Study Design and Genetic Instruments
The study design overview and the assumptions of an MR study are shown in Fig. 1 . Genetic instruments for education [22], coffee and tea consumption [23], relative fat intake [24], BMI [25], physical activities (including three physical activity phenotypes: overall acceleration average, self-reported moderate-to-vigorous physical activity (MVPA) and vigorous physical activity (VPA) [26], smoking initiation [27], lifetime smoking index (taking into account smoking status as well as smoking duration, heaviness, and cessation in ever smokers) [28], alcohol drinking [27], sleep duration [29] and insomnia [30] were selected at genome-wide significance threshold (P < 5×10–8) from corresponding genome-wide association (GWAS) studies. Detailed information on used GWAS studies is presented in Table 1 . Linkage disequilibrium between single-nucleotide polymorphisms (SNPs) was assessed based on the European population of 1000 Genomes reference panel. SNPs without linkage disequilibrium (defined by r2 < 0.01 and clump distance >10,000 kb) were used as instrumental variables.