Abstract
BACKGROUND/OBJECTIVES: Iron is necessary for bodily function, but abnormal levels can increase the risk of chronic diseases. Studies of leukocyte telomere length suggest blood iron levels are positively associated with cellular senescence and accelerated aging. However, associations between blood iron and more robust metrics of biological aging, specifically those based on DNA methylation, have not been examined. METHODS: In a random sample of women from the Sister Study (n = 1260) with measured serum iron (ferritin, iron, transferrin saturation), we used linear regression models to assess cross-sectional associations between standardized serum iron and three methylation-based biological aging metrics (GrimAgeAccel, PhenoAgeAccel, and DunedinPACE), with and without adjustment for smoking, alcohol, menopause status, education, time since menopause, exercise, and diet. RESULTS: In adjusted models, a one standard deviation increase in serum ferritin was positively associated with higher standardized levels of DunedinPACE, GrimAgeAccel, and PhenoAgeAccel (DunedinPACE: 0.05, (0.00, 0.10); PhenoAgeAccel: 0.06 (0.00, 0.11); GrimAgeAccel: 0.06 (0.01, 0.11)). In contrast, higher serum iron and transferrin saturation were inversely associated with the biological aging metrics (serum iron, DunedinPACE: -0.02, (-0.07, 0.03); PhenoAgeAccel: -0.04 (-0.10, 0.01); GrimAgeAccel: -0.05 (-0.10, -0.01); transferrin saturation (DunedinPACE: -0.01, (-0.06, 0.05); PhenoAgeAccel: -0.01 (-0.06, 0.05); GrimAgeAccel: -0.05 (-0.10, -0.01))). CONCLUSIONS: The positive association with ferritin is consistent with the proposed role of oxidative stress in accelerated aging associated with high iron exposure. However, the observed inverse associations with serum iron and transferrin saturation are not consistent with this common explanation, and future studies are needed to examine potential explanations.