Recent genome-wide association studies (GWASs) have revealed new genetic variants behind self-reported individual circadian preference, a distinct biological trait that is fairly stable during adulthood. In this study we analyze whether these genetic variants associate with objectively measured sleep timing from childhood to adolescence, over a nine-year period, with self-reported circadian preference during late adolescence.

The participants (N = 100, 61% girls) came from a community cohort from Finland born in 1998. Sleep midpoint was measured with actigraphy at 8, 12 and 17 years. Circadian preference was self-reported at the age of 17 years. Single nucleotide polymorphisms (SNPs) were extracted at 12 years of age from the Illumina OmniExpress Exome 1.2 bead array data. Weighted polygenic risk scores (PRSs) were calculated based on top SNPs from a recent GWAS for morningness–eveningness in an adult population.

The PRS for circadian preference towards morningness was associated with earlier sleep midpoint from childhood to adolescence. When the time points were analyzed separately, the association between genetic tendency towards morning preference and earlier sleep midpoint was strongest among the 17-year-olds. Furthermore, the shift towards later sleep rhythm from early to late adolescence was milder for those with a higher PRS for morning preference. PRS for morning preference was also associated with self-reported circadian preference towards morningness in late adolescence.

Our results suggest that genetic variants found for circadian preference in adults are already associated with objective sleep timing during childhood and adolescence, and predict individual developmental sleep trajectories from childhood onwards.