Dyslexia and developmental language disorders are significant learning difficulties with a poorly understood genetic basis, particularly in non-European populations. Most genetic studies have focused on European populations, leaving a gap in our understanding of literacy phenotypes in other languages and ancestries. This study addresses this gap by conducting a GWAS on Chinese and English language phenotypes in Hong Kong Chinese bilingual children. The high heritability of literacy and language skills, coupled with the polygenic nature of these traits, necessitates large-scale studies to identify the specific genes and variants involved. Previous GWAS on reading and language phenotypes in European populations have employed different approaches, some focusing on developmental dyslexia as a binary outcome (case-control studies), others treating reading and language abilities as continuous traits. However, these studies often focused on limited numbers of phenotypes or specific domains of language abilities, potentially hindering a comprehensive understanding of the underlying biological mechanisms. Furthermore, studies on the genetic architecture of Chinese literacy and English as a second language (ESL) in Chinese populations are scarce. The current study aims to comprehensively analyze a wide range of literacy and language-related phenotypes (34 in total) in Hong Kong Chinese bilingual children to uncover the genetic basis of these skills in this understudied population. This broader approach aims to maximize the chance of discovering novel genetic associations and provide a valuable resource for future research.

Existing literature highlights the heritability of literacy and language skills, indicating a significant genetic component influencing individual differences in these abilities. However, the specific genes and variants contributing to these complex traits remain largely unknown, likely due to the complexity of the phenotypes and the challenges of assembling sufficiently large and diverse samples. Previous GWAS studies conducted primarily in European populations have identified some associated loci, but these findings may not generalize to other ancestries due to differences in allele frequencies and linkage disequilibrium structures. Furthermore, few GWAS have explored the genetic basis of Chinese literacy or English as a second language (ESL) within a Chinese population. The unique linguistic environment of Hong Kong, with Cantonese as the native language and English as a second language, provides a valuable opportunity to study bilingual language development and its genetic underpinnings. This study builds on previous research by examining a more extensive range of phenotypes, including both Chinese and English language skills, in a well-characterized sample of Hong Kong Chinese children.

This study recruited 1048 typically developing Hong Kong Chinese children aged 5–12 years, including both twins (274 MZ and 350 DZ pairs) and singletons (424). All children were Cantonese-English bilinguals. A comprehensive battery of cognitive and literacy tests was administered, measuring 34 reading/language-related phenotypes in both Chinese and English. Genotyping was performed using the Human Infinium OmniZhongHua-8 v1.3 Beadchip, followed by quality control and imputation using the 1000 Genomes Phase 3 v5 reference panel. Genome-wide association studies (GWAS) were conducted using GEMMA, employing a univariate linear mixed model to account for relatedness and population stratification. Association tests were performed at the single-variant, gene, and pathway levels. Gene-based analyses were carried out using MAGMA, S-Predixcan, and S-Multixcan. Pathway enrichment analysis was performed using GAUSS. Polygenic risk score (PRS) analysis was conducted using PLINK and SBayesR to investigate the genetic overlap of these phenotypes with other neuropsychiatric disorders, cognitive performance (CP), and educational attainment (EA). Several other relevant GWAS on dyslexia and language phenotypes (Doust et al., Wang et al., and Eising et al.) were used for genetic overlap/replication and PRS analyses.

The SNP-based GWAS identified 5 independent loci reaching genome-wide significance (p < 5e-08), associated with various language/literacy traits including Chinese vocabulary, character and word reading, digit rapid naming, and English lexical decision. These loci contained genes previously implicated in educational attainment and neuropsychiatric phenotypes (e.g., MANEA and PLXNC1). S-Predixcan and S-Multixcan analyses identified significant associations between genetically regulated expression (GRex) and phenotypes, highlighting genes like DUS3L (associated with English word reading) and HSD3B7 (associated with English vocabulary). Gene-based tests (MAGMA) revealed significant associations with several genes, including KCNC1 (associated with pure copying) and CATSPERD (associated with English word reading). Pathway enrichment analysis (GAUSS) identified significant associations with several pathways, including the RNA polymerase III transcription pathway (associated with Chinese word order) and the P2Y receptors pathway (associated with Chinese vocabulary knowledge). PRS analysis showed the most consistent and significant polygenic overlap with EA and CP, especially for English literacy skills. Analyses exploring genetic overlap with other GWAS (Doust et al., Wang et al., and Eising et al.) revealed some evidence of shared genetic signals, although discrepancies existed potentially due to sample size, ethnic differences, and phenotypic heterogeneity.

This study provides novel insights into the genetic architecture of Chinese and English language abilities in a Hong Kong Chinese population. The identification of several genome-wide significant loci and the substantial polygenic overlap with EA and CP highlight the complex interplay of genetic factors influencing these skills. The stronger association of EA and CP PRS with English language phenotypes compared to Chinese language phenotypes suggests potential differences in the underlying genetic mechanisms. The observed associations between some language phenotypes and PRS of neuropsychiatric disorders, particularly ASD, warrant further investigation. While some evidence of genetic overlap with previous GWAS on dyslexia and language abilities was observed, discrepancies might be due to limitations such as sample size, ethnic differences, and phenotypic heterogeneity. The study underscores the importance of conducting large-scale GWAS across diverse populations to fully understand the genetic basis of complex traits like language abilities.

This GWAS provides one of the first comprehensive investigations into the genetic architecture of both Chinese and English language abilities in a Hong Kong Chinese population. Several novel genetic loci were identified, along with associated genes and pathways. Significant polygenic overlap with educational attainment and cognitive performance was observed, particularly for English literacy. While the findings are promising, the modest sample size necessitates replication in larger and more diverse samples. Future studies should incorporate larger sample sizes, investigate rare variants, and explore gene-environment interactions to further elucidate the genetic basis of language abilities. This research can inform the development of targeted interventions for children with language learning difficulties.

The relatively modest sample size of this study may have limited the power to detect variants with small effects. The findings might not fully generalize to other populations due to the focus on a specific bilingual Hong Kong Chinese sample. The GWAS summary statistics for CP, EA, and some neuropsychiatric disorders were primarily derived from European populations, which could have impacted the accuracy of the genetic overlap analysis. Additionally, the reliance on self-reported diagnoses in some external datasets could introduce heterogeneity. Further replications in independent samples are crucial to confirm the findings.