ARTICLES nature publishing group EPIDEMIOLOGY Decreased Serum Hepcidin and Improved Functional Iron Status 6 Months After Restrictive Bariatric Surgery Lisa M. Tussing-Humphreys1, Elizabeta Nemeth2, Giamila Fantuzzi3, Sally Freels4, Ai-xuan L. Holterman5, Carlos Galvani6, Subhashini Ayloo6, Joseph Vitello6 and Carol Braunschweig3 Excess adiposity is associated with low-grade inflammation and decreased iron status. Iron depletion in obesity is thought to be mediated by an inflammation-induced increase in the body’s main regulator of iron homeostasis, hepcidin. Elevated hepcidin can result in iron depletion as it prevents the release of dietary iron absorbed into the enterocytes, limiting replenishment of body iron losses. Weight reduction is associated with decreased inflammation; however, the impact of reduced inflammation on iron status and systemic hepcidin in obese individuals remains unknown. We determined prospectively the impact of weight loss on iron status parameters, serum hepcidin, inflammation, and dietary iron in 20 obese premenopausal females 6 months after restrictive bariatric surgery. At baseline, the presence of iron depletion was high with 45% of the women having serum transferrin receptor (sTfR) >28.1 nmol/l. Differences between baseline and 6 months after surgery for BMI (47.56 vs. 39.55 kg/m2; P < 0.0001), C-reactive protein (CRP) (10.83 vs. 5.71 mg/l; P < 0.0001), sTfR (29.97 vs. 23.08 nmol/l; P = 0.001), and serum hepcidin (111.25 vs. 31.35 ng/ml; P < 0.0001) were significantly lower, whereas hemoglobin (Hb) (12.10 vs. 13.30 g/dl; P < 0.0001) and hematocrit (Hct) (36.58 vs. 38.78%; P = 0.001) were significantly higher. Ferritin and transferrin saturation (Tsat) showed minimal improvement at follow-up. At baseline, hepcidin was not correlated with sTfR (r = 0.02); however, at follow-up, significant correlations were found (r = −0.58). Change in interleukin-6 (IL-6) from baseline was marginally associated with decreased log serum hepcidin (Δ IL-6: β = −0.22; P = 0.15), whereas change in BMI or weight was not. No significant difference in dietary iron was noted after surgery. Weight loss in obese premenopausal women is associated with reduced serum hepcidin and inflammation. Reduction in inflammation and hepcidin likely allow for enhanced dietary iron absorption resulting in an improved functional iron profile. Obesity (2010) 18, 2010–2016. doi:10.1038/oby.2009.490 INTRODUCTION Obesity is associated with low-grade inlammation hallmarked by elevated acute-phase proteins including C-reactive protein (CRP) and interleukin-6 (IL-6), and an increased prevalence of iron deiciency (ID) (1–6). he iron depletion present in obesity is thought to manifest due to inlammation-mediated dysregulation of systemic iron metabolism (6–8). Hepcidin, the body’s main regulator of systemic iron homeostasis, is simultaneously regulated by inlammation (increases expression), elevated body iron levels (increases expression), and hypoxia/erythropoiesis (both decrease expression) (9,10). he strength of the opposing signals is likely what determines the expression/suppression of the protein (11). Acutely, hepcidin acts by degrading the iron exporter ferroportin-1 ultimately blocking iron low into plasma from iron-recycling macrophages and iron-absorbing enterocytes, reducing iron bioavailability; both characteristic of the anemia of inlammation (12). In chronic conditions, where inlammation is let untreated, anemia of inlammation and ID can coexist (13). In this scenario, elevated hepcidin over time can result in ID as daily iron losses exceed dietary iron repletion. he impact of chronically elevated hepcidin is well described by the genetic condition of iron-refractory ID anemia. Iron-refractory ID anemia patients sufer from a defect in the TMPRSS6 gene (encodes a protease important for hepcidin suppression), resulting in chronically elevated hepcidin, depleted iron stores, and impaired dietary iron repletion (14–16). Recently, we demonstrated that the geometric mean (95% conidence 1 United States Department of Agriculture-Agricultural Research Service, Southern Regional Research Center, Baton Rouge, Louisiana, USA; 2Department of Medicine, David Geffen School of Medicine, University of California, Los Angeles (UCLA), Los Angeles, California, USA; 3Department of Kinesiology and Nutrition, Applied Health Sciences, University of Illinois at Chicago, Chicago, Illinois, USA; 4Division of Epidemiology and Biostatistics, School of Public Health, University of Illinois at Chicago, Chicago, Illinois, USA; 5Department of Pediatric Surgery, College of Medicine, Rush University Medical Center, Chicago, Illinois, USA; 6Department of Surgery, College of Medicine, University of Illinois at Chicago, Chicago, Illinois, USA. Correspondence: Carol Braunschweig (braunsch@uic.edu) Received 8 June 2009; accepted 9 December 2009; published online 14 January 2010. doi:10.1038/oby.2009.490 2010 VOLUME 18 NUMBER 10 | OCTOBER 2010 | www.obesityjournal.org ARTICLES EPIDEMIOLOGY interval) for serum hepcidin was signiicantly higher in obese compared to nonobese, hemoglobin (Hb)-matched premenopausal women (88.02 ng/ml (53.70–142.03) vs. 9.70 ng/ml (5.24–16.92); P < 0.0001) (17). he obese women were highly inlamed, hypoferremic, had elevated serum transferrin receptor (sTfR) and minimal iron accumulation in both liver and adipose-associated reticuloendothelial cells, suggesting that acute impairment of iron mobilization from stores was unlikely. he pathology and magnitude of hepcidin (range: 10.4–346.1 ng/ml) observed in the obese women was similar to individuals with long-standing untreated chronic disease (E. Nemeth, unpublished data) and iron-refractory ID anemia (14,15,18). Weight reduction has been shown to improve the inlammatory status of obese individuals (19–21). Recently, Anty et al. (8) reported that obese women with higher levels of inlammation had greater iron depletion based on transferrin saturation (Tsat) <20%. Following signiicant weight loss, 6 months ater bariatric surgery, decrease in inlammation and improvement in Tsat occurred. Interpretation of these indings is limited because although Tsat is commonly used as a marker of iron status, it is negatively impacted by inlammation (22). sTfR is not inluenced by acute or chronic inlammation (23); levels are proportional to cellular expression of the membraneassociated transferrin receptor-1 protein and relect the cells needed for iron or rate of erythropoiesis (24,25). Increased sTfR levels relect high-cellular iron need and occur in “true” iron-deicient states; reduced sTfR levels relect low-cellular iron need and an iron-suicient status. he inluence of weight loss on change in sTfR has not been reported. he aim of this study was to investigate the impact of weight loss on iron status assessed by sTfR, serum hepcidin, and inlammation in a group of obese premenopausal women who underwent a restrictive bariatric surgical procedure. We hypothesized that serum hepcidin and markers for inlammation would decrease and iron status would normalize proportionate to the weight loss achieved 6 months ater surgery. METHODS AND PROCEDURES Obese women (BMI >37.0 kg/m2) evaluated and electing to undergo bariatric surgery (limited to restrictive procedures: gastric banding or sleeve gastrectomy) were recruited from the University of Illinois at Chicago bariatric surgery clinics between December 2007 and July 2008. Premenopausal women undergoing laparoscopic gastric banding or gastric sleeve gastroplasty were selected for our obese population because they manifest several characteristics that allow optimal testing of our speciic aims. Speciically, (i) a restrictive bariatric surgery does not involve the small bowel or induce signiicant nutrient malabsorption, and (ii) these women will experience signiicant weight loss in 6 months, enabling discernment of the impact of change in weight on the parameters of interest. Women were excluded if they reported signiicant medical conditions that could inluence iron or inlammatory status (i.e., cancer, HIV/AIDS, inlammatory bowel disease, gastrointestinal bleeding, and rheumatoid arthritis), had weight loss >3% in the past 3 months, were void of at least one menstrual cycle in the past 12 months, had full or partial hysterectomy, donated blood in the past 3 months, were pregnant or gave birth within the past year, had hemochromatosis or Tsat >45%, or consumed excessive amounts of alcohol (>50 g/day). OBESITY | VOLUME 18 NUMBER 10 | OCTOBER 2010 Baseline and 6-month follow-up data collection Baseline data collection occurred prior to subjects starting a required presurgery liquid diet (typically 10–21 days prior to the bariatric procedure) because this severely restrictive diet could afect iron status or hepcidin levels. For both data collection time points, all blood samples were obtained following a minimum of an 8-h overnight fast; dietary iron supplements, vitamins containing iron, or nonsteroidal antiinlammatory drugs were not consumed 48 h prior to each research appointment to eliminate their possible acute efects on inlammation or hepcidin production. Subjects who reported a cold, lu, or urinary tract infection within the preceding 2 weeks were rescheduled for a later date. he research protocol was approved by the University of Illinois at Chicago Institutional Review Board, and participants provided written consent prior to study entry. Subject characteristics Demographic, social, and health history data were collected via selfreport questionnaire and included information on marital status, race, household income, education level, family size, current health status, reproductive history, menstrual status, and disease prevalence. Social variables collected included alcohol consumption and cigarette use. Disease prevalence was deined as follows: obstructive sleep apnea if continuous positive airway pressure use, type 2 diabetes if taking blood glucose–lowering medications, menstrual irregularities (ibroids and polycystic ovary syndrome) based on self-report, hypertension if taking antihypertensive medications, high cholesterol if taking statins, and osteoarthritis based on self-report. Anthropometrics Subjects were weighed to the nearest 0.1 kg in minimal clothing using a digital scale (Tanita BWB-800AS; Tanita, Arlington Heights, IL). Height was measured to the nearest 0.1 mm using a ixed stadiometer (Health o Meter; Sunbeam Products, Alsip, IL), and waist circumference was measured using a lexible tape, to the nearest 0.1 mm, at the umbilicus. Dietary and physical activity assessment Usual dietary iron intake was assessed using the Block Brief 2000 food frequency questionnaire. he questionnaire is self-administered, contains 70 items and is designed to provide estimates of usual and customary dietary intake over the past 12 months (26). Physical activity was assessed using the Kaiser Physical Activity Survey (27). his survey is a self-administered questionnaire based on the Baecke usual physical activity survey and has been validated in women. he last section of the questionnaires asks participants to list three recreational activities they have engaged in most frequently during the past year, along with the frequency and duration of these activities. Activities recorded were assigned a metabolic equivalent value using the standard Compendium of Physical Activities and tallied together resulting in metabolic equivalents per week for each individual (28). Laboratory assays All assays were performed on fasted blood samples. For both Hb and hematocrit (Hct), blood was obtained via ingerstick puncture. Iron, inflammatory, erythropoietic, and metabolic parameters Hb was measured by hemoglobinometer (STAT-Site M Hgb; Stanbio Laboratory, Boerne, TX). Hct was measured using a microcapillary reader (International Microcapillary Reader; International Equipment, Needham Heights, MA) following 3 min of centrifugation. Serum iron, Tsat, serum ferritin, high-sensitivity CRP, glucose, and insulin were performed by Specialty Laboratories (Valencia, CA), respectively. Serum iron was measured by the ferrozine method. Serum iron <50 µg/dl (normal range 50–170 µg/dl) indicated ID 2011 ARTICLES EPIDEMIOLOGY based on reference ranges provided by Specialty Laboratories. Ferritin was measured by chemiluminescence and values <10 ng/ml (normal range for premenopausal women 10–282 ng/ml) were consistent with ID based on laboratory cut-points. Tsat was calculated as iron/total iron-binding capacity × 100; values <20% are consistent with ID based on laboratory cut-points. he analysis of CRP was by immunoturbidity (reference interval <1.0 mg/l), insulin by chemiluminescence (reference range: 3.0–28.0 mU/l), and glucose by hexokinase end-point spectrophotometry (reference range: 74–106 mg/dl). Insulin resistance was determined by the homeostasis model assessment of insulin resistance according to the following formula: ((glucose (mg/dl)/18) × insulin (mU/liter))/22.5. sTfR was measured by Quantikine IVD immunoassay (R&D Systems, Minneapolis, MN). he manufacturer’s expected reference range for this assay is 8.7–28.1 nmol/l with a value >28.1 nmol/l indicative of ID per the manufacturer’s recommendations. his cut-point was used to dichotomize women as iron depleted (sTfR >28.1 nmol/l) or iron suicient (sTfR ≤28.1 nmol/l) for analysis. he manufacturer states that sTfR values for African Americans are higher than those of non-African descent. IL-6 was measured by Quantikine quantitative sandwich enzyme immunoassay (R&D Systems) with an expected reference range of 0.447–9.96 pg/ml for this assay. Erythropoietin was measured by immunoassay (MD Biosciences, St Paul, MN) with an expected reference range of 4.3–32.9 mU/ml in serum. At follow-up, fewer women self-reported hypertension and menstrual irregularity compared to baseline. he baseline and 6-month postsurgery anthropometric, biochemical, diet, and physical activity characteristics are presented in Table 1. he prevalence of ID (i.e., sTfR >28.10 nmol/l) at baseline was 45% (n = 9); at 6-month follow-up, this had decreased to 15% (n = 3). Signiicant changes included decrease in weight, BMI, waist circumference, CRP, IL-6, serum hepcidin, and sTfR. No signiicant diference was observed for Tsat, ferritin, physical activity, or total dietary iron intake, although women acquired more of their dietary iron in the supplemental form ater surgery. Spearman correlations between serum hepcidin, inlammation, anthropometric measures, and iron status were assessed at baseline and 6 months ater surgery. At baseline, serum hepcidin was signiicantly correlated with ferritin (r = 0.88; P < 0.0001) and was not correlated with sTfR (r = 0.02; P = 0.94), serum iron (r = −0.07; P = 0.78), and Tsat (r = 0.07; P = 0.77). Although not signiicant, BMI was moderately associated with serum hepcidin at baseline (r = 0.33; P = 0.15). At follow-up, Serum hepcidin Serum hepcidin was assessed using a competitive enzyme-linked immunosorbent assay developed by Intrinsic LifeSciences, La Jolla, CA. Detailed methods and performance of this assay were recently published (29). he sensitivity for this assay is 0.5 ng/ml, and intra-assay coeicient of variation was 5–19% and median interassay coeicient of variation was 12%. For women with normal iron status, the 5–95% range for this assay is 17–286 ng/ml. Table 1 Baseline and 6-month follow-up clinical characteristics of obese premenopausal women who underwent a restrictive bariatric procedure Statistical analysis Descriptive statistics included mean, standard deviation, median, and interquartile range (IQR) for continuous variables, and frequency for categorical variables. Non-normally distributed variables were log transformed to achieve normality prior to analysis. Diference between baseline and follow-up for normally distributed continuous variables was assessed using paired t-tests. For non-normally distributed variables, a change score was created (e.g., baseline − follow-up = Δ score) and tested using the Wilcoxon signed-rank test. Relationships between anthropometrics, iron status, serum hepcidin, inlammation, diet, and physical activity variables were assessed using Spearman’s correlation coeicients at baseline and 6 months ater surgery. Multivariable linear regression models were used to assess the impact of change in body mass, inlammation, and serum hepcidin from baseline on 6-month postsurgery iron status (sTfR), inlammation, and serum hepcidin as the dependent variables. Finally, a post hoc analysis of biochemical and anthropometric diferences were assessed in women classiied as iron depleted (>28.1 nmol/l) or iron suicient (≤28.1 nmol/l) at baseline and follow-up using independent t-tests and Wilcoxon rank-sum. All P values are two-sided, and the statistical signiicance level was set at α = 0.05. All analyses were performed using SAS (version 9.1, 2002; SAS Institute, Cary, NC). RESULTS Twenty obese women with a mean age of 35.53 (± 7.27) years were studied. Fity percent of the women were African Americans (n = 10), 45% Caucasian (n = 9), and 5% Hispanic (n = 1). At baseline, 25% (n = 5) self-reported obstructive sleep apnea, 60% (n = 12) osteoarthritis, 40% (n = 8) hypertension, 10% (n = 2) type 2 diabetes, and 40% (n = 8) menstrual irregularities. 2012 Variables Baseline (n = 20)a 6-month Follow-up (n = 20)a P valueb BMI (kg/m2) 47.56 (± 7.92) 39.55 (± 7.55) Weight (kg) 130.01 (± 23.46) 106.75 (± 26.34) Waist circumference (cm) 129.01 (IQR 25.30) 116.65 (IQR 21.50) Hemoglobin (g/dl) 12.10 (± 1.29) 13.30 (± 1.22) Hematocrit (%) 36.58 (± 2.87) 38.78 (± 3.09) <0.0001 <0.0001 <0.0001 <0.0001 0.001 Serum ferritin (ng/ml) 28.0 (IQR 27.00) 25.0 (IQR 21.00) 0.81 Serum iron (µg/ml) 51.0 (IQR 24.00) 57.0 (IQR 33.00) 0.20 Transferrin saturation (%) 13.0 (IQR 16.00) 16.5 (IQR 17.50) 0.32 Serum transferrin receptor (nmol/l) Serum hepcidin (ng/ml) HOMAIR 29.97 (IQR 11.00) 111.25 (IQR 132.00) 3.83 (IQR 2.60) 23.08 (IQR 3.68) 31.35 (IQR 37.70) 1.80 (3.25) 0.001 <0.0001 0.001 CRP (mg/l) 10.83 (IQR 9.98) 5.71 (IQR 8.21) <0.0001 IL-6 (pg/ml) 2.90 (IQR 2.06) 1.78 (IQR 1.71) 0.01 Erythropoietin (mU/ml) 21.93 (IQR 7.16) 14.58 (IQR 6.34) 0.001 Total dietary iron (mg) 19.75 (IQR 14.60) 21.2 (IQR 13.60) 0.10 4.41 (IQR 10.21) 1.49 (IQR 7.420) 0.53 Metabolic equivalents/week CRP, C-reactive protein; HOMAIR, homeostasis model assessment of insulin resistance; IL-6, interleukin-6. Plus-minus values are means ± s.d.; median (interquartile range (IQR)). bP values based on paired t-test or Wilcoxon signed-rank as appropriate; unadjusted values. a VOLUME 18 NUMBER 10 | OCTOBER 2010 | www.obesityjournal.org ARTICLES EPIDEMIOLOGY serum hepcidin was signiicantly correlated with ferritin (r = 0.66; P = 0.003), sTfR (r = −0.58; P = 0.01), and Hct (r = 0.45; P = 0.05). Although not signiicant, correlations became stronger between serum hepcidin and Tsat (r = 0.31; P = 0.18) as well as serum iron (r = 0.25; P = 0.29) at follow-up. Correlations between serum hepcidin and IL-6 (r = 0.10; P = 0.64 vs. r = 0.002; P = 0.99) and CRP (r = 0.23; P = 0.34 vs. r = 0.09; P = 0.73) declined between baseline and 6-month follow-up. BMI was weakly correlated with serum hepcidin at follow-up (r = −0.12; P = 0.59). Notably, the strong positive correlation between sTfR and BMI observed at baseline, diminished at follow-up (r = 0.59; P = 0.01 vs. r = 0.22; P = 0.39), whereas CRP remained signiicantly inversely correlated with Tsat at both baseline and 6 months ater surgery (r = −0.64; P = 0.003 vs. r = −0.45; P = 0.05). In separate multivariable linear regression models, change in BMI from baseline to 6 months ater surgery was associated signiicantly with decreased 6-month values for CRP (ΔBMI: β = −0.55; P = 0.04), log homeostasis model assessment of insulin resistance (ΔBMI: β = −0.09; P = 0.05), and marginally with log sTfR (ΔBMI: β = −0.03; P = 0.06) when controlled for baseline BMI. Change in log weight from baseline to 6 months ater surgery was associated with signiicantly decreased 6-month values for CRP (log Δ weight: β = −2.97; P = 0.03) and log sTfR (log Δ weight: β = −0.21; P = 0.01) when controlled for baseline weight. Although not signiicant, change in IL-6 from baseline to 6-month follow-up was associated with decreased follow-up log serum hepcidin (Δ IL-6: β = −0.22; P = 0.15) when controlled for baseline IL-6. Change in serum hepcidin from baseline to 6-month follow-up was associated with follow-up log sTfR (Δ serum hepcidin: β = −0.01; P = 0.02) when controlled for baseline hepcidin. Surprisingly, change in weight, BMI, and CRP from baseline to follow-up were not independently associated with change or 6-month values for log serum hepcidin. Race was not a signiicant confounder in any of the models tested above, therefore, was not controlled for in analysis. To further explore the associations between iron status, and anthropometric and biochemical parameters, participants were categorized as either iron deicient (n = 11; sTfR >28.10 nmol/l) or iron suicient (n = 9; sTfR ≤28.10 nmol/l), and compared at baseline and 6 months ater surgery (Table 2). At baseline, iron- Table 2 Baseline and 6-month clinical characteristics of obese premenopausal women dichotomized by baseline iron status Baseline Iron-deficient sTfR >28.10 nmol/l (n = 9) 2 Iron-sufficient sTfR ≤28.10 nmol/l (n = 11) 6-month Follow-up Pa Iron-deficient sTfR >28.10 nmol/l (n = 9) Pb Iron-sufficient sTfR ≤28.10 nmol/l (n = 11) Pc Pa BMI (kg/m ) 51.48 (± 9.25) 44.35 (± 5.05) 0.04 42.83 (± 9.33) 0.004 36.85 (± 4.57) 0.001 0.11 Weight (kg) 144.65 (± 23.47) 118.03 (± 15.94) 0.01 120.08 (± 21.78) 0.001 95.83 (± 25.46) 0.01 0.04 Waist 136.73 (± 16.71) circumference (cm) 123.6 (± 16.47) 0.13 119.45 (± 15.40) 0.02 109.48 (± 13.85) 0.04 0.14 Hemoglobin (g/dl) 12.36 (± 1.40) 11.82 (± 1.20) 0.40 13.31 (± 1.11) 0.01 12.72 (± 1.32) 0.002 0.35 Hematocrit (%) 37.07 (± 3.24) 36.09 (± 2.54) 0.48 39.10 (± 2.40) 0.02 38.50 (±฀3.72) 0.05 0.99 Serum transferrin receptor (nmol/l) 36.61 (IQR 6.16) 24.78 (IQR 4.92) 0.001 24.33 (IQR 4.86) 0.001 21.30 (IQR 6.04) 0.25 0.09 Serum ferritin (ng/ml) 31.50 (IQR 21.50) 22.00 (IQR 15.0) 0.15 29.00 (IQR 39.00) 0.64 20.00 (IQR 20.00) 0.38 0.20 Serum iron (µg/ml) 56.00 (IQR 39.00) 45.00 (IQR 21.00) 0.77 56.00 (IQR 32.00) 0.30 58.00 (IQR 27.00) 0.47 0.74 Transferrin saturation (%) 18.22 (± 8.69) 16.20 (± 8.77) 0.62 20.11 (± 10.12) 0.32 19.36 (± 11.49) 0.58 0.88 160.60 (IQR 86.80) 84.30 (IQR 195.20) 0.24 26.80 (IQR 29.80) 0.004 34.10 (IQR 54.50) 0.01 0.70 3.33 (IQR 2.45) 3.85 (IQR 2.71) 0.24 1.91 (IQR 4.23) 0.20 1.61 (IQR 2.17) 0.002 0.65 0.88 6.14 (± 5.84) 0.004 6.16 (± 3.52) 0.003 0.99 0.14 2.15 (IQR 2.09) 0.10 1.69 (IQR 1.56) 0.06 0.46 Serum hepcidin (ng/ml) HOMAIRa CRP (mg/l) IL-6 (pg/ml) 11.21 (± 6.35) 3.76 (IQR 1.95) 11.59 (± 4.78) 2.36 (IQR 2.10) Erythropoietin (mU/ml) 22.30 (± 5.21) 23.12 (± 4.70) 0.72 14.96 (± 4.36) 0.02 14.77 (± 6.33) 0.02 0.94 Total dietary iron (mg) 18.83 (± 11.40) 18.85 (± 9.04) 0.99 14.80 (± 7.85) 0.22 19.27 (± 6.40) 0.41 0.26 5.69 (± 6.63) 7.06 (± 6.93) 0.68 4.92 (± 6.03) 0.84 5.17 (± 6.98) 0.56 0.92 Metabolic equivalents/week CRP, C-reactive protein; HOMAIR, homeostasis model assessment of insulin resistance; IL-6, interleukin-6; IQR, interquartile range. Between-group comparison: obese iron deficient vs. iron sufficient; P values based on Student’s t-test or Wilcoxon rank-sum. bWithin-group comparison: obese iron-deficient baseline vs. 6 months; P values are based on paired t-test or Wilcoxon signed-rank. cWithin-group comparison: obese iron-sufficient baseline vs. 6 months; P values are based on paired t-test or Wilcoxon signed-rank. a OBESITY | VOLUME 18 NUMBER 10 | OCTOBER 2010 2013 ARTICLES EPIDEMIOLOGY deicient women had a signiicantly elevated BMI (P = 0.04), waist circumference (P = 0.01), and trended higher for IL-6 (P = 0.14) compared to the iron-suicient women. At follow-up, both groups of women had a signiicant decline in BMI, weight, waist circumference, serum hepcidin, sTfR, CRP, and increase in Hct and Hb compared to baseline. he iron-deicient women had smaller declines in IL-6 and homeostasis model assessment of insulin resistance ater surgery compared to the iron-suicient women. Although not signiicant, serum iron and Tsat trended higher in both groups of women at follow-up. Women classiied as iron deicient at baseline remained heavier and had slightly higher sTfR ater surgery. Six months ater restrictive bariatric procedure, ID resolved in all but three women (sTfR >28.10). hese women (baseline BMI 46.86 ± 13.28 kg/m2) had a signiicantly smaller decline in CRP (−1.75 vs. −5.43 mg/l; P = 0.04) and BMI (−3.82 ± 4.88 vs. −9.30 ± 5.28 kg/m2; P = 0.11) compared to women that were iron-suicient at follow-up. Notably and contrary to baseline, the three iron-deicient women at follow-up had signiicantly lower serum hepcidin (9.40 ng/ml (IQR 7.30) vs. 34.30 ng/ml (IQR 36.10); P = 0.03) compared to iron-suicient women (data not shown). DISCUSSION his is the irst report demonstrating that signiicant weight loss, following restrictive bariatric surgery, in premenopausal women, is associated with decreased serum hepcidin and improved functional iron status (sTfR, Hb, and Hct). he relationship between serum hepcidin and iron status parameters strengthened following weight loss and was similar to what we recently reported in nonobese women (17), suggesting that at baseline, hepcidin regulation by inlammation was dominant despite low iron status. his may be due, in part, to the multimodal regulation of hepcidin. At any time, hepcidin levels are determined by the interplay of pathways controlled by iron status, erythropoietic activity, and inlammation, and the relative strength of each of these signals (9–12). Following weight reduction, inlammation decreased signiicantly in the obese women which likely minimized the strength of the inlammatory signal on hepcidin expression helping to reduce levels. Median serum hepcidin observed in the obese women, at baseline, was within the reference range for nonobese ironreplete healthy women (17–286 ng/ml), as is serum hepcidin in obese compared to nonobese iron-replete children (measured by a diferent method) (29,30). Given this “normal” hepcidin level, the question becomes why are obese individuals iron deicient? We theorize that although hepcidin levels were “normal” in the obese women, they were too high for the degree of ID. It is possible that at an earlier time point, before they became iron-deicient, these women had higher hepcidin due to their heightened inlammatory state. Abnormally elevated hepcidin would irst lead to iron-restricted anemia, similar to the anemia of inlammation, but over time, chronic hepcidin excess would lead to the development of ID as iron losses exceed iron absorption. ID would normally cause a dramatic drop in hepcidin levels, as we showed in nonobese iron-deicient women (17), 2014 but in obese women, hepcidin levels only returned to “normal” presumably because of the continuous opposing efect of inlammation on hepcidin production. Iron-deicient individuals require additional iron for repletion. However, “normal” hepcidin levels will not allow for the enhanced absorption of iron but only maintenance of current levels (31,32). herefore, like individuals with iron-refractory ID anemia, the obese women are not able to suiciently replenish their iron stores through dietary means perpetuating their iron-deicient state. To be conirmed, this theory would require a longitudinal cohort of lean individuals that develop obesity over time. Recent studies have shown that dietary iron absorption is impaired in obese individuals despite adequate dietary intake and bioavailability (30,33). Dietary iron intake in our participants was similar at baseline and follow-up suggesting that increased iron absorption and not improved dietary content was responsible for the rebound in hematological parameters. Furthermore, we already observed that participants included in this study had minimal iron accumulation in the liver and adipose tissue reticuloendothelial cells at baseline (17). Considering that hepcidin is the main determinant of the rate of iron absorption (31,34,35), the postoperative decrease in hepcidin likely accounts for the improved dietary iron uptake in these women. Although signiicant improvement of the hematological parameters was observed, there was only marginal improvement in serum iron, Tsat, and ferritin, indicating that the 6-month period was not suicient for the replenishment of iron stores. he process was likely delayed by postoperative inlammation, excess weight, and slightly elevated hepcidin that remained. Although hepcidin was signiicantly decreased ater surgery, it remained signiicantly higher than levels in our nonobese women with similar iron status (17). Interestingly, another study, in a group of premenopausal women with postmalabsorptive gastric bypass and the gastric banding procedure, showed greater 6-month postsurgical decline in BMI and CRP than in our participants that correlated signiicantly with improvement in Tsat (8). Surprisingly, neither weight loss nor decreased inlammation were signiicantly associated with lower serum hepcidin at follow-up, although a trend between change in IL-6 from baseline and decreased serum hepcidin was observed (P = 0.15). his may be due, in part, to the multimodal regulation of hepcidin that could attenuate the signiicance of this relationship but also by changes in adipokines, such as leptin, which are inluenced dramatically by weight loss (19–21,36). It has been reported that leptin stimulates hepcidin mRNA production via the JAK/STAT3 pathway in human hepatoma cells in a similar fashion as IL-6 (37). herefore, weight reduction and subsequent decline in leptin may be associated with decreased stimulation of hepcidin mRNA production and decreased protein expression. he relationship between leptin and serum hepcidin warrants further investigation in obese populations. At baseline, not all obese women sufered from a functional iron deicit (sTfR >28.10 nmol/l), therefore, we assessed the anthropometric and biochemical diferences in those with and without elevated sTfR. As we hypothesized, the women with VOLUME 18 NUMBER 10 | OCTOBER 2010 | www.obesityjournal.org ARTICLES EPIDEMIOLOGY elevated sTfR had a higher BMI (P = 0.04), waist circumference (P = 0.01), marginally higher IL-6 (P = 0.14) and serum hepcidin, and poorer iron status (10,38). Our study had strengths and limitations. First, despite utilizing a prospective cohort and demonstrating temporality, no direct measure of iron absorption or ferroportin-1 expression at the enterocyte was assessed warranting further investigations to conirm our indings. Second, decreased inlammation could also have hepcidin-independent efect on proteins intricately involved in uptake, storage, and release of iron, and this should be assessed in future studies (39). Finally, even though produced mainly by the liver, adipose tissue hepcidin mRNA expression has also been reported, although the bioactive secretion of the adipose-derived protein in vivo remains unknown (40). herefore, it is possible that reduction in adipose tissue was partially responsible for reduction in serum hepcidin levels despite our failure to show this relationship statistically. In conclusion, weight loss in obese premenopausal women is associated with reduced serum hepcidin and inlammation, and improved functional iron status. An improved inlammatory proile, following weight reduction, likely minimizes the strength of the inlammatory signal on hepcidin production, ultimately lessening hepcidin expression. Reduction in hepcidin likely allowed for enhanced dietary iron absorption that resulted in an improved functional iron proile. Not all obese women experienced iron depletion suggesting heterogeneity among these individuals exists. Further exploration of the interaction between serum hepcidin, enterocyte iron transporter expression, and iron absorption in obese individuals before and ater weight loss is warranted. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. ACKNOWLEDGMENTS We acknowledge Karin Testa and Melissa Farb in aiding with data collection and analysis. Our responsibilities were as follows: L.M.T.-H., E.N., and C.A.B., design of the study; L.M.T.-H., C.G., J.V., S.A., data collection; L.M.T.-H., S.F., and C.A.B., data analysis; L.M.T.-H., C.A.B., E.N., and S.F., interpretation of the data; L.M.T.-H., writing of the manuscript draft; C.A.B., E.N., S.F., A.L.H., C.G., S.A., J.V., critical revision of the manuscript. This project was internally funded by the Department of Kinesiology and Nutrition at the University of Illinois at Chicago. E.N. is affiliated with Intrinsic LifeSciences, the company that assessed the serum samples for hepcidin. 21. 22. 23. 24. DISCLOSURE The authors declared no conflict of interest. © 2010 The Obesity Society 25. 26. REFERENCES 1. 2. 3. 4. 5. Nead KG, Halterman JS, Kaczorowski JM, Auinger P, Weitzman M. Overweight children and adolescents: a risk group for iron deficiency. Pediatrics 2004;114:104–108. Lecube A, Carrera A, Losada E et al. Iron deficiency in obese postmenopausal women. Obesity (Silver Spring) 2006;14:1724–1730. Moayeri H, Bidad K, Zadhoush S, Gholami N, Anari S. Increasing prevalence of iron deficiency in overweight and obese children and adolescents (Tehran Adolescent Obesity Study). Eur J Pediatr 2006;165:813–814. Mohamed-Ali V, Pinkney JH, Coppack SW. Adipose tissue as an endocrine and paracrine organ. Int J Obes Relat Metab Disord 1998;22:1145–1158. Pinhas-Hamiel O, Newfield RS, Koren I et al. Greater prevalence of iron deficiency in overweight and obese children and adolescents. Int J Obes Relat Metab Disord 2003;27:416–418. OBESITY | VOLUME 18 NUMBER 10 | OCTOBER 2010 27. 28. 29. 30. 31. Tussing-Humphreys LM, Liang H, Nemeth E, Freels S, Braunschweig CA. Excess adiposity, inflammation, and iron-deficiency in female adolescents. J Am Diet Assoc 2009;109:297–302. Yanoff LB, Menzie CM, Denkinger B et al. Inflammation and iron deficiency in the hypoferremia of obesity. Int J Obes (Lond) 2007;31:1412–1419. Anty R, Dahman M, Iannelli A et al. Bariatric surgery can correct iron depletion in morbidly obese women: a link with chronic inflammation. Obes Surg 2008;18:709–714. Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of anemia of inflammation. Blood 2003;102:783–788. Nemeth E, Ganz T. Regulation of iron metabolism by hepcidin. Annu Rev Nutr 2006;26:323–342. Darshan D, Anderson GJ. Interacting signals in the control of hepcidin expression. Biometals 2009;22:77–87. Collins JF, Wessling-Resnick M, Knutson MD. Hepcidin regulation of iron transport. J Nutr 2008;138:2284–2288. Cartwright GE, Lee GR. The anaemia of chronic disorders. Br J Haematol 1971;21:147–152. Melis MA, Cau M, Congiu R et al. A mutation in the TMPRSS6 gene, encoding a transmembrane serine protease that suppresses hepcidin production, in familial iron deficiency anemia refractory to oral iron. Haematologica 2008;93:1473–1479. Finberg KE, Heeney MM, Campagna DR et al. Mutations in TMPRSS6 cause iron-refractory iron deficiency anemia (IRIDA). Nat Genet 2008;40:569–571. Silvestri L, Pagani A, Nai A et al. The serine protease matriptase-2 (TMPRSS6) inhibits hepcidin activation by cleaving membrane hemojuvelin. Cell Metab 2008;8:502–511. Tussing-Humphreys LM, Nemeth E, Fantuzzi G et al. Elevated systemic hepcidin and iron depletion in obese premenopausal females. Obesity (Silver Spring) 2009; e-pub ahead of print 8 October 2009. Guillem F, Lawson S, Kannengiesser C et al. Two nonsense mutations in the TMPRSS6 gene in a patient with microcytic anemia and iron deficiency. Blood 2008;112:2089–2091. Bougoulia M, Triantos A, Koliakos G. Effect of weight loss with or without orlistat treatment on adipocytokines, inflammation, and oxidative markers in obese women. Hormones (Athens) 2006;5:259–269. Kopp HP, Kopp CW, Festa A et al. Impact of weight loss on inflammatory proteins and their association with the insulin resistance syndrome in morbidly obese patients. Arterioscler Thromb Vasc Biol 2003;23: 1042–1047. Poitou C, Lacorte JM, Coupaye M et al. Relationship between single nucleotide polymorphisms in leptin, IL6 and adiponectin genes and their circulating product in morbidly obese subjects before and after gastric banding surgery. Obes Surg 2005;15:11–23. Cook JD, Baynes RD, Skikne BS. Iron deficiency and the measurement of iron status. Nutr Res Rev 1992;5:198–202. Ferguson BJ, Skikne BS, Simpson KM, Baynes RD, Cook JD. Serum transferrin receptor distinguishes the anemia of chronic disease from iron deficiency anemia. J Lab Clin Med 1992;119:385–390. Looker AC, Loyevsky M, Gordeuk VR. Increased serum transferrin saturation is associated with lower serum transferrin receptor concentration. Clin Chem 1999;45:2191–2199. Baynes RD. Assessment of iron status. Clin Biochem 1996;29:209–215. Block G, Woods M, Potosky A, Clifford C. Validation of a self-administered diet history questionnaire using multiple diet records. J Clin Epidemiol 1990;43:1327–1335. Ainsworth BE, Sternfeld B, Richardson MT, Jackson K. Evaluation of the kaiser physical activity survey in women. Med Sci Sports Exerc 2000;32:1327–1338. Ainsworth BE, Haskell WL, Whitt MC et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 2000;32:S498–S504. Ganz T, Olbina G, Girelli D, Nemeth E, Westerman M. Immunoassay for human serum hepcidin. Blood 2008;112:4292–4297. Aeberli I, Hurrell RF, Zimmermann MB. Overweight children have higher circulating hepcidin concentrations and lower iron status but have dietary iron intakes and bioavailability comparable with normal weight children. Int J Obes (Lond) 2009;33:1111–1117. Young MF, Glahn RP, Ariza-Nieto M et al. Serum hepcidin is significantly associated with iron absorption from food and supplemental sources in healthy young women. Am J Clin Nutr 2009;89:533–538. 2015 ARTICLES EPIDEMIOLOGY 32. Ruivard M, Lainé F, Ganz T et al. Iron absorption in dysmetabolic iron overload syndrome is decreased and correlates with increased plasma hepcidin. J Hepatol 2009;50:1219–1225. 33. Zimmermann MB, Zeder C, Muthayya S et al. Adiposity in women and children from transition countries predicts decreased iron absorption, iron deficiency and a reduced response to iron fortification. Int J Obes (Lond) 2008;32:1098–1104. 34. Roe MA, Collings R, Dainty JR, Swinkels DW, Fairweather-Tait SJ. Plasma hepcidin concentrations significantly predict interindividual variation in iron absorption in healthy men. Am J Clin Nutr 2009;89:1088–1091. 35. Ganz T. Hepcidin--a regulator of intestinal iron absorption and iron recycling by macrophages. Best Pract Res Clin Haematol 2005;18:171–182. 36. Varady KA, Tussing L, Bhutani S, Braunschweig CL. Degree of weight loss required to improve adipokine concentrations and decrease fat 2016 37. 38. 39. 40. cell size in severely obese women. Metab Clin Exp 2009;58: 1096–1101. Chung B, Matak P, McKie AT, Sharp P. Leptin increases the expression of the iron regulatory hormone hepcidin in HuH7 human hepatoma cells. J Nutr 2007;137:2366–2370. Nemeth E, Rivera S, Gabayan V et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest 2004;113:1271–1276. Sheikh N, Dudas J, Ramadori G. Changes of gene expression of iron regulatory proteins during turpentine oil-induced acute-phase response in the rat. Lab Invest 2007;87:713–725. Bekri S, Gual P, Anty R et al. Increased adipose tissue expression of hepcidin in severe obesity is independent from diabetes and NASH. Gastroenterology 2006;131:788–796. VOLUME 18 NUMBER 10 | OCTOBER 2010 | www.obesityjournal.org