ORIGINAL ARTICLE

Oncology & Hematology

http://dx.doi.org/10.3346/jkms.2016.31.1.25 • J Korean Med Sci 2016; 31: 25-32

Prevalence and Relationships of Iron Deficiency Anemia with

Blood Cadmium and Vitamin D Levels in Korean Women
Young Ju Suh,1 Ji Eun Lee,2 Anemia, iron deficiency (ID), and iron deficiency anemia (IDA) are common disorders. This
Dae Hyung Lee,3 Hyeon Gyu Yi,4 study was undertaken to determine the prevalence of anemia, ID, and IDA in Korean
Moon Hee Lee,4 Chul Soo Kim,4 females. We examined the associations between IDA, heavy metals in blood, vitamin D
Jeung Weon Nah,5 and Soon Ki Kim2 level and nutritional intakes. The study was performed using on data collected from 10,169
1
women (aged ≥ 10 yr), including 1,232 with anemia, 2,030 with ID, and 690 with IDA
Department of Biomedical Sciences, 2Department
of Pediatrics, College of Medicine, Inha University, during the fifth Korea National Health and Nutrition Examination Survey (KNHANES V;
Incheon; 3Future Strategy Team, Inha University 2010-2012). Prevalence and 95% confidence intervals were calculated, and path analysis
Hospital, Incheon; 4Department of Internal was performed to identify a multivariate regression model incorporating IDA, heavy metals
Medicine, College of Medicine, Inha University,
in blood, vitamin D level, and nutritional intakes. The overall prevalence of anemia, ID and
Incheon; 5Department of Biostatistics, College of
Medicine, Korea University, Seoul, Korea IDA was 12.4%, 23.11%, and 7.7%, respectively. ID and IDA were more prevalent among
adolescents (aged 15-18 yr; 36.5% for ID; 10.7% for IDA) and women aged 19-49 yr
Received: 22 April 2015 (32.7% for ID; 11.3% for IDA). The proposed path model showed that IDA was associated
Accepted: 2 September 2015
with an elevated cadmium level after adjusting for age and body mass index (β = 0.46,
Address for Correspondence: P < 0.001). Vitamin D levels were found to affect IDA negatively (β = -0.002, P < 0.001).
Soon Ki Kim, MD This study shows that the prevalence of anemia, ID, and IDA are relatively high in late
Department of Pediatrics, College of Medicine, Inha University,
27 Inhang-ro, Jung-gu, Incheon 22332, Korea
adolescents and women of reproductive age. Path analysis showed that depressed vitamin
Tel: +82.32-890-2843, Fax: +82.32-890-2844 D levels increase the risk of IDA, and that IDA increases cadmium concentrations in blood.
E-mail: pedkim@inha.ac.kr
Our findings indicate that systematic health surveillance systems including educational
Funding: This work was supported by the Inha University campaigns and well-balanced nutrition are needed to control anemia, ID, and IDA.
Research Grant (50470-01).

Keywords:  Anemia; Iron Deficiency; Prevalence; Korean; Women; Cadmium; Vitamin D

INTRODUCTION hemoglobinopathy and malaria are extremely rare.

  Iron deficiency anemia (IDA) can cause fatigue, reduce work
Despite advances in healthcare, anemia and iron deficiency capacity and resistance to infection. Furthermore, IDA has been
(ID) remain the most common disorders in women (1-3). Glob- shown to be correlated with reduced cognitive performance
ally, the prevalence of anemia increased to 30% in non-preg- and mental development in children and adolescents (11,12);
nant women of reproductive age (15 to 49 yr) (4). and detrimental findings have even been reported for those
  The Korea National Health and Nutrition Examination Sur- with non-anemic ID (13). Too little iron in diet causes anemia
vey (KNHANES) was established in 1998 by the Korea Centers and the non-hematologic effects of iron depletion, whereas too
for Disease Control and Prevention to evaluate the health and much dietary iron initiates excessive free radical activity and
nutritional status of the Korean people. This survey is intended causes oxidative injury to cells. The study performed by Lee et
to determine the prevalence of chronic diseases and their risk al. (8) on KNHANES V-1 (2010) data showed the prevalence of
factors (5). KNHANES IV-2 data (2008) showed that the preva- ID in females was 34.7% among 15-17-yr olds and 31.1% among
lence of anemia in females was 13% in 15 to 17 yr olds, 17.9% in 18-49-yr olds, and that prevalence of IDA was 9.0% among 15-
18 to 49-yr olds, 10.9% in 60 to 69-yr olds, and 18.2% in women 17-yr olds and 11.7% among 18-49-yr olds. In addition, preme­
of 70 or older (6,7). Recently, Lee et al. (8) analyzed KNHANES nopausal status, pregnancy, low income, underweightedness,
V-1 (2010) data and found prevalence of 10.1% in 15 to 17-yr and iron- or vitamin C-poor diets were found to be associated
olds, 14.7% in 18 to 49-yr olds, 7.7% in 60 to 69-yr olds, and 20.7% with IDA.
in women 70 or older.   It is known that the mechanisms of iron absorption are close-
  Anemia is an independent risk factor of increased morbidity ly related with levels of heavy metals in blood (14). ID is associ-
and mortality and of diminished quality of life (9,10). The main ated with elevated blood levels of divalent heavy metals, such
causes of anemia worldwide are; ID, hemoglobinopathy, and as, cadmium, lead, and manganese (14-18). In addition, vita-
malaria. However, in Korea, anemia is caused by ID, because min D deficiency has been shown to be associated with an in-

© 2016 The Korean Academy of Medical Sciences. pISSN 1011-8934

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0)
which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. eISSN 1598-6357
 Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

creased risk of IDA in children and adolescents (19,20). Howev- (K2 EDTS tube, Vacutainers). Blood cadmium and lead levels
er, relationships between IDA, blood heavy metals and vitamin were measured by graphite furnace atomic absorption spec-
D levels have not been simultaneously examined. trometry (GF-AAS) using the Zeeman background correction
  In the present study, we aimed at determining the prevalence (PerkinElmer AAS800, PerkinElmer). Blood cadmium and lead
of anemia, iron depletion, and IDA in a female population aged levels were analyzed by the Neodin Medical Institute (a labora-
10 yr or older, and to compare how this prevalence has changed tory certified by the Korea Ministry of Health and Welfare).
over years. In addition, we sought to verify associations between Three commercial reference materials were used for internal
IDA, heavy metals in blood, vitamin D levels, and nutritional quality assurance and control (Lyphochek® Whole Blood Met-
intakes by path analysis. als Control; Bio-Rad, Hercules, CA, USA). The Neodin Medical
Institute had been certified with respect to quality assurance
MATERIALS AND METHODS and control by the German External Quality Assessment Scheme
operated by Friedrich-Alexander University and the Quality As-
Design and study population surance Program operated by the Korea Occupational Safety
This study was performed using data obtained during the KN­ and Health Agency. The limits of detection were 0.062, 0.081
HANES V study (2010 to 2012), a cross-sectional and nationally and 0.088 μg/L (for blood cadmium), and 0.142, 0.148, and 0.172
representative survey conducted on non-institutionalized indi- μg/dL (for blood lead) in 2010, 2011, and 2012, respectively.
viduals by the Division of Chronic Disease Surveillance, Korea   Anemia was defined according to the WHO criteria as: Hgb
Centers for Disease Control and Prevention. KNHANES data is < 11.5 g/dL at 10-11 yr and < 12 g/dL at ≥ 12 yr. ID was defined
obtained from 10,000 individuals randomly chosen from the as either a serum ferritin level of < 15 μg/mL or a transferrin sat-
192 regions in Korea on an annual basis using a stratified, mul- uration (TS) value of < 10%. IDA was defined as anemia with ID.
tistage sampling procedure. KNHANES data is obtained using
health interviews, health examinations, and a nutrition survey. Statistical analysis and model hypotheses
  Women who were pregnant, less than 10 yr old, indefinable Prevalence and 95% confidence intervals (CI) were estimated
for anemia, ID or IDA (due to missing data) were excluded from using the SAS survey procedures. Weighted results were esti-
the study, which was performed on 10,169 Korean women, which mated using survey sample weight variables for associations
included 1,232 anemic, 2,030 ID, and 690 IDA subjects. Adoles- between health interview, health examination, and nutrition
cents were divided into two groups (10 to 14-yr olds and 15 to survey. Associations between IDA and clinical factors (such as
18-yr olds), because 98% of girls reach menarche at 15 yr of age age, body mass index [BMI] and menstruation status), heavy
(21). We considered subjects aged ≤ 18 yr old as adolescents in metals in blood (e.g., blood lead and cadmium levels), vitamin
accordance with the guidelines issued by the Korea Centers for D levels in blood, nutritional intakes (e.g., protein, iron, and
Disease Control and Prevention and the Korean Society of Pe- calcium intakes), and year were identified using SAS surveyreg
diatrics. procedure (for continuous variables) or surveyfreq procedure
(for categorical variables). For blood lead and cadmium levels,
Laboratory measurements and definition values were log transformed due to non-normality. The statisti-
Blood samples were obtained from an antecubital vein after cal analysis was performed using SAS version 9.3 (SAS Institute,
obtaining informed consent and collected in BD Vacutainer Cary, NC, USA). Statistical significance was accepted for P < 0.05
tubes containing EDTA for trace element determinations. Se- (two-tailed).
rum hemoglobin (Hgb) levels were measured using the sodium   The maximum likelihood (ML) based structural equation
lauryl sulfate hemoglobin detection method using a XE-2100D model (SEM) was used to analyze predicted paths between vari-
hematology analyzer (Sysmex, Tokyo, Japan). Serum ferritin, a ables in the model. Relationships between IDA, clinical factors,
surrogate of dietary iron, has been evaluated since 2007. Iron heavy metals in blood, blood vitamin D levels, and nutritional
and total iron binding capacity (TIBC) were added in 2010. Se- intakes were explored. To measure goodness of fit of the path
rum ferritin and serum 25 (OH) vitamin D levels were measured analysis model, normed fit index (NFI), incremental fit index
using an immunoradiometric assay and a 1470 WIZARD gam- (IFI), comparative fit index (CFI), and root mean squared error
ma-counter (PerkinElmer, Turku, Finland). Reference materials of approximation (RMSEA) were calculated. CFI is the indicator
were used for quality assurance and control (Lyphochek® Whole most commonly used to evaluate model fit versus a null model.
Blood Metals Control; Bio-Rad, Hercules, CA, USA). Nutritional A value of > 0.90 was considered a good fit. The RMSEA was
intakes were determined using Dietary Reference Intakes for investigated for lack of fit to avoid problems of sample size when
Koreans by the Korean Nutrition Society (8). we use chi-square statistics; a RMSEA value of ≤ 0.08 is gener-
  To assess heavy metals in blood, 3 mL blood samples were ally taken to be acceptable. Path analysis was performed using
drawn into standard evacuated tubes for trace element EDTA AMOS version 18.0 (SPSS Inc.) software.

26  
http://jkms.org http://dx.doi.org/10.3346/jkms.2016.31.1.25
 Table 1. Prevalence of anemia in the Korean female population

Total Prevalence Total Prevalence Total Prevalence Total Prevalence

Parameters No. 95% CI No. 95% CI No. 95% CI No. 95% CI
No. rate (%) No. rate (%) No. rate (%) No. rate (%)
Year 2010 2011 2012 2010-2012
Total 3,396 443 13.0 11.6-14.4 3,464 427 12.7 11.3-14.2 3,251 362 11.5 10.0-13.0 10,111 1,232 12.4 11.6-13.2
Age group (yr)
10-14 225 6 2.5 0.3-4.8 204 6 3.0 0.2-5.9 204 4 3.0 0.0-6.2 633 16 2.9 1.3-4.5
15-18 125 11 10.7 3.8-17.6 129 15 11.1 4.1-18.1 98 11 13.9 5.9-22.0 352 37 11.9 7.7-16.1
19-49 1,575 262 15.5 13.4-17.6 1,501 249 15.7 13.4-17.9 1,349 201 13.8 11.5-16.1 4,425 712 15 13.7-16.3
50-59 613 51 8.2 5.5-10.9 623 46 7.3 4.9-9.8 571 35 5.7 3.3-8.1 1,807 132 7.0 5.6-8.5
60-69 487 41 8.3 5.6-11.0 504 38 8.8 5.5-12.2 519 29 6.3 3.45-9.1 1,510 108 7.8 6.1-9.5

http://dx.doi.org/10.3346/jkms.2016.31.1.25
≥ 70 371 72 19.5 14.2-24.8 503 73 16.5 12.8-20.3 510 82 17.6 13.7-21.6 1,384 227 17.8 15.3-20.3

Table 2. Prevalence of iron deficiency in the Korean female population

Total Prevalence Total Prevalence Total Prevalence Total Prevalence

Parameters No. 95% CI No. 95% CI No. 95% CI No. 95% CI
No. rate (%) No. rate (%) No. rate (%) No. rate (%)
Year 2010 2011 2012 2010-2012
All 3,430 685 22.8 20.9-24.8 3,475 655 21.6 19.7-23.5 3,264 690 24.9 22.4-27.5 10,169 2,030 23.1 21.9-24.4
Age group (yr)
10-14 234 41 18.7 11.3-26.1 206 35 19.3 12.7-25.9 206 33 16.3 11.0-21.7 646 109 18.2 14.4-21.9
15-18 126 47 39.7 30.1-49.3 131 38 28.7 20.0-37.4 99 42 41.6 29.0-54.2 356 127 36.5 30.4-42.5
19-49 1,584 496 31.1 28.4-33.8 1,502 482 30.9 28.2-33.5 1,352 499 36.2 32.5-39.8 4,438 1,477 32.7 30.9-34.4
50-59 619 55 8.9 6.5-11.3 627 56 10.2 6.8-13.6 573 49 9.1 6.3-11.8 1,819 160 9.4 7.7-11.1
60-69 493 24 5.1 3.1-7.1 505 20 3.2 1.5-4.8 520 23 3.2 1.5-4.9 1,518 67 3.8 2.8-4.8
≥ 70 374 22 7.1 3.5-10.8 503 24 4.7 2.6-6.9 514 44 9.4 6.3-12.5 1,391 90 7.1 5.4-8.8
Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

Table 3. Prevalence of iron deficiency anemia in the Korean female population

Parameters Total Prevalence Total Prevalence Total Prevalence Total Prevalence
No. 95% CI No. 95% CI No. 95% CI No. 95% CI
No. rate (%) No. rate (%) No. rate (%) No. rate (%)
Year 2010 2011 2012 2010-2012
All 3,406 240 8.1 6.8-9.3 3,462 238 7.4 6.3-8.6 3,243 212 7.6 6.4-8.8 10,111 690 7.7 7.0-8.4
Age group (yr)
10-14 233 4 1.8 0.0-3.8 207 6 3.0 0.2-5.8 206 3 2.1 0.0-4.8 646 13 2.3 0.9-3.8
15-18 126 10 10.3 3.4-17.1 131 12 8.8 2.3-15.4 98 10 13.3 5.4-21.3 355 32 10.7 6.7-14.8
19-49 1,562 192 12.0 10.0-13.9 1,487 179 11.0 9.1-12.8 1,332 160 11.1 9.2-13.0 4,381 531 11.3 10.2-12.4
50-59 619 20 3.3 1.8-4.8 628 22 3.8 1.8-5.8 573 14 2.8 0.9-4.7 1,820 56 3.3 2.2-4.3
60-69 493 4 0.8 0.0-1.6 505 7 1.1 0.2-1.9 520 6 1.1 0.0-2.2 1,518 17 1.0 0.4-1.5
≥ 70 373 10 3.0 1.0-5.0 504 12 2.5 0.8-4.1 514 19 4.1 2.1-6.1 1,391 41 3.2 2.1-4.3

http://jkms.org  27
 Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

  The main research hypotheses for path analysis were: 50 to 59, 60 to 69, and ≥ 70-yr olds were 18.2%, 9.4%, 3.8%, and
  Hypothesis 1. Vitamin D level negatively influences IDA. 7.1%, respectively.
 Hypothesis 2. IDA is positively associated with divalent heavy   The overall prevalence of IDA during the study period was
metal levels in blood. 7.7% (Table 3), and was most prevalent in 15-18 (10.7%) and 19-
49 (11.3%)-yr olds. The prevalence of IDA was 2.3% among 10
Ethics statement to 14-yr olds, 3.3% among 50 to 59-yr olds, 1.0% among 60 to
The KNHANES V survey was approved by the institutional re- 69-yr olds, and 3.2% in ≥ 70-yr olds. The prevalence of ID was
view board of the Korea Center for Disease Control and Preven- approximately three times greater than that of IDA.
tion (IRB No. 2010-02CON-21-C, 2011-02CON-06-C and 2012-   For 15-18-yr olds, the prevalence of anemia, ID, and IDA in-
01EXP-01-2C). All individuals that participated in the KNHA­ creased in 2012 as compared with 2010 or 2011, and for 19-49-
NES V survey provide written informed consent. yr olds the prevalence of ID tended to increase in 2012 as com-
pared with 2010 or 2011 (Fig. 1). Although the prevalence rates
RESULTS of anemia (P = 0.322), ID (P = 0.086), and IDA (P = 0.748) show­
ed slight changes as determined by surveys conducted for 3 yr,
The overall prevalence of anemia was 12.4% during the 2-yr there was no statistical significance.
study period (2010 to 2012) in Korean women and adolescents,   Table 4 shows the distributions of clinical factors stratified by
as presented in Table 1. The prevalence was greatest among IDA. Subjects with IDA group were younger than those without
those aged ≥ 70 yr (17.8%) and this was followed by 19-49-yr (P < 0.001). Furthermore, menstruation status differed in the
olds (15.0%). The prevalence of anemia in the other age groups IDA and non-IDA groups (P < 0.001), and blood cadmium (log-
was; 2.9% in 10 to 14-yr olds, 11.9% in 15 to 18-yr olds, 7.0% in transformed) and vitamin D levels were significantly different
50 to 59-yr olds, and 7.8% in 60 to 69-yr olds. in these two groups (P < 0.001). However, BMIs, blood lead lev-
  The overall prevalence of ID between 2010 and 2012 was els, nutritional intakes, and survey years were no different in
23.11% (Table 2). The prevalence of ID for 3 yr in 15-18 and 19- these groups at the α = 0.05 level.
49-yr olds was 36.5% and 32.7%, respectively, and for 10 to 14,   The theoretical path model based on our hypotheses was

25 50

20 40
Prevalence rate (%)

Prevalence rate (%)

15 30

10 20

5 10

0 0
2010 2011 2012 2010 2011 2012
A B
Year Year

14
All
12
10-14 yr
Prevalence rate (%)

10 15-18 yr

8 19-49 yr

6 50-59 yr

4 60-69 yr

≥ 70 yr
2
Fig. 1. Change in the prevalence of anemia, ID and IDA by
0 age in Korean women. (A) Anemia. (B) Iron deficiency. (C) Iron
2010 2011 2012
deficiency anemia.
Year C

28  
http://jkms.org http://dx.doi.org/10.3346/jkms.2016.31.1.25
Figure 2
Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

Vitamin D level

β = -0.002†
Protein intake
0.44
0.52 Iron intake
0.51 IDA
β = 0.48† log Cadmium
Calcium intake level
β = 0.13† (R2 = 0.54)

Menstruation β = 0.001†
β = 0.03† log Lead level
-0.68 (R2 = 0.31)
β = 0.01 †
Fig. 2. Proposed hypothetical path model. The solid and
-0.20 Age
β = -0.001* β = 0.01
†
dotted lines stand for significant and non-significant as-
0.36
sociations between factors, respectively. β is a parame-
BMI ter estimate of each significant association. *P < 0.05;
†
P < 0.001.

Table 4. Distribution of factors stratified by iron deficiency anemia (IDA) in the Korean Table 5. Fit indices of our hypothetical model by path analysis
female population
Model NFI IFI CFI RMSEA
Mean (SE)
Factors P value† Significant goodness of fit 0.90-1 0.90-1 0.90-1 < 0.05: good
IDA Non IDA criteria 0.05-0.08: fair
0.08-1: moderate
Age (yr) 38.15 (0.60) 42.71 (0.29) < 0.001
Hypothetical model 0.904 0.905 0.905 0.084
BMI (kg/m2) 22.57 (0.21) 22.97 (0.06) 0.061
Menstruation < 0.001 NFI, normed fit index; IFI, incremental fit index; CFI, comparative fit index; RMSEA,
No 91 (13.5)* 4,862 (52.7) root mean squared error of approximation.
Yes 582 (86.5)* 4,357 (47.3)
Blood lead level (μg/dL) 1.80 (0.07) 1.81 (0.02) 0.766‡ (R2 = 0.54). However, IDA was not associated with blood lead
Blood cadmium level (μg/L) 1.53 (0.08) 1.02 (0.01) < 0.001‡
levels (P = 0.625).
Vitamin D level (μg/mL) 14.43 (0.24) 16.40 (0.13) < 0.001
Protein intake (g) 62.80 (1.77) 61.76 (0.50) 0.565
Iron intake (mg) 13.40 (0.48) 13.22 (0.17) 0.730 DISCUSSION
Calcium intake (mg) 460.85 (7.72) 464.19 (5.36) 0.397
Year 0.748 Anemia affects an enormous number of women. In this study,
2010 240 (34.8)* 3,166 (33.6)
2011 238 (34.5)* 3,224 (34.2) the overall prevalence of anemia, ID and IDA was 12.4%, 23.1%,
2012 212 (30.7)* 3,031 (32.2) and 7.7%, respectively. In particular, anemia and ID were more
*No. (%); †P values were obtained by surveyreg procedure for continuous variable and common in late adolescent girls and women of reproductive
by surveyfreq procedure for categorical variables; ‡Values were log-transformed lev- age. The prevalence of anemia, ID, and IDA did not change ap-
els for analysis.
preciably over the period 2010 to 2012, although increases in
the prevalence of ID and IDA occurred in females aged 15 to 18
tested (Fig. 2). The hypothesis path model fitted data well with yr. The overall prevalence of iron depletion (defined as serum
NFI, IFI and CFI values of 0.90-0.91 (Table 5). ferritin level < 15 ng/mL) was 23.8% during 2008-2012: 25.5%,
  The model had a RMSEA value of 0.084, indicating model ac- 25.8%, 21.6%, 21.1%, and 24.6% in the year of 2008, 2009, 2010,
ceptability. Results obtained showed vitamin D level (β = -0.002, 2011, and 2012, respectively.
SE = 0.0004, P < 0.001), menstruation (β = 0.13, SE = 0.007, P <   The average woman has approximately 40 mg/kg of iron in
0.001), age (β = 0.001, SE = 0.0001, P < 0.001), and BMI (β = -0.001, body stores, which are maintained by the daily absorption of
SE = 0.001, P = 0.036) significantly associated with IDA, but that 1-2 mg of iron in the duodenum (22). Adolescence is character-
protein (P = 0.261), iron (P = 0.233), and calcium intakes (P = ized by the somatic growth spurt and an increased rate of de-
0.325) did not. velopment with commensurate expansion of total blood vol-
  Moderately positive correlations were observed between ume, which requires a greater supply of iron. Adolescents are
protein, iron and calcium intake (range from 0.44 to 0.52), and highly susceptible to nutritional iron deficiency, not only be-
significant relationships were found for age, BMI, and menstru- cause of their considerably increased nutritional requirements,
ation. Path analyses also showed IDA was significantly associ- but also because of poor dietary habits. The observed increase
ated with an elevated blood cadmium level (log-transformed) in the prevalence of ID in adolescents is worrisome from the
after adjusting for age and BMI (β = 0.48, SE = 0.04, P < 0.001) perspectives of possible developmental and learning problems.

http://dx.doi.org/10.3346/jkms.2016.31.1.25 http://jkms.org  29
 Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

  In non-pregnant females of reproductive age, the prevalence cents (15,18) or adults (14,16,17). Our proposed path model
of anemia in this study was 15.0%, which was not very different showed that IDA was associated with an increase in blood cad-
from the 17.9% of the 2008 KNHANES data (7), the 12.2% of the mium levels (log-transformed) in females aged 10 yr or older
Third National Health and Nutrition Examination Survey (NH­ (β = 0.48, SE = 0.04, P < 0.001). Here we focused on the associa-
ANES III) (23), the 17.8% of studies conducted in America, and tion between IDA and blood cadmium level (standardized β =
the 19.0% of European studies cited by WHO (4). In the present 0.16) after adjusting for age and BMI, even though the strength
study, the prevalence of ID was 32.7% in 19 to 49-yr olds, which of association between age and blood cadmium level was high-
was similar to the 33.0% found using 2008 KNHANES data. It ap- er (standardized β = 0.71).
pears the additional loss of an average of 1 mg iron per day throu­   Vitamin D has an effect on calcium absorption and bone me-
gh menstruation tilts balance toward the development of IDA. tabolism, and its deficiency is viewed as a crucial nutritional
  Despite of relatively high prevalence of anemia in elderly peo- problem (33). In the present study, blood vitamin D levels were
ple ( ≥ 70 yr old; 17.8%), the prevalence of IDA was only 3.2%. found to have a significant negative effect on IDA in those aged
One third of anemia cases in the elderly are attributable to nu- ≥ 10 yr (β = -0.002, SE = 0.0004, P < 0.001). Studies have been
tritional (iron, folate, or vitamin B12 deficiency), one third due performed recently on the association between vitamin D defi-
chronic inflammation or chronic kidney disease, and one-third ciency and IDA in children and adolescents (20,33), and a sig-
to some idiopathic cause (24,25). However, it should be noted nificant association was found between vitamin D and IDA,
no occult blood testing in stools or gastrointestinal examination possibly due to the suppressive effect of vitamin D on IDA via
was performed during the KNHANES V study. Anemia in the hepcidin (an iron-regulatory hormone) (20). Vitamin D supple-
elderly is associated with frailty, increased mortality, poor cog- mentation might control IDA by decreasing hepcidin levels.
nition, and diminished physical performance (26-28), and is These findings suggest that to prevent ID or IDA, women should
especially harmful in the presence of heart disease or chronic consume diet with sufficient available iron and vitamin D to
kidney disease. In view of the rapid increase in the proportion achieve recommended intakes.
of elderly, it is imperative that an intervention be made avail-   Despite progress in nation-wide survey over the past decade,
able to prevent and treat anemia. anemia remains a significant public health problem in Korea,
  The structural equation model (SEM) is a powerful multivari- which begs the question, ‘Why has the prevalence of anemia
ate analysis technique that was devised to analyze complex re- and IDA not decreased?’ The reason may be unbalanced nutri-
lationships among many observed or unobserved variables tion, food restriction including dieting, public unawareness of
(29). The SEM enables the validity of empirical models to be anemia prevention, and no or ineffective feedback for anemia
tested using variable correlation or covariance matrices. Path and IDA. Furthermore, it appears anemia and IDA are relatively
analysis (a special type of SEM) has been used to evaluate di- neglected by health care personnel and policymakers. What is
rect dependencies among a set of observed variables as a form more, many clinicians do not consider IDA a real problem, and
of multivariate regression analyses with focus on causality. In as a result they often overlook offering preventive nutritional
the present study, we examined relationship between IDA, blood guidance. Even when screening is conducted and ID is identi-
cadmium levels, and vitamin D levels simultaneously by path fied, appropriate and effective treatment is often overlooked or
analysis based on the following research hypotheses: Vitamin inadequate (34).
D negatively influences IDA (hypothesis 1), and the presence of   Strong advocacy is required to reduce the incidences of ane-
IDA is positively associated with divalent heavy metals in blood mia and IDA. IDA is compromises physical and cognitive de-
(hypothesis 2). velopment and performance, especially in children and adoles-
  Cadmium is a major environmental pollutant and a carcino- cents. In addition, IDA in adults contributes to low work pro-
genic heavy metal (15,30), and it has been shown cadmium ex- ductivity and has a negative impact on economy according to
posure involves diet, particularly, in children and adults not the WHO (35). Furthermore, treatment of IDA can improve qual-
subjected to occupational exposure (18). Blood cadmium con- ity of life and reduce associated morbidity and mortality
centrations are elevated as cadmium absorption increases thr­   According as the 2013 WHO guidelines (35), optimizing nu-
ough divalent metal transporter 1 (DMT1) in subjects with ID trition with daily or intermittent (1 to 3 times per week) iron sup-
or IDA (31). That is, in those with ID or IDA blood cadmium plementation, should be considered a first-line intervention in
concentrations are elevated, because such individuals absorb high-risk or high-prevalence groups. Although it is probably
more heavy metals. Moreover, Meltzer et al. (32) showed that less effective than daily iron supplementation, intermittent iron
blood cadmium levels were associated with reduced serum fer- appears to be a useful and cost effective way of controlling ane-
ritin levels in women; and in recent investigations of the associ- mia and IDA (36). Iron fortification with micronutrient powders
ation between blood cadmium and ID (14-18), it was found and fortification of staple foods and condiments may be a use-
that ID elevates blood cadmium levels in children and adoles- ful option if the food concerned is widely consumed by at-risk

30  
http://jkms.org http://dx.doi.org/10.3346/jkms.2016.31.1.25
Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

individuals and if the fortificants do not have toxic effects (36). Ji Eun Lee  http://orcid.org/0000-0002-7386-0015
However, targeted anemia surveillance and intervention should Dae Hyung Lee  http://orcid.org/0000-0002-0284-0519
be viewed as a priority in high-risk populations, such as, late Hyeon Gyu Yi  http://orcid.org/0000-0002-3710-0959
adolescents and premenopausal females. On the other hand, Moon He Lee  http://orcid.org/0000-0002-5965-1815
anemia is also an important health concern in the elderly. In Chul Soo Kim  http://orcid.org/0000-0002-1760-0155
this population, anemia should be investigated from a different Jeung Weon Nah  http://orcid.org/0000-0001-7498-8398
perspective, because it is usually multifactorial in origin. Soon Ki Kim  http://orcid.org/0000-0002-4785-1705
  This is the first study to be conducted on simultaneous asso-
ciations between IDA and blood cadmium and vitamin D levels REFERENCES
in a large population of females aged ≥ 10 yr. The results also
provide that vitamin D level is one of possible determinants of 1. Wu AC, Lesperance L, Bernstein H. Screening for iron deficiency. Pedi-
IDA and of elevated blood cadmium level. atr Rev 2002; 23: 171-8.
  However, we acknowledge several limitations of our analysis. 2. Greydanus DE, Patel DR. The female athlete. Before and beyond puber-
Preschool and school age children below 10 yr old were not in- ty. Pediatr Clin North Am 2002; 49: 553-80.
cluded, although they are also known to exhibit high prevalence 3. de Benoist B, McLean E, Egli I, Cogswell M. Worldwide prevalence of

of anemia and iron deficiency. Furthermore, because the acute anemia 1993-2005. WHO Global Database on Anemia 2008;ISBN 978
92 4 159665 7(WH 155): 1-40.
phase reactants, such as, C-reactive protein, were not included,
4. McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B. Worldwide prev-
ID may have been underestimated. In addition, regarding the
alence of anaemia, WHO Vitamin and Mineral Nutrition Information
diagnosis of ID, it would have been better that erythrocyte pro-
System, 1993-2005. Public Health Nutr 2009; 12: 444-54.
toporphyrin (EPP), zinc protoporphyrin, or serum transferrin 5. KNHANES. Korea National Health & Nutrition Examination Survey.
receptor levels had been measured. Also, since this study was Available at https://knhanes.cdc.go.kr/knhanes/eng/index.do [access
based on the cross-sectional KNHANES study, the results ob- on 13, 03, 2015]
tained by path analyses did not demonstrate causal relation- 6. Kim SK, Kang HS, Lee JE, Kwon YS, Jun YH, Kim CS, Kim YT. The preva-
ships between the variables. lence change of anemia in the population aged 10 years or older: 2005
  In conclusion, this study shows that the prevalence of ane- and 2009. Clin Pediatr Hematol Oncol 2011; 18: 8-12.
mia, ID, and IDA was relatively high in late adolescents and in 7. Kim SK, Kang HS, Kim CS, Kim YT. The prevalence of anemia and iron
women of reproductive age. Furthermore, minimal improve- depletion in the population aged 10 years or older. Korean J Hematol
ments in the burdens of anemia and IDA were observed in the 2011; 46: 196-9.
8. Lee JO, Lee JH, Ahn S, Kim JW, Chang H, Kim YJ, Lee KW, Kim JH, Bang
present study from the KNHANES V survey as compared with
SM, Lee JS. Prevalence and risk factors for iron deficiency anemia in the
previous study from the KNHANES IV survey. Path analysis re-
Korean population: results of the fifth Korea National Health and Nutri-
sults showed low vitamin D levels enhance the risk of IDA, and
tion Examination Survey. J Korean Med Sci 2014; 29: 224-9.
that IDA increases cadmium concentrations in blood. Our find-
9. Chaves PH, Ashar B, Guralnik JM, Fried LP. Looking at the relationship
ings indicate that systematic health surveillance systems includ- between hemoglobin concentration and prevalent mobility difficulty in
ing educational campaigns and well-balanced nutrition are need- older women. Should the criteria currently used to define anemia in old-
ed to control anemia, ID, and IDA. er people be reevaluated? J Am Geriatr Soc 2002; 50: 1257-64.
10. Penninx BW, Pahor M, Cesari M, Corsi AM, Woodman RC, Bandinelli S,
DISCLOSURE Guralnik JM, Ferrucci L. Anemia is associated with disability and de-
creased physical performance and muscle strength in the elderly. J Am
The authors have no potential conflicts of interest to declare. Geriatr Soc 2004; 52: 719-24.
11. Sachdev H, Gera T, Nestel P. Effect of iron supplementation on mental
and motor development in children: systematic review of randomised
AUTHOR CONTRIBUTION
controlled trials. Public Health Nutr 2005; 8: 117-32.
12. Congdon EL, Westerlund A, Algarin CR, Peirano PD, Gregas M, Lozoff
Conception, design, study supervision: Kim SK. Analysis and
B, Nelson CA. Iron deficiency in infancy is associated with altered neu-
interpretation of data: Suh YJ, Lee DH, Nah JW. Writing manu-
ral correlates of recognition memory at 10 years. J Pediatr 2012; 160: 1027-
script: Suh YJ, Kim SK. Discussion and review of the manuscript: 33.
Lee JE, Yi HG, Lee MH, Kim CS. Approval of the manuscript: all 13. Bruner AB, Joffe A, Duggan AK, Casella JF, Brandt J. Randomised study
authors. of cognitive effects of iron supplementation in non-anaemic iron-defi-
cient adolescent girls. Lancet 1996; 348: 992-6.
ORCID 14. Kim Y, Lobdell DT, Wright CW, Gocheva VV, Hudgens E, Bowler RM.
Blood metal concentrations of manganese, lead, and cadmium in rela-
Young Ju Suh  http://orcid.org/0000-0002-6816-8067 tion to serum ferritin levels in Ohio residents. Biol Trace Elem Res 2015;

http://dx.doi.org/10.3346/jkms.2016.31.1.25 http://jkms.org  31
 Suh YJ, et al.  •  Prevalence and Relationships between Anemia, Cadmium, and Vitamin D

165: 1-9. Physician 2010; 82: 480-7.

15. Lee BK, Kim SH, Kim NS, Ham JO, Kim Y. Iron deficiency increases blood 26. Denny SD, Kuchibhatla MN, Cohen HJ. Impact of anemia on mortality,
cadmium levels in adolescents surveyed in KNHANES 2010-2011. Biol cognition, and function in community-dwelling elderly. Am J Med 2006;
Trace Elem Res 2014; 159: 52-8. 119: 327-34.
16. Lee BK, Kim Y. Iron deficiency is associated with increased levels of blood 27. Penninx BW, Guralnik JM, Onder G, Ferrucci L, Wallace RB, Pahor M.
cadmium in the Korean general population: analysis of 2008-2009 Ko- Anemia and decline in physical performance among older persons. Am J
rean National Health and Nutrition Examination Survey data. Environ Med 2003; 115: 104-10.
Res 2012; 112: 155-63. 28. Artz AS. Anemia and the frail elderly. Semin Hematol 2008; 45: 261-6.
17. Kim SH, Kim Y, Kim NS, Lee BK. Gender difference in blood cadmium 29. Beran TN, Violato C. Structural equation modeling in medical research:
concentration in the general population: can it be explained by iron de- a primer. BMC Res Notes 2010; 3: 267.
ficiency? J Trace Elem Med Biol 2014; 28: 322-7. 30. Hong YS, Lee BK, Park JD, Sakong J, Choi JW, Moon JD, Kim DS, Kim
18. Silver MK, Lozoff B, Meeker JD. Blood cadmium is elevated in iron defi- BG. Blood cadmium concentration of residents living near abandoned
cient U.S. children: a cross-sectional study. Environ Health 2013; 12: 117. metal mines in Korea. J Korean Med Sci 2014; 29: 633-9.
19. Atkinson MA, Melamed ML, Kumar J, Roy CN, Miller ER 3rd, Furth SL, 31. Andrews NC. Disorders of iron metabolism. N Engl J Med 1999; 341: 1986-
Fadrowski JJ. Vitamin D, race, and risk for anemia in children. J Pediatr 95.
2014; 164: 153-8. 32. Meltzer HM, Brantsaeter AL, Borch-Iohnsen B, Ellingsen DG, Alexan-
20. Lee JA, Hwang JS, Hwang IT, Kim DH, Seo JH, Lim JS. Low vitamin D der J, Thomassen Y, Stigum H, Ydersbond TA. Low iron stores are relat-
levels are associated with both iron deficiency and anemia in children ed to higher blood concentrations of manganese, cobalt and cadmium
and adolescents. Pediatr Hematol Oncol 2015; 32: 99-108. in non-smoking, Norwegian women in the HUNT 2 study. Environ Res
21. Park MJ, Lee IS, Shin EK, Joung H, Cho SI. The timing of sexual matura- 2010; 110: 497-504.
tion and secular trends of menarchial age in Korean adolescents. Korean 33. Yoon JW, Kim SW, Yoo EG, Kim MK. Prevalence and risk factors for vi-
J Pediatr 2006; 49: 610-6. tamin D deficiency in children with iron deficiency anemia. Korean J
22. Shander A, Cappellini MD, Goodnough LT. Iron overload and toxicity: Pediatr 2012; 55: 206-11.
The hidden risk of multiple blood transfusions. Vox Sang 2009; 97: 185- 34. Buchanan G. Paucity of clinical trials in iron deficiency: lessons learned
97. from study of VLBW infants. Pediatrics 2013; 131: e582-4.
23. NHANES. The Third National Health and Nutrition Examination Sur- 35. WHO. Essential Nutrition Actions: improving maternal, newborn, in-
vey (NHANES III 1988–1994). Available at http://www.cdc.gov/nchs/ fant and young child health and nutrition. Geneva, Switzerland: World
nhanes/nh3data.htm [access on 13, 03, 2015] Health Organization 2013.
24. Patel KV. Epidemiology of anemia in older adults. Semin Hematol 2008; 36. Pasricha S, Drakesmith H, Black J, Hipgrave D, Biggs B. Control of iron
45: 210-7. deficiency anemia in low- and middle-income countries. Blood 2013; 121:
25. Bross MH, Soch K, Smith-Knuppel T. Anemia in older persons. Am Fam 2607-17.

32  
http://jkms.org http://dx.doi.org/10.3346/jkms.2016.31.1.25