Heritability of sleep architecture based on home polysomnography

TL;DR

This study shows that certain aspects of sleep, like total sleep time and deep sleep stages, have a genetic component, based on data from over 600 people in Brazil.

Contribution

The first report of heritability of EEG-derived sleep parameters in a large, adult population sample.

Findings

Total sleep time and deep sleep stages (N3) showed significant heritability.

Rapid eye movement sleep and wake after sleep onset showed no significant genetic influence.

Genetic factors contributed to 17% of the variance in the apnea–hypopnea index.

Abstract

We aimed to establish the heritability of polysomnography measures through the analysis of home polysomnography recordings from 648 participants in the Baependi Heart Study, a rural, family‐based, genetically admixed cohort based in the southeast of Brazil. Sleep polysomnography staging variables were computed, and narrow‐sense heritability values were derived. The heritability (h
2) of polysomnography total sleep time was 0.18 ± 0.08 (p = 0.007). Including age and sex did not change the heritability estimates of the model. Wake after sleep onset did not show significant heritability (h
2 = 0.02 ± 0.07, p = 0.39). The unadjusted model for N1 resulted in a heritability estimate of 0.22 ± 0.10 (p < 0.003), and of 0.26 ± 0.10 (p = 0.003) in the adjusted model. Time spent in N2 had an unadjusted heritability of 0.18 ± 0.09 (p = 0.01) and adjusted h
2 of 0.22 ± 0.10 (p = 0.007). The heritability of the total time spent in N3 was 0.35 ± 0.09 (p < 0.001) in the unadjusted and 0.38 ± 0.09 (p < 0.001) when sex, age and age*age were considered in the model. By contrast, no variance in the total time spent in rapid eye movement could be significantly attributed to genetic variance. In terms of the heritability of the apnea–hypopnea index, 17% of its variance could be attributed to genetic factors (0.17 ± 0.08, p = 0.02). This is the first report of heritability of electroencephalographic‐derived sleep parameters from a larger population sample, and the first one performed in a population with a majority above 25 years of age. Our findings indicate the potential feasibility of future genome‐wide association studies of non‐rapid eye movement sleep stages in pooled population samples.