Hereditary Hemochromatosis and Risk of Ischemic Heart Disease
http://www.100md.com
《循环学杂志》
the Department of Clinical Biochemistry, Herlev University Hospital (C.E., B.G.N.)
Department of Clinical Biochemistry (A.T.-H.)
Department of Cardiology (P.G.), Rigshospitalet, Copenhagen University Hospital
The Copenhagen City Heart Study, Bispebjerg University Hospital (A.T.-H., M.A., B.G.N.)
University of Copenhagen, Copenhagen, Denmark.
Abstract
Background— We tested the hypothesis that the hereditary hemochromatosis genotypes C282Y/C282Y, C282Y/H63D, or C282Y/wild-type are risk factors for ischemic heart disease (IHD) and myocardial infarction (MI).
Methods and Results— We performed a prospective study of 9178 individuals from the Danish general population followed up for 24 years, during which 1035 and 511 developed IHD and MI, respectively, and a case-control study of 2441 and 1113 IHD and MI cases versus 8080 controls. C282Y/C282Y, C282Y/H63D, and C282Y/wild-type versus wild-type/wild-type individuals were not associated with increased risk of IHD or MI in prospective studies, overall or stratified by gender. We had 90% power to detect a hazard ratio for IHD of 3.4 for C282Y/C282Y, 1.9 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. Furthermore, these genotypes were not associated with increased risk of IHD or MI in case-control studies, overall or stratified by gender. We had 90% power to detect an odds ratio for IHD of 3.6 for C282Y/C282Y, 1.8 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type.
Conclusions— In these studies, hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacked the power to exclude the possibility that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Key Words: cardiovascular diseases ; prospective studies ; case-control studies ; hemochromatosis
Introduction
The role of increased iron levels as a risk factor for ischemic heart disease (IHD) remains controversial.1,2 The hypothesis is that oxidation of iron, which generates free hydroxyl radicals, leads to an increase in oxidized LDL within the arterial wall and thus accelerated development of atherosclerosis and IHD.3,4
Hereditary hemochromatosis is a genetic disease that leads to lifelong increased iron accumulation throughout the body.3 Therefore, hereditary hemochromatosis provides a model to study the effect of iron overload on the risk of IHD.5–19 Most patients with hereditary hemochromatosis are homozygous for Cys282Tyr (C282Y/C282Y) in the HFE gene3,20; however, individuals who have compound heterozygosity for C282Y and His63Asp (C282Y/H63D) or who are heterozygous for C282Y (C282Y/wild-type) also have biochemical signs of iron overload.21,22 Consequently, the C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes found in 0.3%, 1.4%, and 9.2% of individuals of Danish descent,21,23 respectively, may all lead to oxidation of LDL within the arterial wall, accelerated atherosclerosis, and consequently increased risk of IHD, although this remains controversial.5–19
We tested the hypothesis that the C282Y/C282Y, C282Y/H63D, or C282Y/wild-type genotypes are risk factors for IHD and myocardial infarction (MI). For this purpose, we performed (1) a study of 9178 individuals from the Danish general population followed up for 24 years, during which 1035 and 511 developed IHD and MI, respectively, and (2) a case-control study of 2441 and 1113 IHD and MI cases versus 8080 controls.
Methods
Study Designs
Prospective Study
A total of 9178 participants from the Copenhagen City Heart Study were followed up prospectively from study entry through 1999. Of the 1098 participants recorded with IHD, 63 were diagnosed before study entry, which left 1035 incident IHD cases, of which 511 had MI. Median follow-up time was 22 years (range 0.04 to 24 years). Follow-up was >99.9% complete.
Case-Control Study
A total of 2441 patients with IHD from Copenhagen University Hospital (1113 of whom had MI) were compared with 8080 unmatched controls and 7829 matched controls (matched by 2-year age intervals and by gender) free of IHD from the Copenhagen City Heart Study. Controls in strata without cases were excluded from the analyses in the matched study.
Participants
The Copenhagen City Heart Study
The Copenhagen City Heart Study is a prospective cardiovascular study of individuals randomly selected on the basis of the Danish Central Population Register Code to reflect the adult general population. The participants, age stratified within 10-year age groups from 20 to 80 years, were examined in 1976 to 1978, 1981 to 1983, and 1991 to 1994.24,25 More than 99% were whites of Danish descent. All 3 examinations included a self-administered questionnaire, a physical examination, and blood samples; 9178 participants were genotyped for hereditary hemochromatosis (C282Y, H63D). Among these, none were registered with a diagnosis of hemochromatosis during follow-up.
Clinically overt hemochromatosis was not diagnosed in any of the 23 C282Y/C282Y individuals before the study. One was accidentally diagnosed with increased iron levels, which led to measurement of transferrin saturation of 80% and a ferritin level of 452 μg/L and the subsequent finding of C282Y homozygosity without clinically overt hemochromatosis. Four of the 23 C282Y/C282Y individuals had been blood donors, and 1 had given >50 U of blood. All living C282Y/C282Y individuals were informed about their genotype status, the possibility of a slightly increased risk of developing hemochromatosis, and that phlebotomy could be considered. Of the 20 C282Y/C282Y individuals still alive in 2001, all were referred to a hospital for further control and possible treatment for hemochromatosis (in 2001). The further fate of the C282Y/C282Y individuals will therefore not affect the results in the present study, because our follow-up stopped at the end of 1999.
Information on diagnoses of IHD (World Health Organization International Classification of Diseases, 8th edition, codes 410 to 414; 10th edition, codes I20 to I25) was gathered until the end of 1999 from the Danish National Hospital Discharge Register, the Danish National Register of Cause of Death, and medical records from general practitioners and hospitals. IHD was defined as a previous MI or characteristic symptoms of stable angina pectoris.26 The diagnosis of MI (World Health Organization code 410 or I21/I22) required the presence of at least 2 of the following criteria: (1) characteristic chest pain, (2) elevated cardiac enzymes, and (3) ECG changes indicative of MI. The study was approved by the Danish ethics committee for the City of Copenhagen and Frederiksberg and by Herlev University Hospital. Participants gave informed consent.
Copenhagen University Hospital
During the period 1991 through 2003, we recruited 2441 IHD patients from the greater Copenhagen area referred to Copenhagen University Hospital, Rigshospitalet, for coronary angiography. The diagnosis of IHD was obtained from experienced cardiologists based on angina pectoris plus at least 1 of the following criteria: stenosis/atherosclerosis on coronary angiography, a previous MI, or significant myocardial ischemia on a bicycle exercise test. The diagnosis of MI (n=1113) was established with the same criteria as in the Copenhagen City Heart Study. More than 99% of the subjects were whites of Danish descent. The 28 patients who also participated in the Copenhagen City Heart Study were excluded from the case-control studies. The study was approved by the Danish ethics committee for Copenhagen County and by Herlev University Hospital. Participants gave informed consent.
Analyses
We performed 2 polymerase chain reaction assays with allele-specific amplification for every individual to determine noncarrier states, heterozygosity, or homozygosity for the C282Y and H63D polymorphisms.23 To confirm the diagnosis of H63D, we retyped all those initially genotyped as H63D/wild-type, H63D/H63D, and C282Y/H63D with polymerase chain reaction followed by restriction enzyme digestion. This method was also used to confirm the diagnosis of C282Y in C282Y/C282Y and C282Y/H63D genotypes. Colorimetric and turbidimetric assays were used to measure plasma levels of total cholesterol, HDL cholesterol, triglycerides, fibrinogen (all Boehringer Mannheim), and lipoprotein(a) total mass (DAKO).
Body mass index was calculated as weight in kilograms divided by height in meters squared. Hypertension was defined as (1) systolic blood pressure 140 mm Hg,27 (2) diastolic blood pressure 90 mm Hg,27 or (3) treatment with antihypertensive drugs. Smokers were persons who currently smoked. Diabetes mellitus was self-reported disease, use of insulin, use of oral hypoglycemic drugs, and/or nonfasting plasma glucose >11 mmol/L.
Oxidized LDL, transferrin saturation, and ferritin levels were determined for each genotype group in the prospective study with blood from the 1991 to 1994 examination. In this cross-sectional design, 23 C282Y/C282Y individuals were matched for gender, age (in 10-year groups), and alcohol consumption (in gender-specific tertiles)21 with 2 subjects each from the other 5 genotypes. Because of technical error, the sample from 1 of the C282Y homozygotes was not measured, and consequently, the 2 matched persons in each of the other genotypes were also excluded. Oxidized LDL was measured with a solid-phase, 2-site enzyme immunoassay (Mercodia Oxidized LDL ELISA). Transferrin was measured on a Nephelometer Analyzer II (Behring), ferritin levels were determined on a Technicon Immuno 1 System (Bayer), and iron levels were measured on a Vitros 950 autoanalyzer (Johnson & Johnson). Transferrin saturation (%) was determined as iron levels (in μmol/L) divided by 2xtransferrin levels (in μmol/L).
Statistical Analyses
The statistical software package Stata/S.E. 8.0 was used. For power calculations, NCCS-PASS was used. Two-sided probability values <0.05 were considered significant. Categorical variables were compared with the Pearson 2 test. The Student t test was used for 2-group comparisons; plasma triglycerides, lipoprotein(a), and HDL cholesterol levels were logarithmically transformed to approach normal distribution. Because age was not normally distributed, the Mann-Whitney U test was used to compare age in 2-group comparisons. We used the Mann-Whitney U test for 2-genotype comparisons of ferritin, transferrin saturation, and oxidized LDL.
In prospective studies, plots of cumulative incidence as a function of age are presented. Differences between hemochromatosis genotypes were tested for significance with the log-rank test. Cox regression models with age as the time scale were used to estimate hazard ratios for IHD and MI; this implies that there was automatic adjustment for age. Information on duration of risk factors was included in the prospective models.
In unmatched and matched case-control designs, unconditional and conditional logistic regression models, respectively, were used to estimate ORs for IHD and MI. The matched analysis resulted in 37 2-year strata in both men and women, with an average of 3.2 control subjects per case subject. Risk factor status was based on the last examination date. Box-Tidwell transformation was used to test for linearity in the logit among continuous covariates. The following transformations were used: the inverse to the third square of body mass index, the inverse to the square of total cholesterol, the square root of HDL cholesterol, and logarithmic transformation of triglycerides.
To test for bivariate multiplicative interaction between genotype and the covariates listed in Table 1 on IHD and MI in Cox and logistic regression models, 2-factor interaction terms were included individually in the different models and tested for significance with a likelihood ratio test. No statistically significant interactions were observed. We used Bonferroni correction for multiple comparisons. For each genotype, we calculated the hazard ratio in the prospective study and the OR in the case-control study that could be detected with 90% power assuming a 2-sided P<0.05.
Results
Major cardiovascular risk factors at entry were more prevalent among participants who later developed IHD and MI than among controls with regard to the Copenhagen City Heart Study participants (Table 1). This was also the case for many risk factors in the case-control studies; however, for IHD and MI patients, total cholesterol and smoking frequency were lower than among controls. This is most likely explained by cholesterol-lowering therapy and smoking cessation in IHD and MI patients after diagnosis.
Prospective Study
Iron Status and Oxidized LDL
Ferritin levels were increased in C282Y/C282Y versus wild-type/wild-type genotypes overall and when stratified according to gender (Table 2). However, there was no difference between C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type genotypes, which indicates that individuals with these genotypes did not show evidence of iron overload.
Compared with the wild-type/wild-type genotype, transferrin saturation levels were increased in C282Y/C282Y overall and in men and women separately, in C282Y/H63D overall and in men and women separately, and in C282Y/wild-type overall (Table 2). H63D/H63D versus wild-type/wild-type showed evidence of increased transferrin saturation levels overall and in men, whereas this was not the case for H63D/wild-type. Only the C282Y/C282Y genotype had average transferrin saturation levels >50%. There were no differences in the levels of oxidized LDL between any hemochromatosis genotype and wild-type/wild-type, overall or stratified by gender (Table 2).
Ischemic Heart Disease
Cumulative incidence of IHD as a function of age for hemochromatosis genotypes versus wild-type/wild-type did not differ (Figure). Crude and adjusted hazard ratios for IHD in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 3). There was 90% power to detect a hazard ratio for IHD of 3.4 for C282Y/C282Y, 1.9 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. After stratification by gender, crude and adjusted hazard ratios for IHD in men were also nonsignificant. In women, crude and adjusted hazard ratios for IHD for H63D/wild-type compared with wild-type/wild-type were 1.3 (95% CI 1.0 to 1.6; P=0.03) and 1.3 (95% CI 1.1 to 1.7; P=0.02), respectively. When corrected for multiple comparisons, these hazard ratios were no longer significant.
Cumulative incidence of IHD and MI as function of age for hemochromatosis genotypes vs wild-type/wild-type in general population.
Myocardial Infarction
Cumulative incidence of MI as a function of age for hemochromatosis genotypes versus wild-type/wild-type did not differ (Figure). Crude and adjusted hazard ratios for MI in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant, overall as well as after stratification by gender (Table 4). There was 90% power to detect a hazard ratio for MI of 4.7 for C282Y/C282Y, 2.3 for C282Y/H63D, and 1.5 for C282Y/wild-type versus wild-type/wild-type.
Case-Control Study
Ischemic Heart Disease
In unmatched and matched analyses, crude and adjusted ORs for IHD in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 5). There was 90% power to detect an OR for IHD of 3.6 for C282Y/C282Y, 1.8 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. In the unmatched analysis, men with H63D/wild-type versus wild-type/wild-type had an adjusted OR of 1.2 (1.0 to 1.5; P=0.04), which became nonsignificant after correction for multiple comparisons.
Myocardial Infarction
In unmatched and matched analyses, crude and adjusted ORs for MI in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 6). There was 90% power to detect an OR for MI of 4.5 for C282Y/C282Y, 2.1 for C282Y/H63D, and 1.4 for C282Y/wild-type versus wild-type/wild-type. In the unmatched analysis, men with H63D/wild-type had an adjusted OR of 1.3 (1.1 to 1.7; P=0.02), which became nonsignificant after correction for multiple comparisons.
Discussion
Risk of IHD and MI for individuals with the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes was assessed by a prospective study of the general population and a case-control study. In these studies, the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacks the power to exclude that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Association of the C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes with IHD and MI could be biologically plausible: these genetic variants are associated with iron overload,3,20–22 which could lead to increased oxidation of LDL within the arterial wall and thus accelerated development of atherosclerosis and IHD.3,4 In the present study, however, although C282Y/C282Y individuals had increased ferritin levels and C282Y/C282Y, C282Y/H63D, and C282Y/wild-type individuals had increased transferrin saturation levels, none of these genotypes were associated with elevated levels of oxidized LDL in plasma.
Association between C282Y/C282Y and risk of IHD or MI has not been found in any studies to date, including the present study. Association between C282Y/wild-type and IHD and MI was observed in 2 previous studies, in men14 and in both genders,12 whereas another study reported an association between total cardiovascular death and the C282Y/wild-type genotype in women.13
Associations between H63D/wild-type genotype and IHD or MI do not seem biologically plausible, because this genotype is not associated with iron overload,3,20–22 and such an association has never been observed. It was surprising, therefore, that we observed an association between the H63D/wild-type genotype and IHD among women in the prospective study and between the H63D/wild-type genotype and IHD and MI among men in the unmatched case-control study. After correction for multiple comparisons by the Bonferroni method, each of these findings became nonsignificant; however, if a less stringent correction for multiple comparison were used, then some or all of these findings would remain significant. Therefore, these results either could be due to chance alone or alternatively, an association between the H63D/wild-type genotype and risk of IHD and MI may represent a true phenomenon due to an unidentified biological mechanism. In the present study with 2 different study designs, the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacked the power to exclude that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Acknowledgments
The Danish Heart Foundation and Chief Physician Johan Boserup’s and Lise Boserup’s Fund supported the study. We thank Nina Dahl Kjersgaard and Jesper Schou for technical assistance.
References
Danesh J, Appleby P. Coronary heart disease and iron status: meta-analyses of prospective studies. Circulation. 1999; 99: 852–854.
Sullivan JL, Zacharski LR. Hereditary haemochromatosis and the hypothesis that iron depletion protects against ischemic heart disease. Eur J Clin Invest. 2001; 31: 375–377.
Beutler E, Bothwell TH, Charlton RW, Motulsky AG. Hereditary hemochromatosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease. New York, NY: McGraw-Hill; 2001: 3127–3162.
de Valk B, Marx JJM. Iron, atherosclerosis, and ischemic heart disease. Arch Intern Med. 1999; 159: 1542–1548.
Annichino-Bizzacchi JM, Saad STO, Arruda VR, Ramires JAF, Siqueira LH, Chiaparini LC, Mansur AP. C282Y mutation in the HLA-H gene is not a risk factor for patients with myocardial infarction. J Cardiovasc Risk. 2000; 7: 37–40.
Battiloro E, Ombres D, Pascale E, D’Ambrosio E, Verna R, Arca M. Haemochromatosis gene mutations and risk of coronary artery disease. Eur J Hum Genet. 2000; 8: 389–392.
Candore G, Balistreri CR, Lio D, Mantovani V, Colonna-Romano G, Chiappelli M, Tampieri C, Licastro F, Branzi A, Averna M, Caruso M, Hoffman E, Caruso C. Association between HFE mutations and acute myocardial infarction: a study in patients from Northern and Southern Italy. Blood Cells Mol Dis. 2003; 31: 57–62.
Claeys D, Walting M, Julmy F, Wuillemin WA, Meyer BJ. Haemochromatosis mutations and ferritin in myocardial infarction: a case-control study. Eur J Clin Invest. 2002; 32: 3–8.
Fox CJ, Cullen DJ, Knuiman MW, Cumpston GN, Divitini ML, Rossi E, Gochee PA, Powell LW, Olynyk JK. Effects of body iron stores and haemochromatosis genotypes on coronary heart disease outcomes in the Busselton Health Study. J Cardiovasc Risk. 2002; 9: 287–293.
Franco RF, Zago MA, Trip MD, ten Cate H, van den Ende A, Prins MH, Kastelein JJP, Reitsma PH. Prevalence of hereditary haemochromatosis in premature atherosclerotic vascular disease. Br J Haematol. 1998; 102: 1172–1175.
Hetet G, Elbaz A, Gariepy J, Nicaud V, Arveiler D, Morrison C, Kee F, Evans A, Simon A, Amarenco P, Cambien F, Grandchamp B. Association studies between haemochromatosis gene mutations and the risk of cardiovascular diseases. Eur J Clin Invest. 2001; 31: 382–388.
Rasmussen ML, Folsom AR, Catellier DJ, Tsai MY, Garg U, Eckfeldt JH. A prospective study of coronary heart disease and the hemochromatosis gene (HFE) C282Y mutation: the Atherosclerosis Risk in Communities (ARIC) study. Atherosclerosis. 2001; 154: 739–746.
Roest M, van der Schouw YT, de Valk B, Marx JJM, Tempelman MJ, de Groot PG, Sixma JJ, Banga JD. Heterozygosity for a hereditary hemochromatosis gene is associated with cardiovascular death in women. Circulation. 1999; 100: 1268–1273.
Tuomainen TP, Kontula K, Nyyssonen K, Lakka TA, Helio T, Salonen JT. Increased risk of acute myocardial infarction in carriers of the hemochromatosis gene Cys282Tyr mutation: a prospective cohort study in men in eastern Finland. Circulation. 1999; 100: 1274–1279.
van Aken MO, de Craen AJM, Gussekloo J, Moghaddam PH, Vandenbroucke JP, Heijmans BT, Slagboom PE, Westendorp RGJ. No increase in mortality and morbidity among carriers of the C282Y mutation of the hereditary haemochromatosis gene in the oldest old: the Leiden 85-plus Study. Eur J Clin Invest. 2002; 32: 750–754.
Waalen J, Felitti V, Gelbart T, Ho NJ, Beutler E. Prevalence of coronary heart disease associated with HFE mutations in adults attending a health appraisal center. Am J Med. 2002; 113: 472–479.
Gunn IR, Maxwell FK, Gaffney D, McMahon AD, Packard CJ. Haemochromatosis gene mutations and risk of coronary heart disease: a West Of Scotland Coronary Prevention Study (WOSCOPS) substudy. Heart. 2004; 90: 304–306.
Campbell S, George DK, Robb SD, Spooner R, McDonagh TA, Dargie HJ, Mills PR. The prevalence of haemochromatosis gene mutations in the West of Scotland and their relation to ischaemic heart disease. Heart. 2003; 89: 1023–1026.
Calado RT, Franco RF, Pazin-Filho A, Simoes MV, Marin-Neto JA, Zago MA. HFE gene mutations in coronary atherothrombotic disease. Braz J Med Biol Res. 2000; 33: 301–306.
Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, Dormishian F, Domingo R, Ellis MC, Fullan A, Hinton LM, Jones NL, Kimmel BE, Kronmal GS, Lauer P, Lee VK, Loeb DB, Mapa FA, McClelland E, Meyer NC, Mintier GA, Moeller N, Moore T, Morikang E, Prass CE, Quintana L, Starnes SM, Schatzman RC, Brunke KJ, Drayna DT, Risch NJ, Bacon BR. A novel MHC class I–like gene is mutated in patients with hereditary haemochromatosis. Nat Genet. 1996; 13: 399–408.
Andersen RV, Tybjaerg-Hansen A, Appleyard M, Birgens H, Nordestgaard BG. Hemochromatosis mutations in the general population: iron overload progression rate. Blood. 2004; 103: 2914–2919.
Beutler E, Felitti V, Gelbart T, Ho N. The effect of HFE genotypes on measurements of iron overload in patients attending a health appraisal clinic. Ann Intern Med. 2000; 133: 329–337.
Ellervik C, Mandrup-Poulsen T, Nordestgaard BG, Larsen LE, Appleyard M, Frandsen M, Petersen P, Schlichting P, Saermark T, Tybjaerg-Hansen A, Birgens H. Prevalence of hereditary haemochromatosis in late-onset type 1 diabetes mellitus: a retrospective study. Lancet. 2001; 358: 1405–1409.
The Copenhagen City Heart Study Group. The Copenhagen City Heart Study: ;sterbrounders;gelsen: a book of tables with data from the first examination (1976–78) and a five year follow-up (1981–83). Scand J Soc Med Suppl. 1989; 41: 1–160.
Schnohr P, Jensen G, Lange P, Scharling H, Appleyard M. The Copenhagen City Heart Study: ;sterbrounders;gelsen: tables with data from the third examination 1991–1994. Eur Heart J. 2001; 3: 1–83.
Management of stable angina pectoris: Recommendations of the Task Force of the European Society of Cardiology. Eur Heart J. 1997; 18: 394–413.
Guidelines Subcommittee. 1999 World Health Organization–International Society of Hypertension guidelines for the management of hypertension. J Hypertens. 1999; 17: 151–183.(Christina Ellervik, MD; A)
Department of Clinical Biochemistry (A.T.-H.)
Department of Cardiology (P.G.), Rigshospitalet, Copenhagen University Hospital
The Copenhagen City Heart Study, Bispebjerg University Hospital (A.T.-H., M.A., B.G.N.)
University of Copenhagen, Copenhagen, Denmark.
Abstract
Background— We tested the hypothesis that the hereditary hemochromatosis genotypes C282Y/C282Y, C282Y/H63D, or C282Y/wild-type are risk factors for ischemic heart disease (IHD) and myocardial infarction (MI).
Methods and Results— We performed a prospective study of 9178 individuals from the Danish general population followed up for 24 years, during which 1035 and 511 developed IHD and MI, respectively, and a case-control study of 2441 and 1113 IHD and MI cases versus 8080 controls. C282Y/C282Y, C282Y/H63D, and C282Y/wild-type versus wild-type/wild-type individuals were not associated with increased risk of IHD or MI in prospective studies, overall or stratified by gender. We had 90% power to detect a hazard ratio for IHD of 3.4 for C282Y/C282Y, 1.9 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. Furthermore, these genotypes were not associated with increased risk of IHD or MI in case-control studies, overall or stratified by gender. We had 90% power to detect an odds ratio for IHD of 3.6 for C282Y/C282Y, 1.8 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type.
Conclusions— In these studies, hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacked the power to exclude the possibility that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Key Words: cardiovascular diseases ; prospective studies ; case-control studies ; hemochromatosis
Introduction
The role of increased iron levels as a risk factor for ischemic heart disease (IHD) remains controversial.1,2 The hypothesis is that oxidation of iron, which generates free hydroxyl radicals, leads to an increase in oxidized LDL within the arterial wall and thus accelerated development of atherosclerosis and IHD.3,4
Hereditary hemochromatosis is a genetic disease that leads to lifelong increased iron accumulation throughout the body.3 Therefore, hereditary hemochromatosis provides a model to study the effect of iron overload on the risk of IHD.5–19 Most patients with hereditary hemochromatosis are homozygous for Cys282Tyr (C282Y/C282Y) in the HFE gene3,20; however, individuals who have compound heterozygosity for C282Y and His63Asp (C282Y/H63D) or who are heterozygous for C282Y (C282Y/wild-type) also have biochemical signs of iron overload.21,22 Consequently, the C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes found in 0.3%, 1.4%, and 9.2% of individuals of Danish descent,21,23 respectively, may all lead to oxidation of LDL within the arterial wall, accelerated atherosclerosis, and consequently increased risk of IHD, although this remains controversial.5–19
We tested the hypothesis that the C282Y/C282Y, C282Y/H63D, or C282Y/wild-type genotypes are risk factors for IHD and myocardial infarction (MI). For this purpose, we performed (1) a study of 9178 individuals from the Danish general population followed up for 24 years, during which 1035 and 511 developed IHD and MI, respectively, and (2) a case-control study of 2441 and 1113 IHD and MI cases versus 8080 controls.
Methods
Study Designs
Prospective Study
A total of 9178 participants from the Copenhagen City Heart Study were followed up prospectively from study entry through 1999. Of the 1098 participants recorded with IHD, 63 were diagnosed before study entry, which left 1035 incident IHD cases, of which 511 had MI. Median follow-up time was 22 years (range 0.04 to 24 years). Follow-up was >99.9% complete.
Case-Control Study
A total of 2441 patients with IHD from Copenhagen University Hospital (1113 of whom had MI) were compared with 8080 unmatched controls and 7829 matched controls (matched by 2-year age intervals and by gender) free of IHD from the Copenhagen City Heart Study. Controls in strata without cases were excluded from the analyses in the matched study.
Participants
The Copenhagen City Heart Study
The Copenhagen City Heart Study is a prospective cardiovascular study of individuals randomly selected on the basis of the Danish Central Population Register Code to reflect the adult general population. The participants, age stratified within 10-year age groups from 20 to 80 years, were examined in 1976 to 1978, 1981 to 1983, and 1991 to 1994.24,25 More than 99% were whites of Danish descent. All 3 examinations included a self-administered questionnaire, a physical examination, and blood samples; 9178 participants were genotyped for hereditary hemochromatosis (C282Y, H63D). Among these, none were registered with a diagnosis of hemochromatosis during follow-up.
Clinically overt hemochromatosis was not diagnosed in any of the 23 C282Y/C282Y individuals before the study. One was accidentally diagnosed with increased iron levels, which led to measurement of transferrin saturation of 80% and a ferritin level of 452 μg/L and the subsequent finding of C282Y homozygosity without clinically overt hemochromatosis. Four of the 23 C282Y/C282Y individuals had been blood donors, and 1 had given >50 U of blood. All living C282Y/C282Y individuals were informed about their genotype status, the possibility of a slightly increased risk of developing hemochromatosis, and that phlebotomy could be considered. Of the 20 C282Y/C282Y individuals still alive in 2001, all were referred to a hospital for further control and possible treatment for hemochromatosis (in 2001). The further fate of the C282Y/C282Y individuals will therefore not affect the results in the present study, because our follow-up stopped at the end of 1999.
Information on diagnoses of IHD (World Health Organization International Classification of Diseases, 8th edition, codes 410 to 414; 10th edition, codes I20 to I25) was gathered until the end of 1999 from the Danish National Hospital Discharge Register, the Danish National Register of Cause of Death, and medical records from general practitioners and hospitals. IHD was defined as a previous MI or characteristic symptoms of stable angina pectoris.26 The diagnosis of MI (World Health Organization code 410 or I21/I22) required the presence of at least 2 of the following criteria: (1) characteristic chest pain, (2) elevated cardiac enzymes, and (3) ECG changes indicative of MI. The study was approved by the Danish ethics committee for the City of Copenhagen and Frederiksberg and by Herlev University Hospital. Participants gave informed consent.
Copenhagen University Hospital
During the period 1991 through 2003, we recruited 2441 IHD patients from the greater Copenhagen area referred to Copenhagen University Hospital, Rigshospitalet, for coronary angiography. The diagnosis of IHD was obtained from experienced cardiologists based on angina pectoris plus at least 1 of the following criteria: stenosis/atherosclerosis on coronary angiography, a previous MI, or significant myocardial ischemia on a bicycle exercise test. The diagnosis of MI (n=1113) was established with the same criteria as in the Copenhagen City Heart Study. More than 99% of the subjects were whites of Danish descent. The 28 patients who also participated in the Copenhagen City Heart Study were excluded from the case-control studies. The study was approved by the Danish ethics committee for Copenhagen County and by Herlev University Hospital. Participants gave informed consent.
Analyses
We performed 2 polymerase chain reaction assays with allele-specific amplification for every individual to determine noncarrier states, heterozygosity, or homozygosity for the C282Y and H63D polymorphisms.23 To confirm the diagnosis of H63D, we retyped all those initially genotyped as H63D/wild-type, H63D/H63D, and C282Y/H63D with polymerase chain reaction followed by restriction enzyme digestion. This method was also used to confirm the diagnosis of C282Y in C282Y/C282Y and C282Y/H63D genotypes. Colorimetric and turbidimetric assays were used to measure plasma levels of total cholesterol, HDL cholesterol, triglycerides, fibrinogen (all Boehringer Mannheim), and lipoprotein(a) total mass (DAKO).
Body mass index was calculated as weight in kilograms divided by height in meters squared. Hypertension was defined as (1) systolic blood pressure 140 mm Hg,27 (2) diastolic blood pressure 90 mm Hg,27 or (3) treatment with antihypertensive drugs. Smokers were persons who currently smoked. Diabetes mellitus was self-reported disease, use of insulin, use of oral hypoglycemic drugs, and/or nonfasting plasma glucose >11 mmol/L.
Oxidized LDL, transferrin saturation, and ferritin levels were determined for each genotype group in the prospective study with blood from the 1991 to 1994 examination. In this cross-sectional design, 23 C282Y/C282Y individuals were matched for gender, age (in 10-year groups), and alcohol consumption (in gender-specific tertiles)21 with 2 subjects each from the other 5 genotypes. Because of technical error, the sample from 1 of the C282Y homozygotes was not measured, and consequently, the 2 matched persons in each of the other genotypes were also excluded. Oxidized LDL was measured with a solid-phase, 2-site enzyme immunoassay (Mercodia Oxidized LDL ELISA). Transferrin was measured on a Nephelometer Analyzer II (Behring), ferritin levels were determined on a Technicon Immuno 1 System (Bayer), and iron levels were measured on a Vitros 950 autoanalyzer (Johnson & Johnson). Transferrin saturation (%) was determined as iron levels (in μmol/L) divided by 2xtransferrin levels (in μmol/L).
Statistical Analyses
The statistical software package Stata/S.E. 8.0 was used. For power calculations, NCCS-PASS was used. Two-sided probability values <0.05 were considered significant. Categorical variables were compared with the Pearson 2 test. The Student t test was used for 2-group comparisons; plasma triglycerides, lipoprotein(a), and HDL cholesterol levels were logarithmically transformed to approach normal distribution. Because age was not normally distributed, the Mann-Whitney U test was used to compare age in 2-group comparisons. We used the Mann-Whitney U test for 2-genotype comparisons of ferritin, transferrin saturation, and oxidized LDL.
In prospective studies, plots of cumulative incidence as a function of age are presented. Differences between hemochromatosis genotypes were tested for significance with the log-rank test. Cox regression models with age as the time scale were used to estimate hazard ratios for IHD and MI; this implies that there was automatic adjustment for age. Information on duration of risk factors was included in the prospective models.
In unmatched and matched case-control designs, unconditional and conditional logistic regression models, respectively, were used to estimate ORs for IHD and MI. The matched analysis resulted in 37 2-year strata in both men and women, with an average of 3.2 control subjects per case subject. Risk factor status was based on the last examination date. Box-Tidwell transformation was used to test for linearity in the logit among continuous covariates. The following transformations were used: the inverse to the third square of body mass index, the inverse to the square of total cholesterol, the square root of HDL cholesterol, and logarithmic transformation of triglycerides.
To test for bivariate multiplicative interaction between genotype and the covariates listed in Table 1 on IHD and MI in Cox and logistic regression models, 2-factor interaction terms were included individually in the different models and tested for significance with a likelihood ratio test. No statistically significant interactions were observed. We used Bonferroni correction for multiple comparisons. For each genotype, we calculated the hazard ratio in the prospective study and the OR in the case-control study that could be detected with 90% power assuming a 2-sided P<0.05.
Results
Major cardiovascular risk factors at entry were more prevalent among participants who later developed IHD and MI than among controls with regard to the Copenhagen City Heart Study participants (Table 1). This was also the case for many risk factors in the case-control studies; however, for IHD and MI patients, total cholesterol and smoking frequency were lower than among controls. This is most likely explained by cholesterol-lowering therapy and smoking cessation in IHD and MI patients after diagnosis.
Prospective Study
Iron Status and Oxidized LDL
Ferritin levels were increased in C282Y/C282Y versus wild-type/wild-type genotypes overall and when stratified according to gender (Table 2). However, there was no difference between C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type genotypes, which indicates that individuals with these genotypes did not show evidence of iron overload.
Compared with the wild-type/wild-type genotype, transferrin saturation levels were increased in C282Y/C282Y overall and in men and women separately, in C282Y/H63D overall and in men and women separately, and in C282Y/wild-type overall (Table 2). H63D/H63D versus wild-type/wild-type showed evidence of increased transferrin saturation levels overall and in men, whereas this was not the case for H63D/wild-type. Only the C282Y/C282Y genotype had average transferrin saturation levels >50%. There were no differences in the levels of oxidized LDL between any hemochromatosis genotype and wild-type/wild-type, overall or stratified by gender (Table 2).
Ischemic Heart Disease
Cumulative incidence of IHD as a function of age for hemochromatosis genotypes versus wild-type/wild-type did not differ (Figure). Crude and adjusted hazard ratios for IHD in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 3). There was 90% power to detect a hazard ratio for IHD of 3.4 for C282Y/C282Y, 1.9 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. After stratification by gender, crude and adjusted hazard ratios for IHD in men were also nonsignificant. In women, crude and adjusted hazard ratios for IHD for H63D/wild-type compared with wild-type/wild-type were 1.3 (95% CI 1.0 to 1.6; P=0.03) and 1.3 (95% CI 1.1 to 1.7; P=0.02), respectively. When corrected for multiple comparisons, these hazard ratios were no longer significant.
Cumulative incidence of IHD and MI as function of age for hemochromatosis genotypes vs wild-type/wild-type in general population.
Myocardial Infarction
Cumulative incidence of MI as a function of age for hemochromatosis genotypes versus wild-type/wild-type did not differ (Figure). Crude and adjusted hazard ratios for MI in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant, overall as well as after stratification by gender (Table 4). There was 90% power to detect a hazard ratio for MI of 4.7 for C282Y/C282Y, 2.3 for C282Y/H63D, and 1.5 for C282Y/wild-type versus wild-type/wild-type.
Case-Control Study
Ischemic Heart Disease
In unmatched and matched analyses, crude and adjusted ORs for IHD in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 5). There was 90% power to detect an OR for IHD of 3.6 for C282Y/C282Y, 1.8 for C282Y/H63D, and 1.3 for C282Y/wild-type versus wild-type/wild-type. In the unmatched analysis, men with H63D/wild-type versus wild-type/wild-type had an adjusted OR of 1.2 (1.0 to 1.5; P=0.04), which became nonsignificant after correction for multiple comparisons.
Myocardial Infarction
In unmatched and matched analyses, crude and adjusted ORs for MI in C282Y/C282Y, C282Y/H63D, C282Y/wild-type, H63D/H63D, and H63D/wild-type versus wild-type/wild-type individuals were all nonsignificant (Table 6). There was 90% power to detect an OR for MI of 4.5 for C282Y/C282Y, 2.1 for C282Y/H63D, and 1.4 for C282Y/wild-type versus wild-type/wild-type. In the unmatched analysis, men with H63D/wild-type had an adjusted OR of 1.3 (1.1 to 1.7; P=0.02), which became nonsignificant after correction for multiple comparisons.
Discussion
Risk of IHD and MI for individuals with the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes was assessed by a prospective study of the general population and a case-control study. In these studies, the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacks the power to exclude that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Association of the C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes with IHD and MI could be biologically plausible: these genetic variants are associated with iron overload,3,20–22 which could lead to increased oxidation of LDL within the arterial wall and thus accelerated development of atherosclerosis and IHD.3,4 In the present study, however, although C282Y/C282Y individuals had increased ferritin levels and C282Y/C282Y, C282Y/H63D, and C282Y/wild-type individuals had increased transferrin saturation levels, none of these genotypes were associated with elevated levels of oxidized LDL in plasma.
Association between C282Y/C282Y and risk of IHD or MI has not been found in any studies to date, including the present study. Association between C282Y/wild-type and IHD and MI was observed in 2 previous studies, in men14 and in both genders,12 whereas another study reported an association between total cardiovascular death and the C282Y/wild-type genotype in women.13
Associations between H63D/wild-type genotype and IHD or MI do not seem biologically plausible, because this genotype is not associated with iron overload,3,20–22 and such an association has never been observed. It was surprising, therefore, that we observed an association between the H63D/wild-type genotype and IHD among women in the prospective study and between the H63D/wild-type genotype and IHD and MI among men in the unmatched case-control study. After correction for multiple comparisons by the Bonferroni method, each of these findings became nonsignificant; however, if a less stringent correction for multiple comparison were used, then some or all of these findings would remain significant. Therefore, these results either could be due to chance alone or alternatively, an association between the H63D/wild-type genotype and risk of IHD and MI may represent a true phenomenon due to an unidentified biological mechanism. In the present study with 2 different study designs, the hereditary hemochromatosis C282Y/C282Y, C282Y/H63D, and C282Y/wild-type genotypes were not associated with IHD or MI; however, the study lacked the power to exclude that C282Y/C282Y and C282Y/H63D individuals have a modestly increased risk of IHD or MI.
Acknowledgments
The Danish Heart Foundation and Chief Physician Johan Boserup’s and Lise Boserup’s Fund supported the study. We thank Nina Dahl Kjersgaard and Jesper Schou for technical assistance.
References
Danesh J, Appleby P. Coronary heart disease and iron status: meta-analyses of prospective studies. Circulation. 1999; 99: 852–854.
Sullivan JL, Zacharski LR. Hereditary haemochromatosis and the hypothesis that iron depletion protects against ischemic heart disease. Eur J Clin Invest. 2001; 31: 375–377.
Beutler E, Bothwell TH, Charlton RW, Motulsky AG. Hereditary hemochromatosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inherited Disease. New York, NY: McGraw-Hill; 2001: 3127–3162.
de Valk B, Marx JJM. Iron, atherosclerosis, and ischemic heart disease. Arch Intern Med. 1999; 159: 1542–1548.
Annichino-Bizzacchi JM, Saad STO, Arruda VR, Ramires JAF, Siqueira LH, Chiaparini LC, Mansur AP. C282Y mutation in the HLA-H gene is not a risk factor for patients with myocardial infarction. J Cardiovasc Risk. 2000; 7: 37–40.
Battiloro E, Ombres D, Pascale E, D’Ambrosio E, Verna R, Arca M. Haemochromatosis gene mutations and risk of coronary artery disease. Eur J Hum Genet. 2000; 8: 389–392.
Candore G, Balistreri CR, Lio D, Mantovani V, Colonna-Romano G, Chiappelli M, Tampieri C, Licastro F, Branzi A, Averna M, Caruso M, Hoffman E, Caruso C. Association between HFE mutations and acute myocardial infarction: a study in patients from Northern and Southern Italy. Blood Cells Mol Dis. 2003; 31: 57–62.
Claeys D, Walting M, Julmy F, Wuillemin WA, Meyer BJ. Haemochromatosis mutations and ferritin in myocardial infarction: a case-control study. Eur J Clin Invest. 2002; 32: 3–8.
Fox CJ, Cullen DJ, Knuiman MW, Cumpston GN, Divitini ML, Rossi E, Gochee PA, Powell LW, Olynyk JK. Effects of body iron stores and haemochromatosis genotypes on coronary heart disease outcomes in the Busselton Health Study. J Cardiovasc Risk. 2002; 9: 287–293.
Franco RF, Zago MA, Trip MD, ten Cate H, van den Ende A, Prins MH, Kastelein JJP, Reitsma PH. Prevalence of hereditary haemochromatosis in premature atherosclerotic vascular disease. Br J Haematol. 1998; 102: 1172–1175.
Hetet G, Elbaz A, Gariepy J, Nicaud V, Arveiler D, Morrison C, Kee F, Evans A, Simon A, Amarenco P, Cambien F, Grandchamp B. Association studies between haemochromatosis gene mutations and the risk of cardiovascular diseases. Eur J Clin Invest. 2001; 31: 382–388.
Rasmussen ML, Folsom AR, Catellier DJ, Tsai MY, Garg U, Eckfeldt JH. A prospective study of coronary heart disease and the hemochromatosis gene (HFE) C282Y mutation: the Atherosclerosis Risk in Communities (ARIC) study. Atherosclerosis. 2001; 154: 739–746.
Roest M, van der Schouw YT, de Valk B, Marx JJM, Tempelman MJ, de Groot PG, Sixma JJ, Banga JD. Heterozygosity for a hereditary hemochromatosis gene is associated with cardiovascular death in women. Circulation. 1999; 100: 1268–1273.
Tuomainen TP, Kontula K, Nyyssonen K, Lakka TA, Helio T, Salonen JT. Increased risk of acute myocardial infarction in carriers of the hemochromatosis gene Cys282Tyr mutation: a prospective cohort study in men in eastern Finland. Circulation. 1999; 100: 1274–1279.
van Aken MO, de Craen AJM, Gussekloo J, Moghaddam PH, Vandenbroucke JP, Heijmans BT, Slagboom PE, Westendorp RGJ. No increase in mortality and morbidity among carriers of the C282Y mutation of the hereditary haemochromatosis gene in the oldest old: the Leiden 85-plus Study. Eur J Clin Invest. 2002; 32: 750–754.
Waalen J, Felitti V, Gelbart T, Ho NJ, Beutler E. Prevalence of coronary heart disease associated with HFE mutations in adults attending a health appraisal center. Am J Med. 2002; 113: 472–479.
Gunn IR, Maxwell FK, Gaffney D, McMahon AD, Packard CJ. Haemochromatosis gene mutations and risk of coronary heart disease: a West Of Scotland Coronary Prevention Study (WOSCOPS) substudy. Heart. 2004; 90: 304–306.
Campbell S, George DK, Robb SD, Spooner R, McDonagh TA, Dargie HJ, Mills PR. The prevalence of haemochromatosis gene mutations in the West of Scotland and their relation to ischaemic heart disease. Heart. 2003; 89: 1023–1026.
Calado RT, Franco RF, Pazin-Filho A, Simoes MV, Marin-Neto JA, Zago MA. HFE gene mutations in coronary atherothrombotic disease. Braz J Med Biol Res. 2000; 33: 301–306.
Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, Dormishian F, Domingo R, Ellis MC, Fullan A, Hinton LM, Jones NL, Kimmel BE, Kronmal GS, Lauer P, Lee VK, Loeb DB, Mapa FA, McClelland E, Meyer NC, Mintier GA, Moeller N, Moore T, Morikang E, Prass CE, Quintana L, Starnes SM, Schatzman RC, Brunke KJ, Drayna DT, Risch NJ, Bacon BR. A novel MHC class I–like gene is mutated in patients with hereditary haemochromatosis. Nat Genet. 1996; 13: 399–408.
Andersen RV, Tybjaerg-Hansen A, Appleyard M, Birgens H, Nordestgaard BG. Hemochromatosis mutations in the general population: iron overload progression rate. Blood. 2004; 103: 2914–2919.
Beutler E, Felitti V, Gelbart T, Ho N. The effect of HFE genotypes on measurements of iron overload in patients attending a health appraisal clinic. Ann Intern Med. 2000; 133: 329–337.
Ellervik C, Mandrup-Poulsen T, Nordestgaard BG, Larsen LE, Appleyard M, Frandsen M, Petersen P, Schlichting P, Saermark T, Tybjaerg-Hansen A, Birgens H. Prevalence of hereditary haemochromatosis in late-onset type 1 diabetes mellitus: a retrospective study. Lancet. 2001; 358: 1405–1409.
The Copenhagen City Heart Study Group. The Copenhagen City Heart Study: ;sterbrounders;gelsen: a book of tables with data from the first examination (1976–78) and a five year follow-up (1981–83). Scand J Soc Med Suppl. 1989; 41: 1–160.
Schnohr P, Jensen G, Lange P, Scharling H, Appleyard M. The Copenhagen City Heart Study: ;sterbrounders;gelsen: tables with data from the third examination 1991–1994. Eur Heart J. 2001; 3: 1–83.
Management of stable angina pectoris: Recommendations of the Task Force of the European Society of Cardiology. Eur Heart J. 1997; 18: 394–413.
Guidelines Subcommittee. 1999 World Health Organization–International Society of Hypertension guidelines for the management of hypertension. J Hypertens. 1999; 17: 151–183.(Christina Ellervik, MD; A)