Distinct features of CH and anemia predict for OS. (A) Kaplan-Meier curves for OS, stratified according to the presence of anemia and CH (n = 1298). (B) Forest plot for the risk for death for cases with anemia (n = 635) and control individuals (n = 663) stratified according to the presence of CH. Cox proportional hazards regression included age and sex as covariates. Absence of CH was used as a reference for both cohorts. Circles indicate the HR for each cohort, with horizontal lines corresponding to 95% CIs. (C) Kaplan-Meier curves for OS in the anemia cohort (n = 635) according to largest clone size detected (<5% or ≥5%) or the absence of CH. (D) Kaplan-Meier curves for OS in the anemia cohort (n = 635), according to the number of mutated genes or the absence of CH. (E) Cumulative incidence graphs for mortality from hematological malignancies, stratified according to the presence of anemia and CH (n = 1298). Data on reported primary cause of death were obtained by linkage to the national registry of death statistics. Results for this analysis are based on calculations by the authors using nonpublic microdata from Statistics Netherlands. Colors indicate death from hematological malignancies (red) or other causes (orange).