Music and Genetics

This review addresses how genetic and environmental factors shape individual differences in music-related skills, engagement, and achievement. It outlines the historical development of behavioral genetics, emphasizing twin and molecular genetic methods, and applies these approaches to music as a culturally acquired yet potentially biologically grounded behavior. The goals are to summarize evidence for heritability of musical traits, evaluate emerging molecular genetic findings, and demonstrate how genetically informative designs can illuminate questions of causality, sensitive periods, gene–environment interaction, and gene–environment correlation in the development of musical expertise. Understanding these dynamics is important for refining theories of expertise (e.g., beyond deliberate practice), informing neuroscience and psychology of skill acquisition, and guiding future genomic research in music.

The review synthesizes decades of twin research showing that nearly all complex human traits have a genetic component and applies this to musical phenotypes. Twin studies report heritability for a broad range of music-related outcomes: objectively measured abilities (melody, rhythm, pitch discrimination; singing accuracy), self-reported aptitude and achievement, engagement (practice, listening), and specialized traits (absolute pitch, instrument/genre choice). Heritability estimates vary with age, sex, rater, and measurement modality, with higher genetic influences typically for aptitude than for engagement or achievement, higher in men than women, and increasing with age. The review also covers early molecular work (linkage, candidate genes) with replication challenges, and more recent GWAS/post-GWAS advances, notably a large GWAS of self-reported beat synchronization identifying 69 loci and polygenic scores that predict objective rhythmic and broader musical measures. It highlights pleiotropy and polygenicity, and presents post-GWAS approaches (polygenic scores, GREML) linking genomic variation to musical phenotypes. Finally, it reviews designs leveraging genetically informative data to examine causality and gene–environment interplay, including co-twin control, moderator twin models for G×E, PGS-based G×E, and within-family PGS analyses for rGE.

This is a narrative review integrating findings from twin and molecular genetic studies of music-related traits. The authors: (1) summarize classical twin designs partitioning variance into additive genetic, shared environmental, and unique environmental components, including structural equation modeling; (2) review molecular approaches from linkage and candidate gene studies to GWAS and post-GWAS methods (polygenic scores, GREML); and (3) illustrate causal inference and gene–environment interplay using genetically informative designs: - Co-twin control: comparing monozygotic twins discordant for age-of-onset of music training to assess causality independent of familial confounding. - Moderator twin model (Purcell): testing whether childhood musical enrichment moderates the genetic and environmental components of adult music achievement (G×E). - Polygenic score (PGS) interaction: testing whether associations between music engagement and mental health vary by genetic liability to depression/bipolar disorder (G×E with PGS). - Within-family PGS analyses in dizygotic twins: comparing within- vs between-family prediction to assess passive rGE and familial confounding in associations of a beat-synchronization PGS with musical outcomes and enriched musical environments. Empirical results summarized are drawn from large twin registries (e.g., Swedish Twin Registry), online cohorts, and 23andMe-based GWAS, with objectively measured musical tests (e.g., Swedish Musical Discrimination Test) and self-reports.

- Twin heritability of music-related traits: ranges from 0% to 86%, averaging ~42% across studies. Genetic influence tends to be higher for aptitude/skills than engagement or achievement, higher in men than women, and increases with age. - Objectively measured abilities: pitch recognition heritability ~70–80% (Drayna et al., 2001); melody discrimination ~50%, rhythm ~59%, pitch discrimination ~12% (males) and ~30% (females), with common environment influencing male pitch discrimination ~38% (Mosing et al., 2014; Ullen et al., 2014). Singing ability: heritability 40.7% and shared environment 37.1% (Yeom et al., 2022). - Engagement and achievement: adolescent instrument engagement heritability 78% and singing 43% (Gustavson et al., 2021); parental-rated child engagement heavily shared environmental (e.g., 92% at age 10) (Gustavson et al., 2023). Music achievement heritability: overall ~26% with 61% shared environment (Hambrick & Tucker-Drob, 2015); sex differences: men 55% genetic vs women 13%, with shared environment 14% in men and 40% in women (Wesseldijk et al., 2019). Time spent listening to music: heritability 16–46% in adults and 26% in children (Wesseldijk et al., 2020; Gustavson et al., 2023). Age of onset of training: ~39% genetic (Wesseldijk et al., 2020). Practice time itself shows substantial heritability (40–70%) (Mosing et al., 2014; Hambrick & Tucker-Drob, 2015). - Causality: Associations of practice with musical aptitude and of practice with intelligence are largely explained by shared genetic/familial factors; no evidence of a causal effect of practice on aptitude or intelligence in twin designs (Mosing et al., 2014; 2016). Early age-of-onset relates to higher skill, but co-twin control shows association is fully explained by familial factors, not a direct causal effect (Wesseldijk et al., 2020). - Molecular genetics: Early linkage/candidate findings implicate regions on 8q (absolute pitch/music perception), 4q (music aptitude, pitch accuracy) and genes such as AVPR1A, SLC6A4, though replication is limited. A large GWAS of self-reported beat synchronization (N≈606,825; 555,660 yes vs 51,165 no) identified 69 loci, indicating high polygenicity and pleiotropy (Niarchou et al., 2022). PGS derived from this GWAS predicts objective rhythm perception and other musical abilities and behaviors (practice, listening, achievement, music-domain flow, dance), with no associations to non-music controls, supporting discriminant validity (Wesseldijk et al., 2022). GREML estimates: SNP heritability ~31% for rhythmic perception and ~12% for self-reported music engagement (Gustavson et al., 2023). - Gene–environment interplay: Moderator twin models show that musically enriched childhood environments are associated with higher adult music achievement and higher heritability of achievement (G×E; Wesseldijk et al., 2019). PGS × environment analyses of music engagement and mental health found higher depressive symptoms among those who practice/play more, but no significant interaction with depression/bipolar PGS; power for G×E remains a limitation (Wesseldijk et al., 2023). PGS predicts both musical outcomes and enriched musical childhood environments, consistent with rGE; within-family analyses suggest limited passive rGE inflation, pointing to active/reactive rGE (Wesseldijk et al., 2022).

The review demonstrates that musicality, long thought to be primarily shaped by environmental exposure and practice, has a substantial genetic basis. Twin evidence clarifies that genetic factors contribute to diverse musical outcomes and that their influence varies by age, sex, and phenotype type. Molecular findings are beginning to map specific genomic variation relevant to musical traits, particularly rhythmic synchronization, with polygenic architectures mirroring other complex behaviors. Together, these results challenge strong causal interpretations of training effects derived from observational associations. Co-twin control analyses and genetic covariance findings indicate that observed links between practice and skill, and practice and intelligence, are largely attributable to familial and genetic confounding rather than direct causal effects. At the same time, gene–environment interplay is central: enriched musical environments amplify genetic variance in achievement, and individuals’ genetic propensities correlate with the environments they experience, via passive, reactive, or active rGE. The integration of twin and genomic methods offers powerful tools to distinguish causation from correlation and to quantify how environmental contexts modulate genetic influences, advancing theories of expertise and informing broader behavioral science questions about sensitive periods, enrichment, and health.

Music-related traits and behaviors show robust genetic influences across a range of phenotypes, with average heritability around 42%, greater genetic contributions for aptitude versus engagement/achievement, sex differences favoring higher heritability in men, and increasing genetic influence with age. Molecular genetics is beginning to identify specific loci, most notably 69 loci for beat synchronization, and polygenic scores now show predictive validity for multiple musical outcomes. Genetically informative designs reveal that associations often reflect familial/genetic confounding and that gene–environment interplay, including G×E and rGE, is integral to expertise development. Future work should expand well-powered GWAS across diverse, harmonized musical phenotypes, refine objective measurements of multifaceted musical abilities, and increasingly apply within-family and longitudinal genetically informed designs to clarify causality and mechanisms. This integrated approach will continue to illuminate how biology and environment jointly shape culturally acquired behaviors like music.

- Many early molecular findings (linkage/candidate gene studies) were underpowered and prone to non-replication; effect sizes are small and require large samples. - Measurement limitations: musicality is multifaceted; several studies rely on self-report (e.g., beat synchronization item), parental ratings (susceptible to rater bias), or proxy measures, which can affect heritability estimates and validity. - Power constraints for detecting G×E interactions are substantial; current polygenic scores explain limited variance, reducing sensitivity to interaction effects. - Generalizability: many samples are WEIRD populations; replication across diverse ancestries and cultural contexts is needed. - Review-based synthesis depends on available literature; heterogeneity in age, sex, measures, and designs complicates direct comparison of effect sizes. - Objective assessments of complex musical behaviors (e.g., expressivity, improvisation) remain challenging, potentially underrepresenting important facets in genetic studies.