- Email: [email protected]
Iron stores and HFE genotypes are not related to increased risk of first-time myocardial infarction ☆
International Journal of Cardiology 150 (2011) 169–172
Contents lists available at ScienceDirect
International Journal of Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j c a r d
Iron stores and HFE genotypes are not related to increased risk of first-time myocardial infarction ☆ A prospective nested case-referent study Kim Ekblom a,⁎, Stefan L Marklund a, Jan-Håkan Jansson b, Göran Hallmans c, Lars Weinehall d, Johan Hultdin a a
Clinical Chemistry, Department of Medical Biosciences, Umeå University, Umeå, Sweden Department of Medicine, Skellefteå County Hospital, Skellefteå, Sweden c Nutrition Research, Umeå University, Umeå, Sweden d Epidemiology, Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden b
a r t i c l e
i n f o
Article history: Received 19 October 2009 Received in revised form 27 March 2010 Accepted 2 April 2010 Available online 5 May 2010 Keywords: Myocardial Infarction Iron Ferritin HFE Prospective
a b s t r a c t Objectives: Our objectives were to study the relationship between iron stores, HFE genotypes and the risk for first-ever myocardial infarction. Methods: First-ever myocardial infarction cases (n = 618) and double matched referents from the Northern Sweden Health and Disease Cohort Study were studied in a prospective nested case-referent setting. Plasma iron, total iron binding capacity, transferrin iron saturation and ferritin were analyzed, as well as several confounders. HFE C282Y and H63D genotypes were determined. Results: There was an inverse risk association for myocardial infarction in the highest quartiles of iron (OR 0.68; 95% CI 0.48–0.96) and transferrin iron saturation (OR 0.62; 95% CI 0.42–0.89) in men. This association, however, was lost after adjusting for C-reactive protein. Women homozygous for H63D had a higher risk for myocardial infarction. Conclusions: No risk association between high iron stores and first-ever myocardial infarction was found. The higher risk in female H63D homozygotes is probably not related to iron metabolism. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction High iron stores have been suggested to contribute to the pathogenesis of several diseases. Lipid peroxidation in lipoproteins is believed to be a major factor behind atherosclerosis and subsequent cardiovascular disease [1]. This peroxidation can efficiently be initiated by iron ions [2]. Sullivan hypothesized in 1981 that lower iron stores protect premenopausal women from atherosclerosis [3]. After that, a great number of studies have been published, with contradicting results. A limited proportion of these have a prospective design and a well defined population of myocardial infarction cases. Low serum iron has been associated with both higher [4–6] and lower [7] risk for myocardial infarction. No risk association has been found for transferrin iron saturation [6,8,9]. For total iron binding capacity
☆ The study was financed with grants from the Kempe Foundations, the Swedish Science Council, Västerbotten County Council and Umeå University Medical Research Foundation. ⁎ Corresponding author. Department of Medical Biosciences, Clinical Chemistry, Umeå University, SE-901 85, Umeå, Sweden. Tel.: + 46 90 785 2458; fax: + 46 90 77 81 97. E-mail address: [email protected] (K. Ekblom). 0167-5273/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2010.04.001
(TIBC) and the equivalent transferrin, a positive risk association [5], an inverse risk association [10] and no association [6] has been reported. Ferritin has been suggested to be a better marker for iron stores than transferrin iron saturation [11]. Only one study has found a risk association between ferritin and myocardial infarction [12]. The result from that study has not been replicated [13]. A low ratio between soluble transferrin receptor and ferritin, indicating large iron stores, has been reported to increase the risk for myocardial infarction [14]. Two common hemochromatosis-related polymorphisms in the HFE gene correlate with iron stores in humans; C282Y has a stronger correlation than H63D [15]. Prospective studies on the risk of HFE genotypes and the risk of myocardial infarction are scarce. A Danish group performed a prospective study in addition to a case–control study, and found no risk association between C282Y or H63D, and ischemic heart disease in neither of the studies [16]. A Dutch study found an association between C282Y heterozygosity and cardiovascular death in women [17]. Our aims with this prospective nested case-referent study were to investigate how iron status and HFE genotypes affect the risk of firstever myocardial infarction. We used plasma iron concentration, ferritin and transferrin iron saturation to estimate iron stores and total iron binding capacity (TIBC) as an inverse marker for iron status.
170
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172
2. Materials and methods 2.1. Study cohort: identification of cases and selection of referents Subjects from the Northern Sweden Health and Disease Study Cohort (NSHDS) were included in this nested case-referent study. Recruitment was performed within three projects with identical procedures for blood samples and using the same questionnaire: the Västerbotten Intervention Project (VIP), the Northern Sweden WHO Monitoring of Trends and Cardiovascular Disease (MONICA) Study and the local Mammography Screening Project (MSP). These have previously been described in detail [18]. Subjects in the Västerbotten Intervention Project were recruited through a free health examination offered to all residents of Västerbotten county upon turning 30, 40, 50, and 60 years. The mean participation rate has been 57%. In the MONICA cohort randomized subjects aged 25 to 74 years were invited to participate in health surveys during 1986, 1990, 1994, and 1999. The mean participation rate was 77.2%. Totally, 6952 subjects participated in the MONICA surveys. For the Mammary Screening Program cohort, all women within 40–70 years of age were invited to undergo mammography every 2 or 3 years. Participation rate for screening was 85% and 57% for donation of blood sample. In all three programs the participants were invited to donate a blood sample to be stored in the Medical Biobank, Umeå, Sweden. They also completed a large standardized questionnaire, including a food frequency questionnaire with detailed questions on alcohol intake, except in MSP, where the cases and their matched referents completed a short questionnaire. Alcohol intake (g/day) was calculated from these questionnaires, but this information was available only for a subset (210 cases, 447 referents). First-ever myocardial infarction cases (ICD-9 410, ICD-10 I21-3) registered 1986– 1999 in the counties of Västerbotten and Norrbotten in subjects between the ages of 25 and 64 years, were identified: NSHDS was linked with the Northern Sweden MONICA registry using Swedish personal numbers as the linkage variable. Suspected events were screened according to WHO MONICA criteria and validated using hospital records, general practitioner's reports, death certificates, and, when available, autopsy reports. Cases with a previous myocardial infarction, stroke or cancer diagnosis in the 5 years prior to, or 1 year after diagnosis with myocardial infarction were excluded, as were cases with insufficient plasma at baseline. Referents were excluded if they had myocardial infarction, stroke, cancer or death prior to the time of diagnosis of the index case. Two referents were selected for each case, matched for sex, age, date of health survey, project (VIP, MONICA, or MSP), and geographic area. Plasma specimens from cases and referents were analyzed in triplets of one case and two referents: the position within the triplet was randomly varied to avoid systemic bias and inter-assay variability. The investigators and laboratory staff had no knowledge of case and referent status. The study protocol was approved by the Research Ethics Committee of Umeå University, Umeå.
Table 1 Baseline characteristics of referents and myocardial infarction cases. Median ± inter quartile range for continuous variables. Proportions/percent for non-continuous variables.
Sex (M/F) Age at sample donation (years) Age at event (years) Men Women Lag time to myocardial infarction (years) BMI (kg/m2) Systolic blood pressure (mm Hg) Hypertension (%)b Diabetes (%) Smoking (%) Alcohol intake (g/day) ApoB/ApoA1 Cholesterol hsCRP Iron TIBC Transferrin iron saturation (%) Ferritin Men Iron TIBC Transferrin iron saturation (%) Ferritin Women Iron Transferrin iron saturation (%) TIBC Ferritin
n
Referents
n
Cases
pa
1184 1184
72.8/27.2 59.2 ± 10.1
618 618
72.3/27.3 59.5 ± 10.1
Matched Matched
n.a. n.a. n.a. n.a.
n.a. n.a. n.a. n.a.
618 449 169 618
60.5 ± 11.2 60.1 ± 11.2 62.8 ± 11.6 3.5 ± 3.7
n.a. n.a. n.a. n.a.
1184 1011
26.1 ± 5.6 131.4 ± 22.4
618 527
27.5 ± 6.1 140.0 ± 22.4
b 0.001 b 0.001
1016 855 1148 447 1033 1011 1016 989 988 988
46.1 5.5 19.4 4.0 (2.0–7.0) 0.8 ± 0.3 6.1 ± 1.6 1.2 ± 1.9 19.7 ± 8.0 51.9 ± 9.1 37.5 ± 16.9
527 444 584 210 541 522 539 524 523 523
64.5 10.6 39.7 3.4 (1.4–5.8) 0.9 ± 0.4 6.6 ± 1.7 2.0 ± 2.8 18.6 ± 8.1 52.8 ± 10.1 35.7 ± 15.2
b 0.001 0.001 b 0.001 0.020 b 0.001 b 0.001 b 0.001 0.080 0.162 0.031
987
131.5 ± 142.7
525
130. 6 ± 150.4
0.295
699 699 699
20.4 ± 7.7 51.6 ± 8.3 38.9 ± 16.3
390 370 370
19.4 ± 8.1 52.0 ± 9.3 38.1 ± 14.5
0.060 0.311 0.035
697
156.1 ± 162.4
372
160.1 ± 171.3
0.587
290 289
17.8 ± 7.7 33.7 ± 17.1
154 153
17.6 ± 7.2 31.6 ± 15.6
0.763 0.494
289 290
52.7 ± 10.0 85.0 ± 83.9
153 153
53.9 ± 10.6 87.4 ± 83.0
0.354 0.186
a
Calculated by Mann–Whitney U-test. Defined as systolic blood pressure ≥ 140 and/or diastolic blood pressure ≥90 mm Hg, and/or on anti-hypertensive medication. b
2.2. Sample collection and laboratory procedures Venous blood samples were collected in evacuated glass tubes containing heparin and were centrifuged at 1500 g for 15 min to obtain heparinized plasma. The plasma samples were stored at − 80 °C. Samples were collected in the morning, after at least 4 h of fasting in the VIP and MONICA projects. Samples were collected throughout the day in the MSP project. In summary, 73.8% of the subjects had fasted for 4 h or more; 14.9% had fasted for less than 4 h, and for 11.4% information about fasting status was missing (data not shown). There was no difference in fasting status between cases and referents (Mann–Whitney U-test p = 0.296). A Hitachi 911 multianalyzer was used for analyzing plasma iron parameters using kits from the manufacturer (Roche Diagnostics GmbH, Mannheim Germany). The methods are described in detail in the supplement. TaqMan allelic discrimination methodology was used for genotyping C282Y and H63D in the HFE gene using equipment and assays from Applied Biosystems (Foster City, CA, USA), as previously described [19]. 2.3. Statistical analysis Baseline characteristics for cases and referents were compared using Mann– Whitney U-test. Conditional logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for genotypes and for quartiles of continuous variables. The quartiles were based on the distribution of referent subjects. Quartiles were coded 1–4 (lowest to highest) and genotypes 1–3 (homozygote wild type, heterozygote, and homozygote mutant). When testing for trend, regression analysis was performed treating the variables as continuous variables. Missing values were treated as missing throughout the calculations. Hardy–Weinberg equilibrium was calculated by χ2 test. Statistical tests and corresponding p-values were two-sided. All statistical analyses were performed with SPSS version 17.0 (Chicago, IL, USA).
3. Results The main characteristics of the study population are presented in Table 1. Plasma transferrin iron saturation was higher in referents
Table 2 Risk of myocardial infarction according to HFE C282Y and HFE H63D genotypes and odds ratios (95% CI). N (cases/referents) is given for unadjusted models. ptrenda
HFE C282Y
CC (ref)
CY
YY
N Full study group N Men N Women
(502/928) 1.00 (366/674) 1.00 (136/254) 1.00
(81/162) 0.90 (0.67–1.19) (58/121) 0.84 (0.60–1.18) (23/41) 1.02(0.59–1.75)
(4/4) 2.62 (0.59–11.73) (3/3) 2.95 (0.49–17.67) (1/1) 2.01 (0.12–32.24)
HFE H63D
HH (ref)
HD
DD
ptrenda
N Full study group Adjusted for Iron Iron saturation TIBC Ferritin N Men N Women Adjusted for Iron Iron saturation TIBC Ferritin
(442/854) 1.00
(129/228) 1.09 (0.85–1.41)
(6/15) 2.02 (0.96–4.25)
0.13
1.00 1.00 1.00 1.00 (327/623) 1.00 (115/231) 1.00
1.17 1.17 1.15 1.13
(0.89–1.55) (0.89–1.55) (0.87–1.52) (0.86–1.49) (91/165) 1.04 (0.76–1.40) (38/63) 1.22 (0.77–1.93)
2.42 2.45 2.22 2.27
(1.11–5.30) (1.12–5.36) (1.02–4.86) (1.04–4.94) (7/12) 1.26 (0.48–3.28) (7/3) 4.76 (1.23–18.45)
0.038 0.035 0.060 0.070
1.00 1.00 1.00 1.00
1.52 1.51 1.51 1.62
4.97 5.00 4.85 4.66
0.006 0.006 0.007 0.005
(0.94–2.48) (0.93–2.46) (0.92–2.45) (0.98–2.69)
(1.27–19.41) (1.28–19.50) (1.25–18.83) (1.19–18.30)
0.74 0.60 0.83
0.68 0.041
a Calculated by analyzing the genotypes as continuous variables in conditional logistic regression.
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172 Table 3 Risk for first-time myocardial infarction for quartiles of iron and transferrin iron saturation, OR (95% CI) and p for trend. Q1
Q2
P-Iron Unadjusted Ref 0.84 (0.63–1.12) Adjusteda Ref 0.90 (0.67–1.22) Adjustedb Ref 0.77 (0.51–1.15) c Adjusted Ref 0.78 (0.52–1.18) Men Unadjusted Ref 0.73 (0.52–1.04) Ref 0.75 (0.53–1.06) Adjusteda Adjustedb Ref 0.69 (0.44–1.07) Adjustedc Ref 0.68 (0.44–1.06) Women Unadjusted Ref 1.16 (0.68–1.99) Adjusteda Ref 1.52 (0.85–2.72) Adjustedb Ref 1.48 (0.49–4.41) Adjustedc Ref 1.67 (0.52–5.34) P-Transferrin iron saturation Unadjusted Ref 0.92 (0.69–1.22) Adjusteda Ref 1.00 (0.74–1.38) Adjustedb Ref 0.79 (0.53–1.17) c Ref 0.79 (0.53–1.18) Adjusted Men Unadjusted Ref 0.73 (0.51–1.03) Adjusteda Ref 0.75 (0.52–1.07) Adjustedb Ref 0.68 (0.44–1.06) c Adjusted Ref 0.68 (0.44–1.06) Women Unadjusted Ref 1.50 (0.89–2.52) Adjusteda Ref 2.02 (1.14–3.55) Adjustedb Ref 1.53 (0.53–4.37) Adjustedc Ref 1.61 (0.53–4.88)
Q3
Q4
ptrend
0.82 0.90 0.93 0.95
(0.61–1.10) (0.66–1.22) (0.62–1.39) (0.63–1.44)
0.74 (0.55–1.00) 0.83 (0.61–1.13) 0.64 (0.43–0.97) 0.68 (0.45–1.03)
0.056 0.26 0.076 0.14
0.70 0.75 0.83 0.82
(0.49–1.01) (0.52–1.08) (0.54–1.30) (0.53–1.29)
0.68 (0.48–0.96) 0.73 (0.51–1.05) 0.60 (0.38–0.94) 0.62 (0.39–0.97)
0.033 0.11 0.056 0.082
1.18 1.48 1.49 1.83
(0.68–2.05) (0.82–2.69) (0.48–4.65) (0.56–5.98)
0.93 (0.53–1.64) 1.19 (0.65–2.19) 0.84 (0.29–2.37) 1.10 (0.35–3.48)
0.84 0.60 0.60 1.00
0.92 1.00 0.99 0.99
(0.67–1.25) (0.73–1.38) (0.65–1.51) (0.65–1.53)
0.73 (0.54–1.00) 0.81 (0.59–1.11) 0.63 (0.42–0.96) 0.65 (0.42– 099)
0.065 0.22 0.074 0.10
0.91 0.96 1.06 1.05
(0.63–1.30) (0.66–1.38) (0.67–1.69) (0.66–1.67)
0.62 (0.42–0.89) 0.66 (0.45–0.96) 0.55 (0.34–0.88) 0.56 (0.34–0.89)
0.037 0.092 0.066 0.078
0.90 1.16 0.56 0.72
(0.49–1.65) (0.60–2.24) (0.16–1.93) (0.20–2.60)
1.11 (0.62–1.99) 1.41 (0.76–2.63) 1.00 (0.36–2.77) 1.20 (0.41–3.51)
0.88 0.64 0.60 0.91
Quartile limits were for P-Iron: Men 16.6, 20.4 and 24.3, Women 13.9, 17.8 and 21.6; P-Transferrin iron saturation: Men 32.0, 38.9 and 48.3, Women 25.1, 33.7 and 42.3. a Adjusted for hsCRP. b Adjusted for age, BMI, systolic blood pressure, smoking, Apo B/Apo A1 and diabetes. c Adjusted for age, BMI, systolic blood pressure, smoking, hsCRP and Apo B/Apo A1 and diabetes.
than in cases. Stratified for sex, this finding was only significant in men. No differences between referents and cases were seen for plasma iron, TIBC and ferritin. For HFE C282Y and HFE H63D both referents and cases were in Hardy–Weinberg equilibrium. Associations between the HFE genotypes and iron parameters are shown in Supplementary Table 1. The HFE C282Y genotype did not affect the risk for myocardial infarction (Table 2). Adjustment for iron parameters did not change the outcome. For HFE H63D homozygotes, a 5-fold increase in risk was found for women only, this was unaffected by adjustments. No effects were seen in the full study group or among men, although adjustment for iron parameters revealed a weak risk association in the full study group (Table 2). Quartiles of TIBC and ferritin were not associated with risk for myocardial infarction in univariate analysis or after adjusting for other risk factors (Supplementary Table 2). Plasma iron in the highest quartile was associated with a decrease in risk for myocardial infarction among males in univariate analysis (Table 3). In the full study group there was a tendency of decrease in risk in the highest quartile. No associations were seen among women. Likewise, transferrin iron saturation in the highest quartile was associated with a decrease in risk in the full study group and among men. Adjustment for age, BMI, systolic blood pressure, smoking, diabetes and ApoB/ApoA1 ratio did not change the results. As plasma iron and transferrin saturation levels are known to be influenced by inflammation and alcohol intake, both of which differed between cases and referents, we analyzed the correlations between high-sensitive C-reactive protein (hsCRP), alcohol intake and the iron parameters (Supplementary Table 3). HsCRP showed significant negative correlations with plasma iron, TIBC and transferrin iron
171
saturation. Alcohol intake showed significant positive correlations with plasma iron and ferritin. In multivariate analysis for risk of myocardial infarction, significances for iron and transferrin iron saturation were lost when adjusting for hsCRP (Table 3). Since data on alcohol intake was available only for a subset, the statistical power did not allow reliable adjustment for this parameter. 4. Discussion To our knowledge, this is the first study on a large cohort of well defined first-ever myocardial infarction cases on the risk associations of plasma iron parameters and the clinically important HFE genotypes, C282Y and H63D. We found an inverse risk association with myocardial infarction in the highest plasma iron quartile. This is consistent with the Helsinki Heart Study on dyslipidemic men [4]. It is also consistent with the findings of the NHANES I follow up study and a Finnish cohort study on elderly subjects [5,6], although our findings were significant for men only. Our findings on previously healthy subjects are contradictory to the findings of a Canadian cohort study which included fatal myocardial infarction only [7]. Consistent with our previous finding on stroke [19], we found a lower risk in the highest transferrin iron saturation quartile among men. The lack of risk association between ferritin and myocardial infarction is consistent with most other studies on cardiovascular disease [20,21] and on myocardial infarction [13]. Considering the pro-oxidant properties of iron, our findings were unexpected. As inflammation has an impact on iron metabolism [22], we adjusted our results for hsCRP. This adjustment eliminated the association between high levels of iron and risk for myocardial infarction. This is remarkable considering the low grade of inflammation in the subjects mirrored by low levels of CRP in cases and referents. Another potential confounder is use of alcohol. Alcohol tends to increase iron levels, possibly via down-regulation of hepcidin secretion in the liver [23]. Alcohol intake was higher in referents compared to cases (Table 1), and was correlated with plasma iron levels (Supplementary Table 3). Our data highlight the importance of adjustment for inflammation and alcohol intake when studying impact of iron metabolism on disease. Inflammation has also previously been used as an explanation why low serum iron is associated with higher cardiovascular risk [4]. An interesting finding in this study is that HFE H63D homozygosity was associated with higher risk for myocardial infarction in women. With respect to the plasma markers of iron stores, this effect is probably not due to the iron metabolism, as this polymorphism is a weak determinant for iron levels [15]. The polymorphism may be linked to another locus influencing the risk for cardiovascular disease. H63D homozygosity has previously been reported to be associated with increased self-reported heart disease in women [15]. In a metaanalysis, this homozygosity was associated with borderline increase in risk for heart disease [24]. This study had many advantages compared with previous studies. It had prospective design, and a large number of well defined cases of both sexes. The population was homogenous regarding ethnicity. We were also able to adjust for low-grade inflammation as a confounder. A unique feature was the ability to examine both plasma and DNA samples simultaneously. This study also had some disadvantages. The number of HFE C282Y homozygotes was very low, as were subjects with extremely high iron stores. Therefore we cannot make firm conclusion about the risk of myocardial infarction in such subjects. However, no increase in risk was noted in upper normal range. In summary, iron levels, at least in the upper normal range, were not associated with higher risk for first-time myocardial infarction in this prospective study on Caucasians: upper normal iron levels seemed to be associated with lower risk, but this association was no longer evident after adjusting for inflammation. In some countries iron fortification has been discontinued due to fear of elevated risk for cardiovascular events [25], resulting in increased occurrence of iron
172
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172
deficiency [26]. Our results do not support the hypothesis of increased risk of heart disease associated with iron fortification. Acknowledgements We thank Åsa Ågren at the Medical Biobank and Birgitta Nilsson at the Clinical Chemistry Laboratory at Umeå University Hospital for excellent technical assistance. We also acknowledge Bethany Van Guelpen at the Department of Biomedical Sciences, Pathology for assistance with dietary data on alcohol intake. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [27]. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ijcard.2010.04.001. References [1] Steinberg D. Low density lipoprotein oxidation and its pathobiological significance. J Biol Chem 1997;272:20963–6. [2] Girotti AW. Mechanisms of lipid peroxidation. J Free Radic Biol Med 1985;1:87–95. [3] Sullivan JL. Iron and the sex difference in heart disease risk. Lancet 1981;1:1293–4. [4] Kervinen H, Tenkanen L, Palosuo T, Roivainen M, Manninen V, Manttari M. Serum iron, infection and inflammation; effects on coronary risk. Scand Cardiovasc J 2004;38:345–8. [5] Marniemi J, Alanen E, Impivaara O, et al. Dietary and serum vitamins and minerals as predictors of myocardial infarction and stroke in elderly subjects. Nutr Metab Cardiovasc Dis 2005;15:188–97. [6] Liao Y, Cooper RS, McGee DL. Iron status and coronary heart disease: negative findings from the NHANES I epidemiologic follow-up study. Am J Epidemiol 1994;139:704–12. [7] Morrison HI, Semenciw RM, Mao Y, Wigle DT. Serum iron and risk of fatal acute myocardial infarction. Epidemiology 1994;5:243–6. [8] Baer DM, Tekawa IS, Hurley LB. Iron stores are not associated with acute myocardial infarction. Circulation 1994;89:2915–8. [9] Sempos CT, Looker AC, Gillum RF, Makuc DM. Body iron stores and the risk of coronary heart disease. N Engl J Med 1994;330:1119–24.
[10] Magnusson MK, Sigfusson N, Sigvaldason H, Johannesson GM, Magnusson S, Thorgeirsson G. Low iron-binding capacity as a risk factor for myocardial infarction. Circulation 1994;89:102–8. [11] Salonen JT, Nyyssonen K, Salonen R. Body iron stores and the risk of coronary heart disease. N Engl J Med 1994;331:1159 author reply 60. [12] Salonen JT, Nyyssonen K, Korpela H, Tuomilehto J, Seppanen R, Salonen R. High stored iron levels are associated with excess risk of myocardial infarction in eastern Finnish men. Circulation 1992;86:803–11. [13] Frey GH, Krider DW. Serum ferritin and myocardial infarct. W V Med J 1994;90: 13–5. [14] Tuomainen TP, Punnonen K, Nyyssonen K, Salonen JT. Association between body iron stores and the risk of acute myocardial infarction in men. Circulation 1998;97:1461–6. [15] Adams PC, Reboussin DM, Barton JC, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352:1769–78. [16] Ellervik C, Tybjaerg-Hansen A, Grande P, Appleyard M, Nordestgaard BG. Hereditary hemochromatosis and risk of ischemic heart disease: a prospective study and a case–control study. Circulation 2005;112:185–93. [17] Roest M, van der Schouw YT, de Valk B, et al. Heterozygosity for a hereditary hemochromatosis gene is associated with cardiovascular death in women. Circulation 1999;100:1268–73. [18] Hallmans G, Agren A, Johansson G, et al. Cardiovascular disease and diabetes in the Northern Sweden Health and Disease Study Cohort — evaluation of risk factors and their interactions. Scand J Public Health Suppl 2003;61:18–24. [19] Ekblom K, Hultdin J, Stegmayr B, et al. Iron stores and HFE genotypes are not related to increased risk of ischemic stroke. A prospective nested case-referent study. Cerebrovasc Dis 2007;24:405–11. [20] Aronow WS, Ahn C. Three-year follow-up shows no association of serum ferritin levels with incidence of new coronary events in 577 persons aged N or = 62 years. Am J Cardiol 1996;78:678–9. [21] Manttari M, Manninen V, Huttunen JK, et al. Serum ferritin and ceruloplasmin as coronary risk factors. Eur Heart J 1994;15:1599–603. [22] Ganz T, Nemeth E. Iron sequestration and anemia of inflammation. Semin Hematol 2009;46:387–93. [23] Harrison-Findik DD. Role of alcohol in the regulation of iron metabolism. World J Gastroenterol 2007;13:4925–30. [24] Ellervik C, Birgens H, Tybjaerg-Hansen A, Nordestgaard BG. Hemochromatosis genotypes and risk of 31 disease endpoints: meta-analyses including 66,000 cases and 226,000 controls. Hepatology 2007;46:1071–80. [25] Lynch SR. The impact of iron fortification on nutritional anaemia. Best Pract Res Clin Haematol 2005;18:333–46. [26] Hallberg L, Hulthen L. Perspectives on iron absorption. Blood Cells Mol Dis 2002;29:562–73. [27] Coats AJ. Ethical authorship and publishing. Int J Cardiol 2009;131:149–50.
Contents lists available at ScienceDirect
International Journal of Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j c a r d
Iron stores and HFE genotypes are not related to increased risk of first-time myocardial infarction ☆ A prospective nested case-referent study Kim Ekblom a,⁎, Stefan L Marklund a, Jan-Håkan Jansson b, Göran Hallmans c, Lars Weinehall d, Johan Hultdin a a
Clinical Chemistry, Department of Medical Biosciences, Umeå University, Umeå, Sweden Department of Medicine, Skellefteå County Hospital, Skellefteå, Sweden c Nutrition Research, Umeå University, Umeå, Sweden d Epidemiology, Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden b
a r t i c l e
i n f o
Article history: Received 19 October 2009 Received in revised form 27 March 2010 Accepted 2 April 2010 Available online 5 May 2010 Keywords: Myocardial Infarction Iron Ferritin HFE Prospective
a b s t r a c t Objectives: Our objectives were to study the relationship between iron stores, HFE genotypes and the risk for first-ever myocardial infarction. Methods: First-ever myocardial infarction cases (n = 618) and double matched referents from the Northern Sweden Health and Disease Cohort Study were studied in a prospective nested case-referent setting. Plasma iron, total iron binding capacity, transferrin iron saturation and ferritin were analyzed, as well as several confounders. HFE C282Y and H63D genotypes were determined. Results: There was an inverse risk association for myocardial infarction in the highest quartiles of iron (OR 0.68; 95% CI 0.48–0.96) and transferrin iron saturation (OR 0.62; 95% CI 0.42–0.89) in men. This association, however, was lost after adjusting for C-reactive protein. Women homozygous for H63D had a higher risk for myocardial infarction. Conclusions: No risk association between high iron stores and first-ever myocardial infarction was found. The higher risk in female H63D homozygotes is probably not related to iron metabolism. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction High iron stores have been suggested to contribute to the pathogenesis of several diseases. Lipid peroxidation in lipoproteins is believed to be a major factor behind atherosclerosis and subsequent cardiovascular disease [1]. This peroxidation can efficiently be initiated by iron ions [2]. Sullivan hypothesized in 1981 that lower iron stores protect premenopausal women from atherosclerosis [3]. After that, a great number of studies have been published, with contradicting results. A limited proportion of these have a prospective design and a well defined population of myocardial infarction cases. Low serum iron has been associated with both higher [4–6] and lower [7] risk for myocardial infarction. No risk association has been found for transferrin iron saturation [6,8,9]. For total iron binding capacity
☆ The study was financed with grants from the Kempe Foundations, the Swedish Science Council, Västerbotten County Council and Umeå University Medical Research Foundation. ⁎ Corresponding author. Department of Medical Biosciences, Clinical Chemistry, Umeå University, SE-901 85, Umeå, Sweden. Tel.: + 46 90 785 2458; fax: + 46 90 77 81 97. E-mail address: [email protected] (K. Ekblom). 0167-5273/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2010.04.001
(TIBC) and the equivalent transferrin, a positive risk association [5], an inverse risk association [10] and no association [6] has been reported. Ferritin has been suggested to be a better marker for iron stores than transferrin iron saturation [11]. Only one study has found a risk association between ferritin and myocardial infarction [12]. The result from that study has not been replicated [13]. A low ratio between soluble transferrin receptor and ferritin, indicating large iron stores, has been reported to increase the risk for myocardial infarction [14]. Two common hemochromatosis-related polymorphisms in the HFE gene correlate with iron stores in humans; C282Y has a stronger correlation than H63D [15]. Prospective studies on the risk of HFE genotypes and the risk of myocardial infarction are scarce. A Danish group performed a prospective study in addition to a case–control study, and found no risk association between C282Y or H63D, and ischemic heart disease in neither of the studies [16]. A Dutch study found an association between C282Y heterozygosity and cardiovascular death in women [17]. Our aims with this prospective nested case-referent study were to investigate how iron status and HFE genotypes affect the risk of firstever myocardial infarction. We used plasma iron concentration, ferritin and transferrin iron saturation to estimate iron stores and total iron binding capacity (TIBC) as an inverse marker for iron status.
170
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172
2. Materials and methods 2.1. Study cohort: identification of cases and selection of referents Subjects from the Northern Sweden Health and Disease Study Cohort (NSHDS) were included in this nested case-referent study. Recruitment was performed within three projects with identical procedures for blood samples and using the same questionnaire: the Västerbotten Intervention Project (VIP), the Northern Sweden WHO Monitoring of Trends and Cardiovascular Disease (MONICA) Study and the local Mammography Screening Project (MSP). These have previously been described in detail [18]. Subjects in the Västerbotten Intervention Project were recruited through a free health examination offered to all residents of Västerbotten county upon turning 30, 40, 50, and 60 years. The mean participation rate has been 57%. In the MONICA cohort randomized subjects aged 25 to 74 years were invited to participate in health surveys during 1986, 1990, 1994, and 1999. The mean participation rate was 77.2%. Totally, 6952 subjects participated in the MONICA surveys. For the Mammary Screening Program cohort, all women within 40–70 years of age were invited to undergo mammography every 2 or 3 years. Participation rate for screening was 85% and 57% for donation of blood sample. In all three programs the participants were invited to donate a blood sample to be stored in the Medical Biobank, Umeå, Sweden. They also completed a large standardized questionnaire, including a food frequency questionnaire with detailed questions on alcohol intake, except in MSP, where the cases and their matched referents completed a short questionnaire. Alcohol intake (g/day) was calculated from these questionnaires, but this information was available only for a subset (210 cases, 447 referents). First-ever myocardial infarction cases (ICD-9 410, ICD-10 I21-3) registered 1986– 1999 in the counties of Västerbotten and Norrbotten in subjects between the ages of 25 and 64 years, were identified: NSHDS was linked with the Northern Sweden MONICA registry using Swedish personal numbers as the linkage variable. Suspected events were screened according to WHO MONICA criteria and validated using hospital records, general practitioner's reports, death certificates, and, when available, autopsy reports. Cases with a previous myocardial infarction, stroke or cancer diagnosis in the 5 years prior to, or 1 year after diagnosis with myocardial infarction were excluded, as were cases with insufficient plasma at baseline. Referents were excluded if they had myocardial infarction, stroke, cancer or death prior to the time of diagnosis of the index case. Two referents were selected for each case, matched for sex, age, date of health survey, project (VIP, MONICA, or MSP), and geographic area. Plasma specimens from cases and referents were analyzed in triplets of one case and two referents: the position within the triplet was randomly varied to avoid systemic bias and inter-assay variability. The investigators and laboratory staff had no knowledge of case and referent status. The study protocol was approved by the Research Ethics Committee of Umeå University, Umeå.
Table 1 Baseline characteristics of referents and myocardial infarction cases. Median ± inter quartile range for continuous variables. Proportions/percent for non-continuous variables.
Sex (M/F) Age at sample donation (years) Age at event (years) Men Women Lag time to myocardial infarction (years) BMI (kg/m2) Systolic blood pressure (mm Hg) Hypertension (%)b Diabetes (%) Smoking (%) Alcohol intake (g/day) ApoB/ApoA1 Cholesterol hsCRP Iron TIBC Transferrin iron saturation (%) Ferritin Men Iron TIBC Transferrin iron saturation (%) Ferritin Women Iron Transferrin iron saturation (%) TIBC Ferritin
n
Referents
n
Cases
pa
1184 1184
72.8/27.2 59.2 ± 10.1
618 618
72.3/27.3 59.5 ± 10.1
Matched Matched
n.a. n.a. n.a. n.a.
n.a. n.a. n.a. n.a.
618 449 169 618
60.5 ± 11.2 60.1 ± 11.2 62.8 ± 11.6 3.5 ± 3.7
n.a. n.a. n.a. n.a.
1184 1011
26.1 ± 5.6 131.4 ± 22.4
618 527
27.5 ± 6.1 140.0 ± 22.4
b 0.001 b 0.001
1016 855 1148 447 1033 1011 1016 989 988 988
46.1 5.5 19.4 4.0 (2.0–7.0) 0.8 ± 0.3 6.1 ± 1.6 1.2 ± 1.9 19.7 ± 8.0 51.9 ± 9.1 37.5 ± 16.9
527 444 584 210 541 522 539 524 523 523
64.5 10.6 39.7 3.4 (1.4–5.8) 0.9 ± 0.4 6.6 ± 1.7 2.0 ± 2.8 18.6 ± 8.1 52.8 ± 10.1 35.7 ± 15.2
b 0.001 0.001 b 0.001 0.020 b 0.001 b 0.001 b 0.001 0.080 0.162 0.031
987
131.5 ± 142.7
525
130. 6 ± 150.4
0.295
699 699 699
20.4 ± 7.7 51.6 ± 8.3 38.9 ± 16.3
390 370 370
19.4 ± 8.1 52.0 ± 9.3 38.1 ± 14.5
0.060 0.311 0.035
697
156.1 ± 162.4
372
160.1 ± 171.3
0.587
290 289
17.8 ± 7.7 33.7 ± 17.1
154 153
17.6 ± 7.2 31.6 ± 15.6
0.763 0.494
289 290
52.7 ± 10.0 85.0 ± 83.9
153 153
53.9 ± 10.6 87.4 ± 83.0
0.354 0.186
a
Calculated by Mann–Whitney U-test. Defined as systolic blood pressure ≥ 140 and/or diastolic blood pressure ≥90 mm Hg, and/or on anti-hypertensive medication. b
2.2. Sample collection and laboratory procedures Venous blood samples were collected in evacuated glass tubes containing heparin and were centrifuged at 1500 g for 15 min to obtain heparinized plasma. The plasma samples were stored at − 80 °C. Samples were collected in the morning, after at least 4 h of fasting in the VIP and MONICA projects. Samples were collected throughout the day in the MSP project. In summary, 73.8% of the subjects had fasted for 4 h or more; 14.9% had fasted for less than 4 h, and for 11.4% information about fasting status was missing (data not shown). There was no difference in fasting status between cases and referents (Mann–Whitney U-test p = 0.296). A Hitachi 911 multianalyzer was used for analyzing plasma iron parameters using kits from the manufacturer (Roche Diagnostics GmbH, Mannheim Germany). The methods are described in detail in the supplement. TaqMan allelic discrimination methodology was used for genotyping C282Y and H63D in the HFE gene using equipment and assays from Applied Biosystems (Foster City, CA, USA), as previously described [19]. 2.3. Statistical analysis Baseline characteristics for cases and referents were compared using Mann– Whitney U-test. Conditional logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for genotypes and for quartiles of continuous variables. The quartiles were based on the distribution of referent subjects. Quartiles were coded 1–4 (lowest to highest) and genotypes 1–3 (homozygote wild type, heterozygote, and homozygote mutant). When testing for trend, regression analysis was performed treating the variables as continuous variables. Missing values were treated as missing throughout the calculations. Hardy–Weinberg equilibrium was calculated by χ2 test. Statistical tests and corresponding p-values were two-sided. All statistical analyses were performed with SPSS version 17.0 (Chicago, IL, USA).
3. Results The main characteristics of the study population are presented in Table 1. Plasma transferrin iron saturation was higher in referents
Table 2 Risk of myocardial infarction according to HFE C282Y and HFE H63D genotypes and odds ratios (95% CI). N (cases/referents) is given for unadjusted models. ptrenda
HFE C282Y
CC (ref)
CY
YY
N Full study group N Men N Women
(502/928) 1.00 (366/674) 1.00 (136/254) 1.00
(81/162) 0.90 (0.67–1.19) (58/121) 0.84 (0.60–1.18) (23/41) 1.02(0.59–1.75)
(4/4) 2.62 (0.59–11.73) (3/3) 2.95 (0.49–17.67) (1/1) 2.01 (0.12–32.24)
HFE H63D
HH (ref)
HD
DD
ptrenda
N Full study group Adjusted for Iron Iron saturation TIBC Ferritin N Men N Women Adjusted for Iron Iron saturation TIBC Ferritin
(442/854) 1.00
(129/228) 1.09 (0.85–1.41)
(6/15) 2.02 (0.96–4.25)
0.13
1.00 1.00 1.00 1.00 (327/623) 1.00 (115/231) 1.00
1.17 1.17 1.15 1.13
(0.89–1.55) (0.89–1.55) (0.87–1.52) (0.86–1.49) (91/165) 1.04 (0.76–1.40) (38/63) 1.22 (0.77–1.93)
2.42 2.45 2.22 2.27
(1.11–5.30) (1.12–5.36) (1.02–4.86) (1.04–4.94) (7/12) 1.26 (0.48–3.28) (7/3) 4.76 (1.23–18.45)
0.038 0.035 0.060 0.070
1.00 1.00 1.00 1.00
1.52 1.51 1.51 1.62
4.97 5.00 4.85 4.66
0.006 0.006 0.007 0.005
(0.94–2.48) (0.93–2.46) (0.92–2.45) (0.98–2.69)
(1.27–19.41) (1.28–19.50) (1.25–18.83) (1.19–18.30)
0.74 0.60 0.83
0.68 0.041
a Calculated by analyzing the genotypes as continuous variables in conditional logistic regression.
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172 Table 3 Risk for first-time myocardial infarction for quartiles of iron and transferrin iron saturation, OR (95% CI) and p for trend. Q1
Q2
P-Iron Unadjusted Ref 0.84 (0.63–1.12) Adjusteda Ref 0.90 (0.67–1.22) Adjustedb Ref 0.77 (0.51–1.15) c Adjusted Ref 0.78 (0.52–1.18) Men Unadjusted Ref 0.73 (0.52–1.04) Ref 0.75 (0.53–1.06) Adjusteda Adjustedb Ref 0.69 (0.44–1.07) Adjustedc Ref 0.68 (0.44–1.06) Women Unadjusted Ref 1.16 (0.68–1.99) Adjusteda Ref 1.52 (0.85–2.72) Adjustedb Ref 1.48 (0.49–4.41) Adjustedc Ref 1.67 (0.52–5.34) P-Transferrin iron saturation Unadjusted Ref 0.92 (0.69–1.22) Adjusteda Ref 1.00 (0.74–1.38) Adjustedb Ref 0.79 (0.53–1.17) c Ref 0.79 (0.53–1.18) Adjusted Men Unadjusted Ref 0.73 (0.51–1.03) Adjusteda Ref 0.75 (0.52–1.07) Adjustedb Ref 0.68 (0.44–1.06) c Adjusted Ref 0.68 (0.44–1.06) Women Unadjusted Ref 1.50 (0.89–2.52) Adjusteda Ref 2.02 (1.14–3.55) Adjustedb Ref 1.53 (0.53–4.37) Adjustedc Ref 1.61 (0.53–4.88)
Q3
Q4
ptrend
0.82 0.90 0.93 0.95
(0.61–1.10) (0.66–1.22) (0.62–1.39) (0.63–1.44)
0.74 (0.55–1.00) 0.83 (0.61–1.13) 0.64 (0.43–0.97) 0.68 (0.45–1.03)
0.056 0.26 0.076 0.14
0.70 0.75 0.83 0.82
(0.49–1.01) (0.52–1.08) (0.54–1.30) (0.53–1.29)
0.68 (0.48–0.96) 0.73 (0.51–1.05) 0.60 (0.38–0.94) 0.62 (0.39–0.97)
0.033 0.11 0.056 0.082
1.18 1.48 1.49 1.83
(0.68–2.05) (0.82–2.69) (0.48–4.65) (0.56–5.98)
0.93 (0.53–1.64) 1.19 (0.65–2.19) 0.84 (0.29–2.37) 1.10 (0.35–3.48)
0.84 0.60 0.60 1.00
0.92 1.00 0.99 0.99
(0.67–1.25) (0.73–1.38) (0.65–1.51) (0.65–1.53)
0.73 (0.54–1.00) 0.81 (0.59–1.11) 0.63 (0.42–0.96) 0.65 (0.42– 099)
0.065 0.22 0.074 0.10
0.91 0.96 1.06 1.05
(0.63–1.30) (0.66–1.38) (0.67–1.69) (0.66–1.67)
0.62 (0.42–0.89) 0.66 (0.45–0.96) 0.55 (0.34–0.88) 0.56 (0.34–0.89)
0.037 0.092 0.066 0.078
0.90 1.16 0.56 0.72
(0.49–1.65) (0.60–2.24) (0.16–1.93) (0.20–2.60)
1.11 (0.62–1.99) 1.41 (0.76–2.63) 1.00 (0.36–2.77) 1.20 (0.41–3.51)
0.88 0.64 0.60 0.91
Quartile limits were for P-Iron: Men 16.6, 20.4 and 24.3, Women 13.9, 17.8 and 21.6; P-Transferrin iron saturation: Men 32.0, 38.9 and 48.3, Women 25.1, 33.7 and 42.3. a Adjusted for hsCRP. b Adjusted for age, BMI, systolic blood pressure, smoking, Apo B/Apo A1 and diabetes. c Adjusted for age, BMI, systolic blood pressure, smoking, hsCRP and Apo B/Apo A1 and diabetes.
than in cases. Stratified for sex, this finding was only significant in men. No differences between referents and cases were seen for plasma iron, TIBC and ferritin. For HFE C282Y and HFE H63D both referents and cases were in Hardy–Weinberg equilibrium. Associations between the HFE genotypes and iron parameters are shown in Supplementary Table 1. The HFE C282Y genotype did not affect the risk for myocardial infarction (Table 2). Adjustment for iron parameters did not change the outcome. For HFE H63D homozygotes, a 5-fold increase in risk was found for women only, this was unaffected by adjustments. No effects were seen in the full study group or among men, although adjustment for iron parameters revealed a weak risk association in the full study group (Table 2). Quartiles of TIBC and ferritin were not associated with risk for myocardial infarction in univariate analysis or after adjusting for other risk factors (Supplementary Table 2). Plasma iron in the highest quartile was associated with a decrease in risk for myocardial infarction among males in univariate analysis (Table 3). In the full study group there was a tendency of decrease in risk in the highest quartile. No associations were seen among women. Likewise, transferrin iron saturation in the highest quartile was associated with a decrease in risk in the full study group and among men. Adjustment for age, BMI, systolic blood pressure, smoking, diabetes and ApoB/ApoA1 ratio did not change the results. As plasma iron and transferrin saturation levels are known to be influenced by inflammation and alcohol intake, both of which differed between cases and referents, we analyzed the correlations between high-sensitive C-reactive protein (hsCRP), alcohol intake and the iron parameters (Supplementary Table 3). HsCRP showed significant negative correlations with plasma iron, TIBC and transferrin iron
171
saturation. Alcohol intake showed significant positive correlations with plasma iron and ferritin. In multivariate analysis for risk of myocardial infarction, significances for iron and transferrin iron saturation were lost when adjusting for hsCRP (Table 3). Since data on alcohol intake was available only for a subset, the statistical power did not allow reliable adjustment for this parameter. 4. Discussion To our knowledge, this is the first study on a large cohort of well defined first-ever myocardial infarction cases on the risk associations of plasma iron parameters and the clinically important HFE genotypes, C282Y and H63D. We found an inverse risk association with myocardial infarction in the highest plasma iron quartile. This is consistent with the Helsinki Heart Study on dyslipidemic men [4]. It is also consistent with the findings of the NHANES I follow up study and a Finnish cohort study on elderly subjects [5,6], although our findings were significant for men only. Our findings on previously healthy subjects are contradictory to the findings of a Canadian cohort study which included fatal myocardial infarction only [7]. Consistent with our previous finding on stroke [19], we found a lower risk in the highest transferrin iron saturation quartile among men. The lack of risk association between ferritin and myocardial infarction is consistent with most other studies on cardiovascular disease [20,21] and on myocardial infarction [13]. Considering the pro-oxidant properties of iron, our findings were unexpected. As inflammation has an impact on iron metabolism [22], we adjusted our results for hsCRP. This adjustment eliminated the association between high levels of iron and risk for myocardial infarction. This is remarkable considering the low grade of inflammation in the subjects mirrored by low levels of CRP in cases and referents. Another potential confounder is use of alcohol. Alcohol tends to increase iron levels, possibly via down-regulation of hepcidin secretion in the liver [23]. Alcohol intake was higher in referents compared to cases (Table 1), and was correlated with plasma iron levels (Supplementary Table 3). Our data highlight the importance of adjustment for inflammation and alcohol intake when studying impact of iron metabolism on disease. Inflammation has also previously been used as an explanation why low serum iron is associated with higher cardiovascular risk [4]. An interesting finding in this study is that HFE H63D homozygosity was associated with higher risk for myocardial infarction in women. With respect to the plasma markers of iron stores, this effect is probably not due to the iron metabolism, as this polymorphism is a weak determinant for iron levels [15]. The polymorphism may be linked to another locus influencing the risk for cardiovascular disease. H63D homozygosity has previously been reported to be associated with increased self-reported heart disease in women [15]. In a metaanalysis, this homozygosity was associated with borderline increase in risk for heart disease [24]. This study had many advantages compared with previous studies. It had prospective design, and a large number of well defined cases of both sexes. The population was homogenous regarding ethnicity. We were also able to adjust for low-grade inflammation as a confounder. A unique feature was the ability to examine both plasma and DNA samples simultaneously. This study also had some disadvantages. The number of HFE C282Y homozygotes was very low, as were subjects with extremely high iron stores. Therefore we cannot make firm conclusion about the risk of myocardial infarction in such subjects. However, no increase in risk was noted in upper normal range. In summary, iron levels, at least in the upper normal range, were not associated with higher risk for first-time myocardial infarction in this prospective study on Caucasians: upper normal iron levels seemed to be associated with lower risk, but this association was no longer evident after adjusting for inflammation. In some countries iron fortification has been discontinued due to fear of elevated risk for cardiovascular events [25], resulting in increased occurrence of iron
172
K. Ekblom et al. / International Journal of Cardiology 150 (2011) 169–172
deficiency [26]. Our results do not support the hypothesis of increased risk of heart disease associated with iron fortification. Acknowledgements We thank Åsa Ågren at the Medical Biobank and Birgitta Nilsson at the Clinical Chemistry Laboratory at Umeå University Hospital for excellent technical assistance. We also acknowledge Bethany Van Guelpen at the Department of Biomedical Sciences, Pathology for assistance with dietary data on alcohol intake. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [27]. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ijcard.2010.04.001. References [1] Steinberg D. Low density lipoprotein oxidation and its pathobiological significance. J Biol Chem 1997;272:20963–6. [2] Girotti AW. Mechanisms of lipid peroxidation. J Free Radic Biol Med 1985;1:87–95. [3] Sullivan JL. Iron and the sex difference in heart disease risk. Lancet 1981;1:1293–4. [4] Kervinen H, Tenkanen L, Palosuo T, Roivainen M, Manninen V, Manttari M. Serum iron, infection and inflammation; effects on coronary risk. Scand Cardiovasc J 2004;38:345–8. [5] Marniemi J, Alanen E, Impivaara O, et al. Dietary and serum vitamins and minerals as predictors of myocardial infarction and stroke in elderly subjects. Nutr Metab Cardiovasc Dis 2005;15:188–97. [6] Liao Y, Cooper RS, McGee DL. Iron status and coronary heart disease: negative findings from the NHANES I epidemiologic follow-up study. Am J Epidemiol 1994;139:704–12. [7] Morrison HI, Semenciw RM, Mao Y, Wigle DT. Serum iron and risk of fatal acute myocardial infarction. Epidemiology 1994;5:243–6. [8] Baer DM, Tekawa IS, Hurley LB. Iron stores are not associated with acute myocardial infarction. Circulation 1994;89:2915–8. [9] Sempos CT, Looker AC, Gillum RF, Makuc DM. Body iron stores and the risk of coronary heart disease. N Engl J Med 1994;330:1119–24.
[10] Magnusson MK, Sigfusson N, Sigvaldason H, Johannesson GM, Magnusson S, Thorgeirsson G. Low iron-binding capacity as a risk factor for myocardial infarction. Circulation 1994;89:102–8. [11] Salonen JT, Nyyssonen K, Salonen R. Body iron stores and the risk of coronary heart disease. N Engl J Med 1994;331:1159 author reply 60. [12] Salonen JT, Nyyssonen K, Korpela H, Tuomilehto J, Seppanen R, Salonen R. High stored iron levels are associated with excess risk of myocardial infarction in eastern Finnish men. Circulation 1992;86:803–11. [13] Frey GH, Krider DW. Serum ferritin and myocardial infarct. W V Med J 1994;90: 13–5. [14] Tuomainen TP, Punnonen K, Nyyssonen K, Salonen JT. Association between body iron stores and the risk of acute myocardial infarction in men. Circulation 1998;97:1461–6. [15] Adams PC, Reboussin DM, Barton JC, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352:1769–78. [16] Ellervik C, Tybjaerg-Hansen A, Grande P, Appleyard M, Nordestgaard BG. Hereditary hemochromatosis and risk of ischemic heart disease: a prospective study and a case–control study. Circulation 2005;112:185–93. [17] Roest M, van der Schouw YT, de Valk B, et al. Heterozygosity for a hereditary hemochromatosis gene is associated with cardiovascular death in women. Circulation 1999;100:1268–73. [18] Hallmans G, Agren A, Johansson G, et al. Cardiovascular disease and diabetes in the Northern Sweden Health and Disease Study Cohort — evaluation of risk factors and their interactions. Scand J Public Health Suppl 2003;61:18–24. [19] Ekblom K, Hultdin J, Stegmayr B, et al. Iron stores and HFE genotypes are not related to increased risk of ischemic stroke. A prospective nested case-referent study. Cerebrovasc Dis 2007;24:405–11. [20] Aronow WS, Ahn C. Three-year follow-up shows no association of serum ferritin levels with incidence of new coronary events in 577 persons aged N or = 62 years. Am J Cardiol 1996;78:678–9. [21] Manttari M, Manninen V, Huttunen JK, et al. Serum ferritin and ceruloplasmin as coronary risk factors. Eur Heart J 1994;15:1599–603. [22] Ganz T, Nemeth E. Iron sequestration and anemia of inflammation. Semin Hematol 2009;46:387–93. [23] Harrison-Findik DD. Role of alcohol in the regulation of iron metabolism. World J Gastroenterol 2007;13:4925–30. [24] Ellervik C, Birgens H, Tybjaerg-Hansen A, Nordestgaard BG. Hemochromatosis genotypes and risk of 31 disease endpoints: meta-analyses including 66,000 cases and 226,000 controls. Hepatology 2007;46:1071–80. [25] Lynch SR. The impact of iron fortification on nutritional anaemia. Best Pract Res Clin Haematol 2005;18:333–46. [26] Hallberg L, Hulthen L. Perspectives on iron absorption. Blood Cells Mol Dis 2002;29:562–73. [27] Coats AJ. Ethical authorship and publishing. Int J Cardiol 2009;131:149–50.