About the Sullivan's iron hypothesis

The well known Sullivan’s iron hypothesis is so attractive that it is difficult to ignore it. Its charm derives from the fact that it seems to be involved in everything or is able to explain everything, from the atherogenetic effect of inflammation to plaque formation and instability, from proteinuria to metabolic syndrome, to diagnosis or even genesis of cancer [1–6]. The temptation to employ the iron hypothesis to explain the differences between men and women, and between fertile and menopausal women as well, in terms of blood pressure (BP) and cardiovascular risk, is strong, particularly when the so-called oestrogen hypothesis repeatedly fails, such as in our experience based on the general population [7–11]. Unfortunately, this seems to be up to now nothing more than a hypothesis. The observation that stored iron is lower in normotensive than in hypertensive patients is insufficient; a population-based multivariate prospective model playing ‘stored iron’ as an independent covariate and BP or outcome as the dependent variable is necessary to confirm the iron hypothesis. Moreover, in our consolidated population-based experience [7–11], any differences between fertile and menopausal women in terms of BP values, prevalence and incidence of hypertension, fatal and morbid outcome is no longer present after simple adjustment for age. This seems to indicate that ‘age’ makes the difference between the two categories of women. We are unable to conceive a model in which stored iron could modify the results. Mere sideremia is of no interest in this respect, probably because circulating iron does not reflect body storage. Actually, in our database circulating iron is significantly higher in men [101.835.1mg/dl, 95% confidence interval (CI) 99.3–104.4] than in fertile (82.436.7, 95% CI 77.7–89.0) and menopausal women (86.827.3, 95% CI 84.8–88.9), but not different in menopausal versus fertile women, and does not predict either mortality or nonfatal outcome. We hope this interesting hypothesis can be more directly tested in the future.