factors are associated with different transdiagnostic conditions at different points in development, pointing to some specificity. For example, the same environmental exposure may increase risk for neurodevelopmental conditions in infancy/childhood and for psychosis in adolescence/adulthood. Second, establishing an association between environmental risk factors and neurodevelopmental conditions does not necessarily imply a causal effect. For example, accounting for familial and other confounders often results in attenuated associations^{135, 143}. However, a recent systematic review of twin and sibling studies, which controlled for familial and other shared confounders, did find suggestive evidence of a causal effect of pre-natal exposures for ASD and ADHD¹³⁰.

Developmental course

Additional evidence supporting the validity of the proposed neurodevelopmental spectrum pertains to the similarities in developmental course displayed by features clustering in this dimension. A key characteristic of all neurodevelopmental conditions is that they typically emerge early in development. The first features of some conditions – such as ASD, language and communication disorders, and developmental coordination disorder – are often recognizable before school age^144-150. Other conditions, such as ADHD and learning disabilities, may be detected during the preschool period, but formal diagnosis usually does not occur until the child enters the school system and the skills associated with these conditions are evaluated^{31, 151}. Nevertheless, on average, neurodevelopmental features tend to emerge earlier than most other psychiatric conditions¹⁵² (see Figure 1).

Regarding developmental trajectories, individuals with neurodevelopmental conditions tend to show persisting features over development, with some degree of improvement often seen with maturation^{31, 153, 154}. Rates of diagnostic stability vary widely between studies (from ~20% to 100%)^{31, 155-159}, particularly for ADHD and ASD. This instability may be explained by methodological differences, with lower persistence rates generally reported in population-based samples than in clinical samples and in studies relying on strict diagnostic thresholds^{31, 36, 160, 161}. Nevertheless, research has shown that most individuals with neurodevelopmental conditions show worse functional outcomes into adulthood relative to neurotypical individuals^{31, 154, 162, 163}.

With regard to specific clinical features, inattention and socio-communication difficulties follow a relatively stable developmental pattern^{154, 164}, while other features may change or be manifested differently with age and development, or even be replaced by features of other psychiatric conditions (heterotypic continuity). As an example, motor hyperactivity tends to decline over childhood and adolescence, but is often replaced by a feeling of inner restlessness in adulthood, which is also common in internalizing conditions³¹. In addition, it has been suggested that support strategies and interventions may not alter the natural course of neurodevelopmental conditions, but rather provide skills to compensate for stable underlying problems³¹.

Altogether, evidence suggests similarities in developmental course between different neurodevelopmental conditions, namely an early onset and persistent course, with some degree of heterogeneity among individuals with the same condition^{7, 165, 166}. Transdiagnostic approaches seeking to parse individuals with different neurodevelopmental conditions through more specific data-driven subdimensions cutting across diagnostic boundaries are particularly valuable for characterizing variability in developmental course^{7, 167}. For example, individuals with features of multiple neurodevelopmental conditions – consistent with high score on an overarching neurodevelopmental spectrum – commonly display more persisting symptom trajectories and worse long-term outcomes, including psychiatric comorbidities^{153, 154, 168}. Future research should systematically evaluate the presence of different transdiagnostic neurodevelopmental subdimensions and how they may be linked to other psychiatric conditions, which could inform the design of transdiagnostic early interventions and support strategies.

Temperamental antecedents and personality

There appears to be both consistency and specificity in the temperamental traits that characterize neurodevelopmental conditions, though most of the existing research has been limited to ASD and ADHD, and to a lesser extent intellectual disabilities and communication disorders.

Lower levels of effortful control are common across all neurodevelopmental conditions^169-178. Higher levels of negative emotionality are also frequent across these conditions, though the expression of this domain may partly differ across conditions, with distress, fear, shyness and sadness predominant in ASD and intellectual disabilities, while anger predominates in ADHD^{169, 179}. Positive emotionality/surgency also shows specificity across neurodevelopmental conditions, with higher levels in ADHD but lower levels in ASD and intellectual disabilities^{169, 170, 178-181}.

Although less common, longitudinal prospective studies highlight early temperamental indicators of lower effortful control, and higher negative emotionality and positive emotionality/surgency, as predicting ADHD symptom trajectories¹⁸²; and lower effortful control in toddlerhood as predicting ASD diagnosis in early childhood¹⁸³. Of note, lower effortful control and higher negative emotionality have also been associated with other psychiatric conditions, and may represent more general transdiagnostic indicators^{182, 184}.

Cognitive and socio-emotional profiles

Neurodevelopmental conditions are characterized by difficulties in cognition, social interaction, and communication that emerge early in development and persist into later life^{5, 146}. Early cognitive challenges are a key feature of these conditions, though the nature of challenges may vary among different conditions, ranging from impairments specific to core processes and abilities, while other capacities are relatively preserved, to more generalized problems^185-187.

Executive functioning deficits cut across ASD, ADHD, motor disorders, and learning disabilities^{7, 31, 188, 189}. The ABCD study found that, among transdiagnostic dimensions, the neurodevelopmental factor showed the strongest association with executive functioning¹⁹⁰.

Studies directly comparing children with ASD and ADHD showed similarities during early development in attention, response inhibition, working memory, verbal fluency, preparatory processes, and concept formation, with some differences more likely to be quantitative rather than qualitative^{145, 191, 192}. Similar cognitive challenges have also been reported in other neurodevelopmental conditions, such as language disorders, developmental coordination disorder, and Tourette's syndrome^193-195, but transdiagnostic comparison studies are lacking.

Early atypicalities in sensory processing (e.g., hyper-sensitivity) have been consistently reported in autistic children^{196, 197}, but are also found in children with ADHD¹⁹⁸. Further, motor skills, such as motor coordination, are impaired across developmental coordination disorder, Tourette's syndrome, ASD and ADHD^{145, 158, 199}, and communication/language problems characterize not just language and speech disorders but also to some extent ASD and ADHD^{145, 195, 200}.

With regard to social skills, children with ASD, ADHD, motor disorders and learning disabilities show social impairments relative to their peers^{7, 193, 201, 202}. There does appear to be some specificity in social impairment, with autistic children showing more atypical social skills^{7, 189, 203, 204}, whereas children with ADHD show relatively intact social skills knowledge but more maladaptive social behavior²⁰⁵. Perhaps due to impairments in self-regulation and inhibitory control²⁰⁶, both ASD and ADHD are also associated with elevated emotional dysregulation, including emotional lability and irritability^207-209.

Although there is evidence of characteristic cognitive and socio-emotional profiles for neurodevelopmental conditions, similar cognitive challenges have also been found for other psychiatric conditions^{51, 210}. However, because few studies comprehensively examined cognitive functions across the above conditions (with the notable exception of ADHD and disruptive behavior disorders^{211, 212}), the specificity of these cognitive profiles to neurodevelopmental conditions remains to be clarified in future transdiagnostic studies. Another important direction will be to address the wide heterogeneity in cognitive and socio-emotional profiles across neurodevelopmental conditions more systematically, by developing multidimensional models that integrate both clinical features and functioning abilities.

Neurobiological profiles: neuroimaging

Structural and functional magnetic resonance imaging (MRI) studies suggest largely shared patterns of brain structure and functioning in networks important for executive function and cognitive/motor control across neurodevelopmental conditions^213-217, supporting the proposed neurodevelopmental spectrum.

Specifically, similar patterns of grey matter structure have been observed in frontal, temporal, parietal and striato-thalamic networks supporting cognitive control and goal-directed decision-making across ASD and ADHD^218-224. Mega- and meta-analyses have found associations of ASD and ADHD with lower grey matter volume in ventromedial orbitofrontal cortex and basal ganglia, amygdala and hippocampus, and with lower cortical surface area and thickness in frontal, temporal and parietal regions^{220-222, 225-232}. Neuroanatomical findings may, however, vary with development (e.g., evidence of increased total brain volume in ASD during infancy/early childhood²²⁴) and have been somewhat less consistent in studies of ASD compared to ADHD.

Few studies have compared neurodevelopmental conditions directly, but these findings are consistent with a recent cross-disorder study showing commonalities in cortical regions implicated in socio-emotional and executive functions across ASD and ADHD (as well as OCD)²³³. Additional cross-disorder mega- and meta-analyses further indicate a range of structural brain features shared between ASD and ADHD, such as lower cortical thickness and surface area, though some divergent findings have also been observed, possibly due to the fact that some studies analyzed summary statistics rather than raw data^234-236.

While the majority of studies have focused on ASD and ADHD, structural MRI studies of other neurodevelopmental conditions have shown more mixed findings, likely owing to the use of smaller samples^237-242. Larger comparative studies investigating all neurodevelopmental conditions, as well as other psychiatric conditions, are needed.

As to brain functioning, atypical activations within fronto-striatal brain circuitry and basal ganglia appear to be most commonly implicated across ASD, ADHD, and speech and language disorders. Meta-analyses show underactivation related to cognitive and inhibitory control in lateral/medial fronto-striatal networks in ADHD^{222, 243, 244}. Similarly, a whole-brain meta-analysis of functional MRI studies of cognitive control in ASD found underactivation in salience and executive networks, including lateral and medial frontal regions and left cerebellum, and overactivation in right temporo-parietal regions, including areas of the default-mode network^{218, 221}. Moreover, shared reductions in dorsolateral prefrontal cortex and precuneus activation during higher-order cognitive functions and resting state have been observed across ASD and ADHD^245-248.

Fewer studies have investigated the neural underpinnings of other neurodevelopmental conditions using functional MRI. Regarding specific learning disabilities, dyslexia has been associated with primarily left-lateralized underactivation in lateral temporal, inferior parietal, and fusiform regions, as well as overactivation in motor cortex and the anterior insula^249-251, whereas dyscalculia is associated with number-relevant parietal regions and prefrontal and occipital regions²⁵². Initial data in Tourette's syndrome suggest that functional dysregulation in basal ganglia and limbic regions related to the suppression of tics may overlap with functional abnormalities and related cognitive control difficulties seen in ADHD²⁵³.

Shared alterations across ASD and ADHD have been found in structural connectivity, particularly in white matter tracts connecting prefrontal regions to other regions^{254, 255}, as well as in functional connectivity in large-scale brain networks, particularly in the default-mode network and fronto-parietal attention-related salience networks^{248, 256-261}. Such cross-disorder overlap is also reflected in recent data from the ABCD study showing the association between a neurodevelopmental factor derived from exploratory factor analysis²⁵ and resting state functional connectivity involving the default-model, cingulo-opercular and dorsal attention networks²⁶². Associations with functional connectivity were significant even after controlling for a general psychopathology factor and did not emerge for other psychopathological dimensions (e.g., internalizing, externalizing), suggesting the specificity of these atypical functional connectivity profiles to neurodevelopmental features.

Functional connectivity studies in motor disorders are somewhat more limited, but those directly comparing developmental coordination disorder and ADHD have found evidence for shared abnormalities across disorders in motor circuitry, pointing to a common neural substrate^{263, 264}.

Overall, the available evidence indicates an overlap in candidate diagnostic biomarkers of brain structure and functioning across neurodevelopmental conditions, particularly ASD and ADHD. Structural and functional atypicalities in frontal and striato-thalamic regions implicated in attention and cognitive control, and abnormalities in resting-state functional connectivity networks, particularly the default mode and dorsal attention networks, may be a unifying feature across a neurodevelopmental spectrum. While initial evidence points to brain patterns that may be common across neurodevelopmental features but discriminate them from other psychiatric features²⁶², future studies should systematically assess the extent of possible neurodevelopmental-specific neuroimaging biomarkers.

Neurobiological profiles: neurophysiology

Evidence from neurophysiological studies, particularly based on electroencephalography (EEG), complements knowledge from neuroimaging studies. Reductions in event-related potentials (ERP) amplitudes reflecting early sensory perception, such as N1 and mismatch negativity (MMN), have been commonly reported in ASD^{265, 266} and dyslexia^{267, 268}, whereas associations with ADHD and developmental coordination disorder appear more consistent for reduced MMN^{265, 269} than for N1 components²⁷⁰.

Atypical N170 amplitudes in response to facial stimuli have been studied almost exclusively in ASD, and large studies in other conditions are needed^{271, 272}. Studies of ERPs indexing goal-directed behavior indicate reduced N2 and error-related negativity (ERN) components in ASD^{266, 273}, ADHD^{273, 274}, specific learning disabilities^{275, 276} and developmental coordination disorder²⁷⁷, whereas increased ERNs are commonly associated with Tourette's syndrome²⁷³. The latter finding might be explained by the close link between Tourette's syndrome and OCD, as OCD is also associated with increased ERN, potentially suggesting that Tourette's syndrome might be jointly captured by internalizing and neurodevelopmental spectra. Further, reduced P3 components in response to visual and auditory stimuli have been repeatedly associated with ASD^{278, 279}, ADHD^{280, 281}, specific learning disabilities²⁷⁶, Tourette's syndrome²⁸², and developmental coordination disorder^{269, 283, 284}.

With regard to resting-state EEG studies, elevated power in slow oscillations (delta and theta), thought to reflect delayed cortical maturation, have repeatedly been associated with ASD²⁸⁵, ADHD^{286, 287}, and learning difficulties^{288, 289}. Reduced alpha and beta power have also commonly been associated with these conditions^290-292, although findings for alpha power have been more mixed in ADHD^{287, 293}. Emerging findings suggest shared alterations in intra-individual EEG variability^{294, 295} and EEG connectivity^296-298 between ASD and ADHD. Similarly, a recent study of children with ASD or ADHD and controls found a data-driven cluster largely reflecting a transdiagnostic neurodevelopmental subgroup based on functional connectivity measured with magnetoencephalography (MEG)²⁹⁹.

These findings support the validity of the proposed neurodevelopmental spectrum, in that they delineate neurophysiological profiles mapping onto neural mechanisms shared across neurodevelopmental conditions. However, similar to the neuroimaging literature, the strongest evidence pertains to ASD and ADHD, warranting replication and extension in transdiagnostic studies of neurodevelopmental and other psychiatric conditions.

Other neurobiological profiles

The literature on peripheral diagnostic biomarkers provides further evidence in support of the neurodevelopmental spectrum. Meta-analyses and reviews indicate that reduced levels of serum vitamin D levels^300-303 and blood zinc levels^304-307 are associated with ASD, ADHD, intellectual and learning difficulties. Blood levels of brain-derived neurotrophic factor (BDNF) were found elevated in both ASD and ADHD^{308, 309}. However, the strength of the evidence for these biomarkers varied in these meta-analyses and was often rated as weak, indicating that larger and more rigorous studies are needed.

Service organization and efficacy of support and intervention approaches

Patterns of service organization and efficacy of support and intervention strategies provide additional evidence for the validity of the proposed neurodevelopmental spectrum.

Given the high rates of co-occurrence across neurodevelopmental conditions, the need for multifaceted professional care that adequately addresses varied symptom presentations and needs beyond diagnostic categories has been widely advocated^{310, 311}. In fact, in several countries, clinical services are already organized to provide care for features spanning the neurodevelopmental conditions³¹², with service-user organizations actively advocating for a shift toward transdiagnostic neurodevelopmental services in countries where care is still based on discrete conditions³¹³.

Evidence from treatment studies shows that stimulant medication broadly improves features of inattention, hyperactivity and impulsivity³¹⁴; academic performance in math productivity, math accuracy and reading speed³¹⁵; outcomes related to motor coordination³¹⁶, and cognitive performance in children with ADHD but also ASD³¹⁷. These effects across neurodevelopmental conditions suggest that medication is not acting upon mechanisms specific for one diagnostic category, but instead on shared underlying processes^{315, 317, 318}. It is important to note, however, that some core features of ASD, such as social interaction and stereotypical behaviors, do not show improvements with stimulants³¹⁸, highlighting heterogeneity.

Regarding response to psychosocial interventions, initial evidence indicates that transdiagnostic traits (executive functioning and emotion regulation) predict efficacy of an intensive behavioral treatment program targeting school readiness over and above ASD and ADHD diagnoses²⁰⁶, further supporting the validity of the proposed spectrum. Since most transdiagnostic studies have focused on ASD and ADHD, additional research is needed to conclusively determine whether transdiagnostic psychosocial interventions can provide similarly efficacious benefits across other neurodevelopmental conditions.

Summary of the validity evidence

Overall, there is a broad degree of similarity in genetic, neuroimaging, neurophysiological and other biomarkers, as well as in developmental, cognitive, socio-emotional, temperamental/personality profiles, and patterns of service organization and treatment efficacy, across neurodevelopmental conditions (see Table 2). Evidence is most convincing for ASD and ADHD, as these conditions have been investigated to a much greater extent than other neurodevelopmental conditions.

Many studies support a dimensional approach, highlighting within-diagnosis variability and between-diagnosis overlap, motivating the search for shared patterns of brain-behavior associations across neurodevelopmental conditions. However, most studies have only considered one condition or symptom dimension in small samples, highlighting an acute need for larger studies using transdiagnostic samples with neurodevelopmental and other psychiatric conditions.

Besides broad similarities supporting the coherence of a neurodevelopmental spectrum, findings suggesting some differences between neurodevelopmental conditions, if replicated, would be consistent with the possible delineation of subdimensions (e.g., inattention, social difficulties) within this broad spectrum.

Two points are worth noting. First, multiple and different causes and configurations of underlying mechanisms may give rise to shared phenotypes across neurodevelopmental and other psychiatric conditions (i.e., equifinality³¹⁹). At the same time, the same underlying cause or mechanism may result in phenotypic variation and different outcomes within neurodevelopmental conditions (i.e., multifinality³¹⁹). Further transdiagnostic research using etiological and developmental approaches will be important for elucidating these equifinal and multifinal pathways, and for informing clinical decisions and applications based on underlying mechanisms and developmental processes.

Second, while the reviewed validity evidence points to commonalities across neurodevelopmental conditions, the extent of the overlap with other psychiatric conditions remains largely unclear. For example, similar temperamental, cognitive and neural profiles may also be found in the context of later-onset psychiatric conditions, such as depressive and psychotic disorders⁹. Yet, studies of other psychiatric conditions rarely examine whether these findings may be explained by co-occurrence with earlier-onset features of neurodevelopmental conditions. The systematic inclusion of neurodevelopmental features and conditions in future longitudinal transdiagnostic studies will be crucial to clarify this issue.

Reliability and measurement

To contextualize the potential utility of dimensional alternatives, it is useful to consider evidence for the limited reliability of categorically-defined neurodevelopmental conditions. Most evaluations of reliability of these conditions have focused on ASD and ADHD, whereas data are lacking for other neurodevelopmental conditions (with a few exceptions³²¹). Interrater reliability (i.e., the extent to which two raters agree) has generally been found to be low to barely adequate for both ASD^322-325 and ADHD^326-328, and even lower for specific ADHD subtypes³²⁹.

Interrater reliability estimates for ASD and ADHD diagnoses are also generally lower than for other psychiatric diagnoses^{323, 324, 330, 331}. One important consideration is that, unlike diagnoses of most other psychiatric conditions, diagnoses of neurodevelopmental conditions are often made using parent and other informant (e.g., teacher) reports rather than relying solely on self-report, especially at younger ages. As cross-informant discrepancies must be resolved when making a diagnosis, despite often reflecting unique information that each reporter has access to rather than errors, this can lead to low interrater reliability³³².

Interrater reliability is consistently higher for neurodevelopmental features and specific symptom dimensions. For example, interrater reliability for diagnostic features of ASD (e.g., reciprocal social interaction, communication, repetitive and restricted behavior) has been found to be higher than for the diagnosis^{323, 324, 331}. ADHD symptom dimensions (inattention and hyperactivity-impulsivity dimensions) similarly show higher interrater reliability than the diagnosis³²⁹.

Moreover, estimates of test-retest reliability (a measure of temporal stability) also tend to be higher for dimensional conceptualizations of neurodevelopmental conditions relative to categorical ones. Whereas findings for categorical diagnoses of ASD^{324, 333} and ADHD^334-336 have varied substantially, dimensional studies examining inattention, hyperactivity, and sluggish cognitive tempo features of ADHD^{333, 337, 338}, as well as social, communication, sensory and motor features of ASD^339-341, have consistently shown high stability. Overall, there is extensive evidence that spectra/dimensions of neurodevelopmental and other psychiatric features show superior temporal stability compared to diagnostic categories^{43, 62, 337, 342-344}.

The relatively lower reliability found for neurodevelopmental categories compared to dimensions may be due in part to the high co-occurrence between neurodevelopmental diagnoses and their transdiagnostic features (e.g., similarities in symptoms). These issues pose significant challenges to differential diagnosis among the neurodevelopmental conditions³⁴⁵. For example, the interrater reliability for the DSM-IV autism disorder diagnosis – one of five DSM-IV pervasive developmental disorder (PDD) diagnoses – was generally adequate-to-good for the diagnosis of autism versus non-PDD conditions^{326, 345}, but lower for the differential diagnosis of autism versus other PDD^{326, 345}.

Moreover, it is difficult to partition true dimensional and transdiagnostic features into discrete diagnostic categories. For example, clinicians commonly report uncertainty when determining whether individuals with significant social difficulties meet criteria for ASD versus other neurodevelopmental conditions³⁴⁶, and the feature of elevated inattention that is key for ADHD is also commonly seen in children with other neurodevelopmental conditions, as well as other psychiatric conditions^{97, 347}.

Notably, standard diagnostic measures for children and adolescents – e.g., the Diagnostic Interview for Children and Adolescents (DICA)³⁴⁸ – assess ADHD but rarely include other neurodevelopmental conditions, and most adult diagnostic interviews – e.g., the Structured Clinical Interview for DSM-5 (SCID-5)³⁴⁹ – do not include any neurodevelopmental conditions. As such, these conditions are rarely assessed alongside other psychiatric conditions (because doing so would require the addition of another diagnostic assessment and increase burden on researchers/clinicians and patients). This means that the differential diagnosis of neurodevelopmental conditions versus other psychiatric conditions is rarely done in research or clinical contexts, making it difficult to evaluate or differentiate transdiagnostic features (e.g., attention problems due to ADHD or to depression) and contributing to their artificial separation in research and clinical practices³².

Challenges in the reliable assessment of categorical neurodevelopmental diagnoses have important implications. For example, the DSM-5 introduced several changes in the new ASD diagnosis, now encompassing all five DSM-IV PDD diagnoses. This shift has resulted in more severe presentations in individuals receiving a DSM-5 ASD diagnosis compared to those with a DSM-IV diagnosis, but in fewer DSM-5 diagnoses in individuals who had previously met criteria for DSM-IV PDD diagnoses^{350, 351}. At the same time, many individuals who do not meet criteria for a DSM-5 diagnosis nonetheless still show features of ASD^{350, 351} and other psychiatric conditions³⁰. A shift toward transdiagnostic dimensional assessments that do not rely on strict diagnostic cut-offs would avoid these problems, because dimensional approaches allow for the assessment of key features of neurodevelopmental and other psychiatric conditions, as well as of symptom severity and functional impairment, likely leading to a better identification of treatment targets.

Consistent with recognition of the usefulness of a dimensional approach, there is a long history of dimensional measures for ADHD^352-356, such as the Strengths and Weaknesses of ADHD Symptoms and Normal Behavior Scale (SWAN)³⁵³, the Adult ADHD Self-Report Scale (ASRS)³⁵⁵, and scales included in the ASEBA system²¹. There has also been a recent move toward dimensional measures for ASD^{41, 357}, such as the Social Responsiveness Scale (SRS)³⁵⁸, the Autism Symptom Dimensions Questionnaire (ASDQ)³⁵⁹, and the Repetitive Behavior Scale-Revised (RBS-R)³⁶⁰.

These dimensional measures carry multiple advantages, as they are readily adopted in clinical settings (especially when conducting full diagnostic assessments is not feasible or preferred) and show improved interrater reliability and prediction of functioning and impairment relative to categorical ASD and ADHD diagnoses^358-362. However, most of these assessment tools are still focused on features of individual conditions, rather than providing a transdiagnostic assessment of all neurodevelopmental conditions, as well as other psychiatric conditions.

Dimensional measures for neurodevelopmental conditions other than ASD and ADHD are also still uncommon, with a few exceptions^{363, 364}. In addition, measurement invariance is rarely considered or explicitly evaluated, but it is important to establish when considering assessments across informants, age, sex/gender, and other sociodemographic and clinical characteristics. Notably, initial work with some standard dimensional measures of ASD indicates a lack of measurement invariance for birth sex that obscures sex-specific patterns of ASD symptoms among boys and girls³⁶⁵.

Overall, it is clear from decades of research that – as for dimensional conceptualizations of internalizing, externalizing and psychotic conditions^{8, 366} – dimensional measures of neurodevelopmental conditions show higher reliability than categorical indicators^{7, 41}. The development of new transdiagnostic dimensional measures that span across features of all neurodevelopmental and psychiatric conditions and do not rely on strict diagnostic cut-offs will likely further improve reliability. In turn, better reliability of measures can result in more robust research efforts, greater utility in clinical settings, and improved clinical care.

Explanatory and prognostic utility

A critical aspect of the utility of a model or framework is its ability to explain and predict clinical phenomena, risk factors, and outcomes. The heterogeneity intrinsic to neurodevelopmental diagnostic categories has long been regarded as a significant barrier to explanatory and prognostic utility^367-371. The transdiagnostic nature of risk factors, cognitive and socio-emotional profiles, and developmental course across neurodevelopmental conditions (as reviewed above) further points to the limited explanatory and prognostic value of categorical diagnoses.

Conversely, a growing body of evidence indicates that a transdiagnostic neurodevelopmental spectrum, or narrower dimensions included within this spectrum (e.g., inattention), can improve explanatory and prognostic utility over a categorical approach, in line with meta-analytic evidence showing greater validity of dimensional conceptualizations relative to categorical ones for a wide range of psychiatric conditions³⁶⁶.

In the Canadian 2006 Participation and Activity Limitation Survey (ages 5-14; weighted N=120,700), a factor capturing neurodevelopmental and related functional characteristics fully mediated the relationship between an ASD diagnosis and child and family health and well-being outcomes (e.g., involvement in school activities, family psychological well-being)³⁷². The direct effects of the neurodevelopmental factor on these outcomes were roughly twice as large as the effects of ASD diagnosis, and these latter effects disappeared when the neurodevelopmental factor was added into the model³⁷², indicating the superior explanatory utility of the dimensional approach.

Longitudinal evidence from three cohorts (total N=1,253) further showed that a model including a neurodevelopmental factor primarily capturing ADHD features, measured in early childhood, predicted pre-adolescent psychiatric psychopathology and psychosocial functioning about two times more strongly than meeting criteria for a DSM diagnosis³⁷³.

These findings align with replicated evidence that continuous dimensions of inattention and hyperactivity-impulsivity show greater predictive utility and temporal stability than DSM ADHD inattentive, hyperactive-impulsive, and combined type diagnoses^{329, 374-377}, speaking to the utility of a dimensional approach and the limited validity of categorical diagnostic specifiers.

Several studies also indicate that neurodevelopmental factors show stronger associations with a range of important risk factors and outcomes – including global executive functioning, brain functioning, developmental delays, learning problems, psychosocial functioning, medication use, and service utilization – than general factors comprising other psychiatric conditions and clinical features^{25, 42, 190, 233, 262, 373, 379}. A large Swedish population study (N=1,093,788) showed that a neurodevelopmental factor was associated with reduced fetal growth to a greater extent than internalizing, externalizing, psychosis, and general psychopathology factors⁴⁵. These associations remained significant even in sibling analyses controlling for unmeasured familial confounding.

A transdiagnostic neurodevelopmental spectrum has also been found to show improved utility over specific neurodevelopmental features. For example, a large longitudinal study on a UK birth cohort (N=10,054) found that a broad childhood neurodevelopmental factor predicted adult mood problems more stably and to a larger extent than more narrow dimensions capturing motor and cognitive difficulties⁴⁹.

Overall, these findings indicate that a transdiagnostic characterization of neurodevelopmental conditions can better explain and enhance prediction of important outcomes beyond categorical approaches, and provides incremental validity even over dimensional indicators of other psychiatric conditions and clinical features²⁰².

Clinical utility

Clinical utility includes the extent to which shifting toward a transdiagnostic dimensional approach for neurodevelopmental conditions would reflect clinician acceptability and preference; facilitate practical implementation in clinical settings, patient conceptualization, communication among practitioners, patients and family members, clinical decision-making, and treatment selection/planning; and improve outcomes for individuals, families and clinicians³⁷⁹.

Several lines of evidence indicate that transdiagnostic dimensional approaches in general^17-19, and to neurodevelopmental conditions more specifically^{7, 202, 320, 380}, provide more clinically useful assessment and treatment targets than diagnostic categories.

Research on practitioner ratings has found that clinicians report a preference for and greater acceptability, feasibility and clinical utility of a transdiagnostic dimensional approach to diagnosis in both child and adult patients, including those with ADHD³²⁶. This is consistent with data showing that practitioners focus more on symptom severity than diagnostic categories in clinical decision-making³⁸¹. Notably, child and adult patients (and parents of child patients) also reported that the transdiagnostic approach was likely to be more helpful to the clinician in understanding their problems and needs³²⁶.

Another piece of evidence directly pointing to the clinical utility of a dimensional approach to neurodevelopmental conditions is the widespread implementation of dimensional assessments in clinical settings³⁸², particularly for ADHD^{353-355, 383, 384}, but also more recently for ASD^{41, 357-360, 385} and intellectual disabilities^{363, 386}. As a singular diagnosis does not provide clinicians with an accurate picture of the individual patient, dimensional assessments are routinely used in clinical care to yield a more complete conceptualization of an individual's strengths, weaknesses and needs beyond their categorical diagnosis^{7, 382, 387-390}.

This is illustrated by a recent field trial on the utility of newly-developed behavioral indicators for assessing the severity of intellectual disabilities, included within the Clinical Descriptions and Diagnostic Requirements for ICD-11 neurodevelopmental disorders³⁶³. Findings showed positive practitioner ratings regarding ease of administration, as well as clinical utility for treatment selection, prognostication, and communication with other health care and educational professionals³⁶³.

Regarding support and treatment strategies, there is evidence that targeting the social difficulties common across many neurodevelopmental conditions by designing interventions around social skills rather than diagnoses may be more helpful and cost-effective^{202, 391}. Dimensional assessments can also be useful for monitoring treatment efficacy and clinical improvements over time^{387, 389, 392}. Broadening the transdiagnostic focus across both neurodevelopmental and psychiatric conditions is likely to promote the development of preventive strategies to avert the emergence of later-onset mental health challenges in individuals with neurodevelopmental conditions. This is in line with recent proposals for a stepped-care personalized health model for ASD tailored to autistic individuals’ and families’ needs, strengths and challenges³⁹⁰.

Finally, the clinical utility and acceptability of the proposed transdiagnostic dimensional approach are reflected in stakeholders’ clinical priorities. Advocacy organizations and individuals with neurodevelopmental conditions have highlighted the unmet support needs resulting from the use of traditional categorical approaches in clinical settings^{33, 34, 393}. As a solution to these challenges, they have actively called for a shift toward clinical services supporting multiple neurodevelopmental conditions and including dimensional assessments for neurodevelopmental and co-occurring conditions^{313, 382}.

Emerging evaluations of clinical services adopting such an approach in the UK have shown improvements in capacity for support and satisfaction from patients, families and practitioners, along with reductions in waiting lists, re-referrals to different assessments (e.g., for learning difficulties in a child initially assessed in an ADHD clinic), and economic cost³⁸³. These encouraging data point to the clinical utility of a transdiagnostic dimensional approach to neurodevelopmental conditions and support the extension of this approach to other counties³⁹⁴.

Research implications

From a research perspective, our proposal to situate the neurodevelopmental spectrum in transdiagnostic models of psychiatric conditions can provide a formal framework for characterizing the common developmental continuity and co-occurrence among neurodevelopmental conditions and between them and other psychiatric conditions. Progress in this area has been slow, as research based on categorical approaches has typically considered neurodevelopmental conditions separately from one another and from other psychiatric conditions. This is at odds with how these conditions manifest clinically, and has led researchers, advocacy organizations, and people with lived experience to call for more transdiagnostic work across neurodevelopmental conditions and psychiatric comorbidities^{7, 25, 32-35, 395}. The notion of a neurodevelopmental spectrum and its explicit consideration alongside other transdiagnostic spectra/dimensions can directly respond to these issues by promoting a more holistic characterization of their frequent overlap across development.

Moreover, inclusion of the proposed neurodevelopmental spectrum into transdiagnostic models can have important implications for elucidating the etiological and developmental underpinnings of neurodevelopmental and other psychiatric conditions. Categorical approaches have encouraged etiological and developmental research focused on individual conditions, often not considering whether research participants also present with (or have a prior history of) other conditions. As a result, it has not been possible to clarify to what extent the risk factors and mechanisms identified for any individual condition are specific to that condition, shared across multiple conditions, or explained by unmeasured comorbidities⁹. Incorporating the neurodevelopmental spectrum into transdiagnostic models can allow researchers to pinpoint etiological and developmental processes that are specific to that spectrum (or its constituent subdimensions) and those shared with other transdiagnostic spectra/dimensions. Further, since earlier developmental expressions of neurodevelopmental features often precede more severe presentations and significant impairment¹⁴⁵, the proposed transdiagnostic approach can generate new insights into developmental processes more generally, which are key to early identification and prevention.

In addition, by encouraging a broad assessment of the degree of functioning and challenges displayed by each individual, the dimensional approach intrinsic to the proposed transdiagnostic conceptualization can also help consolidate understanding that neurodevelopmental conditions are on a continuum with typical functioning and subthreshold presentations^{7, 26}, just like their common psychiatric comorbidities^{12, 17-19}. A dimensional approach characterizing the entire population-level distribution of multiple neurodevelopmental features will also encourage researchers to move away from the known limitations of case-control studies^{7, 11, 17}. This is especially valuable for conditions with lower diagnostic prevalence rates (e.g., ASD), for which recruitment of eligible individuals into studies is challenging.

Overall, including the proposed neurodevelopmental spectrum in transdiagnostic frameworks can foster approaches to characterize co-occurring profiles, heterogeneity, mechanisms, and developmental processes, as well as promote significant refinement of theoretical and etiological models of neurodevelopmental and other psychiatric features. This can have important effects on stratification, biomarker discovery, and future clinical practices.

Clinical implications

From a clinical perspective, our proposal has implications for early identification, prevention, treatment planning, and transdiagnostic interventions. First, as neurodevelopmental conditions share many overlapping early features, focusing on transdiagnostic domains rather than single measures of separate diagnostic categories can improve early identification and optimize early intervention³⁹¹. This has direct clinical implications, for example, in addressing issues of diagnostic overshadowing^{33, 396}, such as when receiving a diagnosis of ASD leads to delayed diagnosis of other neurodevelopmental conditions (e.g., ADHD) or other psychiatric conditions (e.g., depression).

Second, a broader assessment of various neurodevelopmental features, alongside features of other psychiatric conditions, can yield a fuller evaluation of an individual's specific strengths, difficulties and needs than possible when focusing on categorical diagnoses. It can also allow for the assessment of gradations in functioning in multiple domains. This can both identify areas of strength, that may be emphasized, and allow for earlier identification of difficulties before they become severe enough to meet the level of a diagnosis. As a result, a transdiagnostic approach can directly address the priorities of individuals with neurodevelopmental conditions and their families, proving more clinically acceptable and less stigmatizing than categorical diagnostic approaches^{10, 32, 33}.

Third, neurodevelopmental conditions increase the risk for other psychiatric conditions in adolescence and adulthood^{31, 49, 151, 153, 395}, but this risk is non-specific across the neurodevelopmental spectrum^{49, 397}. A transdiagnostic approach to assessing the neurodevelopmental spectrum, instead of focusing on individual diagnoses, offers greater explanatory power for predicting important outcomes. Characterizing risk pathways between a broad neurodevelopmental spectrum and subsequent psychiatric conditions can improve prevention and timely diagnostic assessment³⁹⁵.

Finally, the proposed transdiagnostic dimensional approach can increase access to needed services. Clinicians, users and their families, and advocates speak to the need for approaches and services that recognize mental health complexity^{33, 391, 398, 399}, including service delivery based on severity, functioning and prognosis, rather than limiting access to only those who have received a specific diagnosis^{33, 400, 401}. Under the current categorical approach, an individual displaying impairing features of several neurodevelopmental conditions, but not reaching diagnostic cut-offs for any specific diagnosis, can be denied care. Instead, taking a more universal approach to providing services focused on broad neurodevelopmental features can enable prompt recognition and support based on individual needs, helping to address barriers to care, including structural inequities related to immigration, race and ethnicity^402-404. Additionally, such an approach may be destigmatizing, since it will encourage moving away from diagnostic labels towards a focus on individual challenges as well as strengths^{10, 405}.