Elevated serum ferritin levels and infection-related mortality in maintenance hemodialysis patients: 10-year outcomes of the Q-Cohort study

Background Serum ferritin guides iron therapy in hemodialysis patients but also reflects inflammation and systemic illness. We examined the association between baseline serum ferritin and infection-related mortality in maintenance hemodialysis patients. Methods We analyzed 3529 participants in the Japanese Q-Cohort Study. The primary analysis categorized serum ferritin into quartiles. Clinically motivated supplementary analyses used clinical categories, a 300 ng/mL cutoff, and log-transformed ferritin. Infection-related mortality was evaluated using cumulative incidence functions, cause-specific Cox models, Fine–Gray competing-risk models, restricted cubic splines, and sensitivity analyses including C-reactive protein adjustment, adjustment for baseline vitamin D receptor activator use, 3-year restricted follow-up, and a 180-day landmark analysis. Results During a median follow-up of 3199 days, 1748 deaths occurred, including 451 infection-related deaths. In the primary quartile analysis, Q4 was associated with infection-related mortality in unadjusted and age- and sex-adjusted Cox models, but the association was attenuated after multivariable adjustment (model 3: HR 1.32, 95% CI 0.98–1.76; P = 0.065). The corresponding Fine–Gray estimate was also attenuated (sHR 1.23, 95% CI 0.91–1.67; P = 0.182). In supplementary analyses, ferritin ≥ 300 ng/mL was associated with infection-related mortality in the Cox model (HR 1.35, 95% CI 1.09–1.69; P = 0.007) and with all-cause mortality, whereas the Fine–Gray estimate was weaker (sHR 1.20, 95% CI 0.96–1.52; P = 0.114). C-reactive protein adjustment attenuated the estimates, whereas adjustment for baseline vitamin D receptor activator use did not materially change them. The 3-year restricted follow-up and 180-day landmark analyses were broadly consistent with the main findings. Conclusions Higher baseline ferritin was associated with infection-related mortality in several cause-specific Cox analyses, although the primary multivariable quartile and competing-risk results were attenuated. Baseline ferritin should be interpreted as a prognostic marker measured at study entry, not as a stable long-term exposure or a definitive threshold for infection-specific risk or iron overload.