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