Etiological distinction of working memory components in relation to mathematics

• We examined the genetic and environmental underpinnings of working memory and math.
• Genetic factors accounted for overlap of visuo-spatial sketchpad and math measures.
• Genetic influences were shared between phonological loop and math story problems.
• Visuo-spatial, but not phonological component, had genetic effects distinct from g.
• Shared environmental influences on math were not associated with working memory.

Working memory has been consistently associated with mathematics achievement, although the etiology of these relations remains poorly understood. The present study examined the genetic and environmental underpinnings of math story problem solving, timed calculation, and untimed calculation alongside working memory components in 12-year-old monozygotic (n = 105) and same-sex dizygotic (n = 143) twin pairs. Results indicated significant phenotypic correlation between each working memory component and all mathematics outcomes (r = 0.18–0.33). Additive genetic influences shared between the visuo-spatial sketchpad and mathematics achievement were significant, accounting for roughly 89% of the observed correlation. In addition, genetic covariance was found between the phonological loop and math story problem solving. In contrast, despite there being a significant observed relationship between phonological loop and timed and untimed calculations, there was no significant genetic or environmental covariance between the phonological loop and timed or untimed calculation skills. Further analyses indicated that genetic overlap between the visuo-spatial sketchpad and math story problem solving and math fluency was distinct from general genetic factors, whereas g, phonological loop, and mathematics shared generalist genes. Thus, although each working memory component was related to mathematics, the etiology of their relationships may be distinct.