International Journal of Cardiology 183 (2015) 258–262
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
The incidental relationship between serum ferritin levels and hypertension☆ Jae-Hong Ryoo a, Sun Yong Kim b, Chang-Mo Oh c, Sung Keun Park a,b,⁎, Eugene Kim d, Se-Jin Park d, Jae In Yu e, Min-Gi Kim f, Yong-Sung Choi g, Taeg Su Ko d a
Departments of Preventive Medicine, Kyung Hee University School of Medicine, Seoul, Republic of Korea Total Healthcare Center, Kangbuk Samsung Hospital, Sungkyunkwan University, School of Medicine, Seoul, Republic of Korea Korea Central Cancer Registry, National Cancer Control Institute, National Cancer Center, Goyang, Republic of Korea d Department of Orthopaedic Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University, School of medicine, Seoul, Republic of Korea e Department of Medical Management, Graduate School, Gachon University, Incheon, Republic of Korea f Department of Occupational and Environmental Medicine, Dongguk University, Gyeongju Hospital, Gyeongsangbuk-do, Republic of Korea g Department of Pediatrics, Kyung Hee University School of Medicine, Seoul, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 6 June 2014 Received in revised form 14 October 2014 Accepted 22 October 2014 Available online 25 October 2014 Keywords: Ferritin Hypertension
a b s t r a c t Background and objective: Although several studies have shown an association between ferritin level and hypertension, only a few studies have investigated the longitudinal relationship between them. Thus, we evaluated the incidental risk for hypertension according to baseline ferritin level. Patients and methods: A total of 7104 healthy Korean men matched by a propensity score, who had participated in a medical health check-up program in 2005, were followed up from 2005 to 2010. They were divided into four groups according to baseline serum ferritin level (first quartile–fourth quartile). The incidence of hypertension was compared among the four groups, and the Cox-proportional hazard model was used to assess whether the development of hypertension was associated with higher baseline serum ferritin level. Results: A total of 1252 (17.6%) cases had newly developed hypertension during the 26,339.5 person-years of follow-up between 2006 and 2010. The adjusted hazard ratios (HRs) (95% confidence intervals, CIs) for incident hypertension were 1.00 (reference), 1.09 (0.91–1.30), 1.21 (1.01–1.45) and 1.28 (1.07–1.52), respectively (P for trend = 0.003) through the quartiles of serum ferritin levels, respectively, after adjusting for multiple confounders. For the log-transformed serum ferritin levels as a continuous variable, adjusted HRs and 95% CIs for HTN were 1.15 (1.02–1.29). Conclusions: Elevated serum ferritin level was independently associated with the incidental risk for hypertension in Korean men. This finding suggests the value of elevated ferritin level as an early predictor of hypertension. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction High blood pressure is one of the most common risk factors for ischemic heart disease, stroke, hypertensive heart disease, and renal failure [1]. Hypertension (HTN) ranked third as a global and regional disease burden and has been verified as the top risk factor for mortality
Abbreviations: HTN, hypertension; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; BMI,body mass index; HOMA-IR, homeostasismodel assessment of insulin resistance; WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, γ-glutamyltransferase; TIBC, total iron binding capacity. ☆ Financial support and disclosure statement: The authors have nothing to disclose. ⁎ Corresponding author at: Kangbuk Samsung Hospital, 78 Saemunan-gil, Jongro-Gu, Seoul 110-746, Republic of Korea. E-mail address: [email protected] (S.K. Park).
http://dx.doi.org/10.1016/j.ijcard.2014.10.152 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
[1]. Overall, 26.4% of the world's adult population had HTN in 2000, and this prevalence is likely to increase to 29.2% by 2025 [2]. Thus, prevention and prediction of HTN are essential for decreasing the global disease burden and cardiovascular mortality. The health risk posed by iron overload has been attracting much interest with the discovery that the C282Y mutation in the HFE gene is involved in hereditary hemochromatosis [3]. Serum ferritin level is a highly sensitive parameter used to evaluate body iron status [4]. Several studies have investigated the association between elevated serum ferritin and prevalence and risk for HTN [5–7]. However, clinical evidence remains limited for a concrete etiological association between high serum ferritin level and incident HTN because only one study has analyzed the temporal relationship between elevated serum ferritin and development of HTN [7]. Therefore, we conducted this study to assess the effect of elevated serum ferritin level on the future risk of HTN.
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
2. Participants and methods 2.1. Study design A prospective cohort study was conducted to examine the association between serum ferritin levels and the development of HTN in Korean men participating in a medical health check-up program at the Health Promotion Center of Kangbuk Samsung Hospital, Sungkyunkwan University, Seoul, Korea. The study methods and explanation of the medical health check-up program were described in detail in our previous study [8]. 2.2. Study population A total of 43,604 men (age, 30–59 yr), who had visited the Health Promotion Center at Kangbuk Samsung Hospital for a medical check-up in 2005, participated in this study. Among the 43,604 participants, 13,264 were excluded based on the following exclusion criteria that could influence HTN or serum ferritin level: 2120 had a positive serologic marker for hepatitis B surface antigen (HBs Ag); 41 had a positive serologic marker for hepatitis C virus antibody (HCV Ab); 47 had ultrasonographically detected liver cirrhosis; 400 had ultrasonographically detected chronic hepatic diseases; 1491 had a history of blood transfusion; 29 were regarded as probably having hemochromatosis based on abnormal serum ferritin level N 800 ng/mL; 193 had a history of malignancy; 210 had a history of cardiovascular disease; 2990 were receiving lipid-lowering medications; 11 had no baseline HTN information in 2005; and 8139 had baseline HTN at the initial examination. Because some participants had more than one exclusion criteria, the total number of men who were eligible for the study was 30,340. We further excluded 6872 participants who did not attend any follow-up visits between 2006 and 2010. Without the follow-up visit, we could not identify the development of HTN or calculate the individual person year. Among the 23,468 participants, 6043 were excluded, because they have missing values for covariate information. Among the 17,425 participants, 7104 participants were included in the final analysis after matching in a 1:1 ratio by a propensity score. The total follow-up period was 26,339.5 person-years, and the average follow-up period was 3.71 (standard deviation [SD], 1.35) personyears. Ethics approvals for this study protocol and analysis of the data were obtained from the institutional review board of Kangbuk Samsung Hospital. The informed consent requirement was exempted because we only retrospectively accessed a de-identified database for analytical purposes. 2.3. Clinical and laboratory measurements Study data included a medical history, a physical examination, information provided by a questionnaire, anthropometric measurements and laboratory measurements. The medical history and the history of drug prescription were assessed by the examining physicians. All the participants were asked to respond to a questionnaire on healthrelated behavior. Questions about alcohol intake included the frequency of alcohol consumption on a weekly basis and the usual amount that was consumed on a daily basis (≥ 20 g/day). We considered persons reporting that they smoked at that time to be current smokers. In addition, the participants were asked about their weekly frequency of physical activity, such as jogging, bicycling, and swimming that lasted long enough to produce perspiration (≥ 1 time/week). Diabetes mellitus was defined as fasting serum glucose of at least 126 mg/dL or current use of blood glucose-lowering agents. Systolic and diastolic blood pressure (BP) was measured with a standardized mercury sphygmomanometer after at least 5 min of seated rest using the Hypertension Detection and Follow-Up Protocol. According to the JNC-7 guidelines, HTN was defined as a systolic BP of at least
259
140 mm Hg or a diastolic BP at least 90 mm Hg, or current use of antihypertensive agents [9]. The development of HTN was assessed from the annual records of all participants and defined as blood pressure (BP) ≥ 140/90 mm Hg. Also, participants who had a history of HTN or currently used antihypertensive medication based on the self-report questionnaire at each visit were considered to have developed HTN. Anthropometric measurements and procedures for obtaining the blood samples were described in detail elsewhere [8]. Serum levels of ferritin, iron and total iron binding capacity (TIBC) were measured by electrochemiluminescence immunoassay using a Modular E170 analyzer (Roche Diagnostics, Basel, Switzerland). The clinical laboratory has been accredited and participates annually in inspections and surveys by the Korean Association of Quality Assurance for Clinical Laboratories. 2.4. Propensity-score models The highest three quartiles of baseline serum ferritin level were matched to the lowest quartile to balance the major covariates which was associated with the development of hypertension by using a propensity score. Multiple logistic regression was used to estimate propensity scores of highest three quartiles of baseline serum ferritin level on the lowest quartile level of serum ferritin level as the reference group after adjustments for the major covariates. Highest three quartile levels of serum ferritin level were matched with the reference group (1st quartile level of serum ferritin level) using a propensity score with a caliper method using SAS macro program (%PSMatching). The caliper was set as a 0.0004 to match the propensity score and the matching ratio was 1:1. Additional detailed information about propensity score matching is available in the Supplementary Appendix. 2.5. Statistical analyses Data were expressed as means ± (standard deviation) or medians (interquartile range) for continuous variables and percentages of the number for categorical variables. The one-way ANOVA and χ2-test were used to analyze the statistical differences among the characteristics of the study participants at the time of enrollment in relation to the quartile groups of serum ferritin levels. The distributions of continuous variables were evaluated, and log transformations were used in the analysis as required. For incident HTN cases, because we couldn't know the exact time for the development of HTN, the time of HTN was assumed to be the midpoint between the baseline visit (2005) and the visit at which HTN was first detected. The person years were calculated as the sump of follow-up times for the baseline until an assumed time of HTN development or until the final examination of each individual. To evaluate the associations of baseline serum ferritin levels and incident HTN, we used Cox-proportional hazard models stratified by matched pairs to estimate adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for incident HTN comparing the highest 3 quartiles of baseline fasting serum ferritin vs the lowest quartile. Cox-proportional hazard models were adjusted for the multiple confounding factors. In the multivariate models, we included variables that might confound the relationship between serum ferritin and incident HTN, which include: age, body mass index (BMI), white blood cell (WBC), low-density lipoprotein (LDL) cholesterol, log(hsCRP), homeostasis model assessment of insulin resistance (HOMA-IR), estimated glomerular filtration rate (eGFR), total iron binding capacity (TIBC), smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus. For the linear trends of risk, the number of quartiles was used as a continuous variable and tested on each model. We also conducted the sensitivity analysis, after serum ferritin was log-transformed as a continuous variable to assess the robustness of associations with the risk of HTN.
260
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
To test the validity of the Cox-proportional hazard models, we checked the proportional hazard assumption. The proportional hazard assumption was assessed by log-minus-log survival function and found to be graphically unviolated. P values b 0.05 were considered to be statistically significant. Statistical analyses were performed using SPSS Windows version 18.0 (SPSS Inc., Chicago, IL, USA) and SAS (version 9.3, SAS Institute, Cary, NC).
Table 3 shows the HRs and 95% CIs for HTN of the log-transformed serum ferritin level in the sensitivity analysis. In the unadjusted model, the HRs and 95% CIs for HTN of the log-transformed serum ferritin level was 1.18 (1.06–1.32). These associations was consistent both in models 1 and 2 after further adjustments for covariates. In model 2, the adjusted HRs and 95% CIs for HTN were 1.15 (1.02–1.29).
4. Discussion 3. Results We found that elevated serum ferritin level was a risk factor for incident HTN in Korean men. To the best of our knowledge, this is the first study to evaluate the longitudinal influence of baseline serum ferritin level on subsequent development of HTN based on a person-year analysis. Several studies have found significant associations between serum ferritin level and HTN [5–7]. Despite these studies, evidence remains insufficient to elucidate a definite etiological association between serum ferritin level and development of HTN. In particular, interval variables such as person-yr of events could not be counted in cross-sectional studies [5,6] or in a study that assessed the effect of serum ferritin level only on blood pressure taken at baseline and at the end of the study yr [7]. However, as our study reflected the interval process during follow-up, it might be better able to clarify the incidental relationship between the incidental HTN and ferritin level. In this study, the incidence of HTN was proportional to baseline serum ferritin level. In addition, the HRs for HTN increased in proportion to baseline serum ferritin level, even after adjusting for various metabolic factors. These findings suggest that elevated serum ferritin level may be an early predictor for the development of HTN. Ferritin is a major iron storage protein essential for iron homeostasis that is involved in a wide range of physiological and pathological processes [10]. Ferritin reflects systemic inflammatory status and oxidative stress-mediated cellular damage including metabolic syndrome [11,12] and rheumatic disease such as adult-onset Still's disease [13,14].
During the 26,339.5 person-years of follow-up, 1252 (17.6%) incident cases of HTN developed between 2006 and 2010. The baseline characteristics of the study participants in relation to quartile groups of serum ferritin levels are presented in Table 1. At baseline, the mean (SD) age and BMI of the study participants were 40.9 (5.7) years and 24.2 (2.6) kg/m2, respectively. There were no significant differences for major covariates across the quartile groups of serum ferritin levels, except for serum triglyceride, hsCRP, AST, ALT, GGT, TIBC, hemoglobin, iron, and transferrin saturation (%). In contrast to participants without incident HTN, those with incident HTN were slightly older (42.0 vs. 40.6 year) and more likely to have a less favorable metabolic profile at baseline. As expected, all clinical variables were significantly different between the two groups except for HDL-cholesterol level, hsCRP, serum iron level, smoking rate, and regular exercise (Supplementary Table S1). Table 2 shows HRs and 95% CIs for HTN according to the quartile groups of serum ferritin levels. In the unadjusted model, the HRs and 95% CIs for HTN comparing the highest three quartiles of baseline serum ferritin level vs. the lowest quartile were 1.02 (0.86–1.22), 1.16 (0.98–1.38), and 1.33 (1.12–1.57), respectively (P for trend b 0.001). These associations remained significant, even after further adjustments for covariates in models 1 and 2. In model 2, the adjusted HRs and 95% CIs for HTN were 1.09 (0.91–1.30), 1.21 (1.01–1.45), and 1.28 (1.07–1.52), respectively (P for trend = 0.003). Table 1 Baseline characteristics in propensity score matched participants (N = 7104). Characteristic
Age (years) BMI (kg/m2) Systolic BP (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dL) Triglyceride (mg/dL) HDL-cholesterol (mg/dL) LDL-cholesterol (mg/dL) Fasting serum glucose (mg/dL) HOMA-IR Insulin (μU/dL) SCr (mg/dL) eGFR (mL/min per 1.73 m2) hsCRP (mg/L) AST (U/L) ALT (U/L) GGT (U/L) WBC (×109/L) TIBC (μg/dL) Hemoglobin (g/dL) Iron (μg/dL) Transferrin saturation (%) Current smoker (%) Alcohol intake (%) Regular exercise (%) Type 2 diabetes mellitus (%) Development of HTN (%)
Overall
40.9 ± (5.7) 24.2 ± (2.6) 111.0 ± (10.7) 74.1 ± (6.9) 192.0 ± (30.9) 124 (91–173) 49.3 ± (9.6) 113.9 ± (26.0) 95.7 ± (11.0) 1.90 (1.48–2.51) 8.7 ± (3.2) 1.13 ± (0.10) 81.6 ± (9.4) 0.05 (0.03–0.11) 23 (19–28) 25 (19–34) 27 (19–44) 6.2 ± (1.5) 298.5 ± (31.5) 15.6 ± (0.9) 131.8 ± (47.5) 44.6 ± (16.6) 44.2 10.3 10.4 1.8 17.6
Quartiles of serum ferritin Quartile 1 (b70.3 ng/mL)
Quartile 2 (≥70.3, b101.8 ng/mL)
Quartile 3 (≥101.8, b147.4 ng/mL)
Quartile 4 (≥147.4 ng/mL)
P for trend⁎
40.8 ± (5.7) 24.2 ± (2.6) 111.2 ± (10.7) 74.1 ± (6.8) 192.6 ± (30.9) 125 (90–174) 49.0 ± (9.5) 114.7 ± (26.0) 95.8 ± (11.1) 1.93 (1.48–2.54) 8.8 ± (3.1) 1.13 ± (0.11) 81.6 ± (9.5) 0.06 (0.03–0.11) 22 (19–26) 23 (17–30) 24 (17–37) 6.2 ± (1.5) 298.8 ± (31.6) 15.5 ± (0.9) 127.5 ± (50.2) 43.2 ± (17.7) 44.3 10.9 10.4 1.9 16.3
41.0 ± (5.7) 24.1 ± (2.5) 110.6 ± (10.7) 74.1 ± (6.8) 189.6 ± (31.3) 118 (87–164) 49.5 ± (9.7) 112.2 ± (26.4) 95.4 ± (11.1) 1.88 (1.46–2.46) 8.6 ± (3.1) 1.13 ± (0.10) 81.5 ± (9.1) 0.05 (0.03–0.10) 23 (19–27) 23 (18–32) 26 (18–39) 6.2 ± (1.4) 296.2 ± (30.2) 15.5 ± (0.9) 129.1 ± (46.3) 44.0 ± (16.3) 44.0 9.5 10.3 1.4 15.9
40.7 ± (5.6) 24.1 ± (2.6) 110.9 ± (10.6) 74.1 ± (6.9) 193.4 ± (31.0) 125 (93–175) 49.6 ± (9.7) 114.6 ± (26.1) 95.5 ± (11.1) 1.86 (1.47–2.45) 8.6 ± (3.3) 1.13 ± (0.10) 81.6 ± (9.5) 0.05 (0.03–0.11) 23 (20–28) 25 (19–35) 28 (19–45) 6.2 ± (1.5) 298.6 ± (31.4) 15.6 ± (0.9) 134.6 ± (46.8) 45.4 ± (16.2) 45.6 10.6 9.1 2.1 18.2
40.9 ± (5.8) 24.3 ± (2.6) 111.3 ± (10.7) 74.3 ± (7.0) 192.5 ± (30.4) 129 (96–178) 49.2 ± (9.5) 114.3 ± (25.6) 96.1 ± (10.9) 1.96 (1.52–2.60) 8.9 ± (3.2) 1.12 ± (0.10) 81.8 ± (9.6) 0.06 (0.03–0.12) 25 (21–30) 28 (21–41) 34 (22–56) 6.2 ± (1.5) 300.3 ± (32.5) 15.7 ± (0.9) 136.0 ± (46.0) 45.7 ± (16.0) 43.0 10.1 11.9 2.0 20.0
0.99 0.48 0.76 0.26 0.30 0.02 0.60 0.66 0.41 0.30 0.31 0.44 0.45 0.02 b0.01 b0.01 b0.01 0.29 0.04 b0.01 b0.01 b0.01 0.48 0.54 0.06 0.33 b0.001
Data are means (standard deviation), medians (interquartile range), or percentages. ⁎ P-value by ANOVA-test for continuous variables and Chi square test for categorical variables.
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
261
Table 2 Hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of hypertension according to the quartile groups of serum ferritin levels in propensity score matched participants (N = 7104). Hazard ratios (95% confidence interval)
Quartiles of serum ferritin Quartile 1 (b70.3 ng/mL) Quartile 2 (≥70.3 ng/mL, b101.8 ng/mL) Quartile 3 (≥101.8 ng/mL, b147.4 ng/mL) Quartile 4 (≥147.4 ng/mL) P for trend Age (years) BMI (kg/m2) WBC (×109/L) LDL-cholesterol (mg/dL) Log(hsCRP) (mg/L) HOMA-IR eGFR (mL/min per 1.73 m2) TIBC (μg/dL) Smoking status Alcohol intake Regular exercise Type 2 diabetes mellitus
Unadjusted
Model 1
Model 2
1.00 (reference) 1.02 (0.86–1.22) 1.16 (0.98–1.38) 1.33 (1.12–1.57) b0.001
1.00 (reference) 1.08 (0.90–1.29) 1.21 (1.01–1.44) 1.28 (1.07–1.52) 0.003 1.06 (1.04–1.07) 1.10 (1.06–1.14) 1.03 (0.97–1.08) 1.00 (0.99–1.00) 0.98 (0.91–1.06) 1.14 (1.03–1.26) 1.00 (0.99–1.01) 1.01 (1.00–1.01)
1.00 (reference) 1.09 (0.91–1.30) 1.21 (1.01–1.45) 1.28 (1.07–1.52) 0.003 1.05 (1.04–1.07) 1.10 (1.06–1.15) 1.04 (0.99–1.10) 1.00 (1.00–1.01) 0.99 (0.91–1.07) 1.14 (1.03–1.26) 1.00 (0.99–1.01) 1.01 (1.00–1.01) 1.12 (0.96–1.30) 1.24 (0.97–1.58) 0.92 (0.72–1.17) 1.58 (0.99–2.52)
Model 1 was adjusted for age, BMI, WBC, LDL-cholesterol, log(hsCRP), HOMA-IR, eGFR and TIBC. Model 2 was adjusted for model 1 plus smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus.
Nevertheless, the underlying mechanism by which elevated ferritin level contributes to the development of HTN is unclear. Possible mechanisms include atherosclerosis by iron itself, insulin resistance, and a fatty liver-mediated process. Iron overload increases vascular oxidative stress and impairs vasoreactivity [15]. Moreover, atherosclerosis can develop and accelerate due to iron promoting direct endothelial toxicity and oxidizing low-density lipoprotein and isoprostanes [16–18]. In addition, increased total body iron may result in increased rates of adipocyte lipolysis and increased circulating free-fatty acid levels and insulin resistance [19]. Iron is intimately related with oxidative stress and participates in the formation of highly toxic free radicals, such as hydroxide and the superoxide anion [20]. Oxidative stress affected insulin signaling at the cellular level [21]. Hyperinsulinemia and insulin resistance are associated with HTN [22–25]. Thus, incident HTN and serum ferritin level may be mediated by insulin resistance. Insulin resistance induced by iron overload could Table 3 Hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of hypertension according to log-transformed serum ferritin levels in propensity score matched participants (N = 7104). Hazard ratios (95% confidence interval)
Log(ferritin) (ng/mL) Age (years) BMI (kg/m2) WBC (×109/L) LDL-cholesterol (mg/dL) Log(hsCRP) (mg/L) HOMA-IR eGFR (mL/min per 1.73 m2) TIBC (μg/dL) Smoking status Alcohol intake Regular exercise Type 2 diabetes mellitus
Unadjusted
Model 1
Model 2
1.18 (1.06–1.32)
1.15 (1.02–1.29) 1.06 (1.04–1.07) 1.10 (1.06–1.14) 1.03 (0.97–1.08) 1.00 (0.99–1.01) 0.98 (0.91–1.06) 1.14 (1.03–1.25) 1.00 (0.99–1.01)
1.15 (1.02–1.29) 1.05 (1.04–1.07) 1.10 (1.06–1.14) 1.04 (0.99–1.10) 1.00 (1.00–1.01) 0.99 (0.91–1.07) 1.14 (1.03–1.26) 1.00 (0.99–1.01)
1.01 (1.01–1.01)
1.01 (1.00–1.01) 1.11 (0.96–1.29) 1.23 (0.97–1.57) 0.92 (0.72–1.17) 1.58 (0.99–2.52)
Model 1 was adjusted for age, BMI, WBC, LDL-cholesterol, log(hsCRP), HOMA-IR, eGFR and TIBC. Model 2 was adjusted for model 1 plus smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus.
be a causative factor in vascular endothelial dysfunction [26]. Hyperinsulinemia may also evoke sympathetic overactivity by stimulating noradrenaline release, particularly in the skeletal muscle [27]. Chronically elevated sympathetic tone plays a significant role in increasing BP [28]. In addition, hyperinsulinemia can affect the renovascular system, which promotes renal sodium and water retention by increasing tubular reabsorption of sodium [29–31], which can lead to HTN. Several studies have suggested that insulin could interfere with the systemic rennin–angiotensin system (RAS) and intrarenal RAS that might contribute to the development of HTN [32–34]. Taken together, not only the various effects of ferritin itself but insulin resistance and liver-mediated effects of ferritin on the vascular endothelial, endocrine, and autonomic nervous systems might contribute to the development of HTN. Several limitations of the present study should be mentioned. First, our study population was comprised of only Korean men. Serum ferritin levels are generally lower in females due to menstrual iron loss [35]. Therefore, it is unclear whether our findings can be extrapolated to females and other ethnicities. Second, as our study was epidemiological, it was insufficient to clarify the causative relationship between serum ferritin levels and the development of HTN. Therefore, clinical and laboratory studies should be performed to identify the causative relationship. Third, blood pressure was measured only on the day of the health check-up without follow-up or continuous monitoring. This may have caused variability in the measured blood pressure and influenced the proportion of incident HTN. In particular, overestimating hypertension such as white coat hypertension could be a major limitation of this study. However, we think that the proportion of overestimated HTN cases was minimal and did not affect the results. Although white coat hypertension can occur in anyone, it is more prevalent in females and elderly ≥ 60 yr [36–39]. However, all of our study participants were middle-aged Korean men. Accordingly, considering the characteristics of our study population, the effect of white coat hypertension was not substantial. In conclusion, elevated serum ferritin level was independently associated with incident HTN in Korean men. The risk of incident HTN was proportional to the baseline serum ferritin level. These findings suggest the value of elevated serum ferritin level as an early predictor of the development of HTN.
262
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
Conflict of interest No potential conflicts of interest relevant to this article were reported. Author contribution Sung Keun Park collected, analyzed and interpreted the data. Sung Keun Park is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Jae-Hong Ryoo coordinated the study and wrote the manuscript as a first author. Sun Yong Kim contributed to writing and editing the manuscript. Chang-Mo Oh participated in the revising manuscript and statistical analysis. Eugene Kim, Sae-Jin Park, Jae In Yu, Min-Gi Kim and Yong-Sung Choi participated in reviewing and editing manuscript. Taeg-Su Ko had a significant role in revising and editing manuscript. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2014.10.152. References [1] M. Ezzati, A.D. Lopez, A. Rodgers, S. Vander Hoorn, C.J. Murray, Selected major risk factors and global and regional burden of disease, Lancet 360 (2002) 1347–1360. [2] P.M. Kearney, M. Whelton, K. Reynolds, P. Muntner, P.K. Whelton, J. He, Global burden of hypertension: analysis of worldwide data, Lancet 365 (2005) 217–223. [3] A.L. Heath, S.J. Fairweather-Tait, Health implications of iron overload: the role of diet and genotype, Nutr. Rev. 61 (2003) 45–62. [4] C.A. Whittington, K.V. Kowdley, Review article: haemochromatosis, Aliment. Pharmacol. Ther. 16 (2002) 1963–1975. [5] A. Piperno, P. Trombini, M. Gelosa, V. Mauri, V. Pecci, A. Vergani, et al., Increased serum ferritin is common in men with essential hypertension, J. Hypertens. 20 (2002) 1513–1518. [6] B. Choi, K.J. Yeum, S.J. Park, K.N. Kim, N.S. Joo, Elevated serum ferritin and mercury concentrations are associated with hypertension; analysis of the fourth and fifth Korea national health and nutrition examination survey (KNHANES IV-2, 3, 2008– 2009 and V-1, 2010), Environ. Toxicol. (2013). [7] M.K. Kim, K.H. Baek, K.H. Song, M.I. Kang, J.H. Choi, J.C. Bae, et al., Increased serum ferritin predicts the development of hypertension among middle-aged men, Am. J. Hypertens. 25 (2012) 492–497. [8] S.K. Park, M.H. Seo, H.C. Shin, J.H. Ryoo, Clinical availability of nonalcoholic fatty liver disease as an early predictor of type 2 diabetes mellitus in Korean men: 5-year prospective cohort study, Hepatology 57 (2013) 1378–1383. [9] A.V. Chobanian, G.L. Bakris, H.R. Black, W.C. Cushman, L.A. Green, J.L. Izzo Jr., et al., Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure, Hypertension 42 (2003) 1206–1252. [10] A. Dominguez-Rodriguez, M. Carrillo-Perez Tome, C. Hernandez-Garcia, E. ArroyoUcar, R. Juarez-Prera, G. Blanco-Palacios, et al., Serum ferritin and acute coronary syndrome: a strong prognostic factor? Int. J. Cardiol. 152 (2011) 129–130. [11] C. Gabay, I. Kushner, Acute-phase proteins and other systemic responses to inflammation, N. Engl. J. Med. 340 (1999) 448–454. [12] S.K. Park, J.H. Ryoo, M.G. Kim, J.Y. Shin, Association of serum ferritin and the development of metabolic syndrome in middle-aged Korean men: a 5-year follow-up study, Diabetes Care 35 (2012) 2521–2526. [13] S. Franchini, L. Dagna, F. Salvo, P. Aiello, E. Baldissera, M.G. Sabbadini, Adult onset Still's disease: clinical presentation in a large cohort of Italian patients, Clin. Exp. Rheumatol. 28 (2010) 41–48. [14] X.D. Kong, D. Xu, W. Zhang, Y. Zhao, X. Zeng, F. Zhang, Clinical features and prognosis in adult-onset Still's disease: a study of 104 cases, Clin. Rheumatol. 29 (2010) 1015–1019. [15] S.M. Day, D. Duquaine, L.V. Mundada, R.G. Menon, B.V. Khan, S. Rajagopalan, et al., Chronic iron administration increases vascular oxidative stress and accelerates arterial thrombosis, Circulation 107 (2003) 2601–2606.
[16] B.L. Crist, D.L. Alekel, L.M. Ritland, L.N. Hanson, U. Genschel, M.B. Reddy, Association of oxidative stress, iron, and centralized fat mass in healthy postmenopausal women, J. Women's Health (Larchmt) 18 (2009) 795–801. [17] J.L. Cracowski, P. Devillier, T. Durand, F. Stanke-Labesque, G. Bessard, Vascular biology of the isoprostanes, J. Vasc. Res. 38 (2001) 93–103. [18] S.J. Duffy, E.S. Biegelsen, M. Holbrook, J.D. Russell, N. Gokce, J.F. Keaney Jr., et al., Iron chelation improves endothelial function in patients with coronary artery disease, Circulation 103 (2001) 2799–2804. [19] A. Green, R. Basile, J.M. Rumberger, Transferrin and iron induce insulin resistance of glucose transport in adipocytes, Metab. Clin. Exp. 55 (2006) 1042–1045. [20] J.M. Fernandez-Real, A. Lopez-Bermejo, W. Ricart, Cross-talk between iron metabolism and diabetes, Diabetes 51 (2002) 2348–2354. [21] J.L. Rains, S.K. Jain, Oxidative stress, insulin signaling, and diabetes, Free Radic. Biol. Med. 50 (2011) 567–575. [22] B.M. Cheung, N.M. Wat, Y.B. Man, S. Tam, C.H. Cheng, G.M. Leung, et al., Relationship between the metabolic syndrome and the development of hypertension in the Hong Kong Cardiovascular Risk Factor Prevalence Study-2 (CRISPS2), Am. J. Hypertens. 21 (2008) 17–22. [23] N.R. Every, E.J. Boyko, E.M. Keane, J.A. Marshall, M. Rewers, R.F. Hamman, Blood pressure, insulin, and C-peptide levels in San Luis Valley, Colorado, Diabetes Care 16 (1993) 1543–1550. [24] T.A. Manolio, P.J. Savage, G.L. Burke, K.A. Liu, L.E. Wagenknecht, S. Sidney, et al., Association of fasting insulin with blood pressure and lipids in young adults. The CARDIA study, Arteriosclerosis 10 (1990) 430–436. [25] S.K. Park, J.Y. Jung, W.J. Choi, Y.H. Kim, H.S. Kim, W.T. Ham, et al., Elevated fasting serum insulin level predicts future development of hypertension, Int. J. Cardiol. 172 (2014) 450–455. [26] G. Arcaro, A. Cretti, S. Balzano, A. Lechi, M. Muggeo, E. Bonora, et al., Insulin causes endothelial dysfunction in humans: sites and mechanisms, Circulation 105 (2002) 576–582. [27] G. Lembo, R. Napoli, B. Capaldo, V. Rendina, G. Iaccarino, M. Volpe, et al., Abnormal sympathetic overactivity evoked by insulin in the skeletal muscle of patients with essential hypertension, J. Clin. Invest. 90 (1992) 24–29. [28] D. Butlen, S. Vadrot, S. Roseau, F. Morel, Insulin receptors along the rat nephron: [125I] insulin binding in microdissected glomeruli and tubules, Pflugers Arch. — Eur. J. Physiol. 412 (1988) 604–612. [29] R.A. DeFronzo, C.R. Cooke, R. Andres, G.R. Faloona, P.J. Davis, The effect of insulin on renal handling of sodium, potassium, calcium, and phosphate in man, J. Clin. Invest. 55 (1975) 845–855. [30] J.H. Miller, M.D. Bogdonoff, Antidiuresis associated with administration of insulin, J. Appl. Physiol. 6 (1954) 509–512. [31] A. Nizet, P. Lefebvre, J. Crabbe, Control by insulin of sodium potassium and water excretion by the isolated dog kidney, Pflugers Arch. — Eur. J. Physiol. 323 (1971) 11–20. [32] A.P. Rocchini, C. Moorehead, S. DeRemer, T.L. Goodfriend, D.L. Ball, Hyperinsulinemia and the aldosterone and pressor responses to angiotensin II, Hypertension 15 (1990) 861–866. [33] G. Nickenig, J. Roling, K. Strehlow, P. Schnabel, M. Bohm, Insulin induces upregulation of vascular AT1 receptor gene expression by posttranscriptional mechanisms, Circulation 98 (1998) 2453–2460. [34] P.W. Anderson, X.Y. Zhang, J. Tian, J.D. Correale, X.P. Xi, D. Yang, et al., Insulin and angiotensin II are additive in stimulating TGF-beta 1 and matrix mRNAs in mesangial cells, Kidney Int. 50 (1996) 745–753. [35] D.H. Rushton, R. Dover, A.W. Sainsbury, M.J. Norris, J.J. Gilkes, I.D. Ramsay, Why should women have lower reference limits for haemoglobin and ferritin concentrations than men? BMJ 322 (2001) 1355–1357. [36] T.G. Pickering, J.E. Hall, L.J. Appel, B.E. Falkner, J. Graves, M.N. Hill, et al., Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research, Circulation 111 (2005) 697–716. [37] M.G. Myers, R.A. Reeves, White-coat effect in treated hypertensive patients: sex differences, J. Hum. Hypertens. 9 (1995) 729–733. [38] M.B. MacDonald, G.P. Laing, M.P. Wilson, T.W. Wilson, Prevalence and predictors of white-coat response in patients with treated hypertension, Can. Med. Assoc. J. 161 (1999) 265–269. [39] S. Abir-Khalil, S. Zaîmi, M.A. Tazi, S. Bendahmane, O. Bensaoud, M. Benomar, Prevalence and predictors of white-coat hypertension in a large database of ambulatory blood pressure monitoring, East Mediterr. Health J. 15 (2009) 400–407.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
The incidental relationship between serum ferritin levels and hypertension☆ Jae-Hong Ryoo a, Sun Yong Kim b, Chang-Mo Oh c, Sung Keun Park a,b,⁎, Eugene Kim d, Se-Jin Park d, Jae In Yu e, Min-Gi Kim f, Yong-Sung Choi g, Taeg Su Ko d a
Departments of Preventive Medicine, Kyung Hee University School of Medicine, Seoul, Republic of Korea Total Healthcare Center, Kangbuk Samsung Hospital, Sungkyunkwan University, School of Medicine, Seoul, Republic of Korea Korea Central Cancer Registry, National Cancer Control Institute, National Cancer Center, Goyang, Republic of Korea d Department of Orthopaedic Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University, School of medicine, Seoul, Republic of Korea e Department of Medical Management, Graduate School, Gachon University, Incheon, Republic of Korea f Department of Occupational and Environmental Medicine, Dongguk University, Gyeongju Hospital, Gyeongsangbuk-do, Republic of Korea g Department of Pediatrics, Kyung Hee University School of Medicine, Seoul, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 6 June 2014 Received in revised form 14 October 2014 Accepted 22 October 2014 Available online 25 October 2014 Keywords: Ferritin Hypertension
a b s t r a c t Background and objective: Although several studies have shown an association between ferritin level and hypertension, only a few studies have investigated the longitudinal relationship between them. Thus, we evaluated the incidental risk for hypertension according to baseline ferritin level. Patients and methods: A total of 7104 healthy Korean men matched by a propensity score, who had participated in a medical health check-up program in 2005, were followed up from 2005 to 2010. They were divided into four groups according to baseline serum ferritin level (first quartile–fourth quartile). The incidence of hypertension was compared among the four groups, and the Cox-proportional hazard model was used to assess whether the development of hypertension was associated with higher baseline serum ferritin level. Results: A total of 1252 (17.6%) cases had newly developed hypertension during the 26,339.5 person-years of follow-up between 2006 and 2010. The adjusted hazard ratios (HRs) (95% confidence intervals, CIs) for incident hypertension were 1.00 (reference), 1.09 (0.91–1.30), 1.21 (1.01–1.45) and 1.28 (1.07–1.52), respectively (P for trend = 0.003) through the quartiles of serum ferritin levels, respectively, after adjusting for multiple confounders. For the log-transformed serum ferritin levels as a continuous variable, adjusted HRs and 95% CIs for HTN were 1.15 (1.02–1.29). Conclusions: Elevated serum ferritin level was independently associated with the incidental risk for hypertension in Korean men. This finding suggests the value of elevated ferritin level as an early predictor of hypertension. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction High blood pressure is one of the most common risk factors for ischemic heart disease, stroke, hypertensive heart disease, and renal failure [1]. Hypertension (HTN) ranked third as a global and regional disease burden and has been verified as the top risk factor for mortality
Abbreviations: HTN, hypertension; BP, blood pressure; LDL, low-density lipoprotein; HDL, high-density lipoprotein; BMI,body mass index; HOMA-IR, homeostasismodel assessment of insulin resistance; WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, γ-glutamyltransferase; TIBC, total iron binding capacity. ☆ Financial support and disclosure statement: The authors have nothing to disclose. ⁎ Corresponding author at: Kangbuk Samsung Hospital, 78 Saemunan-gil, Jongro-Gu, Seoul 110-746, Republic of Korea. E-mail address: [email protected] (S.K. Park).
http://dx.doi.org/10.1016/j.ijcard.2014.10.152 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
[1]. Overall, 26.4% of the world's adult population had HTN in 2000, and this prevalence is likely to increase to 29.2% by 2025 [2]. Thus, prevention and prediction of HTN are essential for decreasing the global disease burden and cardiovascular mortality. The health risk posed by iron overload has been attracting much interest with the discovery that the C282Y mutation in the HFE gene is involved in hereditary hemochromatosis [3]. Serum ferritin level is a highly sensitive parameter used to evaluate body iron status [4]. Several studies have investigated the association between elevated serum ferritin and prevalence and risk for HTN [5–7]. However, clinical evidence remains limited for a concrete etiological association between high serum ferritin level and incident HTN because only one study has analyzed the temporal relationship between elevated serum ferritin and development of HTN [7]. Therefore, we conducted this study to assess the effect of elevated serum ferritin level on the future risk of HTN.
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
2. Participants and methods 2.1. Study design A prospective cohort study was conducted to examine the association between serum ferritin levels and the development of HTN in Korean men participating in a medical health check-up program at the Health Promotion Center of Kangbuk Samsung Hospital, Sungkyunkwan University, Seoul, Korea. The study methods and explanation of the medical health check-up program were described in detail in our previous study [8]. 2.2. Study population A total of 43,604 men (age, 30–59 yr), who had visited the Health Promotion Center at Kangbuk Samsung Hospital for a medical check-up in 2005, participated in this study. Among the 43,604 participants, 13,264 were excluded based on the following exclusion criteria that could influence HTN or serum ferritin level: 2120 had a positive serologic marker for hepatitis B surface antigen (HBs Ag); 41 had a positive serologic marker for hepatitis C virus antibody (HCV Ab); 47 had ultrasonographically detected liver cirrhosis; 400 had ultrasonographically detected chronic hepatic diseases; 1491 had a history of blood transfusion; 29 were regarded as probably having hemochromatosis based on abnormal serum ferritin level N 800 ng/mL; 193 had a history of malignancy; 210 had a history of cardiovascular disease; 2990 were receiving lipid-lowering medications; 11 had no baseline HTN information in 2005; and 8139 had baseline HTN at the initial examination. Because some participants had more than one exclusion criteria, the total number of men who were eligible for the study was 30,340. We further excluded 6872 participants who did not attend any follow-up visits between 2006 and 2010. Without the follow-up visit, we could not identify the development of HTN or calculate the individual person year. Among the 23,468 participants, 6043 were excluded, because they have missing values for covariate information. Among the 17,425 participants, 7104 participants were included in the final analysis after matching in a 1:1 ratio by a propensity score. The total follow-up period was 26,339.5 person-years, and the average follow-up period was 3.71 (standard deviation [SD], 1.35) personyears. Ethics approvals for this study protocol and analysis of the data were obtained from the institutional review board of Kangbuk Samsung Hospital. The informed consent requirement was exempted because we only retrospectively accessed a de-identified database for analytical purposes. 2.3. Clinical and laboratory measurements Study data included a medical history, a physical examination, information provided by a questionnaire, anthropometric measurements and laboratory measurements. The medical history and the history of drug prescription were assessed by the examining physicians. All the participants were asked to respond to a questionnaire on healthrelated behavior. Questions about alcohol intake included the frequency of alcohol consumption on a weekly basis and the usual amount that was consumed on a daily basis (≥ 20 g/day). We considered persons reporting that they smoked at that time to be current smokers. In addition, the participants were asked about their weekly frequency of physical activity, such as jogging, bicycling, and swimming that lasted long enough to produce perspiration (≥ 1 time/week). Diabetes mellitus was defined as fasting serum glucose of at least 126 mg/dL or current use of blood glucose-lowering agents. Systolic and diastolic blood pressure (BP) was measured with a standardized mercury sphygmomanometer after at least 5 min of seated rest using the Hypertension Detection and Follow-Up Protocol. According to the JNC-7 guidelines, HTN was defined as a systolic BP of at least
259
140 mm Hg or a diastolic BP at least 90 mm Hg, or current use of antihypertensive agents [9]. The development of HTN was assessed from the annual records of all participants and defined as blood pressure (BP) ≥ 140/90 mm Hg. Also, participants who had a history of HTN or currently used antihypertensive medication based on the self-report questionnaire at each visit were considered to have developed HTN. Anthropometric measurements and procedures for obtaining the blood samples were described in detail elsewhere [8]. Serum levels of ferritin, iron and total iron binding capacity (TIBC) were measured by electrochemiluminescence immunoassay using a Modular E170 analyzer (Roche Diagnostics, Basel, Switzerland). The clinical laboratory has been accredited and participates annually in inspections and surveys by the Korean Association of Quality Assurance for Clinical Laboratories. 2.4. Propensity-score models The highest three quartiles of baseline serum ferritin level were matched to the lowest quartile to balance the major covariates which was associated with the development of hypertension by using a propensity score. Multiple logistic regression was used to estimate propensity scores of highest three quartiles of baseline serum ferritin level on the lowest quartile level of serum ferritin level as the reference group after adjustments for the major covariates. Highest three quartile levels of serum ferritin level were matched with the reference group (1st quartile level of serum ferritin level) using a propensity score with a caliper method using SAS macro program (%PSMatching). The caliper was set as a 0.0004 to match the propensity score and the matching ratio was 1:1. Additional detailed information about propensity score matching is available in the Supplementary Appendix. 2.5. Statistical analyses Data were expressed as means ± (standard deviation) or medians (interquartile range) for continuous variables and percentages of the number for categorical variables. The one-way ANOVA and χ2-test were used to analyze the statistical differences among the characteristics of the study participants at the time of enrollment in relation to the quartile groups of serum ferritin levels. The distributions of continuous variables were evaluated, and log transformations were used in the analysis as required. For incident HTN cases, because we couldn't know the exact time for the development of HTN, the time of HTN was assumed to be the midpoint between the baseline visit (2005) and the visit at which HTN was first detected. The person years were calculated as the sump of follow-up times for the baseline until an assumed time of HTN development or until the final examination of each individual. To evaluate the associations of baseline serum ferritin levels and incident HTN, we used Cox-proportional hazard models stratified by matched pairs to estimate adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for incident HTN comparing the highest 3 quartiles of baseline fasting serum ferritin vs the lowest quartile. Cox-proportional hazard models were adjusted for the multiple confounding factors. In the multivariate models, we included variables that might confound the relationship between serum ferritin and incident HTN, which include: age, body mass index (BMI), white blood cell (WBC), low-density lipoprotein (LDL) cholesterol, log(hsCRP), homeostasis model assessment of insulin resistance (HOMA-IR), estimated glomerular filtration rate (eGFR), total iron binding capacity (TIBC), smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus. For the linear trends of risk, the number of quartiles was used as a continuous variable and tested on each model. We also conducted the sensitivity analysis, after serum ferritin was log-transformed as a continuous variable to assess the robustness of associations with the risk of HTN.
260
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
To test the validity of the Cox-proportional hazard models, we checked the proportional hazard assumption. The proportional hazard assumption was assessed by log-minus-log survival function and found to be graphically unviolated. P values b 0.05 were considered to be statistically significant. Statistical analyses were performed using SPSS Windows version 18.0 (SPSS Inc., Chicago, IL, USA) and SAS (version 9.3, SAS Institute, Cary, NC).
Table 3 shows the HRs and 95% CIs for HTN of the log-transformed serum ferritin level in the sensitivity analysis. In the unadjusted model, the HRs and 95% CIs for HTN of the log-transformed serum ferritin level was 1.18 (1.06–1.32). These associations was consistent both in models 1 and 2 after further adjustments for covariates. In model 2, the adjusted HRs and 95% CIs for HTN were 1.15 (1.02–1.29).
4. Discussion 3. Results We found that elevated serum ferritin level was a risk factor for incident HTN in Korean men. To the best of our knowledge, this is the first study to evaluate the longitudinal influence of baseline serum ferritin level on subsequent development of HTN based on a person-year analysis. Several studies have found significant associations between serum ferritin level and HTN [5–7]. Despite these studies, evidence remains insufficient to elucidate a definite etiological association between serum ferritin level and development of HTN. In particular, interval variables such as person-yr of events could not be counted in cross-sectional studies [5,6] or in a study that assessed the effect of serum ferritin level only on blood pressure taken at baseline and at the end of the study yr [7]. However, as our study reflected the interval process during follow-up, it might be better able to clarify the incidental relationship between the incidental HTN and ferritin level. In this study, the incidence of HTN was proportional to baseline serum ferritin level. In addition, the HRs for HTN increased in proportion to baseline serum ferritin level, even after adjusting for various metabolic factors. These findings suggest that elevated serum ferritin level may be an early predictor for the development of HTN. Ferritin is a major iron storage protein essential for iron homeostasis that is involved in a wide range of physiological and pathological processes [10]. Ferritin reflects systemic inflammatory status and oxidative stress-mediated cellular damage including metabolic syndrome [11,12] and rheumatic disease such as adult-onset Still's disease [13,14].
During the 26,339.5 person-years of follow-up, 1252 (17.6%) incident cases of HTN developed between 2006 and 2010. The baseline characteristics of the study participants in relation to quartile groups of serum ferritin levels are presented in Table 1. At baseline, the mean (SD) age and BMI of the study participants were 40.9 (5.7) years and 24.2 (2.6) kg/m2, respectively. There were no significant differences for major covariates across the quartile groups of serum ferritin levels, except for serum triglyceride, hsCRP, AST, ALT, GGT, TIBC, hemoglobin, iron, and transferrin saturation (%). In contrast to participants without incident HTN, those with incident HTN were slightly older (42.0 vs. 40.6 year) and more likely to have a less favorable metabolic profile at baseline. As expected, all clinical variables were significantly different between the two groups except for HDL-cholesterol level, hsCRP, serum iron level, smoking rate, and regular exercise (Supplementary Table S1). Table 2 shows HRs and 95% CIs for HTN according to the quartile groups of serum ferritin levels. In the unadjusted model, the HRs and 95% CIs for HTN comparing the highest three quartiles of baseline serum ferritin level vs. the lowest quartile were 1.02 (0.86–1.22), 1.16 (0.98–1.38), and 1.33 (1.12–1.57), respectively (P for trend b 0.001). These associations remained significant, even after further adjustments for covariates in models 1 and 2. In model 2, the adjusted HRs and 95% CIs for HTN were 1.09 (0.91–1.30), 1.21 (1.01–1.45), and 1.28 (1.07–1.52), respectively (P for trend = 0.003). Table 1 Baseline characteristics in propensity score matched participants (N = 7104). Characteristic
Age (years) BMI (kg/m2) Systolic BP (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dL) Triglyceride (mg/dL) HDL-cholesterol (mg/dL) LDL-cholesterol (mg/dL) Fasting serum glucose (mg/dL) HOMA-IR Insulin (μU/dL) SCr (mg/dL) eGFR (mL/min per 1.73 m2) hsCRP (mg/L) AST (U/L) ALT (U/L) GGT (U/L) WBC (×109/L) TIBC (μg/dL) Hemoglobin (g/dL) Iron (μg/dL) Transferrin saturation (%) Current smoker (%) Alcohol intake (%) Regular exercise (%) Type 2 diabetes mellitus (%) Development of HTN (%)
Overall
40.9 ± (5.7) 24.2 ± (2.6) 111.0 ± (10.7) 74.1 ± (6.9) 192.0 ± (30.9) 124 (91–173) 49.3 ± (9.6) 113.9 ± (26.0) 95.7 ± (11.0) 1.90 (1.48–2.51) 8.7 ± (3.2) 1.13 ± (0.10) 81.6 ± (9.4) 0.05 (0.03–0.11) 23 (19–28) 25 (19–34) 27 (19–44) 6.2 ± (1.5) 298.5 ± (31.5) 15.6 ± (0.9) 131.8 ± (47.5) 44.6 ± (16.6) 44.2 10.3 10.4 1.8 17.6
Quartiles of serum ferritin Quartile 1 (b70.3 ng/mL)
Quartile 2 (≥70.3, b101.8 ng/mL)
Quartile 3 (≥101.8, b147.4 ng/mL)
Quartile 4 (≥147.4 ng/mL)
P for trend⁎
40.8 ± (5.7) 24.2 ± (2.6) 111.2 ± (10.7) 74.1 ± (6.8) 192.6 ± (30.9) 125 (90–174) 49.0 ± (9.5) 114.7 ± (26.0) 95.8 ± (11.1) 1.93 (1.48–2.54) 8.8 ± (3.1) 1.13 ± (0.11) 81.6 ± (9.5) 0.06 (0.03–0.11) 22 (19–26) 23 (17–30) 24 (17–37) 6.2 ± (1.5) 298.8 ± (31.6) 15.5 ± (0.9) 127.5 ± (50.2) 43.2 ± (17.7) 44.3 10.9 10.4 1.9 16.3
41.0 ± (5.7) 24.1 ± (2.5) 110.6 ± (10.7) 74.1 ± (6.8) 189.6 ± (31.3) 118 (87–164) 49.5 ± (9.7) 112.2 ± (26.4) 95.4 ± (11.1) 1.88 (1.46–2.46) 8.6 ± (3.1) 1.13 ± (0.10) 81.5 ± (9.1) 0.05 (0.03–0.10) 23 (19–27) 23 (18–32) 26 (18–39) 6.2 ± (1.4) 296.2 ± (30.2) 15.5 ± (0.9) 129.1 ± (46.3) 44.0 ± (16.3) 44.0 9.5 10.3 1.4 15.9
40.7 ± (5.6) 24.1 ± (2.6) 110.9 ± (10.6) 74.1 ± (6.9) 193.4 ± (31.0) 125 (93–175) 49.6 ± (9.7) 114.6 ± (26.1) 95.5 ± (11.1) 1.86 (1.47–2.45) 8.6 ± (3.3) 1.13 ± (0.10) 81.6 ± (9.5) 0.05 (0.03–0.11) 23 (20–28) 25 (19–35) 28 (19–45) 6.2 ± (1.5) 298.6 ± (31.4) 15.6 ± (0.9) 134.6 ± (46.8) 45.4 ± (16.2) 45.6 10.6 9.1 2.1 18.2
40.9 ± (5.8) 24.3 ± (2.6) 111.3 ± (10.7) 74.3 ± (7.0) 192.5 ± (30.4) 129 (96–178) 49.2 ± (9.5) 114.3 ± (25.6) 96.1 ± (10.9) 1.96 (1.52–2.60) 8.9 ± (3.2) 1.12 ± (0.10) 81.8 ± (9.6) 0.06 (0.03–0.12) 25 (21–30) 28 (21–41) 34 (22–56) 6.2 ± (1.5) 300.3 ± (32.5) 15.7 ± (0.9) 136.0 ± (46.0) 45.7 ± (16.0) 43.0 10.1 11.9 2.0 20.0
0.99 0.48 0.76 0.26 0.30 0.02 0.60 0.66 0.41 0.30 0.31 0.44 0.45 0.02 b0.01 b0.01 b0.01 0.29 0.04 b0.01 b0.01 b0.01 0.48 0.54 0.06 0.33 b0.001
Data are means (standard deviation), medians (interquartile range), or percentages. ⁎ P-value by ANOVA-test for continuous variables and Chi square test for categorical variables.
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
261
Table 2 Hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of hypertension according to the quartile groups of serum ferritin levels in propensity score matched participants (N = 7104). Hazard ratios (95% confidence interval)
Quartiles of serum ferritin Quartile 1 (b70.3 ng/mL) Quartile 2 (≥70.3 ng/mL, b101.8 ng/mL) Quartile 3 (≥101.8 ng/mL, b147.4 ng/mL) Quartile 4 (≥147.4 ng/mL) P for trend Age (years) BMI (kg/m2) WBC (×109/L) LDL-cholesterol (mg/dL) Log(hsCRP) (mg/L) HOMA-IR eGFR (mL/min per 1.73 m2) TIBC (μg/dL) Smoking status Alcohol intake Regular exercise Type 2 diabetes mellitus
Unadjusted
Model 1
Model 2
1.00 (reference) 1.02 (0.86–1.22) 1.16 (0.98–1.38) 1.33 (1.12–1.57) b0.001
1.00 (reference) 1.08 (0.90–1.29) 1.21 (1.01–1.44) 1.28 (1.07–1.52) 0.003 1.06 (1.04–1.07) 1.10 (1.06–1.14) 1.03 (0.97–1.08) 1.00 (0.99–1.00) 0.98 (0.91–1.06) 1.14 (1.03–1.26) 1.00 (0.99–1.01) 1.01 (1.00–1.01)
1.00 (reference) 1.09 (0.91–1.30) 1.21 (1.01–1.45) 1.28 (1.07–1.52) 0.003 1.05 (1.04–1.07) 1.10 (1.06–1.15) 1.04 (0.99–1.10) 1.00 (1.00–1.01) 0.99 (0.91–1.07) 1.14 (1.03–1.26) 1.00 (0.99–1.01) 1.01 (1.00–1.01) 1.12 (0.96–1.30) 1.24 (0.97–1.58) 0.92 (0.72–1.17) 1.58 (0.99–2.52)
Model 1 was adjusted for age, BMI, WBC, LDL-cholesterol, log(hsCRP), HOMA-IR, eGFR and TIBC. Model 2 was adjusted for model 1 plus smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus.
Nevertheless, the underlying mechanism by which elevated ferritin level contributes to the development of HTN is unclear. Possible mechanisms include atherosclerosis by iron itself, insulin resistance, and a fatty liver-mediated process. Iron overload increases vascular oxidative stress and impairs vasoreactivity [15]. Moreover, atherosclerosis can develop and accelerate due to iron promoting direct endothelial toxicity and oxidizing low-density lipoprotein and isoprostanes [16–18]. In addition, increased total body iron may result in increased rates of adipocyte lipolysis and increased circulating free-fatty acid levels and insulin resistance [19]. Iron is intimately related with oxidative stress and participates in the formation of highly toxic free radicals, such as hydroxide and the superoxide anion [20]. Oxidative stress affected insulin signaling at the cellular level [21]. Hyperinsulinemia and insulin resistance are associated with HTN [22–25]. Thus, incident HTN and serum ferritin level may be mediated by insulin resistance. Insulin resistance induced by iron overload could Table 3 Hazard ratios (HRs) and 95% confidence intervals (CIs) for the incidence of hypertension according to log-transformed serum ferritin levels in propensity score matched participants (N = 7104). Hazard ratios (95% confidence interval)
Log(ferritin) (ng/mL) Age (years) BMI (kg/m2) WBC (×109/L) LDL-cholesterol (mg/dL) Log(hsCRP) (mg/L) HOMA-IR eGFR (mL/min per 1.73 m2) TIBC (μg/dL) Smoking status Alcohol intake Regular exercise Type 2 diabetes mellitus
Unadjusted
Model 1
Model 2
1.18 (1.06–1.32)
1.15 (1.02–1.29) 1.06 (1.04–1.07) 1.10 (1.06–1.14) 1.03 (0.97–1.08) 1.00 (0.99–1.01) 0.98 (0.91–1.06) 1.14 (1.03–1.25) 1.00 (0.99–1.01)
1.15 (1.02–1.29) 1.05 (1.04–1.07) 1.10 (1.06–1.14) 1.04 (0.99–1.10) 1.00 (1.00–1.01) 0.99 (0.91–1.07) 1.14 (1.03–1.26) 1.00 (0.99–1.01)
1.01 (1.01–1.01)
1.01 (1.00–1.01) 1.11 (0.96–1.29) 1.23 (0.97–1.57) 0.92 (0.72–1.17) 1.58 (0.99–2.52)
Model 1 was adjusted for age, BMI, WBC, LDL-cholesterol, log(hsCRP), HOMA-IR, eGFR and TIBC. Model 2 was adjusted for model 1 plus smoking status, alcohol intake, regular exercise and type 2 diabetes mellitus.
be a causative factor in vascular endothelial dysfunction [26]. Hyperinsulinemia may also evoke sympathetic overactivity by stimulating noradrenaline release, particularly in the skeletal muscle [27]. Chronically elevated sympathetic tone plays a significant role in increasing BP [28]. In addition, hyperinsulinemia can affect the renovascular system, which promotes renal sodium and water retention by increasing tubular reabsorption of sodium [29–31], which can lead to HTN. Several studies have suggested that insulin could interfere with the systemic rennin–angiotensin system (RAS) and intrarenal RAS that might contribute to the development of HTN [32–34]. Taken together, not only the various effects of ferritin itself but insulin resistance and liver-mediated effects of ferritin on the vascular endothelial, endocrine, and autonomic nervous systems might contribute to the development of HTN. Several limitations of the present study should be mentioned. First, our study population was comprised of only Korean men. Serum ferritin levels are generally lower in females due to menstrual iron loss [35]. Therefore, it is unclear whether our findings can be extrapolated to females and other ethnicities. Second, as our study was epidemiological, it was insufficient to clarify the causative relationship between serum ferritin levels and the development of HTN. Therefore, clinical and laboratory studies should be performed to identify the causative relationship. Third, blood pressure was measured only on the day of the health check-up without follow-up or continuous monitoring. This may have caused variability in the measured blood pressure and influenced the proportion of incident HTN. In particular, overestimating hypertension such as white coat hypertension could be a major limitation of this study. However, we think that the proportion of overestimated HTN cases was minimal and did not affect the results. Although white coat hypertension can occur in anyone, it is more prevalent in females and elderly ≥ 60 yr [36–39]. However, all of our study participants were middle-aged Korean men. Accordingly, considering the characteristics of our study population, the effect of white coat hypertension was not substantial. In conclusion, elevated serum ferritin level was independently associated with incident HTN in Korean men. The risk of incident HTN was proportional to the baseline serum ferritin level. These findings suggest the value of elevated serum ferritin level as an early predictor of the development of HTN.
262
J.-H. Ryoo et al. / International Journal of Cardiology 183 (2015) 258–262
Conflict of interest No potential conflicts of interest relevant to this article were reported. Author contribution Sung Keun Park collected, analyzed and interpreted the data. Sung Keun Park is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Jae-Hong Ryoo coordinated the study and wrote the manuscript as a first author. Sun Yong Kim contributed to writing and editing the manuscript. Chang-Mo Oh participated in the revising manuscript and statistical analysis. Eugene Kim, Sae-Jin Park, Jae In Yu, Min-Gi Kim and Yong-Sung Choi participated in reviewing and editing manuscript. Taeg-Su Ko had a significant role in revising and editing manuscript. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2014.10.152. References [1] M. Ezzati, A.D. Lopez, A. Rodgers, S. Vander Hoorn, C.J. Murray, Selected major risk factors and global and regional burden of disease, Lancet 360 (2002) 1347–1360. [2] P.M. Kearney, M. Whelton, K. Reynolds, P. Muntner, P.K. Whelton, J. He, Global burden of hypertension: analysis of worldwide data, Lancet 365 (2005) 217–223. [3] A.L. Heath, S.J. Fairweather-Tait, Health implications of iron overload: the role of diet and genotype, Nutr. Rev. 61 (2003) 45–62. [4] C.A. Whittington, K.V. Kowdley, Review article: haemochromatosis, Aliment. Pharmacol. Ther. 16 (2002) 1963–1975. [5] A. Piperno, P. Trombini, M. Gelosa, V. Mauri, V. Pecci, A. Vergani, et al., Increased serum ferritin is common in men with essential hypertension, J. Hypertens. 20 (2002) 1513–1518. [6] B. Choi, K.J. Yeum, S.J. Park, K.N. Kim, N.S. Joo, Elevated serum ferritin and mercury concentrations are associated with hypertension; analysis of the fourth and fifth Korea national health and nutrition examination survey (KNHANES IV-2, 3, 2008– 2009 and V-1, 2010), Environ. Toxicol. (2013). [7] M.K. Kim, K.H. Baek, K.H. Song, M.I. Kang, J.H. Choi, J.C. Bae, et al., Increased serum ferritin predicts the development of hypertension among middle-aged men, Am. J. Hypertens. 25 (2012) 492–497. [8] S.K. Park, M.H. Seo, H.C. Shin, J.H. Ryoo, Clinical availability of nonalcoholic fatty liver disease as an early predictor of type 2 diabetes mellitus in Korean men: 5-year prospective cohort study, Hepatology 57 (2013) 1378–1383. [9] A.V. Chobanian, G.L. Bakris, H.R. Black, W.C. Cushman, L.A. Green, J.L. Izzo Jr., et al., Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure, Hypertension 42 (2003) 1206–1252. [10] A. Dominguez-Rodriguez, M. Carrillo-Perez Tome, C. Hernandez-Garcia, E. ArroyoUcar, R. Juarez-Prera, G. Blanco-Palacios, et al., Serum ferritin and acute coronary syndrome: a strong prognostic factor? Int. J. Cardiol. 152 (2011) 129–130. [11] C. Gabay, I. Kushner, Acute-phase proteins and other systemic responses to inflammation, N. Engl. J. Med. 340 (1999) 448–454. [12] S.K. Park, J.H. Ryoo, M.G. Kim, J.Y. Shin, Association of serum ferritin and the development of metabolic syndrome in middle-aged Korean men: a 5-year follow-up study, Diabetes Care 35 (2012) 2521–2526. [13] S. Franchini, L. Dagna, F. Salvo, P. Aiello, E. Baldissera, M.G. Sabbadini, Adult onset Still's disease: clinical presentation in a large cohort of Italian patients, Clin. Exp. Rheumatol. 28 (2010) 41–48. [14] X.D. Kong, D. Xu, W. Zhang, Y. Zhao, X. Zeng, F. Zhang, Clinical features and prognosis in adult-onset Still's disease: a study of 104 cases, Clin. Rheumatol. 29 (2010) 1015–1019. [15] S.M. Day, D. Duquaine, L.V. Mundada, R.G. Menon, B.V. Khan, S. Rajagopalan, et al., Chronic iron administration increases vascular oxidative stress and accelerates arterial thrombosis, Circulation 107 (2003) 2601–2606.
[16] B.L. Crist, D.L. Alekel, L.M. Ritland, L.N. Hanson, U. Genschel, M.B. Reddy, Association of oxidative stress, iron, and centralized fat mass in healthy postmenopausal women, J. Women's Health (Larchmt) 18 (2009) 795–801. [17] J.L. Cracowski, P. Devillier, T. Durand, F. Stanke-Labesque, G. Bessard, Vascular biology of the isoprostanes, J. Vasc. Res. 38 (2001) 93–103. [18] S.J. Duffy, E.S. Biegelsen, M. Holbrook, J.D. Russell, N. Gokce, J.F. Keaney Jr., et al., Iron chelation improves endothelial function in patients with coronary artery disease, Circulation 103 (2001) 2799–2804. [19] A. Green, R. Basile, J.M. Rumberger, Transferrin and iron induce insulin resistance of glucose transport in adipocytes, Metab. Clin. Exp. 55 (2006) 1042–1045. [20] J.M. Fernandez-Real, A. Lopez-Bermejo, W. Ricart, Cross-talk between iron metabolism and diabetes, Diabetes 51 (2002) 2348–2354. [21] J.L. Rains, S.K. Jain, Oxidative stress, insulin signaling, and diabetes, Free Radic. Biol. Med. 50 (2011) 567–575. [22] B.M. Cheung, N.M. Wat, Y.B. Man, S. Tam, C.H. Cheng, G.M. Leung, et al., Relationship between the metabolic syndrome and the development of hypertension in the Hong Kong Cardiovascular Risk Factor Prevalence Study-2 (CRISPS2), Am. J. Hypertens. 21 (2008) 17–22. [23] N.R. Every, E.J. Boyko, E.M. Keane, J.A. Marshall, M. Rewers, R.F. Hamman, Blood pressure, insulin, and C-peptide levels in San Luis Valley, Colorado, Diabetes Care 16 (1993) 1543–1550. [24] T.A. Manolio, P.J. Savage, G.L. Burke, K.A. Liu, L.E. Wagenknecht, S. Sidney, et al., Association of fasting insulin with blood pressure and lipids in young adults. The CARDIA study, Arteriosclerosis 10 (1990) 430–436. [25] S.K. Park, J.Y. Jung, W.J. Choi, Y.H. Kim, H.S. Kim, W.T. Ham, et al., Elevated fasting serum insulin level predicts future development of hypertension, Int. J. Cardiol. 172 (2014) 450–455. [26] G. Arcaro, A. Cretti, S. Balzano, A. Lechi, M. Muggeo, E. Bonora, et al., Insulin causes endothelial dysfunction in humans: sites and mechanisms, Circulation 105 (2002) 576–582. [27] G. Lembo, R. Napoli, B. Capaldo, V. Rendina, G. Iaccarino, M. Volpe, et al., Abnormal sympathetic overactivity evoked by insulin in the skeletal muscle of patients with essential hypertension, J. Clin. Invest. 90 (1992) 24–29. [28] D. Butlen, S. Vadrot, S. Roseau, F. Morel, Insulin receptors along the rat nephron: [125I] insulin binding in microdissected glomeruli and tubules, Pflugers Arch. — Eur. J. Physiol. 412 (1988) 604–612. [29] R.A. DeFronzo, C.R. Cooke, R. Andres, G.R. Faloona, P.J. Davis, The effect of insulin on renal handling of sodium, potassium, calcium, and phosphate in man, J. Clin. Invest. 55 (1975) 845–855. [30] J.H. Miller, M.D. Bogdonoff, Antidiuresis associated with administration of insulin, J. Appl. Physiol. 6 (1954) 509–512. [31] A. Nizet, P. Lefebvre, J. Crabbe, Control by insulin of sodium potassium and water excretion by the isolated dog kidney, Pflugers Arch. — Eur. J. Physiol. 323 (1971) 11–20. [32] A.P. Rocchini, C. Moorehead, S. DeRemer, T.L. Goodfriend, D.L. Ball, Hyperinsulinemia and the aldosterone and pressor responses to angiotensin II, Hypertension 15 (1990) 861–866. [33] G. Nickenig, J. Roling, K. Strehlow, P. Schnabel, M. Bohm, Insulin induces upregulation of vascular AT1 receptor gene expression by posttranscriptional mechanisms, Circulation 98 (1998) 2453–2460. [34] P.W. Anderson, X.Y. Zhang, J. Tian, J.D. Correale, X.P. Xi, D. Yang, et al., Insulin and angiotensin II are additive in stimulating TGF-beta 1 and matrix mRNAs in mesangial cells, Kidney Int. 50 (1996) 745–753. [35] D.H. Rushton, R. Dover, A.W. Sainsbury, M.J. Norris, J.J. Gilkes, I.D. Ramsay, Why should women have lower reference limits for haemoglobin and ferritin concentrations than men? BMJ 322 (2001) 1355–1357. [36] T.G. Pickering, J.E. Hall, L.J. Appel, B.E. Falkner, J. Graves, M.N. Hill, et al., Recommendations for blood pressure measurement in humans and experimental animals: part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research, Circulation 111 (2005) 697–716. [37] M.G. Myers, R.A. Reeves, White-coat effect in treated hypertensive patients: sex differences, J. Hum. Hypertens. 9 (1995) 729–733. [38] M.B. MacDonald, G.P. Laing, M.P. Wilson, T.W. Wilson, Prevalence and predictors of white-coat response in patients with treated hypertension, Can. Med. Assoc. J. 161 (1999) 265–269. [39] S. Abir-Khalil, S. Zaîmi, M.A. Tazi, S. Bendahmane, O. Bensaoud, M. Benomar, Prevalence and predictors of white-coat hypertension in a large database of ambulatory blood pressure monitoring, East Mediterr. Health J. 15 (2009) 400–407.
