We determined the relative contributions of ASD and ADHD to internalising problems using a trait-based approach in a large general population sample. Consistent with the previous literature [e.g.,2,8], both ASD and ADHD traits predicted greater internalising problems. The regression analyses also indicated that ADHD traits may be a stronger predictor of internalising problems than ASD traits. This was formally confirmed with dominance analyses, which showed that ADHD traits dominated ASD traits in predicting internalising problems, with reproducibility estimates highlighting that this finding is almost certain to exist in the general population. Bolstering these results, Bayes Factors made it even clearer that ADHD traits were more predictive of internalising problems than ASD traits. Overall, we report converging evidence that ADHD traits are a stronger and more important predictor of internalising problems than ASD traits.
To our knowledge, this is the first evidence that ADHD traits are more predictive of internalising problems than ASD traits. While its novelty might appear surprising, it is reflective of the dearth of research on this topic. Unfortunately, ASD has often been prioritised over ADHD in both research on internalising problems and clinical practice, particularly for anxiety [see16,17]. This has led to major gaps in the literature, which this study has helped to address. Based on the robustness and consistency of our results, we argue that greater emphasis should be placed on the role of ADHD in mental health research and clinical practice on neurodevelopmental conditions. Our finding that ADHD traits dominated ASD traits in predicting internalising disorder symptoms could inform strategies to identify individuals who are at an increased risk of internalising problems. This might allow preventative measures and interventions to be implemented at an earlier age, which, for example, could focus on managing ADHD symptoms for a greater impact on ameliorating internalising problems and improving mental wellbeing in adults. Developmental, longitudinal research will now be required to advance this line of work, for example, to determine the direction of the relationships between neurodevelopmental and internalising symptoms across development. Equally, addressing other gaps in the literature, this study offers novel insight into the link between neurodevelopmental conditions and mental health in adults, a population that is often understudied in the context of ASD and ADHD15. Our findings suggest that the management of ADHD traits in adults, with or without ASD, has potential to reduce internalising problems, which could supplement clinical interventions directly targeting anxiety and depression (e.g., anti-depressant medications, talking therapies), although replication is, of course, now required in clinically diagnosed samples. Given that ADHD/ASD traits share similar correlates with clinical diagnoses [see27], we hypothesise that a similar pattern of results will be found in such research. Indeed, a recent study in clinically diagnosed autistic people found that ADHD traits, compared to autistic traits, were more strongly associated with a lower quality of life50. As quality of life and internalising problems are correlated51,52, we would expect that our pattern of results will, in future, be found in a sample of clinically diagnosed autistic people.
Before these potential suggestions could be translated into practice, there is a need to investigate the mechanisms underlying our results. While the specific neurocognitive processes that underpin our findings are unknown, a possible explanation for this pattern of results may be the differential executive function difficulties that characterise ASD and ADHD. A systematic review by Craig and colleagues53 revealed that, while both neurodevelopmental conditions are associated with executive dysfunction, autistic individuals tend to have more problems with cognitive flexibility and planning, while those with ADHD have more response inhibition difficulties [see also54]. Crucially, such response inhibition problems have also been documented in depression55,56,57 and anxiety58,59,60. We therefore speculate that response inhibition difficulties, which appear to be a cognitive feature of ADHD, may also potentially underpin the stronger association between ADHD traits and internalising problems. The residual link between autism and internalising problems may be driven by other atypicalities often reported in these conditions, such as alexithymia [see61]. Our study was neither designed nor able to test for such mechanisms, and alexithymia has not been investigated in ADHD controlling for ASD, however it provides some direction for future research on the processes underlying differential contributions of ADHD and ASD to poor mental health.
In considering mechanisms to explain our findings, it may also be important to think about the overlap between ADHD, ASD and internalising problems at the genetic level [see15 for a review]. Family studies have consistently shown an increased prevalence of depression in biological relatives of individuals with ADHD and/or ASD [e.g.,62,63] with similar findings for anxiety [e.g.,64,65]. Further, studies identifying common genetic variants (i.e., via genome-wide association studies; GWAS) demonstrate genetic correlations between internalising disorders and both ADHD and ASD [e.g.,66,67]. Interestingly, a recent GWAS of eight major psychiatric conditions68 showed that, whilst ADHD and ASD show similar genetic correlations with depression (0.44 and 0.45, respectively), significant genetic correlations with anxiety were found with ADHD, but not ASD. That is, a pattern of results in a genetic study in line with our findings. Future research could investigate, more directly, whether genetic factors can explain why ADHD traits are more strongly associated with internalising problems than autistic traits.
Further research is also required to address limitations of the present study. First, internalising problems were assessed using brief depression and anxiety measures, which may not provide a comprehensive index of internalising problems. Additionally, several internalising symptoms measured by these brief screeners (e.g., insomnia, restlessness and changes in appetite) overlap with the symptom profiles of ASD and ADHD15, suggesting that some participants may have been scoring on certain items due to their neurodevelopmental traits rather than experiences of anxiety and/or depression. A future replication of our study using a more in-depth measure that overlaps less with items on the ASRS and AQ-Short (e.g., the Depression, Anxiety and Stress Scales;69), may be helpful in testing the validity and replicability of our findings. Second, ethnicity was not considered as we focused on recruiting a sample that was UK representative by age and sex alone. However, given that there may be a link between ethnicity and internalising problems [e.g.,70,71], it will be important to assess whether the dominance relationship observed holds after accounting for participants’ ethnicity and other socio-demographic factors. Finally, while the trait-based approach was an important step in understanding the unique contributions of ASD and ADHD to internalising problems, the study will require a conceptual replication with clinical samples. Such research would benefit from using measures of internalising symptoms that have been validated in this population, which are currently lacking15. This will be difficult to achieve as it is only with the most recent revision of the Diagnostic and Statistical Manual of Mental Disorders1 that individuals can receive concurrent ASD and ADHD diagnoses72. Consequently, recruiting a sufficiently large sample of clinically diagnosed participants, as required to undertake the analyses we have employed, will require considerable resources. Our study hopefully provides the impetus for such research.
Overall, our findings demonstrate that ASD and ADHD uniquely predict internalising problems, such as anxiety and depression. Crucially, ADHD traits were a much more important predictor of internalising problems and our analyses indicated that this relationship is almost certain to occur in the overall population. While further research is necessary to replicate these findings and elucidate mechanisms that underpin the observed relationships, our study provides important new evidence linking neurodevelopmental conditions and mental health in adulthood.