In addition, we applied a Cox’s proportional hazard model to estimate hazard rate ratios and absolute risk of developing 1) a second episode of depression, 2) anxiety, 3) bipolar disorder, 4) schizophrenia and 5) SUD, among individuals already being diagnosed with their first episode of depression. Analyses were stratified by PRS deciles of the phenotypes described above and using functions from the R-packages survival (https://CRAN.R-proiect.org/package=survival)133,134. while correcting for batch (iPSYCH2012 and iPSYCH2015i) and PCs. The Cox’s proportional hazard models were stratified into three groups based on 1st, 2nd-9th and 10th decile. The middle 2nd–9th decile group, i.e. the 80% prediction interval of the PRS, was used as reference. Two-sided 95% confidence intervals were obtained for the stratified absolute risk trajectories using Cox’s proportional hazards model with covariates135, taking the mean of each covariate (PC 1-5 and iPSYCH sample) and using the survfit.coxph function implemented in the survival R-package. All Cox’s analyses were conducted for both sexes combined and for females and males separately. In addition, HRR and absolute risk of developing 1) anxiety, 2) bipolar disorder, 3) schizophrenia, and 4)
