Fig. 1: Music reward sensitivity is substantially heritable.

a Saturated model (including age as a covariate; not shown) used for assumption testing and comparisons of CTD-informed models fit; for simplicity, only one group (MZ women) is shown. b Twin pair correlations (r_twin) grouped by zygosity and sex (w women, m men, os opposite-sex, see Table 1, Music reward sensitivity (n pairs), for the sample size of complete twin pairs grouped by zygosity); note that MZ twin pairs are more than twice as similar in their music reward sensitivity as DZ twin pairs. Data are presented as correlation estimates with 95% confidence intervals (CI). c The ADE model; note that we identified only additive genetic (A) and non-common environmental (E) components as significant contributors to music reward sensitivity variability (see Supplementary Fig. 3 for the complete diagram). r_α is the expected additive genetic relationship, and r_δ is the expected dominance genetic relationship between pairs (i.e., r_α = 1 or 0.5, r_δ = 1 or 0.25, for MZ and DZ, respectively). d Estimated variance components from the final AE model indicated substantial heritability for music reward sensitivity. The left bar plot shows the estimates obtained from the full ADE model. Model estimates are provided as Source Data. Notes on structural equation models: For simplicity, age is not included in the graphical representation of the model but is included as a covariate; Squares or rectangles represent the measured phenotypes; Circles the latent variable; Double-headed arrows connecting circles with themselves, the variances associated with the latent components; double-headed arrows connected between circles covariances; the triangle indicates the unit vector regressed on the phenotype, which corresponds to the sex-specific phenotypic means, M_sex (already adjusted for age); greyed elements, the component dropped after model comparison. A additive genetic, D dominance genetic, E non-common environmental.