Iron status and risk of cardiovascular disease

ELSEVIER

Iron Status and Risk of Cardiovascular MARIA-CHIARA AND CHARLES

CORTI, MD, MHS, MICHAEL H. HENNEKENS, MD, DRPH

Disease GAZIANO,

MD, MPH,

Free Iron, as well as other transition metals, can catalyze free radical formation. For this reason iron is tightly bound to transport and storage proteins to prevent their involvement in free radical formation. It bus been hypothesized that increased iron intake or iron stores may promote atherogenesis by increasing free radical formation and oxidative stress. While a coherent, plausible hypothesis as to how transition metals, such as iron, might accelerate the progression of atherosclerosis has been generated from basic research, iron status, measured as dietary iron intake, serum iron, serum ferritin, and transfetin saturation, has been inconsistently associated with cardiovascular disease in human epidemiologic research. In addition, limited data suggest that iron overload states do not appear to be strongly associated with increased risk of atherosclerotic disease. One real limitation of the existing data is the lack of a generally agreed upon and logistically feasible means of assessing iron status in free living humans. Further research, including basic research and large-scale epidemiologic studies, is needed to fully assess the association between iron status and the risk of CVD and other advserse outcomes. At present the currently available data do not support radical changes in dietary recommendations or screening to detect high normal levels nor do they support the need for large-scale randomized trials of dietary restriction or phlebotomy as a means of lowering iron stores. 0 1997 by Elsevier Science, Inc. Ann Epidemiol 1997;7:62-68. KEY WORDS:

Iron, ferritin,

oxidative

stress, cardiovascular

INTRODUCTION Basic research, astute clinical observation, and epidemiologic studies have all contributed to an emerging body of evidence on the role of oxidation in the pathogenesis of many chronic diseases including cancer, eye disorders, arthritis, and reperfusion injury during myocardial infarction. Recent evidence suggests that oxidative stress, which results from free-radical formation beyond the level that can be handled by the body’s normal defense mechanisms, may accelerate atherogenesis. Oxidatively modified low-density lipoprotein (Ox-LDL) has been implicated in several important steps in atherogenesis. In addition, free radicals may promote thrombosis and interfere with normal vasomotor regulation. Thus, factors that increase the formation of free radicals may increase the risk of atherosclerotic disease. Transition metals such as iron can cause the formation

From the Epidemiology, Demographyand Biometry Program,National Institute on Aging (MCC); the Division of Preventive Medicine (JMG, CHH) and the Cardiovascular Division, (JMG), Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School; the Department of Ambulatory Care and Prevention, Harvard Medical School (CHH), and the Department of Epidemiology, Harvard School of Public Health (CHH), Boston, MA; and ;he Department of Medicine, Veterans’ Affairs Medical Center, Brockton/West Roxburv, MA (IMG). Address reprint requests to: Dr. j. Michael G&iano, l%vision of Preventive Medicine, Brigham and Women’s Hospital, 900 Commonwealth Avenue East, Boston MA 022151204. Received March 15, 1996; accepted July 31, 1996. 0 1997 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010

disease, coronary heart disease.

of free radicals under certain circumstances, and this had led some to speculate that higher levels of iron may increase the risk of coronary heart disease (CHD) ( 1). In this paper we briefly describe the role of oxidation in atherosclerosis, the ability of iron to induce free radical reactions, and the physiology of iron metabolism. We then review the epidemiologic studies on the relationship of iron with risk of cardiovascular disease and consider future avenues of research.

BASIC RESEARCH Oxidative Stress in Atherogenesis Atherosclerosis is an extremely complex, chronic pathobiology of the arteries (2). The first phase of atherogenesis is the development of the fatty streak, which consists of a large number of lipid-loaded “foam cells” lying beneath an intact layer of endothelial cells. The second is the conversion of the fatty streak to the fibrous plaque. The third-and most clinically significant-is the development of the complex lesion that may result in thrombosis, leading to endorgan damage. The strength and consistency of basic research, epidemiologic studies, and randomized trials strongly support the judgment that hypercholesterolemia is a major causative factor in the development and progression of both the early and later stages of atherosclerosis (3). Elevated LDL is associ1047s2797/97/$17.00 PI1 SlO47-2797(96)00112-3

ated with increased risk of cardiovascular disease but, until recently, the mechanism by which LDL acts was unclear. Data from in-vitro and in-vivo studies now suggest that oxidative damage to LDL significantly increases its atherogenicity (4). Ox-LDL may have several different mechanisms for promoting atherogenesis. First, Ox-LDL may directly alter both the structure and function of endothelial cells (5, 6). Second, Ox-LDL may chemotactically attract monocyte/macrophages to the subendothelium (7), and these monocyte/macrophages may then develop into the lipid-laden foam cells of an atheromatous plaque (8, 9). Third, Ox-LDL is taken up into foam cells via a scavenger receptor ( 10, 11). Fourth, OX-LDL may stimulate the synthesis of autoantibodies, which may play a role in atherogenesis (12). Oxidative stress may be involved in other aspects of atherogeneisis that are not necessarily mediated by the oxidation of LDL. Free radicals may directly damage arterial endothelium (13), promote thrombosis (14), and interfere with normal vasomotor regulation (15). By several mechanisms, then, oxidative damage may initiate and propagate a cascade of reactions that result in atherosclerosis and thrombosis. Those factors that can increase oxidative stress by enhancing the formation of free radicals may increase the risk of atherosclerotic disease. Transition metals, such as iron and copper, in solution are very effective in catalyzing freeradical oxidation of lipids and proteins (16). When incubated with iron or copper, LDL will become oxidized (1719). These data from in-vitro studies raise the possibility that transition metals such as iron could increase oxidative stress in viva and thus could promote atherogenesis, thereby suggesting a possible mechanism by which iron may increase the risk of cardiovascular disease (CVD). However, the degree to which iron-catalyzed oxidation occurs in vivo remains unclear. Limited data from animal studies suggest a role of extreme iron overload states in the development of experimental atherosclerosis (20); however, data on physiologic levels of iron in animals are unavailable. While available data provide a plausible mechanism by which iron could contribute to atherogenesis, the relevance of these basic research findings to free-living humans is not entirelv clear. Effects of Iron on Cellular, Cardiac Metabolism

Muscular,

and

Bound iron is essential in normal redox reactions that occur under tightly controlled conditions. Iron is an essential element to all forms of living organisms, and complex biological systems have been developed to guarantee adequate intake, effective internal distribution, and balanced metabolism of this metal (2 1). Iron-containing proteins are indispensable in order for the electron transport system of the respiratory

chain to provide energy to the cell. Iron is also required for cell proliferation and DNA synthesis (22) and plays a critical role in the proper functioning of the brain and the immune system (23). Iron is essential for hemoglobin function as a means of 02 and CO! transport and for myoglobin synthesis in the skeletal and cardiac muscle cells. Iron-deficiency impairs skeletal muscle function, apparently because of an iron deficiency-induced abnormality of muscle metabolism that leads to increased lactic acidosis and reduction in ol-glycerophosphate dehydrogenase activity (24). In iron-deficient anemic patients, ST-T segment depressions of the ECG, observed during treadmill tests, have been abolished by the infusion of iron dextran well before the blood hemoglobin concentration increased (2 5). These findings suggest that iron deficien<,y can impair cytochrome or enzyme function within the skeletal muscle cells or within the conduction system by a mechanism independent from the anemia-induced cardiac hypoxemla. Extreme excessesof iron in iron overload states are also clearly harmful, resulting in parenchymal damage t3 the liver, heart, and many other organs. Thus, both iron deficiency and extreme overload states have deleterious effects; however, the degree to which iron status in the absence of severe deficiency or overload predicts atheroscl~~rot~cdisease remains an area of some debate. Iron Handling

We humans have evolved a system of Iron-handling to minimize circulating levels of “free” iron, which is necessary for unintended catalysis of free-radical reac.tions (16). Iron must be in its reduced (ferrous) state in order to catalyze free-radical reactions. In its ferrous state ir may exacerbate any oxidative stress and seriously damage cellular integrity (26). These peculiar properties of iron, therefore, could possibly explain why iron metabolism is so carefully controlled. Higher organisms are particularly careful about how iron is handled. In the healthy state, there is rarely any free or loosely chelated ferrous iron. The vast majority of iron in vivo remains protein bound in the oxidized (ferric) state and is therefore less available to be involved m free radical reactions (16). Iron is tightly conserved in a nearly closed system. Each iron molecule cycles repeatedly from plasma and extracellular fluid to the bone marrow, where it is incorporated into hemoglobin, and to the blood, where within erythrocytes it circulates for - 4 months. When senescent erythrocytes are destroyed in the reticuloendothelial system, hemoglobin is digested and iron is released in to the plasma, where the cycle continues. Within each cycle, a :~ruall proportion of iron is transferred to and from storage iron, as ferritin or hemosiderin; a small proportion is lost in urine, feces and blood; and an equivalent proportion is &sorbed from the

64

Cm1et al.

AEP Vd 7, No. I Janum-j 1997: 6248

IRON STATUS AND CVD RISK

intestinal tract and transported, bound to serum transferrin, in the plasma (27). Bound iron is in a complex that is relatively difficult to reduce, and thus, it is not usually available to catalyze free radical reactions. Under certain experimental conditions of oxidative stress, iron can be released from its bound state and may then be available to further propagate ongoing oxidative reactions. However, the degree to which this occurs in vivo is uncertain. It has been postulated that oxidative stress generated in vivo in various micro-environments may result in the liberation of trace amounts of ferrous iron. However, the rate of liberation of ferrous iron from its bound state would likely depend on the degree of local oxidative stress rather than on the amount of total body iron (21).

EPIDEMIOLOGIC

STUDIES

Measures of Iron Stores The first step in evaluating the risk of any factor is establishing a way to quantify the exposure in question. For the purposes of epidemiologic research, the ideal means to assess iron status remains unclear. The most accurate and reliable measure of body iron stores is a histologic evaluation of iron deposits in a bone marrow biopsy specimen (28). However, since bone marrow biopsy is not feasible for use in largescale epidemiologic studies, other less invasive measures of iron stores are generally used. Intake of iron has been used as a measure of iron status, though gastrointestinal absorption of iron varies greatly depending on the iron status of the individual. Thus dietary iron may not be an accurate reflection of total body iron (29). There are several ways to assessiron status clinically (28, 30). The most widely used measures of iron status are blood based, such as the concentration of iron and ferritin in serum. An indirect assay of iron stores is the measurement of the ratio of serum iron to the total iron binding capacity (TIBC), a measure also known as percent transferrin saturation. Iron deficiency depressesserum iron levels and boosts the TIBC, therefore decreasing the transferrin saturation to values generally < 10%. Iron overload increases the serum iron with little effect on TIBC, leading to > 80% saturation of transferrin. Iron stores are also reflected by serum ferritin levels. Serum ferritin represents a secretory form of the storage protein found in cells. The concentration of serum ferritin level rises with iron loading and declines with depletion of tissue iron stores. The ratio of iron to TIBC and serum ferritin level can be affected by conditions unrelated to disturbances of iron metabolism (28). For example, serum iron and transferrin levels are depressed, while ferritin levels are increased, by conditions such as inflammation, cancer, and liver disease. With aging, transferrin levels tend to decrease while ferritin concentration tends to increase. Therefore, no indirect mea-

sure of iron stores can be considered optimal, and several potential confounding factors, such as age and chronic conditions, should be evaluated when serum markers are employed to assessthe role of iron as a possible risk factor for cardiovascular disease morbidity and mortality. Iron and CVD Morbidity

and Mortality

The hypothesis that iron status may play a role in the development of atherosclerotic disease was originally based on several findings from descriptive epidemiology (1, 31). According to the hypothesis, sex differences in CHD mortality and the increase in CHD risk in postmenopausal women are explained on the basis of higher stored iron. The low rates of CHD mortality in third-world countries was explained by the lower mean iron levels observed in these populations. The protective effect of aspirin on CHD incidence and mortality has been postulated to be due to small continued gastrointestinal blood loss resulting from use of aspirin. However, alternative explanations for each observation are far more likely to explain differences in CHD rates. First, postmenopausal hormone replacement therapy with estrogen alone (which does not result in menstruation), or with estrogen alone or in combination with progesterone among women with prior hysterectomy, is associated with protection against CHD comparable to that of regimens that result in menstruation. This observation suggests that it is not menstruation (resulting in iron loss) which is protective. The immediate benefit of aspirin in both primary and secondary prevention suggeststhat it is the antithrombotic effect of aspirin rather than any other effect that is protective. Cross-cultural differences in CHD rates may be the result of a wide range of factors. The possibility that serum iron level is an independent risk factor was given credence and widespread media attention with the publication of the results of a small Finnish study published in 1992 by Saloinen et al. (32). In a cohort of 193 1 middle-aged men, 5 1 experienced an acute myocardial infarction during an average follow-up of 3 years. After controlling for several, but not all, major coronary risk factors for CHD, elevated serum ferritin was associated with risk of myocardial infarction. The association was stronger in men with a LDL-cholesterol level of > 193 mg/dL; this finding raised the possibility of effect modification by lipid levels. A proponent of the hypothesis that elevated iron levels were associated with increased risk of CHD wrote a supportive accompanying editorial (33), and this early report received more widespread publicity than subsequent negative studies. Morrison et al. (34) explored the association between serum iron levels and risk of fatal myocardial infarction in a population of 9920 Canadian men and women. Although based on a small number of fatal events (10 deaths in the groups with the highest iron concentration), a significant

AEP Vol. 7, No. 1 .fanuary 1997. 6248

i :d-ri e, ai IRON STATlJS .%Nl, f MI RISK

65

TABLE 1. Prosuectivecohort studies of iron and cardiovascular disease

Authors

Population and age-range

Outcome

Significant association

Iron store measure

Takkunen et al. (lY8Y) Salonen et al. (1992)

2453 men, 25-74 years old

CVD mortality

Transferrin saturation

18,998 men, 16,926 women, 42-60 years old

CHD

Serum ferritin

Manrtari et al. (1993)

(Nested case-control among dyslipidemic men) 22,000 male physicians (nested case-control) 44,933 men, 40-75 years old 2036 men and women, 25-74 years old 45 18 men and women, 45-74 years old 1827 men, 2410 women, 40-74 years

CHD

Serum ferritin

Myocardial infarction CHD incidence

Serum ferritin

None

Dieray

Slone

New CHD events CHD, CVD, total mortality CHD incidence

TIBC

Stampfer et al. (lY93) Ascherio et al. (1994) Magnusson et al. (1994) Sempos et al. (1994) Lrao et al. (1994)

Morrison et al. (1094)

9920 men and women, 35-79 years old

Incidence of fatal myocardial infarction

Yes

Type of ,rssociation (n) 1nverse

iron intake Ye?, No No (horderline) Yes, mainly in women Yes, mainly in women NO Only if total cholesterol =a 240 mg/dI.

Serum ferritin Transferrin saturation Transferrin saturation Serum iron Dietary iron Serum iron

inverx None inverse trend lnvcrse inverse

Dietary iron

increase in risk of fatal myocardial infarction was found only among those persons with a serum iron level 2 175 pg/dL and a total cholesterol level 2 240 mgfdl. This finding supported the hypothesis of effect modification by lipid levels on the iron-heart disease association. However, the result is based on a subgroup analysis and thus must be interpreted cautiously. Other studies followed (Table 1 and 2) and involved larger and more heterogeneous populations. These failed to confirm the presence of a direct association between iron status and risk of CHD. In a national cohort of 4518 men and women followed for an average of 14 years, Sempos et al. (35) found no evidence of a direct association between transferrin saturation and risk of CHD. In fact, there was a suggestion of an inverse relationship between this measure

TABLE 2. Case-control Authors Aronow er al. (1993) Moore et al. (1995)

studies of iron

and cardiovasculardisease

Study design

Population

Cross-sectional

171 men, 406 women, 362 years old 169 cases, 152 controls, mean age, 60 years

Case control

of iron status and the risk of cardiovascular disease mortality. Similar results were reported in data from the NHANES I Follow-up Study (36). In this study, no a.ssociation between transferrin saturation and CHD risk was found, although a significant and inverse relationship between serum iron and the risk of myocardial infarction was present in women and, less clearly, in men. In a nested case-control study that was part (of the Physicians’ Health Study and included 238 men with myocardial infarction and 238 matched controls (37f, after adjustment for other coronary risk factors, men with .serumferritin levels 2 200 pg/L had a relative risk of myocardial infarction of 1.1 (95% confidence interval, 0.7-1.6) These findings suggest that no increased risk is associated with elevated ferritin

Outcome

Iron btore measure

Prevalent CHD

Serum ferritin

Carotid atherosclerosrs

Serum ferritin, Dietary iron

Siqniiic ,mt association -^Nt, No Nq, ---

Type of association N
66

AEP Vol. 7, No. I January 1997: 6248

Corti et al. IRON STATUS AND CVD RISK

levels. No evidence for effect modification was found by lipid profile and antioxidant status. A possible explanation for the lack of reproducibility of the Finnish results could be in the unique characteristics of the Scandinavian populations, known to have higher rates of CHD and relatively higher levels of serum ferritin as compared with other Western populations (38). The results of the Finnish study, however, were not confirmed by other studies from Scandinavian cohorts (39-41). Magnusson et al. (39) studied the role of several measures of iron stores as a risk factor for CHD in a cohort of 2036 Icelandic men and women aged 24 to 74 who were followed for 8.5 years. When markers of iron stores (e.g., TIBC, and ferritin) were analyzed in this population, no direct association was found between serum ferritin and the risk of CHD incidence. However, a significant negative relationship was demonstrated between TIBC and CHD incidence, both in men and in men and women combined. Similar findings were described by Takkunen et al. (40), who demonstrated an inverse association between transferrin saturation and risk of cardiovascular disease mortality in a cohort of 18,998 Finnish men. Finally, negative results were reported for the Helsinki Heart Study cohort by Manttari et al. (41), who found no association between serum ferritin and CHD risk. Given the small numbers of end points it remains possible that the results of the original Finnish study are due to chance. The study of intermediate end points, such as the degree of carotid intimal thickening, did not clarify the characteristics of the relationship between iron and atherosclerotic disease. Moore et al. (42), in a case-control study from the ARK (Atherosclerosis Risk in Communities study population), investigated the possible association between serum ferritin and carotid arterial intimal thickening, a measure of early atherosclerosis. After adjustment for major cardiovascular risk factors, however, there was no association between serum ferritin concentration and the severity of arterial thickening. Some have suggested that plasma measures are an inadequate way to measure iron status. The role of dietary iron has also been explored in a limited number of studies. In the Health Professionals’ Follow-up Study, Ascherio et al. (43) evaluated the association between dietary iron intake and risk of CHD. In this cohort of 44,933 men with no previous history of CHD who were followed for 4 years, 844 incident cases of CHD were recorded. No significant association was detected between total dietary iron intake and risk of CHD. However, the incidence of CHD was higher among men in the top quintile of heme iron intake. The authors of this report have suggested that the association for heme iron intake may be due to a stronger association with iron stores than that for total iron intake. An alternative explanation for this association is confounding by dietary factors since heme iron comes from animal

sources, which also contain saturated fats. The authors controlled for fat and caloric intake; however, consumption of red meats may be associated with other unhealthy behaviors that increase the risk of CHD for which there was residual confounding despite adjustment for most major coronary risk factors. In fact there was attenuation of the relationship of heme iron with CHD after adjustment for coronary risk factors. While Salonen et al. (32) found a modest effect for total iron intake, no association was seen for dietary iron in the above described study by Morrison et al. (34). If modest increases in iron status are a cause of atherosclerotic disease, it would be reasonable to assume that those with extreme forms of iron overload would be at considerably higher risk of atherosclerotic disease. There is clinical evidence suggesting that iron overload states such as hemochromatosis may contribute to cardiac damage. However, the damage does not result from atherosclerotic disease, but from parenchymal tissue changes in the myocardium. If iron plays a key role in the pathogenesis of ischemic heart disease, subjects homozygous or heterozygous for hemochromatosis should experience high rates of coronary heart disease (44). But evidence of an increased risk is lacking. In fact, the iron overloading observed with hemochromatosis is associated with increased risk of hepatic cirrhosis, diabetes, and congestive heart failure, but not myocardial infarction (45). A recent retrospective autopsy study showed no evidence of increased atherosclerotic disease among those with significant iron overload (hemochromatosis or multiorgan hemosiderosis) (46). In contrast, advanced severe coronary atheroscleroses was present in 12% of iron-overload cases as compared with 38% of age-, sex-, and race-matched controls (I’ = 0.01) (45), suggesting a possible protective effect.

COMMENT In summary, a coherent, plausible hypothesis as to how transition metals, such as iron, might accelerate the progression of atherosclerosis has been generated from basic research. It is clear that iron overload states are associated with a number of adverse sequelae, including myocardial disease, and may enhance susceptibility to reperfusion damage. In addition, there is increasing evidence that alterations in iron metabolism could affect the immunologic (23) and endocrine systems (26) as well as modify the susceptibility to cancer (47). However, iron overload states do not appear to be strongly associated with increased risk of atherosclerotic disease. The evidence for an association between higher normal levels of body iron and atherosclerotic disease is inconsistent. Iron status, measured as dietary iron intake, serum iron, serum ferritin, and transferrin saturation, has been inconsistently associated with cardiovascular disease in epi-

demiologic research. While there was a positive association in some studies (32,34), there was an inverse association in others. In fact, in the larger studies that included populations highly heterogeneous in terms of age, sex, and health status, the relationship between iron and CVD tended to be consistently inverse, i.e., those with lower iron levels were at increased risk of adverse outcomes, and the effect modification by lipids was not detected (34, 35,38, 39). Further, in older populations, low iron levels have been shown to be an important measure of general ill health and a strong risk factor for CHD, CVD, and all-cause mortality (48). One real limitation of the existing data is the lack of a generally agreed upon and logistically feasible means of assessing iron status in free-living humans. Each of the plasma or serum measures have been criticized as inadequate in assessing total body iron status, although one could speculate that the plasma status might be directly relevant to the risk of development of arterial disease. It is not feasible to conduct bone-marrow biopsies in the context of a large prospective cohort study. This inadequacy in assessing iron status would result in biasing epidemiologic studies toward no association. iZvenues for further research include basic research to further our understanding of the risk of oxidation damage that transition metals might pose in vivo and development of better animal models of atherogenesis to test whether physiologic levels of iron are related to disease development or progression. In addition, prospective studies of representative populations, with adequate proportions of women and older adults, that evaluate repeated measures of iron stores over time and that consider long-term follow-up to ascertain cardiovascular disease and other health-related outcomes will he helpful in better defining the relationship. These outcomes would ideally include events such as incidence of CHD, stroke; cancer, and the occurrence of infectious, metabolic and lmmunnlogic diseases, as well as cause-specific and overall mortality. Thus, future research, including basic research and largescale epidemiologic studies, is needed to assess the association hetween iron and the risk of CVD and other adverse outcomes. If the results of these studies confirm the existence of this association, then appropriate clinical trials may be indicated to quantitate the risks and benefits of any interventions before changes in the current definition of ideal or normal levels of iron stores and in the policies for iron fortification of foods can he recommended. However, the current data do not support radical changes either in dietary recommendations or in screening to detect high-normal levels nor do they support the need for large-scale randomized trials of dietary restriction or phlebotomy as a means of lowering iron stores.

REFERENCES 1. Sullivan JL. The iron paradigm of ischemic heart d~se:w. Am Heart J. 1989;117:1177-1188. 2. Steinberg D. Lipoproteins 1983;3:283-301.

and atherosclercws.

3. Consensus Conference: Lowering blood choie,rtwI disease. JAMA. 1985;8:77-88.

Arhcrtaclerosis. ti) prwem heart

4. Steinberg D, Parthasarathy S, Carew ‘T, Khw J, Wttztum J. Beyond cholesterol-modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med. 1989;320:915-924. 5. HesslerJR, Morel DB, JamesLJ, Chisolm GM. Lip~protem oxidation and lipoprotein-induced cytotoxicity. Arterh~scien~si~. 1983;3: 2 15-222. 6. Yagi K. Increased serum lipid peroxides initiate &ro~encsis. says. 1984;1:58-60.

Btoes-

7. Quinn MT, Parthasarathy S, Steinberg D. Endothelial cell-derived chemotactic actwity for mouse peritonwl macrophages and the effect5 of moditied forms of low-density lipoprotein. l’mc Narl Acad Sci IJSA. 1985;82:5949-5953. 8. Schaffner T, Taylor K, Bartuccl EJ, ut al. Arterial foam cells with distinctive immunomorphologic and histochenwal fearurc~ of macrophages. Am ] Pathol. 1980;100:57--73 9. Gerriry RG. The role of the monocyte m atherrrgenesi\. I. Transition of blood-home monocytes into foam cell, in Gf ty Irwm~. Am J I’athol. 1981;103:181-190. 10. Fogelman AM, Schechter I, Hokom M, Chdd JS, Edwards PA. Malondialdehyde alteration ot low density lipoprotems leads to cholesterol zster accumulation in human mtmocvtcmacr~~~d~agr~ 1%)c Nat1 Acarl Sci USA. 1980;77:2214-2218. Il. Goldstein JL, Ho YK, Basu SK, Brown MS. limding site c Nat1 Acad Sci USA. 1990;87:1620-1624. 14. Marcus AJ, Silk ST, Saber LB, Ull man HL. Supwax& production and reducing activity in human platelet<. J Clin Inw,\t. 1977:59:149--158. 15. Saran M, Michael C, Bars W. React~vn cli NC) wtb 0:. Implications for the awon of endothelium-deriwd rrlaxiw i:c ttrr (EDRF). Free Radlc Res Commun. 1990;lL7:211-?.%I. 16. McCord JM. Is mm sufhclency a risk f;~~to~-in i~<~hwnlcbwrt dise.tse! Circulation. 1991;8Z:1112-1 I I!. 17. Fang LG, Parchasarathy S, Witztum JL, Stemlwg 1). Nonenzymatic oxidatlve cleavage of peptide h&s m ap~~pr~~e~~\H-tOO. J Lqid Res. 1987;28:14661477. 18. Morel DW, DiCorleto PE. Chtsolm (;M. En&~tbclial and smooth muscle cells alter Low density lipoprotem m vitrd : !yv ire,: radical ox&non. Arteriosclerosrs. 1984;4:357-364. 19. Cathcart MK, Morel DW, Chisolm GM. Mom ncvtez and nerrtrophils oxidize low-dcnslty hpoprotems makmg it L..tclroxic. J Leukocyte Biol. I985;38: 141.-350. 20. Araq<) JA, Rolnano EL, Brito BE, et al. Iron ~;wrload augments the development <)i atheroscleroric lesion%,in rabbits. ArwrUwIer Throm Vast Btol. 1995;15: 1172-l 180. 21. Huebers HA, Finch CA. The phys~ttlogy 01’tr:msterrm .md transferrin receptor5. Physiol Rev 1987;67:52G58: 22. H&n S;, Jclhansen KS. Adequate mm rtor:. ,md the ‘NII nwere principle. Hernatologia. 1993:25:69-X4

68

AEP Vol. 7, No. I January 1997: 62-68

Corti et al. IRON STATUS AND CVD RISK

23. Johnson MA, Fisher JG, Bowman BA, Gunter EW. Iron nutriture in elderly individuals. FASEB J. 1994;8:609-62 1. 24. Finch CA, Gollnik PD, Hlastala MP, et al. Lactic acidosis as a result of iron deficiency. J Clin Invest. 1979;64:129-137. 25 Mehta BC, Panjwani DD, Jhala DA. Electrophysiology abnormalities of heme in iron deficiency anemia; effect of iron therapy. Acta Hematologica (Basel). 1983;70:189-193. 26 Halliwell B. &peroxide-dependent formation of hydroxyl radicals in the presence of iron chelates. FEBS Lett. 1978;92:321-326. 27. Fairbanks VF. Iron in medicine and nutrition. In: Shils ME, Olson JA, Shike M, eds. Modem nutrition in health and disease. Philadelphia: Lea and Febiger; 1994:185-213.

37.

38. 39.

40. 41.

28. Bridges KR, Bunn HF. Anemias with distrubed iron metabolism. In: Wilson JD et al., eds. Harrison’s Principles of Internal Medicine. 12th ed. New York: McGraw Hill 1991:1518-1523. 29. Willett W. Nutritonal Press, 1990:179-180.

Epidemiology. New York: Oxford University

42.

30. Beaton GH, Corey PN, Steele C. Conceptual and methodological issues regarding the epidemiology of iron deficiency and their implications for studies of the functional consequencs of iron deficiency. Am J Clin Nutr. 1989;50:575-588.

43.

31. Lauffer RB. Iron stores and the international variation in mortality from coronary heart disease. Medical Hypotheses. 1990;35:96102.

44.

32. Salonen JT, Nyssonen K, Korpela H, et al. High stored iron levels are associated with excess risk of myocardial infarction in eastern Finnish men. Circulation. 1992;86:803-811.

45.

33. Sullivan JL. Stored iron and ischemic heart disease. Empirical support for a new paradigm. Circulation. 1992;86:1036-1037.

46.

34. Morrison HI, Semenciw RM, Wigle DT. Serum iron and risk of fatal acute myocardial infarction. Epidemiology. 1994;5:243-246.

47.

36. Sempos CT, Locker AC, Gillum RF, Makuc DM. Body iron stores and the risk of coronary heart disease. N Engl J Med. 1994;330:1119-1124. 36. Liao Y, Cooper RS, MC Gee DL. Iron status and coronary heart diease:

48.

negative findings from the NHANES 1 Epidemiologic Follow-up Study. Am J. Epidemiol. 1994;139:704-712. Stampfer MJ, Grodstein F, Rosenberg I, Willett W, Hennekens C. A prospective study of plasma ferritin and risk of myocardial infarction in U.S. physicans (abstract). Circulation. 1993;87:688. Cook JK, Skikne BS, Lynch SR, et al. Estimates of iron sufficiency in the U.S. population. Blood. 1986;68:72&731. Magnusson MK, Sigfusson N, Sigvaldason H, et al. Low iron binding capacity as a risk factor for myocardial infarction. Circulation. 1994;89:102-108. Takkunen H, Reunanen A, Knekt P, Aromaa A. Body iron stores and the risk of cancer (letter). N Engl J Med. 1989;320:1013-1014. Manttai M, Manninen V, Huttunen JK, Heininen OP, Frick MH. Serum ferritin and ceruloplasmin as coronary risk factors: A nested case control study in Helsinki Heart Study population. Circulation 1993;83:70. Moore M, Folsom A, Barnes RW, et al. No association between serum ferritin and asymptomatic carotid atherosclerosis. The Atherosclerosis Risk in Communities (ARIC) Study. Am ] Epidemiol. 1995;141: 719-723. Ascherio A, Willett WC, Rimm EB, Giovannucci EL, Stampfer MJ. Dietary iron intake and risk of coronary heart disease among men. Circulation. 1994;89:969-974. Burt MJ, Hallyday JW, Powell LW. Iron and coronary heart disease. Iron’s role is undecided. BMJ. 1993;307:575-576. Olson LJ, Edwards WD, Holmes DR, et al. Endomyocardial biopsy in hemochromatosis: clinicopathologic correlates in six cases.J Am Coil Cardiol. 1989;13:116-120. Miller M, Grover M, Hutchins MD: Hemocromatosis, multiorgan hemosiderosis, and coronary artery disease. JAMA. 1994;272:231-233. Stevens RG, Jones DY, Micozzi MS, et al. Body iron stores and the risk of cancer. N Engl J Med. 1988;319:1047-1052. Corti M-C, Guralnik JM, Salive ME, Ferrucci L, Pahor M, Wallace RB, Hennekens CH. Serum iron level, coronary artery disease and all-cause mortality in older men and women. Am J Cardiol. 1997; January, in press.