点击显示 收起
Sahlgren University Hospital, Department of Clinical Nutrition, University of Göteborg, Box 459, Göteborg SE 405 30, Sweden, E-mail: hallberg@medfak.gu.se, E-mail: lena.hulthen@medfak.gu.se
Dear Sir:
In an article in the December issue of the Journal, Fleming et al (1) report that some dietary factors are associated with an increased risk of development of high iron stores in the elderly. This finding contradicts data from our laboratory (2), which indicate that the absorption of dietary iron (heme and nonheme) cannot result in iron overload in otherwise healthy, iron-replete persons.
In the study by Fleming et al, high concentrations of serum ferritin (SF) in their sample of 614 subjects aged 68–93 y were equated with high iron stores. The normal ranges of serum ferritin were arbitrarily set to < 300 µg/L in men and to < 200 µg/L in women. About 13% of the sample had values that exceeded these concentrations, which indicated that the sample was iron overloaded.
Iron stores can only be directly quantitatively determined by measuring the amount of iron removed during repeated phlebotomies until anemia develops and then subtracting both the iron deficit induced and the amount of dietary iron absorbed during the phlebotomy period. In an analysis of the available data (3), an equation was derived to describe the relation between iron stores and SF.
Estimations of iron stores, however, can be made from direct chemical determinations of nonheme iron in liver samples. Total liver iron was determined in necropsy material from 136 healthy male subjects aged 10–95 y who died in accidents or committed suicide (4). Total iron was almost constant between 20 and 90 y. In the classic study of iron stores by Weinfeld (5), the amounts of stored iron were calculated from the concentration of nonheme iron in surgical liver specimens. Weinfeld concluded that iron stores in adult men remain within fairly constant limits with age. The same was found in an extensive study of necropsy materials from the livers of 3983 samples collected in 18 different countries. No effect of age on the concentration of storage iron in the liver was seen in men (6). In the United States, a study of 259 liver samples from persons who died acutely from disease or trauma was conducted. The concentration of nonheme iron in males did not increase with age (7). The conclusion to be drawn, on the basis of consistent evidence from these extensive studies, is that there is no increase with age in iron stores in men.
It was observed in the second and third National Health and Nutrition Examination Surveys that SF increases with age (8). The key question is whether this increase in SF with age is associated with higher iron stores. Several attempts have been made to find an explanation for the increase in SF observed with age. It is well known that the concentration of SF in plasma, which is a strong acute phase reactant, is influenced by various factors such as liver disease, infections, inflammatory conditions, renal failure, cardiovascular diseases, and high alcohol consumption. Thus, there are several reasons to be careful when translating SF concentrations into iron stores.
The findings by Fleming et al that certain dietary factors and intakes of dietary iron supplements are associated with higher SF concentrations seem to fit with our data and the data of others that iron balance is related to diet. Our objection is that there is no evidence of a relation between diet and iron overload and that the control systems for iron absorption from the diet, including from red meat, are overrun in otherwise healthy subjects.
It is reasonable that dietary factors influence serum ferritin concentrations and iron stores, as is outlined in our recent article (9). We object, however, to equating serum ferritin concentrations and iron stores. The question raised by Fleming et al is whether high iron stores in the elderly (ie, high SF concentrations) are associated with disease or whether the opposite—that certain diseases are associated with high SF—is true. We have found that even a very mild disease is associated with elevated concentrations of SF. In a study of 1670 teenagers aged 15–16 y, we found that 24.4% of those with a mild common cold and fever for =" BORDER="0"> 1 d during the preceding 3 wk had a significant shift in SF toward higher values (10).
The report that 13% of elderly Americans have an iron overload contrasts the finding of a low prevalence of iron deficiency (2.7%). This latter value, however, was derived by using multiple criteria that we found to markedly underestimate the true prevalence of iron deficiency (9). The only ways to clarify whether elevated serum ferritin concentrations in the elderly are really associated with increased iron stores are to perform phlebotomies for the measurement of iron stores and to obtain liver biopsy samples or to conduct postmortem analyses for the measurement of liver iron concentrations.
REFERENCES