Circulating Hepcidin Is Independently Associated with Systolic Blood Pressure in Apparently Healthy Individuals

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Archives of Medical Research

-

(2015)

-

ORIGINAL ARTICLE

Circulating Hepcidin Is Independently Associated with Systolic Blood Pressure in Apparently Healthy Individuals Q7

Q2 Q1

Milton Fabian Su
arez-Orteg
on,a,b Alejandra Arbel
aez,b,c Mildrey Mosquera,b,c d Jos
e Maria Moreno-Navarrete, Cecilia Aguilar-de Plata,b,c and Jos
e Manuel Fern
andez-Reald

a Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK Nutrition Group, cPhysiological Sciences Department, Universidad del Valle, Cali, Colombia d Department of Diabetes, Endocrinology and Nutrition, Institut d’Investigaci
o Biomedica de Girona (IdIBGi), CIBEROBN (CB06/03/010) and Instituto de Salud Carlos III (ISCIII), Girona, Spain b

Received for publication April 21, 2015; accepted July 31, 2015 (ARCMED-D-15-00261).

Background and Aims. Few studies have described the association between hepcidin levels and cardiometabolic risk in the general population and more so by considering robust adjustment for confounding factors. Therefore, the aim of the present study was to investigate the associations between circulating hepcidin and anthropometric, biochemical and vascular variables related to cardiometabolic risk in healthy individuals adjusting for relevant covariates. Methods. Two-hundred thirty nine individuals (20e65 years old) were included in this cross-sectional study. Outcome variables were fasting glucose, triglycerides, LDL cholesterol, HDL cholesterol, total cholesterol, waist circumference, systolic and diastolic blood pressures, and the Framingham risk score. Multivariate linear regression and ANCOVA analyses including covariates of body mass index (BMI), menopausal status, physical inactivity, alcohol intake, insulin resistance, subclinical/chronic inflammation, ferritin and soluble transferrin receptors were used to describe the associations between hepcidin and cardiometabolic risk markers. Results. In adjusted linear regression analyses, there was no significant association in men. In women, a relationship between hepcidin and triglycerides became significant after adjustments ( p !0.05). By comparing quartiles of hepcidin levels, systolic blood pressure values in men were significantly higher in the upper quartile of hepcidin vs. the rest of quartiles independently of BMI, chronic inflammation, insulin resistance and other iron markers (ANCOVA, p !0.05). There were no significant independent associations with the Framingham risk score (total points). Conclusion. We found a threshold effect of hepcidin levels on systolic blood pressure specifically in men. Further larger studies and experimental research are required to investigate possible mechanisms for the relationship between hepcidin metabolism and vascular function. Ó 2015 IMSS. Published by Elsevier Inc. Key Words: Hepcidin, Blood pressure, Cardiovascular risk, Insulin resistance.

Introduction Address reprint requests to: Milton Fabian Su
arez-Orteg
on, Centre for Population Health Sciences, College of Medicine and Veterinary Medicine, University of Edinburgh, Teviot Place, Edinburgh EH8 9AG, United Kingdom; Phone: þ44 (0) 131 650 3237; FAX: þ44 (0) 131 650 6909 and Jos
e Manuel Fern
andez-Real, Section of Diabetes, Endocrinology and Nutrition, Hospital of Girona ‘‘Dr Josep Trueta’’, Carretera de Franc¸a s/n, 17007 Girona, Spain; Phone: ---; FAX: ---; E-mail: Milton. [email protected] or [email protected]

Because possible differences by gender in iron status were postulated as a potential explanation for difference in cardiovascular risk between men and women, iron parameters have been evaluated as potential modifiable factors in the general population in the last two decades (1,2). Body iron stores measured as circulating ferritin levels have been

0188-4409/$ - see front matter. Copyright Ó 2015 IMSS. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.arcmed.2015.07.007 ARCMED2039_proof ■ 5-8-2015 15-41-12

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122

2

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177

Su
arez-Orteg
on et al./ Archives of Medical Research

found associated with increased risk of type 2 diabetes and metabolic alterations related to cardiovascular risk, but the causal relationship is still inconclusive (3e5). The exploration of new markers of iron metabolism regarding cardiometabolic risk supposes a further step in the characterization of this relationship. The relationship of hepcidin with cardiometabolic risk has been scarcely investigated. Hepcidin is a peptide produced mainly by the liver in response to increased plasma or tissue iron to homeostatically downregulate its absorption by binding and inactivating intestinal ferroportin (6). Given the potential pleiotropic effects of hepcidin (7), it is important to assess the effect of covariates regarding theoretical relationships with variables of cardiometabolic risk. In the limited studies conducted to date, robust multivariate analyses have not been performed and therefore it is not clear whether the associations with cardiometabolic risk would be similar to those observed with ferritin and whether the associations are independent of insulin resistance, inflammation and/or adiposity. Therefore, the aim of the present study was to investigate whether variables related to cardiometabolic risk were associated with circulating hepcidin in healthy individuals after adjusting for important confounding variables. Materials and Methods Subjects The study population consisted of 239 volunteers (20e65 years old) from the staff of a hospital, a university, a governmental health department and a supermarket chain in Cali, Colombia and who responded to advertisements describing our study. In order to obtain healthy subjects and to avoid bias in estimating variables related to cardiometabolic risk and iron parameters, the following exclusion criteria were considered: clinically significant liver diseases, neurologic or endocrine systems, cardiometabolic diseases (hypertension, history of stroke, myocardial infarction, or type 2 diabetes) or other major systemic disease; smoking; blood transfusion or iron therapies during the previous 6 months; long-term multivitamin or vitamin supplements consumption (two or more days/week); hypolipidemic or oral hypoglycemic drugs; current evidence of acute or chronic inflammatory or infective diseases; and history of disturbances in iron balance (e.g., hemosiderosis from any cause, hemolytic anemia, iron deficiency). The Universidad del Valle Research Ethics Committee approved the study and all participants gave written informed consent. Clinical Measurements Blood pressure was measured using digital sphygmomanometers with an appropriately sized cuff in a sitting position after a 15-min rest. The measurement was repeated

-

(2015)

-

178 179 180 181 182 183 184 185 186 187 188 Biochemical Measurements 189 After 8 h fasting, blood was obtained and serum and plasma 190 samples were stored at 80 C until subsequent analyses. 191 Fasting glucose, triglycerides, total cholesterol and high192 density lipoprotein cholesterol (HDL-C) were determined 193 by using enzymatic-colorimetric assays (Biosystems Inc., 194 Spain). Low-density lipoprotein cholesterol (LDL-C) levels 195 were calculated according to the Friedewald equation: total 196 cholesterol-(HDL-C þ (triglycerides/5) (8). Serum ferritin Q3 197 and high sensitivity C-reactive protein (CRP) were 198 measured using turbidimetry (Biosystems Inc.). Fasting in199 sulin was measured using chemiluminiscence. Levels of 200 hepcidin and soluble transferrin receptor (sTfR) were 201 measured using a double monoclonal sandwich enzyme 202 Ò immunoassay (DRG Hepcidin 25 [Bioactive] ELISA 203 [EIA-5258, DRG International, Inc., Mountainside, NJ]; 204 and Human sTfR ELISA [RD194011100, Heidelberg, Ger205 many], respectively). Intra- and interassay coefficients of 206 variation were !5.5%. HOMA-IR (Homeostatic Model 207 Assessment-Insulin Resistance) was calculated as (insulin 208 mU/mL  [glucose mg-dL]/405) (9). 209 210 Framingham Risk Score 211 212 The Framingham risk score was calculated according to the 213 points-based system (10). Sub-total points derived variables 214 of age, total cholesterol and HDL-C and systolic blood 215 pressure were used obtain the total points. In this calcula216 tion, there were no sub-total points derived from smoking 217 because this was an exclusion criterion of the study. Equiv218 alent 10-year risk was derived from the values of total 219 points (10). 220 221 Statistical Analysis 222 223 All analyses were conducted in each gender and the study 224 variables were described as means and standard deviation 225 or median and interquartile range according to distribution 226 of variables. Differences were estimated via Student t test 227 or ManneWhitney U test. Because 12 subjects had values 228 of hepcidin below the detection limit, the value assigned 229 for these cases for descriptive purposes was the same value 230 of detection limit (0.35 mg/mL). Multivariate linear regres231 sion analysis was conducted to evaluate and adjust the asso232 ciations of hepcidin with cardiometabolic risk factors. after 5 min. The mean of the two measurements was used in the statistical analyses. Body weight and height were measured using standard techniques and instruments and body mass index (BMI) was calculated as weight in kg/ height in m2. Waist circumference (WC) was measured from the midpoint between the lateral iliac crest and the lowest rib using a flexible steel tape measure. A survey to record personal data and lifestyle habits were recorded by trained interviewers.

ARCMED2039_proof ■ 5-8-2015 15-41-12

Hepcidin and Cardiometabolic Risk

233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287

Cardiometabolic risk markers consisted of levels of LDL-C, HDL-C, total cholesterol, triglycerides and glucose, and values of WC, and systolic and diastolic blood pressures. Total points from the Framingham risk score was an additional cardiometabolic outcome. We used total points instead of equivalent 10-year risk values because the skewed distribution and discrete continuous nature (no normalization was possible) of the latter did not allow to run parametric analyses. Regression coefficients were described as non-adjusted and adjusted for age, CRP levels, BMI, alcohol intake (no/yes), physical inactivity (as no physical activity per week), HOMA-IR and other iron markers (levels of ferritin and sTfR). Linear regressions were run with and without cases of hepcidin values below detection limit to compare the effect of these cases on significance of the relationships. In order to evaluate nonlinear relationships with cardiometabolic risk markers, we conducted an additional analysis comparing the values of these variables between sex-specific quartiles of hepcidin by using one-way ANOVA followed by post-hoc analysis and ANCOVA to adjust for covariates. Analyses were performed on transformed values for skewed variables: logarithm of sTfR, systolic blood pressure, triglycerides, HOMA-IR, and ferritin, and 1/hs-CRP. To describe values of systolic blood pressure, triglycerides throughout quartiles of hepcidin, we used median and interquartile ranges but ANOVA and ANCOVA were conducted with transformed values as mentioned above; p value !0.05 was considered statistically significant. Analyses were performed using STATA 8.0. Results Clinical and biochemical characteristics of participants are described in Table 1. Men and women were similar in age and BMI, but hepcidin was lower in men. As expected, men had lower levels of HDL-C and higher values of waist circumference, blood pressure, and serum ferritin than women. Women showed higher levels of CRP and higher proportion of alcohol consumption than men. Linear regression analyses in men did not show significant relationships between variables of cardiometabolic risk and hepcidin levels before or after adjusting for covariates (Table 2). In women, although hepcidin was associated with waist circumference, glucose, total cholesterol and LDL-C, these associations were no longer significant after adjustment for covariates. On the other hand, the relationship between hepcidin and triglycerides became significant in women after these adjustments (Table 2). The significance of the associations with hepcidin was not altered by removing individuals with values below the detection limit of hepcidin (Table 2). Analyses for levels of variables of cardiometabolic risk by quartiles of iron markers showed a significant positive association between systolic blood pressure and hepcidin

3

Q4 288 289 Men Women 290 n 122 117 p 291 292 Age (years) 44.9  7.7 46.1  7.7 0.206 293 Menopause (n) 48 294 26.2  3.4 25.9  3.8 0.588 BMI (kg/m2)b Hepcidin (ng/mL) 12.0  5.0 7.8  5.0 !0.001 295 WC (cm)b 85.1  10.3 75.7  8.2 !0.001 296 SBP (mmHg) 120 (118.8e129.1) 109.5 (101.5e120.8) !0.001 297 DBP (mmHg) 75.2  8.8 72.1  11.2 0.017 298 TG (mg/dL) 163 (117.7e228) 105.5 (79.5e152.5) !0.001 299 Glucose (mg/dL) 91.4  10.4 86.2  8.3 !0.001 HDL-C (mg/dL) 44.0  8.8 53.3  11.6 !0.001 300 LDL-C (mg/dL) 114.7  29.3 118.8  30.3 0.285 301 TC (mg/dL) 196.4  32.1 198.1  36.6 0.697 302 hs-CRP (mg/L) 1.4 (1.1e1.9) 1.6 (1.1e3.0) 0.007 303 Insulin (mU/mL) 9.46 (5.99e13.97) 7.49 (5.69e11.81) 0.056 304 HOMA-IR 2.02 (1.26e2.99) 1.47 (1.05e2.47) 0.013 Framingham score 6.1  4.6 7.0  4.8 0.170 305 (total points) 306 Framingham 10 2.5 (1.0e6.0) 0 (0e1.0) !0.001 307 a year % risk Q5 308 Ferritin (mg/L) 181 (128e269) 70 (24e123) !0.001 309 sTfR (mg/mL) 0.42 (0.12e0.63) 0.54 (0.17e0.78) 0.020 No physical activity/ 52 (42.6) 64 (54.7) 0.062 310 week n (%) 311 Alcohol consumption 45 (36.9) 74 (63.2) !0.001 312 n (%) 313 BMI, body mass index; sTfR, soluble transferrin receptor; TG, triglycer314 ides; TC, total cholesterol; WC, waist circumference; SBP, systolic blood 315 pressure; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein 316 cholesterol; LDL-C, low-density lipoprotein cholesterol; hs-CRP, high317 sensitivity C-reactive protein; HOMA-IR, homeostatic model assessment 318 insulin resistance. Data are mean  standard deviation or median (interquartile range). Values 319 in bold are statistically significant. 320 a Values of zero should be interpreted as !1% risk. 321 b Means: Four missed values for BMI (two in men, two in women) and one 322 for waist circumference (in men). 323 324 325 in men: higher SBP values were observed in those subjects 326 in the upper quartile compared with other quartiles of hep327 cidin (Table 3). The significance of the rest of the associa328 tions comparing quartiles of hepcidin were, in general, 329 congruent with those described by linear regression ana330 lyses (Table 3). 331 Framingham risk score (total points) across quartiles of 332 hepcidin in men and women are shown in Figure 1. 333 Although in women total points increased significantly 334 across quartiles of hepcidin, this association did not remain 335 significant after adjustment for covariates. There were no 336 significant differences either in non-adjusted or adjusted 337 models (Figure 1). Adjusted associations with total points 338 from Framingham risk score are shown with age as part 339 of the covariates despite the fact that age is part of the Fra340 mingham risk score. At any rate, we conducted the analysis 341 without adjustment for age and the significance of the asso342 ciations remained unaltered (data not shown). Table 1. Description of the study population

ARCMED2039_proof ■ 5-8-2015 15-41-12

4

343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397

Su
arez-Orteg
on et al./ Archives of Medical Research

-

(2015)

-

Table 2. b-coefficients (95% confidence interval) for relationships between hepcidin and variables of cardiometabolic risk Hepcidin (ng/mL) Adjusteda

Non-adjusted Men WC (cm) Glucose (mg/dL) HDL-C (mg/dL) Log-TG (mg/dL) DBP (mmHg) log-SBP (mmHg) LDL-C (mg/dL) Total-C (mg/dL) Women WC (cm) Glucose (mg/dL) HDL-C (mg/dL) Log-TG (mg/dL) DBP (mmHg) log-SBP (mmHg) LDL-C(mg/dL) TC (mg/dL)

Hepcidin (ng/mL) (excluding cases below detection limit) Adjusteda

Non-adjusted

0.32 0.39 0.10 0.006 0.30 0.006 0.93 0.88

(0.04 to 0.69) (0.02e0.76) (0.41 to 0.21) (0.01 to 0.02) (0.005 to 0.61) (0.00001 to 0.01) (1.97 to 0.09) (2.02 to 0.24)

0.08 0.10 0.02 0.0001 0.22 0.004 0.43 0.41

(0.08 to 0.25) (0.30 to 0.50) (0.30 to 0.36) (0.02 to 0.02) (0.14 to 0.58) (0.003 to 0.01) (1.66 to 0.79) (1.80 to 0.96)

0.38 0.45 0.09 0.009 0.33 0.007 0.74 0.57

(0.004 to 0.77) (0.07e0.84) (0.42 to 0.22) (0.01 to 0.02) (0.01e0.66) (0.0002e0.01) (1.82-0.34) (1.75 to 0.61)

0.05 0.10 0.041 0.001 0.25 0.004 0.009 0.17

(0.13 to 0.23) (0.33 to 0.54) (0.32 to 0.40) (0.01 to 0.02) (0.14 to 0.64) (0.004 to 0.013) (1.31 to 1.33) (1.30 to 1.65)

0.51 0.41 0.10 0.01 0.22 0.003 1.79 1.60

(0.22e0.80) (0.12e0.71) (0.53 to 0.32) (0.007 to 0.02) (0.18 to 0.63) (0.001 to 0.008) (0.73e2.85) (0.28e2.91)

0.04 0.24 0.55 L0.03 0.12 0.00006 0.73 0.48

(0.17 to 0.26) (0.16 to 0.66) (0.05 to 1.16) (L0.05 to L0.006) (0.39 to 0.64) (0.006 to 0.006) (0.94 to 2.41) (1.87 to 2.83)

0.46 0.50 0.09 0.003 0.17 0.003 1.92 1.65

(0.08e0.78) (0.14e0.86) (0.38 to 0.58) (0.01 to 0.02) (0.31 to 0.66) (0.002 to 0.009) (0.66e3.19) (0.07e3.19)

0.02 0.34 0.74 L0.04 0.14 0.001 1.88 0.48

(0.27 to 0.21) (0.136 to 0.82) (0.04 to 1.43) (L0.06 to L0.01) (0.45 to 0.75) (0.006 to 0.009) (0.02 to 3.79) (1.87 to 2.83)

sTfR, soluble transferrin receptor, WC, waist circumference; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; TC, total cholesterol; DBP, diastolic blood pressure; SBP, systolic blood pressure; CRP, high-sensitivity C-reactive protein; HOMA-IR, homeostatic model assessment-insulin resistance. Significant relationships are shown in bold ( p !0.05). a Adjusted for age, menopause (for women), CRP levels, BMI, alcohol intake (no/yes), physical inactivity (no/yes), HOMA-IR, sTfR and ferritin levels. Analyses were performed on transformed values of skewed variables: logarithm of sTfR, ferritin, triglycerides, SBP and HOMA-IR values, and 1/CRP levels.

By removing from the models and exploring bivariate models, the main covariates attenuated the initial significant unadjusted associations found in women were age, menopause and ferritin levels. This was the case for associations of hepcidin with fasting glucose and total cholesterol. The association found between hepcidin and LDL-C was specifically attenuated by adjustment for ferritin levels. In bivariate models, only ferritin and age attenuated the association between hepcidin and total points from the Framingham risk score in women, although the rest of the covariates as a group also attenuated the association. Discussion The present study explored associations between hepcidin levels and variables related to cardiometabolic risk in apparently healthy individuals with no history of cardiovascular disease or diabetes. We found a significant and independent association between circulating hepcidin and systolic blood pressure in men, which was evident once the highest levels of hepcidin were compared with lower levels. In women, there were unadjusted associations of hepcidin with fasting glucose, cholesterol markers and total points from Framingham score, which were mainly influenced by ferritin levels. A limited number of studies have evaluated the associations of cardiometabolic markers with hepcidin, particularly in the general population. Levels of circulating

pro-hepcidin, a residual prohormone generated during the cleavage of signal peptide, were not associated with any metabolic, anthropometric or vascular marker in subjects with normal glucose tolerance, but significant correlations were observed with fasting glucose and fasting triglycerides in individuals with altered glucose tolerance (11). Similarly, no relationship between hepcidin levels and metabolic and vascular parameters related to metabolic syndrome was observed in a group of non-diabetic or type 2 diabetic individuals, and no relationship was mentioned within each group (12). A recent population-based study by Martinelli et al. reported a unique significant association between circulating hepcidin and glucose levels O100 mg/ dL, specifically in women, even after age and ferritin adjustments (7). In our study, circulating hepcidin was positively associated with systolic blood pressure independently of BMI and chronic/subclinical inflammation. Moreover, persistence of the association after adjustment for insulin resistance and other iron markers (sTfR or hepcidin, and ferritin) suggests alternative mechanisms for the association of hepcidin and blood pressure. The association is consistent with the reported relationship between hepcidin and intima media thickness in 420 men (13) because this vascular marker has been correlated previously with systolic blood pressure (14). An association has also been described between hepcidin levels and MCP-1, a chemoattractant protein involved in atherogenesis and with

ARCMED2039_proof ■ 5-8-2015 15-41-12

398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452

Hepcidin and Cardiometabolic Risk

453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507

5

508 509 p for difference 510 Specific across differences 511 Quartiles of hepcidin quartiles between 512 Q1 Q2 Q3 Q4 quartiles Non513 a (0.35e9.13 ng/mL) (9.14e11.55 ng/mL) (11.56e13.88 ng/mL) (13.89e26.84 ng/mL) with p !0.05 adjusted Adjusted 514 Men 515 WC (cm) 84.2  8.1 81.1  11.6 85.6  10.7 88.9  9.5 — 0.0522 0.4299 516 Glucose (mg/dL) 88.3  7.6 91.6  9.9 90.6  13.8 95.1  8.7 — 0.0902 0.3290 517 HDL-C (mg/dL) 44.3  11.0 44.0  8.7 44.7  7.3 42.9  8.0 — 0.8949 0.5863 518 TG (mg/dL) 124.5 (96.0e261.5) 155 (98e218) 185 (142e252) 160.5 (124.0e231.2) — 0.4139 0.9189 519 DBP (mmHg) 73.5  7.6 73.9  8.7 75.0  6.8 78.8  11.0 — 0.0707 0.2228 SBP (mmHg) 117.8 (111.0e126.5) 117.5 (110.5e126.5) 117.0 (110.5e127.0) 128.2 (119.2e136.9) Q4 vs. Q1, 0.0074 0.0324 520 Q2 and Q3 521 LDL-C (mg/dL) 121.9  22.9 109.9  30.5 115.3  28.3 111.4  34.5 — 0.3828 0.7578 522 Total-C (mg/dL) 200.6  28.4 191.1  33.0 199.7  31.8 193.9  35.3 — 0.6010 0.6950 523 524 Q1 Q2 Q3 Q4 525 (0.35e4.06 ng/mL) (4.07e7.79 ng/mL) (7.80e11.35 ng/mL) (11.52e19.19 ng/mL) 526 Women 527 Q4 vs. Q1 0.0041 0.2072 WC (cm) 72.2  7.2 75.9  7.8 74.7  7.8 79.3  7.9 528 Glucose (mg/dL) 84.9  6.5 84.3  6.5 86.3  9.8 88.9  9.3 — 0.0585 0.7147 529 HDL-C (mg/dL) 56.5  12.9 49.9  8.4 52.2  11.0 55.1  12.8 — 0.1651 0.2928 530 TG (mg/dL) 93.0 (63.0e128.5) 126.0 (90.5e194.0) 105 (77e151) 120.0 (90.5e145.0) — 0.5059 0.0215 531 DBP (mmHg) 71.1  10.6 72.0  12.5 71.9  13.2 73.5  8.6 — 0.8956 0.9701 SBP (mmHg) 109.5 (101.5e118.5) 108.0 (101.2e122.0) 107.5 (98.0e124.5) 110.5 (104.7e122.2) — 0.7970 0.4644 532 LDL-C (mg/dL) 107.6  29.7 115.5  28.0 120.9  23.9 131.8  34.8 — 0.0179 0.3703 533 Total-C (mg/dL) 191.2  33.1 193.6  36.5 197.2  32.4 210  42.3 — 0.1750 0.1013 534 Q, quartile; sTfR, soluble transferrin receptor; WC, waist circumference; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein 535 cholesterol, TC, total cholesterol; TG, triglycerides; DBP, diastolic blood pressure; SBP, systolic blood pressure; CRP, high-sensitivity C-reactive protein; 536 HOMA-IR, homeostatic model assessment insulin resistance. 537 Analyses were performed on transformed values of skewed variables: logarithm of triglycerides, SBP values and HOMA-IR values, and 1/CRP levels. Data 538 are mean  standard deviation or median (interquartile range). Difference across quartiles of iron markers was assessed by one-way ANOVA (non-adjusted) and ANCOVA (adjusted models). Specific differences between quartiles with p !0.05 (Tukey post hoc test). Q6 539 a Adjusted for age, menopause (for women), CRP levels, BMI, alcohol intake (no/yes), physical inactivity (no/yes), HOMA-IR, quartiles of sTfR and quartiles 540 of ferritin. 541 542 543 544 cardiometabolic risk have been supported on the basis of vascular damage in terms of plaque formation detected by 545 higher iron status in men (17). A strong positive correlation echo-color Doppler in individuals with components of 546 has been described between ferritin and hepcidin in the metabolic syndrome (15). Moreover, overexpression of 547 general population, implying that hepcidin increases in hepcidin resulted in destabilization of the carotid athero548 response to increased body iron stores in order to downresclerotic plaque in mice by increasing inflammatory 549 gulate iron absorption (18). However, women from our response, intracellular lipid accumulation and oxidative 550 study had higher levels of hepcidin but lower iron stores stress in parallel to iron retention (16). Interestingly, the 551 than men. This observation, together with the fact that association between serum hepcidin and systolic blood 552 the association with systolic blood pressure was indepenpressure was characterized by a threshold effect because 553 dent of ferritin levels, suggests a difference by gender other the relationship became evident when cases with the high554 than iron status. On the other hand, the power of the study est values of hepcidin were compared with subjects in the 555 could have been limited by the fact of a heterogeneous felowest quartile of this marker. This kind of association 556 male group in terms of pre- and postmenopausal women. could indicate that only high levels of hepcidin and blood 557 Larger samples within each of these subgroups are pressure would influence each other, but this pattern de558 required. serves special attention in future research to infer a causal 559 The cross-sectional design of our study does not allow relationship. 560 inference about causal relationships. In addition, data The relationship between hepcidin and systolic blood 561 regarding dietary iron and antioxidants intake were not pressure was gender-specific. In some cases, differences 562 available and these would allow more robust adjustments. by gender in the relationship of iron markers with Table 3. Levels of cardiometabolic risk factors by quartiles of circulating hepcidin in men and women

ARCMED2039_proof ■ 5-8-2015 15-41-13

6

563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617

Su
arez-Orteg
on et al./ Archives of Medical Research

-

(2015)

-

with individual cardiometanbolic risk markers and even with a clustered variable such as Framingham Risk score (total points) but influenced by ferritin levels. There is a need for larger studies with a multivariate approach in the general population to confirm these findings. Further research is required to investigate the possible mechanisms behind these relationships.

Acknowledgments We want to thank to Dr. Sarah Wild for her comments to improve this manuscript. This work was funded by a grant from the Colombian Administrative Department for Development and Science Technology (COLCIENCIAS) (grant 1106-45921521) and partially supported by research grants from the Ministerio de Econom
ıa y Competitividad (PI11-00214 and PI12/02631). CIBEROBN Fisiopatolog
ıa de la Obesidad y Nutrici
on is an initiative from the Instituto de Salud Carlos III from Spain. Conflict of interest: None of the authors report any conflict of interest.

References

Figure 1. Total points from Framingham scoring across sex-specific quartiles of hepcidin. Data are mean and 95% confidence interval. Equivalent 10-year risk for cardiovascular disease is shown for each mean of total points. aNon-adjusted (ANOVA). bAdjusted (ANCOVA) for menopause (if women), CRP levels, BMI, alcohol intake (no/yes), physical inactivity (no/yes), HOMA-IR, quartiles of sTfR and quartiles of ferritin. Ranges of hepcidin for quartiles (Q) in men were Q1 0.35e9.13 ng/mL; Q2 9.14e11.55 ng/mL; Q3 11.56e13.88 ng/mL; Q4 13.89e26.84 ng/mL. Ranges of hepcidin for quartiles (Q) in women were Q1 (0.35e4.06 ng/ mL); Q2 (4.07e7.79 ng/mL); Q3 (7.80e11.35 ng/mL); Q4 (11.52e19.19 ng/mL).

The findings of this exploratory analysis may not be generalized before replication in large, multicenter studies. On the other hand, the strengths of this study include the use of a well-characterized group of participants and use of a broad set of covariates for multivariate analyses. In summary, in men, serum hepcidin levels were positively associated with systolic blood pressure independently of BMI, chronic inflammation, insulin resistance and other iron markers. In women there were associations of hepcidin

1. Vari IS, Balkau B, Kettaneh A, et al. Ferritin and transferrin are associated with metabolic syndrome abnormalities and their change over time in a general population: Data from an Epidemiological Study on the Insulin Resistance Syndrome (DESIR). Diabetes Care 2007; 30:1795e1801. 2. Chern JP, Lin KH, Lu MY, et al. Abnormal glucose tolerance in transfusion-dependent beta-thalassemic patients. Diabetes Care 2001;24:850e854. 3. Wrede CE, Buettner R, Bollheimer LC, et al. Association between serum ferritin and the insulin resistance syndrome in a representative population. Eur J Endocrinol 2006;154:333e340. 4. Lee BK, Kim Y, Kim YI. Association of serum ferritin with metabolic syndrome and diabetes mellitus in the South Korean general population according to the Korean National Health and Nutrition Examination Survey 2008. Metabolism 2011;60:1416e1424. 5. Forouhi NG, Harding AH, Allison M, et al. Elevated serum ferritin levels predict new-onset type 2 diabetes: results from the EPICNorfolk prospective study. Diabetologia 2007;50:949e956. 6. Ganz T, Nemeth E. Iron imports. IV. Hepcidin and regulation of body iron metabolism. Am J Physiol Gastrointest Liver Physiol 2006;290: G199eG203. 7. Martinelli N, Traglia M, Campostrini N, et al. Increased serum hepcidin levels in subjects with the metabolic syndrome: a population study. PLoS One 2012;7:e48250. 8. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem 1972;18:499e502. 9. Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412e419. 10. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285:2486e2497. 11. Fern
andez-Real JM, Equitani F, Moreno JM, et al. Study of circulating prohepcidin in association with insulin sensitivity and changing iron stores. J Clin Endocrinol Metab 2009;94:982e988.

ARCMED2039_proof ■ 5-8-2015 15-41-14

618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672

Hepcidin and Cardiometabolic Risk 673 674 675 676 677 678 679 680 681 682 683

12. Jiang F, Sun ZZ, Tang YT, et al. Hepcidin expression and iron parameters change in type 2 diabetic patients. Diabetes Res Clin Pract 2011; 93:43e48. 13. Galesloot TE, Holewijn S, Kiemeney LA, et al. Serum hepcidin is associated with presence of plaque in postmenopausal women of a general population. Arterioscler Thromb Vasc Biol 2014;34:446e456. 14. Sander D, Kukla C, Klingelh€ofer J, et al. Relationship between circadian blood pressure patterns and progression of early carotid atherosclerosis: a 3-year follow-up study. Circulation 2000;26:1536e1541. 15. Valenti L, Dongiovanni P, Motta BM, et al. Serum hepcidin and macrophage iron correlate with MCP-1 release and vascular damage

7

in patients with metabolic syndrome alterations. Arterioscler Thromb Vasc Biol 2011;31:683e690. 16. Li JJ, Meng X, Si HP, et al. Hepcidin destabilizes atherosclerotic plaque via overactivating macrophages after erythrophagocytosis. Arterioscler Thromb Vasc Biol 2012;32:1158e1166. 17. Pham NM, Nanri A, Yi S, Kurotani K, et al. Serum ferritin is associated with markers of insulin resistance in Japanese men but not in women. Metabolism 2013;62:561e567. 18. Galesloot TE, Vermeulen SH, Geurts-Moespot AJ, et al. Serum hepcidin: reference ranges and biochemical correlates in the general population. Blood 2011;117:e218ee225.

ARCMED2039_proof ■ 5-8-2015 15-41-14

684 685 686 687 688 689 690 691 692 693 694