Real-world big data to define iron deficiency decision limits: implications for cardiovascular health monitoring

Iron deficiency, with or without anaemia, is increasingly recognized as a relevant factor in cardiovascular health. As it is present in more than 60% of CVD patients, it can be an important therapeutic target to improve quality of life and exercise capacity (1). Accurate and standardized definitions of iron deficiency are critical, not only for patient management but also for population-based screening. Blood donors represent a unique population for studying iron homeostasis, as they are generally healthy but at risk of iron depletion due to repeated donations. Yet, definitions of iron deficiency remain imprecise, with current ferritin cutoffs, such as the WHO limit of ferritin < 15 ng/mL, lacking robust physiologic validation.

To derive data-driven ferritin decision limits by quantifying the association between ferritin levels and hemoglobin (Hb) recovery in whole blood donors in The Netherlands.

We analysed routinely collected donor data from the Dutch national blood service collected between 2017 and 2022. The data was collected during 196,595 whole blood donations. Only donors with baseline (pre-donation) Hb levels were included to avoid confounding by prior donations. The association between log10-transformed ferritin and the change in Hb from baseline (ΔHb) was modelled using segmented linear regression, with changepoints estimated via maximum likelihood. Confidence intervals were derived using bootstrapping across male and female subgroups.

There is a linear association between the drop change in Hb since pre-donation screening and log10 ferritin levels, which changes or disappears when a donor recovers to their initial Hb level (Figure 1). For males, Hb recovery is limited when ferritin is below 30.7 ng/mL (95% CI 30.1-31.7), in females this changepoint lies at 26.3 ng/mL (95% CI 24.4-27.5).

Our findings suggest that iron deficiency, as defined by impaired Hb recovery, occurs at ferritin levels that may differ from the current WHO cutoff. However, due to limited standardization of ferritin assays, caution is warranted when generalizing these thresholds to other settings or populations. The development of internationally standardized ferritin reference materials is essential to ensure harmonized clinical decision limits (2). Blood donors, as a generally healthy population, may serve as a valuable reference group for refining iron deficiency definitions in cardiovascular disease prevention and management

Rolling mean of log ferritin