Iron biomarkers predict peripheral artery disease in females

BACKGROUND AND AIM: Iron overload has been proposed as a risk factor for atherosclerosis, but the available data are controversial. Here, we investigated whether iron status shows sex-specific associations with peripheral arterial disease (PAD). METHODS: Using two different analytical approaches (machine-learning and logistic regression), we studied the association between blood iron biomarkers and PAD in 368 individuals from the Heidelberg Study on Diabetes and Complications (HEIST-DiC) and in 5101 individuals from the National Health and Nutrition Examination Survey (NHANES 1999-2004). Additionally, by analysing data from the UK Biobank, we investigated the odds of PAD in individuals with hemochromatosis genotypes (n = 448,575). RESULTS: We found that iron biomarkers were among the top predictors of PAD in the machine-learning classification in both cohorts. In the HEIST-DiC cohort, ferritin, iron, and transferrin were ranked among the top predictive markers, while in the NHANES cohort, ferritin, total iron binding capacity (TIBC), transferrin saturation (TSAT) and iron showed high predictive power. In the regression analysis, ferritin showed a positive interaction among females in the HEIST-DiC (OR 2.68, 95% CI 0.94-7.61, P = 0.057) and NHANES cohorts (OR 1.76, 95% CI 1.16-2.67, P = 0.008). The multivariable regression analysis of the NHANES cohort detected a nonlinear relationship between ferritin and PAD, in that, certain ferritin ranges (48-97 ng/mL: OR 14.59, 95% CI 1.6-135.93, P = 0.019; 98-169 ng/mL: OR 171.07, 95% CI 1.27-23404, P = 0.039) in females were positively associated with PAD. Nevertheless, we did not detect significant associations between hemochromatosis genotypes and PAD in the UK Biobank. CONCLUSION: Taken together, our data show that iron biomarkers, importantly, elevated ferritin within physiological limits are associated with clinically apparent PAD in females. These findings add to the body of evidence suggesting sex differences in PAD and highlight a possible role of iron (directly or indirectly) in this relationship.

• Ruban Agarvas, A.
• Kopf, S.
• Lopes, T. J. S.
• Atkins, J. L.
• Thalmann, P.
• Fernández-Real, J. M.
• Nawroth, P.
• Muckenthaler, M. U.