Extended Data Fig. 5: Insomnia and hypersomnia are associated with the time to incident of the 110 disease endpoints and mortality in the TriNetX database.

Using insomnia and hypersomnia diagnoses, we replicated the significant associations shown in Fig. 3 within the UK Biobank cohort. Disease endpoints that remained significant after Bonferroni correction (two-sided P-value < 0.05/111, accounting for 110 disease endpoints and all-cause mortality) are indicated by coloured shapes, while non-significant associations are shown in grey. One additional disease category was also tested for all types of dementia. Ctrl: health control. Error bars represent the 95% confidence interval (CI) of the hazard ratio (HR) estimate, which is represented by a central shape.