 Medscape |
 Home Site Map Marketplace My Medscape CME Center Feedback Help Desk |
|
|
NOTE: To view the article with Web enhancements, go to:
http://www.medscape.com/mosby/AmHeartJ/2000/v139.n02/ahj1392.19.zach/ahj1392.19.zach-01.html
[Am Heart J 139(2):337-345, 2000. © 2000 Mosby, Inc.]
Fenton = Fe2+ + H2O2 --> Fe3+ + OH- + OH·
Haber-Weiss = O2·- + H2O2 --> OH- + OH·
Iron stores rise imperceptibly over time because amounts of iron ingested commonly exceed requirements.[23] Iron is absorbed in proportion to the amount ingested, and no homeostatic mechanism exists for excreting excess quantities.[1] This excess iron is taken up by the iron storage protein, ferritin.[24] Storage iron has no known physiologic role but can be released readily from ferritin to generate oxygen free radicals that are highly toxic to biomolecules.[22,24] Reactive oxygen species initiate lipid oxidation required for atherogenesis.[1-3,22,25]
Basic and epidemiologic evidence for a role of iron-induced oxidative stress in the pathogenesis of atherosclerosis has been reviewed in detail elsewhere.[13-21,25] In brief, Sullivan[9] proposed that the reduced risk of myocardial infarction in premenopausal women, compared with both men of comparable age and postmenopausal women, might be explained by the lower iron stores characteristic of premenopausal women. He then examined differences in the incidence of ischemic heart disease for men and women by age and showed that the change in risk over time paralleled changes in iron stores, as reflected by serum ferritin levels.[10,11] Both coronary risk and iron stores remained low during the premenopausal years in women compared with men and were independent of menopausal age.
Subsequently, it has been shown that levels of body iron stores are a strong predictor of coronary risk in men[13] and of carotid atherosclerosis in both men and women.[14,15] Higher iron stores predict poor outcome after stroke.[16] Routine voluntary blood donation is associated epidemiologically with reduced coronary risk.[17,18] Salonen et al[25] showed that reduction of iron stores (represented by a fall in serum ferritin levels) after phlebotomy in a cohort of men older than age 50 resulted in significantly decreased susceptibility of peripheral blood LDL to oxidation. Kiechl et al[15] found a highly significant correlation between serum ferritin concentration and pathologic carotid artery wall thickening in a longitudinal cohort study. These authors found the ferritin concentration to be a strong and independent risk factor for stroke in both men and women, and interactions existed between ferritin levels and both LDL levels and smoking. A rising ferritin level over time was associated with the occurrence of vascular events, whereas a falling ferritin level over time appeared to be protective. An increasing risk of vascular disease was observed with ferritin levels higher than 50 ng/mL, and this increased risk included both the appearance of new disease and extension of pre-existing disease.
Despite this evidence, the role of iron in atherosclerosis remains controversial, primarily because of methodologic inconsistencies and limitations of the epidemiologic approach to this pathophysiologic problem.[20,21] This controversy is susceptible to resolution through prospective randomized trials designed to measure effects of controlled reduction of body iron stores on clinical outcomes. Such studies have not been performed, and appropriate methods are undefined. This article reports the results of a pilot study conducted to evaluate the feasibility and methodologic accuracy of calibrated reduction in iron stores in a cohort of patients with advanced peripheral vascular disease (PVD) considered representative of generalized atherosclerosis. Emphasis was placed on defining demographic characteristics and the extent of disease in study patients. The goal was to establish methods and otherwise facilitate the design of definitive, prospective randomized trials.
To be included in the study, patients with PVD were required to have current intermittent claudication (leg pain that occurred only with walking and that disappeared in less than 10 minutes on standing) and an ankle/arm supine blood pressure ratio (ankle-brachial index) of <=0.85 in either leg at rest on 2 separate days. Alternatively, patients who had a history of intermittent claudication and previous reconstructive surgery or angioplasty of the legs for atherosclerotic disease and at least one leg, but no persistent complications from prior procedures, were also entered. Patients had to be evaluable at the 3-month follow-up end point in the study. Patients who also had atherosclerotic heart disease with angina pectoris had the degree of angina graded as class 0 to IV.[26] Patients with PVD so defined also met the following additional entry criteria:
Patient Stratification
The goal of this pilot study was to enter 50 patients, 30 of whom would be randomly assigned to have blood drawn to reduce body iron stores to an equivalent ferritin level of approximately 25 ng/mL and 20 of whom would serve as control. Having comparable patients monitored concurrently but without iron reduction controlled for possible unforeseen drift in ferritin levels over time. This pilot study was not performed blinded for several reasons. Ethical (human rights) considerations required that patients consent to have the calculated volume of blood drawn. Patient blinding would require sham venipuncture with introduction of a foreign body (a needle) into the veins of control patients without blood removal. A complex follow-up procedure would be required to monitor control patients to compensate for the variable number of blood drawings needed for iron reduction in the intervention group. It was considered unnecessary to blind the outcomes observer because this was a pilot study of short duration with objective (primarily laboratory rather than clinical) end points designed to assess feasibility and methodology.
Patients were stratified before randomization on the basis of their HDL/LDL ratio into 2 groups (high and low). As a patient was recruited for the study, the stratification information was entered into a computer program at the Palo Alto Veterans Administration Cooperative Studies Program Coordination Center (by a voice information system), and the optimal balancing of treatment assignment ascertained by determining which assignment would produce the best balance by HDL/LDL ratio and participating hospital, considering the targeted ratio of patients assigned to iron reduction: control was 3:2. The randomizer would be biased in favor of that assignment, and the patient randomly assigned. The odds were shifted to 3:1 in favor of the optimal allocation. The cutpoint for stratifying the HDL/LDL ratio was determined after the first 7 patients were entered in the study. Before then, the patient assignment was simply by 3:2 randomization.
Intervention
Reduction of body iron stores to a level equivalent to a serum ferritin concentration of approximately 25 ng/mL was chosen because it is typical of premenopausal women[9-12,23] and conditioned athletes[27-30] and because of epidemiologic data suggesting that this level is associated with minimal or baseline vascular disease risk.[15] The formula for calculating the volume of blood to be removed to achieve this ferritin level was based on the assumption that 1 ng/mL of serum ferritin is equivalent to approximately 5 mg of storage iron.[31] Because 1 mL of whole blood contains about 0.5 mg iron, the formula is as follows:
(Initial ferritin - 25) x 10 = Milliliters of blood to be drawn
Patients requiring multiple blood drawings for reduction of iron stores had a hemoglobin and hematocrit determination before each blood drawing. Blood drawing was not done if the hematocrit level was <39% in men and <33% in women. In the event that a value below this threshold was found, the patient was asked to return approximately 1 week later for reassessment and potential blood drawing. If the hematocrit level equaled or exceeded these values, blood was drawn into a phlebotomy bag. The volume of blood removed and the interval between drawings was not precisely mandated, although the goal was to achieve reduction in iron stores within 1 month. It was assumed that at least 150-mL volumes could be removed with the patient in a recumbent position at intervals of no less than 24 hours. Removal of this volume of blood in a single drawing for diagnostic purposes is not unusual in hospitalized patients.[32-34] The volume removed was adjusted as tolerated by the patient, and the frequency of blood drawing modified for the convenience of the patients and staff. No more than 1 unit (500 mL) of blood was removed within a period of 1 week, in keeping with the standard of therapy for hemochromatosis.[35,36] The volume of blood removed was determined by a standard blood scale that measures the weight of the blood removed expressed as volume. The steady state level of iron stores that exists at a given time is determined by the volume of blood removed and the amount of iron ingested in excess of requirements since the last blood drawing. Iron ingested from an ambient diet should produce a rise of approximately 0.2 ng/mL of ferritin/day.[37] Over a 3-month period, an increment in ferritin of about 18 ng/mL would be expected that would oppose reduction of ferritin levels resulting from blood removal. Therefore having a 3-month end point for outcome assessment in this study should allow for physiologic adjustment from iron reduction to occur, whereas the iron stores remained in the presumed protective range.[15] The 3-month time period is consistent with the findings of Salonen et al,[25] who found maximum protection against LDL oxidizability from blood drawing within this time interval.
Achieving the targeted reduction in iron stores in this study would not render patients iron deficient. However, anemia caused by iron deficiency or some other cause might appear in study patients from, for example, the development of a bleeding lesion. In the event that previously unforeseen bleeding or anemia should appear, appropriate steps would be taken to diagnose and treat the condition. Iron replacement therapy sufficient to correct anemia would be prescribed for patients who had iron deficiency anemia develop. However, such events would not necessarily require removing the patient from the study, and intervention would be resumed according to the prescribed treatment plan once the condition had resolved. The possibility was also considered that treatment may have to be discontinued for some reason other than bleeding according to the judgment of the attending physician. For example, a patient may have a myocardial infarction and be at risk for cardiovascular instability. Such discontinuation of intervention and the accompanying reasons were noted on the data sheets. Participation in this study did not preclude any other form of treatment (except for the prohibition against ingesting vitamin and iron supplements during the study period), and all patients would be expected to receive otherwise optimal care for their disease.
Patient Follow-Up
Both the iron reduction and control groups were followed in an identical manner. Evaluation was on an intent-to-treat basis. There were no prior exclusions from outcome analysis for the relative success of blood drawing or degree of reduction of serum ferritin concentration achieved.
Basic demographic, physical examination, and laboratory data were collected at randomization and at 3-month follow-up. The volume of blood drawn for diagnostic testing in control patients was not considered likely to confound conclusions reached in the iron reduction patients because such volumes were relatively small (only about 5 to 10 mL of blood were needed for laboratory test monitoring) and patients were not otherwise acutely ill and would not likely require extensive blood testing for diagnostic purposes. Volumes of blood drawn for both diagnostic and therapeutic purposes for both patient groups were recorded on study data sheets. At the end of the study, patients were asked to assess their overall health status since entry as either improved, stable, or worse.
Statistical Analysis
Categoric data are presented as counts with the percentages. Chi-square or Fisher exact tests were used to compare the 2 strategies. Continuous data are presented as means ± SD. Student t test statistics were used to compare the differences between the 2 strategies. All the statistical tests were 2-tailed.
Removal of the volume of blood targeted was accomplished in a mean of 2.62 visits (minimum 1 and maximum 6 visits). A mean of 311.18 mL (minimum 0, maximum 700 mL) of blood was removed per visit. The phlebotomy procedure was successful in all patients except 1 randomly assigned to have iron reduction. The single exception was a patient with an initial ferritin level of 184 ng/mL who required removal of 1590 mL of blood. He had blood drawn on 3 occasions totaling 1270 mL. On return for his last blood drawing, his hematocrit was 38%. There was not an opportunity to repeat this at a later date and no further blood was drawn before the 3-month completion of the study, at which time his ferritin concentration was 60 ng/mL. Data from this patient were included in the analysis from this study.
Four patients who were randomly assigned dropped out of the study. These included the patient previously mentioned who verbally agreed to participate but did not sign part II of the consent form. Three additional patients, 1 in the control and 2 in the phlebotomy groups, did not keep follow-up appointments for unexplained reasons. Therefore follow-up data were available for analysis on 44 patients (26 iron reduction and 18 control).
Results of laboratory testing were analyzed for both iron reduction and control groups at entry and after 3 months of follow-up. Results are presented in Tables IV through VII. Ferritin levels at entry and after 3 months of follow-up in the control group were comparable to levels expected in men and postmenopausal women of this age.[23] There was little drift in ferritin concentration over the 3 months of this study in the control group. In contrast, blood drawing succeeded in changing measures of iron status, the demonstration of which was a primary goal of the pilot study. In the iron reduction group, the ferritin fell from a mean of 124.5 ng/mL to a mean of 51.54 ng/mL at 3 months on removal of a mean of 909.62 mL of blood. However, iron accumulates daily from the diet and the blood is drawn over time. Assuming that iron accumulates at a rate of approximately 0.2 ng/mL/day,[37] over a 3-month period 18 ng/mL of ferritin would have been gained. Therefore calculation of the theoretical, instantaneous decrease in ferritin concentration would have given a mean nadir ferritin level of 51.54 - 18 = 33.54 ng/mL "true" postdrawing ferritin concentration. This equals a decrease of a mean of 91 ng/mL of ferritin. Thus our results indicate that 1 ng/mL of ferritin concentration represents approximately 5.0 mg of storage iron. This storage iron-ferritin equivalence coincides with the level anticipated,[31] and the nadir ferritin level achieved by the formula is close to the level of 25 ng/mL originally targeted. It would be impractical to refine the iron reduction formula to give a closer value, and the formula for achieving the intended reduction of body iron stores was considered valid. The decline in serum ferritin concentration was accompanied, as expected, by a corresponding rise in total iron binding capacity (Tables V and VI).
The blood glucose concentration was significantly lower in the iron reduction group compared with the control group both at entry and at 3-month follow-up (Tables IV and V). This is likely attributable to the fact that 5 control patients but only 1 phlebotomy patient had diabetes (Table I, P = .0297, Fisher exact test). The difference in triglyceride levels between the groups at 3-month follow-up (Table V) is difficult to explain and may be fortuitous. Levels of C-reactive protein were not elevated at baseline, and phlebotomy had no effect on these levels. In addition, there was no correlation between initial ferritin concentrations and levels of C-reactive protein. There was no significant difference between phlebotomy and control groups for hemoglobin and hematocrit values at 3-month follow-up (Table V). Comparison between groups of changes in these parameters from entry values revealed a decrement at 3-month follow-up in both hemoglobin and hematocrit that was greater in the phlebotomy group (Table VI). However, the absolute magnitude of the change was slight and of no biological significance. Otherwise, no differences in laboratory tests were observed.
New clinical events were recorded on the data sheets at the end of the follow-up period. In the control group, 2 patients sustained a new myocardial infarction, 2 had onset of new unstable angina, 2 had a new dysrhythmia, and 1 had congestive heart failure develop. In the phlebotomy group, 1 patient underwent angioplasty shortly after entry to the study. The time interval from entry to development of these conditions is shown in Table VII. Global health assessment for control patients showed that none were improved, 15 were stable, and 3 were worse. Among phlebotomy patients, 3 were improved, 22 were stable, and 1 was worse. No patient registered a new, nonvascular disease diagnosis and no patient died during the 3-month interval of the study. The small sample size that was appropriate for this pilot study did not have sufficient power to detect statistically significant differences, but these observations suggest that phlebotomy, as prescribed in this study, does not contribute to increased morbidity and mortality in patients with severe atherosclerotic vascular disease.
However, this concept remains controversial.[20,21] It is beyond the scope of this paper to analyze this controversy in detail. However, inconsistencies among epidemiologic studies have been cited to discount the iron hypothesis. Unfortunately, some of these studies used measures of transport iron (such as the percentage of transferrin saturation) as a basis for disease correlations, but these are known to not represent ambient levels of body iron stores as the ferritin concentration does.[23] Another problem is that ferritin levels below 50 ng/mL are uncommon in free-living adult men, and there is no naturally occurring experimental population that simulates the premenopausal state to provide a basis for comparison of risk. Dietary surveys are problematic because many individuals are unaware of the iron content of nutritional supplements and processed foods ingested, especially those consumed outside the home.[43] Dietary loading may occur years to decades before the survey, obscuring temporal relations to disease. The case-control strategy is difficult because the ferritin concentration may be unpredictably altered in ill patients and because such individuals may have considerable blood drawn for diagnostic purposes that would lower ferritin levels.[32-34] Such after-the-fact assessment would indicate little about the premorbid effects of iron.
Nonetheless, available data provide intriguing clues to a potentially important disease mechanism. Although the epidemiologic approach has serious limitations that make it less than ideal for testing the iron hypothesis, this concept can be evaluated definitively by means of prospective, randomized, controlled clinical trials because iron stores can be modified as a single variable. This pilot study was performed to demonstrate feasibility and define methods for such studies. The results of our study confirm that ferritin levels are often in the typical iron-replete range in patients with PVD.[23] The ferritin level in PVD can reasonably be interpreted as a reflection of body iron stores rather than an acute phase reaction[44] because C-reactive protein levels were normal at baseline in these patients. Levels of C-reactive protein have been shown to have predictive value for myocardial infarction.[45] There was no correlation at entry to the study between ferritin and C-reactive protein levels and no change in C-reactive protein levels with phlebotomy. The formula used for calculating the amount of blood to be removed to achieve the targeted reduction in body iron stores is accurate and the goal of reduction in iron stores can be achieved in a timely manner. There was little change in the ferritin level after 3 months in the control group, suggesting that a substantial "drop-in" rate among control patients would not be expected in a prospective, randomized trial. Patient acceptance of intervention was excellent. No patients entered later had iron deficiency develop, and none were lost to follow-up, terminated early, or died because of the intervention. Four patients who entered dropped out of the study before the 3-month follow-up was complete. One of these was in the control group and 3 were in the phlebotomy group. These patients ceased participation because they did not, for unexplained reasons, return for follow-up visits. The 2-part informed consent procedure was successful. Adverse clinical outcomes occur commonly and can be tabulated on the data sheets in patients with generalized atherosclerosis. There was no evidence for a detrimental effect of body iron store reduction to the targeted level on either clinical or laboratory grounds.
If reducing iron stores proves to be beneficial in atherosclerosis, a simple, nonpharmacologic, and inexpensive means requiring no special technology would be at hand for disease prevention and treatment. In this pilot study, 42.5% of screened patients and 57.1% of eligible patients signed part I of the informed consent form, suggesting that patient accrual to subsequent randomized trials should be satisfactory. Reduction of iron stores would not be expected to affect any other aspect of patient care, and iron reduction carries no disease or toxicity tradeoffs. Methodologic and feasibility data reported here may be applicable to diseases other than atherosclerosis attributed to excessive iron-induced oxidative stress[4-8] because toxicities from phlebotomy would be expected to be primarily vascular, and none were observed in patients with overt, generalized vascular disease entered in this pilot study. Although the intervention may not lend itself to a double-blind trial design, single blinding for recording of outcome events would be feasible. Such a study can be executed by prepared nursing staff. These results suggest that definitive clinical trials designed to test effects of reduction of body iron stores by phlebotomy on disease outcomes may be conducted successfully. Such clinical trials would provide a setting for studies of possible mechanisms of the effect of iron on disease.
Acknowledgements
Supported by the Cooperative Studies Program of the Department of Veterans Affairs.
Characteristic Iron reduction
(n = 29)Control
(n = 19)Total
(n = 48)Age (mean ± SD) 64.91 ± 8.22 67.6 ± 6.32 65.98 ± 7.57 Sex (M/F) 29/0 18/1 47/1 Race (white/other) 28/1 15/4 43/5 Height in inches (mean ± SD) 68.66 ± 3.20 68.11 ± 2.85 68.44 ± 3.05 Weight in kilograms (mean ± SD) 79.2 ± 16.46 82.24 ± 19.6 80.37 ± 17.62 No. of years smoked (mean ± SD) 41.62 ± 13.51 38.61 ± 17.45 40.47 ± 15.03 Number of cigarettes/d (mean ± SD) 26.38 ± 12.53 18.89 ± 12.95 23.42 ± 13.09 Alcoholic drinks/d (mean ± SD) 0.45 ± 0.87 0.63 ± 1.89 0.52 ± 1.35 Diabetes mellitus 1 (3.45%) 5 (26.32%) 6 (12.50%) COPD 5 (17.24%) 4 (21.05%) 9 (18.75%) Hypertension (treated) 20 (68.97%) 11 (57.89%) 31 (64.58%)
COPD, Chronic obstructive pulmonary disease.
Condition Iron reduction
(n = 29)Control
(n = 19)Total
(n = 48)PVD 29 (100%) 19 (100%) 48 (100%) Months since diagnosis 51.64 ± 52.89 70.92 ± 68.82 59.28 ± 59.26 Prior reconstructive surgery 11 (37.93%) 7 (36.84%) 18 (37.50%) Prior angioplasty 6 (20.69%) 4 (21.05%) 10 (20.83%) Prior amputation 2 (6.90%) 1 (5.26%) 3 (6.25%) Coronary artery disease 18 (62.07%) 9 (47.37%) 27 (56.25%) Prior myocardial infarction 12 (41.38%) 4 (21.05%) 16 (33.33%) Prior angioplasty 6 (20.69%) 0 (0%) 6 (12.50%) Prior CABG 4 (13.79%) 0 (0%) 4 (8.33%) Prior congestive heart failure 2 (6.90%) 1 (5.26%) 3 (6.25%) Stable angina grade I-IV 11 (37.93%) 5 (26.32%) 16 (33.33%) Cerebral vascular disease 11 (37.93%) 7 (36.84%) 18 (37.50%) Prior stroke 5 (17.24%) 3 (15.79%) 8 (16.67%) Current TIAs 3 (10.34%) 4 (21.05%) 7 (14.58%)
Values are n (%) or mean ± SD.
TIAs, Transient ischemic attacks.
Drug Iron reduction
(n = 29)Control
(n = 19)Total
(n = 48)Estrogen/hormone 0 (0%) 0 (0%) 0 (0%) Lipid lowering 10 (34.48%) 5 (26.32%) 15 (31.25%) Antihypertensive 24 (82.76%) 11 (57.89%) 35 (72.92%) Coronary vasodilator 12 (41.38%) 4 (21.05%) 16 (33.33%) Aspirin 24 (82.76%) 17 (89.47%) 41 (85.42%) Other NSAID 4 (13.79%) 4 (21.05%) 8 (16.67%) Pentoxifylline 5 (17.24%) 2 (10.53%) 7 (14.58%) Prior vitamin C/E use 9 (31.03%) 11 (57.89%) 20 (41.67%) Prior iron supplement use 3 (10.34%) 5 (26.32%) 8 (16.67%)
NSAID, Nonsteroidal anti-inflammatory drug.
Blood
measurementIron reduction
(n = 29)Control
(n = 19)P value Hemoglobin 15.59 ± 1.15 14.98 ± 0.92 .062 Hematocrit 45.62 ± 3.23 44.05 ± 2.98 .097 Fibrinogen 404.3 ± 89.20 405.3 ± 84.87 .97 Ferritin 124.5 ± 72.92 126.2 ± 92.30 .94 Glucose 104.5 ± 15.94 131.3 ± 54.76 .016 Uric acid 6.18 ± 1.12 6.11 ± 1.45 .85 Total cholesterol 221.4 ± 41.68 225.7 ± 37.21 .71 Triglycerides 200.9 ± 95.13 165.5 ± 100.99 .22 HDL cholesterol 36.41 ± 11.05 39.42 ± 12.67 .39 LDL cholesterol 28 ± 41.44 155.8 ± 32.34 .55 Serum iron 75.52 ± 21.88 77.22 ± 35.79 .84 Total iron binding capacity 305.5 ± 48.54 296.2 ± 45.28 .52 C-reactive protein 0.79 ± 0.76 0.52 ± 0.25 .14 LDL/HDL 4.43 ± 1.78 4.31 ± 1.56 .82
Values are mean ± SD.
Blood
measurementIron reduction
(n = 26)Control
(n = 18)P value Hemoglobin 14.86 ± 1.11 15.05 ± 1.05 .58 Hematocrit 43.58 ± 3.48 44.56 ± 3.48 .36 Fibrinogen 408.0 ± 85.44 385.2 ± 99.77 .42 Ferritin 51.54 ± 47.48 128.0 ± 143.21 .015 Glucose 104.5 ± 15.41 131.3 ± 58.19 .030 Uric acid 6.0 ± 1.13 5.94 ± 1.59 .89 Total cholesterol 210.0 ± 32.09 207.8 ± 36.62 .83 Triglycerides 190.5 ± 85.88 133.3 ± 84.95 .038 HDL cholesterol 38.38 ± 12.45 42.47 ± 15.98 .35 LDL cholesterol 138.2 ± 29.57 141.0 ± 29.11 .76 Serum iron 63.12 ± 24.75 68.67 ± 28.14 .49 Total iron binding capacity 344.2 ± 55.35 303.6 ± 43.19 .013 C-reactive protein 0.75 ± 0.50 0.57 ± 0.43 .20 LDL/HDL 4.00 ± 1.59 3.74 ± 1.43 .59
Values are mean ± SD.
Blood
measurementIron reduction
(n = 26)Control
(n = 18)P value Hemoglobin -0.61 ± 0.91 0.028 ± 0.73 .018 Hematocrit -1.85 ± 2.71 0.44 ± 2.59 .0076 Fibrinogen -0.96 ± 76.55 -17.89 ± 86.35 .45 Ferritin -73.38 ± 49.00 9.33 ± 82.47 .00015 Glucose -0.73 ± 15.04 2.89 ± 12.47 .41 Uric acid -0.04 ± 1.06 -0.056 ± 0.87 .95 Total cholesterol -11.73 ± 32.11 -17.72 ± 30.83 .54 Triglycerides -20.15 ± 58.77 -28.76 ± 47.89 .62 HDL cholesterol 2.15 ± 9.35 2.35 ± 7.62 .94 LDL cholesterol -9.89 ± 33.46 -15.88 ± 29.89 .55 Serum iron -13.27 ± 29.43 -7.71 ± 35.14 .58 Total iron binding capacity 39.92 ± 31.79 6.77 ± 21.34 .00055 C-reactive protein 0.0038 ± 0.61 0.039 ± 0.43 .83 LDL/HDL -0.47 ± 1.33 -0.56 ± 1.04 .81
Values are mean ± SD.
Event Interval (days)
from entryControl group (n = 18) Myocardial infarction 91 Myocardial infarction 60 Unstable angina 92 Unstable angina 60 Dysrhythmia 92 Dysrhythmia 63 Congestive heart failure 91 Phlebotomy group (n = 26) Angioplasty 8
All material on this website is protected by copyright. Copyright © 1994-2000 by Medscape Inc. All rights reserved. This website also contains material copyrighted by 3rd parties. CME means Continuing Medical Education credit is available. Medscape requires 3.x browsers or better from Netscape or Microsoft.