nutrients

Article
Relationships between Dietary Patterns and
Erythropoiesis-Associated Micronutrient Deficiencies (Iron,
Folate, and Vitamin B12) among Pregnant Women in Taiwan
Noor Rohmah Mayasari 1,2 , Chyi-Huey Bai 3,4 , Jane C.-J. Chao 1 , Yi-Chun Chen 1 , Ya-Li Huang 3,4 ,
Fan-Fen Wang 5 , Bayu Satria Wiratama 6 and Jung-Su Chang 1,7,8,9,10, *

1 School of Nutrition and Health Sciences, College of Nutrition, Taipei Medical University, Taipei 11031, Taiwan;
rohmah.noor29@gmail.com or noormayasari@unesa.ac.id (N.R.M.)
2 Department of Nutrition, Faculty of Sports and Health Sciences, Universitas Negeri Surabaya,
Surabaya 60213, Indonesia
3 School of Public Health, College of Public Health, Taipei Medical University, Taipei 11031, Taiwan
4 Department of Public Health, School of Medicine, College of Medicine, Taipei Medical University,
Taipei 11031, Taiwan
5 Department of Internal Medicine, Yangming Branch, Taipei City Hospital, Taipei 11146, Taiwan
6 Department of Biostatistics and Epidemiology, Faculty of Medicine, Public Health and Nursing,
Universitas Gadjah Mada, Yogyakarta 55281, Indonesia
7 Graduate Institute of Metabolism and Obesity Sciences, College of Nutrition, Taipei Medical University,
Taipei 11031, Taiwan
8 Nutrition Research Center, Taipei Medical University Hospital, Taipei 11031, Taiwan
9 Chinese Taipei Society for the Study of Obesity (CTSSO), Taipei 100, Taiwan
10 TMU Research Center for Digestive Medicine, Taipei Medical University, Taipei 11031, Taiwan
* Correspondence: susanchang@tmu.edu.tw; Tel.: +886-(2)-27361661 (ext. 6542); Fax: +886-(2)-2737-3112
Citation: Mayasari, N.R.; Bai, C.-H.;
Chao, J.C.-J.; Chen, Y.-C.; Huang, Abstract: Globally, anemia affects 56 million pregnant women, especially women with a low house-
Y.-L.; Wang, F.-F.; Wiratama, B.S.; hold income. Functional erythropoiesis requires a constant supply of micronutrients, and the de-
Chang, J.-S. Relationships between mands significantly increase during fetal development. This study aims to identify dietary patterns
Dietary Patterns and for preventing gestational erythropoiesis-associated micronutrient deficiencies (e.g., iron, folic acid,
Erythropoiesis-Associated and vitamin B12 ). A Nationwide Nutrition and Health Survey in Pregnant Women, Taiwan (NAHSIT-
Micronutrient Deficiencies (Iron, PW), was conducted between 2017 and 2019. Data on baseline information, diet, anthropometrics,
Folate, and Vitamin B12 ) among
and blood biochemistry were collected during a prenatal visit. Dietary patterns were identified using
Pregnant Women in Taiwan.
a reduced rank regression (RRR). Erythropoiesis-related micronutrient deficiencies were defined
Nutrients 2023, 15, 2311.
as single, double, and triple micronutrient deficiencies of an iron deficiency, folate depletion, and
https://doi.org/10.3390/
a vitamin B12 deficiency. In total, 1437 singleton pregnancies aged ≥20–48 years were included in
nu15102311
the analysis. Prevalences of normal nutrition, and single, double, and triple erythropoiesis-related
Academic Editors: Gladys micronutrient deficiencies were 35.7%, 38.2%, 18.6%, and 7.5%, respectively. Anemic pregnant women
Oluyemisi Latunde-Dada and
with a low household income had the highest prevalence rates of double (32.5%) and triple (15.8%)
Jose Lara
erythropoiesis-related micronutrient deficiencies. Dietary pattern scores were positively correlated
Received: 6 March 2023 with nuts and seeds, fresh fruits, total vegetables, breakfast cereals/oats and related products, soy-
Revised: 5 May 2023 bean products, and dairy products but negatively correlated with processed meat products and liver,
Accepted: 13 May 2023 organs, and blood products. After adjusting for covariates, the dietary pattern had 29% (odds ratio
Published: 15 May 2023 (OR): 0.71; 95% confidence interval (CI): 0.055–0.091, p = 0.006)) and 43% (OR: 0.57; 95% CI: 0.41–0.80,
p = 0.001)) reduced odds of having double and triple erythropoiesis-related micronutrient deficiencies
for those pregnant women with a low household income. For those women with anemia, dietary
patterns had 54% (OR: 046, 95% CI: 0.27–0.78) and 67% (OR: 0.33; 95% CI: 0.170.64) reduced odds
Copyright: © 2023 by the authors.
of double and triple erythropoiesis-related micronutrient deficiencies. In conclusion, increased con-
Licensee MDPI, Basel, Switzerland.
This article is an open access article
sumption of breakfast cereals and oats, nuts, and seeds, fresh fruits and vegetables, soybean products,
distributed under the terms and and dairy products may protect women against erythropoiesis-related micronutrient deficiencies
conditions of the Creative Commons during pregnancy.
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/ Keywords: dietary pattern; erythropoiesis-associated micronutrient deficiencies; pregnant women;
4.0/). household income; anemia

Nutrients 2023, 15, 2311. https://doi.org/10.3390/nu15102311 https://www.mdpi.com/journal/nutrients

Nutrients 2023, 15, 2311 2 of 14

1. Introduction
Globally, anemia affects 56 million pregnant women in 2021 [1], especially women
with a low household income level [2]. It was demonstrated that erythropoiesis triggers at
a very early stage and half of the increase may occur within the first trimester using a direct
measurement of red cell mass using a nonradioactive approach [3]. The erythropoietic
system needs a constant supply of micronutrients (e.g., iron, vitamin (vit.) B12 , and folic
acid), and these demands further increase during pregnancy in order to support fetal
growth and development. For example, erythroblasts need iron for hemoglobin (Hb)
synthesis and folate and vit. B12 for proliferation during their differentiation [4].
In developing countries, prevalence rates of erythropoiesis-related micronutrient de-
ficiencies among pregnant women in rural areas were 67.7%, 26.3%, and 74.1%, for iron,
folate, and vit. B12 , respectively [5]. Deficiencies in erythropoiesis-related micronutrients
may lead to the development of gestational anemia [6], including iron deficiency (ID)
anemia (IDA) or megaloblastic anemia [7]. An insufficient iron supply can result in mi-
crocytic hypochromic anemia. Megaloblastic anemia is caused by a deficiency of folic
acid, vit. B6 , and vit. B12 . Deficiencies of those components of the vit. B complex lead to
abnormal cell division and maturation manifested as a larger megaloblast, the precursors
of red blood cells (RBCs) in the bone marrow, and enlarged (macrocytic) erythrocytes in
the blood [8]. Importantly, deficiencies of iron, folate, and vit. B12 not only cause anemia
but are also associated with several short- and long-term health risks for both the mother
and child [9–13]. A low maternal folate status, especially with the concomitant presence
of a vit. B12 deficiency, is recognized as a risk for neural tube defects, intrauterine growth
retardation, preterm birth, and a low birth weight [9–11]. Gestational ID is associated
with fetal growth restriction, offspring obesity, high blood pressure later in life [12], and
impairment of neurodevelopment [13].
Insufficient dietary intake is the major cause of erythropoiesis-related micronutrient
deficiencies [14]. The presence of micronutrient deficiencies during early gestation suggests
that preexisting stores of micronutrients (e.g., iron and folate) were inadequate prior to
pregnancy. Alternatively, the intake of micronutrients, either through dietary or supple-
mental sources, is insufficient to meet the demands of fetal development. Recommended
dietary allowances (RDAs) of pregnant women for iron, folate, and vit. B12 are 27 mg/day,
600 µg/day, and 2.6 mg/day, respectively [15]. Pregnant women consuming vegetable-
or plant-based diets are at increased risk of erythropoiesis-related micronutrient deficien-
cies [16]. Unlike other micronutrients, vit. B12 is mainly derived from meat products, and
consuming a plant-based diet increases the risk of a vit. B12 deficiency. Meat also contains
heme iron, which is more easily absorbed than nonheme iron [17]. Legumes, grains, nuts,
seeds, fruits, and vegetables supply good amounts of iron, vit. B6 , and folate [18]. However,
they also contain phytate, which reduces nonheme iron bioavailability. The concomitant
presence of vit. C increases nonheme iron bioavailability [18,19]. Nonetheless, use of
prenatal supplements must be considered for the wellbeing of the fetus, especially for
pregnant women whose dietary patterns are mainly vegetable- or plant-based [16].
Household income is a primary economic aspect that affects dietary preferences and
the nutritional status of individuals [20]. Results from the National Health and Nutrition
Examination Survey (NHANES) from 1999 to 2002 in the United States (US) showed that
lower proportions of low-income US adults ate above or equal to the adequate intake
(AI) or estimated average requirement (EAR) intake levels for many micronutrients [20].
Data from 18 publications originating primarily from western Europe showed a positive
relationship between socioeconomic status and micronutrient status [21]. Specifically,
women in resource-poor settings often had low micronutrient intake, such as folate, vit. B12 ,
vit. C, vit. A, and Fe (below the EAR) [14,22]. Sociodemographic factors also determined
dietary pattern adherence during pregnancy, for which a healthy diet was associated with
household income [23]. A survey in China found that urban pregnant women had greater
Nutrients 2023, 15, 2311 3 of 14

food diversity than did rural pregnant women [24]. Overall, those studies highlighted
the importance of socioeconomic inequalities in diet-related diseases especially among
pregnant women.
The use of a dietary pattern analysis to understand nutritional risk factors affecting
pregnancy outcomes is becoming more popular [25]. Dietary pattern analyses take into
account cumulative and interactive effects among dietary food groups to reflect the com-
plexity of the human diet [26]. By studying nationwide representative population data in
Taiwan, the broad aims of this study were to investigate (1) the prevalence of erythropoiesis-
related micronutrient deficiencies among pregnant women stratified by household income
and anemic status, and (2) the identification of dietary patterns with protective effects
against erythropoiesis-related micronutrient deficiencies, especially for women with low
household income and an anemic status.

2. Materials and Methods

2.1. Data Source and Participants
A population-scale cross-sectional survey, the Nationwide Nutrition and Health Sur-
vey in Taiwanese Pregnant Women (NAHSIT-PW), was conducted between 2017 and 2019.
Details of the study design can be seen in previous publications [7,19]. Briefly, inclusion
criteria were pregnant women aged ≥ 15 years, with Taiwanese residency who spoke fluent
Chinese or Taiwanese, had received a maternal health checkout booklet, and for whom the
legal guardian had a signed written informed agreement for those who were ≤19 years of
age. For this analysis, we excluded participants who were ≤19 years of age (n = 6), who
had nonsingleton pregnancies (n = 33), and who had missing information on household
income (n = 32). In total, 1437 respondents were included in the current analysis.

2.2. Ethics
The study protocol was approved by the institutional review board of the Taipei Medical
University (TMU-JIRB N201707039). Eligible participants provided written informed consent.

2.3. Data Collection

Baseline data on pregnant women were collected using a self-reported questionnaire
including household income, residential area, anthropometrics, the use of prenatal sup-
plements (such as multivitamins, vit. B12 , folate, and iron), and disease history before
and during the pregnancy. A face-to-face interview with an experienced dietitian was
conducted to collect a 24 h dietary recall (24-HDR) and food frequency questionnaire (FFQ).
The ratio of weight (in kg)-to-height squared (in m2 ) was used as the basis for the estimation
of the pre-pregnancy body mass index (pBMI). To assess nutritional status, blood samples
were taken during prenatal care.

2.4. Blood Biochemistry and Definition of Erythropoiesis-Related Micronutrient Deficiencies

Each participant underwent a phlebotomy in the hospital’s blood collection room. A
complete blood cell count and Hb content in whole blood were measured (Hematology
Analyzer, Sysmex, Kobe, Japan). Serum was used to determine contents of iron (e.g.,
iron, ferritin, and hepcidin), folate, and vit. B12 . Serum ferritin (chemiluminescence
immunoassay using a Beckman Coulter Unicel DxC 800 (Beckman Coulter, Brea, CA,
USA)), total iron-binding capacity (TIBC) (immunoturbidimetric method by Beckman
Coulter), and serum iron (ferrozine-based colorimetric assay using a Beckman Coulter
Unicel DxC 800) were analyzed at Le Zen Clinical Laboratory (Taipei, Taiwan). Transferrin
saturation (TS) was estimated as serum iron divided by the TIBC as a percentage (%).
Hepcidin was measured using a Duo-Set enzyme-linked immunosorbent assay (ELISA)
(R&D System, Minneapolis, MN, USA).
Trimesters (Trs) were defined as (1) Tr1, 17 weeks of gestation that follows the last
normal menstrual cycle; (2) Tr2, weeks 18–28; and (3) Tr3, weeks 29–40 based on recommen-
dations of the Ministry of Health and Welfare, Taiwan. According to the Center for Disease
Nutrients 2023, 15, 2311 4 of 14

Control and Prevention (CDC), anemia was defined as Hb of <110 g/L in the first and third
trimesters and Hb of <105 g/L in the second trimester. Following our previous study, ID
was defined as TS of <16% without serum ferritin as biomarker to avoid ‘false-negative’
diagnoses among pregnant women experiencing chronic inflammation (such as obesity) [7].
Folate depletion was defined as serum folate of 6 ng/mL. A vit. B12 deficiency was defined
as a serum vit. B12 level of <203 pg/mL [7]. Erythropoiesis-related micronutrient deficien-
cies were defined as (1) normal: none of the tested micronutrient deficiencies was present;
and (2) single, double, and triple (multiple) micronutrient deficiencies: the presence of one,
two, or three of the following: ID, folate depletion, or a vit. B12 deficiency.

2.5. Dietary Assessment

Dietary intake was assessed using a 24-HDR. Briefly, participants were asked to recall
detailed information about all food and beverages they had consumed in the past 24 h.
Detailed information (e.g., meal type, meal time, food sources, items consumed, and
cooking techniques) was also collected to help determine the daily intake of nutrient levels.
All dietary data were standardized based on the Taiwan Food Nutrient Database, using
the internet-based program Cofit Pro (Cofit Healthcare, Taipei, Taiwan). An interviewer-
administered FFQ was conducted, and FFQ data were used to generate dietary pattern
scores. Respondents were asked how often (daily, weekly, and monthly) they had consumed
each food item in the previous month. The NAHSIT-PW 2017–2019 uses a 59-item FFQ to
appraise the dietary status of the preceding 1-month period. The detailed FFQ is described
elsewhere [19]. To generate dietary pattern scores, 59 food items were merged into 33 food
groups according to the similarity of food characteristics and nutritional components.

2.6. Statistical Analysis

We calculated the mean ± standard deviation (SD) for continuous data and number
(percentage) for categorical data. General linear models for continuous data and χ-squared
test for categorical data were employed to generate p for trend values. We subsequently
applied a reduced rank regression (RRR) using the PROC PLS function in SAS 9.4 (SAS,
Cary, NC, USA) to derive dietary patterns predictive of erythropoiesis-related micronutrient
deficiencies. Selected nutrient intakes were adjusted using the nutrient density model [27]
and not normally distributed response variables were logarithmically transformed before
the RRR procedure. Briefly, the RRR identified linear functions of predictors (e.g., 33 food
groups) that explained as much of the response variation as possible (e.g., folate, household
income levels, protein (%), vit. C, and TS (%)). We only selected the first factor which
accounted for the highest variation among the response variables. Dietary pattern scores
were categorized into tertiles (Ts), with the highest group (T3) conforming most closely
to the identified dietary pattern. Participants’ characteristics were summarized according
to the T of the dietary pattern score. The odds of erythropoiesis-related micronutrient
deficiencies were examined using a multinomial logistic regression, which generated
adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the risk of erythropoiesis-
related micronutrient deficiencies. p for trend was conducted using generalized linear
models. Data were analyzed using GraphPad Prism 5 (GraphPad Prism 5, GraphPad
Software, San Diego, CA, USA) and SPSS ver. 21 (IBM, Armonk, NY, USA) with p for trend
and p-value < 0.05 considered significant.

3. Results
3.1. Relationships of Anemia and Household Income with Erythropoiesis-Related
Micronutrient Deficiencies
Overall prevalences of normal status, and single, double, and triple erythropoiesis-related
micronutrient deficiencies were 35.7%, 38.2%, 18.6%, and 7.5%, respectively. Figure 1A shows
that compared to nonanemic pregnant women, anemic pregnant women had higher rates
of double (30.9% vs.14.5%) and triple (13.9% vs. 5.2%) erythropoiesis-related micronutrient
deficiencies. We next stratified women according to anemia and household income levels
 Deficiencies
Overall prevalences of normal status, and single, double, and triple erythropoiesis-
related micronutrient deficiencies were 35.7%, 38.2%, 18.6%, and 7.5%, respectively. Fig-
ure 1A shows that compared to nonanemic pregnant women, anemic pregnant women
Nutrients 2023, 15, 2311 had higher rates of double (30.9% vs.14.5%) and triple (13.9% vs. 5.2%) erythropoiesis- 5 of 14
related micronutrient deficiencies. We next stratified women according to anemia and
household income levels (high (≥USD2000) and low (<USD2000). Among nonanemic preg-
(high
nant (≥USD2000)
women, women and lowa(<USD2000).
with high householdAmong nonanemic
income had thepregnant
highestwomen, women
prevalence rate of a
with a status
normal high household income
(47.6% vs. 36.9%)had thethe
but highest prevalence
lowest rates of rate of a (12.5%
double normal status (47.6%
vs. 16.2%) andvs. triple
36.9%) but the lowest rates of double (12.5% vs. 16.2%) and triple erythropoiesis-related
erythropoiesis-related micronutrient deficiencies (3.7% vs. 6.4%) compared with those
micronutrient deficiencies (3.7% vs. 6.4%) compared with those with a low household
with a low household income (Figure 1B,C). Among anemic pregnant women, women
income (Figure 1B,C). Among anemic pregnant women, women with a low household
with a low household income had the highest prevalence rates of double (32.5% vs. 28%)
income had the highest prevalence rates of double (32.5% vs. 28%) and triple (15.8% vs.
and triple
10.4%) (15.8% vs. 10.4%) erythropoiesis-related
erythropoiesis-related micronutrient
micronutrient deficiencies compared to deficiencies compared
those with a high
to household
those withincome
a highlevel
household income
(Figure 1B,C). level (Figure 1B,C).

Figure 1. Cont.
Nutrients 2023, 15, x FOR PEER REVIEW 6 of 14
Nutrients 2023, 15, 2311 6 of 14

Prevalences of
Figure1.1.Prevalences
Figure of erythropoiesis-related
erythropoiesis-related micronutrient
micronutrient deficiencies in relation
deficiencies to household
in relation to household
incomeand
income andthe
the anemic
anemic status
statusamong
amongpregnant
pregnant women.
women. (A) (A)
The Thetotal total
population. (B) Household
population. (B) Household
monthlyincome
monthly income levels of≥≥USD2000.
levels of USD2000. (C) (C)Household
Household monthly
monthlyincomeincomelevelslevels
of <USD2000. Defini-Defini-
of <USD2000.
tions: erythropoiesis-related micronutrient deficiencies: (1) normal: none of the tested
tions: erythropoiesis-related micronutrient deficiencies: (1) normal: none of the tested micronutrient micronutrient
deficiencies
deficiencieswere
werepresent; (2)single
present; (2) single(one),
(one),double
double (two),
(two), andand triple
triple (three)
(three) micronutrient
micronutrient deficiencies:
deficiencies:
thethe
presence
presenceof ofone,
one, two, orthree
two, or threeofofthe
thefollowing:
following: ironiron deficiency
deficiency (ID),(ID),
folatefolate depletion
depletion or a vitamin
or a vitamin
(vit.) B12B12
(vit.) deficiency.
deficiency.Anemia:
Anemia:hemoglobin
hemoglobin (Hb) (Hb) << 11
11 g/dL
g/dL(in(in the
the first
first and thirdtrimesters),
and third trimesters),andand Hb <
10.5
Hbg/dL
< 10.5(in the(in
g/dL second trimester);
the second ID: transferrin
trimester); ID: transferrin saturation
saturation (TS) < 15%;
(TS) < 15%;folate
folatedepletion:
depletion: serum
folate
serum< 6folate
ng/mL; vit. B12 deficiency:
< 6 ng/mL; serumserum
vit. B12 deficiency: vit. B12vit.
< 203
B12 <pg/mL.
203 pg/mL.

3.2. Erythropoiesis-Associated Dietary Pattern Scores Using the RRR

3.2. Erythropoiesis-Associated Dietary Pattern Scores Using the RRR
Erythropoiesis-related dietary pattern scores were derived using the RRR from
Erythropoiesis-related
33 food dietarywere
groups. Response variables pattern scores income,
household were derived using %TS,
serum folate, the RRR from 33
dietary
food
vit. C, and % protein intake (Supplementary Table S1), which were selected based on a vit.
groups. Response variables were household income, serum folate, %TS, dietary
C,literature
and % protein
review. intake (Supplementary Table S1), which were selected based on a litera-
ture review.
Table 1 shows the first dietary pattern scores, of which response variables explained
15.17%
Tableof 1the variation.
shows Dietary
the first pattern
dietary scores
pattern were positively
scores, of which correlated with nuts explained
response variables and
seeds, fresh fruits, total vegetables, breakfast cereals/oats and related products,
15.17% of the variation. Dietary pattern scores were positively correlated with nuts and soybean
products, and dairy products (factor loadings of ≥0.20) but negatively correlated with
seeds, fresh fruits, total vegetables, breakfast cereals/oats and related products, soybean
processed meat products and liver, organs, and blood products (factor loadings of ≤−0.20).
products, and dairy products (factor loadings of ≥0.20) but negatively correlated with pro-
cessed meat products
3.3. Maternal and liver,
Characteristics organs,
Stratified and blood productsDietary
by Erythropoiesis-Associated (factorPattern
loadings of ≤−0.20).
Scores
Table 2 shows that pregnant women with the lowest dietary pattern scores (T1) were
Table 1. Food
younger, groups
heavier, associated
and had the with therates
highest first factor
of lessofthan
erythropoiesis-related dietary (23.9%)
an undergraduate degree pattern scores
identified
and low using a reduced
household rank
incomes regression
(69.5%) (all p (RRR).
for trend < 0.05). Dietary pattern scores were neg-
atively correlated with the prevalence of erythropoiesis-related micronutrient deficiencies
Explained Variation (%) Factor Loading *
but positively correlated with blood biomarkers (serum iron, %TS, vit. B12 , and folic acid)
Food groups (all p for trend < 0.05). Significant positive linear trends were also observed among dietary
Positive association pattern scores, self-reported prenatal dietary supplements used (total supplement and
- Nut and seeds 29.33 0.48
- Fresh fruits 23.22 0.43
- Total vegetables 12.64 0.31
- Breakfast cereals/oats and related products 9.50 0.27
- Soybean products 7.28 0.24
Nutrients 2023, 15, 2311 7 of 14

multivitamin/minerals), and adjusted nutrient intake levels (percentage of protein intake,

fiber intake, total sugar, folate, dietary vit. B12 , vit. C, and vit. A) (all p for trend < 0.05).

Table 1. Food groups associated with the first factor of erythropoiesis-related dietary pattern scores
identified using a reduced rank regression (RRR).

Explained Variation (%) Factor Loading *

Food groups
Positive association
- Nut and seeds 29.33 0.48
- Fresh fruits 23.22 0.43
- Total vegetables 12.64 0.31
- Breakfast cereals/oats and related products 9.50 0.27
- Soybean products 7.28 0.24
- Dairy products 6.57 0.23
Negative association
- Processed meat products 10.57 −0.29
- Liver, organs, and blood products 5.02 −0.20
Total explained variation 104.13
* Factor loadings are correlations between food groups and dietary pattern scores (correlation coefficient for the
RRR-derived pattern ≥ |0.20|).

Table 2. Characteristic of study participants according to tertile (T) groups of erythropoiesis-related

dietary pattern scores.

Dietary Pattern Scores #

Variable p for Trend
T1 (N = 479) T2 (N = 479) T3 (N = 479)
Basic characteristic
Age (years) 31.63 ± 4.93 32.54 ± 4.56 33.51 ± 4.38 <0.001
pBMI (kg/m2 ) 22.87 ± 4.38 22.78 ± 4.14 22.25 ± 3.35 0.016
Underweight (n, %) 51 (10.7) 44 (9.2) 45 (9.4) 0.129
Normal weight (n, %) 282 (59.2) 284 (59.5) 312 (65.3)
Overweight (n, %) 73 (15.3) 77 (16.1) 73 (15.3)
Obesity (n, %) 70 (14.7) 72 (15.1) 48 (10.0)
Trimester
First (n, %) 129 (26.9) 118 (24.6) 111 (23.2) 0.198
Second (n, %) 151 (31.5) 157 (32.8) 155 (32.4)
Third (n, %) 199 (41.5) 204 (42.6) 213 (44.5)
Educational level
Less than undergraduate (n, %) 114 (23.9) 67 (14.0) 35 (7.3) <0.001
Undergraduate or above (n, %) 362 (76.1) 410 (86.0) 443 (92.7)
Household income level (USD)
<2000 (n, %) 333 (69.5) 260 (54.3) 232 (48.4) <0.001
≥2000 (n, %) 146 (30.5) 219 (45.7) 247 (51.6)
Blood biomarkers
Serum hepcidin (ng/mL) 23.03 ± 32.56 24.84 ± 32.43 24.21 ± 32.14 0.573
Serum iron (µg/dL) 67.23 ± 37.13 75.08 ± 43.06 74.37 ± 39.37 0.006
Log TS (%) 15.48 ± 9.61 17.39 ± 10.38 16.84 ± 9.87 0.034
Hemoglobin (g/dL) 11.69 ± 2.01 11.82 ± 2.07 11.72 ± 1.74 0.813
Serum ferritin (ng/mL) 22.92 ± 27.99 24.86 ± 26.32 21.96 ± 22.07 0.564
Folic acid (ng/mL) 10.92 ± 6.84 13.07 ± 7.42 14.19 ± 6.98 <0.001
Log vit. B12 (pg/mL) 309.59 ± 242.63 316.39 ± 207.02 316.95 ± 138.02 0.031
Nutrients 2023, 15, 2311 8 of 14

Table 2. Cont.

Dietary Pattern Scores #

Variable p for Trend
T1 (N = 479) T2 (N = 479) T3 (N = 479)
Erythropoiesis-related nutritional
deficiencies
Normal (n, %) 139 (29.0) 174 (36.3) 201 (42.0) <0.001
Single (n, %) 178 (37.2) 198 (41.3) 176 (36.7)
Double (n, %) 103 (21.5) 78 (16.3) 85 (17.7)
Triple (n, %) 59 (12.3) 29 (6.1) 17 (3.5)
Reported use of prenatal dietary
supplements
Total supplement use (n, %) 364 (77.4) 409 (86.1) 427 (90.1) <0.001
Multivitamin/mineral (n, %) 250 (53.2) 311 (65.8) 325 (69.1) <0.001
Vit. B (n, %) 88 (18.8) 84 (18.1) 87 (18.4) 0.872
Folate (n, %) 217 (46.0) 221 (47.3) 214 (45.6) 0.860
Iron (n, %) 46 (9.8) 42 (9.0) 63 (13.4) 0.073
Adjusted nutrients intake status
Carbohydrates (g/day) 124.15 ± 25.14 124.67 ± 24.56 126.29 ± 23.28 0.174
Protein (g/day) 37.62 ± 9.61 38.10 ± 8.88 39.03 ± 9.47 0.019
Fat (g/day) 40.27 ± 9.98 40.01 ± 10.21 39.09 ± 9.49 0.065
Log dietary Fe (mg/day) 5.75 ± 4.54 5.99 ± 3.83 6.02 ± 2.60 0.001
Log dietary folate (mg/day) 113.27 ± 69.05 117.65 ± 53.71 119.77 ± 48.03 0.006
Log dietary vit. B12 (µg/day) 2.80 ± 5.12 2.99 ± 4.49 3.41 ± 7.02 0.02
Log dietary vit. C (mg/day) 54.00 ± 66.04 63.56 ± 67.75 78.09 ± 83.41 <0.001
Log dietary vit. A (µg RE/day) 404.17 ± 805.96 436.52 ± 692.95 508.50 ± 577.59 <0.001
Under RDA of protein (n, %) 195 (41.0) 199 (41.6) 158 (33.0) 0.011
Under RDA of Fe (n, %) 441 (92.6) 433 (90.6) 440 (91.9) 0.660
Under RDA of folate (n, %) 459 (96.4) 457 (95.6) 459 (95.8) 0.637
Under RDA of vit. B12 (n, %) 231 (48.5) 209 (43.7) 186 (38.8) 0.003
Under RDA of vit. C (n, %) 356 (74.8) 317 (66.3) 252 (52.6) <0.001
Continuous data are represented as the mean ± standard deviation, whereas categorical data are represented as
the number (percentage of the same group). p for trend was tested using a generalized linear model/one-way
ANOVA for continuous data and chi-squared for categorical data. # Dietary pattern scores in tertiles (Ts): Tertile 1
(T1) = −5.55–0.41; Tertile 2 (T2) = −0.40–0.27; Tertile 3 (T3) = 0.28–5.62. p for trend was tested using a general linear
model/one-way ANOVA for continuous data and chi-squared for categorical data. Abbreviations: pBMI, pre-
pregnancy body-mass index; TS: transferrin saturation; vit.: vitamin; RDA: recommended daily amount. In 2019,
the average exchange rate was 1.0 USD ≈ 30 New Taiwanese Dollars (TWD). Between 2017 and 2019, the average
monthly household income per capita was TWD13,855–14,332 ≈ USD1155–1194. Definitions: erythropoiesis-
related micronutrient deficiencies: (1) normal: no nutritional deficiency was present; (2) single (one), double
(two), and triple (three) micronutrient deficiencies: the presence of one, two, or three iron deficiencies (ID); folate
depletion; or a Vit. B12 deficiency. ID: TS < 15%, folate depletion: serum folate < 6 ng/mL; Vit. B12 deficiency:
serum Vit. B12 < 203 pg/mL.

3.4. Protective Effects of Dietary Patterns on Erythropoiesis-Related Micronutrient Deficiencies in

Relation to Household Income and Anemic Status
We next performed a multinomial logistic regression analysis to investigate the protec-
tive effect of dietary patterns on the risk of erythropoiesis-related micronutrient deficiencies.
After adjusting for covariates (age, pBMI, region, trimester, parity, percentage of total di-
etary supplement, and calorie intake), dietary pattern scores showed 22% (OR: 0.78; 95% CI:
0.65–0.94, p = 0.008) and 37% (OR: 0.63; 95% CI: 0.49–0.82, p = 0.001) reduced odds of having
double and triple erythropoiesis-related nutritional deficiencies, respectively (Figure 2, all
populations). Among pregnant women with low household income, dietary pattern scores
showed 29% (OR: 0.71; 95% CI: 0.055–0.091, p = 0.006)) and 43% (OR: 0.57; 95% CI: 0.41–0.80,
p = 0.001)) reduced odds of having double and triple erythropoiesis-related nutritional
deficiencies after adjusting for covariates (age, pBMI, region, trimester, parity, percentage
of total dietary supplement, and calorie intake) (Figure 2). No significant protective effects
of dietary patterns were detected for pregnant women with a high household income after
adjusting for covariates.
Nutrients 2023,15,
2023,
Nutrients 15,2311
x FOR PEER REVIEW 9 of 14 9 of 14
Nutrients 2023, 15, x FOR PEER REVIEW 9 of 14

Adjusted
Figure2.2.Adjusted
Figure oddsodds ratios
ratios (ORs)(ORs)
and 95%andof95% of confidence
confidence intervalsintervals (CI) of erythropoiesis-related
(CI) of erythropoiesis-related
micronutrient
micronutrient deficiencies
deficiencies with with dietary
dietary pattern pattern
scores as scores as astratified
a predictor predictor stratified income.
by household by household in-
Figure 2. Adjusted odds ratios (ORs) and 95% of confidence intervals (CI) of erythropoiesis-related
ORs wereORs
come. calculated
were using a multinomial
calculated using a logistic regression.
multinomial p forregression.
logistic trend was tested
p forusing a general-
trend was tested using
micronutrient deficiencies with dietary pattern scores as a predictor stratified by household income.
ized linear model, which was adjusted for age, pre-pregnancy body mass index, region, trimester,
a generalized
ORs linear
were calculated model,
using which was
a multinomial adjusted
logistic for age,
regression. p for pre-pregnancy body amass
trend was tested using index, region,
general-
parity, (%) total supplement, and calorie intake. In 2019, the average exchange rate was USD1.0 
ized linear model, which was adjusted for age, pre-pregnancy bodyInmass index, region, trimester,
TWD30. Between 2017 and 2019, the average monthly household income per capita was TWD13, rate was
trimester, parity, (%) total supplement, and calorie intake. 2019, the average exchange
parity, (%) total supplement, and calorie intake. In 2019, the average exchange rate was USD1.0 
USD1.0 ≈TWD30.
0.855–14.33 Between
USD1155–1194. 2017 and
p value/p 2019,as
for trend the
** average
p ≤ 0.01. monthly household income per capita was
TWD30. Between 2017 and 2019, the average monthly household income per capita was TWD13,
TWD13, 0.855–14.33 ≈ USD1155–1194. p
0.855–14.33  USD1155–1194. p value/p for trend as ** p ≤for
value/p trend as ** p ≤ 0.01.
0.01.
Figure 3 shows that for those pregnant women with anemia, erythropoiesis-related
dietary patterns
Figure
Figure hadthat
3 shows
3 shows 54%
that(OR:
for for046,
those 95%pregnant
those CI: women
pregnant 0.27–0.78),
women and
with 67% (OR:
with
anemia, 0.33;erythropoiesis-related
anemia, 95% CI: 0.17–
erythropoiesis-related
0.64)
dietaryreduced
dietarypatterns odds
patternshad of
had developing
54%54% (OR:
(OR: double
046,046,
95%95% and
CI: CI:triple erythropoiesis-related
0.27–0.78),
0.27–0.78), and and
67% 67%
(OR: (OR: micronutrient
0.33; 0.33; 95%
95% CI: CI: 0.17–0.64)
0.17–
deficiencies,
reduced
0.64) reduced respectively
oddsodds (all p < 0.01).
ofofdeveloping
developing doubleDietary
double pattern
andtriple
and triple scores also protected
erythropoiesis-related
erythropoiesis-related nonanemic
micronutrient de-
micronutrient
pregnant women
ficiencies,
deficiencies, against (all
respectively
respectively double
(all (OR:
p <p 0.01).0.74;
< 0.01). 95% CI: 0.055–0.099,
Dietary
Dietary patternpattern p = protected
scores scores
also 0.043)
alsoand triple (OR:
protected
nonanemic nonanemic
0.63; 95% CI: 0.40–0.98, p = 0.042) erythropoiesis-related micronutrient deficiencies.
pregnant women against double (OR: 0.74; 95% CI: 0.055–0.099, p = 0.043) and triple (OR:
pregnant women against double (OR: 0.74; 95% CI: 0.055–0.099, p = 0.043) and triple (OR:
0.63;
0.63;95% CI:CI:
95% 0.40–0.98, p = 0.042)
0.40–0.98, erythropoiesis-related
p = 0.042) micronutrient
erythropoiesis-related deficiencies.
micronutrient deficiencies.

Figure 3. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of erythropoiesis-related
micronutrient deficiencies
Adjusted odds with dietary pattern
ratios scores as a predictor stratified(CIs)
by anemia. ORs were
Figure
Figure 3. 3.
Adjusted odds ratios (ORs)(ORs) andconfidence
and 95% 95% confidence
intervalsintervals of erythropoiesis-related
(CIs) of erythropoiesis-related
calculated using a multinomial logistic regression. p for trend was tested by a generalized linear
micronutrient
micronutrient deficiencies
deficiencies withwith dietary
dietary pattern
pattern scores scores as a predictor
as a predictor stratifiedstratified
by anemia.by anemia.
ORs were ORs were
model, which was adjusted for age, region, trimester, and total supplement use (%). Anemia was
calculated
calculated using a multinomial
using a multinomial logistic regression. p for trend
p for was tested by a generalized linear
defined as hemoglobin (Hb) levels oflogistic
<11 g/dLregression.
(in the first and trendtrimesters),
second was tested byHb
and a generalized
levels of linear
model, which was adjusted for age, region, trimester, and total supplement use (%). Anemia was
model,
<10.5 g/dLwhich
(in thewas adjusted
second for age,
trimester). region,
p value/p trimester,
for trend as * p ≤and
0.05,total
** p ≤supplement
0.01. use (%). Anemia was
defined as hemoglobin (Hb) levels of <11 g/dL (in the first and second trimesters), and Hb levels of
defined
<10.5 g/dLas(inhemoglobin (Hb) levels
the second trimester). of <11for
p value/p g/dL
trend(in
asthe
* p ≤first
0.05,and
** psecond
≤ 0.01. trimesters), and Hb levels of
<10.5 g/dL (in the second trimester). p value/p for trend as * p ≤ 0.05, ** p ≤ 0.01.
Nutrients 2023, 15, 2311 10 of 14

4. Discussion
By studying the micronutrient status of a nationwide representative 1430 pregnant
women, the present study shows that the prevalence of erythropoiesis-related micronutri-
ent deficiency was greater among pregnant women with anemia or low household income
compared to those who did not have those features. We identified erythropoiesis-related
dietary patterns, mainly plant-based foods (nuts and seeds, fresh fruits, total vegetables,
breakfast cereals/oats and related products, soybean products, and dairy products, and
a low intake frequency of processed meat and blood/liver/organs and related products)
that can supply adequate micronutrients for erythropoiesis in the bone marrow and re-
duced the risk of micronutrient deficiencies during pregnancy. Importantly, the protective
effect of the dietary pattern was higher for pregnant women with low household income
and anemia.
It is well established that a constant supply of micronutrients such as iron, folate, and
vit. B12 are essential for erythropoiesis, and inadequate supplies of any of these may increase
the risk of gestational anemia [8]. In many developing countries, prevalence rates of anemia
are up to 75%, and micronutrient deficiencies, especially iron deficiency, are believed to be
the main underlying cause of anemia [6]. The current study results agree with findings from
the US National Health and Nutritional Survey [28]. The current study found that among
pregnant women, prevalences of single (38.2%), double (18.6%), and triple micronutrient
deficiencies (7.5%) were similar to those of US pregnant or breastfeeding women, of which
the prevalences of one, two, and multiple (≥3–5) micronutrient deficiencies were 30%, 13%,
and 1.5%, respectively [28]. However, the prevalence rate of multiple (triple) erythropoiesis-
related micronutrient deficiencies among Taiwanese pregnant women (7.5%) was lower
than that of Indian pregnant women living in a rural area (16.2%) [5]. In agreement with a
community based-study in India [5], anemic pregnant women (82.4%) had a higher rate of
developing micronutrient deficiencies than nonanemic pregnant women (58.4%).
The identified erythropoiesis-related micronutrient dietary pattern scores protected the
women against the development of double and triple erythropoiesis-related micronutrient
deficiencies (p < 0.05). Specifically, a one-unit increase in the dietary pattern score was
respectively associated with 22% and 37% reduced risks of having double and triple
erythropoiesis-related micronutrient deficiencies during pregnancy. This suggests that
pregnant women should be encouraged to consume plant-based foods (e.g., nuts and seeds,
fresh fruits, vegetables, breakfast cereals/oats, and soybean products) and dairy products
but should decrease the intake frequency of processed meats and blood/liver/organ-related
products. In general, erythropoiesis-related micronutrient dietary patterns were similar to
recommendations of a healthy diet for pregnant women. Pregnant women are encouraged
to eat a variety of nutrient-dense foods from plant products as these natural food products
are rich in iron minerals and vitamins and also protein and α-linoleic acid compared to
those highly processed meat products [29].
Erythropoiesis-related micronutrient dietary pattern scores accounted for 104.13% of
the total explained variation, with nuts and seeds the strongest at 29.33%, followed by
fresh fruits (23.22 %), vegetables (12.64%), and breakfast cereal and oat-related products
at 9.50% (Table 1). Nuts and seeds were positively correlated with all response vari-
ables, except for %TS (all p < 0.05) (Supplementary Table S2). Fresh fruits were positively
associated with serum folate (ß = 0.022) and vit. C intake frequency (ß = 0.922) but neg-
atively correlated with %TS (ß = −0.027) (Supplementary Table S2). A previous cluster
study [18] showed that nuts and seeds, fruits, vegetables, and cereal grains contain iron
(minimum (min)–maximum (max): 1.64–14.55, 0–1.02, 0.06–8.9, and 0.8–7.61 mg) and folate
(min–max: 11–22, 0–89, 3–338, and 7–281 µg), vit. C (min–max: 0–46.6, 0.4–228, 0–242.5,
0–4.2 mg) per 100 g, respectively. A study among 725 older men and 705 older women
from the Nutrition and Health Survey in Taiwan (1999–2000) (Elderly-NASHIT) found
that vegetables (66%) and fruits (11.8%) were the main sources of folate [30]. This suggests
that a higher folate intake is achievable due to the abundance of a diverse variety of fruit
and vegetables in Taiwan [30]. Dairy products accounted for 6.57% of the total explained
Nutrients 2023, 15, 2311 11 of 14

variation (Table 1), and this food group was significantly correlated with response variables,
except for %TS (Supplementary Table S2). Gille and Schimd showed that dairy products
and meats are sources of vit. B12 [31]. Processed meat products explained 10.57% of the
total explained variation (Table 1), but they were significantly negatively correlated with
%TS, serum folate, and household income levels (Supplementary Table S2). The Biomarker
Reflecting Inflammation and Nutritional Determinant of Anemia (BRINDA) project re-
vealed that higher red and processed meat consumption was positively associated with
increased levels of proinflammatory markers among overweight and obese women [32].
Adiposity-mediated low-grade inflammation is known to reduce iron bioavailability due
to increased hepcidin levels that block iron absorption and iron efflux [33,34].
The present study found that dietary vit. C contributed 3.98% to the total explained
variation of response variables (Supplementary Table S1). Vitamin C intake is known
to influence blood iron status. Vitamin C is an iron enhancer agent that can improve
the nonheme iron bioavailability of plant-based foods such as dark leafy vegetables,
fruits, and cereals and protects against IDA [19]. We found a strong positive trend be-
tween vit. C intake and dietary pattern scores (p < 0.001). We observed that one unit
increase in fresh fruit frequency significantly increased 0.922 mg intake of vit. C (p < 0.001)
(Supplementary Table S2). Abdul-Fattah and colleagues conducted a cross-sectional study
among pregnant women and found that poor dietary intake of vegetables, fresh fruits, and
dry fruits was associated with IDA [35].
Our study found that household income levels were positively associated with dietary
pattern scores (ß = 0.335, 95% CI: 0.214–0.455, p < 0.001). Specifically, we found that dietary
pattern scores protected against double (OR: 0.71; 95% CI: 0.55–0.91, p = 0.006) and triple
(OR: 0.57; 95% CI: 0.41–0.80, p = 0.001) erythropoiesis-related micronutrient deficiencies
among pregnant women with low household income but not high household income.
This is in agreement with De Castro and colleagues, who conducted a cross-sectional
study among pregnant women and found that monthly per capita family income was
a factor related to adherence to a healthy dietary pattern (mostly consisting of legumes,
vegetables, and fruits) [23]. This suggests that women with a high household income level
(≥USD2000) would more likely adhere to a healthy dietary pattern than those with a low
household income level, and this may protect them against micronutrient deficiencies
during pregnancy. Bowman et al. also found a positive relationship between income
and micronutrients intake [20]. In low-income level settings, a possible solution is to
increase dietary diversity [24], especially by diversifying planned food consumption [18].
The positive correlation observed between micronutrient consumption and socioeconomic
status in our study highlights the importance of socioeconomic disparities in diet-related
diseases among disadvantaged populations [21].
Another important finding of our study was the protective effect of the dietary pattern
among anemic and nonanemic pregnant women. The present study found that, among
pregnant women with anemia, dietary pattern scores showed 54% and 67% reduced risks
of double (OR: 0.46, 95% CI: 0.27–0.78, p = 0.004) and triple (OR: 0.33, 95% CI: 0.17–0.064,
p = 0.001) erythropoiesis-related micronutrient deficiencies, while among nonanemic preg-
nant women, dietary pattern scores showed 26% and 37% reduced risks for double
(OR: 0.74, 95% CI: 0.55–0.99, p = 0.043) and triple (OR: 0.63, 95% CI: 0.40–0.98, p = 0.042)
erythropoiesis-related micronutrient deficiencies, respectively. A previous study also sug-
gested that women who did not maintain World Health Organization-recommended levels
of adequate fruit and vegetable intake (five or more servings/day) had an increased risk of
being moderately to severely anemic [36]. Paramastri et al. [37] and Kurniawan et al. [38]
identified an anemia-inflammatory dietary pattern, characterized by high intake frequen-
cies of eggs, meat, processed foods, and sugary beverages, but lower intake of vegetables
and fruits. Taken together, those studies emphasize the importance of the consumption of a
diversity of plant-based foods during pregnancy.
Our current study includes a large sample of pregnant women representing the
Taiwanese population, which allowed us to have more precise and robust estimates of the
Nutrients 2023, 15, 2311 12 of 14

relationship between dietary patterns and erythropoiesis-related micronutrient deficiencies.

However, our study also had several limitations that need to be taken into account when
translating the findings. First, our cross-sectional study design was unable to address
causality. Second, we included only single 24 h recall data, which might not eliminate recall
error. Third, our study relied on self-administered dietary data (e.g., an FFQ), for which
systematic underreporting or recall bias of energy intake may have been present. It was
discovered that pregnant women overreported their food intake when using an FFQ as a
dietary survey tool [39]. Nonetheless, in large-scale epidemiological studies, the FFQ is
regarded as one of the most appropriate methods for dietary assessment, and it presents
useful information on dietary patterns [40]. Our study used transferrin saturation without
ferritin to define ID, which may affect the accuracy of the diagnosis. A previous study
suggested that transferrin saturation has 61% sensitivity and 86% specificity compared
to serum ferritin or bone marrow examination as the gold standard [41]. However, our
definition may also reduce false negative diagnoses of ID for those who are elevated ferritin
concentration because of chronic inflammation.

5. Conclusions
In conclusion, the erythropoiesis-associated dietary pattern, mostly plant-based foods,
characterized by high intake frequencies of nuts and seeds, fresh fruits, total vegetables,
breakfast cereals/oats and related products, soybean products, and dairy products, may
prevent the development of micronutrient deficiencies during pregnancy, especially for
women with low household income or an anemic status.

Supplementary Materials: The following supporting information can be downloaded at: https:
//www.mdpi.com/article/10.3390/nu15102311/s1, Table S1: Response variables identified by the
reduced rank regression; Table S2: Beta (β) coefficients and 95% confidence intervals (CIs) of response
variables for individual for food groups of erythropoiesis-related dietary pattern scores.
Author Contributions: J.-S.C., C.-H.B., J.C.-J.C., Y.-C.C., Y.-L.H. and F.-F.W. designed the search
strategy and translated the findings. N.R.M. conducted the data analysis and wrote the initial
manuscript. B.S.W. performed data visualization. J.-S.C. carefully reviewed the manuscript for
relevant intellectual content and approved the final version of the manuscript. All authors have read
and agreed to the published version of the manuscript.
Funding: This research was funded by Taipei Medical University Hospital (111TMU-TMUH-05-4)
and the Ministry of Science and Technology, Taiwan (MOST111-2320-B-038-030-MY3 and MOST
111-2410-H-038-019-MY2).
Institutional Review Board Statement: This study was conducted in accordance with the guidelines
laid down in the Declaration of Helsinki. The study protocol was approved by the Institutional review
board of the Taipei Medical University (TMU-JIRB N201707039). Eligible participants provided
written informed consent.
Informed Consent Statement: Each participant involved in this study was provided informed consent.
Data Availability Statement: Upon reasonable request, the corresponding author will provide data
described in the manuscript, code book, and analytic code.
Acknowledgments: We would like to express our gratitude to the Ministry of Health and Welfare,
Taiwan for allowing us to access the NAHSIT-PW 2017-2019 data. Jung-Su Chang was financed by
grants from the Ministry of Science and Technology, Taiwan (MOST111-2320-B-038-030-MY3 and
MOST 111-2410-H-038-019-MY2) and Taipei Medical University Hospital (111TMU-TMUH-05-4).
Conflicts of Interest: The authors declare that they have no conflict of interest.

Abbreviations
vit.: vitamin; Hb: hemoglobin; ID: iron deficiency; IDA: iron deficiency anemia; RBCs: red
blood cells; NHANES: the National Health and Nutrition Examination Survey; US: the United States;
AI: adequate intake; EAR: estimated average requirement; NAHSIT-PW: the Nationwide Nutri-
Nutrients 2023, 15, 2311 13 of 14

tion and Health Survey in Taiwanese Pregnant Women; 24-HRD: 24-h dietary recall; FFQ: food
frequency questionnaire; pBMI: pre-pregnancy body mass index; TIBC: total iron-binding capacity;
TS: transferrin saturation; ELISA: enzyme-linked immunosorbent assay; Trs: trimesters; CDC: Center
for Disease Control and Prevention; SD: standard deviation; RRR: reduced rank regression; Ts: tertiles;
ORs: odds ratios; CIs: confidence intervals; NAHSIT: the Nutrition and Health Survey in Taiwan;
BRINDA: the biomarker reflecting inflammation and nutritional determinant of anemia.

References
1. World Health Organization (WHO). The Global Prevalence of Anaemia in 2011; WHO: Genewa, Switzerland, 2015.
2. Balarajan, Y.; Ramakrishnan, U.; Özaltin, E.; Shankar, A.H.; Subramanian, S.V. Anaemia in low-income and middle-income
countries. Lancet 2011, 378, 2123–2135. [CrossRef]
3. Cavill, I. Iron and erythropoiesis in normal subjects and in pregnancy. J. Perinat. Med. 1995, 23, 47–50. [CrossRef]
4. Koury, M.J.; Ponka, P. NEW INSIGHTS INTO ERYTHROPOIESIS: The Roles of Folate, Vitamin B12 , and Iron. Annu. Rev. Nutr.
2004, 24, 105–131. [CrossRef]
5. Pathak, P.; Kapil, U.; Yajnik, C.S.; Kapoor, S.K.; Dwivedi, S.N.; Singh, R. Iron, folate, and Vitamin B12 stores among pregnant
women in a rural area of Haryana State, India. Food Nutr. Bull. 2007, 28, 435–438. [CrossRef]
6. Van den Broek, N. Anaemia and micronutrient deficiencies: Reducing maternal death and disability during pregnancy. Br. Med.
Bull. 2003, 67, 149–160. [CrossRef]
7. Mayasari, N.R.; Hu, T.-Y.; Chao, J.C.J.; Bai, C.-H.; Chen, Y.C.; Huang, Y.L.; Chang, C.-C.; Wang, F.-F.; Hadi, H.; Nurwanti, E.; et al.
Associations of the pre-pregnancy weight status with anaemia and the erythropoiesis-related micronutrient status. Public Health
Nutr. 2021, 24, 6247–6257. [CrossRef]
8. Beckman, B.; Silberstein, P.; Aldoss, I.T. Erythropoiesis. In xPharm: The Comprehensive Pharmacology Reference; Enna, S.J., Bylund,
D.B., Eds.; Elsevier: New York, NY, USA, 2010; pp. 1–4. [CrossRef]
9. Smits, L.J.M.; Essed, G.G.M. Short interpregnancy intervals and unfavourable pregnancy outcome: Role of folate depletion. Lancet
2001, 358, 2074–2077. [CrossRef] [PubMed]
10. Mishra, J.; Tomar, A.; Puri, M.; Jain, A.; Saraswathy, K.N. Trends of folate, Vitamin B12 , and homocysteine levels in different
trimesters of pregnancy and pregnancy outcomes. Am. J. Hum. Biol. 2020, 32, e23388. [CrossRef]
11. Furness, D.; Fenech, M.; Dekker, G.; Khong, T.Y.; Roberts, C.; Hague, W. Folate, Vitamin B12 , Vitamin B6 and homocysteine:
Impact on pregnancy outcome. Matern. Child. Nutr. 2013, 9, 155–166. [CrossRef] [PubMed]
12. Alwan, N.A.; Hamamy, H. Maternal iron status in pregnancy and long-term health outcomes in the offspring. J. Pediatr. Genet.
2015, 4, 111–123. [PubMed]
13. Hernández-Martínez, C.; Canals, J.; Aranda, N.; Ribot, B.; Escribano, J.; Arija, V. Effects of iron deficiency on neonatal behavior at
different stages of pregnancy. Early Hum. Dev. 2011, 87, 165–169. [CrossRef] [PubMed]
14. Lee, S.E.; Talegawkar, S.A.; Merialdi, M.; Caulfield, L.E. Dietary intakes of women during pregnancy in low- and middle-income
countries. Public Health Nutr. 2013, 16, 1340–1353. [CrossRef] [PubMed]
15. Jouanne, M.; Oddoux, S.; Noël, A.; Voisin-Chiret, A.S. Nutrient Requirements during Pregnancy and Lactation. Nutrients 2021, 13, 692.
[CrossRef]
16. Karabulut, A.; Sevket, O.; Acun, A. Iron, folate and Vitamin B12 levels in first trimester pregnancies in the Southwest region of
Turkey. J. Turk. Ger. Gynecol. Assoc. 2011, 12, 153–156. [CrossRef] [PubMed]
17. Young, M.F.; Griffin, I.; Pressman, E.; McIntyre, A.W.; Cooper, E.; McNanley, T.; Harris, Z.L.; Westerman, M.; O'Brien, K.O.
Utilization of Iron from an Animal-Based Iron Source Is Greater Than That of Ferrous Sulfate in Pregnant and Nonpregnant
Women. J. Nutr. 2010, 140, 2162–2166. [CrossRef]
18. Li, Y.; Bahadur, R.; Ahuja, J.; Pehrsson, P.; Harnly, J. Macro-and micronutrients in raw plant foods: The similarities of foods and
implication for dietary diversification. J. Food Compost. Anal. 2021, 102, 103993. [CrossRef]
19. Mayasari, N.R.; Bai, C.-H.; Hu, T.-Y.; Chao, J.C.-J.; Chen, Y.C.; Huang, Y.L.; Wang, F.-F.; Tinkov, A.A.; Skalny, A.V.; Chang, J.-S.
Associations of Food and Nutrient Intake with Serum Hepcidin and the Risk of Gestational Iron-Deficiency Anemia among
Pregnant Women: A Population-Based Study. Nutrients 2021, 13, 3501. [CrossRef]
20. Bowman, S. Low economic status is associated with suboptimal intakes of nutritious foods by adults in the National Health and
Nutrition Examination Survey 1999–2002. Nutr. Res. 2007, 27, 515–523. [CrossRef]
21. Novaković, R.; Cavelaars, A.; Geelen, A.; Nikolić, M.; Altaba, I.I.; Viñas, B.R.; Ngo, J.; Golsorkhi, M.; Medina, M.W.; Brzozowska,
A.; et al. Socio-economic determinants of micronutrient intake and status in Europe: A systematic review. Public Health Nutr.
2014, 17, 1031–1045. [CrossRef]
22. Torheim, L.E.; Ferguson, E.L.; Penrose, K.; Arimond, M. Women in Resource-Poor Settings Are at Risk of Inadequate Intakes of
Multiple Micronutrients. J. Nutr. 2010, 140, 2051S–2058S. [CrossRef]
23. de Castro, M.B.T.; Freitas Vilela, A.A.; Oliveira, A.S.D.d.; Cabral, M.; Souza, R.A.G.d.; Kac, G.; Sichieri, R. Sociodemographic
characteristics determine dietary pattern adherence during pregnancy. Public Health Nutr. 2016, 19, 1245–1251. [CrossRef]
Nutrients 2023, 15, 2311 14 of 14

24. Gao, H.; Stiller, C.K.; Scherbaum, V.; Biesalski, H.K.; Wang, Q.; Hormann, E.; Bellows, A.C. Dietary Intake and Food Habits of
Pregnant Women Residing in Urban and Rural Areas of Deyang City, Sichuan Province, China. Nutrients 2013, 5, 2933–2954.
[CrossRef]
25. Chen, X.; Zhao, D.; Mao, X.; Xia, Y.; Baker, P.N.; Zhang, H. Maternal Dietary Patterns and Pregnancy Outcome. Nutrients 2016, 8, 351.
[CrossRef]
26. Zhao, J.; Li, Z.; Gao, Q.; Zhao, H.; Chen, S.; Huang, L.; Wang, W.; Wang, T. A review of statistical methods for dietary pattern
analysis. Nutr. J. 2021, 20, 37. [CrossRef]
27. Willett, W.C.; Howe, G.R.; Kushi, L.H. Adjustment for total energy intake in epidemiologic studies. Am. J. Clin. Nutr. 1997, 65,
1220S–1228S, discussion 1229S–1231S. [CrossRef]
28. Bird, J.K.; Murphy, R.A.; Ciappio, E.D.; McBurney, M.I. Risk of Deficiency in Multiple Concurrent Micronutrients in Children and
Adults in the United States. Nutrients 2017, 9, 655. [CrossRef]
29. Marshall, N.E.; Abrams, B.; Barbour, L.A.; Catalano, P.; Christian, P.; Friedman, J.E.; Hay, W.W., Jr.; Hernandez, T.L.; Krebs, N.F.;
Oken, E.; et al. The importance of nutrition in pregnancy and lactation: Lifelong consequences. Am. J. Obstet. Gynecol. 2022, 226,
607–632. [CrossRef] [PubMed]
30. Chen, K.J.; Pan, W.H.; Shaw, N.S.; Huang, R.F.; Lin, B.F. Association between dietary folate-rich food intake and folate status of
elderly Taiwanese. Asia Pac. J. Clin. Nutr. 2005, 14, 244–249. [PubMed]
31. Gille, D.; Schmid, A. Vitamin B12 in meat and dairy products. Nutr. Rev. 2015, 73, 106–115. [CrossRef]
32. Shiraseb, F.; Hosseininasab, D.; Mirzababaei, A.; Bagheri, R.; Wong, A.; Suzuki, K.; Mirzaei, K. Red, white, and processed meat
consumption related to inflammatory and metabolic biomarkers among overweight and obese women. Front. Nutr. 2022, 9, 1015566.
[CrossRef] [PubMed]
33. Sangkhae, V.; Nemeth, E. Regulation of the Iron Homeostatic Hormone Hepcidin. Adv. Nutr. 2017, 8, 126–136. [CrossRef]
[PubMed]
34. Köse, S.; Sözlü, S.; Bölükbaşi, H.; Ünsal, N.; Gezmen-Karadağ, M. Obesity is associated with folate metabolism. Int. J. Vitam. Nutr.
Res. 2020, 90, 353–364. [CrossRef]
35. Abdul-Fatah, B.N.; Murshid, R.M.; Ahmed, T.E. Assessment of iron deficiency anemia (IDA) and dietary pattern among pregnant
women in Baghdad city, Iraq. J. Pharm. Sci. Res. 2018, 10, 2279–2284.
36. Ghose, B.; Yaya, S. Fruit and vegetable consumption and anemia among adult non-pregnant women: Ghana demographic and
health survey. PeerJ 2018, 6, e4414. [CrossRef]
37. Paramastri, R.; Hsu, C.Y.; Lee, H.A.; Lin, L.Y.; Kurniawan, A.L.; Chao, J.C. Association between Dietary Pattern, Lifestyle,
Anthropometric Status, and Anemia-Related Biomarkers among Adults: A Population-Based Study from 2001 to 2015. Int. J.
Environ. Res. Public Health 2021, 18, 3438. [CrossRef]
38. Kurniawan, A.L.; Hsu, C.Y.; Rau, H.H.; Lin, L.Y.; Chao, J.C. Inflammatory Dietary Pattern Predicts Dyslipidemia and Anemia in
Middle-Aged and Older Taiwanese Adults with Declined Kidney Function: A Cross-Sectional Population Study from 2008 to
2010. Nutrients 2019, 11, 2052. [CrossRef]
39. Robinson, S.; Godfrey, K.; Osmond, C.; Cox, V.; Barker, D. Evaluation of a food frequency questionnaire used to assess nutrient
intakes in pregnant women. Eur. J. Clin. Nutr. 1996, 50, 302–308. [PubMed]
40. Crozier, S.R.; Inskip, H.M.; Godfrey, K.M.; Robinson, S.M. Dietary patterns in pregnant women: A comparison of food-frequency
questionnaires and 4 d prospective diaries. Br. J. Nutr. 2008, 99, 869–875. [CrossRef]
41. Thompson, W.G.; Meola, T.; Lipkin, M., Jr.; Freedman, M.L. Red Cell Distribution Width, Mean Corpuscular Volume, and
Transferrin Saturation in the Diagnosis of Iron Deficiency. Arch. Intern. Med. 1988, 148, 2128–2130. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.