Iron is an essential trace element present in a number of molecular systems and plays a central role in oxygen transport and utilization as well as for mitochondrial function [1]. In clinical terms, prolonged iron deficiency often leads to anemia and subsequent symptoms such as dizziness and fatigue. In humans, intracellular iron is stored as ferritin, and a cut-off value of 30 µg/l is usually used for the diagnosis of an absolute iron deficiency [2]. However, ferritin is an acute phase reactant and its levels increase during inflammatory processes [3]. Transferrin is a negative acute phase reactant, and its levels decrease during inflammatory processes [4]. Transferrin saturation reflects the utilized amount of iron—i.e., the relative amount of transferrin that is loaded with iron [4].

Recently, iron deficiency has been recognized as an independent entity beyond anemia, with prognostic implications in diseased cohorts, mainly cardiovascular ones [5]. Moreover, a new functional definition of iron deficiency, combining low ferritin levels with a reduced transferrin saturation, was proposed and proved strong efficacy in randomized clinical outcome trials of intravenous iron supplementation and prospective studies [6‐9]. Importantly, these trials and studies proved that the prognostic relevance of iron deficiency is irrespective of anemia [10].

However, there is a gap of knowledge on whether iron deficiency is a risk factor in the general population and which definition should be predominantly used for risk estimation.

The aim of this study is to characterize the impact of iron deficiency on all-cause mortality in the non-anemic general population based on two distinct definitions.

The study followed the recommendations of the Declaration of Helsinki and was approved by the ethics committee of the Chamber of Physicians of Rhineland-Palatinate, Germany (reference no. 837.020.07). Written informed consent was obtained from all participants.

We report three major findings: (1) iron deficiency is very common in the apparently healthy, non-anemic general population. A total of 54.5% of participants showed functional iron deficiency at a mean hemoglobin of 14.3 g/dl, while the rate of absolute iron deficiency was 11.8%, at a mean hemoglobin level of 13.4 g/dl; (2) a relevant proportion of subjects develop new iron deficiency during mid-term follow-up. At year 5, proportion of newly diagnosed subjects with functional iron deficiency was 18.5%, while 4.8% of initially non-iron-deficient individuals had developed absolute iron deficiency; (3) iron deficiency is strongly and independently associated with all-cause mortality in the long term. Rate of all-cause mortality was 7.2% (n = 361) after a median follow-up of 10.1 years. After adjustment for hemoglobin and major cardiovascular risk factors, the hazard ratio with 95% confidence interval of the association of iron deficiency with mortality was 1.3 (1.0–1.6; p = 0.023) for the functional definition, and 1.9 (1.3–2.8; p = 0.002) for absolute iron deficiency.

Anemia is the striking symptom of iron deficiency [13]. Therefore, iron deficiency was reduced to this context for decades, and definition of iron deficiency was limited to very low ferritin levels (usually below 30 ng/l). However, recent research indicates that the impact of iron deficiency goes way beyond subsequent anemia [14]. Ferritin is an acute-phase reactant that increases even during subclinical inflammatory conditions, and the true burden of iron deficiency may have been underestimated. Most renowned is the FAIR-HF trial, which evaluated the effect of iron supplementation on a functional endpoint (self-reported Patient Global Assessment, NYHA functional class, and 6-min walk test) as compared to placebo in individuals with chronic heart failure and iron deficiency. Importantly, this trial did not only show a benefit of iron supplementation on the primary endpoint, but also showed that this association was independent of the baseline hemoglobin levels [5]. The definition of iron deficiency in this trial and other cardiovascular trials differed significantly from the low ferritin definition used in previous studies, with transferrin saturation being added as a diagnostic parameter. The use of this definition was based on the finding that most cardiovascular diseases are associated with a systemic inflammatory response which might elevate ferritin levels and therefore obscure the low ferritin definition of iron deficiency [15]. The present study is the first to transfer the new definition of iron deficiency and the emerging knowledge of beneficial effects of supplemental therapy in those deemed functional iron deficient to the general population. Functional iron deficiency seems more accurate in terms of sex-specific prediction, since absolute iron deficiency is much more frequent in menopausal women. Additionally, this study observed an increasing prevalence of functional iron deficiency in older women, whereas prevalence of absolute iron deficiency in women and both forms in men showed a decreasing prevalence with older age. Although it could be speculated that this relates to hormonal changes in the menopause, the exact mechanism is currently not known and should be further investigated. More surprisingly, the association with mortality was very robust and independent of hemoglobin (used as surrogate for anemia). Experimental and epidemiological evidence proves that, beyond hematopoiesis, iron is required for proper immune function and plays a central role in oxygen transport and utilization as well as for mitochondrial function [1]. Iron deficiency has been shown to compromise cell-mediated immunity, decreasing T-cell numbers and proliferative response and potentially reducing macrophage activity [16], which may reduce host capacity to control infection. Nevertheless, whether it has the capacity to challenge other established and emerging inflammatory and immune biomarkers in prediction of primary and secondary risk is subject to future studies.

Despite all efforts and meticulous design of the GHS, the presented study is subject to some limitations. First of all, although the above-mentioned analyses were adjusted for known confounders, there might be other confounders which are not known or were not measured at all. Secondly, the results can only be seen as descriptive due to the observational design and a causation cannot be derived from this. Thirdly, information on iron intake (e.g., nutritional habits) or iron loss (e.g., menstruation bleeding) were not available and could, therefore, not be factored into this analysis. Lastly, the findings were derived from a population based in the central western part of Germany. Therefore, the findings might not be transferred to a developing country or region.

Iron deficiency is very common in the apparently healthy, non-anemic general population and strongly and independently associated with all-cause mortality in the mid to long term.