In this issue of Clinical Chemistry, Ellervik et al. (1 ) report relationships between baseline ferritin concentrations and risk of all-cause and cause-specific mortality in 8988 individuals enrolled in the Copenhagen City Heart Study. Median follow-up was 23 years, during which 6364 individuals died. Multifactorially adjusted hazard ratios for total mortality in individuals with ferritin values below vs above 200 g/L showed significantly lower mortality overall (P ϭ 0.0008) and separately for men (P ϭ 0.02) and women (P ϭ 0.03) with lower ferritin concentrations. Cause-specific mortality below vs above the 200 g/L threshold was significantly lower for cancer (P ϭ 0.005), metabolic disease (P ϭ 0.002) and cardiovascular disease (P ϭ 0.00006). A highly significant progressive decrease in median survival occurred with increasing ferritin concentrations. The hazard ratio for total mortality increased by 13% for each 100 g/L increase in ferritin concentration. The median survival was 79 years with concentrations Ͻ200 g/L, 76 years with ferritin concentrations between 200 and 399 g/L, 72 years with ferritin concentrations between 400 and 599 g/L, and 55 years with ferritin concentrations in excess of 600 g/L. A stepwise increase in mortality with increasing ferritin concentrations was found for each disease category. Adjusted hazard ratios for increased disease-specific mortality for ferritin concentrations above 600 g/L vs below 200 g/L were significant for malignant (P ϭ 0.01), endocrine (P ϭ 0.0001), and cardiovascular (P ϭ 0.01) disease. The conclusion was that "increased ferritin concentrations represent a biological biomarker predictive of early death in a dose-dependent linear manner in the general population." Increased ferritin concentrations were interpreted by the authors "as a biological marker of pathogenic processes that reflects severity and presence of a variety of disease states leading to premature death." They stated that "causality may be difficult to claim in this study, since, unlike transferrin saturation, increased ferritin is associated with many diseases other than hereditary iron overload."Readers may be challenged by the interpretation of ferritin concentrations reported from this study. Ferritin values in the range of 200 g/L, which were found to be disadvantageous, are common in practice. Reluctance to assign cause-and-effect relationships, or to test the relationship between ferritin concentrations and outcomes at all, may be based on the assumption that high ferritin values might in some sense be "nonspecific" (perhaps "acute phase") in this context. It might also be thought that high ferritin concentrations occur too commonly and in too many different diseases to be useful in terms of determining diagnosis or prognosis or designing treatment strategies. The seemingly discontinuous association between increased ferritin concentrations and clinical outcomes for commonly observed increases in ferritin vs the larger increases seen with hereditary iron overload may lead to the presumptio...