(A) Violin plot of plasma GDF15 concentrations across different cohorts. P < 0.0001 by 1-way ANOVA with Tukey’s post hoc test. (B) ROC curves comparing GDF15 levels between a cohort of individuals acting as healthy controls (Ctrl) and other specified cohorts. (C) Specificity% (95% CI) at 100% sensitivity for distinguishing the control cohort from other specified cohorts. (D) ROC curves for distinguishing PVOD from individual cohorts. (E) ROC curves comparing PVOD with combined cohorts: IPAH, HPAH, and Other PH as well as IPAH and HPAH. (F) Specificity% (95% CI) at 100% sensitivity for distinguishing PVOD from individual or combined cohorts. Numbers associated with individual cohorts indicate the number of individuals within that cohort. (G) Pearson correlation between age-adjusted GDF15 and 6MWD in a combined cohort: PVOD, IPAH, and HPAH. (H–M) Standardized distributions of diagnostic parameters: age-adjusted GDF15 (H), N-terminal pro-B-type natriuretic peptide (NT-proBNP) (I), pulmonary vascular resistance (PVR) (J), mean pulmonary arterial pressure (mPAP) (K), cardiac index (CI) (L), and 6MWD (M) in PVOD, IPAH, and HPAH. Left: Individual values normalized to the median (y = 0). Right: Median values with 95% CI. (N) Cox proportional hazards model showing the association between plasma GDF15 and transplant-free survival in the combined cohort, PVOD, IPAH, and HPAH, adjusted for age and sex. Statistical analysis was performed using C statistics to calculate the AUC, followed by Z tests for pairwise comparisons (B, D, and E). Mixed-effects models with Tukey’s multiple comparisons test were used for analyses shown in panels H and I. P value < 0.05 was considered significant.