European Journal of Nutrition (2019) 58:2011–2018

https://doi.org/10.1007/s00394-018-1759-1

ORIGINAL CONTRIBUTION

Teff consumption and anemia in pregnant Ethiopian women:

a case–control study
Shimels Hussien Mohammed1   · Hailu Taye2 · Tesfamichael Awoke Sissay3 · Bagher Larijani4 ·
Ahmad Esmaillzadeh5,6,7

Received: 6 March 2018 / Accepted: 20 June 2018 / Published online: 23 June 2018
© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Purpose  Teff, an iron-rich staple grain in Ethiopia, is consumed mainly in the form of teff injera (TI). Studies on the associa-
tion of TI consumption with anemia are limited. We aimed to determine the association of frequencies of TI consumption
with anemia, in Addis Ababa, Ethiopia.
Methods  We conducted a hospital-based case–control study involving 592 pregnant women: 187 anemic cases, and 405
non-anemic controls. TI consumption was assessed by a food frequency questionnaire (FFQ). Multiple logistic regression,
adjusted for dietary and non-dietary covariates, was performed to determine the relation of TI consumption to anemia status.
Results  We found that a decrease in frequency of TI consumption was significantly associated with an increase in the like-
lihood of anemia (p-trend = 0.009). Compared with everyday TI consumption, the adjusted odds ratios (AORs) of anemia
associated with the other frequencies of teff consumption were 1.50 [95% confidence interval (CI) 0.71, 3.23; p = 0.285] for
every other day, 2.13 (95% CI 1.03, 4.41; p = 0.04) for 1–2 times a week, and 3.17 (95% CI 1.62, 6.22; p < 0.001) for once
in 2 weeks.
Conclusions  Teff consumption was associated with reduced odds of anemia in pregnant women. Further studies are war-
ranted, including determining the feasibility and effectiveness of TI use for anemia prevention.
Registration  The study was registered as https​://clini​caltr​ials.gov/ct2/show/NCT03​25166​4.

Keywords  Teff · Eragrostis tef · Anemia · Iron deficiency anemia

* Ahmad Esmaillzadeh
a‑esmaillzadeh@tums.ac.ir Introduction
1
Department of Community Nutrition, School of Nutritional Anemia remains among the major threats to the health and
Sciences and Dietetics, Tehran University of Medical
Sciences-International Campus (TUMS-IC), Tehran, Iran survival of pregnant women across many regions of the
2 world [1]. Progress in the reduction of anemia in pregnant
Unit of Reproductive Health, Addis Ababa City
Administration Health Bureau, Addis Ababa, Ethiopia women has been less promising over the last three decades
3 [1], with an estimated 40.6% prevalence of anemia among
Department of Public Health, Wollo University, Dessie,
Ethiopia pregnant women worldwide in the year 2016. The figure
4 was 43.4% in 1990 [2]. Anemia during pregnancy is a sig-
Endocrinology and Metabolism Research Center,
Endocrinology and Metabolism Clinical Sciences Institute, nificant contributing factor to maternal mortality and poor
Tehran University of Medical Sciences, Tehran, Iran fetal outcomes [1, 3, 4]. Prevention and timely treatment of
5
Obesity and Eating Habits Research Center, Endocrinology anemia in pregnant women is a priority public health agenda
and Metabolism Molecular Cellular Sciences Institute, in Ethiopia, where the maternal mortality rate and anemia
Tehran University of Medical Sciences, Tehran, Iran during pregnancy were estimated to be 412 per 100,000 live
6
Department of Community Nutrition, School of Nutritional births and 23%, respectively, in the year 2016 [5].
Sciences and Dietetics, Tehran University of Medical Pregnant women are at a higher risk of developing ane-
Sciences, Tehran, Iran mia, due to various dietary and non-dietary risk factors.
7
Department of Community Nutrition, Food Security The increased physiological demand for nutrients during
Research Center, Isfahan University of Medical Sciences, pregnancy also makes them more vulnerable to anemia.
Isfahan, Iran

13
Vol.:(0123456789)

2012
However, iron deficiency is the most significant contributor
to anemia worldwide [6, 7]. Infection, chronic inflamma-
tion, intestinal parasites, and iron malabsorption disorders
are among the main factors leading to anemia [6, 7]. Cereal-
based diet is often associated with increased risk of anemia,
one of the reasons being its high phytate content. Phytic acid
impairs iron absorption by sequestering the iron in food as
a result of the formation of an insoluble phytate–iron com-
plex. Plant food items contain only the less absorbable iron
form, non-heme iron, which could be another reason for the
high risk of anemia among individuals with a cereal-based
diet [6–8].
Global interest in dietary diversification and fortification
to prevent anemia, preferably using locally available food
items, has increased recently [9]. However, evidence on the
health outcomes of traditional diets in developing coun-
tries, particularly in Africa, is limited. For example, TI con-
sumed only in Ethiopia and some parts of Eritrea, is widely
believed to reduce the risk anemia [10]. TI is the local name
of a flat pancake made of Eragrostis tef, a tropical grain
cultivated mainly in highland areas of East Africa [11, 12].
Despite being the primary staple grain in Ethiopia, teff has
remained largely unknown to the rest of the world. Recently,
research interest in teff has increased noticeably mainly due
to its high iron [10, 12, 13] and gluten-free protein contents
[10–12, 14]. Some studies showed that iron deficiency ane-
mia (IDA) was low in Ethiopia [15, 16], a condition believed
to be partly attributable to teff consumption. On average,
100 gm of teff contains 9 mg of iron, which is more than
the amount in other cereal grains such as rice, wheat, corn,
and sorghum [10, 13]. Teff also contains high fiber, zinc,
calcium, copper, phytochemicals, and antioxidants [10–12].
Previous studies on teff in Ethiopia have focused mainly
on understanding its nutrient content and phytochemical
properties. Studies on its health effects, particularly studies
on human subjects, are limited. To the best of our knowl-
edge, there have been only three studies that examined hema-
tologic response to teff consumption among human subjects,
one conducted in Ethiopia [17] and two in the United King-
dom [18, 19]. The studies conducted in the United Kingdom
were limited in sample size, using less than 50 participants.
Besides, the teff products evaluated by these studies were
different from one widely used in Ethiopia, TI. In this study,
presuming that teff reduces the risk of anemia, we aimed to
determine the association of various frequencies of TI con-
sumption with anemia among pregnant Ethiopian women.
The findings of the study may serve as input for further stud-
ies on the suitability and feasibility of TI use for prevention
and control of anemia.
13
European Journal of Nutrition (2019) 58:2011–2018
Materials and methods
Study setting and population
The study was conducted in Addis Ababa, Ethiopia, in four
health facilities (hospitals and health centers) providing
antenatal care (ANC) for pregnant women. Data were col-
lected from October 2015 to February 2016.
Study design and sample size
We conducted a hospital-based, unmatched case–control
study. We recruited a total of 592 pregnant women, of
whom 187 were cases and 405 controls. The sample size
was estimated based on the assumptions: case to control
ratio 1:2, expected proportion of daily teff consumption
as staple grain among cases 27% and controls 40%, power
80%, two-tailed 95% CI, and 15% non-response rate.
Selection of cases and controls
Classification into cases and controls was achieved based
on the participants’ anemia status at the time of their
recruitment into the study, from October 2015 to February
2016. Cases were those pregnant women diagnosed with
anemia and controls were those diagnosed not anemic.
Hemoglobin concentration (Hb) < 11.0 g/dL, adjusted for
altitude, was used in defining anemia status. Blood sam-
ples were collected and analyzed using HemoCue Hb®201
(HemoCue AB, Ängelholm, Sweden). According to the
World Health Organization (WHO) definition, pregnant
women with Hb < 11.0 g/dL are classified as anemic [20].
While providing regular ANC services, health profession-
als of participating hospitals identified potentially eligible
women who were further contacted by a trained data col-
lector and a nurse of the maternal service unit. The pre-
defined inclusion criteria were (a) laboratory-confirmed
pregnancy status, (b) de facto residence in Addis Ababa
city, and (c) Hb status measured. We excluded those with
any of the following conditions: (a) serious medical condi-
tion requiring in-patient management, (b) history of blood
donation within one year before the study, (c) history of
abortion or trauma within one year before the study, or
(d) currently breastfeeding. Anemic women (cases) who
fulfilled the eligibility criteria were included until the
sample size was achieved. For each case, two consecutive
controls were selected from those who were diagnosed as
non-anemic and fulfilled the eligibility criteria.
European Journal of Nutrition (2019) 58:2011–2018 2013
Exposure assessment
Data collection was conducted by trained data collectors,
together with the midwives or nurses of the participating
health facilities. The main exposure of interest was the
frequency of TI consumption in the year before the study.
‘Injera’ can be prepared from various grains or a mixture
of grains. In this work, we were particularly interested
in injera made of mainly teff flour (> 50% teff content),
referred to in this document as TI. Moreover, a compre-
hensive assessment of anemia determinants was performed
using a pre-tested questionnaire that was adapted from the
Ethiopian Demographic and Health Survey [5] as well
as the World Food and Agriculture Organization (FAO)
dietary assessment [21] tools. Data were collected on
the sociodemographic, anthropometric, dietary, hygiene,
reproductive, and health condition of the study partici-
pants. Weight, height, and mid-upper arm circumference
(MUAC) were measured following the standard procedures
[22]. Stool and blood samples were taken for laboratory
examination, the former for intestinal parasites and the
later for Human Immunodeficiency Virus (HIV) and Hb.
Dietary practice was assessed by a food frequency ques-
tionnaire (FFQ), using the FAO women dietary diversity
scoring (WDDS) tool. The WDDS is a simplified tool
designed for a rapid assessment of micronutrient intake
adequacy. The tool has been widely used by previous stud-
ies in Ethiopia as well as validated in similar setups as a
proxy measure of micronutrient intake [21, 23]. The FAO
WDDS tool categorizes food items into nine groups, namely
(1) cereals, roots, and tubers; (2) legumes and nuts; (3) meat,
poultry, and fish; (4) vitamin A-rich fruits and vegetables;
(5) other vitamin A-rich fruits and vegetables; (6) other fruit
and vegetables; (7) dairy; (8) fats and oils; and (9) eggs.
Participants were asked to recall the usual frequency of con-
sumption of those food groups, including TI, over the last
one year before the date of data collection.
Statistical analysis
Data entry and cleaning was achieved using Epi Info™ Pro-
gram version 3.4 (Center for Disease Control and Preven-
tion, Atlanta, Georgia, United States). Statistical analysis
was conducted using Statistical Package for the Social Sci-
ences (SPSS) Version 23 (IBM Corporation, New York,
United States), with significance determined at a two-tailed
p value < 0.05. Adjusted odds ratios (AORs), with 95% CI,
are reported on the association of frequency of TI consump-
tion with anemia.
Before performing bivariable and multivariable analy-
ses, relevant statistical assumptions were checked, includ-
ing normality test for continuous variables, homogeneity of
variance, multicollinearity among continuous independent
variables, and linearity of the relationship between the inde-
pendent variables and the logit transformation of anemia.
Bivariable analyses were conducted using independent t test
for continuous variables and Chi-square test for categori-
cal variables. The bivariable tests were done to identify the
potential covariates (confounders) which might affect the
association of the variables of interest (TI and anemia).
Sociodemographic, disease, or dietary factors which dem-
onstrated p values ≤ 0.20 during bivariable analyses, were
considered as potential confounders, such that the final
multiple logistic regression analysis was adjusted for these
factors. Interaction of teff with other food items of known
anemia influence was also checked, with the aim of provid-
ing estimates by subgroups if interaction was found. Where
no adequate evidence of interaction was observed, only main
variables were entered in the multiple regression model.
Results
Characteristics of study participants
Table 1 presents the sociodemographic, anthropometric, and
health characteristics of study participants, disaggregated
by cases and controls. In total, 592 pregnant women were
included; 187 anemic cases and 405 non-anemic controls.
There were significant discrepancies between cases and con-
trols by education level, household income, MUAC, and ges-
tational age. Controls were more educated than cases (mean
education: 11 and 9 years among controls and cases, respec-
tively, p < 0.001). The mean monthly household income (in
Ethiopian Birr) of cases was 2966, whereas that of control
was 3159 (p < 0.001). The mean MUAC (cm) of cases was
lower than that of controls (23.6 vs. 24.3, p < 0.001). A
greater proportion of cases (42.2%) than controls (23.2%)
were in the first trimester of gestation; whereas a greater
proportion of controls (38.3%) than cases (10.7%) were in
the second trimester. There were no significant differences
between cases and controls concerning age, height, weight,
religion, occupation, household size, hygiene practice, gra-
vidity, pregnancy interval, HIV status, chronic illness, iron
supplement use, or intestinal parasites.
Table 2 presents comparisons of the frequency of con-
sumption of food groups between cases and controls. A
higher proportion of anemic cases ate TI less frequently
than controls (teff consumption once per 2 weeks: 41.7%
cases vs. 20.5% controls), whereas a greater proportion of
controls ate teff more frequently than cases (everyday teff
consumption: 38.5% controls vs. 20.3% cases). Consump-
tions of dark green leafy vegetables and meat were also
less frequent among cases compared to controls (p < 0.05).
There was no evidence of a significant difference (p > 0.05)
13

2014 European Journal of Nutrition (2019) 58:2011–2018

Table 1  Comparison Variable Cases, n (%) Controls, n (%) pa

of sociodemographic,
anthropometric and health Age group (years)
characteristics of cases (anemic
 15–19 8 (4.3) 12 (3.0) 0.833
pregnant women) and controls
(non-anemic pregnant women),  20–24 52 (27.8) 117 (28.9)
Addis Ababa, 2016  25–29 74 (39.6) 154 (38.0)
 30–34 25 (13.4) 65 (16.0)
 35–39 28 (15.0) 57 (14.1)
Education (years) 8.6 (3.5)b 10.5 (3.8) < 0.001
 Religion
  Muslim 41 (21.9) 78 (19.3) 0.422
  Christian 132 (70.6) 305 (75.3)
  Other 14 (7.5) 22 (5.4)
Occupation
 Employed 39 (20.9) 75 (18.5) 0.919
 Merchant 49 (26.2) 107 (26.4)
 Housewife 57 (30.5) 126 (31.1)
 Student 23 (12.3) 59 (14.6)
 Other 19 (10.2) 38 (9.4)
Hygiene ­practicec
 Poor 15 (8.0) 36 (8.9) 0.430
 Improved 172 (92.0) 369 (91.1)
Household size 4.04 (1.2) 4.1 (1.2) 0.877
Household income (­ Birrd) 2965 (486.9) 3159 (522.5) < 0.001
Height (cm) 163.6 (7.6) 164.0 (8.1) 0.581
Weight (kg) 56.5 (7.4) 57.3 (7.7) 0.224
Mid-upper arm circumference (cm) 23.6 (1.9) 24.3 (1.8) < 0.001
Gestational age (trimester)
 First 79 (42.2) 94 (23.2) < 0.001
 Second 20 (10.7) 155 (38.3)
 Third 88 (47.1) 156 (38.5)
Gravidity (including abortion) 2.2 (1.1) 2.2 (1.06) 0.830
Pregnancy interval (from last birth, years) 2.2 (1.4) 2.3 (1.4) 0.497
Chronic ­illnesse
 No 176 (94.1) 390 (96.3) 0.262
 Yes 11 (5.9) 15 (3.7)
Intestinal parasites
 No 174 (93.0) 383 (94.6) 0.466
 Yes 13 (7.0) 22 (5.4)
Iron ­supplementf
 No 155 (82.9) 317 (78.3) 0.294
 Yes 32 (17.1) 88 (21.7)
a
 Chi-square test or independent t test was used to determine the differences between cases and controls;
i.e., independent t test for household size, household income, height, weight, mid-upper arm circumference,
and gravidity; and Chi-square test for the remaining categorical variables
b
 Mean ± SD (all such values)
c
 Measured by the use of pipe water and improved toilet facilities. ‘Improved’, if the household had both
piped water supply and improved toilet facility within the compound; otherwise ‘Poor’
d
 Ethiopian currency
e
 Tuberculosis or Human Immunodeficiency Virus (HIV)
f
 Use of prescribed iron supplement. ‘Yes’, if the individual used iron supplement for at least 1 month dur-
ing the current pregnancy; otherwise ‘No’

13
European Journal of Nutrition (2019) 58:2011–2018 2015

Table 2  Comparison of the frequency of consumption of food groups Table 2  (continued)

between cases (anemic pregnant women) and controls (non-anemia c
 Statistics not applicable as consumption was universal (all partici-
pregnant women), Addis Ababa (2016) pants reported the same frequency of consumption)
Food group (frequency of) Cases, n (%) Controls, n (%) pa

Teff injera
 Every day 38 (20.3) 156 (38.5) < 0.001 between cases and controls concerning the consumption of
 Every other day 29 (15.4) 86 (21.2) the other food groups shown in the table.
 1–2 times a week 42 (22.5) 80 (19.8) To examine whether eating TI was independently associ-
 Once per 2 weeks 78 (41.7) 83 (20.5) ated with anemia, we calculated AORs for the association of
Cereal ­grainsb frequencies of TI consumption with anemia. The adjustment
 Every day 187 (100) 405 (100) NAc was done for both dietary and non-dietary covariates, namely
Meat educational status, household income, MUAC, and con-
 Every day 17 (9.1) 75 (18.5) < 0.001 sumption of meat and dark green leafy vegetables. Results
 Every other day 34 (18.2) 132 (32.6) are shown in Table 3. Overall, a decrease in the frequency of
 1–2 times a week 54 (28.9) 161 (39.8)
TI consumption demonstrated a significant association with
 Once per 2 weeks 82 (43.9) 37 (9.1)
an increase in anemia likelihood (p for trend = 0.009). Com-
Legumes
pared to the most frequent category (everyday TI consump-
 Every day 187 (100) 405 (100) NA
tion), the odds of anemia associated with the least frequent
Dark green leafy vegetables
 Every day 15 (8) 57 (14.1) < 0.001
category (once every 2 weeks) were 3.17 times higher (95%
 Every other day 28 (15) 117 (28.9) CI 1.62, 6.22; p < 0.001). The AORs for the other catego-
 1–2 times a week 45 (24.1) 195 (48.1) ries of TI consumption, compared to everyday consumption,
 Once per 2 weeks 99 (52.9) 36 (8.9) were 2.13 (95% CI 1.03, 4.41; p = 0.04) for 1–2 times in
Vitamin A-rich fruits and vegetables a week, and 1.50 (95% CI 0.71, 3.23; p = 0.285) for every
 Every day 12 (6.4) 18 (4.4) 0.511 other day.
 Every other day 25 (13.4) 70 (17.3)
 1–2 times a week 113 (60.4) 236 (58.3)
 Once per 2 weeks 37 (19.8) 81 (20.0) Discussion
Other vitamin A-rich fruits and vegetables
 Every day 23 (12.3) 36 (8.9) 0.322 The consumption of teff, in the form of TI, has been prac-
 Every other day 40 (21.4) 102 (25.2)
ticed for a long time in Ethiopia and parts of Eritrea. Teff is
 1–2 times a week 124 (66.3) 267 (65.9)
the primary staple grain, particularly in central and north-
Other fruits and vegetables
ern Ethiopia where it is widely harvested [10, 12]. Teff is
 Every day 31 (16.6) 65 (16.0) 0.338
also widely consumed in urban areas across all regions of
 Every other day 83 (44.4) 154 (38.0)
 1–2 times a week 53 (28.3) 125 (30.9)
Ethiopia. Provided teff has high iron content [10, 13, 24], it
 Once per 2 weeks 20 (10.7) 61 (15.1) could be presumed that teff consumption reduces the risk of
Eggs anemia. However, studies on the relation of frequencies of
 Every day 27 (14.4) 68 (16.8) 0.355 TI consumption to hemoglobin status are limited. The exist-
 Every other day 70 (37.4) 137 (33.8) ing evidence on the relation of teff consumption to anemia
 1–2 times a week 71 (38.0) 172 (42.5) in Ethiopia is largely based on nutrient content analysis and
 Once per 2 weeks 19 (10.2) 28 (6.9) biochemical studies. In this study, we examined the associa-
Dairy tion of different frequencies of TI consumption with anemia
 Every day 35 (18.7) 77 (19.0) 0.479 and found that TI consumption was independently associated
 Every other day 65 (34.8) 124 (30.6) with anemia; i.e., eating TI more frequently was associated
 1–2 times a week 79 (42.2) 193 (47.7) with a significant reduction in the odds of anemia.
 Once per 2 weeks 8 (4.3) 11 (2.7)
The finding of our study is in agreement with our hypoth-
Coffee/tea
esis that TI would be associated with reduced odds of ane-
 No 54 (28.9) 126 (31.1) 0.512
mia. A study in southern Ethiopia [17] also reported a simi-
 Yes 120 (64.2) 242 (59.8)
lar finding that pregnant women who ate teff-based food
 Not using 13 (7.0) 37 (9.1)
products more frequently had lower odds of anemia. Two
NA not applicable studies conducted in the UK also reported that teff-based
a
 Chi-square test was used to determine the differences between cases food products improved Hb and body iron levels [18, 19].
and controls Furthermore, the findings are in concordance with the nutri-
b
 Includes all cereal grains includes bread or injera made of wheat, tional composition of teff, which contains a higher amount
sorghum, corn, rice or mixture of grains

13

2016 European Journal of Nutrition (2019) 58:2011–2018

Table 3  Adjusted ORs of the Variables n Adjusted ­ORa, b (95% CI) P P-trend

association of the frequency
of teff injera consumption and Teff injera consumption
anemia among pregnant women,
 Every day 193 1 0.009
Addis Ababa, 2016
 Every other day 115 1.50 (0.71, 3.23) 0.285
 1–2 times a week 122 2.13 (1.03, 4.41) 0.04
 Once per 2 weeks 161 3.17 (1.62, 6.22) < 0.001
Meat consumption
 Every day 92 1 < 0.001
 Every other day 166 1.58 (0.71, 3.54) 0.263
 1–2 times a week 215 1.56 (0.73, 3.36) 0.251
 Once per 2 weeks 119 4.25 (2.08, 8.68) < 0.001
DGLV consumption
 Every day 72 1 < 0.001
 Every other day 145 0.86 (0.37, 2.00) 0.728
 1–2 times a week 240 0.83 (0.38, 1.82) 0.636
 Once per 2 weeks 135 1.97 (1.32, 2.94) < 0.001
Education (year) 592 0.87 (0.81, 0.94) < 0.001
Household income (­ Birrc) 592 0.98 (0.93, 1.03) 0.426
MUAC (cm) 592 0.83 (0.73, 0.95) 0.007
Pregnancy trimester
 First 173 1.95 (1.13, 3.37) 0.016 < 0.001
 Second 175 0.27 (0.13, 0.54) < 0.001
 Third 244 1
Teff injera*Meat (interaction)d 592 0.96 (0.77, 1.18) 0.669
Teff injera*DGLV (interaction)d 592 0.93 (0.75, 1.15) 0.508

OR odds ratio, CI confidence interval, DGLV dark green leafy vegetables, MUAC​ mid-upper arm circum-
ference
a
 Multiple logistic regression used to determine statistical significance
b
 Variables are adjusted for each other
c
 Ethiopian currency
d
 No evidence of interaction and removed from the multivariable model

of iron than any other cereal grain [10, 13]. Prone to vary
by growing conditions, the average iron contents of different
teff varieties (mg/100 gm) are white teff (9.5–37.7), red teff
(> 11.6), and mixed teff (> 11.5); these are greater amounts
compared to the iron content of 100 gm maize (3.6–4.8), sor-
ghum (3.5–4.1), wheat (3.7), and rice (1.5) [13]. Like other
cereal grains, teff also contains high phytate content [13,
24], which may impair iron absorption by forming insolu-
ble phytate–mineral or peptide–mineral–phytate complexes
in the gastrointestinal tract [24, 25]. However, previous
studies showed that the phytate in teff would be degraded
significantly during the preparation of TI, which involves a
rigorous fermentation process. Before making TI, the teff
dough is fermented for 2–3 days [12, 26]. The fermentation
of grains was shown to degrade phytate by activating endog-
enous phytases [26, 27].
It would be important to relate our result to the con-
text of anemia epidemiology in Ethiopia. Provided that
Ethiopia bears a high level of food insecurity and chronic
undernutrition [28, 29], one could presume a high burden of
anemia among pregnant women. However, over the past five
decades, anemia prevalence among Ethiopian women has
been consistently lower than the global and African averages
[1, 2, 30], even lower than the figures for most of the devel-
oped nations [1, 2]. Besides, recent studies have shown that
the burden of IDA is generally low in Ethiopia [16, 30, 31].
Based on the existing evidence, including the findings of our
study, it could be inferred that the low prevalence of anemia
among women in Ethiopia might be partly attributable to
the consumption of TI, solely of Ethiopians and Eritreans.
However, this proposition should be confirmed by further
studies as genetic factors unique to Ethiopians were also
linked to Hb status [32, 33].
Our finding bears some policy implications. Given
the importance of dietary diversity in combating anemia,
exploring teff as an alternative food for the prevention
of anemia would represent an important consideration.
Pregnant women may benefit more from eating teff, as

13
European Journal of Nutrition (2019) 58:2011–2018 2017
they are at higher risk of anemia and teff may avoid the
problem of intolerance associated with the intake of iron
tablets. Existing dietary and micronutrient assessment
tools, particularly those in use in Ethiopia, may need to
be revised or customized in line with the local dietary
practice. It may also be useful to consider incorporating
teff into the list of iron-rich food items. The commonly
used dietary assessment tools in Ethiopia [5, 21] incor-
porate teff into the group ‘cereals, roots, and tubers’. This
approach could be problematic because teff differs nutri-
tionally from other cereal grains, at least in terms of being
gluten-free and rich in iron.
The main strength of this study was that it was based
on human subjects, unlike many of the previous studies
which focused on assessing teff’s nutrient contents and
chemical properties. Our findings provide insight on the
extent of the association of frequencies of TI consump-
tion with anemia, which may serve as evidence for further
studies on the suitability of teff as an option for anemia
prevention. The comprehensive assessment of anemia
determinants enabled adjustment for various dietary and
non-dietary factors, thereby minimizing the possibility
that the observed association could be due to confound-
ing factors. The study also has many limitations worth
noting to the reader. First, we did not take into account
differences in teff varieties and growing conditions. The
amount of iron in TI varies by the type of the teff grain
used to make it [24]; thus, the extent of TI-anemia asso-
ciation may change according to the type of teff eaten.
Second, in this study, TI made of mainly teff flour (> 50%
teff) was assessed. However, other forms of injera with
smaller proportions of teff, such as ‘mixed injera’, need
to be investigated in future studies. Third, the use of hos-
pital-based cases and controls, and only pregnant women,
would limit the generalizability of the study. Fourth, the
study involved only public health facilities. Teff has
become an expensive commodity, and individuals going
to public health facilities tend to be of low socioeconomic
status. Thus, the level of consumption reported by our
study might be lower than the level in the general popu-
lation and resulted in biased estimates. Population-based
studies will have better generalizability. Fifth, due to the
limitations of the WDDS assessment tool, we did not
conduct dietary pattern and nutrient–nutrient interaction
analyses. Sixth, we did not measure biomarkers such as
hepcidin, ferritin, and transferrin saturation, which are
essential to differentiate types of anemia. Thus, the use
of only Hb precluded us from relating TI to a specific
form of anemia. However, the absence of significant dif-
ferences between cases and controls regarding competing
risks such as HIV and chronic illness might favor the
hypothesis that TI reduces the risk of IDA.
Conclusions
In conclusion, this study aimed to determine the associa-
tion of various levels of TI consumption with anemia in
pregnant Ethiopian women. We found that more frequent
consumption of TI was significantly associated with a lower
likelihood of anemia. Teff consumption may be a contribut-
ing factor to the low level of anemia in Ethiopian women.
To make a more definitive conclusion, further randomized
controlled trials or cohort studies, with better exposure and
outcome assessment, are warranted. The feasibility of TI as
an alternative food for the prevention of anemia needs to be
further explored.
Acknowledgements  SHM is a recipient of postgraduate scholarship,
Tehran University of Medical Sciences-International Campus. We are
also grateful for the study participants, data collectors, and hospital
staff involved in the study.
Author contributions  SHM, HT, and AE designed the study; HKT
coordinated the data collection; SHM and AE conducted the statis-
tical analysis and wrote the manuscript. SHM, HKT, and TAS took
part in the interpretation of findings; AE and BL critically reviewed
the manuscript. All authors read and approved the final version of the
manuscript.
Compliance with ethical standards 
Conflict of interest  The authors declared no conflict of interest.
Ethical standards  Ethical clearance was obtained from the institutional
review boards of the University of Gondar and the Addis Ababa City
Health Bureau (AAHB-217/15). Permission to conduct the study was
also obtained from the participating health facilities. Signed informed
consent was obtained from the study subjects before data collection.
We did not include personally identifiable information in the dataset.
Blood and stool samples were collected for the health facilities’ own
use; thus, we did not hold any biological specimen. Data collectors
advised anemic cases on dietary management of anemia and the impor-
tance of further medical care. Refer to this study’s registration status
through its ClinicalTrials.gov identifier (NCT number: NCT03251664)
available on https​://clini​caltr​ials.gov/ct2/show/NCT03​25166​4.
References
1. Stevens GA, Finucane MM, De-Regil LM, Paciorek CJ, Flax-
man SR, Branca F, Pena-Rosas JP, Bhutta ZA, Ezzati M, Nutri-
tion Impact Model Study G (2013) Global, regional, and national
trends in haemoglobin concentration and prevalence of total and
severe anaemia in children and pregnant and non-pregnant women
for 1995–2011: a systematic analysis of population-representative
data. Lancet Glob Health 1(1):e16–e25. https​://doi.org/10.1016/
S2214​-109X(13)70001​-9
2. The World Bank Group. Prevalence of anemia among pregnant
women (%). https​://data.World​bank.Org/indic​ator/sh.Prg.Anem.
Accessed 29 Nov 2017
3. Allard S (2015) Anaemia in pregnancy–a renewed focus needed.
BJOG 122(2):169. https​://doi.org/10.1111/1471-0528.13226​
13

2018
4. Wessells KR, Ouédraogo CT, Young RR, Faye MT, Brito A,
Hess SY (2017) Micronutrient status among pregnant women in
Zinder, Niger and risk factors associated with deficiency. Nutri-
ents 9(5):430
5. Central Statistical Agency Ethiopia (2016) Demographic and
health survey. https​://dhspr​ogram​.com/pubs/pdf/FR328​/FR328​
.pdf. Accessed 29 Nov 2017
6. Lopez A, Cacoub P, Macdougall IC, Peyrin-Biroulet L (2016)
Iron deficiency anaemia. Lancet 387(10021):907–916. https:​ //doi.
org/10.1016/S0140​-6736(15)60865​-0
7. Camaschella C (2015) Iron-Deficiency Anemia. N Engl J Med
373(5):485–486. https​://doi.org/10.1056/NEJMc​15071​04
8. Schlemmer U, Frolich W, Prieto RM, Grases F (2009) Phytate in
foods and significance for humans: food sources, intake, process-
ing, bioavailability, protective role and analysis. Mol Nutr Food
Res 53(Suppl 2):S330–S375. https​://doi.org/10.1002/mnfr.20090​
0099
9. Nair MK, Augustine LF, Konapur A (2015) Food-based inter-
ventions to modify diet quality and diversity to address multiple
micronutrient deficiency. Front Public Health 3:277. https​://doi.
org/10.3389/fpubh​.2015.00277​
10. Zhu F (2018) Chemical composition and food uses of teff (Era-
grostis tef). Food Chem 239:402–415. https​://doi.org/10.1016/j.
foodc​hem.2017.06.101
11. Cheng A, Mayes S, Dalle G, Demissew S, Massawe F (2017)
Diversifying crops for food and nutrition security—a case of
teff. Biol Rev Camb Philos Soc 92(1):188–198. https​://doi.
org/10.1111/brv.12225​
12. Gebremariam MM, Zarnkow M, Becker T (2014) Teff (Eragrostis
tef) as a raw material for malting, brewing and manufacturing of
gluten-free foods and beverages: a review. J Food Sci Technol
51(11):2881–2895. https​://doi.org/10.1007/s1319​7-012-0745-5
13. Phytate, zinc, iron and calcium content of selected raw and pre-
pared foods consumed in rural Sidama, Southern Ethiopia, and
implications for bioavailability
14. Saturni L, Ferretti G, Bacchetti T (2010) The gluten-free diet:
safety and nutritional quality. Nutrients 2(1):16–34. https​://doi.
org/10.3390/nu201​0016
15. Harika R, Faber M, Samuel F, Kimiywe J, Mulugeta A, Eilander
A (2017) Micronutrient Status and Dietary Intake of Iron, Vitamin
A, Iodine, Folate and Zinc in Women of Reproductive Age and
Pregnant Women in Ethiopia, Kenya, Nigeria and South Africa: A
Systematic Review of Data from 2005 to 2015. Nutrients 9 (10).
https​://doi.org/10.3390/nu910​1096
16. Gebreegziabher T, Stoecker BJ (2017) Iron deficiency was not
the major cause of anemia in rural women of reproductive age
in Sidama zone, southern Ethiopia: a cross-sectional study. PloS
One, 12, e0184742
17. Argaw B, Argaw-Denboba A, Taye B, Worku A (2015) Major
risk factors predicting anemia development during pregnancy:
unmatched-case control study. J Community Med Health Educ
5:2161–0711
18. Alaunyte I, Stojceska V, Plunkett A, Derbyshire E (2014) Dietary
iron intervention using a staple food product for improvement of
iron status in female runners. J Int Soc Sports Nutr 11(1):50. https​
://doi.org/10.1186/s1297​0-014-0050-y
19. Bokhari F, Derbyshire E, Li W, Brennan CS, Stojceska V (2012)
A study to establish whether food-based approaches can improve
13
European Journal of Nutrition (2019) 58:2011–2018
serum iron levels in child-bearing aged women. J Hum Nutr Diet
25(1):95–100. https:​ //doi.org/10.1111/j.1365-277X.2011.01185.​ x
20. World Health Organization. Iron deficiency anaemia: assessment,
prevention and control. http://www.who.int/nutrit​ ion/public​ ation​ s/
micro​nutri​ents/anaem​ia_iron_defic​iency​/WHO_NHD_01.3/en/.
Accessed 29 Nov 2017
21. Food and Agriculture Organization of the United Nations. Guide-
lines for measuring household and individual dietary diversity.
http://www.fao.org/docre​p/014/i1983​e/i1983​e00.htm. Accessed
29 Nov 2017
22. World Health Organization. The use and interpretation of anthro-
pometry. http://www.who.int/child​g rowt ​h/publi​catio​ns/physi​
cal_statu​s/en/. Accessed 29 Nov 2017
23. Arimond M, Wiesmann D, Becquey E, Carriquiry A, Daniels MC,
Deitchler M, Fanou-Fogny N, Joseph ML, Kennedy G, Martin-
Prevel Y, Torheim LE (2010) Simple food group diversity indica-
tors predict micronutrient adequacy of women’s diets in 5 diverse,
resource-poor settings. J Nutr 140 (11):2059S–2069S. https​://doi.
org/10.3945/jn.110.12341​4
24. Gibson RS, Bailey KB, Gibbs M, Ferguson EL (2010) A review
of phytate, iron, zinc, and calcium concentrations in plant-based
complementary foods used in low-income countries and implica-
tions for bioavailability. Food Nutr Bull 31(2 Suppl):S134–S146.
https​://doi.org/10.1177/15648​26510​0312S​206
25. Grases F, Simonet BM, Prieto RM, March JG (2001) Dietary
phytate and mineral bioavailability. J Trace Elem Med Biol
15(4):221–228. https​://doi.org/10.1016/S0946​-672X(01)80037​-7
26. Urga K, Narasimha HV (1997) Effect of natural fermentation on
the HC1-extractability of minerals from tef (Eragrostis tef). Bull
Chem Soc Ethiopia 11:3–10
27. Sanni AI (1993) The need for process optimization of African
fermented foods and beverages. Int J Food Microbiol 18(2):85–95
28. Haile M (2005) Weather patterns, food security and humanitarian
response in sub-Saharan Africa. Philos Trans R Soc Lond B Biol
Sci 360(1463):2169–2182. https:​ //doi.org/10.1098/rstb.2005.1746
29. Labadarios D, McHiza ZJ, Steyn NP, Gericke G, Maunder EM,
Davids YD, Parker WA (2011) Food security in South Africa: a
review of national surveys. Bull World Health Organ 89(12):891–
899. https​://doi.org/10.2471/BLT.11.08924​3
30. Gebre-Medhin M, Killander A, Bahlquist B, Wuhib E (1976)
Rarity of anaemia of pregnancy in Ethiopia. Scand J Haematol
16(3):168–175
31. Abebe Y, Bogale A, Hambidge KM, Stoecker BJ, Arbide I,
Teshome A, Krebs NF, Westcott JE, Bailey KB, Gibson RS (2008)
Inadequate intakes of dietary zinc among pregnant women from
subsistence households in Sidama, Southern Ethiopia. Public
Health Nutr 11(4):379–386. https​://doi.org/10.1017/S1368​98000​
70003​89
32. Alkorta-Aranburu G, Beall CM, Witonsky DB, Gebremedhin
A, Pritchard JK, Di Rienzo A (2012) The genetic architec-
ture of adaptations to high altitude in Ethiopia. PLoS Genet
8(12):e1003110. https​://doi.org/10.1371/journ​al.pgen.10031​10
33. Bigham AW, Lee FS (2014) Human high-altitude adaptation: for-
ward genetics meets the HIF pathway. Genes Dev 28(20):2189–
2204. https​://doi.org/10.1101/gad.25016​7.114