# Genetic disparities in sleep traits and human capital development: A 25-year study in Finnish population-based cohorts

## TL;DR

This study shows that genetic predispositions to certain sleep traits are linked to lower educational attainment, less knowledge-based jobs, and lower income in the Finnish population.

## Contribution

The study demonstrates causal pathways linking genetic sleep traits to human capital outcomes using polygenic indices and instrumental variable analysis.

## Key findings

- Genetic predispositions to insomnia, short sleep, and long sleep are strongly negatively associated with educational attainment and knowledge work.
- These genetic sleep traits are also negatively associated with income, primarily through effects on education and occupation.
- The sleep duration polygenic index was not significantly linked to human capital outcomes.

## Abstract

Sleep supports cognitive performance and recovery, shaping human capital development through education and workplace knowledge application. This study investigates how polygenic indices (PGI) for insomnia (IPGI), short sleep (SSPGI), long sleep (LSPGI), and sleep duration (SDPGI) are associated with educational attainment, occupational group, and income in the Finnish general population.

Genetic and socioeconomic registry data were merged with pooled data from six pentennial (1992–2017) cohorts representative of Finnish regional populations aged 25–64 (N=20 121). Regression models assessed associations between sleep trait PGI and human capital outcomes. In extended regression models, phenotypic sleep traits were treated as endogenous variables—potentially influenced by unobserved confounders—and instrumented with their respective PGI to isolate variation attributable to genetic predisposition.

IPGI, SSPGI, and LSPGI were substantially negatively associated with educational attainment (P<0.001) and selection into knowledge work occupational group (P≤0.005). Their negative association with income (P<0.005) primarily operated through pathways involving education and occupational group. Extended regression models confirmed that these PGI validly predicted their respective phenotypic sleep traits, which, when instrumented, were significantly negatively associated with education and belonging to the knowledge work occupational group, supporting causal pathways linking genetic sleep predispositions to human capital outcomes via phenotypic sleep traits. In contrast, SDPGI—an aggregate proxy for genetically distinct short and long sleep traits—was not significantly associated with any human capital outcome.

Genetic predispositions to insomnia, short sleep, and long sleep were robustly and substantially negatively associated with human capital development. These associations may help to clarify how genetic sleep traits relate to outcomes in work and health contexts.