Language deficits in specific language impairment, attention deficit/hyperactivity disorder, and autism spectrum disorder: An analysis of polygenic risk

Abstract

Language is one of the cognitive domains often impaired across many neurodevelopmental disorders. While for some disorders the linguistic deficit is the primary impairment (e.g., specific language impairment, SLI), for others it may accompany broader behavioral problems (e.g., autism). The precise nature of this phenotypic overlap has been the subject of debate. Moreover, several studies have found genetic overlaps across neurodevelopmental disorders. This raises the question of whether these genetic overlaps may correlate with phenotypic overlaps and, if so, in what manner. Here, we apply a genome‐wide approach to the study of the linguistic deficit in SLI, autism spectrum disorder (ASD), and attention deficit/hyperactivity disorder (ADHD). Using a discovery genome‐wide association study of SLI, we generate polygenic risk scores (PRS) in an independent sample which includes children with language impairment, SLI, ASD or ADHD and age‐matched controls and perform regression analyses across groups. The SLI‐trained PRS significantly predicted risk in the SLI case–control group (adjusted R2 = 6.24%; P = 0.024) but not in the ASD or ADHD case‐control groups (adjusted R2 = 0.0004%, 0.01%; P = 0.984, 0.889, respectively) nor for height, used as a negative control (R2 = 0.2%; P = 0.452). Additionally, there was a significant difference in the normalized PRS between children with SLI and children with ASD (common language effect size = 0.66; P = 0.044). Our study suggests no additive common‐variant genetic overlap between SLI and ASD and ADHD. This is discussed in the context of phenotypic studies of SLI and related disorders. Autism Res 2020, 13: 369–381. © 2019 The Authors. Autism Research published by International Society for Autism Research published by Wiley Periodicals, Inc. Lay Summary Language deficits are characteristic of specific language impairment (SLI), but may also be found in other neurodevelopmental disorders, such as autism spectrum disorder (ASD) and attention deficit/hyperactivity disorder (ADHD). Many studies examined the overlaps and differences across the language deficits in these disorders, but few studies have examined the genetic aspect thereof. In this study, we use a genome‐wide approach to evaluate whether common genetic variants increasing risk of SLI may also be associated with ASD and ADHD in the same manner. Our results suggest that this is not the case, and we discuss this finding in the context of theories concerning the etiologies of these disorders.

Quantitative genome-wide association analyses of receptive language in the Danish High Risk and Resilience Study

“…This suggests that the association of this SNP with receptive language is not part of a larger behavioral phenotype encompassing ASD, ADHD, familial risk of schizophrenia or bipolar disorder. In this context, it would be useful to note that a study of polygenic risk which used the same cohort used in this study [40] found that a polygenic risk score (PRS) trained on a previous GWAS of SLI was not predictive of the risk of ASD or ADHD in this cohort, but it was predictive of the risk of SLI. The latter resulted in a Nagelkerke's R 2 (adjusted for prevalence and the proportion of cases in the sample) of > 5%, compared to values of close to 0% for ASD and ADHD (see footnote in the Methods section for more information about this).…”

“…To this end we used the “POE method” as implemented in QUICKTEST v0.99b [ 95 ]. In this analysis, 391 unrelated children were used, the same sample as used in the PRS study [ 40 ] (TROG-2 scores were available for 389 children of those 391 unrelated children). The underlying principle of this approach is the following: assuming that a locus with alleles A and B has an additive effect on a phenotype (the two alleles have different effects on the quantitative trait), the three genotype groups AA, AB and BB will differ in their mean phenotype scores.…”

“…The QC steps for the genetic data are described in detail in a recent study of PRS for language impairment [40]. QC was performed with PLINK v1.9b5.2 [77].…”

Association between genes regulating neural pathways for quantitative traits of speech and language disorders

“…Because of the strong significance of these findings, this suggests that the genetic architecture of communication disorders maybe largely polygenic, which may additionally explain the lack of replication and/or genome-wide significance. While other studies have examined polygenic risk scores associated with language 15,73 , ours is the first to examine polygenic risk associated with other communication endophenotypes. It is noteworthy that our associated SNPs fell outside of gene coding regions but resided in regulatory regions, even having potential regulatory effects themselves.…”

Research in Speech Science and Voice Disorders: The Promise of Modern Genetic Approaches in Improving Clinical Diagnosis and Treatment

“…A GWA study of SLI identified a nonsense mutation in the gene NOP9, which only resulted in association with the disorder if the gene was paternally inherited, while another locus was identified that was only associated if maternally inherited (Nudel et al, 2014). The same group also demonstrated that genetic variants known to cause SLI do not appear to increase risk of having autism spectrum disorder or attention deficit/hyperactivity disorder (Nudel et al, 2020).…”