Polygenic risk scores and early manifestations of attention‐deficit/hyperactivity disorder

Randomiza-tion methods can also complement findings from more standard PRS association analyses to strengthen causal inference, in particular when based on rigorous MR analyses with smaller numbers of highly significant variants using methods designed to account for pleio-tropic bias.

series of multilevel random-effects meta-analyses suggested that pre-school children with current or later-emerging ADHD are likely to experience difficulties in multiple neurocognitive and behavioural functions.
In contrast to the large number of studies investigating early neurocognitive and behavioural precursors of ADHD, relatively few studies have assessed these as intervention targets. The systematic review by Shephard et al. (2022) included 28 randomized controlled trials (RCTs) and 4 nonrandomized trials testing a range of interventions that targeted relevant early precursors of ADHD.
Multilevel random-effects meta-analyses of the included studies suggested that early interventions show some effectiveness in reducing ADHD symptoms and working memory, but their effects on other neurocognitive and behavioural difficulties remains unclear.
Several of the studies included in the above mentioned metaanalyses used a cross-sectional design. Shephard et al. (2022) therefore highlighted that longitudinal designs are needed to establish the predictive value of early neurocognitive and behavioural precursors as well as their potential causal role for ADHD. Another important focus for future research is to include biological markers, like genetic information, that can contribute to the predictive models (Farhat et al., 2020). A closer look at the included studies in the above mentioned systematic review and meta-analysis reveals that few of the available studies incorporated genetic data in the study design.
Such data could be used in prediction modeling research to develop tools for early detection. Genetic data could also be used to strengthen causal inference for observed association precursors/ early markers and ADHD, which is a critical step towards identifying new early intervention targets for ADHD (Taylor, 2021).

POLYGENIC RISK SCORES AND ADHD
Genome-wide association studies (GWAS) have been instrumental for identifying common genetic variants across the genome associated A recent systematic review explored how the ADHD PRS adds to our understanding of ADHD and associated traits (Ronald, de Bode, & Polderman, 2021). The study found that the ADHD PRS was associated with not only diagnosed ADHD but also with elevated ADHD trait scores, more externalizing behaviours, impaired working memory, lower educational attainment, reduced brain volume, higher BMI and reduced SES. The investigators also highlighted emerging evidence that the signal from the ADHD PRS remained significant after controlling for other PRSs, and that the ADHD PRS primarily appeared to associate with ADHD-relevant phenotypes, which indicates some level of specificity in relation to the genetics of other mental health conditions. Even though the number of PRS studies have increased substantially in recent years, Ronald et al. (2021) and others (Larsson, 2021) have noted that there is still a lack of longitudinal ADHD PRS research. For example, more research is needed on how PRS associates with precursors and early markers of ADHD.  Second, the study by Riglin et al. (2022)    Second, because of a lack of ancestral diversity in currently available GWAS, individuals were excluded from the analyses on the basis of non-European ancestry, which limits generalisability. Investments are needed to perform large-scale GWASs in globally diverse populations. With more diverse samples, the genetic architecture of ADHD and other traits and conditions will be understood better globally and clinical applications of PRS will be more equally applicable across population groups.

LIMITATIONS AND FUTURE DIRECTIONS
Third, while PRSs are useful for studying correlations, they cannot be taken as evidence of causality. Riglin et al. (2022) correctly point out that they cannot differentiate whether motor development and temperament lie on the causal pathway between genetic liability to neurodevelopmental conditions and later neurodevelopmental conditions, if these associations arise independently as the result of (horizontal) pleiotropy, or whether the associations reflect an early manifestation of neurodevelopmental conditions. Implementation of sophisticated study designs and analyses will probably help uncover the mechanisms underlying identified PRS associations. As suggested by Riglin et al. (2022), research could investigate whether parental risk alleles for neurodevelopmental conditions (e.g., ADHD, autism and schizophrenia) are associated with differences in ratings of their children and whether un-transmitted risk alleles using parent-child trios contribute to "genetic nurture" effects. Mendelian Randomization methods can also complement findings from more standard PRS association analyses to strengthen causal inference, in particular when based on rigorous MR analyses with smaller numbers of highly significant variants using methods designed to account for pleiotropic bias.

CONFLICTS OF INTEREST
Henrik Larsson reports receiving grants from Shire Pharmaceuticals; personal fees from and serving as a speaker for Medice, Shire/Takeda Pharmaceuticals and Evolan Pharma AB; and sponsorship for a conference on attention-deficit/hyperactivity disorder from Shire/ Takeda Pharmaceuticals and Evolan Pharma AB, all outside the submitted work. Henrik Larsson is Editor-in-Chief of JCPP Advances.