Increased Plasma Ferritin Concentration and Low-Grade Inflammation-A Mendelian Randomization Study

Research output: Contribution to journalJournal articleResearchpeer-review

Ingrid W. Moen, Helle K. M. Bergholdt, Thomas Mandrup-Poulsen, Børge G. Nordestgaard, Christina Ellervik

BACKGROUND: It is unknown why increased plasma ferritin concentration predicts all-cause mortality. As low-grade inflammation and increased plasma ferritin concentration are associated with all-cause mortality, we hypothesized that increased plasma ferritin concentration is genetically associated with low-grade inflammation.

METHODS: We investigated whether increased plasma ferritin concentration is associated with low-grade inflammation [i.e., increased concentrations of C-reactive protein (CRP) and complement component 3 (C3)] in 62537 individuals from the Danish general population. We also applied a Mendelian randomization approach, using the hemochromatosis genotype C282Y/C282Y as an instrument for increased plasma ferritin concentration, to assess causality.

RESULTS: For a doubling in plasma ferritin concentration, the odds ratio (95% CI) for CRP ≥2 vs <2 mg/L was 1.12 (1.09-1.16), with a corresponding genetic estimate for C282Y/C282Y of 1.03 (1.01-1.06). For a doubling in plasma ferritin concentration, odds ratio (95% CI) for complement C3 >1.04 vs ≤1.04 g/L was 1.28 (1.21-1.35), and the corresponding genetic estimate for C282Y/C282Y was 1.06 (1.03-1.12). Mediation analyses showed that 74% (95% CI, 24-123) of the association of C282Y/C282Y with risk of increased CRP and 56% (17%-96%) of the association of C282Y/C282Y with risk of increased complement C3 were mediated through plasma ferritin concentration.

CONCLUSIONS: Increased plasma ferritin concentration as a marker of increased iron concentration is associated observationally and genetically with low-grade inflammation, possibly indicating a causal relationship from increased ferritin to inflammation. However, as HFE may also play an immunological role indicating pleiotropy and as incomplete penetrance of C282Y/C282Y indicates buffering mechanisms, these weaknesses in the study design could bias the genetic estimates.

Original languageEnglish
JournalClinical Chemistry
Volume64
Issue number2
Pages (from-to)374-385
Number of pages12
ISSN0009-9147
DOIs
Publication statusPublished - 2018

    Research areas

  • Journal Article

ID: 189626074