Food and Nutrition Sciences, 2013, 4, 1271-1280
Published Online December 2013 (http://www.scirp.org/journal/fns)
http://dx.doi.org/10.4236/fns.2013.412163
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron
and Vitamin A Status of Rural Primary School Children in
Uganda
Hedwig Acham, Gaston Ampek Tumuhimbise*, Joyce K. Kikafunda
Department of Food Technology and Nutrition, School of Food Technology, Nutrition and Bioengineering, College of Agricultural
and Environmental Sciences, Makerere University, Kampala, Uganda.
Email: *ampston@yahoo.com
Received September 3rd, 2013; revised October 3rd, 2013; accepted October 10th, 2013
Copyright © 2013 Hedwig Acham et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In accor-
dance of the Creative Commons Attribution License all Copyrights © 2013 are reserved for SCIRP and the owner of the intellectual
property Hedwig Acham et al. All Copyright © 2013 are guarded by law and by SCIRP as a guardian.
ABSTRACT
Children in resource poor settings are at a high risk of inadequate iron and vitamin A intake when diets lack diversity
and are dominated by staple foods. Yet comparative information on diet quality among school children is scarce. The
objective of the study was to assess the potential of simple food group diversity to serve as a proxy indicator of iron and
vitamin A status among rural school children in Uganda. A cross sectional correlation model of associations between
Food Group Diversity (FGD) and iron and vitamin A status was used. We analyzed 8 schools in Kumi District, Uganda,
randomly selected from the 34 schools that participated in the main part of the study. Our sample included primary
school children, aged between 9 - 15 years (n = 172). Food group diversity and food variety (FV) were calculated from
both a food frequency questionnaire (FFQ) and a 24-hour dietary recall. The FGD and FVS were tested against iron (as
serum ferritin) and vitamin A (as serum retinol) status. The FGD (based on FFQ data) was 9.6 (±1.9). There was a posi-
tive correlation between 24-hour recall and FFQ for consumption of cereals (Corr. Coef = 0.28; p < 0.05), which was
also the most highly consumed group (98.9% & 86.9% by FFQ and 24-hour recall; respectively). Consistent with other
studies, increase in the number of food groups significantly increased serum ferritin and serum retinol measures (p <
0.001). Presence of at least one food item in the “roots & tubers”; “cereals”; and “pulses/nuts”, significantly increased
serum ferritin and serum retinol concentrations (p < 0.01). We speculate that simple food group diversity may reflect
intake and serve as a simple indicator of iron and vitamin A status among school children. Strategies aimed at increas-
ing dietary diversity in the community may benefit the families of these children and improve their micronutrient status.
Keywords: Food Group; Diversity; Iron; Vitamin A; Proxy Indicator
1. Introduction
Monotony in diet has been described as the result of
poverty and poor nutrition [1], and indeed; typical child
diets in communities and households with high rates of
malnutrition are monotonous and bulky. Many infants
and young children in Sub-Saharan Africa subsist on
gruel and porridge prepared from staples such as cereals,
roots and tubers accompanied with vegetables and pulses
[2-4]. Plant sourced foods account for more than three-
quarters of energy intake and in some cases similar pro-
portions of protein and micronutrient intakes [5].
Studies have presented information implying ecologi-
cal associations between diversity and overall nutrition
[6], where it has been suggested that the nutritional suc-
cess of the participants is linked with the greater diversity
of their diet [7]. Among school children in South Africa
[8], the diet comprised a limited number of food items,
and the resultant vitamin A deficiency prevalence was a
public health problem. Other studies have similarly
found elevated levels of malnutrition and nutrient defi-
ciencies in populations whose diets are restricted in vari-
ety or lack specific components that characterize high
quality diets [9,10].
In Uganda, two measures of diet quality (lack of meat
*Corresponding author.
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda
1272
and cow’s milk; and low intake of energy from fat) were
shown to have positive associations with nutritional out-
comes (particularly underweight, marasmus and low
Mid-upper-arm circumference measures); whereas meas-
ures of quantity showed no significant relationships with
the same outcomes [11].
Dietary diversity studies in Uganda are limited, and
much less among school children. However, reports in-
dicate that the population diet is generally less diverse [12]
and predominantly vegetarian with only 11% - 13% of the
energy intake is being supplied by foods of animal origin.
All regions of the country have insufficient intake of
micronutrients provided preferentially by meat, fish,
poultry and eggs; including vitamin A, vitamin B-12, iron,
zinc, and calcium. Intra-household food allocation shows
that generally, the dietary intake of children in Uganda is
not any different from that of other household members
[13].
Anaemia in Uganda remains a wide spread problem
(49% and 23% for children and women 15 - 45 y respe-
cively); with the prevalence of iron deficiency being es-
timated at 2 to 2.5 times the prevalence of anaemia [13].
Nearly half the anaemia suffered is reportedly iron-defi-
ciency (IDA); the other half being anaemia is caused by
malaria, worm infestations and chronic disease including
HIV infection [14]. On the one hand, Uganda is among
the African countries reported to be at high risk, with
about 20 percent of children in Uganda estimated to be
vitamin A deficient [15]. VAD can impair immunity and
cause eye damage that can lead to blindness and even
death. The Government has attempted to address these
high levels of micronutrient malnutrition by conducting
iron and vitamin A supplementation for children of 6 - 59
months and for women of 15 - 49 years. Food-based ap-
proaches through diet diversification are also being pro-
moted through food fortification, nutrition education,
particularly at health facilities during antenatal and post-
natal visits. However, more initiatives are needed to
tackle micronutrient deficiency in a sustainable way.
These investments in children and women have led to
developmental successes in Uganda in the recent years,
notably in primary education. There is evidence that pri-
mary enrollment has increased, gender parity has been
reached while adult and youth literacy rates are above
regional averages [16]. However, more than half of the
enrolled school children do not complete primary school.
There is paucity of data on the health and nutrition of
school-age children in Uganda, and even where data is
available, it is not comprehensive. With limited data for
the age group, no interventions can be planned to target
their needs, and yet it is well known that school age and
adolescence represent an additional window during
which growth promoting interventions, possibly initiated
years before puberty, might yield substantial life cycle
and intergenerational effects [17]. This study explored the
potential for use of simple food group diversity to serve as
a proxy indicator of iron and vitamin A status among rural
school children in Uganda aged 9 - 15 years.
2. Materials and Methods
2.1. Study Design
We used a cross-sectional study design, involving 172
primary school children in Grade 4, aged between 9 - 15
years.
2.2. Study Area
The study was conducted in Kumi district, Eastern
Uganda in the period 2006-2007. Kumi district is ap-
proximately 2821 km2 in area. The climate is equatorial
with a bimodal type of rainfall received in the months of
April-May and July-August, and a main dry season run-
ning from December to February. The vegetation is pre-
dominantly savannah, although punctuated by thickets,
some forest plantations, and riparian vegetation [18]. The
district borders Lake Kyoga basin and the wetland sur-
rounding covers 35% of the total area, providing an im-
portant resource to the livelihood of the people in the
district [19]. The settlement patterns in the district reflect
the resources available, a phenomenon in most districts
of Uganda where agriculture is the major backbone of the
economy.
2.3. Sampling Procedure and Sample Size
Determination
This study was part of a larger study earlier reported [20].
Of the 34 schools selected for the main study (following
a modified cluster sampling procedure), a sample of 8
schools were randomly selected to participate in this par-
ticular study. The sample size was based on the estimate
of the initial sample size (n = 1020), using ±10%
precision, a confidence level of 95% and p-value of 0.5.
Considering an original sample size; the required sample
(n) for this particular study was 95. However, to take
care of drop-out, the number was doubled. The data
collected for food intake, iron and vitamin A studies (n =
172) are presented in this paper. All participating schools
were public (government-funded) mixed day schools.
2.4. Study Implementation
2.4.1. Blood Collection, Preparation and Analyses
A sample of 5 ml of blood was obtained from each child
by venipuncture [21]. Haemoglobin (Hb) was assessed in
the field using a portable HemoCue machine (model: 201,
Ängelholm, Sweden). Anaemia was defined as Hb < 12.0
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda 1273
mg/l following recommended classification [22].
A sample of 20 μL of whole blood was transferred di-
rectly onto blood collection cards (Whatman 903, Sch-
leicher & Schuell bioscience; Ref. 10 539 859; GmbH,
Germany), for preparation of dried blood spots (DBS).
The DBS cards were immediately kept in a specially de-
signed dark box and allowed to dry for 3 hours, and later
put in Ziploc bags with a desiccant (minipax sorbent;
1-800-445-9890, Multisorb technologies Inc), kept in the
dark box and transferred into an ice box, transported to
the laboratory at Makerere University where they were
kept at low temperatures (4˚C - 6˚C), until transportation
for retinol analysis. Because of the rural field conditions,
it was not possible to centrifuge the blood; instead blood
was left to stand in darkness in the tubes overnight,
which allowed for separation. After separation had been
achieved, 20 μL of serum was drawn using a capillary
tube, spotted onto filter paper cards and left to dry for 3
hours. The DSS cards were put in Ziploc bags with a
desiccant, kept at low temperatures (4˚C - 6˚C) until
transportation for serum ferritin analysis. Both assays of
Dried Serum Spots and Dried Blood Spots were per-
formed at Vitas AS Analytical Laboratory, Oslo, Nor-
way.
Serum ferritin was analyzed based on ELISA princi-
ples of assays as previously described [23], and modified
according to specifications on the kit. The assay was
done using AssayMax Ferritin ELISA kit (EF2003-1,
MO-USA). Control serum spots were prepared in the
laboratory, with some slight modifications made in vol-
umes of stock and buffer used in standard preparation.
Results were generated on a microplate reader (Mul-
tiskan Ascent, 15018820, version 1.3.2, Thermolab sys-
tems, Helsinki-Finland) at a wavelength of 450 nm. Ad-
justments were made for recovery and dilution by multi-
plying the serum concentrations obtained from the mul-
tiscan reader by the dilution factor of 10. Being an acute
protein whose concentration can rise during inflamma-
tion, the customary threshold (<12 - 15 μg/L) was not
used especially considering the fact that many of the
children had malaria infection. The cut-off for serum
ferritin was raised to 30 μg/L.
Retinol assay was done by isocratic modification of
Bieri’s High Performance Liquid Chromatography. The
modification was carried out using a combination of high
performance liquid chromatography and mass spectro-
metry (LC-MS assay principal). Whole blood retinol
values were validated using the HPLC procedure fol-
lowing Bieri’s high performance liquid chromatography
method, to adjust for recovery, storage effects and volume
of serum used in preparation of DBS. A median factor of
1.07 was obtained for 11 serum samples, which was mul-
tiplied to all the whole blood retinol values to convert
them to serum retinol equivalents. As recommended by a
previous study [24], no adjustments were made for in-
flammation. Therefore, serum retinol levels from labora-
tory assays were used after validating for recovery, stor-
age effects and volumes, without correcting for inflame-
mation. A serum retinol value <0.70 μmol/L, set by WHO
was used as cut off point to define Vitamin A deficiency
[25].
2.4.2. Dietary Assessment
A pre-tested Food Frequency Questionnaire (FFQ) was
used in the household interviews to obtain food con-
sumption data, which was done in home settings. The
questionnaire was designed to capture information of
dietary consumption over a period of 7 days. This was
done so as to have long enough time to offer representa-
tive picture of the diet, but also short enough that the
respondent can accurately recall their consumption [26].
For this study, the 13 food groups recommended for in-
dividual assessment [26], including meat/poultry, fish,
egg, Milk/milk products, Roots/tubers/plantain, cereal,
dark green leafy vegetables, pulses/nuts, vitamin A rich
vegetables, other vegetables, vitamin A rich fruits, other
fruits, fat/oil; were considered for dietary diversity ana-
lysis. The miscellaneous groups (sweets, coffee and tea)
were excluded taking into consideration the poor bio-
availability of micronutrients. Preliminary investigations
revealed 46 main foods (Table 1), which were recorded
into the 13 food groups, and used to determine the Food
Group Diversity (FGD) and food variety scores (FVS) as
shown in Table 3. A calculation of FGD was done by
summing the number of unique food groups reportedly
consumed in the period of the study.
For purposes of comparison, a single 24-h recall was
administered, to provide information of foods consumed
in the last 24 hours. A modified standard sample data
sheet [27] was used in the interviews. A pre-test of the
interview was carried out before the actual interviews
started, and to have the children freely express them-
selves, one teacher in each of the participating schools
was involved in the interview process, as previously
recommended [28]. The teachers underwent thorough
training on questioning techniques, details required, do’s
and don’ts in case of failure of recall. All 24-h recall in-
terviews were conducted at school in the native language
Ateso”, which was chosen to make it easier for the
school children to freely express themselves and to feel
comfortable. The school children were permitted to men-
tion foods in any order.
2.5. Data Analysis
Data was analyzed using the Statistical Package for
ocial Scientists (SPSS, version 15.0; 2006 SPSS Inc; S
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda
Open Access FNS
1274
Table 1. Food groups and food varieties used for construction of dietary diversity scores.
Food group Food varieties Food group Food varieties
1. Cereals/cereal products
Rice
Maize
Millet
Sorghum
Porridge
Bread
2. DGLVs DGLVs (all forms)
3. Roots/ tubers/plantains Cassava
White potato 4. Other vegetables Cabbage
Eggplant
5. Milk/milk products Milk
Milk products 6. Vitamin A rich vegetables Yellow potatoes
Tomatoes
7. Meat/poultry Meat (all forms)
Poultry (all forms) 8. Vitamin A rich fruits Mangoes
Paw paws
9. Fish Fish (all forms) 10. Other fruits
Oranges
Pineapple
Banana
Jack fruit
Guavas
Water melon
Fruit juice
11. Eggs Eggs (all forms) 12. Oil/ oil containing foods
Oil (used in cooking)
Ghee
Margarine
Chapatti
Popcorn
Doughnut
Pancakes
13. Pulses/nuts
Groundnuts
Sesame seeds
Beans
Peas
Soy/soy products
Chicago, USA). Descriptive statistics for frequencies of
food groups consumed and FGD; means and standard
deviations of iron and vitamin A concentrations; and
contribution of different food groups to the biochemical
variables were generated to characterize the intake of the
children. Correlation tests for strength of the relationship
between FGD and iron and vitamin A concentrations
were calculated using linear regression analysis. A
p-value of 0.05 was considered as the level at which tests
were significant.
2.6. Ethical Consideration
Clearance to conduct the study in schools was obtained
from the National Council for Science and Technology
after the proposal was vetted and approved by the Ethical
Committee based at the College of Health Sciences,
Makerere University. Permission was sought from the
Ministry of Education and Sports (MOES) and the Dis-
trict Education Office (DEO), who then gave instructions
to all head teachers in schools where the study was to be
conducted. Written informed consent was further ob-
tained from the parents and guardians and /or caretakers
of all the children in class 4 and assent also from the
children themselves.
3. Results
Over all, 172 children were examined for micronutrient
status, specifically iron and vitamin A as shown in Table
2. The mean concentrations of haemoglobin, serum fer-
ritin and serum retinol were 12.4 ± 1.1 g/dL; 24.1 ± 11.3
μg/L and 0.9 ± 0.3 μmol/L respectively. The means for
the different sexes were 12.5 ± 1.2 g/dL, 24.1 ± 11.6
μg/L, 0.9 ± 0.3 μmol/L; and 12.2 ± 0.9 g/dL, 24.1 ± 11.0
μg/L, 0.8 ± 0.4 μmol/L of haemoglobin, serum ferritin
and serum retinol for boys and girls respectively. Anae-
mia was moderate (28.5%), low serum ferritin and low
serum retinol concentrations were 75.6% and 36.6% re-
spectively, indicating a severe situation in the study
group (Table 2). In all cases, there was not a significant
difference between boys and girls (p < 0.05). However,
more than half of all the anaemia (63.3%) was associated
with iron deficiency (IDA).
Iron deficiency decreased with age among girls, while
he reverse was true for boys. Anaemia data were how- t
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda 1275
Table 2. Estimate of anaemia, iron deficiency, iron deficiency anaemia, vitamin A deficiency, and co-ocurrence of IDA and
VAD among the children (n = 172).
Sex and age category (%) Anaemia % Iron deficiency§ %Iron deficiency anaemia %Vitamin A deficiency % Vitamin A & iron
deficiency¶¶ %
Both sexes 28.5 75.6 18.0 36.6 34.6
Boys, n = 78
9 - 10 (19.5)
11 - 12 (45.5)
13 - 15 (35.1)
25.6
13.3
40.0
14.8
76.9
73.3
74.3
85.2
15.4
6.7
25.7
7.4
29.5
60.0
22.9
18.5
28.3
63.6
23.1
17.4
Girls, n = 94
9 - 10 (23.7)
11 - 12 (50.5)
13 - 15 (25.8)
30.9
31.8
27.7
33.3
74.5
81.8
76.6
62.5
20.2
18.2
17.0
25.0
42.6
54.5
44.7
29.2
40.0
55.6
38.9
26.7
Defined as hemoglobin concentration < 12.0 mg/dL. Defined as serum retinol concentration < 0.70 µmol/L. §Defined as serum ferritin concentration < 30 µg/L.
Defined as hemoglobin concentration < 12.0 mg/dL and serum ferritin concentration < 30 µg/L. ¶¶Defined as serum retinol concentration < 0.70 µmol/L and
serum ferritin concentration < 30 µg/L.
ever inconsistent for both sexes. Vitamin A deficiency
was higher among girls compared to the boys, with re-
duction in prevalence with age for both boys and girls.
Of those deficient in iron, 34.6% were also deficient in
vitamin A, showing co-ocurence of both micronutrient
deficiencies. Co-ocurrence was higher among girls (40.0%)
than boys (28.3%).
Table 3 shows the rates of consumption of different
food groups (percent) as measured by food frequency and
24-h recall. A total of seven food groups were consumed
on the day before the interview, compared to 13 that were
reportedly consumed seven days before the interview.
Food group consumption measured by FFQ indicated that
the majoriy of the respondents (>90%) had consumed
roots/tubers, cereals, dark green leafy vegetables (DGLVs),
pulses/nuts, vitamin A rich vegetables in the last seven
days. Although there was lack of agreement in consump-
tion between FFQ and 24-h recall for many food groups,
the correlation was significantly positive for cereal con-
sumption (p < 0.01). According to the 24-h recall inter-
views, the diet among the school children contained
mostly; cassava, millet, sorghum, maize and sweet/yellow
potato as shown in Tables 3 and 4.
Among the flesh foods, children reported beef and goat
meat as the commonly consumed types of meat, while
Nile perch and Tilapia were the two types of fish com-
monly reported. Other fish types reported were; silver fish,
Oreochromis spp, catfish and lungfish. The 24-h recall
data indicated that there was approximately 15% higher
intake of meat than fish, with less than 35% and 20%
reporting intake of meat and fish in the previous day;
respectively.
Table 4 indicates the food groups and food varieties
that were recorded in the study. Over all, there were 40
food varieties, with a mean of 9.7 food groups and a
mean variety of 16.6 foods for the number of children
assessed. The diet was predominantly plant based mostly
Table 3. Con su mp tio n o f F ood gro up s (pe rce nt) as me asu red
by FFQ and 24-h recall, and correlation between the two (n
= 172).
Food group 24-h FFQ r (*)
1. Meat/poulrty 32.1 52.6 0.062 (ns)
2. Fish 16.7 66.7 0.007 (ns)
3. Egg - 17.4 -
4. Milk/milk products - 62.4 -
5. Roots/tubers/plantain 90.5 95.7 0.063 (ns)
6. Cereal 86.9 98.9 0.283 (**)
7. Dark green leafy vegetables9.5 90.3 -
8. Pulses/nuts 65.5 99.5 -
9. Vitamin A rich vegetables - 92.3 -
10. Other vegetables 2.4 83.0 -
11. Vitamin A rich fruits - 65.4 -
12. Other fruits - 59.9 -
13. Fat/oil - 80.5 -
14. 4 or more FG consumed 2.3 95.1 0.035 (ns)
24-h. 24 hour recall; **Significant at p < 0.01; ns: not significant.
composed of cereals, roots & tubers, pulses and nuts, and
DGLVs. Three staples; cassava, sorghum and millet were
the most frequently reported, both in 24-h recall and
FFQ.
Among the cereals, sorghum (86.5%) and maize meal
(71.9%) used in bread making; were the most dorminant
foods; and in the roots/tubers, cassava (93.6%), also used
in bread making was the most highly consumed food.
Rice and Irish potato (28.7% and 7.6%; respectively)
were not as common as the foodstuffs reported above.
Dark green leafy vegetables were also highly consumed
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda
1276
Table 4. Varieties within food groups and consumption (%), reported by FFQ (n = 172).
Food group Food varieties Consumption (%)Food group Food varieties Consumption (%)
1. Cereals/cereal products
Rice
Maize
Millet
Sorghum
Porridge
Bread
28.7
71.9
59.1
86.5
59.6
39.8
2. Pulses/nuts
Groundnuts
Sesame seeds
Beans
Peas
Soy/soy products
93.0
1.8
82.5
52.0
9.9
3. Roots/tubers/plantains Cassava/Potato 93.6
7.6 4. DGLVs DGLVs (all forms) 93.6
5. Milk/milk products Milk
Milk products
56.1
28.7 6. Other vegetables Cabbage
eggplant
46.8
77.2
7. Meat/poultry Meat (all forms)
Poultry (all forms)
56.7
9.4 8. Vitamin A rich vegetablesYellow potatoes
tomatoes
76.6
71.3
9. Fish Fish (all forms) 68.4 10. Vitamin A rich fruits Mangoes
Paw paws
71.9
23.4
11. Eggs Eggs 15.8
12. Other fruits
Oranges
Pineapple
Banana
Jack fruit
Guavas
Water melon
Fruit juice
39.2
11.7
9.4
15.2
10.5
6.4
12.3
13. Oil/oil containing foods
Oil (used in cooking)
Ghee
Margarine
Chapatti
Popcorn
Doughnut
Pancakes
56.7
5.3
11.1
26.9
12.9
33.9
29.8
Total number of foods = 40
Mean number of food varieties consumed = 16.6
Mean number of food groups = 9.7
(93.6%), vitamin A rich vegetables; mainly yellow pota-
toes (76.6%), tomatoes (71.3%) and fruits (particularly
mangoes, 71.9%) were among the highly consumed
foods. Among the “other vegetables” commonly reported
was egg plant (77%) whose consumption was also quite
high.
Of the foods that were less consumed, poultry (9.4%),
eggs (15.8%) and milk products (28.7%), all of which are
of animal origin and known to be important sources of
iron, were the least consumed varieties of foods. “Other
fruits” and “oils/fats” were also less consumed, with a
consumption rate of less than 50%. Slightly more than
half of the respondents (68.4%) consumed fish, while
milk and meat were consumed by just about half. Both
the 24-h recall and FFQ data indicated that pulses and
nuts were a common part of the diet especially in the
absence of fish and meat, with a reported consumption of
65.5% and 99.5%, respectively as indicated in Table 3.
Although dark green leafy vegetables (DGLVs) were less
reported in the 24-h recall interviews (9.5%), the FFQ
data indicated that in the week prior to the interview,
about 94% of the respondents had consumed DGLVs.
The vegetables in this group included; Amaranthus (pur-
ple-Amaranthus cruentus; green-Amaranthus hybridus),
cowpea leaves (Vigna anguiculata), ecadoi (Cleome gy-
nandra), Alilot (Hibiscus esculentus) and emoros (Cy-
phostemm a a d e nocaule).
As a group, results indicated significantly increased
serum ferritin and serum retinol concentrations with in-
crease in number of food groups consumed (p < 0.001);
Table 5. The effect was however greater for serum fer-
ritin (beta coefficient = 2.0) compared to serum retinol
(beta coefficient = 0.14). Increase in number of foods
(food variety) over the given period (seven days) did not
significantly increase either of the two measures of nutri-
tional status. On the contrary, the relation was negative
(for serum retinol) and even where the relation was seen
to be positive (for serum ferritin), the effect was negligi-
ble. A look into the specific food groups revealed that
presence of atleast one item in the “pulse/nuts” “eggs”,
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda 1277
“cereals” and “vitamin A rich fruit” groups had a positive
effect on serum ferritin and serum retinol concentrations
and this relationship was significant for “pulses/nuts”
with serum retinol and “cereals” with serum ferritin con-
centrations as indicated in Table 5.
Interestingly, presence of tubers in the diet signifi-
cantly increased serum retinol concentration, but the ef-
fect was negligible for serum ferritin. Conversely, intake
of “fish”, “other vegetables” and “oil/oil foods” showed a
positive effect on serum ferritin concentration but not
serum retinol. The general finding for this result has been
that the number of food groups, presence of pulses/nuts
and cereals significantly indicate nutritional measures of
iron and vitamin A.
An analysis of children who were “normal” revealed
quite similar results, where the number of food groups
but not food varieities showed significantly positive as-
sociations with concentartions of both serum ferritin and
serum retinol as shown in Ta ble 6 . Although presence of
pulses/nuts positively increased concentrations of both
nutritional measures, their effect was only significant for
serum retinol (p = 0.02). Cereals also had a positive rela-
tion with both measures but the relation was significant
for only serum ferritin (p = 0.013).
Other food groups (including roots/tubers, vitamin A
rich fruits and oil/oil foods) showed positive effects on
both serum ferritin and serum retinol measures, with ef-
Table 5. Linear regression (by backward elimination) of
serum ferritin and vitamin A concentration with FGD, FV
and presence of different food groups based on FFQ data (n
= 172).
Item Unstandardized coefficient B (sig.)
Serum ferritin Serum retinol
Number of food groups (FGD) 1.996 (0.000) 0.135 (0.000)
Number of food varieties (FV) 0.236 (0.369) 0.029 (0.000)
Presence of
pulses/nuts (1 = yes) 1.745 (0.832) 0.482 (0.058)
Presence of fish (1 = yes) 0.45 (0.816) -
Presence of eggs (1 = yes) 1.02 (0.674) 0.058 (0.444)
Presence of cereals (1 = yes) 16.43 (0.047) 0.102 (0.686)
Presence of
roots/tubers/plantains (1 = yes) - 0.260 (0.023)
Presence of vitamin A
rich fruits (1 = yes) 2.269 (0.245) 0.021 (0.726)
Presence of other
vegetables (1 = yes) 0.635 (0.803) -
Presence of oil foods (1 = yes) 4.136 (0.086) -
Dependent variables are: Serum ferritin (µg/L), Serum retinol (µmol/L).
*Significant at p < 0.05. **Significant at p < 0.01. ***Significant at p < 0.001.
Table 6. Linear regression (by backward elimination) of
serum ferritin and vitamin A concentration with FGD, FV
and presence of different food groups based on FFQ data
for the “normal” children (Normal for SF = > 30.0; normal
for SR = > 0.70).
Item Unstandardized coefficient B (sig.)
Serum ferritin,
n = 42 (sig.)
Serum retinol,
n = 109 (sig.)
Number of food groups (FGD)4.227 (0.000) 0.151 (0.000)
Number of food varieties (FV)0.223 (0.637) 0.027 (0.010)
Presence of
pulses/nuts (1 = yes) 7.096 (0.487) 0.524 (0.020)
Presence of cereals (1 = yes) 28.608 (0.013) 0.214 (0.337)
Presence of
roots/ tubers/plantains (1 = yes)2.104 (0.693) 0.191 (0.236)
Presence of vitamin A
rich fruits (1 = yes) 0.288 (0.931) 0.016 (0.800)
Presence of oil foods (1 = yes)2.341 (0.658) 0.086 (0.180)
Dependent variables are; Serum ferritin (µg/L), Serum retinol (µmol/L).
*Significant at p < 0.05. **Significant at p < 0.01. ***Significant at p < 0.001.
fect seemingly greater for serum ferritin as indicated in
Table 6. From this result, we noted that; the number of
food groups, presence of pulses/nuts, cereals and roots/
tubers indicated the serum ferritin and serum retinol
status of the children, and the effects were much greater
for serum ferritin.
4. Discussion
Malnutrition remains a hidden problem among school-
age children in Uganda since the majority of those af-
fected are most often moderately or mildly malnourished
and yet they are not routinely assessed. In this study, we
found a high rate of iron deficiency, although anemia
was moderately a public health problem. Overlap be-
tween hemoglobin and serum ferritin indicated that more
than half the children with anemia had iron deficiency
anemia. One common practice, however, in assessing
whether or not anaemia is due to iron deficiency involves
monitoring the response in haemoglobin or haematocrit
levels after 1 or 2 months of oral supplementation with
iron [22]. An increase of 10 g/l in haemoglobin or 3% in
haematocrit is indicative of iron deficiency [29]. In this
study, it could not be confirmed whether the anaemia level
was due to iron deficiency since such tests of response
were not conducted. Due to its wide spread nature in
Uganda, region specific causes are varied, but the com-
mon factors that drive the high prevalence of anaemia are
the high disease burdens especially malaria and inade-
quate dietary sources and intake of iron among others
[13].
Similarly, we found severe vitamin A deficiency levels
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda
1278
(20%), based on WHO classification [22]. Anemia/iron
deficiency anemia and vitamin A deficiency are widely
reported as public health problems in Uganda [13] with
mostly children and women of child bearing age affected.
Many reasons have been given for these rates, among
which is the lack of a diverse diet [13]. Other studies
have reported a link between iron, vitamin A intake and
dietary diversity [30]. Studies of dietary intake of school
children have been rarely reported in Uganda. In this
current study, the staple foods reportedly consumed by
the participants, including cassava, sorghum, and millet;
which when ground into a meal and cooked is known as
atap”; were all bulky starchy foods. Sweet potatoes and
maize which are also bulky starchy foods, were often
substituted for “atap” especially among the households
that lived in the townships. This kind of diet is typical of
that consumed over large areas of Africa [4], with high
prevalence of undernutrition.
Meat and fish consumption were not as highly re-
ported as the staples. Given that animal protein foods are
generally expensive, there is limted consumption of these
foods, especially among the rural poor like the commu-
nity in this current study. Our 24-h recall data indicated
quite a low consumption of meat and fish. Findings of
the main study of the same population [20] indicated that,
those in the poor and most poor wealth categories con-
tributed to 70.4% of the entire anemia.
In the absence of meat and fish, groundnuts and beans/
pulses are commonly eaten as relishes on most days. Al-
though groundnuts initially replaced sweet potatoes and
green vegetables during the dry season when these adju-
vants were not available [31], to date it has become a
very popular relish in the diet of the Iteso. Both the FFQ
and 24-h recall data indicated that groundnuts and beans/
pulses were highly consumed (>90%), and were con-
sumed alone or accompanied with greens vegetables
(wild or cultivated), mainly; amaranthus (purple purple/
green). Although dark green leafy vegetables (DGLVs)
were less reported in the 24-h recall interviews (9.5%),
the FFQ data indicated that in the week prior to the inter-
view, about 94% of the respondents had reported con-
suming DGLVs.
Consuming one meal a day was not uncommon in this
study especially among those coming from typically rural
areas. On average, the 24-h recall interview among the
participants revealed that supper was the main meal of
the day, taken in the evenig after returning from school.
Based on their responses, breakfast consisted of mostly
plain tea, or tea with milk and groundnuts. Millet por-
ridge was also reported by some; while others reported
consuming leftover rice or maize porridge, and or wheat
buns next to the tea. A common lunch, for those who
were fortunate to receive lunch at home; was composed
of bread (made of cassava, sorghum/millet) or maize
meal (a stiff porridge made from maize flour). The com-
mon sauce was beans, groundnuts, cabbage or egg plant
with tomato. Supper; the main meal of the day, was often
made of the same foods that were prepared for lunch. At
the same time, there were no organized school feeding
programmes in the region to provide midday meals.
Although World Food Program [32] indicates that
Uganda does not lack food and categorizes 72.4% of
households as food secure, the typical Ugandan diet lacks
diversity and fails to provide sufficient micronutrients.
Rural dietary diversity remains low and tied to harvest
patterns and local availability, with a dietary emphasis on
starchy roots, and cereals. Micronutrient deficiency is
therefore a widespread problem in rural Uganda, affect-
ing school children. Based on the 2011 DHS data, only
34% of children aged 6 -35 months had consumed iron‐
rich foods in the 24 hours preceding the survey [13].
Our findings may have had some limitations, notably
the use of only one 24-h recall, and use of purely qualita-
tive food frequency questionnaire. However, since our
intention was to assess the usual intake and how it related
with the children’s nutritional status, we believe that the
findings of this study are valid and truly indicate the rela-
tionship between diet (diversity) and the children’s health
outcomes. However, we do recommend that for future
studies of this nature, either a repeated 24-h recall or
Semi-Qualitative Food Frequency Questionnaire must be
administered in order to get a better insight and measure
of intake in relation to micronutrient status.
5. Conclusion
From the findings of this study, we conclude that simple
food group diversity may reflect intake and serve as a
simple indicator of iron and vitamin A status among
schoolchildren in rural Uganda. Strategies aimed at in-
creasing dietary diversity in the rural communities may
benefit the families and improve their micronutrient in-
take and status.
6. Acknowledgements
The authors would like to thank the various participants
who took part in this study. Special thanks to technicians
who analysed laboratory samples and Carnegie Corpora-
tion of New York for the financial assistance that enabled
this study to take place.
REFERENCES
[1] M. H. N. Golden, “The Nature of Nutritional Deficiency
in Relation to Growth Failure and Poverty,” Acta Paedi-
atrica, Vol. 80, No. 374, 1991, pp. 95-110.
[2] G. L. Kennedy, M. R. Pedro, C. Seghieri, G. Nantel and I.
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda 1279
Brouwer, “Dietary Diversity Score Is a Useful Indicator
of Micronutrient Intake in Non-Breast-Feeding Filipino
Children,” Journal of Nutrition, Vol. 137, No. 2, 2007, pp.
472-477.
[3] A. Onyango, K. G Koski and G. Tucker, “Food Diversity
versus Breastfeeding Choice in Determining Anthropom-
etric Status in Rural Kenyan Toddlers,” International
Journal of Epidemiology, Vol. 27, No. 3, 1998, pp. 484-
448. http://dx.doi.org/10.1093/ije/27.3.484
[4] A. W. Onyango, “Dietary Diversity, Child Nutrition and
Health in Contemporary African Communities,” Com-
parative Biochemistry and Physiology Part A: Molecular
& Integrative Physiology, Vol. 136, No. 1, 2003, pp. 61-
69. http://dx.doi.org/10.1016/S1095-6433(03)00071-0
[5] A. Tarini, S. Bakari and H. Delisle, “The Overall Nutri-
tional Quality of the Diet Is Reflected in the Growth of
Nigerian Children,” Santé (Montrouge, France), Vol. 9,
No. 1, 1999, p. 23.
[6] A. Hailu and T. Tessema, “Anthropomentric Study of
Ethiopians Pre-School Children,” Ethiopian Medical Jour-
nal, Vol. 35, No. 4, 1997, pp. 235-244.
[7] M. Faber, V. B. Jogessar and A. J. S. Benadé, “Nutri-
tional Status and Dietary Intakes of Children Aged 2 - 5
Years and Their Caregivers in a Rural South African
Community,” International Journal of Food Sciences and
Nutrition, Vol. 52, No. 5, 2001, pp. 401-411.
http://dx.doi.org/10.1080/09637480120078285
[8] M. Faber, “Dietary Intake of Primary School Children in
Relation to Food Production in a Rural Area in KwaZulu-
Natal, South Africa,” International Journal of Food Sci-
ences and Nutrition, Vol. 50, No. 1, 1999, pp. 57-64.
http://dx.doi.org/10.1080/096374899101427
[9] L. E. Torheim, E. L. Ferguson, K. Penrose and M. Ario-
mond, “Women in Resource-Poor Settings Are at Risk of
Inadequate Intakes of Multiple Micronutrients,” Journal
of Nutrition, Vol. 140, No. 11, 2010, pp. 2051S-2058S.
http://dx.doi.org/10.3945/jn.110.123463
[10] J. E. Arsenault, E. A. Yakes, M. M. Islam, M. B. Hossain,
T. Ahmed, C. Hotz, B. Lewis, A. S. Rahman, K. M. Jamil
and K. H. Brown, “Very Low Adequacy of Micronutrient
Intakes by Young Children and Women in Rural Bangla-
desh Is Primarily Explained by Low Food Intake and
Limited Diversity,” The Journal of Nutrition, Vol. 143,
No. 2, 2013, pp. 197-203.
http://dx.doi.org/10.3945/jn.112.169524
[11] J. K. Kikafunda, A. F. Walker, D. Collett and J. K. Tum-
wine, “Risk Factors for Early Childhood Malnutrition in
Uganda,” Pediatrics, Vol. 102, No. 4, 1998, p. e45.
http://dx.doi.org/10.1542/peds.102.4.e45
[12] G. Kennedy, A. Berardo, C. Papavero, P. Horjus, T. Bal-
lard, M. C. Dop, J. Delbaere and I. D. Brouwer, “Proxy
Measures of Household Food Consumption for Food Se-
curity Assessment and Surveillance: Comparison of the
Household Dietary Diversity and Food Consumption
Scores,” Public Health Nutrition, Vol. 13 No. 12, 2010,
pp. 2010-2018.
http://dx.doi.org/10.1017/S136898001000145X
[13] UBOS, “Uganda Demographic and Health Survey
(UDHS) Report,” Uganda Bureau of Statistics and Macro
International, Inc., Calverton, 2011.
[14] F. M. Turyashemererwa, J. Kikafunda, R. Annan and G.
A. Tumuhimbise, “Dietary Patterns, Anthropometric Sta-
tus, Prevalence and Risk Factors for Anaemia among
School Children Aged 5 - 11 Years in Central Uganda,”
Journal of Human Nutrition and Dietetics, Vol. 26, No. 1,
2013, pp. 73-81. http://dx.doi.org/10.1111/jhn.12069
[15] V. M. Aguayo and S. K. Baker, “Vitamin A Deficiency
and Child Survival in Sub-Saharan Africa: A Reappraisal
of Challenges and Opportunities,” Food and Nutrition
Bulletin, Vol. 26, No. 4, 2005, pp. 348-355.
[16] M. Nishimura, T. Yamano and Y. Sasaoka, “Impacts of
the Universal Primary Education Policy on Educational
Attainment and Private Costs in Rural Uganda,” Interna-
tional Journal of Educational Development, Vol. 28, No.
2, 2008, pp. 161-175.
http://dx.doi.org/10.1016/j.ijedudev.2006.09.017
[17] A. M. Prentice, K. A Ward, G. R. Goldberg, L. M. Jarjou,
S. E. Moore, A. J. Fulford and A. Prentice, “Critical Win-
dows for Nutritional Interventions against Stunting,”
American Journal of Clinical Nutrition, Vol. 97, No. 5,
2013, pp. 911-918.
http://dx.doi.org/10.3945/ajcn.112.052332
[18] Kumi District State of the Environment Report, Kumi
District local Government, 2008.
http://nema-ug.org/district_reports/kumi_dsoer_2008.pdf
[19] LAKIMO, “Lake Kyoga Management Plan 2004-2007,”
Lake Kyoga Integrated Management Organisation, Ugan-
da, 2004.
[20] H. Acham, “Nutrition, Health and Academic Achieve-
ment of Primary School Children: A Case of Kumi Dis-
trict Eastern Uganda,” Makarere University, Kampala,
2010.
[21] H. Acham, J. K. Kikafunda, T. Tylleskar and M. K.
Malde, “Nutritional and Health Status of Primary School-
children in Rural Uganda,” African Journal of Agricul-
ture Nutrition and Development, Vol. 12, No. 2, 2012, pp.
5862-5880
[22] WHO/UNICEF/UNU, “Iron Deficiency Anaemia Assess-
ment, Prevention, and Control: A Guide for Programme
Managers,” Geneva, 2001.
[23] G. M. Addison, M. R. Beamish, C. N. Hales, M. Hodg-
kins, A. Jacobs and P. Lewellin, “An Immunoradiometric
Assay for Ferritin in the Serum of Normal Subjects and
Patients with Iron Deficiency and Iron Overload,” Jour-
nal of Clinical Patholology, Vol. 25, No. 4, 1972, pp.
326-329.
[24] J. G. Bieri, T. J. Tolliver and G. L. Catignani, “Simulta-
neous Determination of Alpha Tocopherol and Retinol in
Plasma or Red Cells by High Pressure Liquid Chroma-
tography,” American Journal of Clinical Nutrition, Vol.
32, No. 4, 1979, p. 2143.
[25] WHO, “Indicators of Vitamin A Deficiency and Their
Application in Monitoring and Evaluating Intervention
Programmes,” WHO/NUT/96.10, World Health Organi-
zation, Geneva, 1996.
Open Access FNS
Simple Food Group Diversity as a Proxy Indicator for Iron and Vitamin A
Status of Rural Primary School Children in Uganda
Open Access FNS
1280
[26] J. Coates, B. Colaiezzi, J. Fiedler, J. Wirth, K. Lividini
and B. Rogers, “Applying Dietary Assessment Methods
for Food Fortification and Other Nutrition Programs,”
Global Alliance for Improved Nutrition (GAIN), Geneva,
2012.
[27] R. Gibson, “Principles of Nutrition Assessment,” Oxford
University Press, New York, 1990.
[28] I. Tidemann-Andersen, H. Acham, A. Maage and M. K.
Malde, “Iron and Zinc Content of Selected Foods in the
Diet of Schoolchildren in Kumi District, East of Uganda:
A Cross-Sectional Study,” Nutrition Journal, Vol. 10, No.
81, 2011.
[29] WHO/CDC, “Assessing the Iron Status of Populations,”
Report of a Joint World Health Organization/Centres for
Disease Control and Prevention Technical Consultation
on the Assessment of Iron Status at the Population Level,
Geneva, 2005.
[30] A. W. Onyango, O. Receveur and S. A Esrey, “The Con-
tribution of Breast Milk to Toddler Diets in Western
Kenya,” Bulletin of the World Health Organization, Vol.
80, No. 4, 2002, pp. 292-299.
[31] M. G. Courcy, H. R. Hosking and L. J. A Loewenthal,
“An Investigation into Health and Agriculture in Teso,
Uganda,” Nutrition Report No. 1, Teso, Government
Printer, Entebbe, 1937.
[32] WFP, “Comprehensive Food Security & Vulnerability
Analysis (CFSVA),” World Food Programme, Uganda
Country Office, 2009.